AN EXP L ORATION OF MID - TO HIGH - VALENT TRANSITION ME TAL COMPLEXES FOR APPLICATION TO C ATALYSIS By Kelly E . Aldrich A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements f or the degree of Chemistry Doctor of Philosoph y 2019 ABSTRACT AN EXPLORATION OF MI D - TO HIGH - VALENT TRANSITION ME TAL COMPLEXES FOR APPLICATIONS TO CATALYSIS By Kelly E . Aldrich The valency or oxidation state of a transition metal in a complex plays a large role in determining the reactivity of the complex. With transition metal chemistry, historically accessible chemistry has often focused on metals in a low oxidation state. However, transformations involving transition metals in high o xidation states are of equal importance in providing complex products for use in consumer products. Expanding the applications and understanding of transition metal complexes in high oxidation states is the focus of the research presented in this dissertat ion. Fundamental studies of how ligands interact with high valent metals is presented in chapters 2 and 3, where a chromium(VI) model complex has been used to study bonding interactions between this d 0 transition metal and phosphine ligands. Practical appl ication of high valent titanium(IV) catalysts to C N bond forming reactions is presented in chapters 4 - 6. Finally, chapters 7 and 8 focus on the changes in the character of M N double bonds , with M = Fe and Ru, as the metal is forced to higher oxidation st ates. Collectively, these studies demonstrate different approaches to the same general problems and questions of how chemists can better understand and utilize high valent transitions metals to do catalytically - target desired transformations. iii Dedicated to my family and friends iv ACKNOWLEDGEMENTS I would like to begin by thanking my Ph. D. advisor, Aaron. Over the last 4 years, I have been able to explore my projects in your lab without limitation. You have also supported my manuscripts and participate start to finish in the process of peer - reviewed publication. I appreciate all of the opportunities you have given me. To my committee members, Prof. Mitch Smith, Prof. Ben Levine, and Prof. Tom Hamann, thank you for your guidan ce in navigating my Ph.D. Special thanks are necessary for Ben Levine. On several occasions he was willing to entertain my naïve questions about computational efforts in performing CASSCF calculations with the iron and ruthenium imido systems discussed in Chapter 7. I would like to thank Professor McCracken , both for assistance with EPR spectroscopy, as well as sharing his outlook on graduate school and career over the last several years. Along those same lines, I would also like to thank Dr. Dan Holmes for the user list in the s ubbasement. RIP Mutt. To Dr. Richard Staples, I am so glad that I have had the opportunity to work with you. I have gotten to learn as much crystallography as I can handle and then some from collecting data on some pretty atrocious crystals. While your fo along my path to graduation. v I also have pursued graduate school without the infl uence of instructors that I had early in my undergraduate career. When I started undergrad, I was very lost in terms of what I wanted to do with my life and career . I started attending Bay de Noc Community College out of high school , and eventually gravita ted toward science. The instructors that I encountered at Bay were phenomenal and encouraged me to pursue science further. The collective influence of Marc Labeau and Matt Krynicki is what first made me consider graduate school and a career in research. Al though not my instructor at Bay, I would also like to thank Amy Anderson, my instructor for EMT certification while I was at Bay. She also showed me endless support, and her belief in my abilities helped me figure out that I was ready to explore what I wan ted to study. I need to thank past labmates, including Dhwani Kansal, Corbin Livingston, Nick Boersma, Cody Pasko , and Dr. Amrendra Singh . I want to give special thanks to Dr. Tanner McDaniel (aka Tanner McDanner). Often w ithout realizing it, Tanner is an inspirational chemist. He has enthusiasm and curiosity for all things synthesis, and that enthusiasm is contagious. Talking chemistry with him often helped me find my motivation to tackle hard problems, and I will always be thankful for that. Special thank Lastly, and most imp ortantly, I w ant to thank my family . My sisters, Rachel and Emily, are cats, Balthazar and Mr. Pockets, are always there for me , too , and coming home to their fluffy faces always helped me relax after a long day in lab . vi Finally, I want to thank my parents. Nothing that I have achieved in the last 5 years would should accept are t hose related to my own goals and ambitions, and that hard work can take me really far. Because of the appreciation they instilled in me for education ., and their constant support my Ph. D., and now pursue research as a fulltime career for the rest of my life. Thank you. vii PREFACE Transition metal catalysis is a wide and varied field which evolves every day. The types of transformations, metals used, and supporting fragments for the active metal, span the periodic table. The sheer number of catalyzed reactions that we know can be pe rformed is astounding. Despite this diversity, there are still many important transformations that we 1) cannot do; 2) do not understand; or 3) can do, but very poorly in terms of atom economy, energy expended, and waste generated. Overcoming these limitat ions is one of the primary challenges facing the field today and is responsible for uniting the scientifically diverse community that deals with transition metal catalysis. Over the last 5 years, working in the Odom Group at Michigan State, I have become part of this community. While each of the projects that I have focused on take different tactics in probing the behavior and reactivity of transition metal complexes, this theme underlies all of my experiments (and has helped focus my efforts when I find myself in the glovebox going on a synthesis rampage). Moving forward in my career, catalysis may not ever be the main objective of my research. But, learning how to think about chemistry in the mindset of catalysis, with all its complexities and subtle bal ances, will continue to shape the way I approach any scientific problem. viii TABLE OF CONTENTS LIST OF TABLES ... xi LIST OF FIGURES .. .. . . x v LIST OF SCHEMES . ...... KEY TO SYMBOLS OR ABBREVIATIONS ... ... x l CHAPTER 1. INTRODUCTION ................................ ................................ ................................ ... 1 1.1 Traditional Approaches to Transition Metal Catalysis ................................ .................... 1 1.2 Targeted Design of Catalysts: Ligand Dono r Parameter ................................ ................. 4 1.3 Catalysts ................................ ................................ ................................ ................................ ...... 8 1.4 Electronic Exploration of Unique Transition Metal Complexes: Valency Effects on Metal - Imide Bond Character ................................ ................................ ................................ ..... 10 REFERENCES . 13 CHAPTER 2. PROBING THE IN SITU DYNAMICS OF THE LDP SYSTEM WITH NEUTRAL LIGANDS ................................ ................................ ................................ ................................ ..... 17 2.1 Introduction , ................................ ................................ ................................ ................... 17 2.2 Synthesis of [NCr(N i Pr 2 ) 2 (PR 3 )][X] Complexes, Solvent Choice, and Initial Assessments of Solvent Effects ................................ ................................ ................................ ...................... 23 2.3 Direct Approaches to Characterize Ion Pairing: Diffusion Ordered SpectroscopY (DOSY NMR) ... . 32 2.4 Other Solutio n State Investigations of Ion Proximity in Solution ................................ . 42 2.5 Computational Investigation of Ion Pairing with PF 6 ¯ ................................ ................... 44 2.6 Entropic Complications in Ionic LDP Systems ................................ .............................. 47 2.7 Conclusions ................................ ................................ ................................ .................... 52 2.8 Experimental ................................ ................................ ................................ .................. 54 REFERENCES 191 CHAPTER 3. ANALYSIS OF THE DONOR ABILITIES OF PHOSPHINES TO HIGH VALENT METALS ................................ ................................ ................................ .................... 197 3.1 Introduction , ................................ ................................ ................................ ................. 19 7 3.2 A Comparison of Traditional Phosphine Characteristics from Low - Valent Systems with LDP Results ................................ ................................ ................................ ............................. 201 3.3 Computational Analysis of the Electronic Structure of [NCr(N i Pr 2 ) 2 PR 3 ] + Using Natural Bonding Orbital and Natural Resonance Theory ................................ ................................ .... 208 3.4 Modeling Approximations to Examine Stereoelectronic Control on LDP Value ........ 217 3.5 Conclusions ................................ ................................ ................................ .................. 220 3.6 Experimental ................................ ................................ ................................ ................ 221 REFERENCES 265 ix CHAPTER 4. SILICA - GEL SUPPORTED TITANIUM CATALYSTS FOR C N BOND FORMING REACTIONS 1,2 ................................ ................................ ................................ ....... 269 4.1 Introduction ................................ ................................ ................................ .................. 269 4.2 Preparation of Silica - gels with Varied Surface Hydroxyl Group Density ................... 276 4.3 Performance of [Ti]200 and [Ti]700 as Intermolecular Hydroamination Catalysts .... 279 4.4 Application of [Ti]200 and [Ti]700 to Multicomponent Coupling Reactions ............. 283 4.5 Use of Ti700 to Pro duce Functionalized Heterocycles ................................ ................ 287 4.6 Exploration of Catalyst Reusability and Routes of Deactivation ................................ . 288 4.7 Accidental Discovery of [Ti]700 Activity for Catalytic Guanylation of Carbodiimide . ........................... .. 293 4.8 Conclusions ................................ ................................ ................................ .................. 296 4.9 Experimental ................................ ................................ ................................ ................ 298 REFERENCES CH APTER 5. REACTIVITY AND RATE LAW DETERMINATION OF A LIGAND - FUNCTIONALIZED SILICA - SUPPORTED TITANIUM CATALYST ................................ . 361 5.1 Introduction ................................ ................................ ................................ .................. 361 5.2 Activity of [Ti]700(X) for Three - Component Coupling Chemistry with a Variety of X¯ Ligands ................................ ................................ ................................ ................................ .... 365 5.3 One - pot - two - step Heterocycle Synthesis with [Ti]700(X) ................................ .......... 370 5.4 3CC Reaction Kinetics for [Ti]700(X) ................................ ................................ ......... 372 5.5 Catalyst Poisoning and Recyclability ................................ ................................ ........... 382 5.6 Poisoning Experiments and Controls for [Ti]700 ................................ ........................ 384 5.7 Catalyst Recycling from Ti(X) 3 /SiO 2 700 Precatalysts: Enhanced Recyclability through Poisoning ................................ ................................ ................................ ................................ . 389 5 .8 Conclusions ................................ ................................ ................................ .................. 392 5.9 Experimental ................................ ................................ ................................ ................ 393 REFERENCES 442 CHAPTER 6. HOMOGENEOUS TITANIUM CATALYZED IMINOAMINATION AND CATALYST DISPROPORTIONATION PROCESSES ................................ ............................ 446 6.1 Introduction , ................................ ................................ ................................ ................. 446 6.2 Kinetic Analysis of Iminoamination Catalyzed by Ti(dpm)(NMe 2 ) 2 .......................... 448 6.3 Investigations into Catalyst Deactivation During Ti - Catalyzed Iminoamination ........ 458 6.4 Modification of the Bis - Chelating Ancillary Ligand ................................ ................... 464 6.5 Predicting What Ligands Lead to Stable Catalysts Versus Disproportionation ........... 468 6.6 Conclusions ................................ ................................ ................................ .................. 478 6.7 Experimental ................................ ................................ ................................ ................ 479 REFERENCES . 550 CHAPTER 7. AN EXPLORATION OF THE SYNTHESIS AND ELECTRONIC PROPERTIES OF RUTHENIUM AND IRON IMIDO COMPLEXES ................................ ............................ 554 7.1 Introduction , ................................ ................................ ................................ ................. 554 7.2 Synthesis and Oxidation of Terminal Ru Imido Complexes ................................ ........ 560 7.3 Synthesis and Oxidation of Fe Imide Analogues ................................ ......................... 567 7.4 Exploration of a Trischelating Phosphine Ligand Platform for Fe and Ru Imides ...... 570 7.5 Characterization of Cationic Ru5*, Fe4*, and Fe6 by EPR Spectroscopy (7.5) .......... 573 x 7.6 Computational Analysis Comparing Fe and Ru Analogues ................................ ......... 583 7.7 Reactivity Studies with Fe5 ................................ ................................ .......................... 588 7.8 Conclusions ................................ ................................ ................................ .................. 592 7.9 Experimental ................................ ................................ ................................ ................ 593 REFERENCES .671 CHAPTER 8. PURSUIT OF RUTHENIUM BIS(IMIDO) COMPLEXES AND HIGHER OXIDATION STATES ................................ ................................ ................................ .............. 677 8.1 Introduction ................................ ................................ ................................ .................. 677 8.2 Synthesis of Ru and Os 2 - diphenylhydrazido Complexes from Azobenzene ............ 680 8.3 Thermal and Photochemical Reactivity of ( 2 - diphenylhydrazido)Ru(PMe 3 ) 4 ........... 682 8.4 Synthesis of Ru(II)(1,4 - diaryltetrazene)tris(trimethylphosphine) Complexes ............. 689 8.5 Reactivity of Ru(II)(1,4 - diaryltetrazene)tris(trimethylphosphine) Complexes ........... 692 8.6 Synthetic Alterations of the Ru Platform: A Larger Phosphine Ligand ...................... 695 8.7 A Basic Reactivity Study with Ru1 and Comparison to Known Os Analogues .......... 697 8.8 Conclusions ................................ ................................ ................................ .................. 701 8.9 Experimental ................................ ................................ ................................ ................ 702 REFERENCES .779 xi LIST OF TABLES Table 2.1 The wide variety of complexes accessible through the general synthesis route. a3j and 3o were prepared using an alternate procedure in which the MX salt was TlPF6, the solvent for the reaction was DCM, and the entire procedure was performed as a 1 - pot reaction to facilitate total miscibility of the phosphine with the solvent. bThe synthesis of 3p was conducted under the strict exclusion of MeCN, in DCM with TlBArF24 as a 1 - pot procedure. Upon exposure of 3p to MeCN, 3p converts to 2 and PPh3 ligand. ................................ ................................ .................... 23 Tab le 2.2 Cr1 - P1 bond distances and N2 - Cr1 - N3 bond angles. a For those complexes were multiple bond distances/angles are given, there are two unique molecules in the asymmetric unit cell for these complexes. ................................ ................................ ................................ ............... 26 Table 2.3 LDP values determined in two different NMR solvents with a variety of different PR 3 ligands. All values reported here employ SbF 6 - as the counter ion. ................................ ............. 30 Table 2.4 Results of the DOSY NMR experiments utilizing various anions and the chromium(VI) cations 3a and 3f . ................................ ................................ ................................ .. 39 Table 2.5 LDP values measured for 3a and 3f with a variety of anions in CDCl3 and CD3CN. This illustrates the ion effect on the LDP measurement. ................................ .............................. 41 Table 2.6 Molecular weight calibration results for 3f [PF 6 ] in several solvents and concentrations. ................................ ................................ ................................ ................................ ....................... 44 Table 2.7 Bond orders calculated for the PF 6 ¯ anion, showing H - bo nding interaction effects on P F bond orders. ................................ ................................ ................................ ............................ 46 Table 2.8 LDP values measured in CD 3 CN with SbF 6 ¯ /PF 6 ¯ counter anions. .............................. 47 Table 2.9 Complexes 3f , 3j , and 3m values for the ionic complexes in CD 3 CN with the SbF 6 ¯ counter anion. ................................ ..................... 49 for 3a - p BArF 24 ¯ salts in CDCl 3 . 51 Table 2.1 1 Details from LDP measurements with various X ¯ ligands in CDCl 3 . ........................ 71 Table 2.12 Details from LDP measurements with various X ¯ ligands in CD 3 CN. ....................... 71 Table 2.13 Experimentally determined rate of Cr N i Pr 2 , = - 9 e.u.), and temperature of measurement. ................................ ... 72 Table 2.14 Molecular volumes calculated for various cations and anions from crystal structures using Olex Software. ................................ ................................ ................................ ..................... 75 Table 2.15 Experimental data from DOSY molecular weight calibrations with 3f [PF 6 ]. ............ 76 xii Table 2.16 1 determined across several different temperatures in CD 3 CN. ................................ ................................ .................... 81 Table 2.17 values for 3f [SbF 6 ]. ................................ ................................ ................................ ....................... 83 Table 2.18 Experimentally determined rate of Cr N i Pr 2 bond rota tion and the calculated . ................................ ................................ ................................ ................................ ....... 84 Table 2.19 Experimentally determined rate of rotation (k obs ) for the Cr N i Pr 2 bond in 3a - p salts with BArF 24 ¯ anion in CDCl 3 . ................................ ................................ ................................ ........ 85 values for 3a - p series of compounds in CDCl 3 with BArF 24 ¯ counter anion and by the original method in CD 3 CN with SbF 6 ¯ . ................................ 199 values determined for 3f , 3j , and 3m in both solvent/ion regimes and with the sta values. ................................ ................................ . 201 d electronic effects are 3, 15, 23 204 Table 3.4 AN Values indicating the Lewis acidity of several compounds determined by t he Guttman - Beckett Method. 37 a Measured in DCE, referenced to External O=P(Et) 3 as 41.0 ppm. ................................ ................................ ................................ ................................ ..................... 208 Table 3.5 Natural Charges for NCr(NH 2 ) 2 PE 3 + complexes ................................ ........................ 224 Table 3.6 Summary of Second Order Perturbatio n Interactions ................................ ................. 245 Table 3.7 Total Charges on NCr(NH 2 ) 2 + fragment and PE 3 . ................................ ...................... 246 Table 3.8 Sample variety of literature parameters available for stereoelectronic description of phosphine series under study. 3 - 11 ................................ ................................ ................................ 252 Table 4.1 General conditions for SiO 2 700 preparation from commercially available fumed silica. ................................ ................................ ................................ ................................ ..................... 278 Table 4.2 Substrate scope examined for hydroamination reactivity with [Ti]200 and [Ti ]700. The general conditions outlined in the Scheme below apply to all reactions. a,b ................................ 282 Table 4.3 Iminoamination (3CC) substrate scope examined with [Ti]200 and [Ti] 700. The general conditions outlined in Fig. 7, C apply to all reactions. a,b ................................ ............... 287 Table 4.4 Hydroamination Results for the Coupling of 1 - Octyne and Aniline with Recycl ed [Ti]200. ................................ ................................ ................................ ................................ ....... 289 Table 5.1 Conditions screened for reaction order in each substrate and catalyst for heterogeneous catalyzed 3 - component coupling of aniline, 1 - octyne, and t BuNC. The initial catalyst and reagent amounts are listed by concentration (M). ................................ ................................ ................... 373 xiii Table 5.2 Estimated initial rates (M - 1 min - 1 ) from the various conditions in Table 5.1. These estimates may be artificially low, as the reactions progressed far beyond 10% conversion in ~20 min. ................................ ................................ ................................ ................................ ............. 374 Table 5.3 Ligands examined in Figure 5.8, showing LDP values of each phenoxide and the k obs for the iminoamination reaction catalyzed with the given phenoxide as ligand (X) with [Ti]700(X). ................................ ................................ ................................ ................................ .. 381 Table 5.4 ICP - OES analysis of various treatments for the [Ti]700 precatalyst material. .......... 383 Table 5.5 Yields obtained from the hydroamination of 1 - octyne and aniline using [Ti]700 precatalyst and 10 equiv of HX ligand added. ................................ ................................ ............ 386 Table 5.6 Recycling experiments with 10 equiv of a variety of HX ligands added to the hydroamination of 1 - octyne and aniline under the general conditions. Yields for the initial run with fresh catalyst (run 1), and a subsequent use (run 2) are shown. ................................ ......... 390 Table 5.7 Results of reusing the [Ti]700 catalyst with 10 equiv of 5 - fluoroindole (each trial) to perform the hydroamination of aniline and 1 - o ctyne. ................................ ................................ 391 Table 5.8 Yield and regioselectivity observed for a variety of ligands screened for the iminoamination of aniline, 1 - octyne, and CyNC. ................................ ................................ ....... 4 01 Table 5.9 Yield and regioselectivity observed for a variety of ligands screened for the iminoamination of aniline, 1 - phenylpropyne, and CyNC. ................................ .......................... 402 Table 5.10 Yield and regioselectivity observed for a variety of ligands screened for the iminoamination of aniline, phenylacetylene, and CyNC. ................................ ........................... 403 Table 5.11 Examples of Graphical Method Application to Entries in Table 2. .......................... 432 Table 5.12 Simulated rate constant and equilibrium constant values used to fit each set of experimental reaction traces. Good agreement is noted among the values, with small variances in some rate constants during fitting. ................................ ................................ .............................. 439 Table 5.13 Details of LDP value measurements ) for the 4 - R - phenoxide ligands examined for electronic effects on [Ti]700(2,4,6 - tri - tert - butylphenoxide) iminoamination catalysis. ....... 441 Table 6.1 Experimental conditions examined for kinetic analysis of Ti(dpm)(NMe 2 ) 2 catalyzed iminoamination. The general parameters from the reaction scheme below were applied. Amounts listed in the table are given as cocnetrations (molar). ................................ ................................ . 450 Table 6.2 Initial rates predicted for entries 1 - 3 with different orders in catalyst concentration. The values calculated for a fractional o rder appear to agree better with the experimentally observed initial rates shown in Fig. 6.4. ................................ ................................ ................................ ..... 455 Table 6.3 The ligand combinations listed in the table above, the hete roleptic Ti(X) 2 A 2 complexes are noted quantitatively. This necessitates that K eq is very large. ................................ ............... 469 xiv Table 6.4 Values used to model the relationship of sterics and electronics to the equilibrium constant for the interconversion of the homoleptic and heteroleptic. ................................ ......... 475 Table 6.5 Experimentally determined diffusion coefficients (D) and calculated MW for Ti species for [Ti(OArCH 2 - Ntolyl)] x ·1/2 NHMe 2 in C 6 D 6 . ................................ ................ 532 Table 6.6 Experimental determination of diffusion coefficients (D) and calculated MW for Ti - Ntolyl)] x in C 6 D 6 . ................................ ................................ .................. 533 Table 6.7 Experimentally determined diffusion coeffieicnts (D) and calculated MW for the Ti species for Ti(OArCH 2 ArO)(I) 2 in C 6 D 6 . ................................ ................................ ................... 534 Table 6.8 Experimentally determined diffusion coefficients (D) and calculated MW of the Ti species for Ti(OArCH 2 ArO)(O i Pr) 2 complex in C 6 D 6 . ................................ ............................... 535 Table 6.9 Combinations of parameters examined for fitting the dependence of K eq %V bur . ................................ ................................ ................................ ................................ .......... 548 Table 7.1 Table of rele vant bond lengths and angles for Ru1 . ................................ ................... 559 Table 7.2 Select bond lengths and angles from the single crystal X - ray structures for Ru3 and Ru2 , shown in Fig. 7.3. ................................ ................................ ................................ ............... 563 Table 7.3 Select bond distances and angles from the single crystal X - ray structures of Fe3* and Fe2* shown in Figure 7.7. ................................ ................................ ................................ .......... 570 Table 7.4 Select bond distances and ang les for the single crystal X - ray structures of Fe5 and Fe6 . Note that Fe6 has two unique molecules in the asymmetric unit; several of the bond lengths and angles show statistical differences, so both measurements are shown. Images of these structures are s hown in Fig. 7.8. ................................ ................................ ................................ .................. 571 Table 7.5 Data from the X - ray crystal structures of several Fe and Ru - imide compounds for comparison to the computed optimized structures of Fe7 and Ru6 . ................................ .......... 584 Table 7.6 N1 - M - P angles (°) as optimized at the MP2 level of theory. The range spanned by each set of angles is also given. ................................ ................................ ................................ ........... 585 Table 8.1 Selected bond distances and angles from the single crystal structure of Ru5 in Fig. 8.10. ................................ ................................ ................................ ................................ ............. 692 xv LIST OF FIGURES Figure 1.1 (top) A simple example of an olefin metathesis reaction, cross metathesis (CM) that can be catalyzed by Ru - carbene complexes. (bottom) Examples of several viable Ru catalysts competent for CM and other olefin metathesis reactions. ................................ .............................. 3 Figure 1.2 Illustration of the original Tolman system for phosphine parameterization. (left) The Ni(CO) 3 PR 3 complex was used to d etermine overall donor ability of a phosphine. (right) The system utilized to determine a steric profile for a given PR 3 ligand where the metal P bond ................................ ................................ ................................ .............................. 6 Figure 1.3 The NCr(N i Pr 2 ) 2 X scaffold used to measure the LDP value of a given X ligand. The illustration shows the rotation of one of the Cr N i Pr 2 bonds. The transition state (center) forces the amide lone pair into an antibonding orbital. The highlighted protons on the i Pr groups have unique chemical shifts. Their exchange, monitored by 1 H NMR SST, facilitates the measurement of the rotation rate. ................................ ................................ ................................ .......................... 7 Figure 2.1 ( top ) Cr(VI) molecule used in the LDP system. ( bottom ) A selection of monoanionic the X ligand, reflecting ................................ ................................ ....... 18 Figure 2.2 ( top ) The rotation monitored by 1 H spin saturation transfer experiments to d etermine the LDP value for a ligand, X. The purple and green hydrogens exchange positions through the N Cr bond rotation, allowing for saturation of one of the H signals to carry over to the other signal proportional to the rate of this rotation during the SST experiment ( bottom ). By experimentally determining this rotation rate, the enthalpic barrier to this rotation can be = LDP. ................................ ................................ ................................ ............. 19 Figure 2.3 Genera l synthesis of the 3a - 3p salts. ................................ ................................ ........... 24 Figure 2.4 Plot of Tolman Cone Angle and Cr1 - P1 bond distances in X - ray crystal structures. ................................ ................................ ................................ ................................ ....................... 26 Figure 2.5 Crystal structure of 2, showing the NCCH3 bound end - pair. Ellipsoids are shown at 50% probability. H atoms are omitted for clarity. (Cr1 - N1 1.533 Å; Cr1 - N2 1.817 Å; Cr1 - N3 1.816 Å; N2 - Cr1 - N3 121.85 °; Cr1 - N4 - 13C 176.40 Å; N4 - C13 - C14 178.96; Cr1 - N4 2.004 Å). ................................ ................................ ................................ ............. 28 Figure 2.6 ( top ) Room temperature 1 H NMR spectrum of NC r(N i Pr 2 ) 2 (OPh) showing broad resonances for the i Pr groups due to rapid Cr N i Pr 2 bond rotation. ( bottom ) Room temperature 1 H NMR spectrum of [NCr(N i Pr 2 ) 2 PPhMe 2 ][SbF 6 ] showing sharp well - resolved resonances for the i Pr groups. The high barrier to rotation i n this molecule prevents exchange of the i Pr Hydrogens leading to a static spectrum. ................................ ................................ ....................... 29 Figure 2.7 14 N NMR spectra of 3f in CDCl 3 with SbF 6 - (left) and PF 6 - (right) as counterions. (* = N 2 reference at 309.6 ppm; a = amide shift; b = nitride shift). ................................ ..................... 32 xvi Figure 2.8 A 1 H DOSY NMR spectrum in CD 3 CN that contains several mole cular species of different sizes. The diffusion coefficients are inversely proportional to the molecular sizes, with larger species diffusing more slowly than smaller ones. ................................ .............................. 34 Figure 2.9 system with [NCr(NiPr2)2PR3]+ cations. ................................ ................................ .................... 35 Figure 2.10 ROESY NMR spe ctra for 3a [NCr(N i Pr 2 ) 2 PMe 3 ][BArF 24 - ] in CDCl 3 ( top ) and CD 3 CN ( bottom ). Correlation between the cation and anion are noted in CDCl 3 , but these cross peaks are not observed in CD 3 CN. ................................ ................................ ............................... 38 Figure 2.11 Representation of the optimized 3a [PF 6 ] structure showing electrostatic interactions i Pr groups (pink). ................................ ................................ ................ 46 Figure 2.12 1 H ROESY NMR Spectrum of 3a [BPh 4 ] in CDCl 3 . ................................ ................. 78 Figure 2.13 1 H ROESY NMR Spectrum of 3a [BPh 4 ] in CD 3 CN. ................................ ................ 79 Figure 2.14 The Eyring plot for SST measureme nts of 1 in CD 3 CN. ................................ .......... 82 Figure 2.15 value of Cr N i Pr 2 bond rotation in 3f [SbF 6 ]. ................................ ................................ ................................ ................................ ........ 83 Figure 2.16 1 H NMR of 3b [SbF 6 ] in CDCl 3 . ................................ ................................ ................ 88 Figure 2.17 13 C NMR of 3b [SbF 6 ] in CDCl 3 . ................................ ................................ ............... 89 Figure 2.18 31 P NMR of 3b [SbF 6 ] in CDCl 3 . ................................ ................................ ............... 90 Figure 2.19 19 F NMR of 3b [SbF 6 ] in CDCl 3 . ................................ ................................ ............... 91 Figure 2.20 1 H NMR of 3c [SbF 6 ] in CDCl 3 . ................................ ................................ ................ 92 Figure 2.21 13 C NMR of 3c [SbF 6 ] in CDCl 3. ................................ ................................ ................ 93 Figure 2.22 31 P NMR of 3c [SbF 6 ] in CDCl 3 . ................................ ................................ ................ 94 Figure 2.23 19 F of 3c [SbF 6 ] in CDCl 3 . ................................ ................................ .......................... 95 Figure 2.24 1 H NMR of 3d [SbF 6 ] in CDCl 3 . ................................ ................................ ................ 96 Figure 2.25 13 C NMR of 3d [SbF 6 ] in CDCl 3 . ................................ ................................ ............... 97 Figure 2.26 31 P NMR of 3d [SbF 6 ] in CDCl 3 . ................................ ................................ ............... 98 Figure 2.27 19 F NMR of 3d [SbF 6 ] in CDCl 3 . ................................ ................................ ............... 99 Figure 2.28 1 H NMR of 3e [SbF 6 ] in CDCl 3 . ................................ ................................ .............. 100 xvii Figure 2.29 13 C NMR of 3e [SbF 6 ] in CDCl 3 . ................................ ................................ ............. 101 Figure 2.30 31 P NMR of 3e [SbF 6 ] in CDCl 3 . ................................ ................................ .............. 102 Figure 2.31 19 F NMR of 3e [SbF 6 ] in CDCl 3 . ................................ ................................ .............. 103 Figure 2.32 14 N NMR of 3e [SbF 6 ] in CDCl 3 . ................................ ................................ ............. 104 Figure 2.33 1 H NMR of 3f [SbF 6 ] (PPhMe 2 ) in CDCl 3 . ................................ .............................. 105 Figure 2.34 13 C NMR of 3f [SbF 6 ] (PPhMe 2 ) in CDCl 3 . ................................ .............................. 106 Figure 2.35 31 P NMR of 3f [SbF 6 ] (PPhMe 2 ) in CDCl 3 . ................................ ............................. 107 Figure 2.36 19 F NMR of 3f [SbF 6 ] (PPhMe 2 ) in CD 3 CN. ................................ ............................. 108 Figure 2.37 19 F NMR of 3f [SbF 6 ] (PPhMe 2 ) in CDCl 3 . ................................ ............................. 109 Figure 2.38 14 N NMR of 3f [SbF 6 ] in CDCl 3 . ................................ ................................ .............. 110 Figure 2.39 1 H NMR of 3g [SbF 6 ] in CDCl 3 . ................................ ................................ .............. 111 Figure 2.40 13 C NMR of 3g [SbF 6 ] in CDCl 3 . ................................ ................................ ............. 112 Figure 2.41 31 P NMR of 3g [SbF 6 ] in CDCl 3 . ................................ ................................ ............. 113 Figure 2.42 19 F NMR of 3g [SbF 6 ] in CDCl 3 . ................................ ................................ ............. 114 Figure 2.43 1 H NMR of 3h [SbF 6 ] in CDCl 3 . ................................ ................................ .............. 115 Figure 2.44 13 C NMR of 3h [SbF 6 ] in CDCl 3 . ................................ ................................ ............. 116 Figure 2.45 31 P NMR of 3h [SbF 6 ] in CDCl 3 . ................................ ................................ ............. 117 Figure 2.46 19 F NMR of 3h [SbF 6 ] in CDCl 3 . ................................ ................................ ............. 118 Figure 2.47 1 H NMR of 3i [Sb F 6 ] in CDCl 3 . ................................ ................................ ............... 119 Figure 2.48 13 C NMR of 3i [SbF 6 ] in CDCl 3 . ................................ ................................ .............. 120 Figure 2.49 31 P NMR of 3i [SbF 6 ] in CDCl 3 . ................................ ................................ .............. 121 Figure 2.50 19 F NMR of 3i [SbF 6 ] in CDCl 3 . ................................ ................................ ............... 122 Figure 2.51 1 H NMR of 3j [PF 6 ] in CDCl 3 . ................................ ................................ ................. 123 Figure 2.52 13 C NMR of 3j [PF 6 ] in CDCl 3 . ................................ ................................ ................ 124 Figure 2.53 31 P NMR of 3j [PF 6 ] in CDCl 3 . ................................ ................................ ................ 125 xviii F igure 2.54 19 F NMR of 3j [PF 6 ] in CDCl 3 . ................................ ................................ ................ 126 Figure 2.55 14 N NMR of 3j [PF 6 ] in CDCl 3 . ................................ ................................ ................ 127 Figure 2.56 1 H NMR of 3k [SbF 6 ] in CDCl 3 . ................................ ................................ .............. 128 Figure 2.57 13 C NMR of 3k [SbF 6 ] in CDCl 3 . ................................ ................................ ............. 129 Figure 2.58 31 P NMR of 3k [SbF 6 ] in CDCl 3 . ................................ ................................ ............. 130 Figure 2.59 19 F NMR of 3k [SbF 6 ] in CDCl 3 . ................................ ................................ ............. 131 Figure 2.60 1 H NMR of 3l [SbF 6 ] in CDCl 3 . ................................ ................................ ............... 132 Figure 2.61 13 C NMR of 3l [SbF 6 ] in CDCl 3 . ................................ ................................ .............. 133 Figure 2.62 31 P NMR of 3l [SbF 6 ] in CDCl 3 . ................................ ................................ .............. 134 Figure 2.63 19 F NMR of 3l [SbF 6 ] in CDCl 3 . ................................ ................................ .............. 135 Figure 2.64 14 N NMR of 3l [SbF 6 ] in CDCl 3 . ................................ ................................ .............. 136 Figure 2.65 1 H NMR of 3m [SbF 6 ] in CDCl 3 . ................................ ................................ ............. 137 Figure 2.66 13 C NMR of 3m [SbF 6 ] in CDCl 3 . ................................ ................................ ............ 138 Figure 2.67 31 P NMR of 3m [SbF 6 ] in CDCl 3 . ................................ ................................ ............ 139 Figure 2.68 19 F NMR of 3m [S bF 6 ] in CDCl 3 . ................................ ................................ ............ 140 Figure 2.69 1 H NMR of 3n [SbF 6 ] in CDCl 3 . ................................ ................................ .............. 141 Figure 2.70 13 C NMR of 3n[SbF 6 ] in CDCl 3 . ................................ ................................ ............. 142 Figure 2.71 31 P NMR of 3n [SbF 6 ] in CDCl 3 . ................................ ................................ ............. 143 Figure 2.72 19 F NMR of 3n [SbF 6 ] in CDCl 3 . ................................ ................................ ............. 14 4 Figure 2.73 1 H NMR of 3o [PF 6 ] in CDCl 3 . ................................ ................................ ................ 145 Figure 2.74 13 C NMR of 3o [PF 6 ] in CDCl 3 . ................................ ................................ ............... 146 Figure 2.75 31 P NMR of 3o [PF 6 ] in CDCl 3 . ................................ ................................ ............... 147 Figure 2.76 19 F NMR of 3o [PF 6 ] in CDCl 3 . ................................ ................................ ............... 148 Figure 2.77 14 N NMR of 3o [PF 6 ] in CDCl 3 . ................................ ................................ ............... 149 Figure 2.78 1 H NMR of 3p [BArF 24 ] in CDCl 3 . ................................ ................................ .......... 150 xix Figure 2.79 13 C NMR of 3p [BArF 24 ] in CDCl 3 . ................................ ................................ ......... 151 Figure 2.80 31 P NMR of 3p [BArF 24 ] in CDCl 3 . ................................ ................................ ......... 152 Figure 2.81 19 F NMR of 3p [BArF 24 ] in CDCl 3 . ................................ ................................ ......... 153 Figure 2.82 1 H NMR of 2 [SbF 6 ] in CDCl 3 . ................................ ................................ ................ 154 Figure 2.83 13 C NMR of 2 [SbF 6 ] in CDCl 3 . ................................ ................................ ............... 155 Figure 2.84 19 F NMR of 2 [SbF 6 ] in CDCl 3 . ................................ ................................ ............... 156 Figure 2.85 14 N NMR of 2 [SbF 6 ] in CDCl 3 . ................................ ................................ ............... 157 Figure 2.86 1 H NMR spectrum of 3f [PF 6 ] in CDCl 3 . ................................ ................................ . 158 Figure 2.87 13 C NMR spectrum of 3f [PF 6 ] in CDCl 3 . ................................ ................................ 159 Figure 2.88 19 F NMR spectrum of 3f [PF 6 ] in CDCl 3 ................................ ................................ .. 160 Figure 2.89 31 P NMR of 3f [PF 6 ] in CDCl 3 . ................................ ................................ ................ 161 Figure 2.90 14 N NMR spectrum of 3f [PF 6 ]. ................................ ................................ ................ 162 Figure 2.91 1 H NMR of 3f [BArF 24 ] in CDCl 3 . ................................ ................................ ........... 163 Figure 2.92 13 C NMR spectrum of 3f [BArF 24 ] in CDCl 3 . ................................ .......................... 164 Figure 2.93 31 P NMR spectrum of 3f [BArF 24 ] in CDCl 3 . ................................ .......................... 165 Figure 2.94 19 F NMR spectrum of 3f [BArF 24 ] in CDCl 3 . ................................ .......................... 166 Figure 2.95 14 N NMR spectrum of 3f [BArF 24 ] in CDCl 3 . ................................ .......................... 167 Figure 2.96 1 H NMR spectrum of 3f [BArF 20 ] in CDCl 3 . ................................ ........................... 168 Figure 2.97 13 C NMR spectrum of 3f [BArF 20 ] in CDCl 3 . ................................ .......................... 169 Figure 2.98 31 P NMR spectrum of 3f [BArF 20 ] in CDCl 3 . ................................ .......................... 170 Figure 2.99 19 F N MR spectrum of 3f [BArF 20 ] in CDCl 3 . ................................ .......................... 171 Figure 2.100 14 N NMR spectrum of 3f [BArF 20 ] in CDCl 3 . ................................ ........................ 172 Figure 2.101 1 H NMR spectrum of 3f [Al(OC(CF 3 ) 3 ) 4 ] in CDCl 3 . ................................ ............. 173 Figure 2.102 13 C NMR spectrum of 3f [Al( OC(CF 3 ) 3 ) 4 ] in CDCl 3 . ................................ ............ 174 Figure 2.103 31 P NMR spectrum of 3f [Al(OC(CF 3 ) 3 ) 4 ] in CDCl 3 . ................................ ............. 175 xx Figure 2.104 19 F NMR spectrum of 3f [Al(OC(CF 3 ) 3 ) 4 ] in CDCl 3 . ................................ ............. 176 Figure 2.105 14 N NMR spectrum of 3f [Al( OC(CF 3 ) 3 ) 4 ] in CDCl 3 . ................................ ............ 177 Figure 2.106 1 H NMR spectrum of 3f [BPh 4 ] in CDCl 3 . ................................ ............................. 178 Figure 2.107 13 C NMR spectrum of 3f [BPh 4 ] in CDCl 3 . ................................ ............................ 179 Figure 2.108 31 P NMR spectrum of 3f [BPh 4 ] in CDCl 3 . ................................ ............................ 180 Figure 2.109 14 N NMR spectrum of 3f [BPh 4 ] in CD Cl 3 . ................................ ........................... 181 Figure 2.110 1 H NMR spectrum of 3a [BArF 24 ] in CDCl 3 . ................................ ......................... 182 Figure 2.111 13 C NMR spectrum of 3a [BArF 24 ] in CDCl 3 . ................................ ....................... 183 Figure 2.112 31 P NMR spectrum of 3a [BArF 24 ] i n CDCl 3 . ................................ ........................ 184 Figure 2.113 19 F NMR spectrum of 3a [BArF 24 ] in CDCl 3 . ................................ ........................ 185 Figure 2.114 14 N NMR spectrum of 3a [BArF 24 ] in CDCl 3 . ................................ ....................... 186 Figure 2.115 1 H NMR spectrum of 3a [BPh 4 ] in C DCl 3 . ................................ ............................ 187 Figure 2.116 13 C NMR spectrum of 3a [BPh 4 ] in CDCl 3 . ................................ ........................... 188 Figure 2.117 31 P NMR spectrum of 3a [BPh 4 ] in CDCl 3 . ................................ ........................... 189 Figure 2.118 14 N NMR spectrum of 3a [BPh 4 ] in CDCl 3 . ................................ ........................... 190 Figure 3.1 ( left ) The Ni(CO) 3 PR 3 complex used in to determine the TEP. ( right ) The model used to measure the Tolman Cone Angle of a given phosphine. The spheres represent a variety of R groups, and the P center and block are 2.28 Å apart. ................................ ................................ .. 201 Figure 3.2 Resonance forms that contribute to the ground state electronic structure of traditional low - valent metal - phosphorous interactions. ................................ ................................ ............... 205 Figure 3.3 The synthetic scheme used to generate the O=P(Et) 3 adduct with [NCr(N i Pr 2 ) 2 ] + . A high yield of the desired complex was isolated after recr ystallization. ( right ) Preliminary crystal structure of [NCr(N i Pr 2 ) 2 (OP(Et) 3 )][BArF 24 ]; note, thermal ellipsoids are not shown due to the severe disorder in the structure. It provided only connectivity. ................................ .................. 208 Figure 3.4 Structures examined by NBO/NRT analysis. ................................ ............................ 209 Figure 3.5 Resonance forms inherent to the CrN 3 fragment from rearran gement of the lone pair electron density across the N ligands. These resonance forms and their electron rearrangements do not affect the nature of the Cr P bonds, so they are summed as *Cr to focus resonance form discussion on the Cr P interaction. ................................ ................................ ............................. 210 xxi Figure 3.6 NRT determined resonance forms accounting for 99% of the ground state of 3a* . . 211 Figure 3.7 Contribution of the two resonance forms that compose a dative interaction to the ground state electronic structure of 3m* and 3o* . ................................ ................................ ...... 212 3m* and 3o* . The example is shown here with 3m* . ................................ ................................ ................................ ............................ 213 Figure 3.9 Ar 3m* and 3o* . The example is shown where with 3o* . ................................ ................................ ................................ ........................... 214 Figure 3.10 Ground state electronic structures of 3m* and 3o* , incl hyperconjugation among the - OMe and - NMe 2 hyperconjugative resonance form involving the metal. ................................ .............................. 214 with 3m* and 3o* , electron density is pushed from one of the R groups into the Cr ................................ ................................ ................................ ................................ ..................... 216 values for d ................................ ................................ ... 219 Figure 3.13 All complexes 3a - p fitted with the model in Eq. 2. (Red = trialkyl, orange = monoaryl, green = diaryl, blue = heteroatom substituents). ................................ ....................... 220 Figure 3.14 Ground state resonance assignment to Cr fragment by NRT for [NCr(NH 2 ) 2 PMe 2 Ph] + . ................................ ................................ ................................ ................ 223 Figure 3.15 CHOOSE 1 geometry for NBO analysis. ................................ ................................ 224 Figure 3.16 CHOOSE 2 geometry for NBO analysis. ................................ ................................ 235 Figure 3.17 ( left ) CHOOSE1; ( right ) CHOOSE2 ................................ ................................ ....... 248 Figure 3.18 Correlation of %V bur vs. Tolman Cone Angle for all phosphines used in model building ( 3a - o ). ................................ ................................ ................................ ........................... 254 Figure 3.19 Steric profile from fit of trialkylphosphines ( 3a - e ) using 2 - parameter fit, Eq. 3.6. 255 Figure 3.20 Electronic profile from fit of trialkylphosphines ( 3a - e ) using 2 - parameter fit, Eq. 3.5. ................................ ................................ ................................ ................................ ..................... 256 Figure 3.21 Trialkylphosphine 2 - parameter fit applied to total series: Electronic profile. (Orange squares = PR 2 Ph; Green triangles = PPh 2 R, Red circles = PR 3 , Blue circles = P(OR) 3 /P(NR 2 ) 3 .) ................................ ................................ ................................ ................................ ..................... 257 Figure 3.22 Trialkylphosphine 2 - parameter fit applied to total series: Steric profile. (Orange squares = PR 2 Ph; Green triangles = PPh 2 R, Red circles = PR 3 , Blue circles = P(OR) 3 /P(NR 2 ) 3 .) ................................ ................................ ................................ ................................ ..................... 258 xxii Figure 3.23 1 H NMR of [NCr(N i Pr 2 ) 2 (OP(Et) 3 )][BArF 24 ] in CDCl 3 . ................................ ......... 260 Figure 3.24 31 P NMR of [NCr(N i Pr 2 ) 2 (OP(Et) 3 )][BArF 24 ] in CDCl 3 . ................................ ........ 261 Figure 3.25 19 F NMR of [NCr(N i Pr 2 ) 2 (OP(Et) 3 )][BArF 24 ] in CDCl 3 . ................................ ........ 262 Figure 3.26 13 C NMR of [NCr(N i Pr 2 ) 2 (OP(Et) 3 )][BArF 24 ] in CDCl 3 . ................................ ........ 263 Figure 4.1 Reaction schemes to yield highly substituted nitrogen - based heterocycles using Ti - catalyzed iminoamination followed by simple organic ring - closing reactions. These processes are cond ucted as one - pot - two - step reactions from simple starting materials. 6 ................................ . 270 Figure 4.2 The combination of the LDP system using high valent Cr(VI) and our Ti(IV) hydroami nation catalysts which has facilitated the development of a quantitative model describing the effects of ancillary ligands on catalytic rate. ................................ ....................... 272 Figure 4.3 The specie s that were targeted as potential precatalyst materials for hydroamination and iminoamination chemistry using a silica - supported, heterogeneous catalyst system. ......... 276 Figure 4.4 The binding mode of Ti on the surface of SiO 2 200 upon treatment with Ti(NMe 2 ) 4 . For comparison, the bulk properties of the material reported by the Scott group is given along with the properties of the material we synthesized following their SiO 2 200 preparation. ................... 277 Figure 4.5 (top) Quartz tube used for silica gel prep with tube furnace setup. (bottom) Schematic for the preparation of the precatalyst material [Ti]700. ................................ .............................. 278 Figure 4.6 Possible site variations in the coordination environment and binding modes for the two different catalysts, [Ti]200 and [Ti]700. ................................ ................................ .............. 279 Figure 4.7 Reaction conditions applied to the heterogeneous catalysts. The yields and regioselectivities listed here were observed with [Ti]200. The same trends were observed with [Ti]700, as well, with improvements in regioselectivities and yields at higher temperatures in dramatically shorter reaction times. Moving forward, the conditions for C were adopted as the general procedure. ................................ ................................ ................................ ....................... 281 Figure 4.8 General reaction scheme applied to the iminoamination reactions studied with [Ti]200 and [Ti]700. ................................ ................................ ................................ ................................ . 286 Figure 4.9 Proposed surface morphologies of used [Ti]200, after use in hydroamination reactions. Changes in the coordination environment could lead to changes in the regioselectivity ratio observed in the hydroamination products. ................................ ................................ .......... 291 Figure 4.10 Ti complexes from the Richeson Group used for catalytic guanylation of carbodiimides and imide metathesis reactions. 47,48 ................................ ................................ .... 293 Figure 4.11 1 H NMR of HA1 red in CDCl 3 . ................................ ................................ ................. 319 Figure 4.12 13 C NMR of HA1 red in CDCl 3 . ................................ ................................ ................ 320 xxiii Figure 4.13 1 H NMR of HA2 red in CDCl 3 . ................................ ................................ ................. 321 Figure 4.14 13 C NMR of HA2 red in CDCl 3 . ................................ ................................ ................ 322 Figure 4.15 1 H NMR of HA3 red ................................ ................................ ................................ .. 323 Figure 4.16 1 H NMR of HA4 in CDCl 3 . ................................ ................................ ..................... 324 Figure 4.17 13 C NMR of HA4 in CDCl 3 . ................................ ................................ .................... 325 Figure 4.18 1 H NMR of HA7 red in CDCl 3 . ................................ ................................ ................. 326 Figure 4.19 13 C NMR of HA7 red in CDCl 3 . ................................ ................................ ................ 327 Figure 4.20 1 H NMR of HA8 red in CDCl 3 . ................................ ................................ ................. 328 Figure 4.21 13 C NMR of HA8 red in CDCl 3 . ................................ ................................ ................ 329 Figure 4.22 1 H NMR of HA12 Red in CDCl 3 . ................................ ................................ ............... 330 Figure 4.23 13 C NMR of HA12 red in CDCl 3 . ................................ ................................ .............. 331 Figure 4.24 1 H NMR of HA13 red in C 6 D 6 . ................................ ................................ .................. 332 Figure 4.25 13 C NMR of HA13 red in C 6 D 6 . ................................ ................................ ................. 333 Figure 4.26 1 H NMR of HA14 red in C 6 D 6 . ................................ ................................ .................. 334 Figure 4.27 13 C NMR of HA14 red in C 6 D 6 . ................................ ................................ ................. 335 Figure 4.28 1 H NMR of 3CC1 in CDCl 3 . ................................ ................................ ................... 336 Figure 4.29 13 C NMR of 3CC1 in CDCl 3 . ................................ ................................ .................. 337 Figure 4.30 1 H NMR of 3CC5 in CDCl 3 . ................................ ................................ ................... 338 Figure 4.31 13 C NMR of 3CC5 in CDCl 3 . ................................ ................................ .................. 339 Figure 4.32 GC - MS of HA3 red . ................................ ................................ ................................ ... 340 Figure 4.33 GC - MS of crude HA11 ................................ ................................ ........................... 341 Figure 4.34 GC - MS of crude HA12. ................................ ................................ .......................... 342 Figure 4.35 Crude GC - MS trace of HA13. ................................ ................................ ................. 343 Figure 4.36 HA14 crude GC - MS trace. ................................ ................................ ...................... 344 Figure 4.37 EI - MS Fragmentation Pattern for HA14(A). ................................ .......................... 345 xxiv Figure 4.38 EI - MS Fragmentation Pattern for HA14(B). ................................ ........................... 346 Figure 4.39 GC - MS trace o f HA14 red . ................................ ................................ ........................ 347 Figure 4.40 EI - MS of HA14 red (A). ................................ ................................ ............................. 348 Figure 4.41 EI - MS of HA14 red (B). ................................ ................................ ............................. 349 Figure 4.42 GC - MS of 3CC1. ................................ ................................ ................................ ..... 350 Figure 4.43 GC - MS of 3CC5. ................................ ................................ ................................ ..... 351 Figure 4.44 Fragmentation patterns observed for 1 - phenyl - 3 - cyclohexylurea. .......................... 352 Figure 4.45 Fragmentation patterns observed in the HCl wash of used [Ti]700 catalyst after iminoamination which closely match those for 1 - phenyl - 3 - cyclohexylurea. ............................. 353 Figure 4.46 Fragmentation pattern observed in the HCl wash of used [Ti]700 catalyst after iminoamination, which closely matches 1,3 - dicyclohexylurea. ................................ ................. 354 Figure 5.1 Proposed mechanism and rate law determined for homogeneous Ti(IV) hydroamination catalysts of the (A) 2 Ti(X) 2 variety; The integrated rate law is given, as well as the dependence of k obs on both amine and Ti concentrations. 7 In the active species (A) stays - imide active species. 362 Figure 5.2 (top) General reaction sequence for the 3 - component coupling of an amine, alkyne, and isonitrile to yield 1,3 - diimine tautomers. (bottom) The two catalysts typicall y utilized for this transformation. ................................ ................................ ................................ ............................ 363 Figure 5.3 Effects of different HX ligands on the general 3CC reaction utilizing different alkynes: (top) The variety of HX lig ands screened in the 3CC reactions; (bottom) Plots of the 3CC reaction products showing the regioisomer ratio versus % yield for each HX ligand examined. In the plots, burgundy diamonds correspond to bidentate ligands, while blue circles correspond to mono dentate ligands. The best ligands for each alkyne are specified next to their respective points. ................................ ................................ ................................ ......................... 368 Figure 5.4 The synthetic scheme (left) and single crystal X - ray s tructure of a 2 - amino - 3 - cyanopyridine synthesized using [Ti[700(X). ................................ ................................ ............. 372 Figure 5.5 Plot of initial rates from the kinetic experiments shown in Table 5.1. The initial rates, determined from linear fits shown in the plot above, are listed in the table below. These rates are may be slower than the actual initial rates as they go far beyond 10% converions. ................... 374 Figure 5.6 Order of catalyst and reagent dependence for rate law of the iminoamination reaction catalyzed by [Ti]700(X). Entry 1 represents the light purple points, with the following concentrations: 0.25 M aniline, 0.25 M CyNC, 0.51 M 1 - octyne, and 0.012 M (5 mol%) Ti. It is plotted against the following entries, where the substrate alteration is indicated: 3 (green, 0.024 M (10 mol%) Ti), 5 (blue, 0.51 M CyNC), 4 (maroon, 0.51 M aniline), and 6 (aqua, 1.01 M 1 - xxv octyne) to demonstrate how different reagent concentrations affect the rate of the iminoamination reactions. For rough approximations of the initial rates, refer to Fig. 5.5. ................................ . 376 Figure 5.7 Proposed catalytic cycle for [Ti]700(X) based on experimental kinetic data and KINSIM modeling. Note, in the figure, the green letters, rate constants, and equilibrium constants are from the KINSIM model. The deactivation step from A to J is based on experimental observations of 3CC reactivity with homogeneous Ti analogues (see Chapter 6). ................................ ................................ ................................ ................................ ..................... 378 Figure 5.8 Correlation between reaction rate and donor ability of X - in heterogeneous [Ti]700(X) catalyzed 3CC reactions. More electron - donating X ¯ ligands seem to enhance the rate of the reaction by increasing the rate of formation of the active species from the resting state of the catalyst. (blue = t Bu, grey = H, o range = OMe, yellow = Br). ................................ ................... 381 Figure 5.9 Poisoning experiments with pyrrole and 2 - tert - butyl - 4 - methoxyphenol, showing very different catalyst activity with varying c oncentrations of the two different HX ligands. ........... 385 Figure 5.10 General hydroamination reaction and conditions applied to [Ti]700 catalyzed reactions with excess ligan d additive (HX). Ligands examined in Table 5.5. ............................ 386 Figure 5.11 A traditional homgogeneous mechanism (i.e. Bergman or Doye) of hydroamination shown with a Ti catalyst ( top ), and a modified version of the mechanism where HX may participate in the deprotonation of the aza - titanacyclobutene intermediate and impact the equilibrium formation of the active Ti - imido species ( bottom ). ................................ ................. 388 Figure 5.12 Crystal data and structure refinement for twin5. ................................ ..................... 396 Figure 5.13 Crystal data and structure refi nement for tri_early2_a. ................................ ........... 398 Figure 5.14 1 H NMR of 3CC 1 in CDCl 3 (isomeric mixture of A and B). ................................ .. 416 Figure 5.15 13 C NMR of 3CC 1 in CDCl 3 (isomeric mixture of A and B). ................................ . 417 Figure 5.16 gCOSY NMR of 3CC 1 in CDCl 3 (isomeric mixture of A and B). ......................... 418 Figure 5.17 GC trace of 3CC 1 (A and B) and MS fragmentation pattern for 3CC 1 (A). ........... 419 Figure 5.18 GC trace of 3CC 1 (A and B) and MS fragmentation pattern for 3CC 1 (B). ............ 420 Figure 5.19 13 C NMR of 3CC 2 in CDCl 3 . ................................ ................................ .................. 421 Figure 5.20 13 C NMR of 3CC 2 in CDCl 3 . ................................ ................................ .................. 422 Figure 5.21 gCOSY NMR of 3CC 2 in CDCl 3 . ................................ ................................ ........... 423 Figure 5.22 GC trace and MS fragmentation pattern for 3CC 2 . ................................ ................. 424 Figure 5.23 1 H NMR of 3CC 3 in CDCl 3 . ................................ ................................ ................... 425 xxvi Figure 5.24 13 C NMR of 3CC 3 in CDCl 3 . ................................ ................................ .................. 426 Figure 5.25 gCOSY NMR of 3CC3 in CDCl 3 . ................................ ................................ ........... 427 Figure 5.26 1 H NMR of [Cr](O - Ph - 4 - bromo) in CDCl 3 (room tem perature). ............................ 428 Figure 5.27 1 H NMR of [Cr](O - Ph - 4 - bromo) in CDCl 3 ( - 20 °C). ................................ ............. 429 Figure 5.28 13 C NMR of [Cr](O - Ph - 4 - bromo) in CDCl 3 ( - 20 °C). ................................ ............ 430 Figure 5.29 Results from simulated reactions versus the experimentally determined concentrations of 3CC product over time, Entry 1. ................................ ................................ .... 435 Figure 5.30 Results from simulated reactions versus the experimentally determined concentrations of 3CC product over time, Entry 3. ................................ ................................ .... 436 Figure 5.31 Results from simulated reactions versus the experimentally determined concentrations of 3CC product over time, Entry 4. ................................ ................................ .... 437 Figure 5.32 Results from si mulated reactions versus the experimentally determined concentrations of 3CC product over time, Entry 5. ................................ ................................ .... 438 Figure 5.33 Example simulation from KINSIM program. ................................ ......................... 440 Figure 6.1 Crude GCMS analysis of the iminoamination of 3,5 - dimethylaniline, 1 - phenylpropyne, and CyNC catalyzed by 10 mol% Ti(dpm)(NMe 2 ) 2 ( left ) and 5 mol% [Ti]700(2,6 - dimethylphenylamidate) ( right ). Note that the reaction catalyzed by Ti(dpm)(NMe 2 ) 2 has a substantial peak at 8 min, which is the 3,5 - dimethylaniline starting material, as well as a large peak at 18 min for hydroamination side product. The reac tion catalyzed by [Ti]700 shows no other compounds in the GCMS trace (small peaks on baseline are polysiloxane column material from GC column). ................................ ................................ ...... 447 Figure 6.2 The reac tion traces of two homogeneous Ti(dpm)(NMe 2 ) 2 catalyzed iminoamination reactions. Both reactions reach a maximum yield an after about 12 h. With the high catalyst loading, there is even what appears to be a decrease in concentration of the product from the maximum measured concentration. ................................ ................................ ............................ 451 Figure 6.3 Graphical analysis of Entries 1 - 3 in Table 6.1. The plots are fitted with a fractional order ( top ) and first order ( bo ttom ) dependence. Similar fits result from both analyses. ........... 454 Figure 6.4 Examination of the initial rates for Entries 1 - 3 from Table 6.1. The results suggest that the catalyst concentration may not be simple first - order in these concentration ranges. ........... 455 Figure 6.5 Potential forms of titanium catalyst likely present in the catalytic iminoamination reaction mixture. ................................ ................................ ................................ ......................... 457 xxvii - Ntolyl)] 2 complex. The single crystal X - ray structure is shown with ellipsoids at the 50% probability and H atoms omitted for clarity. ................................ ................................ ................................ ........ 460 Figure 6.7 Various titanium species f ound to (or suspected of) react directly with the iminoamination product. This reactivity provides plausible means of titanium complex deactivation throughout the iminoamination reaction. ................................ ............................... 464 Figure 6.8 ( left ) 1 H NMR spectrum and ( right ) single crystal X - ray structure 4 of Ti(NMe 2 ) 2 (OArCH 2 ArO). The two protons on the methylene linker have unique positions due to the conformation of the 8 - membered ring formed by the ligan d with Ti, which is readily observed by the distinct doublets in the 1 H NMR spectrum. This trait applies to all of the complexes of the general form Ti(X) 2 (OArCH 2 ArO) and makes them easy to observe and distinguish between by 1 H NMR. ................................ ................................ ............................... 466 Figure 6.9 The comproportionation reaction monitored by determination of the equilibrium constant K eq . The 3 possible Ti species in equilibrium in these solutions are shown, including b oth starting materials and the only possible product. ................................ ................................ 473 Figure 6.10 Plot of K eq vs. LDP of the X ¯ ligands in the Ti(OArCH 2 ArO)(X) 2 . ........................ 474 Figure 6.11 A plot show ing model - predicted versus experimental data relating the donor ability of a given X¯ ligand to the equilibrium constant observed for formation of the heteroleptic species, Ti(X) 2 (OArCH 2 ArO) (Fig. 6.15). ................................ ................................ .................. 476 Figure 6.12 Potential ligands of interest that could avoid deactivation via ligand disproportionation as their donor abilities (LDP values) are predicted to be more and less donating than a chelated iminoamination p roduct, potentially disfavoring ligand exchange reactions. ................................ ................................ ................................ ................................ ..... 478 Figure 6.13 1 H NMR of Ti(dpm)(NMe 2 ) 2 and 3CC heated at 80 °C, 40 h in C 6 D 6 . .................. 489 Figure 6.14 1 H NMR of [Ti(dpm)(Ntolyl)] 2 , t BuNC (xs), and 3CC in C 6 D 6 , 80 °C at 6 h. ....... 490 Figure 6.15 1 - Ntolyl)(dpm)] 2 with t BuNC in situ in C 6 D 6 . ........................ 491 Figure 6.16 1 H NMR of [Ti(OArCH 2 A rO)(µ - Ntolyl)] 2 ·NHMe 2 after heation 16 h at 80 ° C. .... 493 Figure 6.17 Proposed decomposition pathway and final products observed (top) and proposed (bottom) for the Ti - imide species upon heating with the iminoamination product in C 6 D 6 . ...... 495 Figure 6.18 1 H NMR of 3CC(A/B) + [Ti(OArCH 2 ArO)(Ntolyl] 2 ·HNMe 2 heated for 0 h at 85 °C showing no Ti(OArCH 2 ArO) 2 . ................................ ................................ ........................... 496 Figure 6.19 1 H NMR of 3CC(A/B) + [Ti(OArCH 2 ArO)(Ntolyl] 2 ·HNMe 2 heated for 3 h at 85 °C showing a small amount of Ti(OArCH 2 ArO) 2 . ................................ ................................ .. 497 xxviii Figure 6.20 1 H NMR of 3CC(A/B) + [Ti(OArCH 2 ArO)(Ntolyl] 2 ·HNMe 2 heated for 48 h at 85 °C showing only Ti(OArCH 2 ArO) 2 as identifiable Ti species. ................................ ................ 4 98 Figure 6.21 1 H NMR of the iminoamination reaction catalyzed by 20 mol% Ti(OArCH 2 ArO)(NMe 2 ) 2 in tol - d 8 . Peaks at 11.2 and 10.4 ppm are for the two different regioisomers of the 3CC product. The large singlet at 3 .97 ppm is Fc as internal standard. The peak at 3.35 ppm belongs to the Ti(OArCH 2 ArO) 2 disproportionation species. ........................ 500 Figure 6.22 1 H NMR of an isomeric mixture of 3C C(A) and (B) in CDCl 3 . ............................. 501 Figure 6.23 13 C NMR of an isomeric mixture of 3CC k (A) and (B) in CDCl 3 . ........................... 502 Figure 6.24 GCMS of 3CC isomers A and B; fragmentation pattern for A isomer. .................. 503 Figure 6.25 GCMS of 3CC isomers A and B; fragmentation pattern for B isomer. .................. 504 Figure 6.26 HRMS for isomeric mixture of 3CC. ................................ ................................ ...... 505 Figure 6.27 1 H NMR of the 4CC product in CDCl 3 . ................................ ................................ .. 506 Figure 6.28 13 C NMR of the 4CC product in CDCl 3 . ................................ ................................ . 507 Figure 6.29 GCMS of the 4CC product and MS fragmentation pattern. ................................ .... 508 Fig ure 6.30 HRMS of the 4CC product. ................................ ................................ ..................... 509 Figure 6.31 gCOSY NMR of the 4CC product in CDCl 3 . ................................ .......................... 510 Figure 6.32 HMBC NMR of 4CC in CDCl 3 . ................................ ................................ .............. 511 Figure 6.33 1 - Ntolyl)(dpm)] 2 in tol - d 8 .(room temperature, high vac grease and hexane impurities) ................................ ................................ ................................ ....................... 512 Figure 6.34 1 - Ntolyl)(dpm)] 2 in tol - d 8 .( - 75 °C, high vac grease and hexane impurities) ................................ ................................ ................................ ................................ ... 513 Figure 6.35 13 - Ntolyl)(dpm)] 2 in tol - d 8 .(room temperature, high vac grease and hexane impurities) ................................ ................................ ................................ ....................... 514 Figure 6.36 1 H NMR of [Ti(OArCH 2 - Ntolyl)]·HNMe 2 in C 6 D 6 . ................................ ... 515 Figure 6.37 13 C NMR of [Ti(OArCH 2 - Ntolyl)]·HNMe 2 in C 6 D 6 . ................................ .. 516 Figure 6.38 1 H NMR of Ti(2,4 - di - tert - butyl - phenoxide) 4 in C 6 D 6 . ................................ ............ 517 Figur e 6.39 13 C NMR of Ti(2,4 - di - tert - butyl - phenoxide) 4 in C 6 D 6 . ................................ ........... 518 Figure 6.40 Crystal data and structure refinement for earlyy2. ................................ .................. 519 xxix Figure 6.41 Crystal data and structure refinement for c2c_early_a. ................................ ........... 521 Fig ure 6.42 Crystal data and structure refinement for p21_c_a. ................................ ................. 523 Figure 6.43 Crystal data and structure refinement for rjs. ................................ .......................... 525 Figure 6.44 Crystal data and structure refinement for uc_a. ................................ ....................... 527 Figure 6.45 Crystal data and structure refinement for early_a. ................................ .................. 529 Figure 6.46 DOSY MW determination for [Ti(OArCH 2 - Ntolyl)] x ·1/2 NHMe 2 in C 6 D 6 . ................................ ................................ ................................ ................................ ..................... 532 - Ntolyl)] x in C 6 D 6 . ................................ 533 Figure 6.48 DOSY MW determinati on of Ti(OArCH 2 ArO)(I) 2 in C 6 D 6 . ................................ .. 534 Figure 6.49 DOSY MW determination of Ti(OArCH 2 ArO)(O i Pr) 2 complex in C 6 D 6 . .............. 535 Figure 6.50 The graphical determination of reaction rate dependence on alkyne concentration. Purple spheres = 0.2 M (Entry 3); Red spheres = 1.0 M (Entry 4); Grey spheres = 0.4 M (Entry 5). ................................ ................................ ................................ ................................ ................ 538 Figure 6.51 The graphical determination of reaction rate dependence on isonitrile concentration. Purple spheres = 0.2 M (Entry 3); Orange spheres = 0.4 M (Entry 6). ................................ ...... 539 Figure 6.52 The graphical determination of reaction rate dependence on amine concentration. Purple spheres = 0.2 M (Entry 3); light blue spheres = 0.4 M (Entry 7); Green spheres = 1.0 M (Entry 8). ................................ ................................ ................................ ................................ ..... 540 Figure 6.53 Reaction progress of two identical kinetics trials run with Ti(dpm)(NMe 2 ) 2 and Ti(OArCH 2 ArO)(NMe 2 ) 2 . Simi lar results were obtained with both catalysts under the conditions used for kinetics despite better performance of the Ti(OArCH 2 ArO)(NMe 2 ) 2 under normal conditions applied to a typical iminoamination reaction. ................................ ........................... 541 Figure 6.54 Equilibrium ligand exchange reaction used to determine K eq experimentally. For reactions where K eq is small, starting materials 1 and 2 were used. For reactions where K eq was large, 3 could be prep ared and isolated cleanly and was utilized in these experiments. ............ 543 Figure 6.55 1 H NMR of Ti(OArCH 2 ArO) 2 in C 6 D 6 . ................................ ................................ .. 544 Figure 6.56 13 C NMR of Ti(OArCH 2 ArO) 2 in C 6 D 6 . ................................ ................................ . 545 Figure 6.57 1 H NMR of the equilibrium mixture of Ti(NMe 2 ) 4 , Ti(NMe 2 ) 2 (OArCH 2 ArO), and Ti(OArCH 2 ArO) 2 . ................................ ................................ ................................ ....................... 546 Figure 6.58 1 H NMR of the equilibrium mixture of Ti(OAr 4 - tert - butyl ) 4 , Ti(OAr 4 - tert - butyl ) 2 (OArCH 2 ArO), and Ti(OArCH 2 ArO) 2 . ................................ ................................ .............. 547 xxx Figure 6.59 Least Squares fit result for predicting K eq from LDP. ................................ .......... 549 Figure 7.1 Examples of terminal m ono - and bis - imido Fe complexes in the literature. Note the prevalence of both bulky and chelating ligands, which stabilize these complexes. 20 - 28 ............. 557 Figure 7.2 (top) X - omitted for clarity (N = blue, Ru = teal, P = pink). (bottom) ................................ ..................... 559 Figure 7.3 (top) Synthetic schemes for the synthesis of Ru2 and Ru3 . (bottom) X - ray crystal structure for Ru3 (left) (Ru(NAr)dppe(PMe 3 )) and Ru2 (right); ellipsoids shown at 50% probability, hydrogens omitted for clarity. For Ru2 , the two enantiomers co - cr ystalize and are disordered across the axis coincident with the P1 Ru1 P3 bond. Select bond lengths and angles are shown. ................................ ................................ ................................ ................................ ... 562 Figure 7.4 ( top ) Synthetic scheme for th e production of Ru4 from Ru3 via oxidation with AgSbF 6 (AgBArF 24 can also be used). ( bottom ) X - ray crystal structure of the dimeric species with ellipsoids shown at 50% probability; H atoms and disordered counter anion omitted for clarity. ................................ ................................ ................................ ................................ ......... 563 Figure 7.5 Proposed resonance contributors for Ru5 with radical distribution across the ortho and para positions of the imide aryl group. ................................ ................................ ................ 564 Figure 7.6 Synthesis procedure for Ru1* and Ru3* from cis - RuCl 2 (PMe 3 ) 4 starting material. The crystal structure of Ru1* is shown with ellipsoids are shown at 50% p robability; H atoms omitted for clarity. ................................ ................................ ................................ ...................... 566 Figure 7.7 (top) Synthesis of Fe analogues of Ru - imido complexes, Fe1 , Fe1* , Fe3 , and Fe3* . The side product, Fe2/2* , als o results from this synthetic route, and is separated from Fe3/3* by several extractions and recrystallizations. (bottom) Crystal structures of Fe2* and Fe3* ; ellipsoids are shown at 50% probability with H atoms omitted for clarity. Note, Fe2* crystallizes in the C2/c spacegroup, with the crystallographic 2 - fold axis bisecting the N1 Fe1 N1 and P1 Fe1 P1 angles. Thus, half of the molecule is symmetry generated. ................................ ........... 569 Figure 7.8 Synthesis of Fe5 and Fe6 by trapping the unstable Fe1 with the trischelating (PMe 2 CH 2 ) 3 Si t Bu ligand. X - ray crystal structures of Fe5 and Fe6 are shown for comparison; ellipsoids are shown at 50% probability with H atoms and counteranion ( Fe6 ) om itted for clarity. ................................ ................................ ................................ ................................ ..................... 572 Figure 7.9 M(III) cationic complexes examined by EPR spectroscopy. Fe6 is reasonably crystalline and has been characterized by single crystal X - ray crystallography. Ru4* and Fe4* are amorphous. The become oily when exposed to polar ethereal solvents, and attempts at crystallization generally produce powder solids. ................................ ................................ ........ 573 Figure 7.10 EPR spectra (black) of 2 different preparations of Fe4* (a and b), utilizing identical synthetic preparations. The insert shows a mixed Fe(III) radical species that seems to form as an impurity (or decomposition product) upon oxidati on of Fe3* to Fe4*. The spectrum shown in a is relatively pure, while b shows the Fe(III) mixed radical impurity superimposed on the spectrum xxxi of Fe4*. Red traces represent simulated spectra. For more simulation details, see Experimental. Spectra were recor ded at 10 K. ................................ ................................ ................................ ... 576 Figure 7.11 EPR spectra of two different preparations of Fe6 (black) and their simulated spectra (red), utilizing identical synthetic methods. Both samples show very similar characteristics, with about 75% Fe(III) - centered character and 25% ligand - centered radical with 14N hyperfine coupling. Two different lineshapes are noted in toluene (top) and 2 - methylTHF (bottom), but the radical ch aracter has very similar properties in both spectra. Spectra were recorded at 10K .... 578 Figure 7.12 EPR spectra of two different samples of Ru5* (black) and their simula ted spectra (red), utilizing identical synthetic methods. The samples are treated as a composite of two distinct paramagnetic centers one centered on Ru(III) and one centered on the imide fragment. The two spectra show dramatically different proportions of each radical center in the sample. Spectrum (a) is about 90% Ru(III) and 10% ligand radical, whereas spectrum (b) shows about 30% Ru(III) character and 70% ligand radical. These spectra were taken from different batches of Ru5* , prepared following the sam e synthetic procedures. The relative distribution of radical character seems to be affected by synthesis and sample preparation of the compounds. ........... 580 Figure 7.13 The f igure shows the radical localization in the Fe4* , Ru5* , and Fe6 cations. Note that with Fe4* , the cationic species is most accurately described by a single paramagnetic center. However, the Ru5* and Fe6 spectra could only be successfully modeled as a compo site of two paramagnetic species, where both a metal - and ligand - centered radical contribute to the entire spectrum. ................................ ................................ ................................ ................................ ..... 583 Figure 7.14 The orbital energies (defined as the negative ionization potential, as described in the text) of Fe7 , Ru6 , and Ru6_mod computed at the CASPT2 level of theory. Insets show the HOMO and HOMO - 1 orbitals (SONOs of the cations, as described in the text). A green arrow indicates the bonding lob e of the HOMO of Ru6 . ................................ ................................ ...... 586 Figure 7.15 Addition of CS 2 to the terminal Fe(II) imide, Fe5 , results in the production of 2,6 - diisopropylphenylthioisocyanate and an insoluble red species proposed to be [Fe( t P 3 - S)] 2 . . 590 Figure 7.16 Reaction of 2 equivalents of benzaldehyde with Fe5 pr oduces a metalacyclic species which is chiral, and appears to be diamagnetic, low spin Fe(II). Based on bond lengths and angles in the crystal structure, the two oxygens coordinated to Fe appear anionic, with N1 best described as an imine. This means that loss of H 2 has occurred during the reaction. The crystal structure ( right - hexane in the lattice were omitted for clarity. (Fe1 O1 = 1.985 Å, Fe1 O2 = 1.858 Å, O1 C1 = 1.291 Å, C1 N1 = 1.297 Å, O2 C9 = 1.416 Å). ................................ ................................ ....................... 591 Figure 7.17 Results demonstrating the electronic basis of the distortion of Ru1 away from C 3v symmetry. 46 ................................ ................................ ................................ ................................ . 596 Figure 7.18 14 N NMR of Ru1 in C 6 D 6 . ................................ ................................ ....................... 609 Figure 7.19 1 H NMR of Ru2 in C 6 D 6 . ................................ ................................ ........................ 610 Figure 7.20 13 C NMR of Ru2 in d 8 - THF. ................................ ................................ ................... 611 xxxii Figure 7.21 31 P NMR of Ru2 in d 8 - THF. ................................ ................................ ................... 612 Figure 7.22 14 N NMR of Ru2 in C 6 D 6 . ................................ ................................ ....................... 613 Figure 7.23 1 H NMR of Ru4 in CD 2 Cl 2 . ................................ ................................ .................... 614 Figure 7.24 31 P NMR of Ru4 in CD 2 Cl 2 . ................................ ................................ ................... 615 Figure 7.25 19 F NMR of Ru4 in CD 2 Cl 2 . ................................ ................................ ................... 616 Figure 7.26 1 H NMR of Ru3 in C 6 D 6 . ................................ ................................ ........................ 617 Figure 7.27 13 C NMR of Ru3 in C 6 D 6 . ................................ ................................ ....................... 618 Figure 7.28 31 P NMR of Ru3 in C 6 D 6 . ................................ ................................ ....................... 619 Figure 7.29 14 N NMR of Ru3 in C 6 D 6 . ................................ ................................ ....................... 620 Figure 7.30 1 H NMR of Ru1* in C 6 D 6 . ................................ ................................ ...................... 621 Figure 7.31 13 C NMR of Ru1* in C 6 D 6 . ................................ ................................ ..................... 622 Figure 7.32 31 P NMR of Ru1* in C 6 D 6 . ................................ ................................ ..................... 623 Figure 7.33 14 N NMR of Ru1* in C 6 D 6 . ................................ ................................ ..................... 624 Figure 7.34 1 H NMR of Ru2* in C 6 D 6 . ................................ ................................ ...................... 625 Figure 7.35 13 C NMR of Ru2* in C 6 D 6 . ................................ ................................ ...................... 626 Figure 7.36 31 P NMR of Ru2* in C 6 D 6 . ................................ ................................ ..................... 627 Figure 7.37 14 N NMR of Ru2* in C 6 D 6 . ................................ ................................ ..................... 628 Figure 7.38 1 H NMR of Ru3* in C 6 D 6 . ................................ ................................ ...................... 629 Figure 7.39 13 C NMR of Ru3* in C 6 D 6 . ................................ ................................ ..................... 630 Figure 7.40 31 P NMR of Ru3* in C 6 D 6 . ................................ ................................ ..................... 631 Figure 7.41 14 N NMR of Ru3* in C 6 D 6 . ................................ ................................ ..................... 632 Figure 7.42 31 P NMR of Fe1 in THF ( in situ ). ................................ ................................ ........... 633 Figure 7.43 14 N NMR of Fe 1 in THF ( in situ ). ................................ ................................ ........... 634 Figure 7.44 1 H NMR of Fe3 in C 6 D 6 . ................................ ................................ ......................... 635 Figure 7.45 13 C NMR of Fe3 in C 6 D 6 . ................................ ................................ ........................ 636 xxxiii Figure 7.46 31 P NMR of Fe3 in C 6 D 6 . ................................ ................................ ........................ 63 7 Figure 7.47 14 N NMR of Fe3 in C 6 D 6 . ................................ ................................ ........................ 638 Figure 7.48 31 P NMR of Fe1* in C 6 D 6 . ................................ ................................ ...................... 639 Figure 7.49 14 N NMR of Fe1* in C 6 D 6 . ................................ ................................ ...................... 640 Figure 7.50 1 H NMR of Fe3* in C 6 D 6 . ................................ ................................ ....................... 641 Figure 7.51 13 C NMR of Fe3* in C 6 D 6 . ................................ ................................ ...................... 642 Figure 7.52 31 P NMR of Fe3* in C 6 D 6 . ................................ ................................ ...................... 643 Figure 7.53 14 N NMR of Fe3* in C 6 D 6 . ................................ ................................ ...................... 644 Figure 7.54 1 H NMR of Fe8 in C 6 D 6 . ................................ ................................ ......................... 645 Figure 7.55 31 P NMR of Fe8 in C 6 D 6 . ................................ ................................ ........................ 646 Figure 7.56 1 H NMR of the C 6 D 6 soluble extracts from the reaction of Fe5 and CS 2 . The 1 H NMR shows H 2 NAr and SCNAr. ................................ ................................ ............................... 647 Figure 7.57 31 P NMR of Fe5 + 1,1 - dimethylhydrazine in C 6 D 6 . The peak at 47 pp m is the starting Fe5 . The peak at 52 ppm is a new compound. ( - 49 is free triphos ligand) .................... 648 Figure 7.58 0.000109 M, THF ................................ ................................ ................................ .... 649 Figure 7.59 0.00022 M THF ................................ ................................ ................................ ....... 650 Figure 7.60 0.00031 M, THF ................................ ................................ ................................ ...... 651 Figure 7.61 0.00031 M, THF ................................ ................................ ................................ ...... 652 Figure 7.62 0.00010 M, THF ................................ ................................ ................................ ...... 653 Figure 7.63 0.00044 M, THF ................................ ................................ ................................ ...... 654 Figure 7.64 0.000068 M, THF ................................ ................................ ................................ .... 655 Figure 7.65 0.000071 M, THF ................................ ................................ ................................ .... 656 Figure 7.66 0.000238 M, THF ................................ ................................ ................................ .... 657 Figure 7.67 0.00016 M, THF ................................ ................................ ................................ ...... 658 Figure 7.68 Crystal data and structure refinement for Pbca. ................................ ...................... 660 Figure 7.69 CV for Fe3* in THF with TBAPF 6 electrolyte . ................................ ...................... 664 xxxiv Figure 7.70 CV for Fe4* in THF with TBAPF6 electrolyte. ................................ ...................... 665 Figure 7.71 CV for Fe5 in THF with TBAPF6 e lectrolyte. ................................ ........................ 666 Figure 7.72 CV for Fe6 in THF with TBAPF6 electrolyte. ................................ ........................ 667 Figure 7.73 CV for Ru3* in THF with TBAPF6 electrolyte. ................................ ..................... 668 Figure 8.1 The synthetic procedure presente d by Wilkinson and coworkers in 1992 which lead to the square planar d 4 Ru species on the right. ................................ ................................ .............. 677 Figure 8.2 Examples of various Group 8 mono - and bis(imido) compo unds. ............................ 679 Figure 8.3 Illustration of the synthetic route proposed to access Ru - imido species in mid - to high oxidation states. ................................ ................................ ................................ ........................... 681 Figure 8.4 ( top) The X - ray crystal structure of Ru2 with ellipsoids shown at 50% probability; H atoms and solvent molecule omitted for clarity. (bottom) Synthetic 2 - diphenylhydrazido)Ru(PMe 3 ) 4 from in situ reduced azobenzene and cis - RuCl 2 (PMe 3 ) 4 . The same procedure can be utilized to produce the Os analogue of this compound. ................................ . 681 Figure 8.5 The Ru(II) terminal hydride (Ru3) species produced upon heating Ru2. The single crystal X - ray structure is shown with ellipsoids at 50% probability; H atoms and solvent omitted for clarity. ................................ ................................ ................................ ................................ .... 683 Figure 8.6 Ru2 and Ru4 . Although the photolysis product absorbs more strongly than the starting material across most of the spectrum, full conversion is still achieved in these photolytic conversions. (Note that the sharp feature at ~15,000 cm - 1 is due to the light source change in the UV - lamp). ..................... 685 Figure 8.7 ( left ) Schematic showing the interconversion of Ru2 , Ru4 , and Ru3 . ( right ) 31 P NMR of photolysis solution to generate Ru4 . The inset shows the new 31 P resonances, while the sharp singlet at - 62 ppm is free PMe 3 . ................................ ................................ ................................ .. 686 Figure 8.8 Several possible products that were considered in identifying Ru4 . The chart shows Ru2 to the product number lis ted on the x - axis. Each complex is numbered and shown structurally on the right. .. 687 Figure 8.9 The single crystal X - ray structure of Ru2 shown ( left ) with thermal el lipsoids and ( right ) as the spacefilling model with van der Walls radii on all atoms. Notice that essentially none of the central Ru atom is visible from the spacefilling perspective, demonstrating the steric crowding in this molecule. ................................ ................................ ................................ .......... 689 Figure 8.10 (top) Single crystal X - ray structure of Ru5. Thermal ellipsoids are shown at 50% probability and H atoms and solvent were omitted for clarity and spacefilling model of Ru5 . (bottom) General synthetic scheme for making Ru(II) tetrazene complexes. ............................ 691 xxxv Figure 8.11 Single crystal X - ray structure of Ru(NAr)2(PMe3)2 with ellipsoids shown at 50% probability and H atoms omitted for clarity. ................................ ................................ ............... 693 Figure 8.12 Absorption spectra for 0.002 M solutions of Ru5 (orange trace) and Ru1 (blue trace) in THF. T he strong absorbance of the product ( Ru1 ) across the UV - Vis spectrum likely contributes to the conversion limit of 25% in solution. ................................ .............................. 694 Figure 8.13 Reaction scheme to produce a Ru(II) imide species with bulkier phosphine ligands, PPhMe 2 , and subsequent lack of reactivity upon addition of aryl azide. ................................ .... 696 Figure 8.14 The formation of a 6 - coordinate, C H activated mesityl anilide species, resulting from an attempt to generate a terminal Ru(=NMes)(PPhMe 2 ) 3 . ................................ ................. 697 Figure 8.15 2 - diphenylacetylene)Os(NAr) 2 ( Os10 nucleophilic or electrophilic behavior via interaction of the unsaturated C C bond participants with the Os N multiple bond. The same synthetic method was applied to produce Ru10 . ....... 698 Figure 8.16 14 N NMR of Ru10 (387.1 ppm) and Os10 (365.6 ppm). ................................ ........ 701 Figure 8.17 Fragment for Ru1 observed by HRMS. ................................ ................................ .. 710 Figure 8.18 1 H NMR spectrum of Ru(PhNNPh)(PMe 3 ) 4 · PhMe ( Ru2 ) in C 6 D 6. ...................... 724 Figure 8.19 13 C NMR spectrum of Ru(PhNNPh)(PMe 3 ) 4 · PhMe ( Ru2 ) in C 6 D 6 . ..................... 725 Figure 8.20 31 P NMR spectrum of Ru(PhNNPh)(PMe 3 ) 4 · PhMe ( Ru2 ) in C 6 D 6. ..................... 726 Figure 8.21 1 H NMR spectrum of Ru4 (in situ) in C 6 D 6 . ................................ ........................... 727 Figure 8.22 13 C NMR spectrum of Ru4 (in situ) in C 6 D 6 . ................................ .......................... 728 Figure 8.23 31 P NMR spectrum of Ru4 (in situ) in C 6 D 6 . ................................ .......................... 729 Figure 8.24 1 H NMR spectrum of Ru3 in C 6 D 6 . ................................ ................................ ........ 730 Figure 8.25 13 C NMR spectrum of Ru3 in C 6 D 6 . ................................ ................................ ....... 731 Figure 8.26 31 P NMR spectrum of Ru3 in C 6 D 6 . ................................ ................................ ........ 732 Figure 8.27 1 H NMR of Ru5 in C 6 D 6 . ................................ ................................ ........................ 733 Figure 8.28 13 C NMR of Ru5 in C 6 D 6 . ................................ ................................ ....................... 734 Figure 8.29 31 P NMR of Ru5 in C 6 D 6 . ................................ ................................ ....................... 735 Figure 8.30 1 H NMR of Ru6 in C 6 D 6 . ................................ ................................ ........................ 736 Figure 8.31 13 C NMR of Ru6 in C 6 D 6 . ................................ ................................ ....................... 737 xxxvi Figure 8.32 31 P NMR of Ru6 in C 6 D 6 . ................................ ................................ ....................... 738 Figure 8.33 1 H NMR of Ru7 in C 6 D 6 . ................................ ................................ ........................ 739 Figure 8.34 13 C NMR of Ru7 in C 6 D 6 . ................................ ................................ ....................... 740 Figure 8.35 31 P NMR of Ru7 in C 6 D 6 . ................................ ................................ ....................... 741 Figure 8.36 1 H NMR of photolysis reaction containing a mixture of Ru1 , Ru5 (starting material), and H 2 NAr (decomposition byproduct). ................................ ................................ ..................... 742 Figure 8.37 31 P NMR spectrum of Ru1 (after extraction and repeated recrystallization) in C 6 D 6. ................................ ................................ ................................ ................................ ..................... 743 Figure 8.38 13 C NMR spectrum of Ru1 in C 6 D 6 . ................................ ................................ ....... 744 Figure 8.39 14 N NMR of Ru1 in C 6 D 6. ................................ ................................ ....................... 745 Figure 8.40 QTOF - HRMS fragmentation patterns (top) calculated and (bottom) experimental for Ru1 . ................................ ................................ ................................ ................................ ............ 746 Figure 8.41 1 H NMR spectrum of Ru(NAr) 2 2 - diphenylacetylene) containing diphenylacetylene (excess) and H 2 NAr impurities. ................................ ................................ ................................ .. 747 Figure 8.42 14 N NMR spectrum of Ru(NAr) 2 2 - diphenylacetylene) in C 6 D 6 (in situ). ............ 748 Figure 8.43 1 H NMR of Ru8 in C 6 D 6 . ................................ ................................ ........................ 749 Figure 8.44 13 C NMR of Ru8 in C 6 D 6 . ................................ ................................ ....................... 750 Figure 8.45 31 P NMR of Ru8 in C 6 D 6 . ................................ ................................ ....................... 751 Figure 8.46 1 H NMR of Ru9 in C 6 D 6 . ................................ ................................ ........................ 752 Figure 8.47 13 C NMR of Ru9 in C 6 D 6 . ................................ ................................ ....................... 753 Figure 8.48 31 P NMR of Ru9 in C 6 D 6 . ................................ ................................ ....................... 754 Figure 8.49 1 H NMR of ArNPMe 3 in C 6 D 6 . ................................ ................................ ............... 755 Figure 8.50 13 C NMR of ArNPMe 3 in C 6 D 6 . ................................ ................................ .............. 756 Figure 8.51 31 P NMR of ArNPMe 3 in C 6 D 6 . ................................ ................................ .............. 757 Figure 8.52 1 H NMR of Os3 in C 6 D 6 . ................................ ................................ ........................ 758 Figure 8.53 31 P NMR of Os3 in C 6 D 6 . ................................ ................................ ........................ 759 Figure 8.54 14 N NMR of Os(NAr) 2 (O) 2 . ................................ ................................ ..................... 760 xxxvii Figure 8.55 14 N NMR of Os(NAr) 2 (PMe 3 ) 2 . ................................ ................................ ............... 761 Figure 8.56 14 N NMR of Os(NAr) 2 2 - diphenylacetylene). ................................ ....................... 762 Figure 8.57 Plo Ru1 (0.000188 M in THF). ................................ ....... 763 Figure 8.58 Ru2 (0.00023 M in THF). ................................ ......... 764 Figure 8.59 Ru4 (0.00019 M in THF). ................................ ......... 765 Figure 8.60 Ru5 (0.000203 M in THF). ................................ ...... 766 Figure 8.61 er for Ru6 (0.00030 M in THF). ................................ ......... 767 Figure 8.62 Ru7 (0.00031 M in THF). ................................ ......... 768 Figure 8.63 Photochemical irradiation setup utilizing a mercury arc lamp. ............................... 769 Figure 8.64 Crystal data and structure refinement for p21c. ................................ ...................... 771 Figure 8.65 Crystal data and structure refinement for early_a. ................................ .................. 773 Figure 8.66 Crystal data and structure refinement for KA_OsBisImido. ................................ ... 775 Figure 8.67 Crystal data and structure refinement for smalltwin5. ................................ ............ 777 xxxviii LIST OF SCHEMES Scheme 4.1 One - pot - two - step quinoline synthesis of 2 - methyl - 3 - phenyl - 6 - ( N,N - dimethy lamino)quinoline utilizing [Ti]700 to perform the initial iminoamination reaction. ..... 288 Scheme 4.2 A typical iminoamination reaction utilizing [Ti]700 as the catalyst material and the organic residues found upon a mild acid - wash of the used catalyst materia ............ 292 Scheme 4.3 [Ti]700 catalyzed guanylation of 1,3 - dicyclohexylcarbodiimide. .......................... 294 Scheme 4.4 HA Entry 1 synthesis and isolation. ................................ ................................ ........ 305 Sche me 4.5 HA Entry 2 synthesis and isolation. ................................ ................................ ........ 306 Scheme 4.6 HA entry 3 synthesis and isolation. ................................ ................................ ......... 307 Scheme 4.7 HA entry 4 synthesis and isolation ................................ ................................ .......... 308 Scheme 4.8 HA entry 7 synthesis and isolation. ................................ ................................ ......... 309 Scheme 4.9 HA entry 8 synthesis and isolation. ................................ ................................ ......... 310 Scheme 4.10 HA entry 12 synthesis and isolation. ................................ ................................ ..... 311 Scheme 4.11 HA entry 13 synthesis and isolation. ................................ ................................ ..... 312 Scheme 4.12 HA entry 14 synthesis and isolation. ................................ ................................ ..... 313 Scheme 4.13 Targeted synthesis of asymmetric urea species. ................................ .................... 318 Scheme 5.1 Addition of Brønsted acidic HX ligands to [Ti]700 to generate [Ti]700(X) species. ................................ ................................ ................................ ................................ ..................... 365 Scheme 5.2 Iminoamination reaction examined with a variety of HX ligands. ......................... 367 Scheme 5.3 Quinoline synthesis using 5 mol% [Ti]700 with 5 mol% 2,6 - dimethylphenylamidate as ligand. ................................ ................................ ................................ ................................ ..... 370 Scheme 5.4 General conditions and substrates used to examine the iminoamination reaction with [Ti] 700(X) catalyst. ................................ ................................ ................................ ..................... 373 Scheme 5.5 Iminoamination reaction catalyzed by [Ti]700 with 2,4,6 - tri - tert - butylphenoxide. 407 Scheme 5.6 Iminoamination reaction catalyzed by [Ti]700 with 2,6 - dimethylphenylamidate ligand and an internal alkyne. ................................ ................................ ................................ ..... 409 Scheme 5.7 Iminoamination reaction catalyzed by [Ti]700 with 2,6 - dimethylphenylamidate and an aromatic alkyne. ................................ ................................ ................................ ..................... 411 xxxix Scheme 5.8 Synthesis of 4 - Br - phenoxide LDP complex. ................................ ........................... 415 Scheme 6.1 General iminoamination reaction and substrates examined to determine the effect of each substrate on the rate of the overall reaction. ................................ ................................ ....... 450 Scheme 6.2 The Ti(dpm)(NMe 2 ) 2 complex reacts with the iminoamination product to yield an intractable mixture. ................................ ................................ ................................ ..................... 459 - Ntolyl)] 2 . ................................ ................................ ................................ ................................ ..... 461 Scheme 6.4 Proposed decomposition pathway observed when both t BuNC and iminioamination - Ntolyl)] 2 . ................................ ............................. 462 - - NAr species upon the addition of an excess of H 2 NAr. With a coordinated H 2 NAr, there are several different pathways conceivable b y which the anilides are exchanged by quick proton transfer steps. ................................ ......... 463 Scheme 6.6 Iminoamination reaction catalyzed by Ti(OArCH 2 ArO)(NMe 2 ) 2 following standard reaction conditions. ................................ ................................ ................................ ..................... 467 Scheme 6.7 General equilibria observed for several Ti(IV) complexes that have be en noted in the literature and through our observations. ................................ ................................ ..................... 469 Scheme 6.8 Production of 4CC product from iminoamination reaction mixture. ...................... 483 Scheme 6.9 Iminoamination reaction examined under different reaction conditions to probe the rate law and suggest optimal reaction conditions. ................................ ................................ ...... 537 Scheme 8.1 Photochemical conversion to yield Ru1 from Ru5 . ................................ ............... 711 xl KEY TO SYMBOLS OR AB BREVIATIONS LDP ligand donor parameter THF tetrahydrofuran DCM dichloromethane DME dimethoxyethane Dppe diphenylphosphino ethane Tol toluene Bnz benzene p - cym para - cymene hex hexanes pent pentanes TEA triethylamine DBU 1,8 - Diazabicyclo[5.4.0]undec - 7 - ene COD cycloactadiene r.t. room temperature NMR nuclear magnetic resonance (spectroscopy) SST spin saturation transfer DOSY diffusion ordered spectroscopy NOESY nuclear overhauser effect spectroscopy ROESY rotating frame overhauser effect spectroscopy MW molecular weight D diffusion coefficient GC gas chromatography xli MS mass spectrome try FID flame ionization detector ICP - OES inductively coupled plasma optical emission spectroscopy EA elemental analysis HRMS high resolution mass spectrometry EI electrospray ionization %V bur percent buried volume Tolman cone angle e.u. entropy units DFT density functional theory NBO natural bonding orbitals NRT natural resonance theory MO molecular orbital (theory) HOMO highest occupied molecular orbital LUMO lowest unoccupied molecular orbital CV cyclic voltammogram UV ultraviolet (radiation) Vis visible (radiation) Std. dev. Standard deviation Dpm - dimethyldipyrrolylmethane Ar 2,6 - diisopropylphenyl Ar* 2,4,6 - triisopropylphenyl Mes 2,4,6 - trimethylphenyl xlii (OArCH 2 ArO) 2,2' - Bis(4 - m ethyl - 6 - tert - butylphenol)methane H enthalpy of activation G free e nerg y of activation S en tropy of activation BArF 24 - tetrakis(3,5 - (trifluoromethyl)phenyl)borane BArF 20 - tetrakis(2,3,4,5,6 - fluoro - phenyl)borane N 4 R 1,4 - disubstituted te trazene ligand 1 CHAPTER 1. INTRODUCTION 1.1 Traditional Approach es to Transition Metal Catalysis the case. The discovery of good catalysts is often the result of serendipity. A typical catalytic reaction might start with the synthesis of an electronically inte resting transition metal complex. Efforts to determine the details of the electronic structure for the complex may then be pursued, and include exploratory reactivity studies with the compound and a variety of reagents with different properties. O ver the c ourse of these reactivity studies, a reaction is discovered in which the transition metal complex or material acts as a catalytic reagent . With the newly discovered catalyst, chemists then turn to catalyst design principles for improvement of the catalytic reaction to achieve faster rate, improved selectivity, expanded substrate scope tolerance, higher yields, etc. There are several good examples of this process of catalyst development in the literature. One well - efin metathesis. It was discovered in the 3 was a precatalyst for olefin metathesis. 1, 2 Despite the fact that the active species in the catalytic cycle was not understood, application of this catalyst to industrial proc esses was undertaken . Some 20 - 30 years later, Grubbs and coworkers published a well - understood molecular precatalyst with phosphine ligands. 3 This species was modified several times by Grubbs and others in the following years until the y develop ed the current generations of these Ru catalyst s . 4 - 9 With these successive changes, expansion of substrate tolerance, impro ved precatalyst tolerance to air and moisture, and additional controls were gained in the catalyst along the way. 2 This journey in catalyst development spans 5 0 years. While it has resulted in great diversificati on of substrates and application s , the optimization of an industrial catalyst that has a direct impact on society. From a chemistry perspective, the evolutionary think about it in terms of society, 60 years is close to a human lifetime. When we consider that many consumer products, such as pharmaceuticals, take human lifetime s to develop, there is strong motivation to make catalyst development faster and more directed. 3 Figure 1 . 1 (top) A simple example of an olefin metathesis reaction, cross metathesis (CM) that can be catalyzed by Ru - carbene complexes. (bottom) Examples of several viable Ru catalysts competent for CM and other olefin metathesis reactions. The real problem facing streamlined transition metal catalyst development is that n o single rule can be applied to the entirety of the transition metal series, and to some extent, the properties of metal reactivity are related to the metal itself and cannot be changed (i.e. electronegativity or ionic radii) . Rather, the most influential and general trends th at exist among transition metal complexes is in the properties of the bonds that they make to other elements, and the subsequent effects that these bonds have on reactivity. Thus, the ligands in a cataly tic system are where chemists can often exert the mos t control over the reactivity of a metal complex. T his applies to ancillary ligands, or ligands which remain bound to a metal throughout the course of a reaction but 4 that do not directly participate in bond making or breaking , as well as ligands that parti cipate directly in reactivity. Collectively, the projects briefly introduced here and discussed in detail in this dissertation, touch on each step of catalyst development mentioned here: (1) identification of trends in metal - ligand bond interactions that correlate to the stereoelectronic nature of the ligands (Chapters 2 and 3); (2) experimental applications of these trends to catalytic systems, and the use of fundamental metal - ligand interaction models to predict catalyst behavior and improve it (Chapter s 4 - 6); and (3) fundamental reactivity and discovery of new reaction s through synthesis of electronically unique complexes (Chapters 7 and 8) . 1.2 Targeted Design of Catalysts: Ligand Donor Parameter Generally, some of the most well - understood transition meta l chemistry is that of the noble metals, such as Pd, Pt, Rh, and Ir. These metals are essential to modern life, providing a variety of commodity chemicals and products from some of the most prolific anticancer drugs (i.e. cis - platin) , to catalytic converte rs . Part of what makes these metals so successful as catalytic metals is their predictability, which is aided by the development of tools that guide their design and improvement (see below). One of the most powerful tools of this type is the ability to pre dict what effect changes to an ancillary ligand will render upon a catalyzed reaction. This predictive ability is contingent upon accurate parameterization of the properties of a series of ligands. When considering ligand properties , t wo main aspects of the ligand metal interaction dominate the minds of organometallic chemists: electronics and sterics of the ancillary ligands. Electronic properties stem from electron donor - acceptor interactions between the vacant or filled orbitals on the trans ition metal and the bonding atom of the ancillary ligand. Sterics (size) affect the amount of space available in the first coordination sphere of a metal , which heavily influences 5 substrate interactions with the metal during the course of a catalyzed react ion. Together, these two properties are the central pillars in tuning reactivity of metals. T o compare the effects of ligand properties on a catalyzed reaction, systematic parameterization of the electronic and steric properties of each ligand is needed t o determine trends . A classic example of ligand parameterization is the Tolman Electronic Parameter and the Tolman Cone Angle (TEP and TCA). 10 - 13 In what is now an iconic system of ligand classification, Tolman focused on quantifying the electron donor ability and size of a wide range of phosphine (PR 3 ) ligands. In his initial study , a series of Ni(CO ) 3 PR 3 complexes were synthesized, and the A 1 s tretching frequencies (breathing mode) of the CO ligands was measured by IR spectroscopy. In this system, a more donating phosphine would result in more electron density on the metal, which - backbonding to the CO ligands. This electron pushing is directly observable as an elongation in the CO bond, which correlates to a lower frequency C - O stretch in the IR spectrum. It is important to note that in this system, a phosphine with electron withdrawing R - groups, such as PX 3 ligands where X = - effects. The PR 3 - acid, accepting electron density from the metal through an interaction dubbed negative hyperconjugation (see Chapter 3) . This generates double - bond character between the ligand and metal (M=P) - donor will decrease the C - - acceptor will increase it by competing with the - backdonation. This makes the TEP a holistic electronic p arameter which - - donor abilities of PR 3 ligands. In a similar way, Tolman also set about quantifying the size of PR 3 ligands. He used a 2.28 Å bond length, representing a typical Ni 0 P bond length, and set the Ni as the origin point of a 6 conical section of space. He then arranged the PR 3 of interest into its tightest formation of the R groups and measured what the angle of the 3 - dimensional cone of space was, occupied by the PR 3 ligand. This measurement became known as the Tolman Cone Angle (TCA) and is still invoked today to draw trends with PR 3 ligand sizes. Figure 1 . 2 Illustration of the original Tolman system for phosphine parameterization. (left) The Ni(CO) 3 PR 3 complex was used to determine overall donor ability of a phosphine. (right) The system utilized to determine a steric profile for a given PR 3 ligand where the metal PR 3 ligands is, there are of course limitations. One m has - valent, late transition metals. There is no complementary system which addresses the donor ability of PR 3 ligands to high oxidation state, early transition metals. This proble m can be more generally addressed to all ligands, not just phosphines. Few examples of systematic ligand effect studies exist with high valent (or high oxidation state) catalyst systems. Generally speaking , h igh valent metals in high oxidation states , particularly those early in the transition metal series, have been neglected in fundamental studies of catalyst development and tool development of catalytic chemists as ca talysts. For example , the group IV metals are utilized in processes such as Sharpless epoxidation and olefin polymerization. 14 - 20 Yet, despite their importance in these and other developed for these early, high valent metals. 7 In response to the lack of predictive tools available to high valent transition metal catalysis, i.e. a complemen are d 0 , the Odom group pioneered the Ligand Donor Parameter (LDP) system in 2012. 21 This system is based on a 4 - c oordinate chromium(VI)nitride bis(diisopropylamide) fragment (NCr(N i Pr 2 ) 2 ) ; the 4 th coordination site to Cr in this fragment can be widely varied using straightforward synthetic techniques (NCr(N i Pr 2 ) 2 X). In the original, and several subsequent studies, X is a monoanionic ligand. This molecule is illustrated below in Fi g. 3 . 22 - 24 Figure 1 . 3 The NCr(N i Pr 2 ) 2 X scaffold used to measure the LDP value of a given X ligand. The illustration shows the rotation of one of the Cr N i Pr 2 bonds. The transition state (center) forces the amide lone pair into an antibonding orbital. The highlighted protons on the i Pr groups have unique chemical shifts. Their exchange, monitored by 1 H NMR SST, facilitates the measurement of the rotation rate. In this system, the two N i Pr 2 ligands and the variable X ligand compete for vacant d - orbitals on Cr in the xy - plane. This electronic competition results in variable donation of the amide lone pairs to Cr. When X is a strong donor, less amide lone pair don ation to Cr occurs, resulting in single - bond character between the Cr and N i Pr 2 ligands . In contrast, when X is a weak donor, more amide lone pair donation to Cr occurs, resulting in double - bond character between Cr and N i Pr 2 ligands . This variable donatio n of the amide lone pairs, leads to variability in Cr N bond character which changes the rate of Cr N bond rotation (for more details, see Chapters 2 and 3). This bond rotation rate is the experimental handle that we can use to make quantitative comparison s of donor ability of X to a high valent metal, analogous to the C system above. More recently, a correlation with the LDP value of ancillary ligands of Ti(IV) hydroamination catalysts and the rate constant of th e catalyzed reaction was established by our 8 group. 25 In fact, a quantitative model was developed that relates the rate constant of the reaction to the electronics (LDP) and sterics (%V bur ) for this specific reaction. Knowing the stereoelectronic parameters of candidate ligands provides the researcher a guid e to making a faster catalyst while bypassing the need to synthesize and screen countless catalyst in order to start finding better ligands. 25 Having established what a valuable tool the LDP system can provide for high valent metal catalyst development, we wanted to broaden the scope of ligands parameterized using the LDP system. One category of ligands which w , including amines, pyridines, nitriles, isonitriles, and phosphines. Such ligands are ubiquitous as ancillary ligands in catalytic systems, and their interactions with high valent metals have been largely ignored. The to suspect it would be an ideal system with which to probe high - valent metal interactions with phosp h ines. We dedicated a lot of time and effort to determinin g how to transition the LDP system to the application of neutral ligands, using phosphines as the guinea pig ligands. The results of these studies, including efforts to establish systematic differences between the neutral Cr complexes with monoanionic X li gands, and the cationic Cr complex with neutral X ligands, are presented herein. 26, 27 1.3 using Silica Supported Titanium Catalysts In addition to studies outlining the donor ability of various ligands to high valent metals, which has been a major focus in the Odom group over the last two decades, C N bond forming reactions are another are of research interest. 28, 29 Specifically, the development of homogeneous Ti species that catalyzed hydroamination, iminoamination, and hydrohydrazination has led to the 9 development of several catalysts which can perform these processes in moderate to high yields. Even when great care is taken to improve these typically solve every problem. For example, problems that remain unsolved by ligand design and manipulation include: off - cycle resting states of catalyst species, catalyst species reactivity with products and substrates (deactivation), and separation of the catalyst from organic reaction mixtures (see C hapters 5 and 6) . 30 These ongoing issues prompted us to pursue a new direction with these organometallic catalysts, specifically those for iminoamination : supported organometallic titanium species. Specifically, we wanted to pursue silic a - gel supported titanium species that retain similar coordination environments to homogeneous systems at the active titanium center , as heterogeneous catalysis of this sort has been shown an effective strategy for solving the problems mentioned above. 31 This dramat ic change in catalyst speciation, going from a typical homogeneous species to a silica - supported variant, we predicted, would render dramatic changes in the reactivity observed, in terms of reaction conditions, substrate tolerance, reaction times, etc. How ever, because the active metal sites retain a similar environment to the homogeneous catalysts, we also wondered if the tools developed with homogeneous systems 25 would apply to these heterogeneous titanium systems. We were able to develop two variants of silica - supported titanium materials competent for C N bond formation catalysis. The more highly dehydroxyla ted silica gel results in a material that can be ligand - functionalized, which provides rapid, selective , and nearly quantitative conversion of several sets of substrates to the iminoamination products. Practical application of this catalyst material has pr oven very promising, as these clean reactions with high conversion have provided heterocycle syntheses and isolation which are an improvement on their known homogeneous 10 analogues. This led to the ligand functionalization studies and kinetic analysis presen ted in Chapter 5 , from which we can propose a mechanism for the catalyzed reaction using the heterogeneous catalyst . Somewhat unexpectedly, these studies also lead us to reexamine the go - to homogeneous catalyst for similar transformations. Kinetic studies undertaken with Ti(dpm)(NMe 2 ) 2 to perform iminoamination reactions catalytically, have guided ligand modifications to improve the rate of this catalyzed reaction. These studies provided a more reactive catalyst but also illustrated a gap in our understand ing of ligand effect in the Ti(IV) hydroamination and iminoamination catalysts. This sparked investigations into ligand exchange reactions observed commonly with high valent transition metal complexes , based on our understanding of ligand donor abilities ( LDP) to high valent transition metals. Collectively, these studies of Ti(IV) C N bond forming catalysts have illustrated that these Ti(IV) systems can benefit from the use of a solid - support. In addition, these studies hav e also demonstrated that new appl ications of the LDP system can improve our understanding of catalysts beyond simple rate and catalyst design correlations. Understanding metal - ligand interactions is as applicable to solid - state heterogeneous catalyst systems with isolated metal sites as i t is in the homogeneous systems where they are typically studied. Indirectly, these studies have also produced a model of how ligand electronics influence ligand exchange reactions in the Ti(IV) catalyst systems, which can be directly applied to improve ca talyst design and performance. 1.4 Electronic Exploration of Unique Transition Metal Complexes : Valency Effects on Metal - Imide Bond Character Targeted ligand studies like those discussed with the Tolman system and LDP are highly useful with systems where a c atalytic species has been identified and its general reactivity can be 11 established. This area of research does nothing to promote the discovery of new catalysts and reactivity, however. Elements toward the middle of the d - block, Groups 5 - 8, tend to access a wide variety of oxidation states, which makes the m unique. However, systematic studies of how the oxidation state of the se metal s might affect the nature of a bond to a given category of ligands are not common, which makes reactivity predictions difficult to make. Specifically, one of the classes of ligands and associated reactivity that we are interested in is imidos. The C N bond forming reactions studied by our research group involve titanium imido species, which are typically recognized as the active catalyst in these types of reactions. The subsequent catalytic activity observed with these systems is a result of [2 + 2] cycload dition with an unsaturated C C bond. 32 - 35 This type of reactivity is typical of a high valent metal imide bond, however, in systems with late, low valent metal - imides, observed reactivity typically involves nucleophilic character at the nitrogen or the metal participating in the bond . With these pieces of knowledge in place, our next question is what happens in between? With a metal in a mid - valent oxidation state will a metal - imide bond react like those found in high or low valent systems? Or is there unique reactivity characterizing this middle - ground? With a late transition metal in a high oxidation state, how does the reactivity change as the oxidation state increases? The group VIII metals are no exception to this knowledge gap , despite how many examples there are of M=NR complexe s where M = Fe, Ru, or Os. Specifically, the chemistry of Fe imides has expanded rapidly in the last few decades, perhaps most especially due to interest in nitrogen fixation and related processes (see Chapter 7) . By comparison, there fewer complexes known with osmium, and fewer still with ruthenium. In trying to make comparisons in reactivity, a lack of synthetic diversity, particularly with Ru became apparent. Thus, interest in these types of 12 complexes became multi - fold. First, we wanted to expand the syn thetic diversity of these types of complexes. This work was begun in 2014 36 and has continued in the group until now. One of the main focuses of this aspect of these studies has been to push the Group VIII metal imide complexes into mid - valent oxidation states, where there are the fewest examples of known complexes. As early studies were pursued, a second goal, in line with several of our overall interests emerged, which was to identify electronic structure changes associated with oxidation of the metal in a series of 4 - coordinate terminal Fe and Ru imide complexes. Hoping to find answers to some of our fundamental questions about the nature of M=NR bonds and how they are affected by the oxidation state in the metal, synthetic and subsequent electronic studies have been undertaken with Fe(II) and Fe(III), as well as Ru(II), (III), and (IV) imide species. Studies focused on electronic structure changes upon oxidation of the metal have provided interesting insight about both Fe and Ru bonds to N. These results also highlight what has been experimentally obser ved across these and related studies, which is that synthetic difficulties are perhaps still the biggest hurdle to answering these basic questions about metal - imide bonds with the group VIII metals, as well as other metals in the middle of the d - block. 13 REFERENCES 14 REFERENCES ( 1 ) Michelotti, F. W.; Keaveney, W. P., Coordinated polymerization of the bicyclo - [2.2.1] - heptene - 2 ring system (norbornene) in polar media. 1965, 3 (3), 895 - 905. ( 2 ) Porri, L.; Rossi, R.; Diversi, P.; Lucherini, A., Ring - Opening polymerization of cycloolefins with catalysts derived from ruthenium and iridium. 1974, 175 (11), 3097 - 3115. ( 3 ) Nguyen, S. T.; Johnson, L. K.; Grubbs, R. H.; Ziller, J. W., Ring - opening me tathesis polymerization (ROMP) of norbornene by a Group VIII carbene complex in protic media. J. Am. Chem. Soc. 1992, 114 (10), 3974 - 3975. ( 4 ) Vougioukalakis, G. C.; Grubbs, R. H., Ruthenium - Based Heterocyclic Carbene - Coordinated Olefin Metathesis Catalyst s. Chem. Rev. 2010, 110 (3), 1746 - 1787. ( 5 ) Mohr, B.; Lynn, D. M.; Grubbs, R. H., Synthesis of Water - Soluble, Aliphatic Phosphines and Their Application to Well - Defined Ruthenium Olefin Metathesis Catalysts. Organometallics 1996, 15 (20), 4317 - 4325. ( 6 ) G rubbs, R. H.; Chang, S., Recent advances in olefin metathesis and its application in organic synthesis. Tetrahedron 1998, 54 (18), 4413 - 4450. ( 7 ) Schwab, P.; France, M. B.; Ziller, J. W.; Grubbs, R. H., A Series of Well - Defined Metathesis Catalysts Synth esis of [RuCl2( 1995, 34 (18), 2039 - 2041. ( 8 ) Nguyen, S. T.; Grubbs, R. H.; Ziller, J. W., Syntheses and activities of new single - component, ruthenium - based olefin metathesis catalysts. Journal of the American Chemical Soci ety 1993, 115 (21), 9858 - 9859. ( 9 ) Schwab, P.; Grubbs, R. H.; Ziller, J. W., Synthesis and Applications of Journal of the American Chemical Society 1996, 118 (1), 100 - 110. ( 10 ) Tolman, C. A., Steric effects of phosphorus ligands in organometallic chemistry and homogeneous catalysis. Chem. Rev. 1977, 77 (3), 313 - 348. ( 11 ) Tolman, C. A., Electron donor - acceptor properties of phosphorus ligands. Substituent additivity. J. Am. C hem. Soc. 1970, 92 (10), 2953 - 2956. ( 12 ) Dias, P. B.; Depiedade, M. E. M.; Simoes, J. A. M., Bonding and Energetics of Phosphorus(Iii) Ligands in Transition - Metal Complexes. Coord. Chem. Rev. 1994, 135 , 737 - 807. ( 13 ) Levason, W., Phosphine complexes of transition metals. In Organophosphorus Comp ounds (1990) , Hartley, F. R., Ed. 1990. 15 ( 14 ) Katsuki, T.; Sharpless, K. B., The first practical method for asymmetric epoxidation. Journal of the American Chemical Society 1980, 102 (18), 5974 - 5976. ( 15 ) Heravi, M. M.; Lashaki, T. B.; Poorahmad, N., Appli cations of Sharpless asymmetric epoxidation in total synthesis. Tetrahedron: Asymmetry 2015, 26 (8), 405 - 495. ( 16 ) Pfenninger, A., Asymmetric Epoxidation of Allylic Alcohols: The Sharpless Epoxidation. Synthesis 1986, 1986 (02), 89 - 116. ( 17 ) Li, J. J., Sha rpless asymmetric epoxidation. In Name Reactions: A Collection of Detailed Mechanisms and Synthetic Applications Fifth Edition , Springer International Publishing: Cham, 2014; pp 552 - 554. ( 18 ) Ziegler - Natta Catalysts. In Technology . ( 19 ) Mülhaupt, R., Catalytic Polymerization and Post Polymerization Catalysis Fifty Years After the Discovery of Ziegler's Catalysts. Macromolecular Chemistry and Physics 2003, 204 (2), 289 - 327. ( 20 ) Alt, H. G.; Köppl, A., Effect of the Nature of Metallocene Complexes of Group IV Metals on Their Performance in Catalytic Ethylene and Propylene Polymerization. Chemical Reviews 2000, 100 (4), 1205 - 1222. ( 21 ) DiFranco, S. A.; Maciulis, N. A.; Staples, R. J.; Batrice, R. J.; Odom, A. L., Evaluati on of donor and steric properties of anionic ligands on high valent transition metals. Inorg. Chem. 2012, 51 (2), 1187 - 200. ( 22 ) Bemowski, R. D.; Singh, A. K.; Bajorek, B. J.; DePorre, Y.; Odom, A. L., Effective donor abilities of E - t - Bu and EPh (E = O, S, Se, Te) to a high valent transition metal. Dalton Transactions 2014, 43 (32), 12299 - 12305. ( 23 ) Billow, B. S.; Bemowski, R. D.; DiFranco, S. A.; Staples, R. J.; Odom, A. L., Synthesis and Structure of Chromium(VI) Nitrido Cyclopentadienyl Complexes. Organometallics 2015, 34 (18), 4567 - 4573. ( 24 ) Beaumier, E. P.; Billow, B. S.; Singh, A. K.; Biros, S. M.; Odom, A. L., A complex with nitrogen single, double, and triple bonds to the same chromium atom: synthesis, structure, and reactivity. Chemical Science 2016 . ( 25 ) Billow, B. S.; McDaniel, T. J .; Odom, A. L., Quantifying ligand effects in high - oxidation - state metal catalysis. Nature Chemistry 2017, 9 , 837. ( 26 ) Aldrich, K. E.; Billow, B. S.; Holmes, D.; Bemowski, R. D.; Odom, A. L., Weakly Coordinating yet Ion Paired: Anion Effects on an Inte rnal Rearrangement. Organometallics 2017, 36 (7), 1227 - 1237. ( 27 ) Aldrich, K. E.; Billow, B. S.; Staples, R. J.; Odom, A. L., Phosphine interactions with high oxidation state metals. Polyhedron 2019, 159 , 284 - 297. 16 ( 28 ) Odom, A. L., New C - N and C - C bond f orming reactions catalyzed by titanium complexes. Dalton Transactions 2005, (2), 225 - 233. ( 29 ) Odom, A. L.; McDaniel, T. J., Titanium - Catalyzed Multicomponent Couplings: Efficient One - Pot Syntheses of Nitrogen Heterocycles. Acc. Chem. Res. 2015, 48 (11), 2822 - 2833. ( 30 ) van Leeuwen, P. W. N. M., Decomposition pathways of homogeneous catalysts. Applied Catalysis A: General 2001, 212 (1), 61 - 81. ( 31 ) Pelletier, J. D. A.; Basset, J. - M., Catalysis by Design: Well - Defined Single - Site Heterogeneous Catalysts. Ac c. Chem. Res. 2016, 49 (4), 664 - 677. ( 32 ) Hydroamination of Alkynes. Angew. Chem. Int. Ed. 2001, 40 (12), 2305 - 2308. ( 33 ) Walsh, P. J.; Hollander, F. J.; Bergman, R. G., Genera tion, alkyne cycloaddition, arene carbon - hydrogen activation, nitrogen - hydrogen activation and dative ligand trapping reactions of the first monomeric imidozirconocene (Cp2Zr:NR) complexes. J. Am. Chem. Soc. 1988, 110 (26), 8729 - 8731. ( 3 4) Substantially Enhanced Reactivity from an Unexpected Cyclopentadienide/Amide Ligand Exchange. J. Am. Chem. Soc. 2001, 123 (12), 2923 - 2924. ( 35 ) Nugent, W. A.; Mayer, J. M ., Metal - ligand multiple bonds: the chemistry of transition metal complexes containing oxo, nitrido, imido, alkylidene, or alkylidyne ligands . Wiley: 1988. ( 36 ) S ingh, A. K.; Levine, B. G.; Staples, R. J.; Odom, A. L., A 4 - coordinate Ru(ii) imido: unusua l geometry, synthesis, and reactivity. Chem. Commun. 2013, 49 (92), 10799 - 10801. 17 CHAPTER 2. PROBING THE IN SITU DYNAMICS OF THE LDP SYSTEM WITH NEUTRAL LIGANDS 2.1 Introduction 1 , 2 The ligand donor parameter (LDP) is an experimentally based system that allows for a quantitative comparison of the donor ability of various ligands to a high valent metal. 1 The system utilizes a chromium(VI)nitride bis(diisopropylamido) fragment, where the 4 th coordination site can be occupied by the ligand of interest, (X ¯ ). The system has been used to quantify the donor ability of a wide variety of anionic ligands, from amides to halogens. 1 - 5 A selection of these ligands and their LD P values are shown in Fig. 2.1. As briefly mentioned in chapter 1, this system has three strongly donating nitrogen ligands. The Cr(VI) metal center, formally a d 0 metal, has empty d and s acceptor orbitals. The orientation of the ligands around the metal creates an electronic competition between the N i Pr 2 and the X ¯ ligand for the xy - plane centered acceptor orbitals. 1 When X ¯ is a strong donor and pushes more nitrogens. Conversely, when X ¯ is a weak donor, Cr is electronically unsaturated, and so the amide N i Pr 2 ligands creates variable single - to double - bond character between N and Cr, which directly i This work was done in collaboration with Dr. Brennan Billow. Both Brennan, and a former postdoc in the Odom Group, Dr. Ross Bemowski, had synthesized and partially characterized several [NCr(N i Pr 2 ) 2 (PR 3 )][X] complexes prior to my inclusion on this project. Their efforts provided me with a good starting point as a first year in the group e is included as an author on the publications related to both chapters. 2 The majority of this work has been previously published. The following articles are related to this research: Aldrich, K. E., et al. (2017). "Weakly Coordinating yet Ion Paired: Ani on Effects on an Internal Rearrangement." Organometallics 36 (7): 1227 - 1237. Aldrich, K. E., et al. (2019). "Phosphine interactions with high oxidation state metals." Polyhedron 159 : 284 - 297. 18 effects the energetics of the barrier to rotation of the Cr N i Pr 2 bond. The weaker the X ¯ donor, the more double - bond character there is in the i Pr 2 N Cr interaction and the higher the barrier to rotation, and vice versa. Figure 2 . 1 ( top ) Cr(VI) molecule used in the LDP system. ( bottom high valent Cr(VI). This barrier to rotation has a simple molecular orbital (MO) origin, which has been previously presented with DFT calculations. 1 The transition state for the Cr N i Pr 2 rotation puts the lone pair from the amide in an orientation t o overlap with the - bonding orbital of the Cr nitride bond. This forces the amide lone pair into a * - orbital, evidenced by the pyramidalization of the optimized transition states for these molecules. 1 The energy of the ground state, specifical ly, how 8 9 10 11 12 13 14 15 16 NMe2 OAd NPhMe NiPr2 OBn Carbazole OPhOMe OPhtBu OPh OPhSMe OPhF OPhCl OPhCF3 OSiPh3 OPhtBu F Indole OtBuF6 NO3 OC6F5 Pyrrole SPh PyrC6H5 PyrAr(CF3)2 CN CO2Ph NCO NCS Cl Br OTf I LDP (kcal/mol) 19 localized the lone pair of the amide is in a bonding interaction, determines this energy difference and thus the energy of this rotation barrier with a given X ¯ ligand. 1 Figure 2 . 2 ( top ) The rotation monitored by 1 H spin saturation transfer experiments to determine the LDP value for a ligand, X. The purple and green hydrogens exchange positions through the N Cr bond rotation, allowing for saturation of one of the H signals to carry over to the other signal proportional to the rate of this rotation during the SST experiment ( bottom ). By experimentally determining this rotation rate, the enthalpic barrier to this rotation can be assessed, and = LDP. The energies of these barriers to rotation for the X ¯ ligands examined have been easily studied over the temperature range amenable to NMR spectroscopy, the primary method used to measure LDP values. The methyne hydrogens on the i Pr - groups si t in distinct positions in the ground state orientation of the NCr(N i Pr 2 ) 2 (X) molecules, giving them unique chemical shifts syn and anti to the Cr nitride. During the rotation of the - N i Pr 2 bond, these two Hydrogens exchange positions (and subsequently NMR shifts). With a simple 1 H NMR experiment, called spin saturation transfer (SST), the rate of exchange between these two sets of Hydrogens can be measured directly ( Fig. 2.2 ). The rate of this rotation can be used in conjunction with the Eyring equation to find G of rotation. 1 20 With monoanionic X ¯ ligands, the entropy of rotation, S , was determined by measuring over a wide temperature range with several different X ¯ ligands. These measurements provided the assessment that, with small X ¯ is a small negative number (in e.u.). The for X = iodide was the most accurate over the widest temperature range and was adopted as the offici ( - 9 e.u.). With a value assigned to the entropic , a temperature - independent enthalpic barrier to rotation. This value is what we refer to as LDP. Note, because the temperature dependent factor has been remov ed, we can now directly compare the experimental values measured at any temperature. The NCr(N i Pr 2 ) 2 X measurements are done in solution, with little complication. When X ¯ is a monoanionic ligand, the complexes are neutral species overall. The complexes ar e soluble and stable in a variety of NMR solvents. Typically, the NMR solutions need to be chilled in order to obtain a reliable rate of rotation for the Cr N i Pr 2 bonds, so CDCl 3 became the solvent of choice for these measurements. With a neutral LDP molec ule in a relatively nonpolar solvent, the solution behavior of these complexes has been consistent during LDP experiments. In real catalytic systems, though, ligands have much greater variety than the ligand selection that has been examined with LDP. Som e of the most common sets of ligands are those which are neutral and interact with metals via dative formation of - bonds. These ligands include things like CO, pyridines, phosphines, amines, nitriles, etc. Although some of these ligands, such as CO, are m ore common in low valent systems due to their propensity for - backbonding with electron - rich metal centers, ligands like pyridines and phosphines are ubiquitous. 6, 7 Additionally, other species such as nitriles and isonitriles are relevant to catalytic systems as substrates, including with high valent catalyst species. 8, 9 In these cases, such substrates may interact with the active catalytic species via dative bond formation in various steps of a catalytic cycle. 21 To find an example of this, we need not look far. The Ti(IV) - catalyzed hydroamination an d multicomponent coupling reactions studied by our group (and others), for example, contain high valent homogeneous Ti catalysts, as well as a variety of possible dative - donor ligands in the catalytic mixture. 8, 10 Particularly, the donor ability of isonitriles in our iminoamination reaction (3CC) may affect the rates of our catalysts and bear relevance to future ligand design (vide supra, Chapters 5 and 6). 8 We, therefore, wanted to turn our attention towar d neutral ligands in the implementation of the LDP system. While several comprehensive ligand studies have been conducted that quantify ligand donor abilities to low valent metals (Tolman system 11, 12 and spectrochemical series 13, 14 ), a demonstrated above, with X - ligands. The bonding differences between M L bonds (whe re L is a neutral ligand) in high valent versus low valent systems has never been systematically examined in a quantifiable manner. To approach this task, we started with a class of neutral ligands which are well understood sterically and electronically, in both the bound and unbound configuration with low valent metals: phosphines (PR 3 ). Chadwick Tolman pioneered the parameterization of these ligands in terms of 11, 12 Since then, his system has been cited thousands of t imes and can be used to rationalize everything from reaction rates in catalyses where the phosphine is an ancillary ligand to simple substitution reactions. 6 Of course, with improvements in instrumentation and computational methods, adjustments to his original parameterizations have been made. A trend in the field has moved toward differentiation of the electr onic component of phosphine donor abilities into separate - and - terms. 15 - 18 This trend has produced systems such as the Quantitative Analysis of Ligand Effects, 22 established by Giering, Prock, and Poe. 19 - 22 It has also sparked similar parameterization efforts for 23 While today, the Tolman values for new ligands are primarily determined using computational analyses in simple Ir or Rh molecular systems, the usefulness of these lig and classification schemes is difficult to overstate. 24 Give n the vast knowledge of low - valent metal phosphine interactions and more specifically, the detailed efforts that have gone into substituent effects from changing the PR 3 R - groups phosphines seemed like an ideal place to start when crossing over to high val ent systems. We not only wanted to assess the quantitative donor ability of a variety of phosphine ligands to a high valent metal, but also address fundamental differences in high valent versus low valent bonding in these systems. For example, in a high v alent metal - phosphine bond, the bond should be more covalent due to the high electronegativity of d 0 early transition metals (i.e. s of Cr(VI) = 3.37 ; Ti(IV) = 1.53 14, 25 ). Does this affect the - donor ability of phosphines relative to their donation in low valent metal i nteractions? Do changes to R groups produce the same trends in high valent interactions as have been observed in low valent systems? We sought to address these questions directly using the experimental donor ability of phosphines determined by measuring t he LDP values for these ligands. However, before we could put too much stock into the LDP measurements for a variety of PR 3 ligands using our Cr(VI) system, we needed to establish what effect the ionicity of these neutral ligand complexes would have on the solution state behavior of the LDP molecule. A detailed picture of ionicity based interactions in the solution - based system was critical to understanding our LDP measurements. Exchanging X - for L ligands makes the Cr(VI) fragment a monocationic species wi th a counter anion. This change causes potentially massive differences in solution behavior of [NCr(N i Pr 2 ) 2 PR 3 ][X] salts versus neutral NCr(N i Pr 2 ) 2 X complexes. 23 2.2 Synthesis of [NCr(N i Pr 2 ) 2 (PR 3 )][X] Complexes, Solvent Choice, and Initial Assessments of Solvent Effects Synthesis of the [NCr(N i Pr 2 ) 2 (PR 3 )][X] salts was first undertaken by Brennan and Ross (see above), and proved to be straightforward and broadly applicable to a wide range of PR 3 complexes. They utilized AgSbF 6 to precipitate AgI from the NCr(N i Pr 2 ) 2 I starting material ( 1 ) in MeCN. The AgI was removed by filtration, and in a second step, PR 3 can be added to displace MeCN and yield the [NCr(N i Pr 2 ) 2 PR 3 ][SbF 6 ] complex ( 3 ). The same general synthetic scheme, shown in Fig. 2.3 , was utilized to produce almost all the derivatives shown in the table below . By altering the precipitation agent, the identity of the counterion, X ¯ , is easily changed. Table 2 . 1 The wide variety of complexes accessible through the general synthesis route. a3j and 3o were prepared using an alternate procedure in which the MX salt was TlPF6, the solvent for the reaction was DCM, and the entire procedure was performed as a 1 - pot reaction to facilitate t otal miscibility of the phosphine with the solvent. bThe synthesis of 3p was conducted under the strict exclusion of MeCN, in DCM with TlBArF24 as a 1 - pot procedure. Upon exposure of 3p to MeCN, 3p converts to 2 and PPh3 ligand. PMe 3 3a P n Bu 3 3b P i Bu 3 3c P i Pr 3 3d PCy 3 3e PPhMe 2 3f PPh 2 Me 3g PPhEt 2 3h PPh 2 Et 3i PPh 2 n Bu a 3j PPh 2 Cy 3k PPhCy 2 3l P(OEt) 3 3m P(O i Pr) 3 3n P(NC 4 H 8 ) 3 a 3o PPh 3 b 3p 24 There were three exceptions to this pathway, which required synthetic alterations. For both 3j and 3o , the starting phosphines, PPh 2 n Bu and P(pyrrolidino) 3 , were immiscible with NCCH 3 . Thus, the yields for these phosphines utilizing the general procedure f or synthesis were very poor. The purity of the compounds in these attempts was also compromised. Therefore, AgSbF 6 was substituted for TlPF 6 and the reaction was conducted as a one - pot - one - step synthesis using DCM as the solvent (Ag coordinates PR 3 ligands competitively in 1 - pot procedures). This produced high yields of both 3j and 3o as the PF 6 - salts rather than SbF 6 - salts. However, as will be shown below, in terms of solution behavior and LDP, PF 6 - and SbF 6 - demonstrate identical behavior as counterions . The other synthetic challenge was the preparation of 3p , where PPh 3 is the ligand. The PPh 3 cannot displace NCCH 3 from the intermediate species, 2 . Even with a large excess of PPh 3 added to 2 , no conversion to 3p is noted (observed by NMR). Therefore, the synthesis was performed with strict exclusion of NCCH 3 , using TlBArF 24 as the precipitation agent and DCM as solvent. If NCCH 3 (1 equiv) is added to an isolated solution of 3p (1 equiv), it rapidly yields 2 and P Ph 3 in quantitative yields. This demonstrates that 3p behaves very differently from the other 3 Figure 2 . 3 General synthesis of the 3a - 3p salts. 25 variants. Anomalous behavior with PPh 3 relative to other phosphines has been previously reported in the literature, where it is the lone derivative in a series of phosphine complexes that does not follow reaction kinetics characteristic of other phosphines. 26 These synthetic complications aside, we were able to synthesize 15 different [NCr(N i Pr 2 ) 2 (PR 3 )][X] complexes with which to study phosphine donor properties with a d 0 metal. These complexes readily crystalize from DCM or chloroform layered with a nonpolar solvent (Et 2 O or pentane) at - 35 ° C, and all the PR 3 derivatives have been structurally characterized using X - ray diffraction. The complexes demonstrate predictable s tructural properties, with normal bond lengths observed for the N i Pr 2 ligands and the nitride nitrogen. Additionally, there was no trend observed in the solid - state in terms of where the counterion sits relative to the Cr(VI) cation. Despite this, two no teworthy elements of the structural data demonstrate the steric crowding in the system with PR 3 ligands. First, the Cr P bond length increases roughly with the size of the phosphine (cone angle). A plot of this is shown in Fig. 2.4 . Second, the N2 Cr1 N3 b ond angle appears to get smaller as the PR 3 bound to Cr increases in size (Table 2.2) . The results of the steric impacts on LDP measurement will be discussed in more detail in the next chapter. 26 Figure 2 . 4 Plot of Tolman Cone Angle and Cr1 - P1 bond distances in X - ray crystal structures. Table 2 . 2 Cr1 - P1 bond distances and N2 - Cr1 - N3 bond angles. a For those complexes were multiple bond distances/angles are given, there are two unique molecules in the asymmetric unit cell for these complexes. 3a a 3i P - Cr 2.363/2.368 P - Cr 2.401 N2 - Cr1 - N3 124.22 N2 - Cr1 - N3 124.05 3b a 3j P - Cr 2.399/2.392 P - Cr 2.413 N2 - Cr1 - N3 122.96/123.15 N2 - Cr1 - N3 123.1 3c 3k P - Cr 2.435 P - Cr 2.441 N2 - Cr1 - N3 121.41 N2 - Cr1 - N3 120.94 3d a 3l P - Cr 2.452/2.472 P - Cr 2.401 N2 - Cr1 - N3 120.25/119.75 N2 - Cr1 - N3 121.49 3e 3m P - Cr 2.461 P - Cr 2.342 N2 - Cr1 - N3 119.24 N2 - Cr1 - N3 125.83 3f 3n a P - Cr 2.461 P - Cr 2.381/2.383 N2 - Cr1 - N3 119.24 N2 - Cr1 - N3 125.08/122.77 3g 3o a P - Cr 2.399 P - Cr 2.451/2.442 N2 - Cr1 - N3 124.02 N2 - Cr1 - N3 118.62/119.58 3h 3p P - Cr 2.395 P - Cr 2.44 N2 - Cr1 - N3 123.97 N2 - Cr1 - N3 121.73 Perhaps predictably, these [NCr(N i Pr 2 ) 2 (PR 3 )] + salts are not very stable. Stored as solids at reduced temperatures, these complexes are stable for several weeks; they are generally yellow or orange powders, with the hue and intensity of the colors of the salts changing with different X ¯ 100 110 120 130 140 150 160 170 180 2.32 2.34 2.36 2.38 2.4 2.42 2.44 2.46 2.48 Tolman Cone Angle ( ) Cr1 - P1 Bond Distance (Å) Cone Angle Vs. Bond Length for 3a - 3p 27 counter anions . When stored at room temperature or in solution, however, the complexes show signs of decomposition within 1 week. The counterion appears to have some bearing on stability such that SbF 6 - < PF 6 - < B(Ar) 4 - ; this trend was observed both in the solid state a s well as in solution (Ar = C 6 F 5 , 3,5 - bis(CF 3 ) - C 6 H 3 , or Ph) by examining old samples prepared with the different counter anions. The relative instability of these complexes has also made accurate elemental analysis unreliable. Samples prepared using standa rd air - sensitive EA techniques available at MSU have not resulted in reliable characterization using this method. To our knowledge, these are the first synthetic examples of isolated Cr(VI) cations and their extreme Lewis acidity (see Chapter 3) results in highly reactive complexes susceptible to decomposition pathways. One other result of the syntheses of these complexes has been the isolation of 2 (the intermediate acetonitrile adduct produced from iodide extraction) and its full characterization. Nitri les by themselves are an interesting class of neutral ligands, and the successful isolation of 2 demonstrates that the synthesis used for the 3a - p complexes here could easily be applied to the preparation of a series of L = NCR ligands. The crystal structu re of 2 ( Fig. 2.5 ) shows the nitrile coordinated to Cr datively through the N, with the Cr1 N4 C13 angle being approximately linear (176.04 ° ). Additionally, the N4 C13 bond distance is consistent with a covalent N C bond length (1.157 Å in free acetonitrile vs 1.137 Å in the bound form). Collectively, these two structural parameters suggest little to no - interaction between Cr and NCCH 3 , which characterizes NCCH 3 as a purely - donor in this interaction. 28 With a reliable synthetic route established to produce the [NCr(N i Pr 2 ) 2 (PR 3 )][X] complexes, with [X] = SbF 6 ¯ , we began studying them in solution. Initially, we started with conditions similar to those used for the NCr(N i Pr 2 ) 2 X complexes. NMR samples were prepared in CDCl 3 with concentrations of 0.025 M. Looking at a room temperature 1 H NMR spectrum for a phosphine salt ( 3f ) compared to a moderately strong X ¯ liga nd (OPh), we can see a dramatic difference. In 3f , the - N i Pr 2 methyne ( - CH - ) and methyl doublet peaks (CH 3 ) are well resolved and static, whereas these peaks in the NCr(N i Pr 2 ) 2 (OPh) spectrum are broad and fluctional due to their fast rotation ( Fig. 2.6). Q ualititatively this demonstrates that the phenol, in which rapid rotation occurs at room temperature, is a much better donor than the phosphine, with static N i Pr 2 groups at room temperature. Therefore, in order to get the Cr N bond rotation in the phosphin e complexes into a time regime that agrees with the NMR time scale, we needed to heat the samples. For a handful of the 3a - p salts, this was not an issue, but for the majority, we were heating the samples to within a few degrees of the boiling point of the NMR solvent. In a few cases, even at 60 ° C, bond rotation was still too slow to get measurable rotation rates. Figure 2 . 5 Crystal structure of 2, showing the NCCH3 bound end - shown at 50% probability. H atoms are omitted for clarity. (Cr1 - N1 1.533 Å; Cr1 - N2 1.817 Å; Cr1 - N3 1.816 Å; N2 - Cr1 - N3 121.85 °; Cr1 - N4 - 13C 176.40 Å; N4 - C13 - C14 178.96; Cr1 - N4 2.004 Å). 29 Figure 2 . 6 ( top ) Room temperature 1 H NMR spectrum of NCr(N i Pr 2 ) 2 (OPh) showing broad resonances for the i Pr groups due to rapid Cr N i Pr 2 bond rotation. ( bottom ) Room temperature 1 H NMR spectrum of [NCr(N i Pr 2 ) 2 PPhMe 2 ][SbF 6 ] showing sharp w ell - resolved resonances for the i Pr groups. The high barrier to rotation in this molecule prevents exchange of the i Pr Hydrogens leading to a static spectrum. These challenges prompted us to switch NMR solvent to one that would support higher temperature s for NMR experiments. Due to the polar nature of the salts, however, a nonpolar alternative such as C 6 D 6 was not a good option because of poor solubility. Since the syntheses of 30 the 3a - p salts takes place in NCCH 3 , we reasoned that solubility would be mor e than adequate in CD 3 CN; additionally, this change gave us an additional 24 ° C temperature threshold. With a new solvent for measurement, we began remeasuring the LDP values for several 3 salts. However, the solvent change caused large differences in the LDP value between the two solvents. A selection of LDP values measured in both CDCl 3 and CD 3 CN is shown in Table 2. 3 . In the most severe cases, the LDP is 0.62 kcal/mol. Given the error margin typically quoted for LDP measurement (0.1 kcal/mol), these differences were far too large to amount to experimental errors. In previous experiments with NCr(N i Pr 2 ) 2 X complexes, no solvent dependence in the measured LDP values had been established. However, with the iconicity of the neutral ligand complexes, we suspected that ionicity - based solvent effects were impacting the measurements. Table 2 . 3 LDP values determined in two different NMR solvents with a variety of different PR 3 ligands. All values r eported here employ SbF 6 - as the counter ion. Phosphine Complex (SbF 6 - ) LDP CDCl 3 (kcal/mol) LDP CD 3 CN (kcal/mol) LDP P(OEt) 3 16.08 15.73 0.35 P(Me) 3 17.23 16.64 0.59 PPhMe 2 16.99 16.53 0.46 P(O i Pr) 3 16.29 15.91 0.38 PPh 2 Me 16.68 16.16 0.52 PPh 2 Cy 17.04 16.43 0.61 P( i Pr) 3 17.79* 17.17 0.62 *The temperature required to obtain an accurate rate of N - Cr bond rotation exceeded the boiling point of the solvent. This value is an estimate based on the rotation rate at the highest achievable temperature in CDCl 3 . Utilizing 14 N NMR to characterize the complexes provided a clue as to why the LDP values measured in nonpolar CDCl 3 were higher than those measured in CD 3 CN. A typical NCr(N i Pr 2 ) 2 X complex shows two resonances: the nitride appears be tween 900 - 1100 ppm and the amide appears between 200 - 400 ppm. The 14 N NMR spectrum of several [NCr(N i Pr 2 ) 2 PR 3 ][SbF 6 ] complexes 31 exhibit only one resonance in CDCl 3 , which appears at about 450 ppm. This resonance has been assigned as the amide nitrogen. Deviation from the normal behavior for these complexes raised significant difference between the neutral NCr(N i Pr 2 ) 2 X complexes and the [NCr(N i Pr 2 ) 2 PR 3 ][X] complexes, so the lack of a nitride resonance seemed to be related to the ionicity. 14 N is a quadrupolar nucleus with a fast relaxation rate, so the most likely reason for the lack of a nitride resonance was enhance relaxation which would broaden the signal into the baseline of the spectrum. The most abundant isotope of antimony is 121 Sb, which is also quadrupolar. Proximity of two quadrupolar nuclei can lead to coupling and subsequent enhancement of quadrupolar relaxation. It seemed likely that t he nitride signal is diminished in CDCl 3 due ion pairing which positions the SbF 6 ¯ counterion close the nitride nitrogen in solution. 27, 28 Unfortunately, due to the large background resonance o bserved when CD 3 CN is used as the NMR solvent, we could not compare between the two solvents directly. However, when the counterion is exchanged for one lacking a quadrupolar nucleus, the 14 N resonance for a nitride signal is observed in CDCl 3 . This was ob served with both the PF 6 ¯ and B(Ar) 4 ¯ anions, and generally supports the conclusion that in the SbF 6 ¯ salts, quadrupolar relaxation diminishes the nitride signal. This is illustrated by the spectra in Fig. 2.7. 32 Figure 2 . 7 14 N NMR spectra of 3f in CDCl 3 with SbF 6 - ( left ) and PF 6 - ( right ) as counterions. (* = N 2 reference at 309.6 ppm; a = amide shift; b = nitride shift). 2.3 Direct Approaches to Characterize Ion Pairing: Diffusion Ordered SpectroscopY (DOSY NMR) With this piece of preliminary experimental evidence suggesting that, in CDCl 3, tight ion pairing occurs, we needed a more direct approach to compare the solution state behaviors of the salts ( 3 ) in CDCl 3 versus CD 3 CN. Fro m initial observations and our understanding of the LDP measurement, we hypothesized that the difference was primarily that, in moving to the more polar CD 3 CN solvent system, we were disrupting ion pairing. Thus, in the paired regime (CDCl 3 ) the counterion is close to the Cr cation in solution, inhibiting the rotation of the N i Pr 2 groups by sterically blocking them, while in the unpaired regime (CD 3 CN) the ions are separated by fluctional solvation spheres and the N i Pr 2 is unhindered in its rotation. This difference in the ion pairing behavior would then directly account for the differences observed in the LDP values measured in the two solvents with a given PR 3 . 33 The new problem became measuring ion pairing interactions . This has been a historically challenging problem to tackle experimentally, with early efforts focusing on conductivity measurements, potentiometry, or UV - Vis spectroscopy. 29 - 32 In some cases, correlation can even be drawn from solid state structures. 33 Advances in NMR spectroscopy in the last 15 years, however, have improved our ability to study complexes in solution. Specifically, methods that allow for through - space correlation (i.e. Nuclear Overhauser Effect) can show close contacts in tight ion pairs. 34 - 37 Another NMR advancement, Diffusion Order SpectroscopY (DOSY) NMR, allows for the determination of the diffusion rate of species in solution. This can be used as an indirect way to determine if a cation and an anion are paired in solution, provided both ions have a readily observed NMR signal. 34, 37 - 40 Two ions of different size should diffuse through solution at different rates. A simple equation relating diffusion rate and molecular size, the Stokes - Einstein equation, is shown in Eq. 2.1 and is a good approximation for regularly shaped molecules (i.e. spherical). From this equation, we can observe the size dependence. ( Eq. 2. 1 ) Here the particle (molecular species) is assumed to be spherical, with a radius r, traveling through a medium with a viscosity of . More elaborate models of diffusion that describe the motion of various shapes have been developed, but this equation can st ill be applied to simple molecules. 41, 42 More importantly, it illustrates the inverse relationship between particle size (r) and diffusion rate (D). A practical demonstration of this effect is shown in Fig. 2.8, below, where several molecular species are in solution, and their diffusion rates have been determined by DOSY NMR spectroscopy. The larger the molecule, the more slowly it diffuses through solution, such that we know Et 2 O ~ CHD 2 CN < Ferrocene < ((H 3 C) 3 Si) 4 Si < 3f [PF 6 ]. 34 Figure 2 . 8 A 1 H DOSY NMR spectrum in CD 3 CN that contains several molecular species of different sizes. The diffusion coefficients are inversely proportional to the molecular sizes, with larger specie s diffusing more slowly than smaller ones. If two ions are separated, they should diffuse through solution at a rate inversely proportional to their independent hydrodynamic radii. However, if they are paired, they should exhibit the same hydrodyanamic ra dius and diffuse at the same rate. Using a combination of these techniques (DOSY and NOE experiments) we set about determining whether the Cr(VI) cations and their counter anions were paired in the two NMR solvents examined. Anions such as the B(Ar) 4 ¯ wer e designed to replace anions such as SbF 6 ¯ , PF 6 ¯ and OTf ¯ , - 43 The B(Ar) 4 ¯ are - coord achieved through extreme delocalization of charge, imparted to electronegative functional groups on the periphery of the anion. For example, in BArF 24 ¯ , 8 CF 3 groups are spread out across the anion, boun d to aromatic rings to delocalize the formal - 1 charge of the molecular unit. In the LDP 3f [PF] 6 35 system, we also hoped that perhaps these more diffuse anions would exhibit less ion pairing and solve the ion pairing issues in solutions simply. We had already undert aken synthetic efforts with three different anions: SbF 6 ¯ , PF 6 ¯ , and BArF 24 ¯ . With these different counter anions, we had observed different 14 N NMR behaviors, and wondered if there were further differences in the solution state behaviors of these salts ( 3 ) dependent on the identity of the counterions. Thus we 3 and CD 3 CN. The anions selected are shown in Fig. 2.9 below. Figure 2 . 9 [NCr(NiPr2)2PR3]+ cations. While diversifying the types of anions cons idered, we expanded the scope of the study beyond application to the LDP system. 44 - 47 In doing so, however, we created a large number of potential complexes to make and study (>50). Therefore, we carefully selected two PR 3 species to N i PR 2 ) 2 PR 3 ][X] complexes. Due to their simplicity in the 1 H NMR spectra, we chose the ligands PPhMe 2 ( 3f ) and PMe 3 ( 3a ) for our model complexes. For accurate DOSY or ROESY measurements, each species must present distinguishable peaks without overlap from o ther signals, and these ligands had provided resolved, well - separated spectra in which the 1 H peaks are easily distinguished and assigned. 36 Considering the available NOE experiments that allow for detection of close contact between species in solution, we tried examining correlation between 19 F and 1 H signals from the anion and cation respectively using Heteronuclear Overhauser Effect Spectros copY, (HOESY NMR). However, with many of the fluorinated anions, coupling induced odd and rapid relaxation properties of the 19 F nuclei, so we struggled to find adequate parameters for these NMR experiments. Additionally, due to the intermediate size of th ese species (1000 g/mol), their tumbling behavior in solution precludes good signal in a typical NOE experiment orientation. 48 To solve this problem, we looked at the [NCr(N i Pr 2 ) 2 PMe 3 ][BArF 24 ] salt, using 1 H Rotating - frame Overhauser Effect SpectroscopY (ROESY NMR); the difference in the orientation and pulse sequence timing, going to the rotating frame experiment ( 1 H) versus a traditional NOESY or HOESY experiment, solves the signal intensity issues caused by molecular size. The BArF 24 ¯ anion can be observed by its two sharp and unique 1 H NMR signals for the aromatic protons rather than by observing the 19 F signal. With this salt, 3a [BArF 24 ], in CDCl 3 , we were able to observe pronounced cross peaks between the aliphatic signals of the Cr(VI) cation and the aromatic signals of the BArF 24 ¯ anion in solution. Th ese spectroscopic results are presented in Fig. 2.10. This provides direct evidence of proximity (<5 Å) for these two ions in solution and indicates a tight ion pair. By switching the solvent to CD 3 CN, the cross peaks disappear. Of course, while negative evidence is often not informative, the results of thes e two experiments considered side - by - side certainly supported our hypothesis. More interesting than the identification of ion pairing, this experiment also provides an idea of how the ions are paired. Cross peaks are observed between the BArF 24 ¯ anion and the aliphatic Hydrogens of the i Pr - groups as well as the PMe 3 ligand. This indicates that the ion 37 pairing, while tight, is non - site - specific. The pairing exists across the entire surface of both the cation and the anion rather than via one specific site o n each anion with strong electrostatic attraction. Additionally, we were able to observe this tight ion pairing behavior in CDCl 3 with the BArF 24 ¯ , which is one of the best and most diffuse anions in terms of minimizing its charge density. This direct obse rvation made it seem highly likely that the ion pairing with diffuse anions is just as tight and substantial as that observed with the less diffuse anions. It also strongly indicates that s and ion pairing. 38 Figure 2 . 10 ROESY NMR spectra for 3a [NCr(N i Pr 2 ) 2 PMe 3 ][BArF 24 - ] in CDCl 3 ( top ) and CD 3 CN ( bottom ). Correlation between the cation and anion are noted in CDCl 3 , but these cross peaks are not observed in CD 3 CN. 39 To verify this experimentally, we progressed to a DOSY study and examined each anion with one or both model cations, 3a and 3f . The results of the DOSY studies were very straightforward and informative. These results are highlighted in Table 2. 4 below. Table 2 . 4 Results of the DOSY NMR experiments utilizing various anions and the chromium(VI) cations 3a and 3f . Complex Anion Solvent Ion D std D ion Ratio (D ion :D Std ) Ratio (D +/s :D - /s ) PPhMe 2 PF 6 ¯ CDCl 3 + 17.53 10.05 0.573 0.98 ±0.05 - 14.92 8.73 0.584 CD 3 CN + 25.81 17.90 0.691 0.66 ±0.04 - 28.2 29.63 1.051 BArF 24 ¯ CDCl 3 + 14.79 6.35 0.429 1.04 ±0.07 - 15.69 6.47 0.412 CD 3 CN + 21.69 14.21 0.658 1.14±0.03 - 20.27 11.68 0.576 BArF 20 ¯ CDCl 3 + 16.38 7.34 0.449 1.08 ±0.04 - 12.25 5.10 0.416 CD 3 CN + 20.26 13.36 0.659 1.15 ±0.06 - 19.75 11.30 0.572 Al(O t BuF 9 ) 4 ¯ CDCl 3 + 15.17 6.08 0.397 1.01 ±0.04 - 13.65 5.35 0.392 CD 3 CN + 24.02 15.7 0.654 0.93 ±0.04 - 20.46 14.48 0.708 BPh 4 ¯ CDCl 3 + 14.68 6.07 0.413 1.01 ±0.02 - 14.68 6.03 0.410 CD 3 CN + 22.53 14.89 0.661 0.98 ±0.02 - 22.53 15.12 0.671 PMe 3 BArF 24 ¯ CDCl 3 + 15.72 6.47 0.41 1.02 ±0.04 - 15.72 6.34 0.40 CD 3 CN + 21.29 11.68 0.56 0.76 ±0.02 - 21.29 15.03 0.73 BPh 4 ¯ CDCl 3 + 14.75 6.27 0.43 0.99 ±0.02 - 14.75 6.32 0.43 CD 3 CN + 21.76 15.36 0.71 1.04 ±0.03 - 21.76 14.75 0.68 40 From the results in the table above (which are standardized against 1,3,5 - tris(trifluoromethyl)benzene as internal standard when 19 F DOSY was used for the anion measurement), a common trend is observed for the CDCl 3 standardized diffusion coefficient ratio s (cation : anion): they are all very close to 1. This indicates that the anionic and the cationic species are diffusing with identical rates in solution regardless of differences in ionic size. These results strongly indicate tight ion pairing with all an ions examined in the nonpolar CDCl 3 system. Here A common trend is also observed from these results upon shifting to the more polar CD 3 CN solvent system and repeating the DOSY measurements. The ratio of the standardized diffusion coefficients (D +/s :D - /s ) diverge from 1 with a magnitude and direction proportional to the size difference of the ions. For example, PF 6 ¯ is considerably smaller than the chromium(VI) cation in 3f , so in acetonitrile the diffusion coefficient for the cation is smaller (i.e. diffuses slower) than that of the anion. The ratio of the diffusion coefficients therefore diverges from 1 and gets sm aller, reflecting that the cation now diffuses considerably slower than the anion in solution. The only pair where the differences are less pronounced is the 3f cation with BPh 4 - as the anion. However, due to the similarity of their molecular volumes ( ~ 400 vs. 300 Å 3 ), this seems to be due to the detection limits of the technique. Thus, the results are not conclusive. The results of the ROESY and DOSY NMR experiments are reinforced by the measured LDP values for each salt with different X ¯ in both CDCl 3 a nd CD 3 CN. These values are shown in Table 2.5 below. With the different X ¯ examined, the LDP values in CDCl 3 are generally higher compared to the LDP values in CD 3 CN. What is most interesting about these differences between solvents is that the magnitude s eems to be completely dependent on the identity of X ¯ . For the compact anions (i.e. SbF 6 ¯ and PF 6 ¯ ) a very large difference, > 0.5 kcal/mol, is observed on 41 switching solvent. With the B(Ar) 4 ¯ or aluminate anions, however, these differences are much smaller in fact, within experimental error. Table 2 . 5 LDP values measured for 3a and 3f with a variety of anions in CDCl3 and CD3CN. This illustrates the ion effect on the LDP measurement. Complex Anion LDP (CDCl 3 ) (kcal/mol) LDP (CD 3 CN) (kcal/mol) 3f SbF 6 ¯ 16.99 16.53 PF 6 ¯ 19.96 16.53 BArF 24 ¯ 16.60 16.58 BArF 20 ¯ 16.64 16.62 BPh 4 ¯ 16.57 16.47 Al(OC(CF 3 ) 3 ) 4 ¯ 16.66 16.62 3a SbF 6 ¯ 17.24 16.64 BPh 4 ¯ 16.71 16.66 BArF 24 ¯ 16.86 16.65 From the ROESY results, BArF 24 ¯ demonstrated non - site - specific ion pairing. From the similar impact on the LDP value observed with other borate and aluminate counterions, this behavior appears to be the norm for large diffuse anions in the system. The SbF 6 ¯ counter ion, on the other hand, diminishes the 14 N signal of the nitride ligand preferentially. Based on these results, we suspected that the large difference in the measured LDP variance between solvents with the small X ¯ anions was primarily caused by site - speci fic ion pairing. The SbF 6 ¯ and PF 6 ¯ anions have a specific part of the Cr cation where the electrostatic interaction is strongest, so the ions pair together in one specific orientation. If that orientation puts the X ¯ next to the - N i Pr 2 ligands, as would b e the case if the X ¯ sits above the Cr nitride bond vector, the ion pairing would influence the measured LDP value more than ion pairing where the interaction is spread across the entire cation. Afterall, the LDP measurement is a bulk average of all the molecules in solution. Experimentally, this demonstrates that the difference in the mechanism of ion pairing can greatly impact the effects of ion pairing on processes such as internal rearrangements in the first 42 coordination sphere of an ionic complex. T impacts ion pairing, where more charge - diffuse ions pair non - specifically and ions with more localized charge pair specifically. Overall these experiments provide significant evidence that the Cr(IV) salts ( 3 ) are tightly ion paired to their counter anion in CDCl 3 , while in CD 3 CN, the ions are separated and exhibit diffusion as free species in solution. These results agree well with the initial hypothesis, suggesting that steric interference from the paired counter anion is artificially raising the LDP values observed for the phosphine complexes in CDCl 3 . 2.4 Other Solution State Investigations of Ion Proximity in Solution DOSY NMR to determine the diffusion rate of the ionic species in solution provided us with a clear qualitative picture of the fate of the Cr(VI) salts in solution with regards to their ion pairing. However, in order to get a more quantifiable sense of how unpaired the salts are in CD 3 CN, we turned to an alternate DOSY technique in which the molecular weight of a species in solution can be calibrated with a series of internal standards. 49 - 52 This technique can be useful when determining if a molecular species is monomeric or dimeric in solution, fo r example. Similarly, with these ionic species in solution, there should be a noticeable difference in the calibrated molecular weight dependent on the pairing of the ions. To utilize this technique, three internal standards which are inert, have unique NMR signals, and a wide range of molecular weights across the series of standards, were chosen. The diffusion coefficient of each molecule can then be plotted against the log of the molecular weight (log(MW)) to give a linear relationship, shown in Eq. 2.2 The experimentally determined diffusion coefficient for the unknown species of interest can then be calculated from the linear regression to provide the molecular weight in solution. 43 ( Eq. 2. 2 ) Here, D is the measured diffusion coe fficient, MW = molecular weight, and m and b are the coefficients derived from linear regression. Of course, there is error associated with this experimental method, typically on the order of 5 - 20% relative error from the algorithm used to calculated diffu sion from NMR signal attenuation. This raises an important point about the in situ when the possible species are highly weight discrepant, such as a monomer vs. a dimer, where the molecular weight doubles. For the purposes of determining the ion pairing behaviors within our system, we anticipated some challenges. The internal standards used are neutral species. Introducing additional ionic species would like ly introduce non - innocence between the Cr salts in solution and the standards. However, because our species is charged and the standards are neutral, we suspected that there might be some complications with the absolute accuracy of the measurement. Specifi cally, the solvent interactions affecting the 3a - p species in solution should be markedly different than those experienced by the neutral, high C H composition standards examined due to the charge. We suspected that the association of a solvation sphere ar ound the ionic species in solution would make these species manifest artificially high masses in solution. The molecular weight determination of 3f with PF 6 ¯ as the counter anion was performed in CDCl 3 , C 6 D 5 Cl, and CD 3 CN. We observed that the molecular weight determined for the chromium(VI) cation was highest in CDCl 3 and decreased as the polarity of the solvent increased. The weights are higher than the molecular weight of 3f [PF 6 ] (549.5 g/mol) or [ 3f] + (404 g/mol) in both solvents, suggesting there is a systematic difference in the behavior of the ionic compound 44 versus the neutral standards. However, a clear trend is still observed going from a nonpolar to a polar solvent. Table 2 . 6 Molecular wei ght calibration results for 3f [PF 6 ] in several solvents and concentrations. Solvent Concentration (M) Dielectric Constant Molecular Weight (g/mol) CDCl 3 0.025 4.91 693 ± 114 C 6 D 5 Cl 0.025 5.62 628 ± 21 CD 3 CN 0.025 37.5 542 ±7 6 0.01 578 ± 93 0.10 525 ± 68 Looking at the series, in CDCl 3 the Cr(VI) cation fragment demonstrates a molecular weight of 693 g/mol, decreasing to 628 g/mol in C 6 D 5 Cl, and 542 g/mol in CD 3 CN. This trend aligns with a reduction in ion pairing as the polarity of the solvent i s increased. A molecular weight reduction of 151g/mol from 693 to 542 g/mol corresponds closely with the loss of PF 6 ¯ (144 g/mol) on switching to the polar system. This trend supports the experimental observations made directly by analyzing the diffusion c oefficients in Table 2. 4 . There is a second possibility for why the ionic species have high molecular weights observed by this method. Aggregation, or the conglomeration of several pairs of ions, is a dynamic process which could also increase the observe d molecular weight. 40, 53 This process typically demonstrates concentration dependence, such that when the concentration of 3f [PF 6 ] is increased in solution, more aggregation occurs. From Table 2. 4 , we can see that changing the concentration does not change the measured molecular weight, thus the high overall weights are not likely caused by aggregation. Solvent interaction differences between the ionic species of interest and the internal standards is the most likely culprit for these high molecular weights. 2.5 Computational Investigation of Ion Pairing with PF 6 ¯ From experimental investigations, we had discovered several important pieces of informat ion about the behaviors of these ionic species in solution, and the effect that this renders 45 on LDP measurements. As discussed above, large, diffuse anions such as BArF 24 ¯ appear to pair to the Cr(VI) cations in a non - site - specific manner. However, the sma ller, less diffuse counter anions have demonstrated different behavior according to LDP measurements and the effects observed with 14 N NMR. Based on the 14 N NMR results with SbF 6 ¯ , we suspected that the small, compact counter anions pair to the Cr(VI) cat ions in a site - specific manner. The point of contact between the two ions likely involves the nitride nitrogen on the Cr(VI) cation. From such an orientation, the N i Pr 2 seemed likely. As no crystallographic interaction had been observed, and the properties of the anion and the inherent molecular weight of the compounds had le d to difficulties in conducting HOESY NMR experiments, we turned to computational analysis to probe the nature of the ionic interaction with EF 6 ¯ counter anions (E = Sb of P). The structure of the Cr(VI) cation, 3f , was optimized (starting from the crysta llographic coordinates) using DFT, with the B3PW91 (and also with b97xd, which gave identical results) functional and 6 - 31G+(d,p) basis set on all atoms. With this optimized cation structure, the PF 6 ¯ anion was then included in proximity to the cation; st arting from several different orientations (i.e. above the Cr nitride bond vector, below the Cr nitride bond vector, next to the PR 3 group, and next to the N i Pr 2 groups), reoptimization was attempted. 54 These different orientations were each submitted for further optimizations. The only orientation from which the calculation would converge to an energy minimum was the orientation in w hich the PF 6 ¯ anion was placed above the Cr nitride bond vector. In this orientation, the structure quickly converged on an optimized geometry in which hydrogen bonding interactions exist between the i Pr - groups and the fluorines of the PF 6 ¯ . 46 This electros tatic interaction is most readily observed by considering the Mayer Bond Order for the P F bonds (table 2.7) . The 3 - centered - 4 - electron bond that is oriented away from the Cr(VI) cation, F6 P1 F5, serves as a good reference point for comparison. In this bo nd, the bond order is calculated as 0.88. Relatively, the P F1 and P F2 bond orders are reduced (0.74 and 0.80 Figure 2 . 11 Representation of the optimized 3a [PF 6 and i Pr groups (pink ). Table 2 . 7 Bond orders calculated for the PF 6 ¯ anion, showing H - bonding interaction effects on P F bond orders. P F 1 bond Calculated Mayer Bond Order F 1 0.74 F 2 0.80 F 3 0.94 F 4 0.94 F 5 0.88 F 6 0.88 The fluorines trans (F3 and F4) to these electron - density donor fluorines (F1 and F2) compensate for the shift of electron density in the P F1/F2 bonds, by increasing their bond orders, such that both the P F3 and P F4 bond orders are 0.94 each. These substantial effects on the bonding in the PF 6 ¯ anion show the significance of these interactions, providing a logical reason 47 for the severe inhibition that ion pairing places on the rotation of the i Pr - groups with EF 6 ¯ counterions. The computational results demonstrate clearly what the experimental r esults suggest, that with a compact PF 6 ¯ , the charge distribution on the cation and the anion lead to a specific orientation between the cation and the anion. This site - specific ion pairing leads to differences in the solution state of the ionic complexes compared to those where non - site - specific pairing occurs. 2.6 Entrop ic Complications in Ionic LDP Systems Based on our efforts thus far to thoroughly understand the nature of ionic interactions in solution with the 3a - p salts in the LDP system, it was deter mined that with a compact anion, in CD 3 CN, ion pairing interactions were minimized. From this assessment, we thought that the system would lead to most purely electronic LDP values with the least amount of interference from ionic interactions in solution d uring the rotation barrier measurements. The LDP values of the series of Cr(VI) cations, 3a - o , were measured with SbF 6 ¯ or PF 6 ¯ as the counter anion in CD 3 CN (excluding 3p due to instability with CD 3 CN). The experimental values are listed in Table 2. 8 . Table 2 . 8 LDP values measured in CD 3 CN with SbF 6 ¯ /PF 6 ¯ counter anions. Compound Number Phosphine Complex (kcal/mol ) 3a PMe 3 16.64 3b P n Bu 3 16.77 3c P i Bu 3 17.13 3d P i Pr 3 17.17 3e PCy 3 17.27 3f PPhMe 2 16.53 3g PPh 2 Me 16.16 3h PPhEt 2 16.65 3i PPh 2 Et 16.15 3j PPh 2 n Bu 16.31 3k PPh 2 Cy 16.43 3l PPhCy 2 16.37 3m P(OEt) 3 15.73 3n P(O i Pr) 3 15.91 3o P(NC 4 H 8 ) 3 16.21 48 Before we could fully analyze these values, there was one other aspect of these complexes in solution that warranted further investigation: entropy. With the monoanionic X ¯ ligands typically examined in this system, it was previously determined that the entropy associated with N i Pr 2 rotation is a small negative number (e.u.) that is relatively constant regardless of the identity of X ¯ . In the 3a - p salts, however, we had obse rved some concerning results that suggested either 1) the - 9 e.u. determined with neutral Cr(VI) complexes was substantially different from S in the ionic systems, or 2) the entropy for each 3a - p complex was somehow dependent on the identity of PR 3 , and was an oversimplification in these systems. The first piece of evidence that caused concern about the accuracy of the entropy assumption within this system was that, for a few phosphine complexes, different LDP values were obtained at different temperatures. Furthermore, these differences were substantial. While a claimed error of ± 0.1 kcal/mol is typically assigned to LDP numbers, the measurement of these values is extremely precise; this error is largely attributed t o our estimate of temperature calibration affecting the accuracy of the measurement. Two different researchers measuring the LDP value of two independently prepared batches of a given complex, on different instruments, making the measurement at different t emperatures and on different days, typically yields numbers within 0.04 kcal/mol of one another. Additionally, these measurements are made in triplicate, with standard deviations typically below 0.02 kcal/mol. Consequently, differences on the order of thos e we had source of error that we could attribute these differences to was entropy, a temperature dependent of rotation. Of course, measuring entropy experimentally is time - consuming and associated with strict limitations. In order to determine the entropy of the N i Pr 2 group rotation about the N Cr bond, the 49 LDP value must be measured at several different temperatures. Generally, the wider the r ange of . 1, 55, 56 (Note, ranges that come closer to absolute zero also make these calculations more accurate, but this property is not something that can be controlled by the experimenter) With the phosphine complexes 3a - p , ther e is a very small range of temperatures over which the rotation is measurable on the NMR time scale, yet the complexes are stable enough to obtain accurate measurements without decomposition. Heating these complexes for extended time over 65 ° C has been noted to release HN i Pr 2 , and even produce the bridging nitride - mixed valent Cr dimer that has been previously structurally characterized. 57 Consequently, for many of the phosphine complexes examined, this reliable temperature wind is a mere 10 - 20 ° C. By carefully selecting a few phosphine complexes in which the rotation barrier can be measured at lower temperatures, a wide enough variety of complexes was examined experimentally to elucidate the entr for 3f , 3j , and 3m , with SbF 6 ¯ as the counterion and CD 3 CN solvent, are shown in Table 2. 9 below. Table 2 . 9 Complexes 3f , 3j , and 3m values for the ionic complexes in CD 3 CN with the SbF 6 ¯ counter anion. Complex Entropy (e.u.) Temperature Range ( ° C) Real H (kcal/mol) 3f 4.7(0.5) 21.81 17.90 3j 25(5) 30.94 10.99 3m 38(4) 24.63 6.76 The complexes for which S was measured span a relatively large range of electronic properties. They are all small phosphines, which was necessary to achieve the broader temperature ranges desired. The bulky phosphines require higher temperatures in order to achieve measurable rate s of rotation in solution (see Chpt. 3). Even with only these 3 values, it is apparent that the entropy values are changing drastically depending on the PR 3 group in the complex under the 50 conditions that reduce ion pairing. These differences are far too si gnificant to attribute to error in the measurement technique. We suspected that these dramatic entropy differences, are a direct result of ionicity. In the polar solvent system utilizing CD 3 CN, disruption of the electrostatic interaction between the cati on and anion is achieved by replacing these charge - based interactions with directed dipole moment interactions with the solvent molecules. A specifically oriented arrangement of solvent molecules around each cation and anion in solution solvation spheres w ould likely impose a high degree of order in these solutions. Upon rotation of the N i Pr 2 group, the dipole moment of the Cr(VI) cation changes dramatically, which could in turn lead to a rearrangement of the solvation sphere. All these solvent rearrangemen ts associated with the Cr N i Pr 2 rotation via the disruption of the solvation sphere would make the entropy of this rearrangement in each system unique; and in some cases this much order would likely produce a large value for S . Therefore, it seemed likel y that the LDP values determined with the SbF 6 ¯ /CD 3 value. These experiments answered several questions about the entropic behavior of the ionic Cr(VI) complexes in solution and there was a logical explanati on for the entropic differences observed with these complexes versus NCr(N i Pr 2 ) 2 X complexes. However, this raised several new questions, as well. Is this entropic variability a property of the ligands (PR 3 )? Is it due to differences in the separation of th e ion pair? Does pairing the ions make the variable entropic behavior more values determined with X ¯ ligands on Cr? Potentially, answering some of these questions would assist in determining how to alleviate both ion pairing and entropic complications simultaneously. Thus, several additional entropy measurements were made. 51 3 character, ion pair separation, or both, we examined the experimental 3a - p salts with the BArF 24 - anion. These ionic complexes exhibit ion pairing in CDCl 3 , however, the mechanism of the ion pairing (non site - specific) allows for bond rotation at lower temperatures than SbF 6 ¯ , opening up the experimental t emperature range. In fact, only 1 value in the 3a - p series was not accessible within the temperature range provided by the BArF 24 ¯/ CDCl 3 system ( 3e , PCy 3 ). These entropy values are shown in Table 2. 10 . From these values, it appears that the complexes have small, negative entropies of activation when ion - paired in CDCl 3 . The entropy ranges from - 0.6 to - 8.7 e.u. with an average of - 4.7 e.u. Additionally, regardless of the PR 3 ligand, the values measured for S are also tightly grouped. Table 2 . 10 for 3a - p BArF 24 ¯ salts in CDCl 3 . Phosphine Complex Experimental H (kcal/mol) Experimental S (e.u.) b Phosphine Complex Experimental (kcal/mol) Experimental S (e.u.) b 3a, PMe 3 18.71 - 3.4 (1.0) 3i, PPh 2 Et 17.98 - 4.0 (1.0) 3b, P n Bu 3 18.91 - 2.8 (1.0) 3j, PPh 2 n Bu 18.17 - 3.4 (1.0) 3c, P i Bu 3 17.76 - 5.7 (1.0) 3k, PPh 2 Cy 18.11 - 5.5 (3.2) 3d, P i Pr 3 19.47 - 3.0 (1.5) 3m, P(OEt) 3 16.99 - 6.1 (1.3) 3e, PCy 3 19.46 - a 3n, P(O i Pr) 3 17.12 - 6.0 (1.2) 3f, PPhMe 2 18.79 - 2.1 (1.1) 3o, P(NC 4 H 8 ) 3 19.52 - 0.6 (1.5) 3g, PPh 2 Me 16.90 - 8.2 (1.2) 3p, PPh 3 18.00 - 4.1 (2.1) 3h, PPhEt 2 16.96 - 6.7 (1.4) Average - - 4.7 a The value could not be determined experimentally because a temperature range of <8 ° C was available over which to measure bond rotation. b Values listed in parentheses are error approximations. These results support the hypothesis that the deviations observed with 3f , 3j , and 3m in CD 3 CN are caused by the solvation interactions upon separation of the ion pair. However, whether is caused by an additional ion effect or the p roperties of the PR 3 ligand remains to be seen. As one final piece of support for this hypothesis, we wanted to make sure that 52 the entropic differences observed between the two solvents were not simply the result of doing these measurements in CD 3 CN, but r ather originate from the ionic interactions with the solvent values for NCr(N i Pr 2 ) 2 I ( 1 ) were measured in CDCl 3 and CD 3 CN. The values that were obtained for this neutral Cr(VI) species in the two solvents were the same within error, - 0.5 ± 1.0 and - 1.3 ± 0.5 respectively. These values were determined over >40 ° C temperature determined in triplicate. The indistinguishable entropy behavior between solvents with a neutral molecule support our hy pothesis that the deviations observed in the CD 3 CN measurements of the [NCr(N i Pr 2 ) 2 PR 3 ][X] salts is due to ionic interactions with a polar solvent. 2.7 Conclusions The solution state rotation barrier measurement, used to quantify ligand donor ability to a hig h valent metal in the LDP system, is highly sensitive to changes in solvent interactions. Neutral NCr(N i Pr 2 ) 2 X complexes analyzed in different solvents using 1 H spin saturation transfer demonstrate essentially no change in behavior, solvent to solvent. LDP values ( H ) and the associated entropy of activation for the Cr N i Pr 2 bond rotation ( S ) are consistent across highly different NMR solvents, suggesting little to no interaction between the Cr(VI) complexes and solvent during the bond rotati on of interest. Upon switching to a ligand set where X ¯ is replaced by neutral ligands, L = PR 3 , we were able to synthesize a series of [NCr(N i Pr 2 ) 2 PR 3 ][X] salts, where X ¯ can be one of several weakly coordinating anions. These salts exhibit very differen t H values for Cr N i Pr 2 bond rotation in different solvents. Upon strategic investigation with a variety of experimental techniques, several interesting ionic effects were observed that affect the LDP measurement of PR 3 ligands in these complexes . 53 Highly diffuse, weakly coordinating anions such as B(Ar) 4 ¯ ions or aluminates ion pair with the Cr(VI) cation tightly in nonpolar solvents. This is contrary to the behavior of these anions portrayed by those who advocate the weak interaction or inertness that these anions exhibit with other species in solution. T he ion pairing behavior is disrupted in a sufficiently polar solvent, such as CD 3 CN, to the limits of our experimental detection. A charge - localized, weakly coordinating anion, such as SbF 6 ¯ also forms a tight ion pair with the Cr(VI) cation in a nonpolar solvent. Also, similar to the diffuse anions, the ion pairing with SbF 6 ¯ is disrupted in polar solvents. What is much more interesting about the ion pairing behavior discovered with these two types of weakly coordinating anions is the difference in the me chanism of the pairing itself. With charge - diffuse anions, the ion pairing is non - site - specific, meaning that electrostatic attraction between the cation and the anion is dispersed across the surfaces of both molecules. With SbF 6 ¯ , however, site specific p airing occurs, in which the ion pair exhibits a preferred contact point for pairing due to stronger electrostatic interactions between specific portions of the ions. In this system, the anion sits above the Cr nitride bond vector, which causes the anion to directly impinge upon the N i Pr 2 N bonds. In several ways, the LDP technique itself has proven to be a useful experimental tool for probing the ionic effects and solvent behaviors of these cationic Cr(VI) LDP complexes with s everal common weakly coordinating anions, X ¯ . This type of indirect observation with in situ ion effects is a valuable method for identifying 1) when ion pairing is affecting solution behaviors of organometallic complexes, and 2) the specifics of these effects, such as the mechanism of pairing. 54 2.8 Experimental Instrumentation and facilities All NM R spectra, including LDP and routine characterization data, were recorded utilizing the Max T. Rogers NMR Facility at Michigan State University. These include a UNITYplus 500 spectrometer equipped with a 5 mm switchable broadband probe operating at 36.12 M Hz (14N); a Varian Inova 500 spectrometer equipped with a 5 mm Pulse Field Gradient (PFG) switchable broadband probe operating at 499.84 MHz (1H) and 470.28 MHz(19F); a Varian Inova 600 spectrometer equipped with a 5 mm PFG switchable broadband probe opera ting at 599.89 MHz ( 1 H) and 564.30 MHz ( 19 F); and an Agilent DDR2 500 MHz NMR spectrometer equipped with a 5 mm PFG OneProbe operating at 499.84 MHz ( 1 H), 125.73 MHz ( 13 C), 469.96 MHz ( 19 F), and 202.35 ( 31 P). 1 H NMR chemical shifts are reported relative to residual CHCl 3 in CDCl 3 as 7.26 ppm. 13 C NMR chemical shifts are reported relative to natural abundance 13 CDCl 3 in d - chloroform as 77.16 ppm. Single crystal X - ray diffraction data was collected in the Center for Crystallographic Research at MSU.4.2 operat ing with either Mo - or Cu - K . General considerations All syntheses were carried out under an N 2 atmosphere, using standard Schlenk techniques or in an MBraun glovebox. All reagents were stored in a glovebox after purification. Diethyl ether, acetonitrile , and dichloromethane were purified by passing them over a neutral alumina column under N 2 and stored over 3 Å molecular sieves. Chloroform was distilled from P 2 O 5 under N 2 and stored over molecular sieves. Deuterated chloroform from Cambridge Isotope Labo ratories was distilled from P 2 O 5 under N 2 and stored over molecular sieves. Deuterated acetonitrile from 55 Cambridge Isotope Laboratories was distilled under N 2 from calcium hydride and stored over 3 Å molecular sieves. The complex NCr(N i Pr 2 ) 2 I ( 1 ) was prepared according to the literature procedure. 2 Trimet hyl - , dimethylphenyl - , and diphenylmethylphosphinewere purchased from Aldrich Chemical Co. and used as received. Triethylphosphite and triisopropylphosphite were purchased from Aldrich Chemical Co. and distilled from Na 2 SO 4 under reduced pressure. Triisob utylphosphine, diphenylcyclohexylphosphine, and phenyldicyclohexylphosphine were purchased from Strem Chemical Co. and used as received. Triisopropylphosphine purchased from Strem Chemical Co. was distilled from a 10 wt% solution in hexanes and stored over 3 Å molecular sieves. Tri( n - butyl)phosphine purchased from Strem Chemical Co. was distilled under N 2 and stored over 3 Å molecular sieves. Triphenyl - , phenyldiethyl - , diphenylethyl - , and tricyclohexylphosphine were purchased from Alfa Aesar and used as re ceived. Silver hexafluoroantimonate (AgSbF 6 ) and thallium hexafluorophosphate (TlPF 6 ) were purchased from Sigma - Aldrich Chemical Co. and used as received. The KBArF 20 was supplied as a gift from Boulder Chemical Co. and was used as received. Thallium(I) BA rF 24 ¯ was prepared using the literature procedure. 58 The precipitation agent Ag[Al(OC(CF 3 ) 3 ) 4 ] was prepared according to literature procedures. 59 Tris(pyrrolidino)phosphine was synthesized by adding TMS - pyrrolidine (3.3 equiv) to trichlorophosphine (1 equiv) in cold ( - 78°C) diethyl ether solution and stirred for 3 h, over which time it was allowed to warm to room temperature. 60 Diphenyl(n - butyl)phosphine was synthesized by adding 1.7 M n BuLi solution (1 equiv) to PPh 2 Cl (1 equiv) in cold ( - 78°C) diethyl ether. 61 In the literature preps for these phosphines, they were purified by distillation. However, the syntheses were carried out on much smaller scales than in the literature. Thus as an alternative method of 56 purification, the diphenyl( n - butyl)phosphine and the tris(pyrrolidino)phosphine, were run over a short plug of alumina for purification, which provided colorless oils that were pure by multi - nuclear NMR spec troscopy. Adequate CHN was not obtained on the complexes under study despite many attempts. Their instability has been noted both in the solid state and in solution, and repeated attempts to obtain adequate CHN have demonstrated decomposition during attem pts to transfer the sample to the combustion analysis instrument. The cationic Cr(VI) complexes have been characterized by NMR ( 1 H, 13 C, 31 P, 14 N, and 19 F), X - ray diffraction, and melting point. All experiments carried out with the chromium complexes in this s tudy were conducted with X - ray quality single crystals to ensure purity. Synthetic Procedures General procedure for the synthesis of [NCr(N i Pr 2 ) 2 PR 3 ][SbF 6 ] (3) A 20 mL scintillation vial was charged with 1 equiv of 1 , acetonitrile (3 mL), and a Teflon - coated stirbar. This mixture was stirred at room temperature giving a dark red - orange solution. Separately, a solution of AgSbF 6 (1 equiv) was prepared in acetonitr ile (1 2 mL). The AgSbF 6 solution was then added dropwise to the stirred solution of 1 . Upon addition, copious amounts of off - white precipitate formed, and the solution became dark brown. The resultant mixture was stirred for 20 min after complete addition of the Ag solution. The mixture was then filtered over Celite to remove the precipitate (AgI). The dark brown solution of 2 (filtrate) was once again stirred at room temperature and a solution of PR 3 (1 2 equiv) in acetonitrile (1 2 mL) was added. (Note: 1 equiv of the phosphine was used if it was a solid or high - boiling liquid phosphine that is 57 difficult to remove by recrystallization. 2 equiv of phosphine were used if PR 3 is a low - boiling liquid, easily removed under reduced pressure.) Upon addition of PR 3 , the solution quickly became yellow - orange. The reaction solution was stirred for 1 h at room temperature. The volatiles were then removed under reduced pressure to give a dark residue. This residue was rinsed with small aliquots of cold Et 2 O (3x1 mL) to remove any unreacted 1 . The residue was once more dried under reduced pressure. The residue was dissolved in a minimal amount of CH 2 Cl 2 or CHCl 3 and layered with Et 2 O or n - pentane. The layered solution was then stored overnight at - 35 °C to yield yellow - orange X - ray quality crystals. Note: [NCr(N i Pr 2 ) 2 PR 3 ][BArF 24 ] derivatives were synthesized in a similar fashion for LDP measurements in CDCl 3 and experimental determination of the entropies of activation. Instead of acetonitrile, DCM was generally used a s solvent and TlBArF 24 replaced AgSbF 6 as the precipitation agent. With Tl + instead of Ag + the phosphine could be added to the initial solution for a 1 - pot synthesis. The NMR data for each of these analogues is not reported, as all spectra except the 19 F s pectrum are very similar. Synthesis of 3a [NCr(N i Pr 2 ) 2 PMe 3 ][SbF 6 ] Following the general procedure, the reaction was carried out with 1 (89 mg, 0.226 mmol), AgSbF 6 (78 mg, 0.226 mmol), and PMe 3 (35 mg, 0.460 mmol). This yielded 3a 1 H NMR (500 MHz, CDCl 3 ): 1.61 (d, J = 6.3 Hz, 6H), 1.39 (d, J = 6.3 Hz, 6H), 1.27 (d, J = 9.0 Hz, 6H). 13 C NMR (126 MHz, CD 3 19 F NMR (470 MHz, CDCl 3 31 P NMR (202 MHz, CDCl 3 Synthesis of [NCr(N i Pr 2 ) 2 P n Bu 3 ][SbF 6 ] (3b ) F ollowing the general procedure, the reaction was carried out with 1 (100 mg, 0.254 mmol), AgSbF 6 (87 mg, 0.254 mmol), and P n Bu 3 (51.4 mg, 58 0.51 mmol). This yielded 3b (83.1 mg, 46.4%). M.p.: 50 °C (dec.). 1 H NMR (500 MHz, CDCl 3 ): 7.70 (s, 1H), 7.52(s, 1H), 7 .47 (ddd,J= 19.8, 9.9, 5.1 Hz, 1H), 4.95 (sept,J= 12.8,6.4 Hz, 1H), 3.88 (sept,J= 12.5, 6.3 Hz, 1H), 1.87 (d,J= 10.3 Hz,1H), 1.65 (d,J= 6.3 Hz, 1H), 1.56 (d,J= 6.3 Hz, 1H), 1.15 (d,J= 6.4 Hz, 1H), 1.10 (d,J= 6.4 Hz, 1H). 13 C NMR (126 MHz, CDCl 3 ): 59.85 (d), 58.92, 32.48, 30.30, 25.79 (d), 24.92 (d), 24.56 (d), 23.87,23.79, 13.85. 31 P NMR (202 MHz, CDCl 3 ): 30.0. 19 F NMR (470 MHz, CDCl 3 ): 106.55 to 137.63 (m). Synthesis of [NCr(N i Pr 2 ) 2 P i Bu 3 ][SbF 6 ] (3c) Following the general procedure, the reaction was carried out with 1 (50 mg, 0.127 mmol), AgSbF 6 (44 mg, 0.127 mmol), and P i Bu 3 (36 mg, 0.254 mmol). This yielded 3c (58.6 mg, 64.7%). M.p.: 150 °C (dec.). 1 H NMR (500 MHz, CDCl 3 ): 5.36 (sept, J= 12.4, 6.2 Hz, 2H), 4.00 (sept, J= 12.4, 6.2 Hz, 2H), 2.16 (dt, J= 18.8, 6.3 Hz, 3H), 1.97 (dd, J= 8.5, 6.3 Hz, 6H), 1.84 (d, J= 6.3 Hz, 6H), 1.57 (d, J= 6.2 Hz, 6H), 1.37 (d, J= 6.2 Hz, 6H),1.32 (d, J= 6.2 Hz, 6H), 1.10 (d, J= 6.6 Hz, 18H). 13 C NMR(126 MHz, CDCl3): 59.50 (d), 59.09, 32.56 (d), 29.40, 25.02, 24.87 ,23.94 (d), 23.74, 19.46. 31 P NMR (202 MHz, CDCl 3 ): 32.8. 19 F NMR (470 MHz, CDCl 3 ): - 124.6 (m). Synthesis of [NCr(N i Pr 2 ) 2 P i Pr 3 ][SbF 6 ] (3d) Following the general procedure, the reaction was carried out with 1 (52 mg, 0.132 mmol), AgSbF 6 (45 mg, 0.132 mmol), and P i Pr 3 (33.5 mg, 0.210 mmol). This yielded 3d (33.5 mg, 40%). M.p.: 155 °C (dec.). 1 H NMR (500 MHz, CDCl 3 ): 5.41 (sept, J= 12.5,6.2 Hz, 2H), 4.02 (sept, J= 12.6, 6.3 Hz, 2H), 2.46 2.34 (m, 3H),1.84 (d, J= 6.4 Hz, 6H), 1.58 (d, J= 6.3 Hz, 6H), 1.43 (d, J= 7.2 Hz,9H), 1.40 (d, J= 6.9 Hz, 15H), 1.34 (d, J= 6.2 Hz, 6H). 13 C NMR (126 MHz, CDCl 3 ): 59.71, 59.24, 32. 74, 29.58, 25.12 (d), 24.14, 23.94, 19.65. 59 31 P NMR (202 MHz, CDCl 3 ): 67.0. 19 F NMR (470 MHz, CDCl 3 ): - 105.15 to - 139.23 (m, J= 1809.1, 1654.8,1457.8, 1258.2 Hz). Synthesis of [NCr(N i Pr 2 ) 2 PCy 3 ][SbF 6 ] (3e ) Following the general procedure, the reaction was c arried out with 1 (55 mg, 0.140 mmol), AgSbF 6 (48 mg, 0.140 mmol) and PCy 3 (40.5 mg, 0.140 mmol). This yielded 3e (66 mg, 60%). M.p.:117 °C (dec.). 1 H NMR (500 MHz, CDCl 3 ): 5.35 (sept, J= 12.4,6.2 Hz, 2H), 4.02 (sept, J= 12.6, 6.3 Hz, 2H), 2.06 (dt, J= 21. 0,10.5 Hz, 6H), 1.99 1.86 (m, 16H), 1.83 (d, J= 6.3 Hz, 8H), 1.78 (s,4H), 1.58 (d, J= 6.3 Hz, 10H), 1.56 1.41 (m, 6H), 1.39 (d, J= 6.2 Hz, 8H), 1.36 (d, J= 6.2 Hz, 8H), 1.35 1.11 (m, 16H). 13 C NMR (126 MHz, CDCl 3 ): 59.84, 59.22, 34.84 (d), 32.70, 29.75,29. 65, 27.60 (d), 25.84, 24.30, 23.75. 31 P NMR (202 MHz, CDCl 3 ): 56.9. 19 F NMR (470 MHz, CDCl 3 ): - 123.19 (d, J= 3323.8 Hz). Synthesis of [NCr(N i Pr 2 ) 2 PPhMe 2 ][SbF 6 ] (3f) Following the general procedure, the synthesis was carried out with 1 (50 mg, 0.127 mmol, 1 equiv), AgSbF 6 (44 mg, 0.127 mmol, 1 equiv), and PPhMe 2 (35 mg, 0.253 mmol, 2 equiv). This yielded 3f °C dec. 1 H NMR (500 MHz, CDCl 3 23.2 Hz, 2H), 3.91 (dt, J = 12.0, 5.9 Hz, 2H), 1.99 (d, J = 10.3 Hz, 6H), 1.57 (dd, J = 11.1, 6.2 Hz, 13 C NMR (126 MHz, CDCl 3 ), 59.81 (d), 58.73 (s), 32.05 (s), 29.68 (s), 23.40 (s), 22.63 (s), 14.56 (s), 14.32 (s). 19 F NMR (470 MHz, CDCl 3 31 P NMR (202 MHz, CDCl 3 14 N NMR (36 MHz, CDCl 3 Synthesis of [NCr(N i Pr 2 ) 2 PPh 2 Me] [SbF 6 ] (3g) Following the general procedure, the reaction was carried out with 1 (50 mg, 0.127 mmol), AgSbF 6 (43 mg, 0.127 mmol), and PPh 2 Me (50 mg, 0.250 mmol). This yielded 3g (50.1 mg, 54.9%). M.p.: 138 °C. 1 H NMR (500 MHz, CDCl 3 ): 7.94 7.27 (m, 10H), 5 .40 (sept, J= 11.6, 5.7 Hz, 2H), 3.89 (sept, J= 12.0, 5.9 Hz, 2H), 60 2.34 (d, J= 9.2 Hz, 3H), 1.89 (s, 1H), 1.55 (dd, J= 15.3, 6.2 Hz, 12H), 1.40(s, 3H), 1.27 (d, J= 6.0 Hz, 7H), 1.07 (d, J= 6.0 Hz, 6H). 13 C NMR (126 MHz, CDCl 3 ): 132.42 (d), 132.09, 131.79 ( d), 130.70, 129.92(d), 129.21, 128.31, 127.93, 59.97 (d), 58.89, 31.99, 29.11, 23.05,22.25, 12.67 (d). 31 P NMR (202 MHz, CDCl 3 ): 20.7. 19 F NMR (470 MHz, CDCl 3 ): - 107.41 to - 139.91 (m). The X - ray diffraction study was carried out on the BArF 24 ¯ salt, which gave single crystals and was made analogously to the SbF 6 ¯ salt. Synthesis of [NCr(N i Pr 2 ) 2 PPhEt 2 ][SbF 6 ] (3h) Following the general procedure, the reaction was carried out with 1 (50 mg, 0.127 mmol), AgSbF 6 (43 mg, 0.127 mmol), and PPhEt 2 (34 mg, 0.246 mmol). This yielded 3h (28.1 mg 33.1%). M.p.: 115 °C (dec.). 1 H NMR (500 MHz, CDCl 3 ): 7.64 7.50 (m, 5H),5.21 (sept, J= 12.5, 6.3 Hz, 2H), 3.90 (sept, J= 12.5, 6.3 Hz, 2H),2.48 2.33 (m, J= 15.0, 10.7, 7.5 Hz, 2 H), 2.33 2.21 (m, 2H), 1.65 (d, J= 6.3 Hz, 6H), 1.57 (d, J= 6.3 Hz, 6H), 1.31 1.18 (m, 14H), 1.14 (d, J= 6.2 Hz, 6H). 13 C NMR (126 MHz, CDCl 3 ): 132.09,130.87 (d), 130.25 (d), 59.65, 58.51, 31.88, 23.53, 22.34, 17.60(d), 7.16. 31 P NMR (202 MHz, CDCl 3 ): 34.9 . 19 F NMR (470 MHz, CDCl 3 ): - 123.30 (m). Synthesis of [NCr(N i Pr 2 ) 2 PPh 2 Et][SbF 6 ] (3i) Following the general procedure, the reaction was carried out with 1 (100 mg, 0.254 mmol), AgSbF 6 (87 mg, 0.254mmol), and PPh 2 Et (63 mg, 0.298 mmol). This yielded 3i (94 mg, 51.6%). M.p.: 150 °C (dec). 1 H NMR (500 MHz, CDCl 3 ): 7.60 (dt,J= 11.6, 6.1 Hz,6H), 7.45 7.28 (m, 4H), 5.20 (sept, J= 6.3 Hz, 2H), 3.93 (sept, J= 6.4 Hz, 2H), 2.62 (p, J= 7.3 Hz, 2H), 1.70 (d, J= 6.2 Hz, 6H), 1.56 (d, J= 6.2 Hz, 6H), 1.24 (d, J= 6.3 Hz, 6H), 1.11 (dt, J= 18.0,7.4 Hz, 3H), 1.01 (d, J= 6.3 Hz, 6H). 13 C NMR (126 MHz, CDCl 3 ): 132.98 61 (d), 132.85 (d), 130.20 (d), 125.36 (d), 60.06, 59.07, 32.31,29.49, 23.39, 22.60 (d), 22.34, 7.98 (d). 31 P NMR (202 MHz, CDCl 3 ): 35.3. 19 F NMR (470 MHz, CDCl 3 ): - 122.12 (m). Synthesis of [NCr(N i Pr 2 ) 2 PPh 2 n Bu][PF 6 ] (3j) A 20 mL scintillation vial was charged with 1 (100 mg,0.254 mmol), CH 2 Cl 2 (5 mL), diphenyl( n - butyl)phosphine (61 mg,0.510 mmol), and a Teflon - coated stir bar. This solution was stirred at room temperature to give a dark red - orange solution. Separately, a suspension of TlPF 6 was prepared in 2 mL of CH 2 Cl 2 . The TlPF 6 suspension was then added dropwise to the stirred solution of 1 and tris(pyrrolidinyl)phosphine. A yellow precipitate began to form on addition. After addition, the solution was stirred 3 h at room temperature. Then, the reaction mixture was filtered through Celite to remove the precipitate, and the bright orange filtrate was collected. The volatiles were removed from the filtrate unde r reduced pressure, leaving a dark residue. The residue was washed with cold Et 2 O (31 mL), and the solution was again dried under reduced pressure. The residue was dissolved in a minimal amount of CH 2 Cl 2 and layered with pentane. The layered solution was s tored at - 35 °C overnight to get X - ray quality orange crystals. This yielded 3j (104 mg, 63%). M.p.: 74 77°C (dec). 1 H NMR (500 MHz, CDCl 3 ): 7.78 7.47 (m,10H), 5.22 (sept, J= 6.3 Hz, 2H), 3.93 (sept, J= 6.4 Hz, 2H), 2.53(q, J= 7.8 Hz, 2H), 1.69 (d, J= 6.3 Hz, 6H), 1.56 (d, J= 6.2 Hz, 6H),1.38 (dq, J= 23.9, 8.0, 7.6 Hz, 4H), 1.25 (d, J= 6.3 Hz, 6H), 0.99 (d, J= 6.3 Hz, 6H), 0.84 (t, J= 7.1 Hz, 3H). 13 C NMR (126 MHz, CDCl 3 ):132.89 (d), 132.80 (d), 130.23 (d), 125.81 (d), 60.11, 59.04, 32.29,29.44, 29.05 (d), 25.67 (d), 23.99 (d), 23.33, 22.32, 13.60. 31 P NMR (202 MHz, CDCl 3 ): 32.7. 19 F NMR (470 MHz, CDCl 3 ): - 73.41 (d, J= 712.4 Hz). Synthesis of [NCr(N i Pr 2 ) 2 PPh 2 Cy][SbF 6 ] (3k) Following the general procedure, the reaction was carried out with 1 (75 mg, 0.191 mm ol), AgSbF 6 (68 mg, 0.195 mmol), and PPh 2 Cy (63 mg, 0.230 mmol). This yielded 3k (62.5 mg, 63.8%). M.p.: 168 °C (dec). 1 H NMR (500 MHz, 62 CDCl 3 ): 7.63 (m, J= 15.0,7.8 Hz, 10H), 4.94 (sept, J= 6.2 Hz, 2H), 3.92 (sept, J= 6.3 Hz, 2H),2.50 2.33 (m, 1H), 2.31 2.14 (m, 2H), 1.93 1.77 (m, 2H), 1.72 (d, J= 6.3 Hz, 6H), 1.57 (d, J= 6.2 Hz, 6H), 1.45 1.33 (m, 6H), 1.20 (d, J= 6.3 Hz, 6H), 1.02 (d, J= 6.2 Hz, 6H), 0.93 0.72 (m, 2H). 13 C NMR (126 MHz, CDCl 3 ): 133.77 (d, J= 10.2 Hz), 132.89, 129.85 (d, J= 10.2 Hz), 123.03 (d, J= 42.8 Hz), 59.92 (d, J= 1.8 Hz), 58.90,36.94 (d, J= 22.4 Hz), 32.14 (d, J= 1.8 Hz), 29.05, 26.63 (d, J= 12.5 Hz), 25.46, 23.34, 22.39 . 31 P NMR (202 MHz, CDCl 3 ): 45.2. 19 F NMR (470 MHz, CDCl 3 ): 106.76 - 136.05 (m). The X - ray diffraction study was done with the BPh 4 ¯ salt, which gave single crystals and was made analogously to the SbF 6 ¯ salt. Synthesis of [NCr(N i Pr 2 ) 2 PPhCy 2 ][SbF 6 ] (3l) Fol lowing the general procedure, the reaction was carried out with 1 (52 mg, 0.132 mmol), AgSbF 6 (45 mg, 0.132 mmol), and PPhCy 2 (48.3 mg, 0.176 mmol). This yielded 3l (58.2 mg, 56.1%). M.p.: 115 °C (dec). 1 H NMR (500 MHz, CDCl 3 ): 7.79 7.37 (m, 5H), 5.38 (sep t, J= 12.2, 6.1 Hz, 2H), 3.99 (sept, J= 12.2, 6.0 Hz, 2H), 2.34 (d, J= 8.7 Hz, 2H), 1.91 (d, J= 21.6 Hz, 8H), 1.77 (d, J= 6.2 Hz,8H), 1.60 (d, J= 6.2 Hz, 6H), 1.44 1.05 (m, 24H). 13 C NMR (126 MHz, CDCl 3 ): 132.89 (d), 132.49, 130.01 (d), 122.36 (d),59.96, 5 9.02, 34.31 (d), 32.28, 29.37, 28.32, 28.04, 27.17 26.56 (m), 25.52, 23.50, 23.19. 31 P NMR (202 MHz, CDCl 3 ): 49.8. 19 F NMR (470 MHz, CDCl 3 ): - 108.70 to - 142.59 (m). Synthesis of [NCr(N i Pr 2 ) 2 P(OEt) 3 ][SbF 6 ] (3m ) Following the general procedure, the reaction was carried out with 1 (100 mg, 0.254 mmol), AgSbF 6 (87 mg, 0.254 mmol), and P(OEt) 3 (43 mg, 0.26 mmol). This yielded 3m (48.2 mg, 28.3%). M.p.: 149 150 °C. 1 H NMR (500 MHz, CDCl 3 ): 5.36 (sept, J= 12.6,6.3 Hz, 2H), 4.24 (p, J= 7.1 Hz, 6H), 4.03 (sept, J= 12.5, 6.1 Hz,2H), 1.87 (d, J= 6.3 Hz, 6H), 1.60 (d, J= 6.3 Hz, 6H), 1.45 1.35 (m,18H), 1.28 (d, J= 6.3 Hz, 9H). 13 C 63 NMR (126 MHz, CDCl 3 ): 65.06(d), 60.32 (d), 59.19, 32.14, 30.46, 23.17, 22.81, 16.29 (d). 31 P N MR (202 MHz, CDCl 3 ): 122.6. 19 F NMR (470 MHz, CDCl 3 ): - 122.86 (m). Synthesis of [NCr(N i Pr 2 ) 2 P(O i Pr) 3 ][SbF 6 ] (3 n) Following the general procedure, the reaction was carried out with 1 (50 mg, 0.127 mmol), AgSbF 6 (43 mg, 0.127 mmol), and P(O i Pr) 3 (38 mg, 0.1 82 mmol). This yielded 3n (44.3 mg, 51%). M.p.: 138 140 °C. 1 H NMR (500 MHz, CDCl 3 ): 5.35 (dt, J= 12.6,6.3 Hz, 2H), 4.89 4.70 (m, 3H), 4.05 (dt, J= 12.5, 6.2 Hz, 2H), 1.92(d, J= 6.3 Hz, 6H), 1.56 (d, J= 6.3 Hz, 6H), 1.41 (d, J= 6.2 Hz, 18H),1.38 (d, J= 6.3 Hz, 6H), 1.30 (d, J= 6.3 Hz, 6H). 13 C NMR (126 MHz, CDCl 3 ): 75.15 (d), 59.98 (d), 59.26, 32.48, 29.97, 24.11(d), 23.00 (d). 31 P NMR (202 MHz, CDCl 3 ): 119.1. 19 F NMR (470 MHz, CDCl 3 ): - 122.86 (m). Synthesis of [NCr(N i Pr 2 ) 2 P(NC 4 H 8 ) 3 ][PF 6 ] (3o) A 20 mL scintillation vial was charged with 1 (100 mg,0.254 mmol), CH 2 Cl 2 (5 mL), tris(pyrrolidinyl)phosphine (61 mg,0.510 mmol), and a Teflon - coated stir bar. This solution was stirred at room temperature to give a dark red - orange solution. Separately, a suspension of TlPF 6 was prepared in CH 2 Cl 2 (2 mL). The TlPF 6 suspension was then added dropwise to the stirred solution of 1 and tris(pyrrolidinyl)phosphine. A yellow precipitate began to form on addition. Upon complete addition, the solution was stirred 3 h at room temperature. Then, the reaction mixture was filtered over Celite to remove the precipitate, and the orange filtrate was collected. The volatiles were removed from the filtrate under reduced pressure, leaving a dark residue. The residue was wash ed with cold Et 2 O (3x1 mL) and was again dried under reduced pressure. The residue was dissolved in a minimal amount of CH 2 Cl 2 and layered with pentane. The layered solution was stored at - 35 °C overnight to get X - ray quality orange crystals of 3o (102 mg, 61.5%). M.p.: 145 147 °C (dec). 1 H NMR (500 MHz, CDCl 3 ): 5.29 (sept, J= 12.1, 6.0 Hz, 2H),3.99 (sept, J= 12.0, 5.9 Hz, 2H), 3.14 (d, J= 4.5 Hz, 12H), 1.89 (s,12H), 1.85 (d, J= 6.2 Hz, 6H), 1.48 (d, J= 6.1 Hz, 6H), 1.37 1.29 (m, 12H). 13 C NMR (126 64 MHz, CD Cl 3 ) 58.81, 58.62, 47.56 (d),33.04, 29.17, 25.97 (d), 23.54, 22.75. 31 P NMR (202 MHz, CDCl 3 ):103.1. 19 F NMR (470 MHz, CDCl 3 ): 73.85 (d, J=712 MHz). Synthesis of [NCr(N i Pr 2 ) 2 PPh 3 ][BArF 24 ] (3p) A 20 mL scintillation vial was charged with 1 (100 mg,0.254 mmol), a Teflon - coated stir bar, CH 2 Cl 2 (2 mL) and PPh 3 (64.5 mg, 0.250 mmol). Separately, TlBArF 24 (282 mg, 0.254 mmol) was dissolved in 2 mL of CH 2 Cl 2 . The TlBArF 24 solution was then added drop wise to the stirred chromium solution, resulting in rapid formation of a yellow precipitate. Upon complete addition, the reaction was stirred for 8 h at room temperature. Then, there action was filtered over Celite to remove the precipitate, and the orange filtrate was collected. The volatiles were removed from the filtrate under reduced pressure resulting in a brown residue. The residue was washed with cold pentane, and the volatiles were removed under reduced pressure. The residue was dissolved in minimal amount of Et 2 O and layered with pentane and stored at - 35 °C overnight to get X - ray quality orange crystals of 3p (292 mg, 35.4%). From the X - ray diffraction study the structure of the cation could be gleaned but had full molecule disorder; in addition, t he BArF 24 ¯ was severely disordered. Consequently, the structure is of poor quality. M.p.:110 112°C (dec.). 1 H NMR (500 MHz, CDCl 3 ): 7.71 (s, 8H), 7.66 7.42 (m, 16H), 4.95 (sept, J= 12.6, 6.3 Hz, 2H), 3.87 (sept, J= 12.6,6.3 Hz, 2H), 1.76 (d, J= 6.3 Hz, 6H) , 1.54 (d, J= 6.3 Hz, 6H), 1.13(d, J= 6.3 Hz, 6H), 0.76 (d, J= 6.3 Hz, 6H). 13 C NMR (126 MHz,CDCl 3 ): 161.83 (dd), 134.93, 133.61 (d), 133.13 (d), 130.11 (d),129.03 (d), 127.93, 126.31, 125.85 (d), 123.60, 121.43, 117.61,59.97 (d), 59.77 (d), 32.51 (d), 29. 19, 22.95, 21.96. 31 P NMR (202 MHz, CDCl 3 ): 35.0. 19 F NMR (470 MHz, CDCl 3 ): 62.42 (s). Synthesis of [NCr(N i Pr 2 ) 2 NCCH 3 ][SbF 6 ] (2) A 20 mL scintillation vial was charged with 1 (50 mg,0.127 mmol), a Teflon - coated stir bar, CH 2 Cl 2 (4 mL), and acetonitrile ( 60 L). The solution was stirred at room temperature to give a dark red - orange solution. Separately, AgSbF 6 65 (43 mg,0.125 mmol) was suspended in CH 2 Cl 2 (2 mL). The AgSbF 6 suspension was then added dropwise to the chromium solution, resulting in rapid formation of an off - white precipitate. Upon complete addition, the solution was stirred 3 h at room temperature. Then, the reaction mixture was filtered over Celite to remove the precipitate, and the red filtrate was collected. The volatiles were removed from the filtrate under reduced pressure, leaving a dark brown residue. The residue was washed with cold Et 2 O (3x1 mL), and once again, the volatiles were removed under reduced pressure. The residue was dissolved in a minimal amount of CH 2 Cl 2 and layered with Et 2 O. This solution was stored overnight at - 35 °C to give X - ray quality red - orange crystals of 2 (30.9 mg, 43.5%). M.p.: 126 129 °C (dec.). 1 H NMR (500 MHz,CDCl 3 ): 5.57 (s ept, J= 7.5, 6.7 Hz, 2H), 4.07 (sept, J= 6.2 Hz, 2H), 2.58 (s, 3H), 2.02 (d, J= 6.2 Hz, 6H), 1.52 (d, J= 6.2 Hz, 6H), 1.39 (d, J= 6.2 Hz, 6H), 1.20 (d,J= 6.3 Hz, 6H). 13 C NMR (126 MHz, CDCl 3 ): 60.34 (d), 59.87, 31.15, 30.91, 22.40, 22.25. 19 F NMR (470 MHz, CDCl 3 ): 123.79 (m). Additional Complexes Studied for Ion Pairing Effects Synthesis of 3a[BArF 24 ]. A 20 mL scintillation vial was charged with 1 (50 mg, 0.127 mmol, 1 equiv), PMe 3 (0.301 mmol, 2.4 equiv), 3 mL of DCM, and a magnetic stir bar. To the stirre d solution was added a solution of TlBArF 24 (135 mg, 0.127 mmol, 1 equiv) in 1 mL of DCM. Upon addition, a copious amount of yellow precipitate formed, and the solution went from dark red to bright orange. This solution was stirred for 3 h at room temperat ure. The TlI precipitate was removed by filtration through Celite, and the bright orange filtrate was concentrated under reduced pressure. The concentrated solution was layered with pentane and chilled to - 35 °C, which afforded X - ray - quality crystals (92.6 mg, 60.4%). Mp: 93 °C (dec.). 1 H NMR (500 MHz, CDCl 3 ): 7.69 (s, 8H), 7.53 (s, 4H), 5.01 (sept, J = 12.8, 6.4 Hz, 2H), 3.95 (sept, J = 12.5, 6.3 Hz, 2H), 1.82 (d, J = 6.3 Hz, 6H), 1.59 (d, J = 6.3 Hz, 6H), 1.51 (d, J = 10.6 Hz, 9H), 1.22 (dd, J = 14.2, 6.4 Hz, 14H). 66 13 C NMR (126 MHz, CDCl 3 59.61 (s), 59.04 (s), 31.92 (s), 30.23 (s), 22.97 (s), 15.89 (s), 15.64 (s). 19 F NMR (470 MHz, 31 P NMR (202 MHz, CDCl 3 ): 7.16 (s). 1 4 N NMR (36 MHz, CDCl 3 ): 1010.7 (s), 449.4 (s). Synthesis of 3a[BPh 4 ]. A scintillation vial was charged with 1 (75 mg, 0.191 mmol, 1 equiv), 3 mL of acetonitrile, and a magnetic stir bar. To this stirred solution was added a suspension of AgBPh 4 (82 mg, 0. 191 mmol, 1 equiv) in 1 mL of DCM. This solution was stirred for 3 h at room temperature. During this time the solution changed color from dark red - orange to dark brown, and an off - white precipitate was formed. The precipitate was removed by filtration thr ough Celite. The dark brown filtrate was stirred, and a solution of 30 mg of PMe 3 (0.394 mmol) in 1 mL of acetonitrile was added. This solution was stirred for 1 h at room temperature and then dried in vacuo. The residue was rinsed with several small aliquots of diethyl ether and dried under reduced pressure once again. The complex was dissolved in a minimal amount of DCM, layered with diethyl ether, and chilled to - °C (dec.). 1 H NMR (500 MHz, CDCl 3 ): 7.44 (s, 8H), 7.05 (s, 8H), 6.90 (s, 4H), 4.84 (sept, J = 12.6, 6.3 Hz, 2H), 3.88 (dt, J = 12.6, 6.1 Hz, 2H), 1.76 (d, J = 6.2 Hz, 6H), 1.55 (d, J = 6.2 Hz, 7H), 1.18 (t, J = 6.1 Hz, 13H), 1.08 (d, J = 10.8 Hz, 9H). 13 C NMR (126 MHz, CDCl 3 ): 164.37 (q), 136.47 (s), 125.78 (s), 121.89 (s), 59.56 (s), 58.79 (s), 32.11 (s), 30.4 1 (s), 23.39 (s), 15.70 (s), 15.45 (s). 31 P NMR (202 MHz, CDCl 3 ): 8.02 (s). 14 N NMR (36 MHz, CDCl 3 ): 1015.9 (s), 446.4 (s). Synthesis of 3f[PF 6 ]. A 20 mL scintillation vial was charged with 1 (50 mg, 0.127 mmol, 1 equiv), a magnetic stir bar, PPhMe 2 (35 m g, 0.253 mmol, 2 equiv), and 3 mL of acetonitrile. To this stirred solution was added a solution of TlPF 6 (44.3 mg, 0.127 mmol, 1 equiv) in 1 mL of 67 acetonitrile. Upon addition, copious amounts of yellow precipitate formed, and the solution went from dark r ed - orange to a lighter orange. This solution was stirred for 3 h at room temperature. The TlI precipitate was filtered off the bright orange solution through Celite, and the filtrate was pumped to dryness under reduced pressure. The residue was crystallize d with CH 2 Cl 2 /pentane at - 35 °C, which afforded X - ray - 1 H NMR (500 MHz, CDCl 3 6.3 Hz, 2H), 3.91 (sept, J = 12.5, 6.2 H z, 2H), 1.99 (d, J = 10.3 Hz, 6H), 1.56 (dd, J = 8.7, 6.4 Hz, 12H), 1.24 (d, J = 6.2 Hz, 12H). 13 C NMR (126 MHz, CDCl 3 ): 132.26 (d), 130.68 (d), 130.13 (d), 59.98 (d), 58.94 (s), 32.28 (d), 29.92 (s), 23.65 (s), 22.89 (s), 14.69 (d). 19 F NMR (470 MHz, CDCl 3 31 P NMR (202 MHz, CDCl 3 (sept). 14 N NMR (36 MHz, CDCl 3 ): 1016.1 (s), 451.1 (s). Synthesis of 3f[BArF 24 ]. A 20 mL scintillation vial was charged with 1 (50 mg, 0.127 mmol, 1 equiv), PPhMe2 (35 mg , 0.253 mmol, 2 equiv), 3 mL of DCM, and a magnetic stir bar. To this stirred solution was added a solution of TlBArF 24 (135 mg, 0.127 mmol, 1 equiv) in 1 mL of DCM. Upon addition, copious amounts of yellow precipitate formed, and the solution went from da rk red - orange to transparent bright orange. The solution was stirred for 3 h at room temperature. The TlI precipitate was filtered through Celite, and the bright orange filtrate was concentrated in vacuo. The concentrated solution in DCM ( 1 mL) was layere d with pentane and chilled to - 35 ° C overnight to obtain X - ray - quality orange crystals (62 mg, 38.5%). Mp: 115 ° C (dec.). 1 H NMR (500 MHz, CDCl 3 ): 7.70 (s, 8H), 7.52 (s, 5H), 7.51 7.40 (m, 4H), 4.95 (sept, J = 12.8, 6.4 Hz, 2H), 3.88 (sept, J = 12.5, 6.3 Hz, 2H), 1.87 (d, J = 10.3 Hz, 6H), 1.65 (d, J = 6.3 Hz, 6H), 1.56 (d, J = 6.3 Hz, 6H), 1.15 (d, J = 6.4 Hz, 6H), 1.10 (d, J = 6.4 Hz, 6H). 13 C NMR (126 MHz, CDCl 3 (m), 134.76 (s), 132.65 (s), 130.15 (d), 129.65 (d), 128.87 (d), 68 127.77 (s), 125.60 (s), 123.43 (s), 121.26 (s), 117.46 (s), 59.70 (s), 59.00 (s), 31.85 (s), 29.71 (s), 23.07 (s), 22.27 (s), 14.86 (d). 19 F NMR (470 MHz, CDCl 3 31 P NMR (202 MHz , CDCl 3 ): 12.74 (s). 14 N NMR (36 MHz, CDCl 3 ): 1008.5 (s), 445.8 (s). Synthesis of 3f[BArF 20 ]. A 20 mL scintillation vial was charged with 1 (60 mg, 0.153 mmol, 1 equiv), PPhMe 2 (42 mg, 0.303 mmol, 2 equiv), 3 mL of DCM, and a magnetic stir bar. To this sti rred solution was added a solution of KBArF 20 (110 mg, 0.153 mmol) in 1 mL of DCM. This reaction mixture was stirred for 8 h at room temperature. Over the course of the reaction, the dark red - orange, cloudy solution slowly cleared and became bright orange as a yellow precipitate formed. The KI precipitate was removed by filtration through Celite, and the bright orange solution was concentrated under reduced pressure. The concentrated filtrate ( 1 mL) was layered with pentane ( 2 mL) and chilled to 30 ° C ov ernight to yield X - ray quality orange crystals (46 mg, 27.8%). Mp: 135 °C (dec.). 1 H NMR (600 MHz, CDCl 3 6.4 Hz, 2H), 3.91 (sept, J = 12.5, 6.2 Hz, 2H), 1.90 (d, J = 10.3 Hz, 6H), 1.66 (d, J = 6.3 Hz, 6H), 1.58 (d , J = 6.3 Hz, 6H), 1.21 (d, J = 6.4 Hz, 6H), 1.16 (d, J = 6.4 Hz, 6H). 13 C NMR (126 MHz, CDCl 3 ): 148.30 (d), 137.28 (t), 132.78 (s), 130.31 (d), 129.92 (d), 128.52 (d), 59.97 (d), 59.15 (s), 32.08 (d), 29.88 (s), 23.35 (s), 22.47 (d), 14.88 (d). 19 F NMR (4 70 MHz, CDCl 3 31 P NMR (202 MHz, CDCl 3 ): 12.92 (s). 14 N NMR (36 MHz, CDCl 3 ): 1010.3 (s), 448.5 (s). Synthesis of 3f[Al(OC(CF 3 ) 3 ) 4 ]. A 20 mL scintillation vial was charged with 1 (100 mg, 0.254 mmol, 1 equiv), 3 mL of acetonitrile, and a magnetic stir bar. To this was added a solution of AgAl(OC(CF 3 ) 3 ) 4 (273 mg, 0.254 mmol, 1 equiv) in 1 mL of acetonitrile. The resultant solution was s tirred for 1 h at room temperature, during which time a yellowish precipitate formed, and the solution darkened from red - orange to brown. The AgI was removed by filtration through Celite, 69 and the filtrate was once again stirred. To the stirring filtrate wa s added a solution of PPhMe 2 (35 mg, 0.254 mmol, 1 equiv) in 1 mL of acetonitrile. This solution was stirred at room temperature for 2 h, changing slightly in color from dark brown to dark orange - brown. The solution was dried under reduced pressure, and th e residue was washed with several aliquots of cold pentane. The solids were again dried under reduced pressure and dissolved in a minimal amount of chloroform. This concentrated solution was layered with pentane and chilled to - 35 °C, which afforded yellow crystals (183 mg, 52.5%). Removal of all traces of solvent from the aluminate complex (without decomposing the complex) was difficult, and solvent peaks were identified in the NMR spectra. 1 H NMR (500 MHz, CDCl 3 sept, J = 6.4 Hz, 2H), 3.93 (sept, J = 6.1 Hz, 2H), 1.94 (d, J = 10.3 Hz, 7H), 1.68 (d, J = 6.3 Hz, 6H), 1.60 (d, J = 6.3 Hz, 6H), 1.23 (d, J = 6.4 Hz, 8H), 1.20 (d, J = 6.4 Hz, 6H). 13 C NMR (126 MHz, CDCl 3 ): 132.85 (d), 130.35 (d), 129.88 (d), 128.51 (d), 121.35 (q), 60.00 (d), 59.19 (s), 32.03 (d), 29.85 (s), 23.22 (s), 22.40 (d), 14.80 (s), 14.21 (s). 19 F NMR (470 MHz, CDCl 3 31 P NMR (202 MHz, CDCl 3 ): 12.71 (s). 14 N NMR (36 MHz, CDCl 3 Synthesis of 3f[BPh 4 ]. A 20 mL scintillation vial was charged with 1 (100 mg, 0.254 mmol, 1 equiv), 3 mL of acetonitrile, and a magnetic stir bar. To this was added a suspension of AgBPh 4 (108 mg, 0.254 mmol, 1 equiv) in 1 mL of DCM. The resultant solution was stirred for 3 h at room te mperature, during which time an off - white precipitate formed, and the solution darkened from red - orange to brown. The AgI was removed by filtration through Celite, and the filtrate was once again stirred. To the stirred filtrate was added a solution of PPh Me 2 (70 mg, 0.506 mmol, 2 equiv) in 1 mL of acetonitrile. This solution was stirred at room temperature for 1 h, changing slightly in color from dark brown to dark orange - brown. The solution was dried under reduced pressure, and the residue was rinsed with several small aliquots of diethyl ether. The residue was dried again and 70 dissolved in a minimal amount of DCM. This concentrated solution was layered with diethyl ether and chilled to - 35 °C overnight, affording bright orange crystals (23.4 mg, 12.7%). Mp : 84 °C (dec.). 1 H NMR (600 MHz, CDCl 3 ): 7.49 (dt, J = 6.5, 3.2 Hz, 1H), 7.43 (dd, J = 11.4, 3.6 Hz, 11H), J = 12.5, 6.3 Hz, 2H), 1.52 (dd, J = 11.9, 6.3 Hz, 13H), 1.39 (d, J = 10.3 Hz, 6H), 1.14 (d, J = 6.4 Hz, 6H), 1.04 (d, J = 6.3 Hz, 6H). 13 C NMR (126 MHz, CDCl 3 ): 164.38 (q), 136.49 (d), 132.16 (d), 130.23 (m), 125.65 (d), 121.80 (d), 59.88 (s), 58.75 (s), 32.06(s), 29.74 (s), 23.48 (s), 23.71 (s), 14.0 8 (s). 31 P NMR (202 MHz, CDCl 3 ): 13.60 (s). 14 N NMR (36 MHz, CDCl 3 ): 1006.8 (s), 445.9 (s). 71 Details on NMR Techniques and Analyses Additional Experimental Data for LDP Measurements Table 2 . 11 Details from LDP measurements with various X ¯ ligands in CDCl 3 . LDP Measurements CDCl 3 Cation Anion Experimental Rate (kcal/mol) (kcal/mol) Standard Deviation Temp (°C) 3f PF 6 0.38 19.90 16.96 0.07 -- a BArF 24 0.41 19.50 16.60 0.02 49.28 BArF 20 0.24 19.50 16.64 0.01 43.70 Al(O t BuF 9 ) 4 0.23 19.51 16.66 0.01 43.84 BPh 4 0.40 19.47 16.57 0.01 48.44 SbF 6 0.16 19.87 16.99 -- b 45.86 3a BArF 24 0.32 19.78 16.86 0.01 51.32 BPh 4 0.46 19.64 16.71 0.01 52.77 SbF 6 0.11 20.11 17.24 -- b 45.86 Table 2 . 12 Details from LDP measurements with various X ¯ ligands in CD 3 CN. LDP Measurements CD 3 CN Cation Anion Experimental Rate (kcal/mol) H (kcal/mol) Standard Deviation Temp (°C) 3f PF 6 0.56 19.46 16.53 0.01 52.01 BArF 24 0.38 19.47 16.58 0.01 48.02 BArF 20 0.25 19.47 16.62 0.01 43.92 Al(O t BuF 9 ) 4 0.25 19.47 16.62 0.01 43.84 BPh 4 0.46 19.36 16.47 0.01 48.44 SbF 6 0.57 19.46 16.53 0.01 52.04 3a BArF 24 0.38 19.54 16.65 0.03 45.96 BPh 4 0.34 19.55 16.66 0.01 48.15 SbF 6 0.58 19.59 16.64 0.03 54.31 72 Table 2 . 13 Experimentally determined rate of Cr N i Pr 2 = - 9 e.u.), and temperature of measurement. Phosphine Compound number Rate Constant (s 1 ) (kcal/mol) LDP (kcal/mol) Std. Dev. Temperature (°C) PMe 3 3a a 0.58 19.59 16.64 0.03 54.31 P n Bu 3 3b 0.36 19.69 16.77 0.005 51.13 P i Pr 3 3c a 0.46 20.14 17.13 0.02 62.90 P i Bu 3 3d 0.52 20.19 17.17 0.02 60.72 PCy 3 3e a 0.43 20.32 17.29 0.03 62.96 PPhMe 2 3f 0.57 19.46 16.53 0.01 52.04 PPh 2 Me 3g 0.33 18.96 16.16 0.01 34.71 PPhEt 2 3h 0.45 19.58 16.65 0.02 51.46 PPh 2 Et 3i 1.16 19.09 16.15 0.02 53.62 PPh 2 n Bu 3j 0.4 19.16 16.31 0.008 43.71 PPh 2 Cy 3k 1.01 19.41 16.43 0.02 57.36 PPhCy 2 3l 0.54 19.26 16.37 0.02 48.25 P(OEt) 3 3m a 0.48 18.5 15.73 0.02 34.88 P(O i Pr) 3 3n a 1.15 18.81 15.91 0.04 48.82 P(NC 4 H 8 ) 3 3o 0.68 19.1 b 16.21 0.05 b PPh 3 3p 0.65 19.13 16.24 (15.95) c 0.03 48.10 NCCH 3 2 1.02 17.95 15.19 0.01 33.33 a LDP value in CD 3 CN was measured via in situ generated species stabilized with excess phosphine. b Measured LDP was taken in multiple trials, taking a single measurement on three different samples due to compound instability. As a result, three different temperatures were calibrated, one for each sep is approximate, as , for each run was fully calculated with the calibrated temperature for that run, before an average was taken of the final LDP values. c The LD P value for PPh 3 was measured in CDCl 3 , as the chromium complex with PPh 3 is completely unstable in CD 3 CN. The value reported in parenthesis is the model predicted LDP value (CD 3 CN). 73 DOSY Experiments: General Considerations The Varian Dbppste_cc (DOSY bipolar pulse pair simulated spin echo convection corrected) pulse sequence was utilized for all experiments except for the 19 F DOSY measurements of the 3f [BArF 20 [ salt. For this complex, the large chemical shift range led to p hasing and modulation issues under the normal Dbppste_cc strategy. As a result, an alternate experiment was developed by adapting the Oneshot_CHORUS pulse sequence developed by Morris, et. al. 62 This experiment was utilized in order to obtain reliable 19 F DOSY data without suffering from the limited bandwidth of excitation of standard DOSY experiments. All spectra were multiplied by weighted exponential of 10 Hz and baseline corrected before applying DOSY Processing. Standard DOSY processing as supplied by the vendor was used based on peak heights and with compensation for non - uniform gradients. Convection Effects Due to the need to compare each compound via two different nuclei, two separate DOSY experiments were run for each compound (except 3f [BPh 4 ], 3a [BPh 4 ] , and 3a [BArF 24 ]). The parameter differences between the 1 H and 19 F DOSY experiments for a given compound (relaxation delay, acquisition time, gradient length, diffusion delay, etc.) gave rise to inconsistencies in data. These inconsistencies were traced back to convection differences between successive experiments. By running convection corrected pulse sequences and including i nternal standards in each experiment, the convection effects have been eliminated, such that multiple trials of the same compound provide diffusion coefficients within experimental error. The internal standard for all experiments was chosen based on the s implicity of its resonance in both 1 H and 19 F NMR (one sharp singlet by each method), and the uniqueness of the resonances relative to the resonances of the chromium complexes. For these purposes, 1,3,5 - tris(trifluoromethyl)benzene was utilized as the inte rnal standard. The diffusion coefficients 74 experiment. This ratio can then be used to compare the diffusion rate of the anion and cation for a chromium complex w ithout the complication of convection differences between experiments. As a result of the much more complex pulse sequence required by the 3f [BArF 20 ] complex to obtain DOSY signal, convection correction in the pulse sequence was not feasible. As a result, the experiment was run at 18 °C in CDCl 3 , and utilizing a capillary tube for both solvents (as capillary tubes have been shown to reduce the effects of convection substantially compared to a standard 5 mm NMR tube). 2 Consistent diffusion values were verified by 1 H NMR of the compound running the non - convection - corrected experiment under the same conditions, allowing for the 19 F measurement to continue with minimal residual convection effects anticipated. 75 Table 2 . 14 Molecular volumes calculated for various cations and anions from crystal structures using Olex Software. OLEX Molecular Volume (Å 3 ) 326 3f 301 3a 68 PF 6 - 293 BPh 4 - 564 BArF 24 - 536 Al(O t BuF 9 ) 4 - 411 BArF 20 - Molecular Weight Calibration Following literature methods, the molecular weight of 3f[ PF 6 ] was probed using DOSY techniques. The internal molecular weight standards chosen for the experiment included diethyl ether, ferrocene, and tetrakis(trimethylsilyl)silane. The experiments were performed utilizing a capillary tube (2 mm) to further reduce any convection errors in the experiments and improve accuracy. 76 Table 2 . 15 Experimental data from DOSY molecular weight calibrations with 3f [PF 6 ]. Compound MW (g/mol) Diffusion Coefficient (m 2 /s *10 - 10 ) Error log MW log D 0.025 M CD 3 CN Diethyl Ether 74.12 37.64 0.09 1.869935 1.57565 Ferrocene 180.04 25.83 0.07 2.255369 1.412124 TMS 4 Si 320.84 18.24 0.05 2.506289 1.261025 Cr - fragment 542 14.47 0.04 2.734353 1.160469 0.1 M CD 3 CN Diethyl Ether 74.12 38.18 0.09 1.869935 1.581836 Ferrocene 180.04 25.73 0.07 2.255369 1.41044 TMS 4 Si 320.84 18.45 0.05 2.506289 1.265996 Cr - fragment 525 14.745 0.04 2.720234 1.168645 0.01M CD 3 CN Diethyl Ether 74.12 37.12 0.09 1.869935 1.569608 Ferrocene 180.04 26.62 0.07 2.255369 1.425208 TMS 4 Si 320.84 19.59 0.05 2.506289 1.292034 Cr - fragment 578 15.54 0.04 2.761599 1.191451 0.025 M CDCl 3 Diethyl Ether 74.12 25.35 0.75 1.869935 1.403978 Ferrocene 180.04 18.79 0.51 2.255369 1.273927 TMS 4 Si 320.84 14.33 0.54 2.506289 1.156246 Cr - fragment 694 10.93 0.53 2.841356 1.03862 0.025 M C 6 D 5 Cl Diethyl Ether 74.12 19.21 0.09 1.869935 1.283527 Ferrocene 180.04 11.83 0.07 2.255369 1.072985 TMS 4 Si 320.84 8.46 0.05 2.506289 0.92737 Cr - fragment 628 5.84 0.04 2.797865 0.766413 0.025 M CD 3 CN (45 °C) CD3CN residual 44.07 59.12 0.09 1.644143 1.771734 Ferrocene 180.04 34.15 0.07 2.255369 1.533391 TMS 4 Si 320.84 24.24 0.05 2.506289 1.384533 Cr - fragment 579 19.425 0.04 2.762472 1.288361 77 ROESY Experiments with 3a[BArF 24 ] and 3a[BPh 4 ] A Varian Inova 600 spectrometer equipped a 5 mm pulse - field - gradient (PFG) switchable broadband HCN probe, operated at 25 °C. The ROESY spectra were acquired with a spectral width of 9596 Hz in both F2 and F1. The preacquisition delay was set to 2 s. The mixing time was 0.35 ms. A total of 256 increments with 64 transients per increment were collected, containing 2878 data points. Linear prediction to 512 data points was applied to F1 prior to 2D processing. Gaussian multiplication was applied to both F1 and F2 dimensions. For both complexes examined by ROESY NMR( 3a [BArF 24] and 3a [BPh 4 ]), it appears that in CDCl 3 there are small correlation peaks between the aromatic protons of the anion (BArF 24 ¯ and BPh 4 ¯ ) and the aliphatic peaks of the phosphine and diisopropylamido methyl groups. This suggests that in a non - polar solvent, the anion is sufficiently close to the cation in solution to allow for observable NOE signal (~7Å). Such behavior supports the analysis of ion pairing in CDCl 3 , which was by DOSY and LDP analysis. The results in CD 3 CN for the two complexes show no observable correlation peaks under the same ex perimental regime as applied to the samples in CDCl 3 . While lack of signal is not a positive identifier, there is no evidence of ion pairing between the chromium fragment and the counterion. This also agrees with DOSY and LDP analyses for the compounds. Th e 3a [BArF 24 ] results are shown in above in the text. The spectra of 3a [BPh 4 ] are shown below. 78 Figure 2 . 12 1 H ROESY NMR Spectrum of 3a [BPh 4 ] in CDCl 3 . 79 Figure 2 . 13 1 H ROESY NMR Spectrum of 3a [BPh 4 ] in CD 3 CN. 80 Experimental Determinations of S Experimental measurements of entropy with the LDP system have previously been conducted by two different methods. 1 In this study, due to the temperatures at which measurement occurs, it was most practical to utilize the determination of the rate of - N i Pr 2 by Spin Saturation Transfer 1 H NMR, over a range of temperatures, and then use the Eyring equat as a function of temperature (K). Experimentally, this consists of performing the Spin Saturation Transfer 1 H NMR experiment at 4 or more different temperatures, and plotting ln(k obs /T) vs. 1/T, where k o bs is the were then derived from the slope and intercept of the Eyring plots, respectively. Standard treatment of the data was used to approximate errors in these va lues, which were relatively small for the series of measurements presented here. 2 The values determined in CD 3 CN for the 3f , 3j , and 3m salts of SbF 6 ¯ were the complexes that could be measured accessible temperature range over which the SST measurement could be performed. At low temperatures, the amide rotation is too slow to observe the rate accurately, while at higher temperatures, the samples thermal ly decompose in solution during the measurement. Similar issues were faced with the 3a - o salts with BArF 24 ¯ as the anion. Lower limits of rotation, thermal instability, and the low boiling point of the NMR solvent (CDCl 3 ) all limited the accessible experim ental temperature range for these derivatives. 81 Table 2 . 16 1 determined across several different temperatures in CD 3 CN. Temp (K) k (s - 1 ) 281.93 0.0225 18.60 0.0235 18.57 0.0230 18.58 289.49 0.0528 18.62 0.0533 18.61 0.0539 18.61 296.05 0.1098 18.62 0.1120 18.61 0.1135 18.60 302.45 0.2289 18.60 0.2227 18.61 0.2282 18.60 308.99 0.3920 18.68 0.3863 18.69 0.3879 18.69 315.32 0.8478 18.59 0.8615 18.58 0.8332 18.61 318.65 1.0545 18.66 1.0398 18.67 1.0444 18.67 323.14 1.7011 18.62 1.6799 18.63 1.6618 18.64 82 Figure 2 . 14 The Eyring plot for SST measurements of 1 in CD 3 CN. y = - 9171.9x + 23.099 R² = 0.9988 -10 -9 -8 -7 -6 -5 -4 -3 0.003 0.0031 0.0032 0.0033 0.0034 0.0035 0.0036 ln(k/T) (s - 1 K - 1 ) 1/T(K) Eyring Plot for 1 in CD 3 CN 83 Figure 2 . 15 value of Cr N i Pr 2 bond rotation in 3f [SbF 6 ]. Table 2 . 17 values for 3f [SbF 6 ]. T k 1/T ln(k/T) 1 306.91 0.39765 18.54415 0.003258 - 6.63454 2 303.87 0.333047 18.4615 0.003291 - 6.8276 3 292.46 0.131258 18.28714 0.003419 - 7.70892 4 323.29 0.912213 19.0339 0.003093 - 5.88755 Avg 306.6325 y = - 5531.5x + 11.298 R² = 0.9829 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 0.0025 0.0027 0.0029 0.0031 0.0033 0.0035 ln(k/T)(s - 1 K - 1 ) 1/T (K) Eyring Plot for 3f [SbF 6 ] in CD 3 CN 84 Table 2 . 18 Experimentally determined rate of Cr N i Pr 2 . Phosphine Compound No. Temp (K) k obs (s - 1 ) (kcal/mol) (e.u.) (kcal/mol) PPhMe 2 3f 325.21 0.5585 17.90 - 4.76 19.47 315.33 0.2395 19.39 339.07 2.0544 19.44 303.40 0.0728 19.35 PPh 2 n Bu 3j 306.91 0.3977 10.99 - 24.68 18.53 303.87 0.3330 18.47 292.46 0.1313 18.30 323.29 0.9122 19.04 P(OEt) 3 3m 308.08 0.4840 6.76 - 37.88 18.50 299.79 0.3800 18.13 291.51 0.22410 17.92 283.45 0.1762 17.54 85 Table 2 . 19 Experimentally determined rate of rotation (k obs ) for the Cr N i Pr 2 bond in 3a - p salts with BArF 24 ¯ anion in CDCl 3 . Phosphine Compound No. Temp (K) k obs (s - 1 ) (kcal/mol) (e.u.) (kcal/mol) PMe 3 3a 306.72 0.0546 18.71 - 3.38 19.74 314.28 0.1112 19.80 321.75 0.2339 19.81 331.44 0.5776 19.83 P n Bu 3 3b 316.58 0.1385 18.91 - 2.83 19.81 323.91 0.2928 19.80 328.85 0.4429 19.84 331.50 0.5650 19.85 P i Bu 3 3c 311.08 0.1229 17.76 - 5.66 19.53 317.52 0.2324 19.55 325.68 0.4635 19.62 332.79 0.8692 19.64 P i Pr 3 3d 321.19 0.0831 19.48 - 2.95 20.44 325.46 0.1305 20.42 330.03 0.1965 20.45 333.42 0.2657 20.47 PPhMe 2 3f 306.87 0.0862 18.79 - 2.13 19.47 313.81 0.1725 19.50 321.80 0.3916 19.48 329.32 0.7432 19.54 PPh 2 Me 3g 309.76 0.3302 16.29 - 8.22 18.84 321.19 0.8478 18.95 325.46 1.2066 18.99 330.03 1.8105 18.99 PPhEt 2 3h 306.80 0.1828 16.96 - 6.66 19.01 314.15 0.3635 19.05 321.99 0.6844 19.14 327.88 1.1946 19.14 PPh 2 Et 3i 306.94 0.1340 17.97 - 3.99 19.21 314.10 0.2757 19.22 321.78 0.5657 19.25 329.55 1.0834 19.30 PPh 2 n Bu 3j 302.66 0.08587 18.18 - 3.38 19.20 310.09 0.1818 19.22 316.58 0.3413 19.24 323.91 0.6663 19.27 86 Table 2.19 ( ) PPh 2 Cy 3k 312.26 0.0879 18.11 - 5.54 19.81 321.05 0.1753 19.95 328.85 0.3896 19.93 331.50 0.5007 19.93 P(OEt) 3 3m 298.42 0.1010 16.98 - 6.13 18.83 308.83 0.2896 18.86 319.22 0.7419 18.92 329.60 1.6656 19.02 P(O i Pr) 3 3n 311.08 0.2987 17.12 - 6.03 18.98 317.52 0.5025 19.06 325.68 1.0674 19.08 332.79 1.9091 19.12 P(NC 4 H 8 ) 3 3o 306.80 0.0599 19.52 - 0.52 19.69 314.15 0.1387 19.65 321.99 0.2852 19.70 327.88 0.5127 19.69 PPh 3 3p 306.87 0.1139 18.00 - 4.13 19.30 313.81 0.2493 19.27 321.80 0.4866 19.35 329.32 0.9393 19.38 87 X - ray Crystallography All crystal structures have been deposited with the Caimbridge Structural Database. The following CCDC numbers have been assigned to the structures referenced in this work 1552070 - 84 ( 2 and 3b - 3e and 3g - 3p ), 1544906 - 10 ( 3a and 3f salts with various X ¯ ). 88 NMR Spectra of 2 and 3a - 3 Figure 2 . 16 1 H NMR of 3b [SbF 6 ] in CDCl 3 . 89 Figure 2 . 17 13 C NMR of 3b [SbF 6 ] in CDCl 3 . 90 Figure 2 . 18 31 P NMR of 3b [SbF 6 ] in CDCl 3 . 91 Figure 2 . 19 19 F NMR of 3b [SbF 6 ] in CDCl 3 . 92 Figure 2 . 20 1 H NMR of 3c [SbF 6 ] in CDCl 3 . 93 Figure 2 . 21 13 C NMR of 3c [SbF 6 ] in CDCl 3. 94 Figure 2 . 22 31 P NMR of 3c [SbF 6 ] in CDCl 3 . 95 Figure 2 . 23 19 F of 3c [SbF 6 ] in CDCl 3 . 96 Figure 2 . 24 1 H NMR of 3d [SbF 6 ] in CDCl 3 . 97 Figure 2 . 25 13 C NMR of 3d [SbF 6 ] in CDCl 3 . 98 Figure 2 . 26 31 P NMR of 3d [SbF 6 ] in CDCl 3 . 99 Figure 2 . 27 19 F NMR of 3d [SbF 6 ] in CDCl 3 . 100 Figure 2 . 28 1 H NMR of 3e [SbF 6 ] in CDCl 3 . 101 Figure 2 . 29 13 C NMR of 3e [SbF 6 ] in CDCl 3 . 102 Figure 2 . 30 31 P NMR of 3e [SbF 6 ] in CDCl 3 . 103 Figure 2 . 31 19 F NMR of 3e [SbF 6 ] in CDCl 3 . 104 Figure 2 . 32 14 N NMR of 3e [SbF 6 ] in CDCl 3 . 105 Figure 2 . 33 1 H NMR of 3f [SbF 6 ] (PPhMe 2 ) in CDCl 3 . 106 Figure 2 . 34 13 C NMR of 3f [SbF 6 ] (PPhMe 2 ) in CDCl 3 . 107 Figure 2 . 35 31 P NMR of 3f [SbF 6 ] (PPhMe 2 ) in CDCl 3 . 108 Figure 2 . 36 19 F NMR of 3f [SbF 6 ] (PPhMe 2 ) in CD 3 CN. 109 Figure 2 . 37 19 F NMR of 3f [SbF 6 ] (PPhMe 2 ) in CDCl 3 . 110 Figure 2 . 38 14 N NMR of 3f [SbF 6 ] in CDCl 3 . 111 Figure 2 . 39 1 H NMR of 3g [SbF 6 ] in CDCl 3 . 112 Figure 2 . 40 13 C NMR of 3g [SbF 6 ] in CDCl 3 . 113 Figure 2 . 41 31 P NMR of 3g [SbF 6 ] in CDCl 3 . 114 Figure 2 . 42 19 F NMR of 3g [SbF 6 ] in CDCl 3 . 115 Figure 2 . 43 1 H NMR of 3h [SbF 6 ] in CDCl 3 . 116 Figure 2 . 44 13 C NMR of 3h [SbF 6 ] in CDCl 3 . 117 Figure 2 . 45 31 P NMR of 3h [SbF 6 ] in CDCl 3 . 118 Figure 2 . 46 19 F NMR of 3 h [SbF 6 ] in CDCl 3 . 119 Figure 2 . 47 1 H NMR of 3i [SbF 6 ] in CDCl 3 . 120 Figure 2 . 48 13 C NMR of 3i [SbF 6 ] in CDCl 3 . 121 Figure 2 . 49 31 P NMR of 3i [SbF 6 ] in CDCl 3 . 122 Figure 2 . 50 19 F NMR of 3i [SbF 6 ] in CDCl 3 . 123 Figure 2 . 51 1 H NMR of 3j [PF 6 ] in CDCl 3 . 124 Figure 2 . 52 13 C NMR of 3j [PF 6 ] in CDCl 3 . 125 Figure 2 . 53 31 P NMR of 3j [PF 6 ] in CDCl 3 . 126 Figure 2 . 54 19 F NMR of 3j [PF 6 ] in CDCl 3 . 127 Figure 2 . 55 14 N NMR of 3j [PF 6 ] in CDCl 3 . 128 Figure 2 . 56 1 H NMR of 3k [SbF 6 ] in CDCl 3 . 129 Figure 2 . 57 13 C NMR of 3k [SbF 6 ] in CDCl 3 . 130 Figure 2 . 58 31 P NMR of 3k [SbF 6 ] in CDCl 3 . 131 Figure 2 . 59 19 F NMR of 3k [SbF 6 ] in CDCl 3 . 132 Figure 2 . 60 1 H NMR of 3l [SbF 6 ] in CDCl 3 . 133 Figure 2 . 61 13 C NMR of 3l [SbF 6 ] in CDCl 3 . 134 Figure 2 . 62 31 P NMR of 3l [SbF 6 ] in CDCl 3 . 135 Figure 2 . 63 19 F NMR of 3l [SbF 6 ] in CDCl 3 . 136 Figure 2 . 64 14 N NMR of 3l [SbF 6 ] in CDCl 3 . 137 Figure 2 . 65 1 H NMR of 3m [SbF 6 ] in CDCl 3 . 138 Figure 2 . 66 13 C NMR of 3m [SbF 6 ] in CDCl 3 . 139 Figure 2 . 67 31 P NMR of 3m [SbF 6 ] in CDCl 3 . 140 Figure 2 . 68 19 F NMR of 3m [SbF 6 ] in CDCl 3 . 141 Figure 2 . 69 1 H NMR of 3n [SbF 6 ] in CDCl 3 . 142 Figure 2 . 70 13 C NMR of 3n[SbF 6 ] in CDCl 3 . 143 Figure 2 . 71 31 P NMR of 3n [SbF 6 ] in CDCl 3 . 144 Figure 2 . 72 19 F NMR of 3n [SbF 6 ] in CDCl 3 . 145 Figure 2 . 73 1 H NMR of 3o [PF 6 ] in CDCl 3 . 146 Figure 2 . 74 13 C NMR of 3o [PF 6 ] in CDCl 3 . 147 Figure 2 . 75 31 P NMR of 3o [PF 6 ] in CDCl 3 . 148 Figure 2 . 76 19 F NMR of 3o [PF 6 ] in CDCl 3 . 149 Figure 2 . 77 14 N NMR of 3o [PF 6 ] in CDCl 3 . 150 Figure 2 . 78 1 H NMR of 3p [BArF 24 ] in CDCl 3 . 151 Figure 2 . 79 13 C NMR of 3p [BArF 24 ] in CDCl 3 . 152 Figure 2 . 80 31 P NMR of 3p [BArF 24 ] in CDCl 3 . 153 Figure 2 . 81 19 F NMR of 3p [BArF 24 ] in CDCl 3 . 154 Figure 2 . 82 1 H NMR of 2 [SbF 6 ] in CDCl 3 . 155 Figure 2 . 83 13 C NMR of 2 [SbF 6 ] in CDCl 3 . 156 Figure 2 . 84 19 F NMR of 2 [SbF 6 ] in CDCl 3 . 157 Figure 2 . 85 14 N NMR of 2 [SbF 6 ] in CDCl 3 . 158 NMR Spectra of Additional Complexes Utilized for Ion Pairing Studies Figure 2 . 86 1 H NMR spectrum of 3f [PF 6 ] in CDCl 3 . 159 Figure 2 . 87 13 C NMR spectrum of 3f [PF 6 ] in CDCl 3 . 160 Figure 2 . 88 19 F NMR spectrum of 3f [PF 6 ] in CDCl 3 161 Figure 2 . 89 31 P NMR of 3f [PF 6 ] in CDCl 3 . 162 Figure 2 . 90 14 N NMR spectrum of 3f [PF 6 ]. 163 Figure 2 . 91 1 H NMR of 3f [BArF 24 ] in CDCl 3 . 164 Figure 2 . 92 13 C NMR spectrum of 3f [BArF 24 ] in CDCl 3 . 165 Figure 2 . 93 31 P NMR spectrum of 3f [BArF 24 ] in CDCl 3 . 166 Figure 2 . 94 19 F NMR spectrum of 3f [BArF 24 ] in CDCl 3 . 167 Figure 2 . 95 14 N NMR spectrum of 3f [BArF 24 ] in CDCl 3 . 168 Figure 2 . 96 1 H NMR spectrum of 3f [BArF 20 ] in CDCl 3 . 169 Figure 2 . 97 13 C NMR spectrum of 3f [BArF 20 ] in CDCl 3 . 170 Figure 2 . 98 31 P NMR spectrum of 3f [BArF 20 ] in CDCl 3 . 171 Figure 2 . 99 19 F NMR spectrum of 3f [BArF 20 ] in CDCl 3 . 172 Figure 2 . 100 14 N NMR spectrum of 3f [BArF 20 ] in CDCl 3 . 173 Figure 2 . 101 1 H NMR spectrum of 3f [Al(OC(CF 3 ) 3 ) 4 ] in CDCl 3 . 174 Figure 2 . 102 13 C NMR spectrum of 3f [Al(OC(CF 3 ) 3 ) 4 ] in CDCl 3 . 175 Figure 2 . 103 31 P NMR spectrum o f 3f [Al(OC(CF 3 ) 3 ) 4 ] in CDCl 3 . 176 Figure 2 . 104 19 F NMR spectrum of 3f [Al(OC(CF 3 ) 3 ) 4 ] in CDCl 3 . 177 Figure 2 . 105 14 N NMR spectrum of 3f [Al(OC(CF 3 ) 3 ) 4 ] in CDCl 3 . 178 Figure 2 . 106 1 H NMR spectrum of 3f [BPh 4 ] in CDCl 3 . 179 Figure 2 . 107 13 C NMR spectrum of 3f [BPh 4 ] in CDCl 3 . 180 Figure 2 . 108 31 P NMR spectrum of 3f [BPh 4 ] in CDCl 3 . 181 Figure 2 . 109 14 N NMR spectrum of 3f [BPh 4 ] in CDCl 3 . 182 Figure 2 . 110 1 H NMR spectrum of 3a [BArF 24 ] in CDCl 3 . 183 Figure 2 . 111 13 C NMR spectrum of 3a [BArF 24 ] in CDCl 3 . 184 Figure 2 . 112 31 P N MR spectrum of 3a [BArF 24 ] in CDCl 3 . 185 Figure 2 . 113 19 F NMR spectrum of 3a [BArF 24 ] in CDCl 3 . 186 Figure 2 . 114 14 N NMR spectrum of 3a [BArF 24 ] in CDCl 3 . 187 Figure 2 . 115 1 H NMR spectrum of 3a [BPh 4 ] in CDCl 3 . 188 Figure 2 . 116 13 C NMR spectrum of 3a [BPh 4 ] in CDCl 3 . 189 Figure 2 . 117 31 P NMR spectrum of 3a [BPh 4 ] in CDCl 3 . 190 Figure 2 . 118 14 N NMR spectrum of 3a [BPh 4 ] in CDCl 3 . 191 REFERENCES 192 REFERENCES ( 1 ) DiFranco, S. A.; Maciulis, N. A.; Staples, R. J.; Batrice, R. J.; Odom, A. L., Evaluation of donor and steric properties of anionic ligands on high valent transition metals. Inorganic chemistry 2012, 51 (2), 1187 - 200. ( 2 ) Billow, B. S.; Bemowski, R. D. ; DiFranco, S. A.; Staples, R. J.; Odom, A. L., Synthesis and Structure of Chromium(VI) Nitrido Cyclopentadienyl Complexes. Organometallics 2015 . ( 3 ) Bemowski, R. D.; Singh, A. K.; Bajorek, B. J.; DePorre, Y.; Odom, A. L., Effective donor abilities of E - t - Bu and EPh (E = O, S, Se, Te) to a high valent transition metal. Dalton transactions 2014, 43 (32), 12299 - 12305. ( 4 ) Beaumier, E. P.; Billow, B. S.; Singh, A. K.; Biros, S. M.; Odom, A. L., A complex with nitrogen single, double, and triple bonds t o the same chromium atom: synthesis, structure, and reactivity. Chemical Science 2016 . ( 5 ) Billow, B. S.; McDaniel, T. J.; Odom, A. L., Quantifying ligand effects in high - oxidation - state metal catalysis. Nature Chemistry 2017, 9 , 837. ( 6 ) Dias, P. B.; de Piedade, M. E. M.; Simões, J. A. M., Bonding and energetics of phosphorus (III) ligands in transition metal complexes. Coordination Chemistry Reviews 1994, 135 136 (0), 737 - 807. ( 7 ) Levason, W., Phosphine complexes of transition metals. In Organophosphorus Compounds (1990) , Hartley, F. R., Ed. 1990. ( 8 ) Odom, A. L.; McDaniel, T. J., Titanium - Catalyzed Multicomponent Couplings: Efficient One - Pot Syntheses of Nitrogen Heterocycles. Acco unts of Chemical Research 2015, 48 (11), 2822 - 2833. ( 9 ) Porter, J. R.; Wirschun, W. G.; Kuntz, K. W.; Snapper, M. L.; Hoveyda, A. H., Ti - Catalyzed Regio - - Amino Nitriles, Amides, and Acids. Catalyst Identif ication through Screening of Parallel Libraries. Journal of the American Chemical Society 2000, 122 (11), 2657 - 2658. ( 10 ) Odom, A. L., New C - N and C - C bond forming reactions catalyzed by titanium complexes. Dalton transactions 2005, (2), 225 - 233. ( 11 ) Tol man, C. A., Electron donor - acceptor properties of phosphorus ligands. Substituent additivity. Journal of the American Chemical Society 1970, 92 (10), 2953 - 2956. ( 12 ) Tolman, C. A., Steric effects of phosphorus ligands in organometallic chemistry and homoge neous catalysis. Chemical Reviews 1977, 77 (3), 313 - 348. 193 ( 13 ) Tsuchida, R., Absorption Spectra of Co - ordination Compounds. I. Bull. Chem. Soc. Jpn. 1938, 13 (5), 388 - 400. ( 14 ) Wulfsberg, G., Inorganic Chemistry. 1987 . ( 15 ) Drago, R. S.; Dadmun, A. P., Dual - parameter substituent constants, .DELTA.Ex and .DELTA.Cx, for the correlation of physicochemical measurements. Journal of the American Chemical Society 1993, 115 (19), 8592 - 8602. ( 16 ) Vogel, G. C.; Drago, R. S., The ECW Model. Journal of Chemical Education 1996, 73 (8), 701. ( 17 ) Suresh, C. H.; Koga, N., Quantifying the Electronic Effect of Substituted Phosphine Ligands via Molecular Electrostatic Potential. Inorganic chemistry 2002, 41 (6), 1573 - 1578. ( 18 ) Allman, T.; Goel, R. G., The basicity of phosphines. Canadian Journal of Chemistry 1982, 60 (6), 716 - 722. ( 19 ) Golovin, M. N.; Rahman, M.; Belmonte, J. E.; Giering, W. P., QUANTITATIVE SEPARATION OF SIGMA - COMPONENTS AND PI - COMPONENT S OF TRANSITION - METAL PHOSPHORUS BONDING AND THE APPLICATION OF LIGAND EFFECTS IN ORGANOMETALLIC CHEMISTRY. Organometallics 1985, 4 (11), 1981 - 1991. ( 20 ) Wilson, M. R.; Woska, D. C.; Prock, A.; Giering, W. P., THE QUANTITATIVE - ANALYSIS OF LIGAND EFFECTS (QALE) - THE ARYL EFFECT. Organometallics 1993, 12 (5), 1742 - 1752. ( 21 ) Poe, A. J., The tuning of P - donor ligands: the aryl and other pendent group effects (PGEs) revisited. Dalton transactions 2009, (11), 1999 - 2003. ( 22 ) Rahman, M. M.; Liu, H. Y.; Proc k, A.; Giering, W. P., QUANTITATIVE - ANALYSIS OF LIGAND EFFECTS (QALE) .2. STERIC AND ELECTRONIC FACTORS INFLUENCING TRANSITION - METAL PHOSPHORUS(III) BONDING. Organometallics 1987, 6 (3), 650 - 658. ( 23 ) Nelson, D. J.; Nolan, S. P., Quantifying and understand ing the electronic properties of N - heterocyclic carbenes. Chemical Society Reviews 2013, 42 (16), 6723 - 6753. ( 24 ) Setiawan, D.; Kalescky, R.; Kraka, E.; Cremer, D., Direct Measure of Metal Ligand Bonding Replacing the Tolman Electronic Parameter. Inorganic chemistry 2016 . ( 25 ) Sanderson, R. T., Principles of electronegativity Part I. General nature. Journal of Chemi cal Education 1988, 65 (2), 112. ( 26 ) Davis, M. F.; Clarke, M.; Levason, W.; Reid, G.; Webster, M., Tin(IV) Fluoride Complexes with Tertiary Phosphane Ligands A Comparison of Hard and Soft Donor Ligands. 2006, 2006 (14), 2773 - 2782. ( 27 ) Mason, J., Mul tinuclear NMR . Plenum Press: New York [etc.], 1987. 194 ( 28 ) Mason, J., Nitrogen nuclear magnetic resonance spectroscopy in inorganic, organometallic, and bioinorganic chemistry. Chemical Reviews 1981, 81 (3), 205 - 227. ( 29 ) Marcus, Y.; Hefter, G., Ion Pairing. Chemical Reviews 2006, 106 (11), 4585 - 4621. ( 30 ) Hefter, G., When spectroscopy fails: The measurement of ion pairing. In Pure and Applied Chemistry , 2006; Vol. 78, p 1571. ( 31 ) Fuoss, R. M., Conductance - concentration function for the paired ion model. The Journal of Physical Chemistry 1978, 82 (22), 2427 - 2440. ( 32 ) Kameda, Y.; Ebina, S.; Amo, Y.; Usuki, T.; Otomo, T., Microscopic Structure of Contact Ion Pairs in Concentrated LiCl - and LiClO4 - Tetrahydrofuran Solutions Studied by Low - Frequency Isotropic Raman Scattering and Neutron Diffraction with 6Li/7Li Isotopic Substitution Methods. The Journal of Physical Chemistry B 2016, 120 (20), 4668 - 4678. ( 33 ) Diaz - Torres, R.; Alvarez, S., Coordinating ability of anions and solvents towards transition metals and lanthanides. Dalton transactions 2011, 40 (40), 10742 - 10750. ( 34 ) Pregosin Paul, S., NMR spectroscopy and ion pairing: Measuring and understanding how ions interact. In Pure and Applied Chemistry , 2009; Vol. 81, p 615. ( 35 ) Pregosin, P. S., NMR diffusion methods in inorganic and organometallic chemistry. In Spectros copic Properties of Inorganic and Organometallic Compounds: Techniques, Materials and Applications, Volume 42 , The Royal Society of Chemistry: 2012; Vol. 42, pp 248 - 268. ( 36 ) Fernández, I.; Martínez - Viviente, E.; Pregosin, P. S., Multinuclear PGSE Diffusi on and Overhauser NMR Studies on a Variety of Salts in THF Solution. Inorganic chemistry 2005, 44 (15), 5509 - 5513. ( 37 ) Pregosin, P. S., Applications of NMR diffusion methods with emphasis on ion pairing in inorganic chemistry: a mini - review. Magnetic Reso nance in Chemistry 2016 , n/a - n/a. ( 38 ) Macchioni, A., Ion Pairing in Transition - Metal Organometallic Chemistry. Chemical Reviews 2005, 105 (6), 2039 - 2074. ( 39 ) Brand, T.; Cabrita, E. J.; Berger, S., Intermolecular interaction as investigated by NOE and di ffusion studies. Progress in Nuclear Magnetic Resonance Spectroscopy 2005, 46 (4), 159 - 196. ( 40 ) Bellachioma, G.; Ciancaleoni, G.; Zuccaccia, C.; Zuccaccia, D.; Macchioni, A., NMR investigation of non - covalent aggregation of coordination compounds rangi ng from dimers and ion pairs up to nano - aggregates. Coordination Chemistry Reviews 2008, 252 (21 22), 2224 - 2238. ( 41 ) Macchioni, A.; Ciancaleoni, G.; Zuccaccia, C.; Zuccaccia, D., Determining accurate molecular sizes in solution through NMR diffusion spectroscopy. Chemical Society Reviews 2008, 37 (3), 479 - 489. 195 ( 42 ) Chen, H. C.; Chen, S. H., Diffusion of crown ethers in alcohols. The Journal of Physical Chemistry 1984, 88 (21), 5118 - 5121. ( 43 ) Honeychuck, R. V.; Hersh, W. H., Coordination of "noncoordinating" anions: synthesis, characterization, and x - ray crystal structures of fluorine - bridged hexafluoroantimonate(1 - ), t etrafluoroborate(1 - ), and hexafluorophosphate(1 - ) adducts of [R3P(CO)3(NO)W]+. An unconventional order of anion donor strength. Inorganic chemistry 1989, 28 (14), 2869 - 2886. ( 44 ) Metz, M. V.; Sun, Y.; Stern, C. L.; Marks, T. J., Weakly Coordinating Al - , Nb - , Ta - , Y - , and La - Based Perfluoroaryloxymetalate Anions as Cocatalyst Components for Single - Site Olefin Polymerization. Organometallics 2002, 21 (18), 3691 - 3702. ( 45 ) Hamilton, G. L.; Kang, E. J.; Mba, M.; Toste, F. D., A Powerful Chiral Counterion St rategy for Asymmetric Transition Metal Catalysis. Science 2007, 317 (5837), 496. ( 46 ) Brak, K.; Jacobsen, E. N., Asymmetric Ion - Pairing Catalysis. Angewandte Chemie International Edition 2013, 52 (2), 534 - 561. ( 47 ) Ohmatsu, K.; Ito, M.; Kunieda, T.; Ooi, T., Ion - paired chiral ligands for asymmetric palladium catalysis. Nature Chemistry 2012, 4 , 473. ( 48 ) Claridge, T., High Resolution NMR Spectroscopy in Organic Chemistry. Elsevier 2008 , Chpt 8. ( 49 ) Lane, E. M.; Chapp, T. W.; Hughes, R. P.; Glueck, D. S.; Feland, B. C.; Bernard, G. M.; Wasylishen, R. E.; Rheingold, A. L., Synthesis of Gold Phosphido Complexes Derived from Bis(secondary) Phosphines. Structure of Tetrameric [Au(MesP(CH2)3PMes)Au]4 . Inorganic chemistry 2010, 49 (8), 3950 - 3957. ( 50 ) Li, D.; Kagan, G.; Hopson, R.; Williard, P. G., Formula Weight Prediction by Internal Reference Diffusion - Ordered NMR Spectroscopy (DOSY). Journal of the American Chemical Society 2009, 131 (15), 5627 - 5 634. ( 51 ) Li, D.; Keresztes, I.; Hopson, R.; Williard, P. G., Characterization of Reactive Intermediates by Multinuclear Diffusion - Ordered NMR Spectroscopy (DOSY). Accounts of Chemical Research 2009, 42 (2), 270 - 280. ( 52 ) Neufeld, R.; Stalke, D., Accurat e molecular weight determination of small molecules via DOSY - NMR by using external calibration curves with normalized diffusion coefficients. Chemical Science 2015, 6 (6), 3354 - 3364. ( 53 ) Rocchigiani, L.; Macchioni, A., Disclosing the multi - faceted world o f weakly interacting inorganic systems by means of NMR spectroscopy. Dalton transactions 2016, 45 (7), 2785 - 2790. ( 54 ) Clot, E., Ion - Pairing in Organometallic Chemistry: Structure and Influence on Proton Transfer from a Computational Perspective. 2009, 200 9 (16), 2319 - 2328. 196 ( 55 ) Leskowitz, G. M.; Ghaderi, N.; Olsen, R. A.; Pederson, K.; Hatcher, M. E.; Mueller, L. The Journal of Physical Chemistry A 2005, 109 (6), 1 152 - 1158. ( 56 ) Schneider, A.; Hommel, G.; Blettner, M., Linear regression analysis: part 14 of a series on evaluation of scientific publications. Deutsches Arzteblatt international 2010, 107 (44), 776 - 782. ( 57 ) Odom, A. L.; Cummins, C. C.; Protasiewicz, J. D., Nitric Oxide Cleavage: Synthesis of Terminal Chromium(VI) Nitrido Complexes via Nitrosyl Deoxygenation. Journal of the American Chemical Society 1995, 117 (24), 6613 - 6614. ( 58 ) Park, J. G.; Jeon, I. - R.; H arris, T. D., Electronic Effects of Ligand Substitution on Spin Crossover in a Series of Diiminoquinonoid - Bridged FeII2 Complexes. Inorganic chemistry 2015, 54 (1), 359 - 369. ( 59 ) Krossing, I., The Facile Preparation of Weakly Coordinating Anions: Structure and Characterisation of Silverpolyfluoroalkoxyaluminates AgAl(ORF)4, Calculation of the Alkoxide Ion Affinity. Chemistry A European Journal 2001, 7 (2), 490 - 502. ( 60 ) Wiesler, W. T.; Caruthers, M. H., Synthesis of Phosphorodithioate DNA via Sulfur - Linke d, Base - Labile Protecting Groups1. The Journal of Organic Chemistry 1996, 61 (13), 4272 - 4281. ( 61 ) Kermagoret, A.; Braunstein, P., SHOP - type nickel complexes with alkyl substituents on phosphorus, synthesis and catalytic ethylene oligomerization. Dalton tr ansactions 2008, (6), 822 - 831. ( 62 ) Power, J. E.; Foroozandeh, M.; Moutzouri, P.; Adams, R. W.; Nilsson, M.; Coombes, S. R.; Phillips, A. R.; Morris, G. A., Very broadband diffusion - ordered NMR spectroscopy: 19F DOSY. Chemical communications 2016, 5 2 (42), 6892 - 6894. 197 CHAPTER 3. ANALYSIS OF THE DONOR ABILITIES OF PHOSPHINES TO HIGH VALENT METALS 3.1 Introduction 3 , 4 In the previous chapter, several of the complications created by using neutral ligands in the LDP system were presented. Ion pairing effects in the [NCr(N i Pr 2 ) 2 PR 3 ][X] salts in nonpolar solvents artificially increase the measured barrier to Cr N i Pr 2 bond rotations. This effect is not systematic, which causes variable increases in the experimentally determined donor ability, precluding a n easy correction to the H values. Conversely, when the rotation barrier of the Cr N i Pr 2 bond in these ionic complexes are examined in a more polar solvent system (CD 3 CN) the ion pairing effect is reduced. This is reflected in the reduction of the measur ed LDP values in CD 3 CN versus CDCl 3 . The polar solvent disrupts the ion pairing in the [NCr(N i Pr 2 ) 2 PR 3 ][X] salts, removing the steric and electrostatic inhibition of Cr N i Pr 2 bond rotation by the anion. However, ion pair disruption is accomplished at the e xpense of the consistency associated with the entropy of activation, S , when the LDP complex is neutral (NCr(N i Pr 2 ) 2 X) or an ion paired salt. 1 Due to the instability of the [NCr(N i Pr 2 ) 2 PR 3 ][X] salts and experimental limitations cannot be determined for each complex in the 3a - p series (Fig. 2.3). Even if these measurements we re experimentally feasible, it is not clear what exactly the activation energies mean in these measurements where the ions are unpaired and solvation appears to be the controlling factor. 3 The work presented in this chapter was done in collaboration with Dr. Brennan Billow. 4 The results presented in this chapter have been published in the following article: Aldrich, K. E., Billow, B. S . , Staples, R. J., Odom, A. L., ( 2019 ). "Phosphine int eractions with high oxidation state metals." Polyhedron 159 : 284 - 297. 198 Thus, while we had collected two complete sets of LDP values, one in CD 3 CN with the SbF 6 ¯ counter anion and the other in CDCl 3 with the BArF 24 ¯ counter anion, we were uncertain exactly what the LDP values meant. In CDCl 3 , we suspected that the ion pairing interference could cause the measurements to be errant by up to 0. 7 kcal/mol (See Chapter 2), but because these 3 CN, several of the measured PR 3 LDP values vary with temperature and our investigations into the entropy of activation for several of these varies greatly in this system on a case - by - case basis. For reference, both sets of LDP values are shown in Table 3.1. 199 Table 3 . 1 values for 3a - p series of compounds in CDCl 3 with BArF 24 ¯ counter anion and by the original method in CD 3 CN with SbF 6 ¯ . Compound Number Phosphine Complex (kcal/mol) CDCl 3 (e.u.) CDCl 3 (kcal/mol) c CD 3 CN 3a PMe 3 18.71 - 3.4 (1.0) 16.64 3b P n Bu 3 18.91 - 2.8 (1.0) 16.77 3c P i Bu 3 17.76 - 5.7 (1.0) 17.13 3d P i Pr3 19.47 - 3.0 (1.5) 17.17 3e PCy 3 19.46 - a 17.27 3f PPhMe 2 18.79 - 2.1 (1.1) 16.53 3g PPh 2 Me 16.9 - 8.2 (1.2) 16.16 3h PPhEt 2 16.96 - 6.7 (1.4) 16.65 3i PPh 2 Et 17.98 - 4.0 (1.0) 16.15 3j PPh 2 n Bu 18.17 - 3.4 (1.0) 16.31 3k PPh 2 Cy 18.11 - 5.5 (3.2) 16.43 3l PPhCy 2 - - 16.37 3m P(OEt) 3 16.99 - 6.1 (1.3) 15.73 3n P(OiPr) 3 17.12 - 6.0 (1.2) 15.91 3o P(NC 4 H 8 ) 3 19.52 - 1.0 (1.5) 16.21 3p PPh 3 18 - 4.1 (2.1) - b Average - - 4.7 - a Measurable rotation rate did not occur until ~55 ° temperatures could not be performed. b 3p decomposes to 2 and PPh 3 in the presence of acetonitrile so an LDP of the complex in CD 3 CN could not be determined. c values determined in the typical manner of LDP, was set equal to - 9 e.u. without focusing on the exact values, two major trends struck us as significant. First, adding more phenyl substit uents to a phosphine appears to make the phosphine a stronger donor. This trend is contrary to the steric effects we would predict in the system, as adding phenyl groups to PR 3 ligands makes them bulkier and more rigid, which should raise the LDP value. 1, 2 In terms of electronics this, is also contra ry (vide supra), as phenyl groups are typically regarded as electron withdrawing. Second, the P(OR) 3 - acids with low valent metals (i.e. accept electron density - backbonding), are two of the best donors in eit her system. 200 Additionally, if we consider these trends in light of the three [NCr(N i Pr 2 ) 2 PR 3 ][SbF 6 ] was determined experimentally, in the absence of ion pairing effects 3f , 3j , and 3m the relative ordering of these LDP values remain s unchanged in all 3 systems. These results are shown in Table 3.2 determination, the value for 3m is lowest, meaning it is the best donor, and the value for 3f is the highest, indicating that it is the worst overall donor. Again, we hesitated to put too much confidence in the LDP values in Table 3.1 or 3.2, because we knew that complications across both sets of values may interfere with accurate analysis and su bsequent interpretations of bonding interactions. However, if we consider specific values in values, with CDCl 3 and X ¯ = BArF 24 ¯ , several direct comparisons make these perceived trends seem substantial and worth further con sideration. values for 3a and 3m are 18.71 and 16.99 kcal/mol, respectively. The phosphines are roughly the same size (vide supra) so the primary difference in their donor abilities should be electronic. Despite the fact that PMe 3 i - electron - rich than P(OEt) 3 , P(OEt) 3 is a far better donor according to these values. 3 Even if we consider an exaggerated scenario, and say the for P(OEt) 3 ( 3m ) is unaffected by ion pairing but that the value for PMe 3 ( 3a ) is raised artificially by up to 1 kcal/mol, there is still a substantial difference between the two values (0.7 kcal/mol), which suggests P(OEt) 3 is a better donor in our h igh valent system. Taking a more realistic approach to the likely effects that ion pairing with BArF 24 ¯ measurement, both values are likely increased by 0.1 - 0.5 kcal/mol with this non - site - specific ion pairing counter anion (refer to Chapte how much better P(OEt) 3 is than PMe 3 in terms of donor ability, the 1.7 kcal/mol margin seems significant enough that the trend is real. 201 Table 3 . 2 values determined for 3f , 3j , and 3m in both solvent/ion regimes and with the standard and values. PR 3 Complex H (CD 3 CN, SbF 6 ¯ ) a S (CD 3 CN, SbF 6 ¯ ) b H (CD 3 CN, SbF 6 ¯ ) c S (CD 3 CN, SbF 6 ¯ ) c H (CDCl 3 , BArF 24 ¯ ) c S (CDCl 3 , BArF 24 ¯ ) c PPhMe 2 3f 16.53 - 9 17.90 - 5 18.79 - 2.1 PPh 2 n Bu 3j 16.31 - 9 10.99 - 25 18.17 - 3.4 P(OEt) 3 3m 15.73 - 9 6.76 - 38 16.99 - 6.1 a = - 9 e.u. b S determined from NCr(N i Pr 2 ) 2 X system. c The fully experimentally determined enthalpic and entropic parameters for activation derived from Eyring plots with the e xperimental rates of rotation. 3.2 A Comparison of Traditional Phosphine Characteristics from Low - Valent Systems with LDP Results As mentioned briefly in chapter 2, PR 3 ligands are some of the most well - understood ligands used in transition metal chemistry. Their characterization as ligands is primarily based on an electronic parameter and a steric parameter. This concept was first presented in a systematic the Tolman Ni 0 (CO) 3 PR 3 complex (or representative model for sterics). 4 - 6 Figure 3 . 1 ( left ) The Ni(CO) 3 PR 3 complex used in to determine the TEP. ( right ) The model used to measure the Tolman Cone Angle of a given phosphine. The spheres represent a va riety of R groups, and the P center and block are 2.28 Å apart. 1 CO stretching frequency. Donation from the PR 3 group increases the electron density at the already electron rich Ni 0 metal center; the Ni 0 - backbonding interactions between Ni and the CO ligands, elongating the CO bonds and decreasing the frequency of the CO stretches. By 202 comparison, a less donating PR 3 ligand exhibits less backbonding, less C O bond elongation, and to higher frequency CO stretches. This property is readily observed and serves as a direct indicator of the donor ability of PR 3 ligands. 4 The steric parameter was based on CPK models (spacefilling) in which the P atom of a given PR 3 ligand was centered on a pin above a pivot 2.85 cm (scaled from 2.28 Å). A protractor was used to determine the angle subtended throughout the rotation of the ligand around the pin. lity in the PR 3 model, approximations that were generally reproducible within a few degrees. 5, 6 ameterization, IR spectroscopy has advanced, as has the treatment of discrete electronic effects. This has led - - electronic contrib utions as well as other expansions in the way we classify phosphines as electron donors. 7 - 21 Despite these criticisms and alterations , the utility of this system of ligand parameterization is hard to dispute considering the number of citations for both the TEC and TCA. 5 In fact, the TCA is still widely used for a relative comparison of size, and the use of alternate steric descriptors ( i.e. %V bur or solid G) mostly stems from the specifics of the complex used and/or personal preferences. 9, 22 Again, these parameterizations have been accomplished using low valent metals (Ni, Pd, Fe 0/+2 , etc) in mind. However, to determine whether these PR 3 ligands are behaving the same way in our high valent system as they do in these low valent systems in the literature, these parameters serve as a useful tool. If there are no differences in the M P interaction caused by the valency of 5 References 2 - 4 have over 3,000 citations. 203 the metal, the stereoelectronic parameters used to describe PR 3 ligands in low valent systems should also describe the interactions in high valent systems. In our high valent Cr(VI) system with low (~C s ) symmetry, we know that bot orbitals participate and are heavily mixed. 1 Therefore, for comparison, we wanted to consider a system in which these two parameters are separated so we could distinguish the role of each of these electronic factors separately in the h separately, was selected for comparison purposes between low and high valent interactions. These values for the PR 3 ligands across the series of 3a - p complexes are listed in Table 3.3, below. 204 Table 3 . 3 d s 3, 15, 23 Compound Number PR 3 d Aryl ° ) H (CDCl 3 ) H (CD 3 CN) 3a PMe 3 8.55 0 0 118 18.71 16.64 3b P n Bu 3 5.25 0 0 136 18.91 16.77 3c P i Bu 3 5.7 0 0 143 17.76 17.13 3d P i Pr 3 3.45 0 0 160 19.47 17.17 3e PCy 3 1.4 0 0 170 19.46 17.27 3f PPhMe 2 10.5 0 1.0 122 18.79 16.53 (17.90) 3g PPh 2 Me 12.6 0 2.2 136 16.90 16.16 3h PPhEt 2 8.6 0 1.1 136 16.96 16.65 3i PPh 2 Et 11.1 0 2.3 140 17.98 16.15 3j PPh 2 n Bu 11.3 0 2.1 143 18.17 16.31 (10.99) 3k PPh 2 Cy 9.1 0 1.6 153 18.11 16.43 3l PPhCy 2 5.7 0 1.6 162 - 16.37 3m P(OEt) 3 15.8 2.9 1.1 109 16.99 15.73 (6.76) 3n P(O i Pr) 3 13.4 2.9 1.3 130 17.12 15.91 3o P(NC 4 H 8 ) 3 - 1.4 0.9 - 0.6 146 19.52 16.21 3p PPh 3 13.25 0 2.7 145 18.00 - It is important to clarify that, like the - d , is based on the donor ability of P t Bu 3 as the zero point reference. 4, 12 - d value correlates with a less donating phosphine. Thus, by these parameters, of the PR 3 ligands listed above, P(NC 4 H 8 ) 3 is the most donating while P(OEt) 3 is the least donating. or electron withdrawing effects in low valent systems. Considering these parameters from the literature, which specifically describe traditional low - valent M abilities) of ligan ds such as aryl phosphines and especially the phosphites ( 3m and 3n ) relative to the trialkyls ( 3a - e ) are so surprising. Both types of PR 3 donors have electron withdrawing R groups - f we consider the manner of the bonding involved in an aryl or - OR substituted PR 3 205 - - acidic PR 3 , a fil led d - orbital on the metal pushes electron density into an appropriately oriented P in Fig. 3.2 and leads to the resonance form dativ P bond in a low valent complex. 21, 24 - 30 Figure 3 . 2 Resonance forms that contribute to the ground state electronic structure of traditional low - valent metal - phosphorous interactions. Presumably, the interaction is heavily diminished in high valent M P bonds, especially when the metal is d 0 . Formally, any participation of in the electronic structure of a high valent M P bond is populated by donation from or mixing with an M X bonding orbital in these complexes (where X is another ligand). 31 Where symmetry or energy discrepa ncies prevent such a bonding - to - antibonding orbital interaction, the participation of would be essentially 0%. - - a cidic PR 3 ligands. Based on the experimental evidence, it seemed likely that some other interaction, not readily observed with low valent metals, was leading to an enhancement of these PR 3 206 Considering the fundamental properties o f low and high valent metals, their inherent Lewis acid - base characteristics are markedly different. Low valent metals are electron rich and can delocalize their d (and s) orbital electron density onto their ligands. High valent metals, on the other hand, are typically regarded as Lewis acids. In fact, simple molecular compounds of high valent transition metals (i.e. TiCl 4 ) are utilized for Lewis acid catalyzed (mediated) organic transformations. 31 - 33 The exact Lewis acidity or basicity of highly dissimilar compounds is a hard comparison to make outside of an experimental handle such as pK a or ionization potentials. How ever, this gave us the idea that perhaps in the M P bonds with high valent metals there is a role reversal observed relative to low valent M P bonds. Specifically, we suspected that this - interactions. With a high valent m etal, lacking electron density in the s and d - acid) and the PR 3 ligand may act as - base). If this were the case, a PR 3 ligand such as P(OEt) 3 , which has electronegative - electron density (oxygen lone pairs). - electron density. With an electron deficient metal bound to P, these substituents, which act as electron - withdrawing groups in low v alent interactions, may act as electron donating groups, through simple negative hyperconjugation interactions where the metal is the recipient of additional electron density. Such an interaction could strongly enhance the donor ability of such ligands, de - lone pair of electrons. Based on this idea, we decided to pursue computational investigations that might elucidate the electronic interactions of these ligands with a high valent metal. Before we got carried away wit h interpreting calculations, we sought experimental evidence to support our hypothesis that the Cr(VI) metal center utilized in the LDP system is in fact a strong Lewis acid. A piece of experimental evidence supporting the proposed inversion of 207 the Lewis a cid - base dynamic with PR 3 ligands one free of the solvent effects impacting the LDP measurements seemed necessary in light of our losing fight against ion pairing and unpredictable values. It would unify all of the qualitative pieces of evidence that w e had collected while probing the in situ dynamics of the 3a - p salts. With this motivation looked for a way to determine the Lewis acidity of the Cr(VI) in our system with a quantitative experimental technique. To determine the Lewis acidity of [NCr(N i Pr 2 ) 2 ] + a convenient method for probing the Lewis acidity of various species in solution, which utilizes 31 P NMR, was pursued. This method, known as the Gutmann Parameter or Gutmann - Beckett Method, requires the synthesis of the tri(ethyl)phosphine oxide adduc t of the Lewis acid in question. 34 - 38 Upon binding through t he oxygen of the O=P(Et) 3 , inductive effects caused by the donation of electron density from the O to the Lewis acidic species produce a shift in the 31 P NMR signal for the O=P(Et) 3 . Therefore, when the O=P(Et) 3 is bound to a Lewis acid, the 31 P signal shi fts downfield; the extent of this shift can be related to the Lewis acidity by the following equation (Eq. 3. 1). ( Eq. 3. 1 ) In this equation AN, or acceptor number, is the value indicating Lewis acidity (aka Gutmann parameter). The higher the value, the more Lewis acidic a species is. Since the system was originally designed to probe the Lewis acidity of various solvents, the 3 1 P NMR shift of 41.0, which is the value observed for O=P(Et) 3 n - hexane, serves as the zero - point reference. The coefficient of 2.21 scales the 31 P NMR shifts such that the AN for SbCl 5 is set to acids is listed in Table 3.4 with Fig. 3.3 . 208 Table 3 . 4 AN Values indicating the Lewis acidity of several compounds determined by the Guttman - Beckett Metho d. 37 a Measured in DCE, referenced to External O=P(Et) 3 as 41.0 ppm. Compound 31 AN B(C 6 F 5 ) 3 78.0 82 BF 3 ·Et 2 O 80.9 88.5 SbCl 5 86.1 100 BCl 3 88.7 106 BBr 3 90.3 109 BI 3 92.9 115 TiCl 4 72.7 70 AlCl 3 80.3 87 NCr(N i Pr 2 ) 2 + 86.2 100 a Figure 3 . 3 The synthetic scheme used to generate the O=P(Et) 3 adduct with [NCr(N i Pr 2 ) 2 ] + . A high yield of the desired complex was isolated after recrystallization. ( right ) Preliminary crystal structure of [NCr(N i Pr 2 ) 2 (OP(Et) 3 )][BArF 24 ]; note, thermal ellipsoids are not shown due to the severe disorder in the structure. It provided only connectivity. Using this technique, we were able to synthesize and isolate the [NCr(N i Pr 2 ) 2 (O=P(Et) 3 )][BArF 24 ] salt and determine the 31 P NMR shift for the bound phosphine oxide species. The shift of 86.2 ppm correlates to an AN of 100, which is the same as the highly acidic SbCl 5 . The experimental verification of such strongly Lewis acidic character from the [NCr(N i Pr 2 ) 2 ] + fragment offers support for the following computational results. The relative elect ron deficiency of the metal is substantial enough that it is one of the primary factors determining the bonding interactions between the Cr(VI) and its ligands. 3.3 Computational Analysis of the Electronic Structure of [NCr(N i Pr 2 ) 2 PR 3 ] + Using Natural Bonding Orbital and Natural Resonance Theory Calculations to examine the details of the bonding interactions between Cr and the PR 3 purposes of our studies, we found it nece ssary to abridge the N i Pr 2 ligands, substituting the i Pr - 209 groups for hydrogens. 6 Similarly, two of the PR 3 ligand R groups ( 3m* and 3o* ) were slightly abridged for ease of optimization of the structures. With 3m* and 3o* , the extended alkyl portions of the R groups on the PR 3 ligand caused difficulties in geometry optimization, so they were appended to smaller versions of electronically s imilar PR 3 ligands. In all cases, the first coordination sphere of the Cr and the atoms bound directly to P were unaltered. Our model complexes followed the form [NCr(NH 2 ) 2 PR 3 ] + . To achieve diversity of the R groups in the PR 3 ligands, the following model PR 3 ligands were examined: PMe 3 ( 3a* ), PPhMe 2 ( 3f* ), P(OMe) 3 ( 3m* ), and P(NMe 2 ) 3 ( 3o* ). These model complexes are shown in Fig. 3.4 . Figure 3 . 4 Structures examined by NBO/NRT analysis. The general approach to these calculations started with geometry optimization for each structure. The initial structural models input for optimization were based on modified versions of the X - ray structures of the closest 3 derivatives ch aracterized. These model compound structures were optimized using DFT with the B3PW91 functional and 6 - 311G+(d,p) basis set on all atoms, with Gaussian09. 39 Upon successful optimization, each structure was analyzed using the NBO6 program. 40 Through this program, lo the P atom, and its direct substituent atoms on each molecule, were performed (see experimental for more specifics). 6 Brennan performed several sets of calculations looking at G values for the - NR 2 rotations, where R = H, Me, or i Pr. From his efforts, little difference was noted among the different R groups in the trends and absolute values calculated for G , but increasing the R group complexity dramatically increased computational time. Therefore, we lculations, especially because we were considering only orbital interactions rather than energies. 210 Figure 3 . 5 Resonance forms inherent to the CrN 3 fragment from rearrangement of the lone pair electron density across the N ligands. These resonance forms and their electron rearrangements do not affect the nature of the Cr P bonds, so they are summed as *Cr to focus resonance form discussion on the Cr P interaction. Before discussing the results of the NRT calculations specifically, it is worth noting that, However, the various rearrangements that can occur via electron reorganization of the N lone pairs and bonds to Cr do not directly affect the character of the Cr P bond. For example, whether N2 versus N3 has a lone pair, the same electronic interaction occurs between Cr and P. Therefore, when electro n density shifts among the N ligands, but the Cr P bond and the PR 3 bonds are the same, we have summed the total contributions of these resonance forms to focus discussion on the Cr P bond. This treatment is summarized in Fig. 3. 5. By most descriptions, PMe 3 is a simple ligand. In trialkyl phosphines, the P are high in energy and participate very little in their bonding interactions, even in low valent systems. Consequently, these PR 3 - donors. 24 Knowing this we started our NBO/NRT analysis with 3a* . Unsurprisingly, with the PMe 3 ligand bound to our Cr* model, the result ing bonding picture is very simple. According to NBO/NRT calculations, 211 >99% of the ground state electronic structure can be described by two resonance forms. The first, - bond formation from dative donation of the P lone pair to Cr, whic h accounts no lone pair donation from P to Cr occurs (alternatively, population of Cr electronic description is a nearly perfect example o f a purely dative interaction and these results are summarized in Fig. 3. 6. 26 Figure 3 . 6 NRT determined resonance forms accounting for 99% of the ground state of 3a* . This bonding picture, generated by NBO and represented by NRT, served as a nice check for the system and application of resonance theory to this l ow symmetry system. With these results, we next sought to analyze slightly more complicated systems, 3m* and 3o* . In both systems, we 3a* . The contribution of these reson ance forms, the components of a classic dative bond, are shown in Fig. 3. 7. 212 Figure 3 . 7 Contribution of the two resonance forms that compose a dative interaction to the groun d state electronic structure of 3m* and 3o* . According to the NRT analysis of these complexes only ~60% of the ground state electronic structure is described by these resonance forms. Most interesting is that the contribution ly decreased, making up less than 5% of the ground state structure in both compounds. This alone lends support to the increased donor ability of these phosphines, as it suggests an increased bond order between the PR 3 and Cr in these structures relative to 3a* . But what accounts for the remaining 30% of the electronic ground state? Closer inspection of the NRT results shows two additional resonance forms that contribute substantially to the overall ground state. The f systems, as well. When heteroatom - containing R groups are substituents on P in a PR 3 ligand, negative hyperconjugation allows for lone pair donation from one substituent into the P orbital with one of the other substituents. This creates double bond character between P and R 1 (donor), while R 2 dissociates with negative charge (acceptor). 24, 30 The introduction of this resonance form, was not surp rising, but can be considered to contribute to the increased donor 213 Fig. 3.10 and is shown by the arrow pushing diagram in Fig. 3.8 , below. Figure 3 . 8 3m* and 3o* . The example is shown here with 3m* . One final resonance form that appears to contribute substantially to the ground state electronic structures with P(OMe) 3 and P(NMe 2 ) 3 appear in literature describing PR 3 interactions with low valent metals. In this resonance form, a lone pair on a heter oatom substituent donates electron density into the Cr negative hyperconjugation with the metal! This pushes a lone pair of elect ron density onto the Cr* fragment. Arrow pushing that represents this electron redistribution is shown below in Fig. 3.9 , as well as in Fig. 3.10 as the 214 Figure 3 . 9 3m* and 3o* . The example is shown where with 3o* . Figure 3 . 10 Ground state electronic structures of 3m* and 3o* among the - OMe and - NMe 2 involving the metal. The observation of this new re increases the electron density on the Cr* fragment by a full electron pair. This resonance form is s acidity of the Cr(VI) metal center. Another way to think about this phenomenon is to consider what happens to the electronegativity of an atom as it is oxidized. Looking at the Sanderson electronegativities of the different oxidation states of Cr, for ex ample, the following values are assigned: Cr(II) = 1.24, Cr(III) = 1.66, Cr(IV) = 2.29, Cr(V) = 2.83, and Cr(VI) = 3.37. 41 As the metal becomes increasingly oxidized, the electronegativity increases substantially. In fact, on the Sanderson scale, the electronegativity of Cr(VI) is comparable to that of oxygen or nitrogen. When the metal in an M 215 where the electronegativity of the highly oxidized metal is comparable to that of the heteroatom substituents on P, the metal begins to participate in negative hyperconjugation. These calculations have provided electronic structures which support the po ssibility that, in interactions with high valent metals, bonding interactions occur which increase the electron irectly increases the electron density donated from the PR 3 fragment to the CrN 3 fragment via negative hyperconjugation from the substituents on P structure wou ld result in more lone pair electron density on the amide ligands and a reduced barrier to Cr N i Pr 2 bond rotation. With a clear picture of the bonding interactions that make up a total of ~90% of the ground state electronic structure in 3m* and 3o* , we e xamined 3f* . We suspected that in the aryl - 3 ligands of 3m* and 3o* might be increasing their electron donor abilities. NRT calculations were performed on 3f* in a slightly differ ent manner compared to the other complexes. Due to the massive number of possible resonance forms enabled by the inclusion of the entire phenyl ring on the PPhMe 2 ligand, the NBO6 program could not run the NRT calculation. Inclusion of the entire PPhMe 2 li gand was crucial to the analysis, so in order to free - up more space in the NRT calculations, we froze the nitrogen ligands using the CHOOSE keyword (see Experimental). Only the Cr, P, and P substituent groups were included in the local NRT calculation. W ith the Cr N interactions frozen, the calculation involving the aromatic ring was able to run successfully. The results of the NRT calculation for 3f* are shown below in Fig. 3.11. Similar 216 - electron density from the ph enyl ring into the P orbital, was observed for 3f* . We would note that, because a slightly different method was used for 3f* , due to the limits of the NRT calculation, comparing the percentages from this calculation to those for 3a* , 3m* , and 3o* may not be valuable. Also of note is the fact that, because the other Figure 3 . 11 NRT results for 3m* and 3o* , electron density is pushed from one of the R groups into the Cr The NRT results are consistent with the energies of several orbital interactions in the Se - backbonding interactions within all 4 complexes (~3 kcal/mol = E2). Interestingly, these donations from filled Cr N nitride and amide bonds into P espite radically different R substituents, in terms of their stabilization abilities for this type of interaction. Additionally, there (~12 kcal/mol) inte ractions for 3m* and 3o* . These substantial energies support the interpretation of the NRT results with these complexes. 217 Finally, examination of the optimized structures 3a* , 3m* , 3o* , and 3f* , by Mayer Bond Order 42 analysis suggests that the Cr P bond orders increase in the order 3a* < 3f* < 3m* ~ 3o* . There is consistency across several aspects of the calculation s suggesting that, in fact, based on the electronic structure, there is an explanation for how PR 3 ligands which are poor donors in low valent metal systems are much better donors to high valent metals. 3.4 Modeling Approximations to Examine Stereoelectronic Control on LDP Value values, determined for 3a - p , calculations had pointed to real electronic effects causing improved donation from traditionally poor PR 3 values determined in CD 3 CN, where ion pairing effects were eliminated, the main concern was entropic inconsistencies. Knowing this, we were still unable to determine if the experimentally measured entropies in this system for 3f , 3j , and 3m were actually relevant to the Cr N i Pr 2 values determined from the Eyring plot of the rotation rate of the Cr N i Pr 2 bond in 3m [SbF 6 ] at value of 6.76 kcal/mol. In the simplest values measured here have the same meaning as those in the NCr(N i Pr 2 ) 2 X system, this means that P(OEt) 3 is a better donor than (NMe 2 ) (See Chapter 2, Fig. g at the room temperature 1 H NMR spectra of the two value seems similarly affected by solvent. are consistent when the ions are paired in the [NC r(N i Pr 2 ) 2 PR 3 ][BArF 24 ] salts with CDCl 3 , and similar to the values observed in the NCr(N i Pr 2 ) 2 X system. 1 Therefore, we value for the [NCr(N i Pr 2 ) 2 PR 3 ] + bond rotation is actually different in this system. With this suspicion, we set about trying to model the stereoelectronic effects of 218 each PR 3 values observed with X ¯ = SbF 6 ¯ in CD 3 CN. For this purpose, we left the = - 9 e.u. in place. Given the similarity of the PR 3 ligands where R = value for these complexes should be similar. Additionally, all 5 experimental measurements were conducted at similar temperatures (within 10 ° C), so small calculated. The stereoelectronic properties of the trialkyl phosphines, as mentioned above are - donors, bu t due to the covalency of the P C bonds and very little in bonding. Therefore, if we consider even an elaborate parameterization system, such as QALE, only tw o terms are needed to parameterize 3a - e . This assumption was verified with the [NCr(N i Pr 2 ) 2 PR 3 ] + system via the computational analysis with NRT. By this method, we found that 99% of the electronic ground state is composed of a pure dative interaction. Ther efore, based on our own experimental and literature evidence, we elected to examine a simple 2 - parameter fit, - electron donor ability and sterics. d to represent the electronic properties of the PR 3 represen fitted as the dependent term. The model then fits Eq. 3.2, below. ( Eq. 3. 2 ) According to the model, both electronic donor ability and sterics make sizable value for each PR 3 ligand. We can verify that the values (or the steric and electronic profiles). The fit shown in Fi g. 3.12 demonstrates good correlation between the 219 values suggesting that the least squares fit in Eq. 3.2 is an accurate predictor of the donation ability for the trialkyl phosphines. Furthermore, the model makes chemic al sense. According to Eq. 2, as the size of PR 3 also increases. This is the logical results as a larger ligand on Cr will lead to steric hindrance of the measured bond d parameter increases and the PR 3 ligand bec - electron value also increases. Since dative donation of the lone pair is the only significant bonding interaction between Cr and PR 3 with a trialkyl substituted P, this correlation is also logical. These trends are signified b y the (+) coefficients in Eq. 3.2. Figure 3 . 12 values for trialkyl phosphines d We tried to apply this simple, 2 - parameter model, with the coefficients fitted from the trialkyls in Eq. 3.2, to the entire series of electronically diverse PR 3 ligands in the series 3a - p . This does not produce good correlation. In fact, the data series looks like a nearly random scatter plot (Fig. 3.13). When we begin dissecting this scatter more closely, it resembles several of the plots shown by Giering and Prock in the QALE system, where there are stepwise deviations from the model based on the types of R groups. 15 This is highlighted in the color coding in the figure below. This observation directly supports the conclusion that diff erent electronic factors are likely y = 0.9503x + 0.8453 R² = 0.9503 16 16.5 17 17.5 18 16 16.5 17 17.5 18 Experimental H Model - Predicted H Experimental Vs. Model - Predicted 220 affecting the measured H values here, and that these effects appear to be related to systematic alterations in the R groups on the PR 3 ligands. While further quantitative fittings with these data were not pursued, cons idering the remaining issues with entropy complications, these findings are in agreement with both computational and experimental observations. Figure 3 . 13 All complexes 3a - p fitted with the model in Eq. 2. (Red = trialkyl, orange = monoaryl, green = diaryl, blue = heteroatom substituents). 3.5 Conclusions stereoelectronic model detailing the exact quanti tative effect of each electronic property of a PR 3 ligand on the donor ability of that ligand. However, the fact that the aryl and heteroatom substituted phosphines do not fit Eq. 2, supports the assessment that these phosphines have additional electronic properties affecting their bonding interactions with Cr. Qualitatively, this agrees with values examined. Despite ongoing difficulties in experimentally determining the true donor abilit ies of these PR 3 ligands, without solvent effects caused by ionicity, the methods presented here that examine the Cr PR 3 bonding interaction consistently point toward the same conclusion. PR 3 ligands that are poor -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 -5 0 5 10 15 20 LDP A C( d Electronic Profile 221 donors for traditional low valent metals a re strong donors for high valent metals due to the enhanced Lewis acidity and electronegativity of a highly oxidized metal. Specifically, from the NRT calculations, orbital interactions which facilitate this enhancement in donor ability were discovered. W ith our high valent metal system, a form of - bonding or lone pair electron density from an R group populates the Cr - bonding (aryl) or lone pair (heteroatom) electron density better donors to metal centers which are Lewis acidic, electronegative, or electron deficient (i.e. high valent). These findings demonstrate that the interactions of ligands with metal centers in drastically different oxidation states can completely change the role of the ligand in the bonding interaction with the metal. It also points toward the risks of extending chemical intuition built on an understanding of metal - ligand interactions with low valent systems to high valent ones. The fundamental differences in the types of interactions are potentially quite substantial and could lead the unsuspecting researcher completely off target when attempting to manipulate metal complex behaviors through an cillary ligand selection. Experimental investigations to probe the differences in high valent metal interactions with a wider variety of ligand types is a necessary endeavor to advance catalysis with these metals. For the purposes of quantifying the fundam ental bonding effects discovered in these studies, it is likely that a neutral high valent metal system is needed to experimentally verify the enhanced donor abilities of phosphites and aryl phosphines. 3.6 Experimental General Considerations All complexes d iscussed in this chapter were previously characterized and discussed in Chapter 2. Additionally, details on ion pairing behavior and other NMR measuremnets are 222 included in Chapter 2, which should be referenced for further information. The exception to this is the phosphine oxide complex utilized for determination of the Gutmann Paramater, the AN. Details on this complex are provided below. Calculations All calculations were carried out at the High Performance Computing Center (HPCC) through Michigan State - Enabled Research. DFT optimizations were performed using Gaussian09 with B3PW91 and the 6 - 311G+(d,p) basis for all chromium and main group compounds. The NBO and local NRT ( vide infra ) calculations were performed using NBO6 . The phosphine complexes were analyzed using NCr(NH 2 ) 2 PE 3 + as a model in order to reduce the amount of computational time required for optimization. To focus on the interactions of the bound ligand atom (P) to the metal and simplify the NRT analysis, we chose to localize the NRT calculations on the metal and all atoms bound directly to it. The first substituent atoms on the phosphorus atom, directly bound in E through P - bond, were also included. Primarily, the localization only eliminates inclusion of the hydrogens on the amide ligands in the Cr(VI) model and alkyl carbons and hydrogens on the distal portions of the E groups. So, for example, in the [NCr(NH 2 ) 2 PMe 3 ] + cation, the NRT included the chromium, three nitrogens, phosphorus, and the three carbo ns in the methyl groups. This method allowed for the calculations to focus on the first coordination sphere interactions with chromium and the phosphine with charge distribution among distal protons and similar effects which do not involve rearrangement of the electron density at the metal. The only exception to this NRT method was made for [NCr(NH 2 ) 2 PMe 2 Ph] + . The full phenyl ring was crucial to examining the impact of an aromatic substituent on phosphorus, so all carbons in the Ph group were included in the calculation. However, too many resonance forms 223 were found during these calculations. Consequently, the structure below was used with the CHOOSE keyword, and resonance forms localized on the chromium, phosphorus, and all carbons in the phosphine ligand were calculated. The default applications of NRT then locked the N interactions of the CrN 3 portion of the fragment as shown below. This arrangement is roughly the average for the resonance forms s intended to focus on the Cr P interaction. Additionally, the resonance of the Cr N bonds behaves predictably based on other calculations performed, and do not alter the interactions of Cr and P. Thus, we do not believe that this adjustment to the method lead to substantially different treatment of those atoms included in the local NRT. Figure 3 . 14 Ground state resonance assignment to Cr fragment by NRT for [NCr(NH 2 ) 2 PMe 2 Ph] + . Additional Analysis from the NBO Calculations: Below is a list of the atomic charges assigned by NBO for the atoms considered in the NBO calculations. Note that since any change in the electron density at Cr is delocalized across the entire CrN 3 fragment, the changes in the charg es should be relatively small in magnitude. Thorough analysis and overemphasis of the differences noted in these calculated charges, especially in the 224 absence of a valid assignment of error for these calculated values, is neither highly informative nor rec ommended. However, the values are listed below for transparency. Table 3 . 5 Natural Charges for NCr(NH 2 ) 2 PE 3 + complexes NBO Charge by Atom Complex Cr P N1 a N2 b N3 b R c R c R c [Cr]PMe 3 + 0.57790 1.05257 - 0.05995 - 0.85567 - 0.92391 - 0.92391 - 0.91721 - 0.91721 [Cr]POMe 3 + 0.53725 1.94464 - 0.07349 - 0.8703 - 0.86708 - 0.84327 - 0.83297 - 0.83121 [Cr]PPhMe 2 + 0.58370 1.08700 - 0.07261 - 0.84814 - 0.86036 - 0.92910 - 0.91149 - 0.38288 [Cr]P(NMe 2 ) 3 + 0.54956 1.74058 - 0.06175 - 0.87353 - 0.86802 - 0.84567 - 0.86371 - 0.87116 a N1 = nitride nitrogen. b N2/3 = amide nitrogens. c R = C or E bound to P. For the default bonding in the NCr(NH 2 ) 2 PE 3 + molecules, NBO chose the following structure for 3 of the 4 molecules: CHOOSE 1 (picked by NBO) Cr - N1 triple bond Cr - N2 double bond Cr - N3 double bond Cr - P single bond Figure 3 . 15 CHOOSE 1 geometry for NBO analysis. 225 This arrangement was not the default for P(NMe 2 ) 3 but was easily selected with an appropriate CHOOSE command added to the input file. NALYSIS OF 3 examined using this above bonding configuration. The type of interaction, Donor - acceptor, and reported E2 value (kcal/mol) are reported. 226 [NCr(NH 2 ) 2 PMe 3 ] + ( 3a * ) N1 = nitride, N2/3 = Donor Acceptor E (kcal/mol) LP1 N1 BD*1 P2 - C6 0.74 BD2 Cr - N2 BD*1 P2 - C14 2.33 BD2 Cr - N1 BD*1 P2 - C6 3.07 BD3 Cr - N1 BD*1 P2 - C14 0.64 BD3 Cr - N1 BD*1 P2 - C10 0.64 BD2 Cr - N3 BD*1 P2 - C10 2.33 Total 9.75 ( avg 3.25/CH 3 ) BD1 P2 - C6 BD*1 P2 - C10 1.31 BD1 P2 - C6 BD*1 P2 - C14 1.31 BD1 P2 - C10 BD*1 P2 - C14 1.27 BD1 P2 - C10 BD*1 P2 - C6 1.33 BD1 P2 - C14 BD*1 P2 - C10 1.27 BD1 P2 - C14 BD*1 P2 - C6 1.33 Total 7.82 ( avg 2.60/CH 3 ) 227 - R to Cr BD1 P2 - C14 BD*1 Cr1 - P2 1.50 BD1 P2 - C10 BD*1 Cr1 - P2 1.50 Total 3.00 228 [NCr(NH 2 ) 2 P(OMe) 3 ] + ( 3m* ) Donor Acceptor E (kcal/mol) LP1 N1 BD1* P1 - O3 0.72 BD2 Cr - N2 BD1* P1 - O4 2.53 BD2 Cr - N3 BD1* P1 - O3 1.27 BD3 Cr - N1 BD1* P1 - O4 0.82 BD3 Cr - N1 BD1* P1 - O15 0.75 BD2 Cr - N2 BD1* P1 - O15 1.98 Total 8.07 ( avg 2.7/OMe ) 229 BD1 P1 - O15 BD1* P1 - O3 4.28 BD1 P1 - O15 BD1* P1 - O3 1.98 BD1 P1 - O4 BD1* P1 - O3 2.28 BD1 P1 - O 4 BD1* P1 - O14 4.12 BD1 P1 - O3 BD1* P1 - O4 2.63 BD1 P1 - O3 BD1* P1 - O15 1.23 LP2 O15 BD1* P1 - O3 1.31 LP2 O15 BD1* P1 - O4 8.91 LP1 O15 BD1* P1 - O3 6.46 LP1 O15 BD1* P1 - O4 1.39 LP2 O4 BD1* P1 - O3 13.61 LP1 O4 BD1* P1 - O15 6.74 LP2 O3 BD1* P1 - O15 6.79 LP2 O3 BD1* P1 - O4 11.90 LP1 O3 BD1* P1 - O15 2.43 LP1 O3 BD1* P1 - O4 1.51 Total 77.57 ( avg 25.9/OMe ) 230 - R to Cr BD1 P1 - O15 BD1* P1 - Cr 1.21 BD1 P1 - O3 BD1* P1 - Cr 1.69 LP2 O15 BD1* P1 - Cr 3.59 LP1 O15 BD1* P1 - Cr 0.56 LP2 O4 BD1* P1 - Cr 2.14 LP1 O4 BD1* P1 - Cr 1.52 LP1 O3 BD1* P1 - Cr 0.84 Total 11.55 231 [NCr(NH 2 ) 2 PMe 2 Ph] + ( 3f* ) Donor Acceptor E (kcal/mol) LP1 N1 BD1* P - C5 0.82 BD2 Cr - N2 BD1* P - C20 2.63 BD2 Cr - N3 BD1* P - C9 2.36 BD2 Cr - N1 BD1* P - C5 3.26 BD3 Cr - N1 BD1* P - C5 0.56 BD3 Cr - N1 BD1* P - C5 0.71 Total 10.3 (avg 3.4/C) BD1 P - C5 BD1* P - C9 1.04 BD1 P - C5 BD1* P - C20 1.24 BD1 P - C9 BD1* P - C5 1.46 BD1 P - C9 BD1* P - C20 1.28 BD1 P - C20 BD1* P - C9 0.91 BD1 P - C20 BD1* P - C5 1.17 BD2 C9 - C10 BD1* P - C20 2.90 Total 10 (avg 3.3/C) 232 - R to Cr BD1 P - C5 BD1* Cr - P 1.59 BD1 P - C9 BD1* Cr - P 1. 23 BD1 P - C20 BD1* Cr - P 1.70 BD2 C9 - C10 BD1* Cr - P 1.46 Total 5.98 233 [NCr(NH 2 ) 2 P(NMe 2 ) 3 ] + ( 3o* ) N6 = nitride, N5/7 = amide, N3/4/8 = P(NMe 2 ) 3 Donor Acceptor E (kcal/mol) LP1 N6 BD1* P - N4 1.50 BD2 Cr - N5 BD1* P - N3 1.74 BD2 Cr - N6 BD1* P - N4 1.19 BD3 Cr - N6 BD1* P - N3 0.99 BD1 Cr - N7 BD1* P - N8 0.54 BD2 Cr - N7 BD1* P - N3 0.53 BD2 Cr - N7 BD1* P - N8 1.96 Total 8.5 (avg 2.8/NMe 2 ) 234 hyperconjugation among R groups LP1 N3 BD1* P - N4 2.32 LP1 N3 BD1* P - N8 6.23 LP1 N4 BD1* P - N3 11.79 LP1 N4 BD1* P - N8 5.35 LP1 N8 BD1* P - N3 1.59 LP1 N8 BD1* P - N4 14.61 BD1 P - N3 BD1* P - N4 2.82 BD1 P - N3 BD1* P - N8 2.95 BD1 P - N4 BD1* P - N3 2.13 BD1 P - N4 BD1* P - N8 1.78 BD1 P - N8 BD1* P - N3 2.60 BD1 P - N8 BD1* P - N4 1.89 Total 53.7 (avg 17.9/NMe 2 ) 235 - R to Cr LP1 N3 BD1* Cr - P 4.64 LP 1 N8 BD1* Cr - P 1.55 BD1 P - N3 BD1* Cr - P 0.97 BD1 P - N4 BD1* Cr - P 1.35 BD1 P - N8 BD1* Cr - P 0.93 Total 9.44 We know in the LDP system (NCr(N i Pr 2 ) 2 PE 3 + ) the Cr - N i Pr 2 bonds are between double and single bonds in character. 13 In light of this, we wanted to see how the second order perturbation theory interactions and their energies would change with an alternate bonding arrangement. The CHOOSE command was used to assign each of the NCr(NH 2 ) 2 PE 3 + complexes the following bonds and lone pairs: CHOOSE 2 Cr - N1 triple bond Cr - N2 single bond Cr - N3 single bond Cr - P single bond Figure 3 . 16 CHOOSE 2 geometry for NBO analysis. 236 3 examined using this above bonding configuration. The type of inter action, Donor - acceptor, and reported E2 value (kcal/mol) are reported. 237 [NCr(NH 2 ) 2 PMe 3 ] + ( 3a* ) Donor Acceptor E (kcal/mol) LP1 N3 BD1* P - C6 0.85 LP1 N4 BD1* P - C6 0.74 LP1 N5 BD1* P - C10 0.85 BD1 Cr - N3 BD1* P - C14 0.78 BD3 Cr - N4 BD1* P - C6 2.65 BD1 Cr - N5 BD1* P - C10 0.78 Total 6.65 (2.2/Me) Negative hyperconjugation among R groups BD1 P - C6 BD1* P - C10 1.31 BD1 P - C6 BD1* P - C14 1.31 BD1 P - C10 BD1* P - C6 1.33 BD1 P - C10 BD1* P - C14 1.27 BD1 P - C14 BD1* P - C6 1.33 BD1 P - C14 BD1* P - C10 1.27 Total 7.82 (2.6/Me) 238 Negative hyperconjugation from P - R to Cr BD1 P - C6 BD1* Cr - P 1.74 BD1 P - C10 BD1* Cr - P 1.64 BD1 P - C14 BD1* Cr - P 1.64 Total 5.02 239 [NCr(NH 2 ) 2 P(OMe) 3 ] + ( 3m* ) Donor Acceptor E (kcal/mol) LP1 N12 BD1* P - O4 1.13 LP1 N13 BD1* P - O3 0.72 LP1 N14 BD1* P - O15 0.99 BD1 Cr - N12 BD1* P - O4 0.65 BD3 Cr - N13 BD1* P - O3 0.93 Total 4.42 (1.5/OMe) 240 Negative hyperconjugation among R groups LP1 O3 BD1* P - O4 1.51 LP1 O3 BD1* P - O15 2.43 LP2 O3 BD1* P - O4 11.90 LP2 O3 BD1* P - O15 6.79 LP1 O4 BD1* P - O15 6.74 LP2 O4 BD1* P - O3 13.61 LP1 O15 BD1* P - O3 6.46 LP1 O15 BD1* P - O4 1.39 LP2 O15 BD1* P - O3 1.31 LP2 O15 BD1* P - O4 8.91 BD1 P - O3 BD1* P - O4 2.63 BD1 P - O3 BD1* P - O15 1.23 BD1 P - O4 BD1* P - O3 2.28 BD1 P - O4 BD1* P - O15 4.12 BD1 P - O15 BD1* P - O3 4.28 BD1 P - O15 BD1* P - O4 1.98 Total 77.57 (25.9/OMe) 241 Negative hyperconjugation from P - R to Cr LP1 O3 BD1* Cr - P 0.93 LP1 O4 BD1* Cr - P 1.63 LP2 O4 BD1* Cr - P 2.39 LP2 O15 BD1* Cr - P 4.03 BD1 P - O3 BD1* Cr - P 1.92 BD1 P - O15 BD1* Cr - P 1.34 Total 12.24 242 [NCr(NH 2 ) 2 P(NMe 2 ) 3 ] + ( 3o* ) N6 = Nitride, N5/7 = amide, N3/4/8 = P(NMe 2 ) 3 Donor Acceptor E (kcal/mol) BD3 Cr - N6 BD1* P - N4 1.17 LP1 N7 BD1* P - N8 1.55 LP1 N6 BD1* P - N4 1.50 LP1 N5 BD1* P - N3 1.38 Total 5.6 (1.9/NMe 2 ) Negative Hyperconjugation among R groups LP1 N3 BD1* P - N4 2.32 LP1 N3 BD1* P - N8 6.23 LP1 N4 BD1* P - N3 11.79 LP1 N4 BD1* P - N8 5.35 LP1 N8 BD1* P - N3 1.59 LP1 N8 BD1* P - N4 14.61 BD1 P - N3 BD1* P - N4 2.82 BD1 P - N3 BD1* P - N8 2.95 BD1 P - N4 BD1* P - N3 2.13 BD1 P - N4 BD1* P - N8 1 .78 BD1 P - N8 BD1* P - N3 2.60 BD1 P - N8 BD1* P - N4 1.89 Total 53.7 (avg 17.9/NMe 2 ) 243 Negative hyperconjugation from P - R to Cr LP1 N3 BD1* Cr - P 5.61 LP1 N8 BD1* Cr - P 1.82 BD1 P - N3 BD1* Cr - P 1.19 BD1 P - N4 BD1* Cr - P 1.60 BD1 P - N8 BD1* Cr - P 1.10 Total 11.32 244 [NCr(NH 2 ) 2 PPhMe 2 ] + ( 3f* ) N24 = nitride, N3/4 = amide, C5/20 = Me, C9 = Ph Donor Acceptor E (kcal/mol) LP1 N3 BD1* P - C20 0.92 LP1 N4 BD1* P - C9 0.99 LP1 N24 BD1* P - C5 0.82 BD1 Cr - N3 BD1* P - C20 0.89 BD1 Cr - N4 BD1* P - C9 0.77 BD3 Cr - N24 BD1* P - C5 2.79 Total 7.18 (2.4/C) Negative hyperconjugation among R groups BD1 P - C5 BD1* P - C9 1.04 BD1 P - C5 BD1* P - C20 1.24 BD1 P - C9 BD1* P - C5 1.46 BD1 P - C9 BD1* P - C20 1.28 BD1 P - C20 BD1* P - C9 0.91 BD1 P - C20 BD1* P - C5 1.17 BD2 C9 - C10 BD1* P - C20 2.90 Total 10 (3.3/C) 245 Negative hyperconjugation from P - R to Cr BD1 P - C5 BD1* Cr - P 1.75 BD1 P - C9 BD1* Cr - P 1.41 BD1 P - C20 BD1* Cr - P 1.84 BD2 C9 - C10 BD1* Cr - P 1.74 Total 6.74 Table 3 . 6 Summary of Second Order Perturbation Interactions Discussion of Natural Charges The electronic competition created between the two amide ligands and the PE 3 ligand, for should result in very similar natural charges in Cr for all four of the NCr(NH 2 ) 2 PE 3 + cations examined by NBO. In fact, this seems to be the cas e, as there are only very small differences in the natural charge assigned to Cr in the four different cations, with the largest difference being PR 3 Interaction typ e CHOOSE1 E2 (kcal/mol) CHOOSE2 E2 (kcal/mol) PMe 3 - backbonding 9.8 (3.3) 6.7 (2.2) 38.1 27.21 7.8 (2.6) 3.0 5.0 P(OMe) 3 - backbonding 8.1 (2.7) 4.4 (1.5) 38.98 22.21 77.6 (25.9) 11.5 12.2 P(NMe 2 ) 3 - backbonding 8.5 (2.8) 5.6 (1.9) 41.66 38.76 53.7 (17.9) 9.4 11.3 PPhMe 2 - backbonding 10.3 (3.4) 7.2 (2.4) 39.13 29.23 10.0 (3.3) 6.0 6.7 246 only 0.046. Additionally, similar contributions of negative charge are located on the N ligands in each struct ure as well. Generally, we know that the charge on Cr should remain relatively constant despite shifts in the electron densities between the metal and the ligands. Thus, it is most informative if we consider the NCr(NH 2 ) 2 + fragment separate from the PE 3 fragment. Examining the molecule in this piecewise fashion, the charge distribution agrees with the bonding interactions suggested by NRT. In terms of total charge, there is more total positive charge on the NCr(NH 2 ) 2 + fragm ent in the PMe 3 derivative, suggesting that PMe 3 is the least donating PE 3 . The amount of positive charge decreases in the order PMe 3 > P(OMe) 3 ~ PPhMe 2 > P(NMe 2 ) 3 . These results are summarized in the table below. Again, while the differences here are smal l in absolute terms, the trend agrees with the NRT calculations, with PMe 3 being the poorest donor to Cr. Table 3 . 7 Total Charges on NCr(NH 2 ) 2 + fragment and PE 3 . PE 3 Charge on NCr(NH 2 ) 2 + Charge on PE 3 PMe 3 0.38051 0.61949 P(OMe) 3 0.33423 0.66577 P(NMe 2 ) 3 0.29951 0.70049 PPhMe 2 0.32444 0.67556 Potentially the largest absolute differences come from the charges localized on the P atoms. The natural charges are most informative when we compare the charge on P in the unbound ligands relative to the charge when the ligand is bound to the CrN 3 fragment. Similar changes are noted for P(OMe) 3 and P(NMe 2 ) 3 . The free phosphite has a charge of +1.504 localized on P, while the phosphite bound to Cr has a cha rge of +1.945; the P(NMe 2 ) 3 has a free charge of +1.312 on P, and a charge of +1.740 when bound to Cr. In both cases, the P has increased in positive charge by about 0.43. By comparison, the charge on P in free PMe 3 is +0.737, while the charge on the bound phosphine is +1.053. The difference with the P charge with PMe 3 is only about +0.316. At the 247 same time, the substituent atoms bound directly to P in all 3 cases have essentially the same charge - 0.03 in total cha rge per E). This suggests that the increased positive shift in the natural charge on P in P(OMe) 3 and P(NMe 2 ) 3 is related to the Cr P interaction and is the result of greater donation (higher degree of oxidation) from P to Cr in these cases. The charge den sity analysis provided by NBO is in agreement with the suggested bonding about the nature of the orbitals or types of bonds that facilitate donation from each P E 3 to the metal center. Discussion of Second Order Perturbation Theory Interactions CHOOSE Bonding Arrangements The second order perturbation theory (SOPT) list was analyzed after performing the NBO calculations with two different bonding arrangements. The first, which was the default chosen by the NBO program for 3 of the 4 input structures (CHOOSE1), displays a Cr N triple bond with the nitride, Cr N double bonds with both amide groups, and a Cr P single bond. In this arrangement, the maximum potential - backbonding will be reached, as this places - bonding orbitals in proximity to the P Because we know from experiments using NCr(N i Pr 2 ) 2 PE 3 + that the Cr N amide bonds are most accurately described with a bond order between 1 and 2, 13 we show the same NBO results for a different bonding arrangement. In CHOOSE2, the model cation has a Cr N triple bond with the nitride and a Cr P single bond; how ever, the two Cr N bonds to the amide ligands were specified as single bonds, which results in a lone pair on each amide nitrogen. 248 Figure 3 . 17 ( left ) CHOOSE1; ( right ) CHOOSE2 Comparing the SOPT analysis between the two different structural arrangements is quite - bonding electrons in the Cr N double bonds with the amides, out on the - backbonding in each structure to ~2 kcal/mol per P electronic check on the system as it is interpreted by NBO. In these types of NCr(NR 2 ) 2 X systems, a decrease in electronic donation from the X ligand results in an increase in donation from the amides; conversely, an increase in the electronic donation fr om X results in a decrease in donation from the amide ligands. This push - and - pull or electronic competition is directly observable in these forced bond changes of the initial NBO structure. Overall these results agree very well with the previous MO calcula tions done with the NCr(NR 2 ) 2 X system, as well as chemical intuition. 1 A Comparison of SOPT and NRT Results Perhaps most important in terms of the second order perturbation theory results, is the fact that the predominant interactions from the NRT result s are readily recognized in the SOPT participate to different amounts for each cation. In the SOPT list, the starting structure has a Cr P single bond. The stabiliz ation energies found for the Cr 3 ligands in CHOOSE1 are close to 40 kcal/mol. Upon switching to CHOOSE2, there is a common trend in these stabilization energies, as they all 249 decrease going to the less electronically satu rated Cr. This means that as electron donation from the amides is removed, the contribution of the Cr increased bond order between Cr and P. , supporting the fact that these delocalizations are primarily ligand based and dependent on the properties of the R group on PR 3 . The magnitudes of these energies are much smaller (order of magnitude) in the PMe 3 and PPhMe 2 compared to the P(OMe) 3 and P(N Me 2 ) 3 cases. For P(OMe) 3 and P(NMe 2 ) 3 , the total stabilization energy from these interactions is quite substantial, calculated to be 77 and 54 kcal/mol, respectively. Considered per P - R interaction, these energies are still quite large, at 26 and 17 kcal/mol. These large energies arise from a well - established electronic behavior associated with heteroatom substituted phosphine ligands; these values serve as a comparative tool to evaluate the magnitude of the stabilization energies of other interactions. participate in each of the 4 structures. This contribution is the largest for P(OMe) 3 and P(NMe 2 ) 3 , in the range of 10 - 12 kcal/mol. In both cases, the stabilization energy of this interaction increases in CH OOSE2 compared to CHOOSE1. This trend reflects the need to increase donation from the PR 3 is about half as large in PPhMe 2 3 and P( NMe 2 ) 3 and is only 3 kcal/mol with PMe 3 CHOOSE2, thus the overall trend is the same among the series. The same delocalizations of electron density, whether via donor acceptor inter actions in the second order perturbation list, or with formal resonance forms in the NRT calculations, are readily observed with both methods. 250 In addition to the resonance forms found in NRT, we also examined the SOPT list for interactions - b ackbonding. NRT does not show any evidence of from the CrN 3 fragment to the PE 3 ligand. However, these interactions are observed in the SOPT list, with smaller energies than those observed for any of the other interactions (1 - 3 kcal/mol). It is interesti ng to note that the magnitude of the interaction decreases in the order PPhMe 2 >PMe 3 >P(OMe) 3 >P(NMe 2 ) 3 . This is somewhat counterintuitive when considering the electronegativities of the R groups bound to P, but perhaps is a reflection of the error in the c alculations rather than a real trend. Overall, it seems that these interactions are small, potentially contributing very little to the overall electronic structure of the molecule. The point raised above regarding the magnitude of the stabilization energi es for the resonance form raises another valid point, which is: how much does each of these interactions matter to the overall electronic structure in these cations? The NRT calculations give weighted % values indicating the piecewise contribution of eac h resonance form, while the SOPT list gives stabilization energies that correlate to the amount of delocalizations. discrepancy in these forms, and it comes from th 2 ) 3 case. The stabilization energy in CHOOSE2 is still rather large (> 38 kcal/mol), in contrast to the low percentage that NRT gives for this same delocalization (3%). The trends observed for the other three ph overall electronic structure when other delocalizations are available to participate. Absolute quantification from these methods, however, seems unwise. It seems th - backbonding for P(OMe) 3 and P(NMe 2 ) 3 (10 - 12 vs 1 - 2 kcal/mol respectively). If we compare both 251 interactio ns are about 10% of 3 and P(NMe 2 ) 3 . Based on this Perhaps like the conclusions drawn about the co struct ure. Modeling Approach General Considerations value as a holistic stereoelectronic parameter follows the general form ( Eq. 3.3 ): ( Eq. 3 .3 ) value is broken into a constant value, a , and a series of terms x 1 through x i , each representing unique stereoelectronic properties of the ligand. These properties are weighted by coefficients (b - m n ) to scale relative importance of each property in determ value. values were determined experimentally. However, in order to fully solve the equation, we then need a series of stereoelectronic parameters, to take the place of x 1 - x i in the equation. Available in the literature are a wide variety of descriptors of the steric and electronic properties of phosphines. In terms of electronics, these properties include Tolman Electronic a acidity, etc. 3 - 11 In terms of sterics, descriptors like cone bur ), or the solid G angles are adequate descriptors of ligand 252 term, we can start ( Table 3. 8 ). Table 3 . 8 Sample variety of literature parameters available for stereoelectronic description of phosphine series under study. 3 - 11 a Taken with the standard 3.5 Å radius from the metal center, including hydrogens. b P(OMe) 3 treated as electronic surrogate. c P(NMe 2 ) 3 treated as electronic surrogate In the equation shown above, the coefficients can then be solved for by applying a least - sq uares fit to the chosen series of parameters for x i values for the series of phosphines. With the least squares fit giving real number values to the weighting coefficients, we can rationalize, based on the properties with which each p arameter correlates, whether the effect makes chemical sense. Goodness of fit with a given set of parameters can generally be determined by looking at three different metrics: 1) Electronic profile 2) Steric profile 3) Model predicted LDP value Phosphine FT pK a E° V min PA (kJ/mol) d TCA E ar a %Vbur a PMe 3 8.55 8.55 0.3593 43.02 945 8.55 118 0 0 22.6 P( n Bu) 3 5.25 8.43 0.3994 43.71 5.25 136 0 0 25.8 P( i Bu) 3 5.7 7.97 0.3939 44.8 5.7 143 0 0 28.6 P( i Pr) 3 3.45 0.4406 44.47 3.45 162 0 0 28.7 PCy 3 1.4 9.7 0.4597 44.99 1018 1.4 170 0 0 32.4 PPhMe 2 10.6 6.5 0.317 40.41 961 10.5 122 1 0 23.7 PPh 2 Me 12.1 4.56 0.2674 36.76 964 12.6 130 2.2 0 25.5 PPhEt 2 9.3 6.25 0.3426 40.76 8.6 136 1.1 0 25.3 PPh 2 Et 11.3 4.9 11.1 140 2.3 0 25.6 PPh 2 n Bu 11.1 5 11.3 142 2.1 0 25.2 PPh 2 Cy 5.05 9.1 153 1.6 0 26.9 PPhCy 2 5.7 162 1.6 0 28.0 P(OEt) 3 20.9 3.35 0.1551 27.85 924 b 15.7 109 1.1 2.9 23.4 P(O i Pr) 3 19.05 4 13.4 130 1.3 2.9 25.7 P(NC 4 H 8 ) 3 1015 c 1.2 146 0.6 0.9 28.1 253 Th electronic parameter is demonstrated in the following equation: ( Eq. 3. 4 ) then show very strong correlation if the fit is a good descriptor, with the slope of the line equal to the weighting parameter, b. Any fitting parameter can be examined in this manner. Provided the profiles demonstrat value based on the coefficients and the stereoelectronic parameters ( x 1 x i values calculated by solving the full equation ( Eq. 3.3) against the experimentally mea values will ideally give a plot with intercept of zero and a slope of one. Evaluating our model by examination of the steric, electronic, and model - allows for quantitative determination of the accuracy of the parameter system sele . 12 Parameter Selection As mentioned above, there is an abundance of literature parameters that serve as both electronic and steric descriptors of the phosphine ligands. With such a large selection available, it can be a challenge to select parameters and incorporate them into a model for phosphine behavior. Selection of the best electronic donation parameter was based the trialkylphosphines examined in the system. It is generally assumed that trialkylphosphines are the si mplest phosphines electronically and should provide the simplest interaction with the NCr(N i Pr 2 ) 2 - cation. Good agreement with the trialkyls should - donor ability. Steric parameters were evaluated similarly, by examining the trialkylphosphines for the best correlation with each steric descriptor available. 254 As an illustration of how important the starting stereoelectronic parameters are, Fig. 3.18 shows the differences between two of the steric parameters con sidered for the series of phosphines ( 3a - 3o ). In this case, poor correlation is noted between %V bur and TCA. Again, this illustrates that careful parameter selection is needed to get appropriate representation of the ligands in our system. Based on this me - donor descriptor) and d . Figure 3 . 18 Correlation of %V bur vs. Tolma n Cone Angle for all phosphines used in model building ( 3a - o ). Model Building Unfortunately, due to entropic and ion pairing complications with the variety of solvent/anion combinations examined with the salts 3a - o , a full stereoelectronic model, which e 5 trialkylphosphines 3a - e, however, were examined under the assumption that their similar polarity R groups would interact in a similar manner with solvents, and these compounds would have a similar entropic contribution. A stereoelectronic fit accountin - donor ability and size, was performed using a least squares fit, following the shortened form of the general equation ( Eq. 3.3 ) in Eq. 3.5 . R² = 0.6988 20 22 24 26 28 30 32 34 100 120 140 160 180 %Vbur Tolman Cone Angle ( ) %Vbur vs. Tolman Cone Angle 255 ( Eq. 3. 5 ) d - donation term derived from the QALE system of phosphine d excellent correlation with the LDP system, with th e solved values for the weighting parameters shown in Eq. 3. 6 . ( Eq. 3.6 ) In addition to the high R 2 values observed below in the steric and electronic profiles for this fit of 3a - e , these plots also demonstrate good correlation through the intercepts, each close to zero, and the slopes of each profile matching closely to the fitted coefficients c and b respectively. y = 0.03x - 3E - 12 R² = 0.9898 2 2.5 3 3.5 4 4.5 5 5.5 100 120 140 160 180 C( ) Steric Profile Figure 3 . 19 Steric profile from fit of trialkylphosphines ( 3a - e ) using 2 - parameter fit, Eq. 3.6. 256 Figure 3 . 20 Electronic profile from fit of trialkylphosphines ( 3a - e ) using 2 - parameter fit, Eq. 3 .5 . The application of the trialkylphosphine - derived 2 - parameter model ( Eq. 3.5 ) to the entire series of phosphines shows significant deviation in both the steric and electronic profiles (Fig. 3.21 and 3.22 - donor ability must be considered when describing the donati on of more diverse phosphines to a high valent metal. In the Cr(VI) system here, these deviations with more complex aryl and heteroatom substituents on - interactions, or these could also be ma in the unpaired regime (CD 3 CN with X - = SbF 6 - ). Consequently, expansion of the model to include more diverse stereoelectronic parameters of these phosphines is not a reliable method of determining i nteractions with the metal distinct from solvent effects. y = 0.1347x - 1E - 12 R² = 0.9708 0 0.2 0.4 0.6 0.8 1 1.2 1.4 0 2 4 6 8 10 B( ) d Electronic Profile 257 Figure 3 . 21 Trialkylphosphine 2 - parameter fit applied to total series: Electronic profile. (Orange squares = PR 2 Ph; Green triangles = PPh 2 R, Red circles = PR 3 , Blue circles = P(OR) 3 /P(NR 2 ) 3 .) -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 -5 0 5 10 15 20 LDP A C( d Electronic Profile 258 Figure 3 . 22 Trialkylphosphine 2 - parameter fit applied to total series: Steric profile. (Orange squares = PR 2 Ph; Green triangles = PPh 2 R, Red circles = PR 3 , Blue circles = P(OR) 3 /P(NR 2 ) 3 .) 0 1 2 3 4 5 6 100 120 140 160 180 LDP - A - B( d) Steric Profile 259 Synthetic considerations All manipulations were preformed in an MBraun glovebox under N 2 atmosphere. NCr(N i Pr 2 ) 2 I was prepared according to literature procedures. 43 TlBArF 24 was prepared according to literature proce dures. 44 The triethylphosphine oxide was purchased from Alfa Aesar and used as received. The solvents DCM and Et 2 O were dried by passage over activated alumina and sparged with N 2 prior to use. The NMR solvent CDCl 3 as well as the DCE used to record the Gutmann parameter measurement were dri ed over P 2 O 5 and distilled under N 2 prior to use. Synthesis of [NCr(N i Pr 2 ) 2 (OP(Et) 3 )][BArF 24 ]: A scintillation vial was charged with 50 mg of NCr(N i Pr 2 ) 2 I (0.127 mmol, 1 equiv), 3 mL DCM, a stir bar, and 17 mg OP(Et) 3 (0.127 mmol, 1 equiv). The solution w as stirred at room temperature and to it was added 135 mg TlBArF 24 (0.127 mmol, 1 equiv), as a solution in 2 mL DCM. The dark brown - orange solution rapidly turned bright orange and precipitated a yellow solid. The mixture was stirred 4 h at room temperatur e. The precipitate was removed by filtration over Celite and the filtrate was dried under reduced pressure. The precipitate was dissolved in a minimum amount of DMC, filtered over Celite, and layered with n - pentane. The layered solution was stored at - 35 ° C overnight to yield small, twinned (and disordered) crystals of [NCr(N i Pr 2 ) 2 (OP(Et) 3 )][BArF 24 ] (115 mg, 72%). 1 H NMR (500 MHz, Chloroform - d ): 7.77 7.67 (m, 9H), 7.54 (s, 4H), 5.09 (sept, J = 6.5 Hz, 2H), 3.89 (sept, J = 6.3 Hz, 2H), 2.00 1.80 (m, 12H), 1.44 (d, J = 6.3 Hz, 6H), 1.15 (dd, J = 6.5, 1.5 Hz, 12H), 1.14 1.03 (m, 9H). 13 C NMR (126 MHz, Chloroform - d ) 163.70 160.02 (m), 134.74, 128.92 (d, J = 30.1 Hz), 125.57, 123.40, 117.47, 59.68, 58.20, 30.43 (d, J = 39.4 Hz), 21.50, 21.00, 17.86, 17.35, 4.82 (d, J = 5.0 Hz). 31 P NMR (202 MHz, Chloroform - d ) 86.15. 19 F NMR (470 MHz, Chloroform - d ) - 62.42. 260 Figure 3 . 23 1 H NMR of [NCr(N i Pr 2 ) 2 (OP(Et) 3 )][BArF 24 ] in CDCl 3 . 261 Figure 3 . 24 31 P NMR of [NCr(N i Pr 2 ) 2 (OP(Et) 3 )][BArF 24 ] in CDCl 3 . 262 Figure 3 . 25 19 F NMR of [NCr(N i Pr 2 ) 2 (OP(Et) 3 )][BArF 24 ] in CDCl 3 . 263 Figure 3 . 26 13 C NMR of [NCr(N i Pr 2 ) 2 (OP(Et) 3 )][BArF 24 ] in CDCl 3 . 264 REFERENCES 265 REFERENCES ( 1 ) DiFranco, S. A.; Maciulis, N. A.; Staples, R. J.; Batrice, R. J.; Odom, A. L., Evaluation of donor and steric properties of anionic ligands on high valent transition metals. Inorganic chemistry 2012, 51 (2), 1187 - 200. ( 2 ) Billow, B. S.; McDaniel, T. J. ; Odom, A. L., Quantifying ligand effects in high - oxidation - state metal catalysis. Nature Chemistry 2017, 9 , 837. ( 3 ) Fernandez, A. L.; Reyes, C.; Prock, A.; Giering, W. P., The stereoelectronic parameters of phosphites. The quantitative analysis of liga nd effects (QALE). Journal of the Chemical Society - Perkin Transactions 2 2000, 5 , 1033 - 1041. ( 4 ) Tolman, C. A., Electron donor - acceptor properties of phosphorus ligands. Substituent additivity. Journal of the American Chemical Society 1970, 92 (10), 2953 - 2 956. ( 5 ) Tolman, C. A., Steric effects of phosphorus ligands in organometallic chemistry and homogeneous catalysis. Chemical Reviews 1977, 77 (3), 313 - 348. ( 6 ) Tolman, C. A., Phosphorus ligand exchange equilibriums on zerovalent nickel. Dominant role for s teric effects. Journal of the American Chemical Society 1970, 92 (10), 2956 - 2965. ( 7 ) Bartik, T.; Himmler, T.; Schulte, H. G.; Seevogel, K., Substituenteneinflüsse auf die basizität von phosphorliganden in R3P - Ni(CO)3 - komplexen. Journal of Organometallic Chemistry 1984, 272 (1), 29 - 41. ( 8 ) Based on Local Vibrational Modes. Inorganic chemistry 2014, 53 (1), 478 - 495. ( 9 ) Dias, P. B.; de Piedade, M. E. M.; Simões, J. A. M., Bonding and energetics of phosphorus (III) ligands in transition metal complexes. Coordination Chemistry Reviews 1994, 135 136 (0), 737 - 807. ( 10 ) Drago, R. S., .DELTA.E - .DELTA.C Analysis of Phosphine Basicity. Organometallics 1995, 14 (7), 3408 - 3417. ( 11 ) Vogel, G. C.; Drago, R. S., The ECW M odel. Journal of Chemical Education 1996, 73 (8), 701. ( 12 ) Golovin, M. N.; Rahman, M.; Belmonte, J. E.; Giering, W. P., QUANTITATIVE SEPARATION OF SIGMA - COMPONENTS AND PI - COMPONENTS OF TRANSITION - METAL PHOSPHORUS BONDING AND THE APPLICATION OF LIGAND EF FECTS IN ORGANOMETALLIC CHEMISTRY. Organometallics 1985, 4 (11), 1981 - 1991. 266 ( 13 ) Rahman, M. M.; Liu, H. Y.; Prock, A.; Giering, W. P., Quantitative analysis of ligand effects. 2. Steric and electronic factors influencing transition - metal - phosphorus(III) bonding. Organometallics 1987, 6 (3), 650 - 658. ( 14 ) Rahman, M. M.; Liu, H. Y.; Eriks, K.; Prock, A.; Giering, W. P., Quantitative analysis of ligand effects. Part 3. Separation of phosphorus(III) ligands into pure .sigma. - donors and .sigma. - donor/.pi. - a cceptors. Comparison of basicity and .sigma. - donicity. Organometallics 1989, 8 (1), 1 - 7. ( 15 ) Wilson, M. R.; Woska, D. C.; Prock, A.; Giering, W. P., THE QUANTITATIVE - ANALYSIS OF LIGAND EFFECTS (QALE) - THE ARYL EFFECT. Organometallics 1993, 12 (5), 1742 - 1752. ( 16 ) Poe, A. J., The tuning of P - donor ligands: the aryl and other pendent group effects (PGEs) revisited. Dalton transactions 2009, (11), 1999 - 2003. ( 17 ) Fernandez, A. L.; Reyes, C.; Prock, A.; Giering, W. P., The stereoelectronic para meters of phosphites. The quantitative analysis of ligand effects (QALE). J. Chem. Soc., Perkin Trans. 2. 2000, (5), 1033 - 1041. ( 18 ) Allman, T.; Goel, R. G., The basicity of phosphines. Canadian Journal of Chemistry 1982, 60 (6), 716 - 722. ( 19 ) Ardizzoia, G. A.; Brenna, S., Interpretation of Tolman electronic parameters in the light of natural orbitals for chemical valence. PCCP 2017, 19 (8), 5971 - 5978. ( 20 ) Kühl, O., Predicting the net donating ability of phosphines do we need sophisticated theoretical met hods? Coordination Chemistry Reviews 2005, 249 (5 6), 693 - 704. ( 21 ) Leyssens, T.; Peeters, D.; Orpen, A. G.; Harvey, J. N., Insight into metal - phosphorus bonding from analysis of the electronic structure of redox pairs of metal - phosphine complexes. New J ournal of Chemistry 2005, 29 (11), 1424 - 1430. ( 22 ) Wu, K.; Doyle, A. G., Parameterization of phosphine ligands demonstrates enhancement of nickel catalysis via remote steric effects. Nature Chemistry 2017, 9 , 779. ( 23 ) Fernandez, A.; Reyes, C.; YingLee, T.; Prock, A.; Giering, W. P.; Haar, C. M.; Nolan, S. P., Assessing the stereoelectronic properties of pyrrolyl phosphines and related ligands. The quantitative analysis of ligand effects (QALE). J. Chem. Soc., Perkin Trans. 2. 2000, (7), 1349 - 1357. ( 24 ) - - Acceptor Properties of Phosphorus Ligands: An Insight from the Natural Orbitals for Chemical Valence. Inorganic chemistry 2010, 49 (2), 578 - 582. ( 25 ) Leyssens, T.; Peeters, D.; Orpen, A. G.; Harvey, J. N., Back - Bonding toward YX3 Ligands (Y = N, P, C, Si)? An NBO Analysis. Organometallics 2007, 26 (10), 2637 - 2645. 267 ( 26 ) Parkin, G., Valence, Oxidation Number, and Formal Charge: Three Related but Fundamentally Different Concepts. J ournal of Chemical Education 2006, 83 (5), 791. ( 27 ) Orpen, A. G.; Connelly, N. G., Structural systematics: the role of P - A .sigma.* orbitals in metal - phosphorus .pi. - bonding in redox - related pairs of M - PA3 complexes (A = R, Ar, OR; R = alkyl). Organometal lics 1990, 9 (4), 1206 - 1210. ( 28 ) Orpen, A. G.; Connelly, N. G., Structural evidence for the participation of P in metal PX3 bonding. Journal of the Chemical Society, Chemical Communications 1985, (19), 1310 - 1311. ( 29 ) Fey, N.; Orpen, A. G. ; Harvey, J. N., Building ligand knowledge bases for organometallic chemistry: Computational description of phosphorus(III) - donor ligands and the metal phosphorus bond. Coordination Chemistry Reviews 2009, 253 (5), 704 - 722. ( 30 ) Pacchioni, G.; Bagus, P. S. , Metal - phosphine bonding revisited. .sigma. - Basicity, .pi. - acidity, and the role of phosphorus d orbitals in zerovalent metal - phospine complexes. Inorganic chemistry 1992, 31 (21), 4391 - 4398. ( 31 ) Kobayashi, S.; Manabe, K., Development of Novel Lewis Acid Catalysts for Selective Organic Reactions in Aqueous Media. Accounts of Chemical Research 2002, 35 (4), 209 - 217. ( 32 ) Houk, K. N.; Strozier, R. W., Lewis acid catalysis of Diels - Alder reactions. Journal of the American Chemical Society 1973, 95 (12), 4094 - 4096. ( 33 ) Jung, M. E.; Chang, J. J., Enantiospecific Formal Total Synthesis of (+) - Fawcettimine. Organic Letters 2010 , 12 (13), 2962 - 2965. ( 34 ) Mayer, U.; Gutmann, V.; Gerger, W. J. M. f. C. C. M., The acceptor number A quantitative empirical parameter for the electrophilic properties of solvents. 1975, 106 (6), 1235 - 1257. ( 35 ) Gutmann, V.; Steininger, A.; Wychera, E . J. M. f. C. u. v. T. a. W., Donorstärken in 1,2 - Dichloräthan. 1966, 97 (2), 460 - 467. ( 36 ) Beckett, M. A.; Strickland, G. C.; Holland, J. R.; Sukumar Varma, K., A convenient n.m.r. method for the measurement of Lewis acidity at boron centres: correlatio n of reaction rates of Lewis acid initiated epoxide polymerizations with Lewis acidity. Polymer 1996, 37 (20), 4629 - 4631. ( 37 ) Beckett, M. A.; Brassington, D. S.; Coles, S. J.; Hursthouse, M. B., Lewis acidity of tris(pentafluorophenyl)borane: crystal an d molecular structure of B(C6F5)3·OPEt3. Inorganic Chemistry Communications 2000, 3 (10), 530 - 533. ( 38 ) Sivaev, I. B.; Bregadze, V. I., Lewis acidity of boron compounds. Coordination Chemistry Reviews 2014, 270 - 271 , 75 - 88. 268 ( 39 ) Frisch, M. J.; Trucks, G. W .; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery Jr., J. A.; Peralta, J. E.; Ogliaro, F.; Bearpark, M. J.; Heyd, J.; Brothers, E. N.; Kudin, K. N. ; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A. P.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö.; For esman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09 , Gaussian, Inc.: Wallingford, CT, USA, 2009. ( 40 ) NBO 6.0. E. D. Glendening, J., K. Badenhoop, A. E. Reed, J. E. Carpenter, J. A. Bohmann, C. M. Morales, C. R. Landis, and F. Weinhold., Theoretical Chemistry Institute, University of Wisconsin, Madison (2013). ( 41 ) Sanderson, R. T., Principles of electrone gativity Part I. General nature. Journal of Chemical Education 1988, 65 (2), 112. ( 42 ) Mayer, I., Program "BORDER" version 1.0. Program "BORDER" version 1.0, (Chemical Research Center, Hungarian Academy of Sciences), Budapest 2005 . ( 43 ) Billow, B. S.; Bem owski, R. D.; DiFranco, S. A.; Staples, R. J.; Odom, A. L., Synthesis and Structure of Chromium(VI) Nitrido Cyclopentadienyl Complexes. Organometallics 2015 . ( 44 ) Park, J. G.; Jeon, I. - R.; Harris, T. D., Electronic Effects of Ligand Substitution on Spin Crossover in a Series of Diiminoquinonoid - Bridged FeII2 Complexes. Inorganic chemistry 2015, 54 (1), 359 - 369. 269 CHAPTER 4. SILICA - GEL SUPPORTED TITANIUM CATALYSTS FOR C N BOND FORMING REACTIONS 1,2 4.1 Introduction In 2001, the Odom group published studies that demonstrated the electronic structure of Ti(NMe 2 ) 2 - pyrrolylmethane) complexes and their potential applications to hydroamination - based reaction chemistry. 3 - 5 These efforts mark the starting point for what has become an extensive body of research conducted in the following 2 decades, by various members of the Odom group, into Ti(IV) catalysts for C N bond formation reactions. 6 - 8 Primarily, three different varieties of C N bond forming reactions have been the focus of these catalyst development efforts: (1) hydroamination, coupling amines and alkynes; (2) hydro hydrazination, coupling hydrazines and alkynes; and (3) multicomponent coupling reactions originating from 1 and 2, where more complex organic products are yielded via the inclusion of an additional coupling partner. In particular, the multicomponent coupl ing reactions are of interest to the broader organic synthesis and biological chemistry communities, as many of these multicomponent coupling products can be readily functionalized to yield highly - substituted heterocyclic compounds. 6,9 - 15 A highlight of some of this chemistry, based on iminoamination, is featured in Fig. 4.1 below. 270 Figure 4 . 1 Reaction schemes to yield highly substituted nitrogen - based heterocycles using Ti - catalyzed iminoamination followed by simple organic ring - closing reactions. These processes are conducted as one - pot - two - step reactions from simple starting materials. 6 Several of the products shown in Fig. 4. 1, including substituted quinolines and 2 - amino - 3 - cyanopyridines, have led to collaborations within the department and the university t o explore the biological activity of these complex organic products. Many of the quinoline compounds, which can be produced in 1 - pot - 2 - step reaction sequences using Ti(IV) catalyzed C N bond forming reactions starting from simple, cheap, commercially avail able starting materials, are sub - micromolar inhibitors of the human proteasome. 9 This finding is very interestin g due to the relevance of the human proteasome in many human diseases. Additionally, moieties such as these heterocycles are commonly observed in natural products; employment of these catalyses is a potential tool to basic organic synthesis problems pursin g natural products. 10 Sev eral other groups have been pursuing similar chemistry with Group - 4 transition metal catalysts and deserve mention here. The Schafer group at UBC has pursued ligand design and 271 product functionalization, in addition to expanded substrate tolerance in hydroa mination and similar C N bond forming reactions utilizing homogeneous Ti and Zr catalysts. 16,17 The Tonks group at University of Minnesota has recently been expanding the C N bond forming chemistry - pot synthesis of tetra - substituted pyrroles from alkynes and azobenzen e. Of course, these more recent advances have benefited from early work in the field by the Bergman, Mountford, Bercaw, and Doye groups, which provided fundamental mechanistic investigations and footholds for future catalyst design and development. 18 - 21 For several of the heterocyclic compounds accessible through Group - 4 transition metal catalyzed reaction pathways, complex organic reaction routes could likely lead to the same or similar compounds. However, using the Ti(IV) catalysts developed in our group and other groups, these reactions are 1 to 2 steps and frequently provide moderate to high yields of the desired heterocy cles. 6 Unfortunately, there are limitations with thes e homogeneous Ti(IV) catalysts, which have been discovered in the last several years. Thus, some significant barriers remain along the pathway to optimization of these catalytic reactions with a broad range of substrates, higher turnover frequencies, and increased regioselectivity. A very fundamental example of one such limitation has been demonstrated by the multidisciplinary study undertaken by Dr. Brennan Billow and Dr. Tanner McDaniel during their time in the Odom Group. In this study, the rates of v arious Ti(IV) hydroamination catalysts were found to correlate with the electronic donor ability (LDP) and size (%V bur ) of the ancillary ligands on Ti. 22 With the electronic and steric parameters for each ligand determined utilizing our Cr(VI) LDP system, 23 the rate of a Ti - catalyst bearing a giv en ancillary ligand can be predicted using the equation shown below (Fig. 3.2). 272 Figure 4 . 2 The combination of the LDP system using high valent Cr(VI) and our Ti(IV) hydroamin ation catalysts which has facilitated the development of a quantitative model describing the effects of ancillary ligands on catalytic rate. According to this model, a more electron deficient ancillary ligand results in faster catalyst turnover, as demons trated by the positive coefficient for the LDP (electronic) term; as LDP increases, the X ligand is a poorer donor, and multiplied by the positive coefficient this increases k obs . This fits mechanistically with the proposed rate - limiting step of imine for mation via hydroamination being protonation of the intermediate metallacycle; a more Lewis acidic Ti may enhance the rate of that protonation step. Conversely, as the steric term gets larger, indicating more crowding in the first coordination sphere of Ti, the rate of the catalyst decreases. This is indicated by the negative coefficient modulating the effect of the steric descriptor on k obs . 22 While the successful development of this predicti ve model was a triumph, several complications resulted from these studies which highlight some of the limits of homogeneous systems. For example, based on its relatively high LDP value ( ~ 14 kcal/mol) and its small steric profile (%V bur ~ 23%), 2 - thionapthol was predicted to be an excellent ancillary ligand, resulting in 273 a fast k obs . However, when the Ti(NMe 2 ) 2 - di(thionapthol)) catalyst was prepared and the k obs measured experimentally, the performance was quite poor with next to no product formation. U pon further analysis it was determined that the complex exists as a thiol - bridged dimer in the solid state. Even in the presence of excess aniline at elevated temperatures, the dimerized form persists in solution,. 24 Consequently, a definitive form fo r the catalytically active species cannot be assigned, as one of two scenarios (or both at the same time) are likely occurring. If the active species is actually monomeric, its concentration in solution is fleeting and low due to the equilibrium process wi th the dimeric species. Again, by in situ experimental indications, the dimer is the major species in solution. If the active species is in fact the dimer, its kinetics are very different from the monomeric analogues that were used to build the model for t he study, and therefore the predictive capability of the model fails to account for these differences. Regardless of the active species, a ligand which was predicted to yield a fast k obs turned out to be the worst catalyst (slowest) examined in the study. Dimerization processes are by no means exclusive to the catalysts in this system and are a common issue across a wide variety of homogeneous transition metal catalyzed reactions, reducing the concentration of active species in situ . Easy ways to circumven t these processes, however, are not common and typically involve synthetic strategies related to ligand design that disfavor dimer formation (i.e. adding steric bulk to block molecular contacts between two molecular metal complexes in solution). Many oth er examples of ligand - design - limited alterations to homogeneous Ti(IV) catalysts for these and similar reactions include ligand non - innocence, ligand comproportionation and disproportionation reactions (see Chapter 6), and introduction of competing side re actions with radical changes in ligand design. These sorts of results are only discovered after careful ligand 274 design, synthesis, isolation, and attachment to the transition metal has occupied hours or days of a Many of the problems plaguing these homogeneous Ti systems have a simple solution. Dimerization and ligand exchange interactions are not problems in heterogeneous systems. In some types of heterogeneous catalysts, the active metal is immobilized on a solid support or within a porous 3 - dimensional matrix. In covalent materials like MOFs or many types of nanomaterials, itself making bonds to a metal. They are not nea rly as susceptible to the same types of exchange reactions and dynamic processes that a homogeneous catalyst might undergo. There are a handful of recent examples in the literature where heterogeneous catalyst systems have been used to perform hydroaminati on reactions, including nanoparticles and gel - supported systems. 25 - 27 One type of heterogeneous catalyst that has appeal for application to C N bond forming reactions is silica - gel supported organometallic catalysts. Over the last few decades , both the Coperet, Basset, and Scott groups (among others) have pioneered thorough characterization and mechanistic understanding of catalytically active organometallic complexes immobilized on silica surfaces via Si O M linkages. 28 - 34 For example, in collaboration with Schrock, the Coperet group has extensively characterized the properties and catalytic ability of a silica - gel supported high - olefin metathesis. 35 - 38 Even examples of silica - supported la nthanide catalysts and a silica supported Zn catalyst have recently been reported for intramolecular hydroamination. 39,40 This type of catalyst system has the potential to di rectly solve many of the problems that so heavily diminish the success of the homogeneous catalysts discussed above. Unlike many heterogeneous catalysts, however, silica - supported systems offer the benefit of maintaining a 275 similar coordination environment around the active metal to homogeneous analogues. The mechanistic and intuitive understanding built on homogeneous analogues over the last several decades, to the development of similar reaction chemistry with these heterogeneous materials. 35 It has also been demonstrated that silica - supported catalysts are good candidates for things l ike catalyst recycling and reactivation. 41 These are features we sought to extend to hydroamination and iminoamination catalysts but which are also relatively impracti cal with homogeneous systems. Put simply, heterogeneous catalysts are far more robust than homogeneous counterparts, so even simple things like higher reaction temperatures or higher reaction concentrations seemed like potential routes to improve catalyst performance, in terms of rate. With these potential benefits offered by moving to a catalyst system using a silica - support, we decided to explore the reactivity of silica - supported Ti materials. Our only synthetic requirement in these supported systems w as the preservation of two protolytically active sites in the Ti coordination sphere. Generally, we would expect a Ti O bond, such as those that would be produced by grafting Ti(NR 2 ) 4 precursors to a silica surface, are going to resist protonation by the r elatively weak proton sources such as the primary amines utilized in hydroamination and iminoamination reactions (pK a : H 2 NPh ~ 30; (Si)O - H ~ 5). With this forethought, we knew that a maximum of 2 surface interactions per Ti was likely needed to preserve the desired reactivity at the metal. We also suspected that there would be large differences in the reactivity of Ti sites bound through 1 versus 2 Ti O Si linkages. In the simplest organometallic view, each bond to the surface is like a large siloxide ligan d. When two of these interactions are achieved, the Ti then has a very 276 additional protolytically active ligand, such as an - NHPh fragment, in the reaction soluti on. The differences in this ligand set are dramatic both in terms of sterics and electronics. Keeping these ideas in mind, we set about preparing two different materials that would offer the two different surface site environments. The targeted species are shown below in Fig. 4.3 to assess the reactivity of such surface - bound species for C N bond forming reactions. Figure 4 . 3 The species that were targeted as potential precatalyst materials for hydroamination and iminoamination chemistry using a silica - supported, heterogeneous catalyst system. 4.2 Preparation of Silica - gels with Varied Surface Hydroxyl Group Density In 2001, the Scott group at UCSB published the preparation of a material that they characterized as a Ti(IV) bis(dimethylamido) complex bound to the silica surface through two Ti O Si linkages each. A depiction of this surface is shown in Fig. 4.4. 42 277 Figure 4 . 4 The binding mode of Ti on the surface of SiO 2 200 upon treatmen t with Ti(NMe 2 ) 4 . For comparison, the bulk properties of the material reported by the Scott group is given along with the properties of the material we synthesized following their SiO 2 200 preparation. - gel is str aightforward and requires minimal dehydroxylation. After compacting the fumed silica (surface area 200 ± 25 m 2 /gram) , it is heated to 200 ° C under vacuum for 8 h to provide SiO 2 200 a material which liberates 1.99 ± 0.04 equiv HNR 2 per mole of Ti(NR 2 ) 4 consumed, and a Ti content of 1.93 ± 0.4 wt% when saturated with Ti(NMe 2 ) 4 . 42 In our hands, this preparation led to a very similar material with 0.52 mmol Si O H sites per gram or 1.6 Si O H sites per nm 2 ; this correlates with liberation of 2.03 ± 0.12 e quiv HNR 2 per mole of Ti(NR 2 ) 4 consumed according to NMR titration of SiO 2 200 with Ti(NEt 2 ) 4 . When treated with Ti(NMe 2 ) 4 , our material shows 2.33 ± 0.12 wt% Ti by ICP - OES analysis. The properties of the two materials are the same within error. Preparati on of the silica - gel leading to a much lower concentration of terminal Si O H sites requires much harsher conditions. Fortunately, using a quartz tube fitted with a gas adapter ing literature procedures resulted in highly dehydroxylated silica - gel. 28 This involves heating compacted silica - gel under vacuum at a temperature of 700 ° C for several hours; repeat batches of SiO 2 700 prepared in this manner have demonstrated consistent results following the protocol outlined in Table 4.1 , bel ow. 278 Table 4 . 1 General conditions for SiO 2 700 preparation from commercially available fumed silica. Temperature (°C) Ramp rate (°C/min) Time (min) Atmosphere 20 - 500 5 96 Air 500 240 Air 500 720 Vacuum 500 - 700 1.33 150 Vacuum 700 480 Vacuum 700 - 20 180 Vacuum Figure 4 . 5 (top) Quartz tube used for silica gel prep with tube furnace setup. (bottom) Schematic for the preparation of the precatalyst material [Ti]700. Calibration of the surface with Ti(NEt 2 ) 4 and quantification of the Ti content of the SiO 2 700, when treated with Ti(NMe 2 ) 4 and measured by ICP - OES spectroscopy, provides values consistent with a low Ti concentration. The calibration stoichiometry suggests that, on average, each Ti is bound to the silica - gel surface through a single Ti O Si; these NMR titrations show that 0.98 ± 0.04 equiv of HNR 2 is liberated per equiv of Ti(NR 2 ) 4 consumed. The subsequent Si O H concentration was determined to be 0.31 ± 0.05 mmol per gram of SiO 2 700 or 0.9 ± 0.1 Si O H sites per nm 2 ; when saturated with Ti(NMe 2 ) 4 this correlates with 1.53 ± 0.0 7 wt% Ti. These properties provide characterization data for the material in terms of its bulk properties. Of course, 279 there is still the possibility that each individual Ti site on the silica - gel surface could possess one of several variations in the local Ti environment. Figure 4 . 6 Possible site variations in the coordination environment and binding modes for the two different catalysts, [Ti ]200 and [Ti]700. These possible differences in the local Ti environment on the silica surface could induce differences in the potential active catalytic species, however, they do not preclude catalysis (i.e. each Ti has at least 2 proteolytically cleavab le sites). Thus, we began screening this material, [Ti]700, as well as the [Ti]200 for catalysis. In the future, it would be very interesting to look at these and other relevant materials (vide supra) by advanced solid - state NMR techniques, which could pro vide insight related to Ti speciation on the surface. 30,43 This technology is not a vailable at MSU, and would require collaboration with an external group, such as a national lab. With this acknowledgement out of the way, note that our chemistry and rationale proceeds with the simplest assumption that the stoichiometry of the bulk materi al represents the average Ti site in the material. 4.3 Performance of [Ti]200 and [Ti]700 as Intermolecular Hydroamination Catalysts As a starting point for hydroamination reactions, similar conditions were used to those employed previously with homogeneous catalysts. Using 10 mol% catalyst loading at 110 ° C in toluene, both catalysts, [Ti]200 and [Ti]700, demonstrated slow production of the hydroamination product of aniline and 1 - hexyne. Since one of the main benefits of a heterogeneous catalyst over a homo geneous one is the increased stability, we increased the temperature of the reaction to 140 280 ° C. This dramatically accelerated the rate of the reactions in addition to increasing the with this dimerized form persisting in solution, regioselectivity observed in the products with both the [Ti]200 and [Ti]700 catalyst materials. With increased temperatures demonstrating such a marked improvement in catalyst performance, the temperature was again increased to 180 °C. The solvent for the reaction was changed fr om toluene to p - cymene, to retain similar properties but raise the boiling point of the solvent; likewise, the alkyne substrate was swapped for 1 - octyne so higher concentrations of alkyne would remain in solution rather than vaporizing into the reaction he adspace, provided by the increased boiling point. These changes provided another dramatic improvement in catalyst performance. The [Ti]200, for example, now performs the hydroamination of aniline and 1 - octyne in 98% yield in under 40 min with only 5 mol% c atalyst loading. Similar improvements were noted with the [Ti]700, as well. Fundamentally, needing higher reaction temperatures to achieve catalyst activation in a system with mixed - phase substrates and catalyst is not uncommon; interaction of catalyst and substrate across phases can increase the barrier to initiate reactions. 281 Figure 4 . 7 Reaction conditions applied to the heterogeneous catalysts. The yields and regioselectivities listed here were observed with [Ti]200. The same trends were observed with [Ti]700, as well, with improvements in regioselectivities and yields at higher temperat ures in dramatically shorter reaction times. Moving forward, the conditions for C were adopted as the general procedure. Using these conditions ( C ), a series of hydroamination reactions were performed with both catalysts to evaluate functional group tolerance in both the amine and alkyne substrates. Both catalysts demonstrate high yields for hydroamination with aniline, and bulky, electron - rich anilines (entries 1 - 4, 7, 8, and 11 - 14). This is true with both 1 - octyne and 1 - phenylpropyne as coupling pa rtners. Both catalysts also demonstrate high regioselectivities for these reactions, as well, showing selectivity to the limits of our detection in many cases (Table 4.2) . These results provided promising reactivity that suggested many of the reactions we were interested in pursuing with iminoamination (i.e. aniline as the amine coupling partner) could lead to high yields in regioselectivities in heterocyclic products down the line. 282 Table 4 . 2 Substrate scope examined for hydroamination reactivity with [Ti]200 and [Ti]700. The general conditions outlined in the Scheme below apply to all reactions. a,b a Yields reported were measured by GCFID using external calibrations of the reduced derivatives of the hydroa mination products shown in the figure. For those species with low yield, the next closest derivative was used to estimate the yield. b Times for each reaction varied, from under 40 min for entry 1 to 14 h for entry 4. For specific reaction times, see experi mental. 283 Both catalysts also show similar limitations. Specifically, tolerance toward alkyl amines and electron deficient aniline as coupling partners is low, with incredibly low yields; in fact, yields were so poor that no attempt was made to isolate p roducts from these reactions. The regioisomers drawn in the table, therefore, are representative examples rather than an assignment of the observed regioisomer preference for these reactions. Overall, catalyst performance for hydroamination is very similar between the two materials, [Ti]200 and [Ti]700. While there are many catalysts available in the literature that are capable of hydroamination, many at much milder conditions, we think this catalyst system offers a unique advantage. If imines are needed as intermediates in a multi - step synthesis, these r eactions are clean and high yielding with the [Ti]700/200 systems, such that in situ reactions from their imine products would likely proceed successfully. At the same time, use of a solid catalyst means that after the reaction is complete, removal of the catalyst requires only a simple filtration. The organic filtrate can then be conveniently taken on to the next step in a reaction pathway without fear of remaining contaminants from the HA catalyst in solution with the products. Especially given how easy i t is the make [Ti ] 200, these benefits make its use for even routine hydroamination catalysis attractive, especially as the first step in a multi - step sequence. 4.4 Application of [Ti]200 and [Ti]700 to Multicomponent Coupling Reactions In the previous secti on, both [Ti]200 and [Ti]700 show comparable performance in hydroamination reactions. When using these catalysts for iminoamination, the three - component coupling (3CC) of an amine, an alkyne, and an isonitrile, their abilities are vastly different. Similar to the hydroamination reactions, several sets of conditions were probed, and while the catalysts showed product formation at lower temperatures (110 ° C and 140 ° C), this product formation was 284 extremely slow. Thus, similar reaction conditions as those pres ented for hydroamination were adopted for iminoamination, as shown in Fig. 4.8 . These reaction conditions limit the choice of isonitrile coupling partner. At 180 ° C, isonitrile decomposition was noted with aryl and bulky alkyl (octyl, 1,1 - 2,2 - tetramethyl - propyl, and tert - butyl) groups on the isonitrile. Only CyNC, with its high boiling point and reduced propensity to undergo alkyl eliminations (which had been observed when reactions were attempted with t BuNC), was able to withstand the high reaction tempe ratures. With the [Ti]200 catalyst, however, even at elevated temperatures, relatively small amounts of product were observed with many sets of substrates. This is demonstrated by several of the entries in Table 4. 3 , below. Examining the series of reacti ons in entries 1 - 3, it seems that steric protection of the aniline derivative at either or both ortho - ring is necessary to promote formation of the 3CC product. This is because the side reaction to generate formamidine, wh ich is an off - cycle product, or the hydroamination products are competing with 3CC production. 6,8 The rates of side product formation are particularly high with an unprotected aniline, generating a very substantial amount of formamidine with these substrates. Typically, in homogeneous reactions, the amount of off - cycle formam idine production can be reduced with the use of t BuNC, which slows the rate of formamidine production relative to other processes in the iminoamination reaction. However, as noted above, this substrate had already proven incompatible with the reaction cond itions. As such, the bulky aniline derivatives are really the only substrates that are compatible with [Ti]200 for iminoamination chemistry. Switching catalyst to the [Ti]700, every aspect of the catalysis is improved. The [Ti]700 catalyst material gives even higher yields than the [Ti]200 catalyzed reactions with the bulky aniline derivatives, with entries 1 and 2, giving 94% and 72% yield, respectively, in 48 h. 285 Additionally, the [Ti]700 can perform couplings with 1 - phenylprophyne much more readily than the [Ti]200 material (entry 5), and it can even provide high yields of the 3CC product coupling aniline and 1 - phenylpropyne with CyNC. Note, this reaction also appears much faster, as 88% yield is generated in only 16 h. Overall, the iminoamination reacti ons catalyzed by [Ti]700 are much cleaner than any of the reactions catalyzed by the [Ti]200 as well, with very little formamidine or hydroamination observed by GCMS and GCFID of the crude reaction mixtures. There is only one real limitation remaining wit h the [Ti]700 as a catalyst for iminoamination: the catalyst does not couple small aniline derivatives (i.e. aniline or 4 - NMe 2 - aniline) with terminal alkylalkynes well (1 - octyne). These substrates show very low yields with substantial side product formatio n. The catalyst also falls short with alkyl amine derivatives such as CyNH 2 ; this is true even when considering just the simple hydroamination reaction, so this is a direct failure of the iminoamination reaction so much as an inherent substrate intoleranc e of these types of catalysts. To qualify [Ti]700 as a universal iminoamination catalyst for a broad range of substrates, these substrate intolerances need to be overcome. As an iminoamination catalyst for anilines with 1 - phenylpropyne or phenylacetylene, however, this material appears to perform with similar success to the best homogeneous catalysts for these reactions. This class of compounds alone is quite valuable for application to quinoline synthesis, for example, and certainly demonstrates the potent ial of this type of catalyst in performing these C N bond forming reactions quickly and cleanly. 286 Figure 4 . 8 General reaction scheme applied to the iminoamination reactions studied with [Ti]200 and [Ti]700. 287 Table 4 . 3 Iminoamination (3CC) substrate scope examined with [Ti]200 and [Ti]700. The general conditions outlined in Fig. 7, C apply to all reacti ons. a,b a Yields reported were measured by GCFID using external calibrations of the reduced derivatives of the hydroamination products shown in the figure. For those species with low yield, the closest derivative in terms of molecular formula, was used to estimate t he yield by GCFID. b Times for each reaction varied, from under 40 min for entry 1 to 14 h for entry 4. For specific reaction times, see experimental. 4.5 Use of Ti700 to Produce Functionalized Heterocycles To demonstrate the practicality of [Ti]700 as an i minoamination catalyst, the 1 - pot - 2 - step quinoline synthesis, previously developed by the Odom group with homogeneous Ti catalysts, was repeated here with [Ti]700. 9,44 The reaction scheme shown below produced the t argeted quinoline in 43% yield utilizing 5 mol% [Ti]700. The total reaction time of both steps together was 36 hours. 288 Scheme 4 . 1 One - pot - two - step quinoline synthesis of 2 - me thyl - 3 - phenyl - 6 - ( N,N - dimethylamino)quinoline utilizing [Ti]700 to perform the initial iminoamination reaction. Compared to the original synthesis of this quinoline, the synthesis of the quinoline utilizing [Ti]700 is, in many ways, an improvement. The 3C C reaction was catalyzed with 10 mol% of the homogeneous catalyst (Ti(NMe 2 ) 2 (dpm)) and took 24 hours to complete, followed by a 24 h conversion to the quinoline using acetic acid. This provided 50% yield of the targeted quinoline. The yields of the two Ti - catalyzed reactions are comparable, but the [Ti]700 catalyzed reaction took less time and half the catalyst loading. The results of the quinoline synthesis not only demonstrates that [Ti]700 performs well enough to be practically employed in routine synthe sis of heterocycles, but also shows that [Ti]700 performs these iminoamination based reactions on the same level as some of our best homogeneous variants. 4.6 Exploration of Catalyst Reusability and Routes of Deactivation As mentioned in the introduction, a nother appealing aspect of heterogeneous catalysts, specifically those supported on a silica or metal - oxide surface, is their potential to be recycled or reused. The physical ease with which the catalyst can be removed from a complex, solvated organic reac tion mixture facilitates recovery of the catalyst material, something which is a far greater challenge with homogeneous systems. Additionally, a recent example from the Sadow group demonstrates the ability to reactivate a silica - supported Zr catalyst with a mild reductant (HBpin) after it has been exposed to air. 41 289 Attempts were made with both [Ti]200 and [Ti]700 catalysts to recycle the material. Despite the similari ty in structure of the materials, two very different responses were noted for the when reusing them in subsequent catalytic reactions. With [Ti]200 catalyzing the hydroamination of aniline and 1 - octyne (entry 1, Table 1), the catalyst can be filtered off t he crude reaction mixture, washed with pentane, dried under reduced pressure, and added to a second hydroamination reaction. The second round provides a yield comparable to the first round. This process was repeated for a total of 5 uses of the catalyst, w ith the third round showing a slight decrease in yield; by the 4 th round, the yield has been reduced to about half of the original yield; and by round 5, the catalyst is close to complete deactivation, showing fewer than 5 turnovers relative to the origina l loading. Interestingly, as the yield gradually tapers off, the regioselectivity also decreases (by about 1 order of magnitude total) with each use of the catalyst. When the used catalyst material was 2 ) 4 after the 5 th use of the catalyst, the yield makes a dramatic relatively low (7.3:1 vs. >100:1) compared to the initial use. These results are highlighted in Table 4. 4 . Table 4 . 4 Hydroamination Results for the Coupling of 1 - Octyne and Aniline with Recycled [Ti]200. Trial Number Yield (%) a Regioselectivity Ratio 1 98 >100:1 2 96 32:1 3 90 14:1 4 44 10:1 5 23 10:1 6 b 71 7.3:1 a Reported yields are GC - FID yields. b Prior to running experiment 6, the catalyst material was treated with Ti(NMe 2 ) 4 290 surface that may be changing the nature of the catalyst, reducing regioselectivity and yield across multiple uses of the catalyst. One detrimental effect of catalyst recycli ng that we were able to identify is that over the course of a single hydroamination reaction, a measurable quantity of the Ti metal is leached from the surface of the catalyst. ICP - OES analysis of used [Ti]200, that has been through 1 round of hydroaminati on, demonstrates a 10% reduction in Ti (wt%) content. Successive uses of the same catalyst material strip more of the active metal away, leading to a decrease in catalyst loading with each reuse. This phenomenon agrees with the observations presented in th e table above, including a recovery in the reaction yield upon re - exposure of the SiO 2 200 support to a molecular Ti source. possible contribution to this process may be the a lkyne trimerization that appears as an off - cycle side reaction in hydroamination catalyses. One of the proposed steps in this type of catalyze bond to the sur face. This hypothesis is currently very speculative and is simply our best guess as to what is different between the [Ti]200 and the [Ti]700, which does not demonstrate loss of the metal (see below). The only aspect of the [Ti]200 recycling experiments th at is not necessarily explained by the Ti leaching from the catalyst material is the change in the regioselectivity noted. While a change in the catalyst loading could certainly alter the observed regioisomer ratio of the products, this change could also b e the result of other changes in the material. For example, if the Ti on the catalyst material is labile, it may be rearranging to form different coordination environments (Fig. 4. 9 ). One could even envision the formation of dimers, like those observed by Scott and coworkers 291 when they used Ti(O i Pr) 4 to synthesize Ti(O i Pr) 2 /SiO 2 200 . 45 typical in a heterogeneous system like this, we know that Ti comes off the surface during these reactions with [Ti]200. T he liberated Ti, once in solution, could exhibit behaviors more typical of homogeneous species. This type of rearrangement could certainly result in changes in the reactivity, including regioselectivity. Figure 4 . 9 Proposed surface morphologies of used [Ti]200, after use in hydroamination reactions. Changes in the coordination environment could lead to changes in the regiosele ctivity ratio observed in the hydroamination products. When we tried to reuse the [Ti]700 material, very different results were observed. There was no loss of Ti, determined by ICP - OES analysis of used catalyst material, within the error of our detection techniques. Despite this, only a trace (<5%) of the hydroamination or iminoamination products were observed upon a second use of the catalyst material. This was the case with a variety of different substrates examined for reuse with either reaction. Car catalyst poisoning. As of yet, we have no clear experimental indication of why the catalyst is ation. However, there is some precedence with homogeneous Ti systems that catalyze hydroamination reactions, to deactivate via the generation of thermodynamically stable, metallacyclic species. Specifically, Mountford and coworkers have demonstrated that d ouble - insertion of alkyne into the Ti imide double bond can result in an isolable product that is not readily cleaved from Ti in the presence of excess amine (i.e. a proton source). 21,46 While this product (aniline + 2 equiv alkyne) has not been 292 d irectly observed in the crude reaction solutions of any hydroamination reactions catalyzed by [Ti]700, if the product were irreversibly bound to the Ti on the silica surface, it may not appear in solution. Thus, the lack of GC evidence means that a deactivation event of this n ature is still possible in the [Ti]700 system. There is slightly more experimental evidence suggesting the source of the catalyst poisoning in the iminoamination. Again, there were no obvious indications of catalyst poisoning products in the crude reacti on mixtures by GCMS analysis. With [Ti]700 used for iminoamination reactions, surface extractions or washing with dilute HCl solutions provided evidence of new organic residues. While several of the new masses observed by GCMS remain unidentified, two mass es match those of 1 - phenyl - 3 - cyclohexylurea and 1,3 - dicyclohexylurea. This suggests that the isonitrile is potentially non - innocent toward the active Ti metal, as CyNC is the only source for a cyclohexyl functional group in the reaction. Unfortunately, w ith only these complexes identified in surface extractions, a working mechanism for the deactivation process has not been fully pieced - together. Additionally, because Scheme 4 . 2 A typical iminoamination reaction utilizing [Ti]700 as the catalyst material and the organic residue s found upon a mild acid - 293 an acidic, aqueous wash is required to remove the surface residues, it is likely that wha t is on the surface, actually poisoining the catalyst, is transformed upon removal by the species present in the HCl solution. Thus, although we have identified the urea species by GCMS, they may exist in a different form on the surface of the material. He re again, the application of solid - state NMR, or even (stringently) air - free IR spectroscopy would be highly beneficial to our understanding of these poisoning events. 4.7 Accidental Discovery of [Ti]700 Activity for Catalytic Guanylation of Carbodiimide Wh ile trying to figure out how the [Ti]700 is poisoned by iminoamination, we came across a few interesting studies in the literature. In 2003, Richeson and coworkers published two studies in which the interactions of carbodiimides and terminal Ti imido speci es were examined. In the first study, they observed metathesis of the C N double bonds of the carbodiimide with the Ti N double bond. 47 In a second study, catalytic production of substituted guanidines was achieved with a primary amine, a carbodiimide, and catalytic Ti imide species b earing guanidine ancillary ligands. 48 This catalytic reaction is depicted in Scheme 4. 3 . Similar guanylation reactions have been reported using alternate Ti imide systems since that time. Figure 4 . 10 Ti complexes from the Richeson Group used for catalytic guanylation of carbodiimides and imide metathesis reactions. 47,48 We wondered if in fact a guanidine - like species was able to form in the crude reaction mixture from CyN C and aniline, and if, subsequently, guanidine or a carbodiimide could shut down the catalytic activity of [Ti]700. A stoichiometric reaction between [Ti]700, carbodiimide, and an excess of aniline was heated for 2 h, and the crude solution analyzed by GCM S. Two 294 products were noted. First, a small amount of 1 - cyclohexyl - 3 - phenyl - carbodiimide was noted, the metathesis product of Ti=NPh and 1,3 - dicyclohexylcarbodiimide. Second, a very large amount of 1,2 - dicyclohexyl - 3 - phenylguanidine was also observed. When the reaction was repeated with 5 mol% [Ti]700, 1 equiv 1,3 - dicyclohexylcarbodiimide, and 1.2 equiv H 2 NPh, an isolated yield of 79% of the 1,2 - dicyclohexyl - 3 - phenylguanidine was obtained in 2 h of reaction time (Scheme 4.3) . Scheme 4 . 3 [Ti]700 catalyzed guanylation of 1,3 - dicyclohexylcarbodiimide. The fact that the [Ti]700 catalyst can produce guanidines catalytically from aniline and carbodiimides is an interesting discovery on its own and certainly supports exploratory reactivity studies with the catalyst material. For example, with this specific reaction, the homogeneous catalyst Ti(dpm)(NMe 2 ) 2 can also generate guanidine when combined with aniline and carbodiimide, but it cannot do so catalytically, as the dpm ligand is displaced by the guanidine product. This happens even stoichiometrically, where the crude reaction solution shows only H 2 dpm and the guanidine produced can only be observed after washing the reaction solution with water and performing an organic work - up. Because the [Ti]700 is supported on a silica surface, interactions between the product and the Ti metal are only of concern if they constitute irreversible bonding interactions. In the case of something like a guanidine product, however, binding of the gua nidine to Ti and its removal by protonation are likely equilibria processes, that appear to avoid shutting down the catalyst irreversibly. Thus, this direct comparison in reactivity between a homogeneous catalyst and the heterogeneous [Ti]700 system demons trates one of the major strategic advantages of the [Ti]700 295 material as a C N bond forming catalyst. Exploration of other reactions where the product may be destabilizing to a homogeneous catalyst could be similarly productive with [Ti]700. With the knowl edge that our catalyst can generate guanidine following the pathway demonstrated by Richeson, we returned to the original hypothesis, that guanidine could be generated in the reaction as a side product and that it could inhibit the catalyst from performing iminoamination. Upon addition of 20 mol% guanidine to a 3CC reaction (Entry 6, Table 2), we did note a decreased yield of the 3CC product over the same reaction time. Relative to entry 6, where 88% yield was observed, the same reaction with 20 mol% 1,3 - di cyclohexyl - 2 - phenylguanindine added provides a yield of only 45% in the same amount of time. While this suggests that guanidine does in fact inhibit the reaction, presumably by reducing the concentration of active Ti sites available by acting as a chelatin g ligand, it does not shut down catalysis completely. Since the recycling experiments demonstrate complete deactivation, we concluded this was not likely the source of the catalyst poisoning (or at least not the major source). One other aspect of the cata lytic reaction mixture that seemed necessary to consider since we have experimental evidence suggesting that isonitrile contributes to the poisoning phenomena in the 3CC reactions is an interaction between H 2 NPh and CyNC, perhaps mediated by Ti but unique from the pathway through which formamidine is generated. In recent years, reports of these types of reactions have been made. In 2015, Ji and coworkers published a report in which a primary amine and an isonitrile can be coupled to generate carbodiimides i n up to 93% yield using catalytic I 2 and an oxidant (cumene hydroperoxide, CHP). 49 Using a similar approach, Bez reported in 2018 that ureas can be synthesized directly through an I 2 mediated coupling of an amine and an isonitrile, using DMSO as oxidant. 50 One step in this proposed pathway includes an off - cycle equilibrium between one of the intermediates and carbodiimide. 296 To get an idea if this type of process could be responsible for the poisoning we observe in the [Ti]700 system, we looked at the potential of carbodiimide to inhibit the 3CC reaction. S imilar to the experiment with guanidine, described above, 0.2 equiv of CyNCNCy was added to the reaction mixture to generate 3CC Entry 6 ( Table 4.2 ). Here again, some product was observed, but with carbodiimide in the reaction mixture, a yield of only abou t 10% was observed, relative to 88% without carbodiimide. This is a much more dramatic reduction in catalyst activity than was observed with the guanidine added directly. This evidence, when considered with the observation that urea species were removed fr om the surface of the used catalyst upon a mild acid wash, makes the formation of carbodiimide - like intermediates forming in the reaction mixture seem like a probable source of catalyst deactivation. Obviously, many details of how this process may occur a re unclear. Specifically, in our system, there is no clear source of oxidant, aside from the SiO 2 support itself. At the same time, generation of these poisonous species in no way appears catalytic; so perhaps we are witnessing our catalyst material taking the reactants down one of these coupling pathways (amine + isonitrile) where the whole process gets stuck prior to oxidation of an unsaturated intermediate species. Of course, much of this is speculative, and short of stringent 13 C NMR on the surface pre - and post - reaction, it is difficult to draw many well - founded conclusions. Further efforts to access recyclability with the catalyst material may then meet with the greatest success through condition optimization efforts, whereby the conditions for poisoni ng would be avoided altogether. 4.8 Conclusions - supported Ti catalysts for C N bond formation. The easily prepared [Ti]200, which requires only access to a basic vac uum - oven is an excellent hydroamination catalyst for aromatic primary amines 297 and a variety of alkynes. These reactions provide high yields and regioselectivities with these substrates and offer the benefit of easily removed catalyst material. For both the rapid synthesis of imines and the synthesis of imines for further functionalization, [Ti]200 offers an attractive means of producing the desired product. Additionally, these reactions can be performed with lower catalyst loadings than similar homogeneous s ystems. selective for production of the 3CC product over other possible byproducts in the reaction. While the best results were achieved with bulky derivatives ( i.e. entries 1 and 2 above , Table 4.2), even with these substrates, large amounts of the hydroamination and formamidine byproducts were still observed in relatively high concentrations in the reaction mixtures. This makes the prospect of using the catalyst for these iminoamination reactions much less attractive. By contrast, the [Ti]700 catalyst is average in terms of its hydroamination ability, it is a good catalyst for iminoamination reactions between aromatic primary amines and aromatic alkynes. The pr oduct yields are moderate to high and the byproduct yields are low. Additionally, with aniline as the amine and aromatic alkynes, these reactions are much faster than previously thought, with reaction times around 16 h showing high yields and almost comple te consumption of the starting materials. Using iminoamination catalyzed by [Ti]700, we were able to demonstrate the utility of this catalyst system for practical synthesis of heterocycle synthesis. Following the 1 - pot - 2 - step synthesis of quinolines which we developed with homogeneous Ti catalyst, the [Ti]700 system was able to provide a similar final yield with a faster total reaction time and half the catalyst loading. These results demonstrate that even a very simple silica - supported Ti catalyst is capab le of improving upon the known homogeneous systems. 298 Of course, even with the [Ti]700 system, shortcomings have still been discovered. Primarily, we would still like to overcome the substrate limitations (sterically unprotected anilines coupled to 1 - octyn e and tolerance of alkyl primary amines). Achieving a highly reusable catalyst, that provides high yields and regioselectivities across several uses of the same batch of catalyst material is still an end goal in switching to a heterogeneous catalyst system . These goals encouraged us to explore further modification of the catalyst that enhanced its performance (Chapter 5). 4.9 Experimental General Considerations All syntheses and handling of materials were carried out under an inert N 2 atmosphere, either in an MBraun glovebox or by standard Schlenk technique. Any handling of materials in air is specified. Generally, this was limited to column chromatography and preparation of some GC and all ICP samples. Fumed SiO 2 was purchased from Sigma Aldrich and used as received (200 ± 25 m 2 /g, Lot # SLBT0198). The following solvents were purchased commercially and dried prior to use: para - cymene was dried over CaH 2 and distilled under vacuum prior to use; pentanes and toluene were dried by passage over activated alumina and sparged with N 2 prior to use; tetrahydrofuran was purchased commercially, dried over sodium, and distilled under N 2 prior to use. The NMR solvent C 6 D 6 was purchased from Sigma Aldrich, dried over CaH 2 , and distilled under N 2 prior to use. All dried sol vents were stored in an N 2 glovebox until use. For routine isolated product characterization, CDCl 3 was purchased from Cambridge Isotopes and used as received. Ti(NMe 2 ) 4 and Ti(NEt 2 ) 4 were purchased from Gelest and used as received. Aniline, 2,6 - dimethylaniline, 2,5 - dimethylaniline, 3,5 - dimethylaniline and 3,5 - bis(trifluoromethyl)aniline 299 were purchased commercially, dried over an appropriate drying agent (see purification of laboratory chemicals, 7 th ed.), and distilled under vacuum prior to use. The amines NH 2 Cy and 1 - NH 2 Ad were purchased commercially and dried before use. The alkynes phenylacetylene, 1 - phenylpropyne, and 1 - octyne were purchased from Alfa and dried over Na 2 SO 4 , then di stilled under N 2 before use. Diphenylacetylene was recrystallized from dry solvents before use. Cyclohexylisonitrile was prepared according to literature procedures. 51 SiO 2 200 was prepared 42 NMR Titration A J - Young tube was loaded with SiO 2 200/700 (approx. 100 mg), Ti(NEt 2 ) 4 , and hexamethyldisiloxane internal standard as a 2.0 mL solution in C 6 D 6 . The tube was sealed and transferred to a sonicator. The mixture was sonicated for approximately 1 h, and the solids allowed to settle. The mixture was then examined by 1 H NMR (gain = 36, relaxation delay = 30 s) and the integral for NEt 2 H versus Ti(NEt 2 ) 4 in solution was evaluated relative to the internal standard. This allowed determination of how much NHEt 2 had evolved from the reaction of Ti(NEt 2 ) 4 , and by correlation, terminal Si - OH site quantity on the silica surface. This also indicated the surface densi ty of Ti, and by comparing NHEt 2 generated versus Ti(NEt 2 ) 4 consumed, the average binding mode for the Ti was determined. The surface loading based on this titration was verified by ICP - OES. ICP - OES Analysis for Ti Content Sample preparation consisted of d igesting a known mass of [Ti]200 or [Ti]700 catalyst (approx. 100 mg) in 2 mL of concentrated HNO 3 . The digests were allowed to sit for 4 h under ambient conditions before dilution with deionized water and centrifuging. The liquid portion of the sample was removed, and the solids washed with 2 aliquots of deionized water, which was again centrifuged, and the liquid portions collected. The combined liquid portions were diluted to a known volume, and further dilution with a 2% HNO 3 was carried 300 out as needed t o get sample concentrations within the limits of detection of the instrument (0 - 6 ppm for Ti). A 1000 ppm Ti ICP standard in 2% HNO 3 was purchased from Sigma - Aldrich and used as received to prepare an external calibration curve. The [Ti]200/700 samples were then measured in triplicate, and quantified from the external calibration, allowing for the mass of titanium in each sample to be determined. Preparation of SiO 2 700 An OTF - 1200X - S (MTI Corporation) high temperature furnace was utilized in the preparation of the SiO 2 700 . The fumed SiO 2 purchased from Sigma, was poured into a 1 L beaker. To this was added DI water, and this mixture was stirred until it formed a homogenous slurry. This slurry was air - dried for 48 h, and then transferred to a 140 °C glassware oven for an additional 48 h. The resulting SiO 2 was compact and clumpy. The material was then ground with a mortar and pestle until a finely divided, free - flowing powder resulted. 15 g of this material could be loaded into a quartz tube closed at one end and fitted with a gas - adapted ball - and - socket joint at the other end (borosilicate glass). The loaded quartz tube was placed in the tube furnace, taking care to center the SiO 2 over the heating element. The heating and atmosphere protocol listed in Fig. 4.5 was then followed to de - hydroxylate the SiO 2 . After the tube was cooled to room temperature, it was sealed under vacuum and transf erred to an N 2 atmosphere glovebox. The material was stored in sealed containers in the glovebox until further use. Ti(NEt 2 ) 4 titrations of the material following reported procedure (above) provide a surface density Si - OH determination of 0.00031 ± 0.0000 5 mol/g SiO 2 700 or 0.90 ± 0.11 Si OH sites/nm 2 . This correlates to a predicted Ti loading of 1.46 ± 0.12 wt %. Preparation of [Ti]200 From Ti(NEt 2 ) 4 titrations, the surface abundance of Si - OH sites was estimated (0.00052 mol/g). A 125 mL Erlenmeyer flask w as charged with 4g of SiO 2 200 and 30 mL of pentanes. The slurry was stirred and 1.2 equiv of Ti(NMe 2 ) 4 (560 mg) was added 301 dropwise. The mixture was stirred for 4 h at room temperature during which time the colorless silica turned yellow. The solids were co llected by filtration, rinsed with 20 mL of benzene, and dried under vacuum. The material was stored in an N 2 glovebox in a sealed container and used as needed. ICP - OES : 2.33 % Ti ( ± 0.12). Preparation of [Ti]700 From Ti(NEt 2 ) 4 titrations the surface abundance of Si - OH sites was estimated (0.00031 mol/g). A 125 mL Erlenmeyer flask was charge with 6 g of SiO 2 700 and 30 mL of pentanes. The slurry was stirred and 1.2 equiv of Ti(NMe 2 ) 4 (500 mg) was added dropwise. The mixture was stirred for 4 h a t room temperature during which time the colorless silica turned yellow. The solids were collected by filtration, rinsed with 20 mL of benzene, and dried under vacuum. The material was stored in an N 2 glovebox in a sealed container and used as needed ICP - O ES: 1.50 % Ti ( ± 0.07). General Hydroamination Procedure with [Ti]200 A pressure tube was charged with 5 mol % [Ti]200 (100 mg) and a Teflon stir bar. A 1.0 mL solution containing 1 mmol NH 2 R and 1 - 2 mmol Alkyne in p - cymene was then added to the pressure t ube, which was sealed and transferred desired amount of time with magnetic stirring. Upon reaction completion, the pressure tube was centrifuged, compacting the [Ti]200 into a pellet at the bottom of the tube. This leaves a transparent yellow to orange solution which was decanted and used for GC/MS or GC/FID analysis. General 3CC Prodcedure with [Ti]200 A pressure tube was charged with 5 mol % [Ti]200 (100 mg) an d a Teflon stir bar. A 1.50 mL solution containing 1 mmol NH 2 R, 1 - 2 mmol alkyne, and 1.5 mmol CyNC in p - cymene was then added to the pressure tube, which was sealed was heated for the desired amount of time with magnetic stirring. Upon reaction completion, the 302 pressure tube was centrifuged, compacting the [Ti]200 into a pellet at the bottom of the tube and leaving a transparent yellow to orange solution which was dec anted and used for GC/MS or GC/FID analysis. General Hydroamination Procedure with [Ti]700 A 15 mL pressure tube was charged with 163 mg Ti(NMe 2 ) 3 /SiO 2 700 (5 mol%), a stir bar, and 1.0 mL p - cymene. Separately, a volumetrically prepared 1.0 mL solution of 1 mmol NH 2 R and 2 mmol alkyne was prepared. This solution was added to the catalyst mixture in the pressure tube, which was sealed and transferred from the glovebox to a preheated aluminum block (180 ° C). The reaction was heated with magnetic s tirring for 40 min - 12 h. The pressure tube was ambiently cooled to room temperature, and centrifuged to compact the Ti(NMe 2 ) 3 /SiO 2 700 into an orange - brown pellet at the bottom of the tube. The pressure tube was transferred back to the glovebox and the liqu ids decanted. The crude solution was utilized for GC analysis. General 3CC Procedure with [Ti]700 A 15 mL pressure tube was charged with 163 mg Ti(NMe 2 ) 3 /SiO 2 700 (5 mol%), a stir bar, and 1.0 mL p - cymene. Separately, a volumetrically prepared 1.0 mL solu tion of 1 mmol NH 2 R, 1.5 mmol CyNC, and 2 mmol alkyne was prepared. This solution was added to the catalyst mixture in the pressure tube, which was sealed and transferred from the glovebox to a preheated aluminum block (180 °C). The reaction was heated wit h magnetic stirring for 12 - 36 h. The pressure tube was ambiently cooled to room temperature, and centrifuged to compact the Ti(NMe 2 ) 3 /SiO 2 700 into an orange pellet at the bottom of the tube. The pressure tube was transferred back to the glovebox and the li quids decanted. The crude solution was utilized for GC analysis. General Procedure: Recycling Experiments The general procedure for HA or 3CC was followed for set - up of the initial reaction with either [Ti]200 or [Ti]700. Upon completion of the 303 initial r eaction (run 1), the pressure tube was cooled ambiently, centrifuged, and returned to the glovebox. The catalyst material was collected by filtration and thoroughly rinsed (20 mL aromatic solvent followed by 20 mL hexane). The catalyst material was dried u nder vacuum and loaded into a new pressure tube with a stir bar. The reagent solution for a second reaction (run 2) was then loaded into the tube. The tube was sealed and returned to heat (180 °C, aluminum well plate), and the process repeated as many time s as needed. The filtrate solutions were reserved for GC analysis. General Notes about Quantification and Product Characterization: Reduced derivatives of the hydroamination products were isolated from catalyzed reactions with [Ti]200. These isolated der ivatives provided evidence for the favored regioisomer (NMR) and allowed for verification of retention times by GC - MS. The isolated derivatives were also utilized to generate GC - FID calibration curves of the hydroamination products, which allowed for furth er quantification of hydroamination yields by GC analysis in crude reaction solutions and products identification by GC retention times. For those hydroamination products which could not be isolated from the catalyzed reaction mixtures due to poor yield, q uantification was performed using the next closest isolated hydroamination derivatives. Similarly, iminoamination products were quantified in the crude reaction mixture using GC - FID analysis. Both 3CC1 and 3CC5 were successfully isolated from these catal yzed iminoamination reactions. The compounds were characterized by NMR, which allowed for regioisomer identification. They were used to provide GC - FID calibration curves, allowing for further quantification of similar 3CC derivatives in the crude reaction solutions, using GC - FID analysis. Retention times confirmed product identity by GC - MS. Reduction of Hydroamination Products and Isolation of Amine Derivatives HA products (with which GC - FID calibrations were performed) were isolated from 3 mmol scales of 304 the starting amine, following the General Procedure listed in the Experimental section. The resulting imine products were reduced to their respective amine derivatives for isolation. In a glovebox, the crude reaction mixture was decanted from the solid cat alyst, which was rinsed with 5 mL of THF. The THF wash and crude p - cymene solution were combined with 2 equiv of Na[B(CN)H 3 ] (6 mmol) and another 10 mL of THF in a Schlenk flask, and the reaction was transferred to a Schlenk line. The reaction mixture was stirred and 15 mol% p - toluenesulfonic acid was added. The reaction was stirred 4 h at room temperature, after which time 5 mL of 2 M HCl solution was added and stirring was continued for an additional hour. The solution was neutralized (pH 7 - 8) and extract ed with Et 2 O (3 x 30 mL). The volatiles were removed from the extracted organic layer by rotary evaporation to give a viscous yellow oil, which was purified by column chromatography (typically silica gel basified with NEt 3 using a hexane/EtOAc gradient as eluent. 305 Amine Derivative for Hydroamination Entry 1: 52 1 3 ): 7.17 (dd, J = 8.6, 7.3 Hz, 2H), 6.67 (t, J = 7.2 1H), 6.61 6.57 (m, 2H), 3.46 (sextet, J = 6.2 Hz, 1H), 1.63 1.52 (m, 1H), 1.47 1.23 (m, 11H), 1.19 (d, J = 6.3 Hz, 3H), 0.89 (t, 3H). 13 3 ): 147.83, 129.39, 116.84, 113.17, 48.58, 37.38, 32.01 , 29.53, 26.29, 22.82, 20.96, 14.27. Yield: 51% (yellow oil) Scheme 4 . 4 HA Entry 1 synthesis and isolation. 306 Amine Derivative for Hydroamination Entry 2: 53,54 1 3 major isomer A (aliphatic): 6.80 6.71 (t, J = 7.3 Hz, 3H), 6.68 (d, J = 8.7, 2H), 3.91 3.76 (m, 1H), 2.99 (dd, J = 13.4, 4.7 Hz, 1H), 2.75 (dd, J = 13.4, 7.3 Hz, 1H), 1.20 (d, J = 6.4 Hz, 3H) . 1 3 minor isomer B : 7.15 7.10 (m, 2H), 6.56 (d, J = 7.6 Hz, 2H), 4.27 (t, J = 6.7 Hz, 1H), 1.94 1.72 (m, 2H), 1.00 (t, J = 7.4 Hz, 3H). Aromatic region of isomeric mixture also shows the following overlapping multiplets 7.41 7.31 (m), 7.29 7.26 (m), 7.26 7.22 (m) . 13 3 ): isomeric mixture 126.50, 126.30, 117.19, 117.12, 113.36, 113.24, 59.72, 49.34, 42 .30, 31.71, 20.22, 10.89. Yield: 33% (yellow oil) Scheme 4 . 5 HA Entry 2 synthesis and isolation. 307 Amine Derivative for Hydroamination Entry 3: 55 HA3 red : Both isomers of HA3 red have been previously reported in the literature. Difficulties were encountered in isolating one or both regioisomers of HA3 red from the alkyne trimerization byproduct that inevitably forms due to the excess of alkyne needed to force the reaction to comple tion. After repeated column chromatography, the major isomer was determined from matching peaks in 1 H NMR to reports by Kato, et. al., referenced to the retention times of the two isomers in the crude reduction mixture. Identification is shown below, by GC MS and 1 H NMR identification in the impure mixture. This sample was not utilized for in situ quantification of the yield of this reaction. Rather it served only to identify the major regioisomer. 1 H NMR (500 MHz, Chloroform - d): 7.36 7.32 (m, 2H), 7.28 (t, J = 8.5, 6.8 Hz, 2H), 7.20 (m, J = 5.5, 1.8 Hz, 1H), 7.06 (t, J = 7.3 Hz, 2H), 6.61 (t, J = 7.3 Hz, 1H), 6.52 6.43 (m, 2H), 4.52 4.40 (m, 1H), 3.99 (s(br),1H), 1.48 (d, J = 6.7 Hz, 3H). Scheme 4 . 6 HA entry 3 synthesis and isolation. 308 Hydroamination Entry 4 (Imine product): 56 1 H NMR (500 MHz, CDCl 3 ): 7.97 (dd, J = 8.0, 1.7 Hz, 2H), 7.40 (d, J = 7.7 Hz, 3H), 7.37 7.31 (t, 2H), 7.25 (s, 2H), 7.21 7.14 (m, 1H), 7 .12 (m, 3H), 6.90 6.85 (d, J = 7.0 Hz, 2H), 4.12 (s, 2H). 13 3 ): 166.27, 151.11, 138.32, 137.14, 130.39, 129.03, 128.65, 128.43, 128.38, 128.11, 126.30, 123.43, 119.14, 36.27. Yield: 57 %, pale yellow crystalline solid Scheme 4 . 7 HA entry 4 synthesis and isolation 309 Amine Derivative of Hydroamination Entry 7: 57 1 CDCl 3 J = 7.5 Hz, 2H), 6.79 (t, J = 7.5 Hz, 1H), 3.31 3.22 (m, 1H), 2.82 (s, 1H), 2.26 (s, 6H), 1.62 1.48 (m, 1H), 1.45 1.35 (m, 3H), 1.32 1.24 (m, 7H), 1.05 (d, J = 6.3 Hz, 3H), 0.89 (t, J = 6.8 Hz, 3H). 13 3 128.86, 128.79, 121.00, 52.42, 38.48, 31. 87, 29.51, 26.45, 22.64, 21.37, 19.11, 14.10. Yield: 34% (yellow oil) Scheme 4 . 8 HA entry 7 synthesis and isolation. 310 Amine Derivative of Hydroamination Entry 8: 58 1 3 7.27 (m, 2H), 7.24 7.16 (m, 2H), 7.00 (d, J = 7.4 Hz, 2H), 6.82 (t, J = 7.5 Hz, 1H), 3.66 3.41 (sextet, J = 6.2 Hz, 1H), 2.95 (dd, J = 13.0, 4.7 Hz, 2H), 2.56 (ddd, J = 12.9, 8.5, 1.8 Hz, 1H), 2.25 (s, 6H), 1.06 (d, J = 6.3 Hz, 3H). 13 C NMR (126 3 126.23, 121.53, 54.18, 44.53, 20.93, 19.19. Yield: 52% (yellow oil) (1 mmol scale) Scheme 4 . 9 HA entry 8 synthesis and isolation. 311 Amine Derivative of Hydroamination Entry 12: No spectral data for HA12 red was found reported in the literature. The spectral data support the structural assignment shown below. 1 H NMR (500 MHz, Chloroform - d): 6.33 (s, 1H), 6.22 (s, 2H), 3.43 (pent, J = 6.3 Hz, 1H), 3.33 (s, 1H), 2.23 (d, J = 2.1 Hz, 6H), 1.63 1.45 (m, 2H), 1.42 1.21 (m, 14H), 1.15 (d, J = 6.3 Hz, 3H), 0.89 (m, 6H). 13 C NMR (126 MHz, Chloroform - d): 147.77, 138.87, 118.72, 110.93, 48.33, 37.31, 31.84, 29.36, 26.13, 22.64, 21.53, 20.89, 14.11. EI - MS: m/z HA12 231 (base 160); HA12 re d 233 (base 148). Yield: 120 mg (1 mmol scale), 52 % (yellow oil). Scheme 4 . 10 HA entry 12 synthesis and isolation. 312 Amine Derivative of Hydroamination Entry 13: 1 H NMR - d 6 ): 6.77 (d, J = 8.8 Hz, 2H), 6.58 (d, J = 8.8 Hz, 2H), 3.31 (sextet, J = 6.2 Hz, 1H), 2.61 (s, 6H), 1.47 1.37 (m, 2H), 1.34 1.19 (m, 10H), 1.03 (d, J = 6.2 Hz, 3H), 0.90 (t, J = 7.0 Hz, 3H). 13 - d 6 ): 144.43, 140.81, 116.32, 115.40, 49.69, 42.20, 37.69, 32.31, 29.89, 26.60, 23.09, 21.10, 14.39, 1.44. Yield: 19 % (reddish oil) . Scheme 4 . 11 HA entry 13 synthesis and isolatio n. 313 Amine Derivative of Hydroamination Entry 14: 1 - d 6 ): ( major A) 7.23 (d, J = 7.9 Hz, 1H), , 7.12 7.04 (m, 4H), 6.76 (d, J = 8.2 Hz, 2H), , 6.58 (d, J = 8.4 Hz, 2H), 3.58 (sextet, J = 6.0 Hz, 1H), 2.77 (dd, J = 13.3, 4.7 Hz, 1H), 2.62 (s, 6H), 2.46 (dd, J = 13.6, 7.6 Hz, 1H), 0.95 (d, J = 6.3 Hz, 3H). 1 H NMR - d 6 ): ( minor B) 7.14 (s, 2H ( 5 )), 6.62 (d, J = 8.7 Hz, 0.67H ( 2 )), 6.51 (d, J = 8.3 Hz, 0.57H ( 2 )), 4 .07 (t, J = 6.7 Hz, 0.29H ( 1 )), 2.53 (s, 1.65H ( 6 )), 1.55 (quintet, J = 7.5 Hz, 0.65H ( 2 )), 1.18 (dt, J = 6.9, 1.9 Hz, 0.22H ( 1 )), 0.78 (t, J = 7.4 Hz, 0.90H ( 3 )). 13 C NMR - d 6 ) isomeric mixture: 144.18, 144.06, 140.02, 139.74, 138.9 8, 129.50, 128.29, 128.15, 127.96, 126.58, 126.01, 115.82, 115.60, 115.25, 114.90, 60.40, 50.20, 42.13, 41.68, 41.64, 31.60, 19.89, 10.54. Yield: 24% (orange oil) Scheme 4 . 12 HA entry 14 synthesis and isolation. 314 Isolation of 3CC1 Isomer: A sample of one of the isomers of 3CC1 was isolated from the reaction of Ti(NMe 2 ) 2 dpm (20 mol%) as catalyst, 2,6 - dimethylphenylaniline (1) (1 equiv, 3 mmol), 1 - octyne (1 equiv, 3 mmol), and cyclohexylisonitrile (1.5 equiv, 4.5 mmol) in toluene at 110 °C. The crude reaction mixture was concentrated, and 3CC1 was separated by column chromatography (basified alumina, gradient pentane:Et 2 O) to yield 235 mg of the p roduct. This sample was used to quantify GC/FID yields for the other 3CC compounds synthesized by [Ti]200. The NMR and MS characterization are consistent with similar derivatives isolated in previous studies by our group. Yield: 235 mg, (24%, orange oil). 1 H NMR (500 MHz, CDCl 3 ): 9.89 (s, 1H), 7.02 (d, J = 7.4 Hz, 2H), 6.85 (t, J = 7.5 Hz, 1H), 6.71 (d, J = 7.8 Hz, 1H), 4.68 (d, J = 7.7 Hz, 1H), 3.00 (s, 1H), 2.05 (s, 6H), 1.93 - 1.84 (m, 3H), 1.75 - 1.68 (m, 2H), 1.55 (m, 2H), 1.41 - 1.08 (m, 18H), 0.94 - 0.87 (m , 2H), 0.82 (t, J = 7.1 Hz, 3H). 13 C NMR (126 MHz, CDCl 3 ): 170.52, 149.54, 145.16, 128.46, 127.65, 121.77, 90.73, 57.00, 34.63, 34.38, 31.64, 29.53, 27.74, 25.64, 24.84, 22.57, 18.71, 14.17. EI - MS: m/z 340 (base 255). Assessment of purity for use as GCFID quantification sample: for C 23 H 36 N 2 (3CC product above): C, 81.12; H, 10.66; N, 8.23. Based on 1 H and 13 C NMR, an impurity of 13 mol% triethylamine (TEA) is evident in the isolated 3CC product, as a result of basifying the column material used to separate the crude reaction mixture. Even after extended time under reduced pressure, the TEA remains present in the 3CC sample, likely due to hydrogen bonding interactions with the 3CC amine moiety. This amount of TEA impurity correlates to a modified molecular formula of C 23 H 36 N 2 ·0.13(C 6 H 15 N). Comparison of the anticipated elemental analysis results with this impurity versus the experimentally determined elemental analysis is 23 H 36 N 2 ·0.13(C 6 H 15 N): C, 80.75; H, 10.81; N, 8.43. 315 Found: C, 80.34; H, 10.52; N, 7.99. The results match closely with the molecular formula including the proportional amount of TEA observed by NMR. The FID calibration samples were adjusted for this impurity. 316 Isolation of 3CC5: An authentic sample of one of the regioisomers of 3CC5 was isolated in a similar way as described above. This compound was used to calibrate the GCFID response for 3CC5. Characterization data for this product are shown below. Isolated 3CC Isomer of Iminoamination Entry 5: Yield: 185 mg (53%, 1 mmol). 1 H NMR (500 MHz, CDCl 3 ) 10.42 (s, 1H), 7.32 (t, J = 7.6 Hz, 2H), 7.28 7.23 (m, 2H), 7.23 7.17 (m, 1H), 7.08 (d, J = 7.5 Hz, 2H), 6.90 (t, J = 7.5 Hz, 1H), 6.85 (s, 1H), 3.07 (sextet, J = 9.2, 4.5 Hz, 1H), 2.13 (s, 6H), 1.98 1.87 (m, 2H), 1.79 1.70 (m, 3H), 1.60 (s, 3H), 1.31 (m, 6H). 13 C NMR (126 MHz, CDCl 3 ) 165.82, 150.63, 147.53, 142.32, 130.52, 128.68, 127.98, 125.36, 122.31, 121.58, 107.65, 57.81, 34.52, 25.50, 24.96, 24.62, 20.33. EI - MS: m/ z 346 (base 331). Elemental Analysis 24 H 30 N 2 : C, 83.19; H, 8.73; N, 8.08. Found: C, 82.82; H, 8.56; N, 7.62. 317 Investigations into Catalyst Deactivations After running a 3CC reaction following the general procedure, the contents of the pressure tube were centrifuged to compact the catalyst material into a tight pellet at the bottom of the tube. In air, the organic phase was removed, and the catalyst material washed several times with hexanes (3 * 5 mL), vortexing and centrifuging with each wash. The catalyst material was air - dried and then washed with 3 mL of a dilute (10%) HCl solution. The HCl aliquot was then neutralized to pH 7 - 8 with sodium bicarbonate solution, and extracted with Et 2 O, followed by EtOAc. The organic extracts were combined a nd examined by GCMS. The GC results show several new compounds in the HCl wash not observed in the organic phase while some remain unidentified, two masses appear with m/z 218 and 224. In particular, the mass of 218 was determined to closely match that of 1 - phenyl - 3 - cyclohexyl - urea. The urea was independently synthesized by literature procedures, 59 and the retention time and fragmentation pattern determined on the same instrument as the authentic samples. The synthesized urea and the peak observed with the same mass in the HCl wash display very similar retention times and fragmentation patterns. Thus, we believe this is likely the identity of the organic residue. Likewise, the m/z 224 peak has the same mass as 1,3 - dicyclohexylurea. GC/MS fragmentation patterns are shown below ( Fig. 4.44 - 4.46 ). While other species are also examined in the HCl wash extract of the catalyst material, from the identification of the urea species, and the observation that both cyclohexyl and phenyl groups are involved, we know that an off - cycle interaction between anil ide and the cyclohexylisonitrile are leading to new and previously undetected reactivity. Also of note is the fact that these species are not observed in the hexanes washes of the catalyst material. Only once an aqueous acid is added to the catalyst materi al do these species appear, which suggests that they 318 are tightly bound to the surface. The types of ligand - metal interactions that can be drawn between an carbodiimide - like species or a urea species and a metal would make for good ligands for a Ti(IV) meta l center, and may be related to the catalyst poisoning observed in these reactions. Scheme 4 . 13 Targeted synthesis of asymmetric urea species. 319 Spectra of Isolated Hydroamination and 3CC Derivatives Figure 4 . 11 1 H NMR of HA1 red in CDCl 3 . 320 Figure 4 . 12 13 C NMR of HA1 red in CDCl 3 . 321 Figure 4 . 13 1 H NMR of HA2 red in CDCl 3 . 322 Figure 4 . 14 13 C NMR of HA2 red in CDCl 3 . 323 Figure 4 . 15 1 H NMR of HA3 red 324 Figure 4 . 16 1 H NMR of HA4 in CDCl 3 . 325 Figure 4 . 17 13 C NMR of HA4 in CDCl 3 . 326 Figure 4 . 18 1 H NMR of HA7 red in CDCl 3 . 327 Figure 4 . 19 13 C NMR of HA7 red in CDCl 3 . 328 Figure 4 . 20 1 H NMR of HA8 red in CDCl 3 . 329 Figure 4 . 21 13 C NMR of HA8 red in CDCl 3 . 330 Figure 4 . 22 1 H NMR of HA12 Red in CDCl 3 . 331 Figure 4 . 23 13 C NMR of HA12 red in CDCl 3 . 332 Figure 4 . 24 1 H NMR of HA13 red in C 6 D 6 . 333 Figure 4 . 25 1 3 C NMR of HA13 red in C 6 D 6 . 334 Figure 4 . 26 1 H NMR of HA14 red in C 6 D 6 . 335 Figure 4 . 27 13 C NMR of HA14 red in C 6 D 6 . 336 Figure 4 . 28 1 H NMR of 3CC1 in CDCl 3 . 337 Figure 4 . 29 13 C NMR of 3CC1 in CDCl 3 . 338 Figure 4 . 30 1 H NMR of 3CC5 in CDCl 3 . 339 Figure 4 . 31 13 C NMR of 3CC5 in CDCl 3 . 340 Figure 4 . 32 GC - MS of HA3 red . 341 Figure 4 . 33 GC - MS of crude HA11 342 Figure 4 . 34 GC - MS of crude HA12. 343 Figure 4 . 35 Crude GC - MS trace of HA13. 344 Figure 4 . 36 HA14 crude GC - MS trace. 345 Figure 4 . 37 EI - MS Fragmentation Pattern for HA14(A). 346 Figure 4 . 38 EI - MS Fragm entation Pattern for HA14(B). 347 Figure 4 . 39 GC - MS trace of HA14 red . 348 Figure 4 . 40 EI - MS of HA14 red (A). 349 Figure 4 . 41 EI - MS of HA14 red (B). 350 Figure 4 . 42 GC - MS of 3CC1. 351 Figure 4 . 43 GC - MS of 3CC5. 352 Figure 4 . 44 Fragmentation patterns observed for 1 - phenyl - 3 - cyclohexylurea. 353 Figure 4 . 45 Fragmentation patterns observed in the HCl wash of used [Ti]700 catalyst after iminoamination w hich closely match those for 1 - phenyl - 3 - cyclohexylurea. 354 Figure 4 . 46 Fragmentation pattern observed in the HCl wash of used [Ti]700 catalyst after iminoamination, which closely matches 1,3 - dicyclohexylurea. 355 REFERENCES 356 REFERENCES Note: The original research presented in this chapter has been, in parts, published in the peer - reviewed articles under references 1 and 2 below. (1) Aldrich, K. E.; Odom, A. L. Titanium - Catalyzed Hydroamination and Multicomponent Coupli ng with a Simple Silica - Supported Catalyst. Organometallics 2018, DOI:10.1021/acs.organomet.8b00313 10.1021/acs.organomet.8b00313. (2) Aldrich , K. E. , O dom, A. L., Low surface density silica - supported Ti precatalyst for C - N bond forming reactions. Submitte d, 2019. (3) - and Structural Characteristics Imposed by the N,N - Di(pyrrolyl - - methyl) - N - methylamine (dpma) Ligand. Inorganic chemistry 2001, 40 (9), 1987. (4) Cao, C.; Ciszewski, J. T.; Odom, A. L. Hydroamination of Alkynes Catalyzed by a Titanium Pyrrolyl Complex. Organometallics 2001, 20 (24), 5011. (5) Shi, Y.; Hall, C.; Ciszewski, J. T.; Cao, C.; Odom, A. L. Titanium dipyrrolylmethane derivatives: rapid intermolecula r alkyne hydroamination. Chemical communications 2003, DOI:10.1039/B212423H 10.1039/B212423H(5), 586. (6) Odom, A. L.; McDaniel, T. J. Titanium - Catalyzed Multicomponent Couplings: Efficient One - Pot Syntheses of Nitrogen Heterocycles. Accounts of Chemical R esearch 2015, 48 (11), 2822. (7) Odom, A. L. New C - N and C - C bond forming reactions catalyzed by titanium complexes. Dalton transactions 2005, DOI:10.1039/B415701J 10.1039/B415701J(2), 225. (8) Cao, C.; Shi, Y.; Odom, A. L. A Titanium - Catalyzed Three - Compo nent Coupling To - - Iminoamines. Journal of the American Chemical Society 2003, 125 (10), 2880. (9) McDaniel, T. J.; Lansdell, T. A.; Dissanayake, A. A.; Azevedo, L. M.; Claes, J.; Odom, A. L.; Tepe, J. J. Substituted quinolines as noncovalent proteasome inhibitors. Bioorganic & Medicinal Chemistry 2016, 24 (11), 2441. (10) Majumder, S.; Gipson, K. R.; Staples, R. J.; Odom, A. L. Pyrazole Synthesis Using a Titanium - Catalyzed Multicomponent Coupling Reaction and Synthesis of Withasomn ine. 2009, (11) Majumder, S.; Odom, A. L. Titanium catalyzed one - pot multicomponent coupling reactions for direct access to substituted pyrimidines. Tetrahedron 2010, 66 (17), 3152. 357 (12) Barnea, E.; Majumder, S.; Staples, R. J.; Odom, A. L. One - Step Route to 2,3 - Diaminopyrroles Using a Titanium - Catalyzed Four - Component Coupling. Organometallics 2009, 28 (13), 3876. (13) Dissanayake, A. A.; Odom, A. L. Regioselective conversion of alkynes to 4 - substituted and 3,4 - disubstituted isoxazoles u sing titanium - catalyzed multicomponent coupling reactions. Tetrahedron 2012, 68 (3), 807. (14) Dissanayake, A. A.; Staples, R. J.; Odom, A. L. Titanium - Catalyzed, One - Pot Synthesis of 2 - Amino - 3 - cyano - pyridines. Advanced Synthesis & Catalysis 2014, 356 (8) , 1811. (15) Pasko, C. M.; Dissanayake, A. A.; Billow, B. S.; Odom, A. L. One - pot synthesis of pyrroles using a titanium - catalyzed multicomponent coupling procedure. Tetrahedron 2016, 72 (9), 1168. (16) Lui, E. K. J.; Brandt, J. W.; Schafer, L. L. Regio - and Stereoselective Hydroamination of Alkynes Using an Ammonia Surrogate: Synthesis of N - Silylenamines as Reactive Synthons. Journal of the American Chemical Society 2018, 140 (15), 4973. (17) Hao, H.; Thompson, K. A.; Hudson, Z. M.; Schafer, L. L. Ti - Catalyzed Hydroamination for the Synthesis of Amine - - Conjugated Materials. 2018, 24 (21), 5562. (18) Walsh, P. J.; Baranger, A. M.; Bergman, R. G. Stoichiometric and catalytic hydroamination of alkynes and allene by zirconium bisamides Cp2Zr (NHR)2. Journal of the American Chemical Society 1992, 114 (5), 1708. (19) Substantially Enhanced Reactivity from an Unexpected Cyclopentadienide/Amide Ligand Exchange. Jo urnal of the American Chemical Society 2001, 123 (12), 2923. (20) Pohlki, F.; Doye, S. The Mechanism of the [Cp2TiMe2] - Catalyzed Intermolecular Hydroamination of Alkynes. 2001, 40 (12), 2305. (21) Hazari, N.; Mountford, P. Reactions and Applications of Tit anium Imido Complexes. Accounts of Chemical Research 2005, 38 (11), 839. (22) Billow, B. S.; McDaniel, T. J.; Odom, A. L. Quantifying ligand effects in high - oxidation - state metal catalysis. Nature Chemistry 2017, 9, 837. (23) DiFranco, S. A.; Maciulis, N. A.; Staples, R. J.; Batrice, R. J.; Odom, A. L. Evaluation of donor and steric properties of anionic ligands on high valent transition metals. Inorganic chemistry 2012, 51 (2), 1187. (24) Aldrich, B., Holmes, Bemowski, Odom. (25) Ng, E. - P.; Law, S. - P.; Muk ti, R. R.; Juan, J. - C.; Adam, F. Hydroamination of cyclohexene enhanced by ZnCl2 nanoparticles supported on chiral mesoporous silica. Chemical Engineering Journal 2014, 243, 99. 358 (26) Davaasuren, B.; Emwas, A. - H.; Rothenberger, A. MAu2GeS4 - Chalcogel (M = Co , Ni): Heterogeneous Intra - and Intermolecular Hydroamination Catalysts. Inorganic chemistry 2017, 56 (16), 9609. (27) Sengupta, M.; Bag, A.; Das, S.; Shukla, A.; Konathala, L. N. S.; Naidu, C. A.; Bordoloi, A. Reaction and Mechanistic Studies of Heterogen eous Hydroamination over Support - Stabilized Gold Nanoparticles. 2016, 8 (19), 3121. (28) Copéret, C.; Comas - Vives, A.; Conley, M. P.; Estes, D. P.; Fedorov, A.; Mougel, V.; Nagae, H.; Núñez - Zarur, F.; Zhizhko, P. A. Surface Organometallic and Coordination Chemistry toward Single - Site Heterogeneous Catalysts: Strategies, Methods, Structures, and Activities. Chemical Reviews 2016, 116 (2), 323. (29) Copéret, C.; Chabanas, M.; Petroff Saint - Arroman, R.; Basset, J. - M. Cover Picture: Angew. Chem. Int. Ed. 2/2003 . 2003, 42 (2), 129. (30) Blanc, F.; Copéret, C.; Lesage, A.; Emsley, L. High resolution solid state NMR spectroscopy in surface organometallic chemistry: access to molecular understanding of active sites of well - defined heterogeneous catalysts. Chemical S ociety Reviews 2008, 37 (3), 518. (31) Centers on Oxide Surfaces. Chemical Reviews 2010, 110 (2), 656. (32) Barman, S.; Merle, N.; Minenkov, Y.; De Mallmann, A.; Samantaray, M. K.; Le Quéméner, F.; Szeto, K. C.; Abou - Hamad, E.; Cavallo, L.; Taoufik, M.et al. Well - Defined Silica Grafted Molybdenum Bis(imido) Catalysts for Imine Metathesis Reactions. Organometallics 2017, DOI:10.1021/acs.organomet.7b00115 10.1021/acs.organomet. 7b00115. (33) Samantaray, M. K.; Dey, R.; Kavitake, S.; Basset, J. - M. In C - H Bond Activation and Catalytic Functionalization II; Dixneuf, P. H.;Doucet, H., Eds.; Springer International Publishing: Cham, 2016, DOI:10.1007/3418_2015_139 10.1007/3418_2015_139 . (34) Goldsmith, B. R.; Peters, B.; Johnson, J. K.; Gates, B. C.; Scott, S. L. Beyond Ordered Materials: Understanding Catalytic Sites on Amorphous Solids. ACS Catalysis 2017, DOI:10.1021/acscatal.7b01767 10.1021/acscatal.7b01767, 7543. (35) Schrock, R. R .; Copéret, C. Formation of High - Oxidation - State Metal Carbon Double Bonds. Organometallics 2017, 36 (10), 1884. (36) Conley, M. P.; Forrest, W. P.; Mougel, V.; Copéret, C.; Schrock, R. R. Bulky Aryloxide Ligand Stabilizes a Heterogeneous Metathesis Cataly st. 2014, 53 (51), 14221. (37) Solans - Monfort, X.; Copéret, C.; Eisenstein, O. Metallacyclobutanes from Schrock - Type d0 Metal Alkylidene Catalysts: Structural Preferences and Consequences in Alkene Metathesis. Organometallics 2015, 34 (9), 1668. 359 (38) Pucin o, M.; Inoue, M.; Gordon, C. P.; Schowner, R.; Stöhr, L.; Sen, S.; Hegedüs, C.; Robé, E.; Tóth, F.; Buchmeiser, M. R.et al. Promoting Terminal Olefin Metathesis with a Supported Cationic Molybdenum Imido Alkylidene N - Heterocyclic Carbene Catalyst. 2018, 57 (44), 14566. (39) Le Roux, E.; Liang, Y.; Storz, M. P.; Anwander, R. Intramolecular Hydroamination/Cyclization of Aminoalkenes Catalyzed by Ln[N(SiMe3)2]3 Grafted onto Periodic Mesoporous Silicas. Journal of the American Chemical Society 2010, 132 (46), 1 6368. (40) Cook, A. K.; Copéret, C. Alkyne Hydroamination Catalyzed by Silica - Supported Isolated Zn(II) Sites. Organometallics 2018, 37 (9), 1342. (41) Eedugurala, N.; Wang, Z.; Chaudhary, U.; Nelson, N.; Kandel, K.; Kobayashi, T.; Slowing, I. I.; Pruski, M.; Sadow, A. D. Mesoporous Silica - Supported Amidozirconium - Catalyzed Carbonyl Hydroboration. ACS Catalysis 2015, 5 (12), 7399. (42) Beaudoin, M.; Scott, S. L. Spontaneous Evolution of Silica - Supported Ti Amide Fragments to Imine and Imido Complexes. Organ ometallics 2001, 20 (2), 237. (43) Walder, B. J.; Berk, C.; Liao, W. - C.; Rossini, A. J.; Schwarzwälder, M.; Pradere, U.; Hall, J.; Lesage, A.; Copéret, C.; Emsley, L. One - and Two - Dimensional High - Resolution NMR from Flat Surfaces. ACS Central Science 2019 , DOI:10.1021/acscentsci.8b00916 10.1021/acscentsci.8b00916. (44) Majumder, S.; Gipson, K. R.; Odom, A. L. A Multicomponent Coupling Sequence for Direct Access to Substituted Quinolines. Organic Letters 2009, 11 (20), 4720. (45) Bouh, A. O.; Rice, G. L.; Scott, S. L. Mono - and Dinuclear Silica - Supported Titanium(IV) Complexes and the Effect of TiOTi Connectivity on Reactivity. Journal of the American Chemical Society 1999, 121 (31), 7201. (46) Vujkovic, N.; Fillol, J. L.; Ward, B. D.; Wadepohl, H.; Mou ntford, P.; Gade, L. H. Insertions into Azatitanacyclobutenes: New Insights into Three - Component Coupling Reactions Involving Imidotitanium Intermediates. Organometallics 2008, 27 (11), 2518. (47) Ong, T. - G.; Yap, G. P. A.; Richeson, D. S. Catalytic C N bo nd metathesis of carbodiimides by group 4 and 5 imido complexes supported by guanidinate ligands. Chemical communications 2003, DOI:10.1039/B308710G 10.1039/B308710G(20), 2612. (48) Ong, T. - G.; Yap, G. P. A.; Richeson, D. S. Catalytic Construction and Reco nstruction of - Mediated Guanylation of Amines and Transamination of Guanidines. Journal of the American Chemical Society 2003, 125 (27), 8100. (49) Zhu, T. - H.; Wang, S. - Y.; Tao, Y. - Q.; Ji, S. - J. Synthesis of Carbodiimides by I2/CHP - Mediated Cross - Coupling Reaction of Isocyanides with Amines under Metal - Free Conditions. Organic Letters 2015, 17 (8), 1974. (50) Bora, P.; Bez, G. Chemoselective isocyanide insertion into the N H bond using iodine DMSO: metal - free access to substituted ureas. Chem ical communications 2018, 54 (60), 8363. 360 (51) Ugi, I. M., R.; Lipinski, M.; Bodesheim, F.; Rosendahl, F. Cyclohexyl Isocyanide. Org. Synth. 1961, 41, 13. (52) Khedkar, V.; Tillack, A.; Beller, M. A Dramatic Effect of Aryloxo Ligands on the Titanium - Catalyz ed Hydroamination of Alkynes. Organic Letters 2003, 5 (25), 4767. (53) Heydari, A.; Khaksar, S.; Esfandyari, M.; Tajbakhsh, M. A novel one - pot reductive amination of aldehydes and ketones with lithium perchlorate and zirconium borohydride piperazine comple xes. Tetrahedron 2007, 63 (16), 3363. (54) Pei, D.; Wang, Z.; Wei, S.; Zhang, Y.; Sun, J. S - Chiral Sulfinamides as Highly Enantioselective Organocatalysts. Organic Letters 2006, 8 (25), 5913. (55) Kato, H.; Shibata, I.; Yasaka, Y.; Tsunoi, S.; Yasuda, M.; Baba, A. The reductive amination of aldehydes and ketones by catalytic use of dibutylchlorotin hydride complex. Chemical communications 2006, DOI:10.1039/B610614E 10.1039/B610614E(40), 4189. (56) Anderson, L. L.; Arnold, J.; Bergman, R. G. Catalytic Hydroa mination of Alkynes and Norbornene with Neutral and Cationic Tantalum Imido Complexes. Organic Letters 2004, 6 (15), 2519. (57) Buil, M. L.; Esteruelas, M. A.; López, A. M.; Mateo, A. C. Preparation of Half - Sandwich by a Cyclopentadienyl Ligand with a Pendant Phosphine Tether and Their Use in the Catalytic Hydroamination of Aliphatic and Aromatic Alkynes. Organometallics 2006, 25 (17), 4079. (58) Heutling, A.; Pohlki, F.; Doye, S. [Ind2TiMe2]: A General Catalyst for t he Intermolecular Hydroamination of Alkynes. 2004, 10 (12), 3059. (59) Lin, B.; Waymouth, R. M. Urea Anions: Simple, Fast, and Selective Catalysts for Ring - Opening Polymerizations. Journal of the American Chemical Society 2017, 139 (4), 1645. 361 CHAPTER 5. R EACTIVITY AND RATE LAW DETERMINATION OF A LIGAND - FUNCTIONALIZED SILICA - SUPPORTED TITANIUM CATALYST 5.1 Introduction An introduction to some of the possible Ti - catalyzed C N bond forming reactions was discussed in Chapter 4. 1 - 6 While many simple C N bond forming reactions have relatively simple mechanisms and simple rate laws, others are quite complex. As an example, the rate law determined by Doye for Ti - catalyzed hydroamination (Fig. 5.1) suffers from not one, but two equilibrium processes which reduce the effective concentration of the active species (Ti - imide) and hinder the rate of the ove rall reaction. 7 Of note is that o ne of these processes is a dimerization, a process inherent to many active homogeneous catalyst systems, of which several examples with Ti are known. 8 - 13 362 Figure 5 . 1 Proposed mechanism and rate law determined for homogeneous Ti(IV) hydroamination catalysts of the (A) 2 Ti(X) 2 variety; The integrated rate law is given, as well as the dependence of k obs on both ami ne and Ti concentrations. 7 In the Ti - imide active species. The kinetics of these types of reactions only become more complicated as the C N bond forming reactions under study are made more c omplex with the addition of a third coupling partner. As mention in Chapter 4, a reaction that has been a focal point in our group is the hydroamination - based multicomponent coupling reaction where an amine, an alkyne, and an isonitrile are combined to yie ld 1,3 - diimine tautomers (Fig. 5.2). 1, 14, 15 These products serve as organic building blocks to a host of different heterocyclic complexes, of interest for a variety of reasons. However, a look at the proposed mechanism for their formation reveals multiple competing side reactions. 14, 16, 17 These side reactions put strict limitations on the catalyst, as both the ancillary ligand and substrates utilized need to be very well balanced to result in formation of 3CC at the exclusion of other products. 363 Figure 5 . 2 (top) General reaction sequence for the 3 - component coupling of an amine, alkyne, and isonitrile to yield 1,3 - diimine tautomers. (bottom) The two catalysts typically utilized for this transformation. A wide variety of iminoamination (3CC) products can be synthesized using the Ti - catalysts shown above, with dipyrrolyl ligands under homogeneous conditions. However, in these homogeneous catalyst systems, little can be done if a substrate leads to side reactions, and the regioselectivity is not readily manipulated. Additi onally, with the homogeneous nature of these systems, it is likely that similar catalyst deactivation pathways and potential off - cycle occupation of the Ti catalyst to those demonstrated by Doye 18 also contribute to catalyst deactivation and effective reduction of loading in these reactions (vide supra). With the goal of improving on some of these catalyst issues, we set out to perform this same type of chemistry with heterogeneous catalysts. Thus, the work highlighted in Cha pter 4 was inspired, to move to a heterogenous system but maintain similar coordination environment and corresponding reactivity to the well - studied homogeneous catalysts. The initial results obtained with the two different catalyst variants that we synthe sized, [Ti]200 and [Ti]700, suggested that this was a fruitful direction to explore for Ti - catalyzed iminoamination. While these catalysts are both competent for hydroamination, [Ti]200 is a poor catalyst for iminoamination. On the other hand, [Ti]700 is a decent iminoamination catalyst for both terminal and internal aromatic alkynes coupled with aromatic primary amines. In fact, with aniline, 1 - phenylpropyne, and CyNC, a yield of 88% was noted in only 16 h. This result not only 364 demonstrates a dramatic imp rovement in iminoamination ability relative to the [Ti]200, which is severely substrate - scope - limited, but also demonstrates catalytic ability comparable to the best homogeneous catalysts for the same reaction at lower catalyst loadings. Some challenges s till remained to making the material a general iminoamination catalyst capable of dramatically improving upon the ability of the homogeneous systems. The first issue remaining with these silica - supported catalysts, specifically [Ti]700 is a limited substr ate scope tolerance. The catalyst cannot perform iminoamination between an unsubstituted aniline and a terminal alkyl alkyne, such as 1 - octyne. Additionally, with both heterogeneous catalysts, intolerance to alkyl primary amines had been observed. Finally, we also had evidence for catalyst deactivation, as the material was not able to be recycled (i.e. multiple runs of the reaction catalyzed by a single batch of material). Considering traditional high - valent transition metal mechanisms for these reactions, i.e. 19, 20 7 shown above, only 2 protolytically active sites are necessary on the metal in order to form the active metal - - based hydroamination, requires a single Ln N bond where olefin or alkyne insertion can occur. 21 Following either path, the Ti sites in [Ti]700 have more protolytically cleavable NMe 2 ligand s than each site needs to undergo these types of C N bond forming reactions. This leads to the possibility of ligand functionalization of the surface - bound metal to tune reactivity while maintaining the active sites necessary for reactivity. Capitalizing o n this idea, we decided to pursue ligand functionalization of the [Ti]700 catalyst for three component coupling chemistry. We suspected this would be an easy route to tuning the substrate tolerance and subsequent selectivity in the products of iminoaminati on yielded in [Ti]700 catalyzed reactions. 365 5.2 Activity of [Ti]700(X) for Three - Component Coupling Chemistry with a Variety of X¯ Ligands Operating under the assumption that the surface - bound Ti sites are catalyzing HA and 3CC, with a Ti - imide active species , we hypothesized that two protolytically active sites are needed to maintain the catalytic activity of the silica - supported Ti species. With [Ti]700, 3 protolytically active sites remain on Ti after binding to the surface. Theoretically, we can add 1 stoi chiometric equivalent (relative to Ti) of an irreversibly bound Bronsted acidic species, HX, to our Ti catalyst to protonate one NMe 2 ligand to liberate NHMe 2 . Removal of one NMe 2 still maintains a Ti site that is active for the desired catalysis. In this manner, a variety of HX species can be utilized as ligand additives to functionalize the surface - bound titanium sites, providing the catalyst species [Ti]700(X). The addition of these ligands should result in dramatic differences in the reactivity of the T i species, just as ancillary ligand design affects the performance of homogeneous catalysts. This approach is illustrated in Schem e 5.1 . Scheme 5 . 1 Addition of Brønsted acidic HX ligands to [Ti]700 to generate [Ti]700(X) species. We began screening the 3CC reactions of aniline and CyNC with 4 different alkynes, employing a variety of HX species as ligand additives. Note, in terms of the overall reaction stoichiom etry, we are adding 5 mol% HX, or a 1:1 ratio of Ti to HX. Overall, the ligands we chose to examine in these initial studies are similar to those which have yielded competent catalysts in homogeneous systems, such as phenoxides, pyrrolides, or amidate liga nds. 2, 22, 23 The goal was to examine commercially available or easily synthesized ligands (1 step reactions from cheap starting 366 catalyst material and the intent to avoiding time - intensive ligand design and preparation. Even with these simple ligands, the effects of HX addition to these Ti - catalyzed iminoamination reactions is dramatic. The reaction times with HX additives are incredibly fast. With many combinations of substrates and HX ligands, the 3CC reactions now provide yields of >90% in less than 1 h of reaction time! For comparison, even with the best homogeneous catalyst, these re actions typically take on the order of 24 h to complete. Additionally, large changes in the regioselectivity of the catalyst can be observed depending on the identity of HX, within each set of substrates. This is demonstrated well by Scheme 5.2 and Fig. 5 . 3 , which shows the yields of the different 3CC reactions examined at 1 h reaction times, using a variety of different HX ligands. Tables that list each of these results can be found in the Experimental but are shown graphically here for ease of discussion . Because the parent precatalyst, [Ti]700, catalyzes the 3CC reactions of aniline and CyNC with both 1 - phenylpropyne and phenylacetylene, in the absence of HX, we can compare the overall regioselectivity and rates of these reactions with HX included, to t he parent catalyst. 24 This provides a complete picture of how the ligand additives affect catalyst performance. In both cases, it is apparent that the rate of these reactions is dramatically enhanced with the inclusion of a ligand additive. For example, even though 1 - phenylpropyne is an internal alkyne, with the right ligands, we see over 90% yield of the iminoamination products in under 1 h. Namely , [Ti]700(X), where X = 2,6 - dimethylphenylamidate, gives the 3CC products in 99% yield in only 1 h with a regioselectivity of 10.2:1. In direct comparison to the [Ti]700, which takes 16 h to reach 88% completion with a very similar regioselectivity of 10.3 :1, the rate increase upon inclusion of HX is readily apparent. 367 Scheme 5 . 2 Iminoamination reaction examined with a variety of HX ligands. 368 Figure 5 . 3 Effects of different HX ligands on the general 3CC reaction utilizing different alkynes: ( top ) The variety of HX ligands screened in the 3CC reactions; ( bottom ) Plots of the 3CC reaction products showing the regioisomer ratio versus % yield for each HX ligand examined. In the plots, burgundy diamonds correspond to bidentate ligands, while blue circles correspond to monodenta te ligands. The best ligands for each alkyne are specified next to their respective points. A very comparable trend is noted with phenylacetylene as the coupled alkyne. The [Ti]700 provides a yield of 52% in 18 h, with a regioselectivity of 6.3:1. Yet whe n [Ti]700(X) is used as the catalyst, again where X = 2,6 - dimethylphenylamidate, we observe a yield of 99% in 1 h with a regioselectivity of 5.6:1. It is also worth noting that with phenylacetylene, the inclusion of the 0 2 4 6 8 10 0 20 40 60 80 100 Regioisomer Ratio Yield (%,FID) Phenylacetylene 3CC 0 5 10 15 20 0 20 40 60 80 100 Regioisomer Ratio Yield (%,FID) 1 - Phenylpropyne 3CC 0 1 2 3 4 0 20 40 60 80 100 Regioisomer Ratio Yield (%,FID) 1 - Octyne 3CC 17 17 1 8 1 1 369 ligand greatly reduces the propensit y for side reactions. With [Ti]700, we see 18% of the aniline remaining after 18 h; between the 3CC product and the unreacted aniline, however, only 70% of the initial aniline has been accounted for, with the remaining 30% going to hydroamination (~15%), f ormamidine production (~5%), and other side products. Additionally, alkyne trimerization is also observed by GC analysis. While the reduction in side product formation with 1 - phenylpropyne is less substantial than it is with phenylacetylene upon addition of HX ligand to the catalytic 3CC reactions, it is still observed (on the order of <5%). Thus overall, the 3CC reactions using [Ti]700 catalysts appear to offer 3 major advantages: (1) faster rates, (2) higher yields, and (3) reduction in side product form ation. The advantages of ligand inclusion continue to grow when we consider one of the substrate - limited reactions with [Ti]700: the coupling of aniline and CyNC with a terminal alkyl alkyne, 1 - octyne. This reaction with [Ti]700 had previously yielded onl y side products, but did not show formation of the desired 3CC products. 17, 24 Switching to [Ti]700(X), however, has prov ided a number of catalytic species that can do this transformation in high yields (See Fig. 5. 3 ). Now the main limitation in this reaction is the relatively low regioselectivity, which has previously been observed when utilizing a terminal alkyl alkyne, su ch as 1 - hexyne as the coupling partner (vide supra). 14 Despite this regioselectivity limitation, we do see vast opportunity to increase the regioselectivity, as changing the HX ligands in this reaction results in large changes in this ratio. Thus, finding a selective catalys t may simply require a slightly different ligand from the classes examined here. Also, it is still an improvement over homogeneous systems and their inherent regioselectivity, which is close to 1:1 (vide supra). 370 5.3 One - pot - two - step Heterocycle Synthesis with [Ti]700(X) Quantification of organic products in catalyzed reactions via in situ techniques (i.e. GC analysis calibrated against authentic isolated products) is valuable and provides detailed information about the species in reaction solutions. However, for the Ti - catalyzed iminoamination reactions discussed above, product isolation r emains the primary objective for the functionalized heterocycles that can proceed directly from the iminoamination products. To evaluate the practical usability of the [Ti]700(X) catalyst system, we pursued product isolation of a previously synthesized qui noline (using the one - pot - two - step method) produced with homogeneous systems, to provide direct comparison with a [Ti]700(X) reaction. 25, 26 Scheme 5.2 describes the reaction, which provided an isolated quinoli ne yield of 72% after column chromatography (previously 53%). 25 This synthesis demonstrates that t he high yields, high regioselectivities, and limited amounts of side products formed with reactions catalyzed by the silica - supported Ti could result in higher yields and more readily purified products than many previous homogeneous catalysts. Scheme 5 . 3 Quinoline synthesis using 5 mol% [Ti]700 with 5 mol% 2,6 - dimethylphenylamidate as ligand. Another functionalization of interest is the formation of 2 - amino - 3 - cyanopyridines i n a similar 1 - pot - 2 - step reaction. 27 The original report of this complex, produced through Ti - catalyzed iminoamination, employed Ti(dpm)(NMe 2 ) 2 (10 mol%) as the catalyst and required a 24 h iminoamination step, followed by a 2 h ring - closure reaction. Additi onally, in the original report, the product pyridine was characterized as a waxy brown solid. Attempts to repeat the synthesis, as originally reported, have failed in the hands of several researchers. Highly impure reaction 371 mixtures result and isolation of the product from the reaction mixtures is quite challenging. This synthetic method, utilizing 10 mol% Ti(dpm)(NMe 2 ) 2 does provide product, as an oily, colored residue with residual impurities by NMR, GC - MS, and TLC. There are several places where the synthesis of the 2 - amino - 3 - cyano pyridines could be encountering challenges to reproducibility on each experimental run. Specifically, the second step to transform the iminoamination product into the final pyridine has not been optimized and still includes 3 Å molecular sieves (perhaps to prevent unintentional hydration). It has been directly observed by GCMS of the crude i minoamination reaction mixture that with Ti(dpm)(NMe 2 ) 2 , the additional reaction steps and their influence in compromising the yield, this observation suggests fundamen tal problems with the formation of the iminoamination product in the first step. However, the impure product and reproducibility issues are a major concern. Using [Ti]700(Amidate 2,6 - dimethyl ), the coupling of 3,5 - dimethylaniline, CyNC, and 1 - phenylpropyne , followed by base - catalyzed pyridine formation with malononitrile takes a total of 4 h of reaction time ( Fig. 5.4). The resulting pyridine product was isolated in 70% yield after these 2 steps. While a slight impurity was noted in the product by NMR, recr ystallization of the pale - yellow solid from a dilute hexane solution yields extremely pure, X - ray quality crystals of the pyridine. This synthesis as well as the crystal structure are shown in Fig. 5.4, below, and have provided a very reliable route to iso lation of a clean, crystalline pyridine product, suitable for use in biological studies. 372 Figure 5 . 4 The synthetic scheme (left) and single crystal X - ray structure of a 2 - a mino - 3 - cyanopyridine synthesized using [Ti[700(X). This investigation and application of the [Ti]700(X) system to the synthesis of complex heterocycles clearly demonstrates that the catalyst system is of practical use in the laboratory for isolation of t argeted organic molecules. It also demonstrates that the improvements in the iminoamination step cleaner reactions with fewer byproducts, consumption of limiting reagents, inertness of the catalyst toward subsequent functionalization steps all add up to be tter results with the heterogeneous catalyst system compared to the homogeneous systems with certain substrates. 5.4 3CC Reaction Kinetics for [Ti]700(X) To gain insight into how the [Ti[700(X) system is outperforming the homogeneous systems, we wanted to pr obe the rate law for the Ti - catalyzed iminoamination reaction. Following the simple graphical method of Bures, to analyze a reaction for order in reagents and catalyst, the modifications listed in Table 2 were made to the reaction under study. 28, 29 Utilizing X ¯ = 2,4,6 - tri - tert - butylphenoxide as the ancillary ligand, a very clean and rapid reaction suitable for kinetic analysis, was achieved with aniline, 1 - octyne, and CyNC as the coupling partners. Note that the concentration of the reactions examined for kinetic analysis were about ½ of the concentration typically used for one of our standard heterogeneous catalyzed reaction. This allowed for a slower 373 reaction that was more amenable to sampling the amount of product over time (i.e. the reaction continued to run for longer than 1 h), as well as conservation of materials. Scheme 5 . 4 General conditions and substrates used to examine the iminoamination reaction with [Ti]700(X) catalyst. Table 5 . 1 Conditions screened for reaction order in each substrate and catalyst for heterogeneous catalyzed 3 - component coupling of aniline, 1 - octyne, and t BuNC. The initial catalyst and reagent amounts are listed by concentration (M). Entry Ti(NMe 2 ) 2 (X)/SiO 2 700 (mol%) H 2 NPh CyNC 1 - Octyne Total Conversion to 3CC (%) b Regioisomer Ratio (A:B) Hydroamination Byproduct (%) 1 0.012 (5) 0.25 0.25 0.51 83 2.9:1 16 2 a 0.006 (2.5) 0.25 0.25 0.51 8 1.8:1 3 3 0.024 (10) 0.25 0.25 0.51 83 3.4:1 15 4 0.012 (5) 0.51 0.25 0.51 54 2.1:1 9 5 0.012 (5) 0.25 0.51 0.51 61 3.2:1 8 6 0.012 (5) 0.25 0.25 1.01 58 3.0:1 7 7 c 0.012 (5) 0.25 0.25 0.51 59 2.9:1 8 a With less than 5 mol% catalyst loading, a large amount of formamidine is produced; in fact the forward rate of formamidine formation is in excess of that of 3CC formation (18% vs. 8%). With 5 mol% or more catalyst loading, essentially no formamidine format ion is observed, and 3CC is the only reaction with substantial forward progress after the initial heating period. b This is the conversion observed after approximately 100 minutes of reaction time. Typically, the reaction progress has begun to plateau at th is point; note in entries 1 and 3 this yield correlates to consumption of the limiting reagent. c In addition to the reagents listed in the table, an initial 3CC concentration was included in this reaction mixture (0.02 M) to probe for 3CC inhibition on the reaction. The graphical analyses provided by the experiments listed in Table 5. 1 , p resent some complications. First, there appears to be hydroamination occurring rapidly at the beginning of the reaction while the solution is reaching thermal equilibrium w ith the aluminum well - plate. In Entry 1 for example, the first reaction sample at 10 minutes shows a HA concentration of 0.02 M (8%), while at the end of the reaction at 100 minutes, there is a HA product concentration of 0.04 M (16%). As much HA product i s produced in the first 10 minutes as is produced throughout the remaining 90 minutes of reaction time. This indicates that HA initiates with the heterogeneous 374 catalyst at a lower temperature relative to iminoamination. However, when the temperature for im inoamination is reached, it becomes the predominant reaction catalyzed in this complex mixture, by an order of magnitude. Figure 5 . 5 Plot of initial rates from the kinetic experiments shown in Table 5.1. T he initial rates, determined from linear fits shown in the plot above, are listed in the table below . These rates are may be slower than the actual initial rates as they go far beyond 10% converions. Table 5 . 2 Estimated initial rates (M - 1 min - 1 ) from the various conditions in Table 5.1. These estimates may be artificially low, as the reactions progressed far beyond 10% conversion in ~20 min. Entry Number Initial Rate R 2 1 0.004 0.99 3 0.008 0.93 4 0.005 0.97 5 0.004 0.99 6 0.004 0.99 7 0.004 1 Also note that total conversion was not being reached under several of the conditions listed in Table 5.1. With the conditions in Entries 1 and 3, we noted essentially quantitative conversion of the H 2 NPh starting material into a combination of 3CC products and HA side product (99% and 98%) respectively. However, in many of the other entries in Table 5.1, we n oted a similar problem, with the average reaction consuming only about 70% of the initial H 2 NPh starting material. This, 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0 10 20 30 concentration 3CC (M) time (min) Investigation of initial rates for Kinetic Runs Entry 1 Entry 3 Entry 4 Entry 5 Entry 6 Entry 7 Linear (Entry 1) Linear (Entry 3) Linear (Entry 4) Linear (Entry 5) Linear (Entry 6) Linear (Entry 7) 375 along with other observations from the graphical analysis of the kinetics, suggests that catalyst deactivation is an issue in this syst em. Given that Entries 1 and 3 reached total conversion, the relative rate of deactivation with the heterogeneous catalyst under these conditions is certainly a better match when compared to homogeneous catalysts (see Chapter 6), but it complicates the rat e law analysis substantially. For these reasons, a plot of the initial rates extracted from the kinetic runs shown in Table 5.1, is presented in Fig. 5.5 and Table 5.2, for comparison. There are some clear relationships between the rate law and the subst rates used in the reaction. Primarily, upon graphical analysis of Entries 1 and 4 and 1 and 5, we noted that there appears to be inverse 1 st - order on both the concentration of aniline and isonitrile in the reaction (Fig. 5.6). There appears to be a zeroth order dependence in alkyne (Entries 1 and 6), looking at the graphical order and initial rates observed. All three of these reactions present less total conversion than Entry 1. The graphically determined rate law is summarized in Eq. 5.1, below. With the aniline and isonitrile concentrations and their perceived inverse first order dependence, this makes sense. Increasing the concentrations of these substrates decreases the overall rate of the reaction and makes the deactivation process (or processes) catch up with the catalyst faster than it did under the conditions in Entry 1. ( Eq. 5.1 ) Initially it seemed less obvious how a zeroth order dependence was possible with alkyne, yet lower conversion was achieved. This may suggest that the alkyne concentration directly affects the rate of the deactivation of the catalyst; this interaction remai ns a possibility, as the means of catalyst deactivation are not yet fully understood (See Chapter 4). 376 Figure 5 . 6 Order of catalyst and reagent dependence for rate law of the iminoamination reaction c atalyzed by [Ti]700(X). Entry 1 represents the light purple points, with the following concentrations: 0.25 M aniline, 0.25 M CyNC, 0.51 M 1 - octyne, and 0.012 M (5 mol%) Ti. It is plotted against the following entries, where the substrate alteration is ind icated: 3 (green, 0.024 M (10 mol%) Ti), 5 (blue, 0.51 M CyNC), 4 (maroon, 0.51 M aniline), and 6 (aqua, 1.01 M 1 - octyne) to demonstrate how different reagent concentrations affect the rate of the iminoamination reactions. For rough approximations of the i nitial rates, refer to Fig. 5.5. Finally, according to the graphical analysis method, it seems that the rate is first - order relationship with the [3CC] and reaction rate as both Entries 1 and 7 demonstrate similar rates, but again, similar to Entry 6 (excess 1 - octyne), a lower conversion is achieved with Entry 7 compared to Entry 1. 0 0.05 0.1 0.15 0.2 0.25 0 0.001 0.002 0.003 Concentration 3CC (M) t*[cat] Catalyst Order Analysis 0 0.05 0.1 0.15 0.2 0.25 0 500 1000 1500 Concentration 3CC (M) t*[ t BuNC] - 1 Order in Isonitrile 0 0.05 0.1 0.15 0.2 0.25 0 200 400 600 Concentration 3CC (M) t*[aniline] - 1 Order in Aniline 0 0.05 0.1 0.15 0.2 0.25 0 50 100 150 Concentration 3CC (M) t*[1 - octyne] 0 Order in Alkyne 377 We specify that the rate law shown above describes the initial rate of the reaction because this simple equation assumes a constant catalyst concentration. As a consequence, it fails to accurately represent the catalyst deactivation that occurs throughout the course of the reaction. To obtain a more accurate idea of what happens in the re action, start - to - finish, we turned to kinetic simulations. Using the kinetic data outlined in Table 5.1, we attempted to model the proposed mechanism of the iminoamination reaction using KinSim software. The mechanism we propose, and the corresponding rat e law probed using KINSIM are shown in Fig. 5.6 (See Experimental for more information). Based on the experimental observations and the rate law, the rate determining step appears to be generation of the active the catalyst , which we approximate as species C in Fig. 5.7, below . This resting state appears to be a SiO Ti(NHPh) 2 (CNCy) species such that generation of the Ti - imide species, which initiates the coupling for generating the 3CC product, is inhibited by CyNC and aniline. We propose that dissociation of 1 equivalent of H 2 NPh and 1 equivalent of CNCy is required before t he [Ti]700 supported sites reach the active imido form of the catalyst, from which the catalytic cycle can be entered. This interpretation is consistent with the inverse - first - order or inhibitory dependence of the rate upon the concentrations of H 2 NPh and CyNC. As a result, the individual rates of any given step in the presumably multi - step conversion of the three substrates (aniline, alkyne, and isonitrile) into the coupled product may not play a determining role in the overall rate. Since these steps appe ar fast relative to the formation of the active catalytic species, the actual 3 - component coupling can therefore be treated effectively as a single rapid step. This results in the assumption that the first step in the iminoamination cycle ([2+2] cycloaddit ion of alkyne to the Ti - - imide is generated. 378 where k is determined by k 4 , K 1 , and K 2 Figure 5 . 7 Proposed catalytic cycle for [Ti]700(X) based on experimental kinetic data and KINSIM modeling. Note, in the figure, the green letters, rate constants, and equilibrium constants are from the KINSIM model. The deactivation step from A to J is based on e xperimental observations of 3CC reactivity with homogeneous Ti analogues (see Chapter 6). From a practical standpoint, of improving the reaction rate, we took a more qualitative approach. Thus, regardless of the full rate equation, the suggestion that for mation of the active imido is the rate - limiting step, and that this step is dependent on CyNC dissociation and H 2 NPh, led us to reconsider the effects of the X¯ ancillary ligand on the rates of the [Ti]700(X) catalysts. Specifically, we wondered if increas ing the electronic donor ability of the X ¯ ligand would lead to more labile interactions with datively bound CyNC and facilitate a faster equilibrium between a (SiO 2 700 )Ti(NHPh) 2 (CNCy)(X) species and the proposed active Ti - imide. According to the rate 379 law derived from our graphical analyses, rendering these changes stands to greatly improve the rate of the 3CC reaction. In order to test this hypothesis, we turned to the Ligand Donor Parameter system. This system uses a Cr(VI), d 0 model complex to experiment ally determine quantitative electronic donor abilities of various X - ligands to a high valent metal, as discussed in Chapters 1 - 3. 30, 31 Previously, it has been demonstrated that these LDP values, determined for X ¯ ligands on Cr(VI), correlate very well with the electronic effects of ancillary l igands on Ti(IV) hydroamination catalyst. 22 The exten sion of LDP analysis of ligand effects in this heterogeneous system aligns well with the previous studies, as we are still examining the electronic effects of X ¯ ancillary ligands on a Ti(IV) catalyst. The easiest way to make these comparisons while maint aining similar conversions and preventing the reintroduction of side reaction competition in the catalyzed 3CC reaction, is to alter the periphery of the X ¯ ligand to render subtle electronic changes. This allows the ligands to remain isosteric and result in similar bond distances and steric profiles around the active metal. A series of 2,6 - di - tert - butyl - 4 - R - phenols were used as the X ¯ ligands in the [Ti]700(X) catalyst material. A list of the phenols used is displayed in Fig. 5. 8. Our strategy of maintain ing the same steric profile around the metal and changing only the R group on the phenols did result in comparable conversions, without the observation of additional byproducts in the reaction mixtures (relative to 2,4,6 - tri - tert - butylphenol). Any observed changes in the rates of the catalysts is also directly related to the electronic donor ability of the ligand alone, without the additional complication of steric effects on the rate. Comparing the LDP values for the ligands examined, listed in Fig. 5.8 , the relative donor ability is ordered such that R = OMe > H ~ t Bu > Br. Looking at the measured k obs (initial) for 380 each of the catalyst generated with these ancillary ligands, or the plots of the reaction traces also shown in Fig. 5.8 and Table 5.3 , we see that the rate of the catalyst follows the same order. This supports our hypothesis, that making the metal center more electron - rich by increasing the donor ability of the ancillary X ¯ ligand increases the rate of active species formation, and thus increas es the rate of the overall reaction. This also supports the possibility that, with the right ligand(s), the catalyst may be able to achieve rates that allow it to outpace catalyst deactivation processes. 381 Figure 5 . 8 Correlation between reaction rate and donor ability of X - in heterogeneous [Ti]700(X) catalyzed 3CC reactions. More electron - donating X ¯ ligands seem to enhance the rate of the reaction by increasing the rate of formation of the active species from the resting state of the catalyst. (blue = t Bu, grey = H, orange = OMe, yellow = Br). Table 5 . 3 Ligands examined in Figure 5.8, showing LDP values of each phenoxide and the k obs for the iminoamination reaction catalyzed with the given phenoxide as ligand (X) with [Ti]700( X). Entry R LDP (kcal/mol) k obs (initial) (s - 1 ) Trace color 1 t Bu 12.01 0.00006 Blue 2 H 11.98 0.00008 Grey 3 OMe 11.71 0.00010 Orange 4 Br 12.18 0.00004 Yellow 0 0.05 0.1 0.15 0.2 0.25 0 50 100 [3CC] (M) Time (min) Reaction Progress for Various X Ligands 382 5.5 Catalyst Poisoning and Recyclability Recyclability of the catalyst material was one of our initial goals in moving toward heterogeneous Ti based catalysts for these hydroamination and multicomponent coupling reactions. Previously, utilizing [Ti]200, it was determined that the catalyst cannot be perfectly recycled, as Ti is leeched from the s urface of the SiO 2 200 support. It was also determined that this leaching of the active metal originates from interaction with alkyne substrate, and while modest reusability is noted for a few iterations of a given reaction, both the yield and regioselectiv ity of the catalyst suffer with repeated used of the catalyst material. The recyclability of [Ti]700 species was also examined, and initial results demonstrated only trace amounts of product generated upon a second used of the catalyst material. This was the case for both HA or 3CC reactions. Even with the inclusion of a variety of HX type ligands, the [Ti]700(X) is not recyclable. Our initial suspicion was that the [Ti]700(X) and the [Ti]700 material may suffer from loss of the active metal via interactio n with substrates, similar to the [Ti]200. However, by examining the Ti concentration in a variety of [Ti]700(X) samples (and [Ti]700, see Chapter 4), after use in a catalytic reaction, there appeared to be no statistically significant loss of Ti from the material. The results of these analyses, which utilized ICP - OES spectroscopy to determine the Ti concentration in the used catalyst samples, is shown in Table 5. 4 . 383 Table 5 . 4 ICP - OES analysis of various treatments for the [Ti]700 precatalyst material. Ti(X) 3 /SiO 2 700 species ICP determined Ti wt% (±) Material Evaluation [Ti]700(OPh 2,4,6 - tri - tbutyl ) 1.53 (0.06) Within error of calculated Ti wt% [Ti]700(pypyr) 1.38 (0.05) [Ti]700(2,6 - dimethylphenylamidate) 1.36 (0.04) [Ti]700(OSiPh 3 ) 1.38 (0.04) [Ti]700 1.42 (0.03) Ti(X) 3 /SiO 2 700 species ICP Ti wt% calculated (±) As - prepared [Ti]700 1.46 (0.12) Used [Ti]700(X) a 1.46 - 1.35 (0.11) a A range is provided for the calculated wt% Ti for used [Ti]700(X) because different speciation may occur on the Ti sites after a reaction. (i.e. there are 3 coordination sites that could be occupied by X ¯ , NHPh ¯ , CyNC, etc.). Without evidence of Ti loss, we then suspected that, simil ar to our suspicions about [Ti]700, stable surface species were forming on the active Ti sites. Under the standard 3CC conditions, there is both excess alkyne and excess cyclohexylisonitrile. Either of these species, or any product generated in the reactio n mixture, could form non - innocent interactions with the metal that would lead to metallocycles forming on the catalyst surface, for example. Such species may not easily re - enter our catalytic cycle via generation of a Ti - imido simply by addition of more s ubstrates, similar to what we observed with [Ti]700 (See Chapter 4). With non - innocent substrates, potentially adjusting the ratio of the three substrates in the 3CC reaction would reduce these side reactions that lead to deactivation of the catalyst. Both decreasing the amount of CyNC used (1 equiv) in the reaction and increasing the amount of aniline so that it was no longer the limiting reagent (2 equiv), resulted in lower yields than those achieved with the general 3CC conditions. Additionally, the cata lyst material recovered from these reactions had still been deactivated. Unfortunately, this simple strategy was not effective at preserving the 384 activity of the Ti sites. In light of the kinetic analyses, this result is somewhat unsurprising, as both CyNC and H 2 NPh demonstrate inverse first order dependence in the rate law; thus, with these conditions, the total reaction rate is likely lower than it was with the initial reaction conditions. 5.6 Poisoning Experiments and Controls for [Ti]700 Note, for these po isoning experiments and regeneration attempts, a simpler hydroamination reaction was examined, as opposed to iminoamionation. The general reaction scheme for this hydroamination reaction is outlined in the header for Table 5.3 ents, the equilibrium formation of the active Ti imido species is systematically altered by the addition of an excess of HX compounds. The HX compounds compete with NH 2 Ph, as ligands that occupy the active sites on Ti. We would predict that as more of the HX species is added, the concentration of inactive species such as Ti(X) 3 /SiO 2 700 will increase relative to the proposed active species, Ti(=NR)X/SiO 2 700 . This shift should be directly observed by a reduction in the rate of catalysis and/or a decreased ove rall yield. With the closely related [Ti]200, addition of 1.2 equiv of pyrrole to the catalyst decreased the performance of the catalyst to about 20% of the activity of the un - poisoned catalyst for the hydroamination of 1 - phenylpropyne and aniline. This f inding, in combination with the observation that N - methyl - aniline does not react with [Ti]200, served to support the assumption that generation of an imido species is needed to facilitate the catalysis. With [Ti]700, the addition of 1 equiv of HX results in the formation of [Ti]700(X), which retains two protolytically active sites (two NMe 2 ligands), and is an active catalyst with comparable activity to the original [Ti]700, resulting in high overall yields for hydroamination of aniline and 1 - octyne. The a ddition of more equivalents of HX, does indeed decrease the overall yields of these reactions, but by a surprisingly small amount. As an example, Fig. 5.9 sh ows the yield obtained 385 with 3 different amounts of pyrrole added to the [Ti]700 catalyst. We see th at the [Ti]700 and the [Ti]700(pyrrolide) give similar overall results, both with yields over 90%. When 10 equivalents of >80% yield in less than 1 h. Even with the inclusion of up to 40 equiv of pyrrole, which is 2.5 times the amount of H 2 NPh in the reaction mixture, we still observed 70% yield in 1 h. Figure 5 . 9 Poisoning experiments with pyrrole and 2 - tert - butyl - 4 - methoxyphenol, showing very different catalyst activity with varying concentrations of the two different HX ligands. These results starkly contrast the observations with [Ti]200, and we sought to determine whether this resilience to poisoning was a general trait with the [Ti]700 or was inherently related to some property of the specific HX ligand used, in this instance pyrrole. Since in the 3CC ligand screenings, phenols generally performed well as ligands, addition of a substituted phenol to the HA reaction was first examined for comparison. Similar to the pyrrole reaction, 1 equivalent of the phenol results in catalyst performance similar to that of [Ti]700. However, the catalytic activity is eliminated when 10 equivalents of phenol was added . This experiment suggests that the properties of HX determine whether the equilibria favor inactive species such as Ti(X) 3 /SiO 2 700 or the active Ti(=NR)X/SiO 2 700 , where aniline displaces X ¯ and the Ti species can enter the catalytic cycle. 0 20 40 60 80 100 0 20 40 60 yield (%) Time (min) Pyrrole Poisining Ti-700 10 equiv 40 equiv 1 equiv 0 20 40 60 80 100 0 20 40 60 Yield (%) Time(min) 4 - Methoxy - Phenol Poisoning Ti700 1equiv 10 equiv 386 Figure 5 . 10 General hydroamination reaction and conditions applied to [Ti]700 catalyzed reactions with excess ligand additive (HX). Ligands examined in Table 5.5. Table 5 . 5 Yields obtained from the hydroamination of 1 - octyne and aniline using [Ti]700 precatalyst and 10 equiv of HX ligand added. Entry HX (10 equiv) LDP (kcal/mol) pK a %V bur % yield HA1 Regioisomer Ratio 1 pyrrole 13.64 17(20) 20.4 81 52:1 2 2 - tert - butyl - 4 - methoxyphenol 11.82 a ~10 21.5 0 - 3 tert - butanol 10.59 17 21.0 c 0 - 4 2 - aryl (CF3) - pyrrole 14.32 b - 27.9 86 11.8:1 5 2 - thionaphthol 13.99 ~5 22.3 33 46.4:1 6 2,6 - dimethylphenylamidate 15.02 b - 30.4 27 4.7:1 7 - pyridinylpyrrole 13.64(pyr)/>1 5 (Py) - 25.6 d 76 4.4:1 8 5 - fluoroindole 13.16 - 22.2 88 56:1 9 3 - methylindole 12.49 - 22.6 74 6.5:1 10 2,3 - dimethylindole 11.38 - 25.1 73 >100:1 11 3 - methyl - 5 - methoxyindole 12.22 - 22.6 62 13.7:1 a LDP listed is that for 2 - methyl - 4 - methoxyphenol. b LDP value is artificially increased by steric affects in the measurement. c Approximated from a close derivative. d Calculated %V bur based on DFT optimized structure. See the SI for more details. Looking at the two HX ligands initially examined, there are four major differences that could likely affect the equilibrium processes: boiling point, pK a of the acidic H, donor ability of X ¯ as a ligand to a high valent metal, and sterics. 22, 30 With these four factors in mind, a series of experiments in whi ch different HX ligands, where the above 4 properties were varied, are used in excess (10 equiv) in the hydroamination of aniline and 1 - octyne. The results, shown in Table 5. 5 span a considerable range for these 4 properties and suggest some correlations between the properties listed above and the equilibrium. For these comparisons, pK a values for each ligand (or a close derivative) were referenced from the literature. The ter ms describing donor ability to Ti (LDP) and size (%V bur ) are parameters derived from a Cr(VI) model complex in the Ligand Donor Parameter system. 22, 30 This system was thoroughly discussed in chapters 2 and 3. These parameters to describe donor ability and size 387 of a given X ¯ ligand on a high valent metal have previously shown excellent correlation to homogeneous Ti hydroamination catalysts. 22 Thus, it seems these ligand property descriptors could be informative here, in determining what ligand properties are affecting the formation of the active Ti species with this heterogeneous system. Two of the l igand properties appear to have little effect on the equilibrium properties of the X ¯ ligands in this system. First, the data provided by this small group shows no correlation between size and the metric which provides an estimate of the equilibrium betwee n Ti(X) 3 /SiO 2 700 and Ti(=NR)X/SiO 2 700 , which here is the yield of hydroamination product (%) after 1 h. Even taken to the extreme, we note that the bidentate pypyr ligand outperforms many substantially smaller ligands, yet at the same time, many of the sma ll ligands are also high - yielding. Additionally, it does not appear that the boiling point of the ligand, or relative volatility, is hugely important. Both pyrrole and t BuOH, for example boil well below the reaction temperature, at 130 °C and 83 °C respect ively. Despite the fact that both ligands will be vaporized to some extent, when free in solution, under the reaction conditions, pyrrole facilitates equilibrium with aniline, while t BuOH halts catalytic activity. The donor ability of the X ¯ ligand and th e pK a seem to be exhibiting the greatest affect over the equilibrium processes here. To some extent, this makes sense, as pK a - donor ability of X ¯ ects. There is a degree of inherent relation between the two descriptors. However, as ligand exchange in this system also involves protonation/deprotonation events, pK a may also play other roles in the thermodynamic and kinetic controls of the ligand excha nges on Ti, specific to proton behavior . An illustration of how the X ¯ ligands may be impacting the catalytic cycle is shown in Fig. 5.1 1 , which is a modified version of the Bergman mechanism for hydroamination. 388 Figure 5 . 11 A traditional homgogeneous mechanism (i.e. Bergman or Doye) of hydroamination shown with a Ti catalyst ( top ), and a modified version of the mechanism where HX may participate in the deprotonation of the aza - titanacyclobutene intermediate and impact the equilibrium formation of the active Ti - imido species ( bottom ). Comparing Entries 1, 2, 3, and 10 from Table 5. 5 provides a few insights. First, looking at 1 and 3, t he two ligands have similar pK a values but different donor abilities as ligands. This suggests that the donor abilities of these ligands to Ti impact the equilibrium, with the stronger donor shutting down the equilibrium entirely (Entry 3), favoring the Ti (X) 3 /SiO 2 700 species by 389 equilibrium. Now looking at Entries 2 and 10, these HX ligands have much more similar donor abilities but very different pK a values. HX in Entry 10 has an LDP value lower than that of HX in Entry 2, meaning it is a slightly better o verall donor to the metal; however, the pK a of HX in Entry 10 is much higher than that of HX in Entry 2. In this case, we see that Entry 10 allows for much more product formation than Entry 2, which completely shuts down the catalysis (74% vs 0%). Collecti vely, these comparisons suggest that a higher LDP (weaker overall donor) and a higher pK a - donating X - ) are needed to access the equilibrium regime in which the active Ti(=NR)X/SiO 2 700 species can be formed in the presence of exc ess HX (10 equiv). If the HX of choice is too donating to Ti as the conjugate base X , or the conjugate base is not easily protonated by aniline, no equilibrium is observed, and thus no hydroamination product is observed. 5.7 Catalyst Recycling from Ti(X) 3 /S iO 2 700 Precatalysts: Enhanced Recyclability through Poisoning Perhaps more interesting than identifying trends in these equilibria presented in the previous section, is the application of these equilibria to the catalyst as a means of recycling it for multiple uses. We know that even with an excess of these HX type ligands, some portion of the Ti sites can reach the Ti(=NR)X/SiO 2 700 active species through equilibrium formation, and the reaction can progress. However, when the Ti sites are not active in the cycle, they presumably resemble Ti(X) 3 /SiO 2 700 . By using the right X ¯ ligand, which has a weak enough Ti - X interaction (in terms of equilibria) that the imido can be formed, but a strong enough Ti - X interaction to outcompete the binding of excess subst rate(s) when the reaction nears completion, we may be able to prevent irreversible surface - bound species from forming, and thus prevent the deactivation of the catalyst. 390 Based on the ligand property screening for active HX equilibria, we were able to nar row in on the range of LDP values and pK a values that will likely lead to productive hydroamination as opposed to shutting down the catalysis. Ligands such as pyrroles and indoles are in the ideal range for both donor ability and pK a of the acidic proton. With a selection of several of these ligands, we examined the ability of the catalysts to be recycled. Based on the results in Table 5.4, we can see that this technique, of including excess ligand in the hydroamination reaction, does in fact preserve catal ytic activity for a second use of the catalyst material. Interestingly, upon a second use, we observed changes in the regioselectivity of the product with all but one of the X ¯ ligands employed. The best result was observed with 5 - fluoroindole; more reuses of this catalyst material were pursued. Table 5 . 6 Recycling experiments with 10 equiv of a variety of HX ligands added to the hydroamination of 1 - octyne and aniline under the general conditions. Yields for t he initial run with fresh catalyst (run 1), and a subsequent use (run 2) are shown. HX Run Number Yield (%) Regioisomer Ratio pyrrole 1 81 51.8 :1 2 9 4.2 : 1 pypyr 1 76 48.8 :1 2 39 7.2 : 1 3 - methyl - 5 - methoxyIndole 1 62 13.7 : 1 2 15 9.1 :1 2,3 - dimethylindole 1 74 >100 :1 2 56 5.1 : 1 3 - methylindole 1 73 6.5 : 1 2 56 6.6 : 1 5 - fluoroindole 1 88 55 :1 2 74 6.5 :1 Upon additional reuses of the 5 - fluoroindole doped catalysts we see the yield progressively decrease ( Table 5. 6 ). This result is similar to the observed reusability with [Ti]200, where loss of the catalytically active metal was observed across 5 uses of the material. Likewise, with the [Ti]700(5 - F - indole) species, after 5 runs, ICP - OES analysis indicates about a 30% loss of the mass of Ti present in the catalyst material. This suggests that two different mechanisms of catalyst 391 deactivation, minimally, are contributing to the loss of catalytic activity upon several uses of the material. The first is a slow loss of the catalytically active metal and the second is a buildup of inhibitory species. In this case, the inhibitory species could be interaction of substrates with the catalyst material in a non - innocent way or from an interaction with the high concentrations of X ¯ on the heterogeneous catalyst. Table 5 . 7 Results of reusing the [Ti]700 catalyst with 10 equiv of 5 - fluoroindole (each trial) to perform the hydroamination of aniline and 1 - octyne. Catalyst Trial % yield of hydroamination product Regioselectivity Ratio 1 86 >100 : 1 2 73 6.6 : 1 3 56 5.6 : 1 4 2 1 isomer 5 Not observed N/A of the material shows promise but is far from a perfect solution to catalyst recyclability. Experimentally, these observations demonstrate that there can be compe titive binding to the active metal between the doped ligand and the other substrates in solution. While the high concentrations of ligand can preserve some of the metal sites, it is likely via an equilibrium exchange with protic species in solution. Eventu ally (i.e. with enough usage of the catalyst) the majority of the Ti sites succumb to irreversible binding with other species in solution; while the excess ligand concentration slows this process down, it still catches up with the material after several us es. At the same time, this also enables surface extraction of the active metal. This suggests that adjusting the ligand concentration in solution during the catalytic reactions may be one means by which to optimize the reactions and lead to improved reusab ility, but even with ligands that show high rates of exchange with the catalyst, there is a limit to the amount that can be productively added. 392 5.8 Conclusions Exploration of the catalytic activity of the [Ti]700 material with the addition of ligands has p rovided several big improvements to the overall performance of this material. Reaction times have been dramatically reduced, such that iminoamination with several different sets of substrates provides over 90% yield of the desired product(s) in less than 1 hour of reaction time. Additionally, the regioselectivity of the catalyst can be dramatically altered based on the identity of the ligand. The practical advantage of these catalysts has also been demonstrated by the high yields obtained with the one - pot - t wo - step production of complex heterocycles (quinolines and 2 - amino - 3 - cyanopyridines) utilizing [Ti]700(X) species. Experimental determination of the rate law governing the iminoamination reaction with [Ti]700(X) has also given us insight into what strateg ies may improve the rates of these catalysts even further. It appears that formation of the active Ti - imide species is rate - limiting, with inverse - first - order dependence on both aniline and isonitrile concentration. However, rate increases were noted by in creasing the electron - donor ability of the ancillary X ¯ ligand. This suggests that a more donating ancillary ligand increases the equilibrium - based formation of the active Ti - imide relative Based on the observation of slow catalyst deactivation over the course of the iminoamination reaction, even small rate increases stand to improve the overall conversion observed in these reactions. This observation of catalyst deactivation agrees with previous observations that bo th [Ti]700 and [Ti]700(X) are not reusable after a single application to iminoamination catalysis. Non - innocence of both products and reactants are suspected to contribute to these deactivation pathways. 393 Preliminary investigations show that running hydroa mination reactions with [Ti]700(X) in the presence of a substantial excess of HX ligand species facilitates the some reusability of the material or rather that the excess ligand slows down the rate of deactivation. Eventually, i.e. after several runs, all of the Ti sites do still become deactivated, and this is accompanied by a loss of some of the active metal. So, while this strategy is far from a perfect system by which to facilitate catalyst reuse, it provides experimental evidence that by manipulating t he reaction conditions, we may be able to prolong the life of the catalyst in solution. Achieving perfect reusability, especially with iminoamination, will require further study. 5.9 Experimental General Considerations All manipulations involving catalytic reaction set - up and catalyst material handling were carried out under inert atmosphere (N 2 ), either in an MBraun glovebox or using standard Schlenk technique. Manipulations in air were primarily limited to preparation of ICP samples, organic product handli ng, and catalytic product isolation via column chromatography. The catalyst material, [Ti]700, was prepared according to procedures listed in Chapter 4. The preparation of the [Ti]700X variants was generally performed in situ , as described below. The sol vents n - hexane, toluene, and n - pentane were dried by passage over an activated alumina column and sparging with N 2 prior to use. The solvents p - cymene and C 6 D 6 were dried over CaH 2 and distilled under vacuum and N 2 , respectively, prior to use. The solvents used for column chromatography, organic workup, and routine complex characterization (GC or NMR) included hexanes, ethyl acetate, ether, triethylamine, and CDCl 3 . These complexes were purchased commercially and used as received. 394 All substrates employed in catalytic reactions were dried prior to use. The alkynes, 1 - octyne, 1 - phenylpropyne, and phenylacetylene were purchased from Alfa Aesar, drired over Na 2 SO 4, and distilled under N 2 prior to use. Aniline and 3,5 - dimethylaniline were dried over CaH 2 and di stilled under vacuum prior to use. Cyclohexylisonitrile was synthesized according to literature procedures. 32 The following phenol ligands were purchased from commercial vendors: phenol; 4 - methoxyphenol; 2 - tert - butylphenol; 2,4,6 - tri - tert - butylphenol; 2,6 - di - tert - butyl - 4 - bromophenol; 2,6 - di - tert - butylphenol; 2 - phenylphenol; and 8 - hydroxyquinoline. All phenols were purified by sublimation under reduced pressure prior to use. The following pyrrole and indole ligands were purchased from commercial vendors: pyrrole, 2 ,3 - dimethylindole, 3 - methylindole, and 5 - fluoroindole. These ligands were dried azeotropically with toluene using a Dean - Stark apparatus prior to use. Additionally, 1 - adamantanol, triphenylsilanol, 2 - thionaphthol, and benzoic acid were also purchased from commercial vendors. These ligands were purified by sublimation prior to use. The amidate and thioimidate ligands were synthesized from published procedures. 33, 34 2 - ( N,N - dimethylaminomethyl) - 4,6 - di - tert - butylphenol was synthesized via a Mannich reaction. 35 3 - phenyl - 1 - naphthalenol was donated by the Wulff group at MSU. The 2 - aryl - substituted pyrroles, 2 - (3,5 - bis(trifluoromethyl)phenyl)pyrrole and 2 - pyridinyl - pyrrole, were synthesized using Suzuki reactions between t he Boc - protected - boronic acid - substituted pyrrole derivatives and halogenated aryl groups. 36, 37 These ligands are the only ones considered that were more in - catalyzed C N bond formation chemistry. 395 NMR Solution phase NMR was utilized to perform the Si O 2 700 surface titration experiments, as well as routine characterization of isolated products from catalytic reactions. Routine characterization spectra were obtained using an Agilent DDR2 500 MHz NMR spectrometer equipped with a 5 mm PFG OneProbe operatin g at 499.84 MHz ( 1 H) and 125.73 MHz ( 13 C). 1 H NMR titrations of the SiO 2 700 with Ti(NEt 2 ) 4 was performed using a Varian Inova 500 spectrometer equipped with a 5mm pulse - field - gradient (PFG) switchable broadband probe operating at 499.84 MHz ( 1 H). 1 H NMR chemical shifts were referenced to residual CHCl 3 in CDCl 3 as 7.26 ppm, or residual C 6 HD 5 in C 6 D 6 as 7.16 ppm. 13 C NMR chemical shifts are reported relative to 13 CDCl 3 as 77.16 ppm, or ( 13 C)C 5 D 6 as 128.06 ppm. X - ray Crystallography X - ray crystal structure data was collected at the Center for Crystallographic Research at MSU. The data was collecte d using either Mo or Cu - 396 NCr(N i Pr 2 ) 2 (OPh - 4 - Br) Single Crystal X - ray Data Details Figure 5 . 12 Crystal data and structure refinement for twin5. Identification code twin5 Empirical formula C 18 H 32 BrCrN 3 O Formula weight 438.37 Temperature/K 173(2) Crystal system monoclinic Space group P2 1 /n a/Å 9(2) b/Å 9.228 397 c/Å 25.647 90 93.26 90 Volume/Å 3 2150(473) Z 4 calc g/cm 3 1.355 - 1 2.400 F(000) 912.0 Crystal size/mm 3 0.498 × 0.375 × 0.144 Radiation 3.182 to 55.204 Index ranges - Reflections collected 8059 Independent reflections 8059 [R int = ?, R sigma = 0.1118] Data/restraints/parameters 8059/0/196 Goodness - of - fit on F 2 1.066 R 1 = 0.1189, wR 2 = 0.3042 Final R indexes [all data] R 1 = 0.1578, wR 2 = 0.3279 Largest diff. peak/hole / e Å - 3 1.70/ - 1.36 398 (2 - [(3,5 - dimethylphenyl)amino] - 6 - methyl - 5 - phenyl - nicotinonitrile) Single Crystal X - ray Data Details Figure 5 . 13 Crystal data and structure refinement for tri_early2_a. Identification code tri_early2_a Emp irical formula C 21 H 19 N 3 Formula weight 313.39 Temperature/K 172.99 Crystal system triclinic Space group P - 1 399 a/Å 8.0856(3) b/Å 9.8636(3) c/Å 12.1249(5) 66.347(2) 82.136(3) 70.711(2) Volume/Å 3 836.03(6) Z 2 calc g/cm 3 1.245 - 1 0.579 F(000) 332.0 Crystal size/mm 3 0.655 × 0.302 × 0.12 Radiation 7.96 to 144.658 Index ranges - - - Reflections collected 14176 Independent reflections 3142 [R int = 0.0280, R sigma = 0.0239] Data/restraints/parameters 3142/0/293 Goodness - of - fit on F 2 1.027 R 1 = 0.0363, wR 2 = 0.0959 Final R indexes [all data] R 1 = 0.0403, wR 2 = 0.1000 Largest diff. peak/hole / e Å - 3 0.19/ - 0.19 400 GC Experiments and in situ Quantification GCMS data was collected on an Agilent 5973 MSD with a 6890N series GC. GCFID data was collected on a Hewlett Packard 6890 series GC system, and standardized against dodecane as an internal standard. The iminoamination products were quantified in situ by ut ilizing GCFID calibration curves generated with authentic samples of the isolated products for each derivative, standardized against internal dodecane. ICP - OES ICP data was collected on a Varian 710es ICPOES spectrometer. A 1000 ppm Ti ICP standard in 2% HNO 3 was purchased from Sigma and used as received to prepare an external calibration curve. The Ti - SiO 2 200 samples were then measured in triplicate, and quantified from the external calibration, allowing for the mass of Ti in each sample to be determined. Catalytic Reactions General Iminoamination Procedure with [Ti]700(X): A 15 mL pressure tube was charged with [Ti]700 (0.05 mmol, 163 mg), HX ligand (0.05 mmol), 1.0 mL p - cymene, and a stir bar. This mixture was stirred for 5 min at room te mperature. To the catalyst mixture was added a separate 1.0 mL solution of H 2 NPh (1 mmol, 93 mg), alkyne (2 mmol), and CyNC (1.5 mmol, 164 mg) in p - cymene. The pressure tube was sealed and transferred from the glovebox to a preheated 180 °C aluminum block. The tube was heated with stirring for 1 h; after the reaction was complete, the tube was removed from heat and cooled ambiently to room temperature. The pressure tube was centrifuged, compacting the catalyst material into a pellet at the bottom of the tub e. The organic reaction solution was removed (by pipette) from the catalyst material for GC analysis. 401 Table 5 . 8 Yield and regioselectivity observed for a variety of ligands screened for the iminoamination of aniline, 1 - octyne, and CyNC. Ligand Yield (%) Regioselectivity monodentate pyrrole 0 na 2,4 - di - t Buphenol 77 1.41:1 2 - thionapthol 16 1.35:1 2 - Ar CF3 - pyrrole 15 1.4:1 2,4,6 - tri - t Buphenol 95 2.98:1 p - CF 3 - phenol 0 Na Ph 3 SiOH 96 2.24:1 1/2 vapol 71 1.32:1 AdOH 62 1.54:1 No ligand 0 Na 2 - phenylphenol 53 1.7:1 bidentate 6 - dimethylAmino - 2,4 - di - t Bu - phenol 98 2.29:1 2,6 - dimethylphenylamidate 99 1.31:1 8 - hydroxyquinoline 77 1.65:1 pypyr 79 1.2:1 2,6 - dimethylphenylthioamidate 95 2.08:1 dipp - thioamidate 65 1.97:1 dipp - amidate 99 1.4:1 Ar(CF 3 ) 2 - amidate 55 1.45:1 Benzoic acid 0 a na a Large amount of formamidine product noted in this reaction, but no iminoamination product. 402 Table 5 . 9 Yield and regioselectivity observed for a variety of ligands screened for the iminoamination of aniline, 1 - phenylpropyne, and CyNC. Ligand Yield (%) Regioselectivity monodentate 2,4,6 - tri - t Buphenol 67 5.4: 1 HOSiPh 3 54 7.5:1 NHAdAr 69 8.34:1 2,4 - di - t Buphenol 42 8.6:1 NH 2 Ph or CyNC 89 10.35:1 bidentate 2,6 - dimethylphenylamidate 99 10.28:1 6 - dimethylaminomethyl - 2,4 - di - t Bu - phenol 47 15.8:1 2,6 - dimethylphenylthioamidate 66 13.1:1 8 - hydroxyquinoline 57 11.5:1 pypyr 76 9.64:1 Dipp - Amidate 96 14.43:1 Dipp - Thioamidate 77 9.04:1 403 Table 5 . 10 Yield and regioselectivity observed for a variety of ligands screened for the iminoamination of aniline, phenylacetylene, and CyNC. Ligand Yield (%) regioselectivity monodentate 2,4,6 - tri - t BuPhOH 72 1.5:1 2,4 - di - t Buphenol 21 1.7:1 NHAdAr 47 4.89:1 2 - phenyl - phenol 31 2.9:1 aniline/CyNC 52 6.3:1 bidentate 6 - dimethylaminomethyl - 2,4 - di - t Bu - phenol 31.5 2.7:1 diip - amidate 67 3.6:1 pypyr 51 4.3:1 2,6 - dimethylphenylamidate 99 5.6:1 2,6 - dimethylphenylthioamidate 36 2.1:1 404 General Procedure for Catalyst Recycling (Iminoamination) An initial reaction with the catalyst material was set up and performed according to the general procedure above. After the reaction was finished and centrifuged, the pressure tube was transferred b ack to the glovebox and the organic reaction solution was decanted for GC analysis. The catalyst material was rinsed with benzene (5 mL) and pentane (5 mL) on a fritted funnel. The material was briefly dried under vacuum and transferred to a new pressure t ube. 1.0 mL of p - cymene was added and the catalyst material was stirred. To this mixture was added a volumetrically prepared 1.0 mL solution containing H 2 NPh (1 mmol, 93 mg), alkyne (2 mmol), and CyNC (1.5 mmol, 164 mg) in p - cymene. The pressure tube was s ealed and transferred to an aluminum block (180 °C) and the reaction heated, with stirring, for 1 hour. The reaction solution was centrifuged and the organic solution was then decanted and analyzed by GCMS and GCFID. (Note, even when the rinsing and filtra tion steps were omitted, the same results were observed on a second use with a variety of ligands and different substrates, such that <5% 3CC yield was observed on subsequent runs). General Procedure for Catalyst Recycling (Hydroamination with excess HX Ad ded ) A 15 mL pressure tube was charged with 163 mg of [Ti]700 (0.05 equiv, 0.05 mmol), a stir bar, the HX ligand of interest (0.5 equiv, 0.5 mmol) and 1.0 mL of p - cymene. This mixture was stirred for 5 min at room temperature and a 1.0 mL solution containi ng 93 mg H 2 NPh (1 equiv,1 mmol) and 220 mg 1 - octyne (2 equiv, 2 mmol) in p - cymene was added. The pressure tube was sealed and transferred to a 180 °C aluminum well plate. The reaction was heated and stirred for 1 h; it was cooled ambiently, centrifuged and returned to the glovebox. The organic solution was decanted for GC analysis (trial 1). The catalyst material was rinsed with benzene (5 mL) and pentane (5 mL) on a fritted funnel. The catalyst was dried under reduced pressure and transferred to a new pres sure 405 tube. The HX ligand and substrates were added in the same manner as the initial run and the reaction performed again (trial 2). This process was repeated again as necessary. General Procedure for Kinetic Analysis For kinetics runs, monitored by GC, 6 - 8 trials of each set of reaction conditions in Table 1 were prepared. To achieve consistency across the reactions, the catalyst material was prepared prior to the reactions. In an Erlenmeyer flask, 2 g of [Ti]700 (0.62 mmol) was stirred as a suspension in toluene (20 mL). To this suspension was added 162 mg of 2,4,6 - tri - tert - butylphenol (0.62 mmol). The suspension was stirred for a total of 1 h at room temperature after addition. The catalyst material was collected by filtration and dried under reduced pre ssure. This pre - formed catalyst was used in the kinetics experiments. The following description of a reaction set - up utilizes the amounts from Entry 1 of Table 5.1 as a representative example of how these reactions were run. Representative Kinetic Procedure with Conditions from Entry 1 8 separate pressure tubes were charged with 88 mg of [Ti]700(OPh 2,4,6 - tri - tbutyl ) (0.025 mmol, 0.05 equiv) and a stir bar. To each tube 2.0 mL of a volumetrically prepared solution in p - cymene t hat was 0.25 M H 2 NPh (0.5 mmol, 1 equiv), 0.25 M CyNC (0.5 mmol, 1 equiv) and 0.51 M 1 - octyne (1 mmol, 2 equiv) and 0.05 M dodecane (as internal standard for GC analysis), was added. The tubes were sealed and transferred from the glovebox to a 180 °C alumi num well plate. At timed intervals from 0 - 100 min, samples were removed from heat and analyzed by GC - MS and GC - FID to identify reaction products and quantify the amount of iminoamination product in solution. Note, the tubes contain small amounts of soluti on and cool rapidly. Based on previous observations, we know that by the time the reaction temperature has reached 140 °C, the catalyzed reaction will have slowed by 1 - 2 orders of magnitude. No specific quenching step was taken upon removal of reaction ves sels from heat, as cooling the reaction sufficiently eliminates further 406 catalytic activity, and opening the reaction vessel at 180 °C is not advisable. As soon as the tube was cool enough to handle, it was opened to air, thus killing any residual active ca talyst. 407 Product Isolation from Catalytic Reactions : Iminoamination with H 2 NPh, 1 - octyne, and CyNC Scheme 5 . 5 Iminoamination reaction catalyzed by [Ti]700 with 2,4,6 - tri - tert - butylphenoxide. Synthetic Procedure A 15 mL pressure tube was charged with 320 mg [Ti]700 (0.05 equiv, 0.1 mmol), a stir bar, 1 mL p - cymene, and 26 mg of 2,4,6 - tri - tert - butylphenol (0.05 equ iv, 0.1 mmol). This mixture was stirred for 5 min at room temperature. Then a 2 mL solution containing 186 mg H 2 NPh (1 equiv, 2 mmol), 327 mg CyNC (1.5 equiv, 3 mmol), and 440 mg 1 - octyne (2 equiv, 4 mmol), was added to the pressure tube. The tube was seal ed and transferred to a 180 °C aluminum block, where it was heated and stirred for 1 h. The reaction was cooled and centrifuged to compact the catalyst material at the bottom of the tube. The crude reaction solution was decanted and the 3CC compound isolat ed by column chromatography (Alumina, Hexanes(1%TEA) gradient from 0 - 30% Et 2 O). The 3CC product was isolated as a yellowish oil as an isomeric mixture of A and B (yield: 330 mg, 53%). HRMS 21 H 33 N 2 : 313.2644; found: 313. 2641. 1 H NMR (500 MHz, CDCl 3 ) 9.90 (s, 1H), 6.98 (q, J = 8.4 Hz, 1H), 6.80 (d, J = 7.8 Hz, 1H), 6.72 (d, J = 7.8 Hz, 1H), 4.67 (d, J = 7.8 Hz, 1H), 3.04 (dt, J = 17.3, 5.6 Hz, 1H), 2.57 (t, J = 7.8 Hz, 0H), 2.21 2.09 (m, 2H), 1.93 1.70 (m, 1H), 1.63 1.54 (m, 1H), 1.45 (dd, J = 10.0, 5.1 Hz, 2H), 1.30 (s, 4H), 1.23 1.15 (m, 3H), 0.89 (t, J = 6.5 Hz, 2H), 0.82 (t, J = 7.1 Hz, 2H). 13 C NMR (126 MHz, CDCl 3 ) 170.83, 153.60, 151.46, 147.52, 147.09, 145.38, 129.32, 128.61, 122.01, 121.53, 117.60, 91.82, 408 62.96, 57.11, 34.94, 34.55, 33.77, 33.13, 32.71, 31.78, 31.60, 31.33, 29.45, 29.02, 28.92, 28.80, 25.87, 25.66, 24.82, 24.64, 22.79 (d, J = 12.9 Hz), 22.61, 14.30, 14.19. gCOSY NMR correlation list: 1 H NMR (500 MHz, CDCl 3 ) 7.28, 7.31, 6.81, 7.27, 7.27, 4.67, 1.82, 1.86, 1.89, 1.29, 1.60, 1.60, 1.49, 1.50, 1.46, 1.46, 1.45, 1.30, 3.02, 3.06, 1.30, 2.58, 2.58, 2.58, 1.18, 1.17, 1.85, 1.17, 1.84, 2.15, 3.06, 1.84, 2.15, 1.82, 2.14, 1.77, 1.79, 1.76, 3.02 , 0.88, 1.87, 3.02, 0.83, 0.82, 0.82, 1.45, 1.30, 1.30, 1.22, 1.22 . 1 H NMR (500 MHz, CDCl 3 ) 8.26, 7.49, 7.25, 6.98, 6.80, 6.72, 3.05, 3.05, 3.02, 3.01, 2.59, 2.56, 2.37, 2.33, 2.17, 2.14, 2.14, 1.90, 1.88, 1.83, 1.72, 1.62, 1.59, 1.56, 1.56, 1.50, 1.49, 1.48, 1.46, 1.45, 1.44, 1.43, 1.42, 1.41, 1.39, 1.33, 1.33, 1.32, 1.32, 1.30, 1.28, 1.28, 1.24, 1.21, 1.19, 1.16, 0.90, 0.87, 0.84, 0.81. 409 Iminoamination with H 2 NPh, 1 - phenylpropyne, and CyNC: Scheme 5 . 6 Iminoamination reaction catalyzed by [Ti]700 with 2,6 - dimethylphenylamidate ligand and an internal alkyne. Synthetic Procedure: A 15 mL pressure tube was charged with 326 mg [Ti]700 (0.05 equiv, 0.1 mmol), 28 mg of 2,6 - dimethylphenylamidate (0.05 equiv, 0.1 mmol), a magnetic stir bar, and 1 mL p - cymene. The mixture was stirred for 5 minutes, and then a 2.0 mL p - cymene solution containing 186 mg aniline (2 mmol, 1 equiv), 327 mg of CyNC (3 mmol, 1.5 equiv), and 348 mg of 1 - phenylpropyne (3 mmol, 1.5 equiv) was also added to the pressure tube. The tube was sealed and transferred from the glovebox to a preheated, 180 °C aluminum block. The tube was heated, with stirring, for 1 h. The pressure tube was removed f rom heat and allowed to ambiently cool to room temperature. The contents were centrifuged and the liquid portion decanted. This crude reaction solution was separated by column chromatography (Al 2 O 3 , Hex(1%TEA), gradient Et 2 O from 0 to 20%) to yield 410 mg (64%) of the major regioisomer (A) as a yellow oil. Note, the isolated product contains only the main regioisomer (A) to the limits of detection of our GC instruments and NMR. HRMS 22 H 27 N 2 : 319.2174; found: 319.2164. 1 H NMR (500 MHz, CDCl 3 ) 10.75 (s, 1H), 7.34 7.28 (m, 4H), 7.27 (d, J = 1.6 Hz, 2H), 7.22 7.17 (m, 1H), 7.03 (m, 1H), 6.92 (s, 1H), 6.91 6.87 (m, 2H), 3.09 3.01 (m, 1H), 1.95 1.87 (m, 3H), 1.83 (s, 3H), 1.75 (d, J = 3.3 Hz, 4H), 1.57 (s, 4H), 1.43 1.27 (m, 4H). 13 C NMR (126 MHz, CDCl 3 410 165.82, 150.63, 147.53, 142.32, 130.52, 128.68, 127.98, 125.36, 122.31, 121.58, 107.65, 34.52, 25.50, 24.96, 24.62, 20.33. gCOSY NMR Correlation List: 1 H NMR (500 MHz, CDCl 3 4, 6.93, 7.32, 7.31, 7.31, 7.30, 7.31, 1.36, 1.33, 1.91, 3.08, 1.81, 1.94, 1.71 . 1 H NMR (500 MHz, CDCl 3 7.18, 7.05, 7.01, 6.89, 1.92, 1.75, 1.39, 1.37, 1.36, 1.33, 1.30. 411 Iminoamination of H 2 NPh, phenylacetylene, and CyNC: Scheme 5 . 7 Iminoamination reaction catalyzed by [Ti]700 with 2,6 - dimethylphenylamidate and an aromatic alkyne. Synthetic Procedure: A 15 mL pressure tube was charged with 326 mg [Ti]700 (0.05 equiv, 0.1 mmol), 28 mg of 2,6 - dimethylphenylamidate (0.05 equiv, 0.1 mmol), a magnetic stir bar, and 1 mL p - cymene. The mixture was stirred for 5 minutes, and then a 2.0 mL p - cymene solution containing 186 mg aniline (2 mmol, 1 equiv), 327 mg of CyNC (3 mmol, 1.5 equiv), and 306 mg of phenylacetylene (3 mmol, 1.5 equiv) was also added to the pressure tube. The tube was sealed and transferred from the glovebox to a preheated, 180 °C aluminum block. The tube was heated, with stirring, for 1 h. The pressure tube was removed from heat a nd allowed to ambiently cool to room temperature. The contents were centrifuged and the liquid portion decanted. This crude reaction solution was separated by column chromatography (Al 2 O 3 , Hex(1%TEA), gradient Et 2 O from 0 to 20%) to yield 312 mg (51%) of t he major regioisomer (A) as a yellow oil. Note, the isolated product contains only the main regioisomer (A) to the limits of detection of our GC instruments and NMR. HRMS 21 H 25 N 2 : 305.2018; found: 305.2015. 1 H NMR (500 M Hz, CDCl 3 ) 11.65 (s, 1H), 8.17 (d, J = 2.9 Hz, 1H), 7.59 (d, J = 2.9 Hz, 1H), 7.40 7.28 (m, 6H), 7.23 7.16 (m, 1H), 7.12 (dd, J = 8.5, 1.1 Hz, 2H), 7.08 (td, J = 7.3, 1.2 Hz, 1H), 3.25 (ddt, J = 13.5, 9.3, 3.9 Hz, 1H), 2.02 1.91 (m, 2H), 1.86 1.78 (m, 2H), 1.68 1.62 (m, 2H), 1.52 (m, 2H), 1.48 1.38 (m, 2H). 13 C NMR (126 MHz, CDCl 3 ) 152.48, 150.72, 149.33, 140.95, 129.23, 128.60, 125.48, 124.67, 123.24, 119.10, 106.34, 34.55, 31.62, 25.60, 24.40. 412 gCOSY NMR Correlation List: 1 H NMR (500 MHz, CDCl 3 18, 7.27, 7.11, 7.32, 7.35, 1.06, 1.52, 1.42, 1.36, 1.85, 3.24, 2.09, 1.82, 1.62, 2.55, 2.56, 1.32 . 1 H NMR (500 MHz, CDCl 3 1.34, 1.07, 1.04, 0.89. 413 1 - Pot - Two - Step Quinoline Synthesis: A 15 mL pressure tube was charged with 163 mg of [Ti]700 (0.05 equiv, 0.05 mmol), 13 mg 2,6 - dimethylphenylamidate (0.05 equiv, 0.05 mmol), a stir bar, and 1.0 mL p - cymene. The mixture was stirred for 5 min at room tempe rature. A 1.0 mL solution of 93 mg H 2 NPh (1 equiv, 1 mmol), 232 mg 1 - phenylpropyne (2 equiv, 2 mmol), and 164 mg CyNC (1.5 equiv, 1.5 mmol) was added to the catalyst mixture and the pressure tube was sealed. The tube was transferred from the glovebox to a preheated 180 °C aluminum block and heated with stirring for 2 h. The tube was cooled ambiently to room temperature and then 2 mL of glacial acetic acid was added to the pressure tube. The tube was resealed and heated at 120 ° C for an additional 10 h. Th e tube was removed from heat and the contents neutralized with sodium bicarbonate solution once cooled (pH 7 - 8). The neutralized mixture was extracted with EtOAc and the organic layer concentrated by rotary evaportation to give provide a viscous brown resi due. The crude residue was purified by column chromatography (silica, Hexanes(1%TEA) - Hexanes(1%TEA)/EtOAc) yielding the product as an red oil (158 mg, 72%). The compound matches literature reports by 1 H NMR, 13 C NMR, and HRMS. HRMS : QTOF EI (positive ion 16 H 14 N: 220.1126; found: 220.1129. 1 H NMR (500 MHz, CDCl 3 ) 8.08 (d, J = 8.4 Hz, 1H), 7.97 (s, 1H), 7.79 (d, J = 8.0 Hz, 1H), 7.70 (t, J = 8.4, 6.9 Hz, 1H), 7.53 7.45 (m, 3H), 7.45 7.39 (m, 3H), 2.68 (s, 3H). 13 C NMR (126 MHz, CDCl 3 ) 157.38, 1 46.98, 139.87, 136.14, 135.74, 129.37, 129.21, 128.43, 128.36, 127.60, 127.45, 126.84, 126.06, 24.56. 414 One - pot - Two - Step Synthesis of a 2 - amino - 3 - cyano - pyridine (2 - [(3,5 - dimethylphenyl)amino] - 6 - methyl - 5 - phenyl - nicotinonitrile): A 15 mL pressure tube was charged with 163 mg of [Ti]700 (0.05 equiv, 0.05 mmol), 13 mg of 2,6 - dimethylphenylamidate (0.05 equiv, 0.05 mmol), a stir bar, and 1.0 mL of p - cymene and the mixture was stirred for 5 min at room temperature. A 1.0 mL solution containing 121 mg of 3,5 - dim ethylaniline (1 equiv, 1 mmol), 116 mg of 1 - phenylpropyne (1 equiv, 1 mmol), and 109 mg CyNC (1 equiv, 1 mmol) was then added to the catalyst mixture. The pressure tube was sealed and transferred from the glovebox to a heated aluminum block (180 ° C). The r eaction was heated for 2 h with stirring before it was cooled ambiently to room temperature. To the cooled reaction solution was added 125 mg of malononitrile (2 equiv, 2 mmol) and 75 mg of DBU (0.5 equiv, 0.5 mmol) with 2 mL of dry EtOH and 200 mg of acti vated 3 Å molecular sieves. This provided 228 mg (70%) of the product as a yellow solid that matched the reported 1 H and 13 C NMR spectra, but which also contained p - cymene and hexanes. The product was washed with cold hexanes and then recrystallized from h exane at - 20 ° C. This yielded 88 mg (28%) of X - ray quality single crystals. 1 H and 13 C NMR spectra of the crystalline solid were extremely pure. 1 H NMR (500 MHz, CDCl 3 ) 7.62 (s, 1H), 7.44 (dd, J = 8.0, 6.6 Hz, 3H), 7.39 (d, J = 7.1 Hz, 1H), 7.33 (s, 2H), 7.32 7.26 (m, 2H), 6.91 (s, 1H), 6.76 (s, 1H), 2.48 (s, 4H), 2.35 (s, 7H). 13 C NMR (126 MHz, CDCl 3 ) 160.29, 154.31, 142.18, 138.85, 138.70, 138.29, 129.18, 128.70, 128.01, 127.71, 125.36, 118.05, 116.82, 90.39, 24.23, 21.62. 415 Synthesis of NCr(N i Pr 2 ) 2 (OPh - 4 - Br) Scheme 5 . 8 Synthesis of 4 - Br - phenoxide LDP complex. Synthetic Procedure : A scintillation vial was charged with 50 mg of NCr(N i Pr 2 ) 3 (0.127 mmol, 1 equiv), a stir bar, and 4 mL of Et 2 O. This solution was chilled in a liquid nitrogen Coldwell for 10 minutes. A solution containing 24 mg of 4 - Br - phenol in 1 mL of Et 2 O was added to the chilled chromium solution in a dropwise manner. The re action rapidly changed color from beet to an orangish - red hue. The reaction was allowed to come to room temperature, with stirring, for 2 h. The volatiles were removed under reduced pressure to yield a dark red powdery residue. This residue was dissolved i n a minimal amount of pentane and filtered over Celite. This solution was chilled at - 35 °C for 2 days to yield a fine, powdery precipitate, which was the product. 43 mg of the product (72%) was collected and used for LDP measurements, elemental analysis, and NMR characterization of the complex. X - ray quality crystals were grown from a concentrated solution of the Cr complex in toluene, layered with n - hexane at - 35 °C, over the course of 7 days. 1 H NMR (500 MHz, CDCl 3 ) 7.22 (d, J = 8.8 Hz, 2H), 6.83 (d, J = 8.8 Hz, 2H), 5.04 (septet, J = 6.5 Hz, 2H), 3.75 (septet, J = 6.4 Hz, 2H), 1.85 (d, J = 6.2 Hz, 6H), 1.46 (d, J = 6.3 Hz, 6H), 1.15 (d, J = 6.4 Hz, 12H). 13 C NMR (126 MHz, CDCl 3 ) 165.93, 131.60, 119.56, 110.65, 58.46, 55.59, 30.50, 30.24, 21.48, 21.20. E lemental analysis 18 H 32 N 3 : C, 49.32; H, 7.36; N, 9.59. Found: C, 49.09; H, 7.41; N, 9.33. 416 Spectral Data Figure 5 . 14 1 H NMR of 3CC 1 in CDCl 3 (isomeric mixture of A and B). 417 Figure 5 . 15 13 C NMR of 3CC 1 in CDCl 3 (isomeric mixture of A and B). 418 Figure 5 . 16 gCOSY NMR of 3CC 1 in CDCl 3 (isomeric mixture of A and B). 419 Figure 5 . 17 GC trace of 3CC 1 (A and B) and MS fragmentation pattern for 3CC 1 (A). 420 Figure 5 . 18 GC trace of 3CC 1 (A and B) and MS fragmentation pattern for 3CC 1 (B). 421 Figure 5 . 19 13 C NMR of 3CC 2 in CDCl 3 . 422 Figure 5 . 20 13 C NMR of 3CC 2 in CDCl 3 . 423 Figure 5 . 21 gCOSY NMR of 3CC 2 in CDCl 3 . 424 Figure 5 . 22 GC trace and MS fragmentation pattern for 3CC 2 . 425 Figure 5 . 23 1 H NMR of 3CC 3 in CDCl 3 . * * * * Et 2 O, Hex 426 Figure 5 . 24 13 C NMR of 3CC 3 in CDCl 3 . * * * *Et 2 O, Hex 427 Figure 5 . 25 gCOSY NMR of 3CC3 in CDCl 3 . 428 Figure 5 . 26 1 H NMR of [Cr](O - Ph - 4 - bromo) in CDCl 3 (room temperature). 429 Figure 5 . 27 1 H NMR of [Cr](O - Ph - 4 - bromo) in CDCl 3 ( - 20 °C). 430 Figure 5 . 28 13 C NMR of [Cr](O - Ph - 4 - bromo) in CDCl 3 ( - 20 °C). 431 Kinetic Analysis Graphical Method to Determine Order in Catalyst and Reagents: The method utilized to analyze the experimental data graphically applies time - normalization to the x - axis, such that the variable concentration for each reagent is accounted for . The time - normalized reagent concentration is plotted against the concentration of product or consumption of a starting material (other than the one under study) across the time points sampled. The catalyst concentration dependence determined via the same method is treated differently, as catalyst concentration is assumed to be constant throughout the reaction. Using the following formula ( Equation 5.2), the x - axis values for time - normalized concentration in the reagent under study can be calculated for th e period of time between two sampled points. 29 ( Eq. 5. 2 ) In this equation, [A] is the concentration of A, the species under study. This effectively demonstrates the way the change in species A over time changes the rate of product formation. According to the literature, application of this method is not advisab catalyst is unknown or if it changes in an unknown way during the reaction. This problem is 28 In this system, we know that the catalyst deactivates under certain reaction conditions; however, this deactivation process does not appear to be rapid relative to the generation of product under many conditions ( Table 5.1). This is specifically ap parent at the beginning of each reaction. Comparisons between the full graphical analysis for order in catalyst, and traditional k obs (initial) observation, agree quite well. Thus, we feel confident in the assessment that these reactions are first - order in catalyst. Plots of the graphical method results for each reagent, as well as catalyst concentration, are shown in Fig. 5. 7. Examples of the applied graphical method functions are shown below in Table 5.9. 432 Table 5 . 11 Examples of Graphical Method Application to Entries in Table 2. Entry sample time 3cc t*[cat] n 1 0 0 0 0 1 10 0.045079 0.00022316 2 20 0.076953 0.00044632 3 30 0.113497 0.00066948 4 40 0.139191 0.00089263 6 60 0.187779 0.00133895 8 100 0.21032 0.00223159 2 0 0 0 0 1 10 0.087487 0.00043111 2 20 0.121113 0.00086223 3 30 0.159726 0.00129334 4 45 0.200862 0.00194001 6 60 0.208921 0.00258668 Entry sample time 3cc x n (t) 1 - octyne 1 0 0 0 10 1 10 0.045079 20 2 20 0.076953 30 3 30 0.113497 40 4 40 0.139191 60 6 60 0.187779 100 8 100 0.21032 - 6 0 0 0 10 1 10 0.043931 20 2 20 0.088563 30 3 30 0.104307 40 4 40 0.129061 50 5 50 0.143897 - Entry sample time 3cc x n (t) H 2 NPh 1 0 0 0 38.0300973 1 10 0.045079 84.4857829 2 20 0.076953 142.683592 3 30 0.113497 218.742693 4 40 0.139191 457.667275 6 60 0.187779 1731.10317 8 100 0.21032 - 433 4 0 0 0 22.9614895 1 10 0.062778 49.5168329 2 20 0.097242 91.6754708 3 35 0.100878 117.994244 4 45 0.11943 227.298338 5 85 0.128889 297.690438 6 110 0.135292 - 434 Application of KINSIM Modeling Program: The KINSIM modeling program, published by JPlus Consulting was used to simulate kinetics experiments, analyzing for rate constants and equilibrium processes following the proposed mechanism ( Fig. 5.7). The simulated trace showing concentration of 3CC is represented by the bright blue line (H). Plotting this trace against the experimen tal data points, for product concentration vs. time, allows for a direct comparison of the modeled kinetic behavior versus the experimentally observed behavior. With this method, and the reaction processes outlined in Fig . 5.7, the f ollowing plots were o btained, where K1, K2, and the forward rate constants described are kept relatively constant between the different reaction conditions examined. The values assigned to each equilibrium or rate constant are provided in the associated Table. 435 Figure 5 . 29 Results from simulated reactions versus the experimentally determined concentrations of 3CC product over time , Entry 1. 0 0.05 0.1 0.15 0.2 0.25 0 2000 4000 6000 8000 10000 3CC Concentration (M) Time (s) Experimental and Simulated Reaction Trace: Entry1 Experimental Simulation 436 Figure 5 . 30 Results from simulated reactions versus the experimentally determined concentrations of 3CC product over time , Entry 3. 0 0.05 0.1 0.15 0.2 0.25 0 2000 4000 6000 8000 10000 3CC Concentration (M) Time (s) Experimental and Simulated Reaction Trace: Entry 3 Experimental Simulation 437 Figure 5 . 31 Results from simulated reactions versus t he experimentally determined concentrations of 3CC product over time , Entry 4. 0 0.05 0.1 0.15 0.2 0.25 0 2000 4000 6000 8000 10000 3CC Concentration (M) Time (s) Experimental and Simulated Reaction Traces: Entry 4 Experimental Simulation 438 Figure 5 . 32 Results from simulated reactions versus the experimentally determined concentrations of 3CC product over time , En try 5. 0 0.05 0.1 0.15 0.2 0.25 0 2000 4000 6000 8000 10000 3CC Concentration (M) Time (s) Experimental and Simulated Reaction Trace: Entry 5 Experimental Simulation 439 Table 5 . 12 Simulated rate constant and equilibrium constant values used to fit each set of experimental reaction traces. Good agreement is noted among the values, with small variances in some rate constants during fitting. E5 E4 A+I = C K1 1 A+I = C K1 1 B+D = A K2 1.55 B+D = A K2 1.5 B+E > F k1 0.58 B+E > F k1 0.8204 F+I > G k2 0.3472 F+I > G k2 0.3472 G+D > H+B k3 0.367 G+D > H+B k3 0.367 B+H > J k4 0 B+H > J k4 0 A+H+D > J k5 0.3584 A+H+D > J k5 0.1469 E3 E1 A+I = C K1 1 A+I = C K1 1 B+D = A K2 1.55 B+D = A K2 1.55 B+E > F k1 0.3952 B+E > F k1 0.5507 F+I > G k2 0.28 F+I > G k2 0.3472 G+D > H+B k3 0.296 G+D > H+B k3 0.367 B+H > J k4 0 B+H > J k4 0 A+H+D > J k5 0.296 A+H+D > J k5 0.2278 440 Figure 5 . 33 Example simulation from KINSIM program. 441 LDP Data: All other complexes were synthesized as previously reported. 30 These complexes include NCr(N i Pr 2 ) 2 (OPh - 4 - t Bu), NCr(N i Pr 2 ) 2 (OPh), and NCr(N i Pr 2 ) 2 (OPh - 4 - OMe). The Cr complexes were then used to perform a series of 1 H NMR experiments that allow for the Cr - N i Pr 2 barrier to rotation to be measured exp erimentally. 30 Table 5 . 13 Details of LDP value measurements ) for the 4 - R - phenoxide ligands examined for electronic effects on [Ti]700(2,4,6 - tri - tert - butylphenoxide) iminoamination catalysis. Complex Rate(s - 1 ) (kcal/mol) T (K) (kcal/mol) Std. Deviation 4 - Br - OPh 1.17 14.41 249.1 12.17 0.009 4 - piperazino - OPh 1.76 13.90 243.8 11.70 0.005 OPh - 14.27 254.7 11.98 - 4 - t Bu - OPh 1.86 14.26 250.5 12.01 0.004 4 - OMe - OPh - 13.91 245.3 11.71 0.008 442 REFERENCES 443 REFERENCES ( 1 ) Odom, A. L., New C - N and C - C bond forming reactions catalyzed by titanium complexes. Dalton transactions 2005, (2), 225 - 233. ( 2 ) Odom, A. L.; McDaniel, T. J., Titanium - Catalyzed Multicomponent Couplings: Efficient One - Pot Syntheses of Nitrogen Heterocycle s. Acc. Chem. Res. 2015, 48 (11), 2822 - 2833. ( 3 ) Davis - Gilbert, Z. W.; Tonks, I. A., Titanium redox catalysis: insights and applications of an earth - abundant base metal. Dalton transactions 2017, 46 (35), 11522 - 11528. ( 4 ) Gilbert, Z. W.; Hue, R. J.; Tonks , I. A., Catalytic formal [2+2+1] synthesis of pyrroles from alkynes and diazenes via TiII/TiIV redox catalysis. Nature Chemistry 2015, 8 , 63. ( 5 ) Vujkovic, N.; Ward, B. D.; Maisse - François, A.; Wadepohl, H.; Mountford, P.; Gade, L. H., Imido - Alkyne Co Hydroamination Reaction. Organometallics 2007, 26 (23), 5522 - 5534. ( 6 ) Hazari, N.; Mountford, P., Reactions and Applications of Titanium Imido Complexes. Accounts of Chemical Research 2005, 38 (11 ), 839 - 849. ( 7 ) Pohlki, F.; Doye, S., The Mechanism of the [Cp2TiMe2] - Catalyzed Intermolecular Hydroamination of Alkynes. 2001, 40 (12), 2305 - 2308. ( 8 ) Rosner, T.; Le Bars, J.; Pfaltz, A.; Blackmond, D. G., Kinetic Studies of Heck Coupling Reactions Using Palladacycle Catalysts: Experimental and Kinetic Modeling of the Role of Dimer Species. Journal of the American Chemical Society 2001, 123 (9), 1848 - 1 855. ( 9 ) van Leeuwen, P. W. N. M., Decomposition pathways of homogeneous catalysts. Applied Catalysis A: General 2001, 212 (1), 61 - 81. ( 10 ) Proutiere, F.; Aufiero, M.; Schoenebeck, F., Reactivity and Stability of Dinuclear Pd(I) Complexes: Studies on the Active Catalytic Species, Insights into Precatalyst Activation and Deactivation, and Application in Highly Selective Cross - Coupling Reactions. Journal of the American Chemical Society 2012, 134 (1), 606 - 612. ( 11 ) Evans, D.; Yagupsky, G.; Wilkinson, G., Th e reaction of hydridocarbonyltris(triphenylphosphine)rhodium with carbon monoxide, and of the reaction products, hydridodicarbonylbis(triphenylphosphine)rhodium and dimeric species, with hydrogen. Journal of the Chemical Society A: Inorganic, Physical, Theoretical 1968, (0), 2660 - 2665. ( 12 ) Bonds, W. D.; Brubaker, C. H.; Chandrasekaran, E. S.; Gibbons, C.; Grubbs, R. H.; Kroll, L. C., Polystyrene attached titanocene species. Preparation and rea ctions. Journal of the American Chemical Society 1975, 97 (8), 2128 - 2132. 444 ( 13 ) Bercaw, J. E.; Marvich, R. H.; Bell, L. G.; Brintzinger, H. H., Titanocene as an intermediate in reactions involving molecular hydrogen and nitrogen. Journal of the American C hemical Society 1972, 94 (4), 1219 - 1238. ( 14 ) Cao, C.; Shi, Y.; Odom, A. L., A Titanium - Catalyzed Three - Component Coupling To - - Iminoamines. Journal of the American Chemical Society 2003, 125 (10), 2880 - 2881. ( 15 ) Vujkovic, N.; F illol, J. L.; Ward, B. D.; Wadepohl, H.; Mountford, P.; Gade, L. H., Insertions into Azatitanacyclobutenes: New Insights into Three - Component Coupling Reactions Involving Imidotitanium Intermediates. Organometallics 2008, 27 (11), 2518 - 2528. ( 16 ) Barnea , E.; Majumder, S.; Staples, R. J.; Odom, A. L., One - Step Route to 2,3 - Diaminopyrroles Using a Titanium - Catalyzed Four - Component Coupling. Organometallics 2009, 28 (13), 3876 - 3881. ( 17 ) Aldrich, K. E.; Odom, A. L., Titanium - Catalyzed Hydroamination and M ulticomponent Coupling with a Simple Silica - Supported Catalyst. Organometallics 2018 . ( 18 ) Severin, R.; Doye, S., The catalytic hydroamination of alkynes. Chemical Society Reviews 2007, 36 (9), 1407 - 1420. ( 19 ) Walsh, P. J.; Baranger, A. M.; Bergman, R. G. , Stoichiometric and catalytic hydroamination of alkynes and allene by zirconium bisamides Cp2Zr(NHR)2. Journal of the American Chemical Society 1992, 114 (5), 1708 - 1719. ( 20 ) Johnson, J. S.; Bergman, R. G., Imidotitanium Complexes as Hydroamination Cataly Substantially Enhanced Reactivity from an Unexpected Cyclopentadienide/Amide Ligand Exchange. Journal of the American Chemical Society 2001, 123 (12), 2923 - 2924. ( 21 ) Gagne, M. R.; Marks, T. J., Organolanthanide - catalyzed hydroamination. Facile, regi ospecific cyclization of unprotected amino olefins. Journal of the American Chemical Society 1989, 111 (11), 4108 - 4109. ( 22 ) Billow, B. S.; McDaniel, T. J.; Odom, A. L., Quantifying ligand effects in high - oxidation - state metal catalysis. Nature Chemistry 2017, 9 , 837. ( 23 ) Ayinla, R. O.; Schafer, L. L., Bis(amidate) titanium precatalyst for the intermolecular hydroamination of allenes. Inorg. Chim. Acta 2006, 359 (9), 3097 - 3102. ( 24 ) Aldrich, O., Low surface density silica - supported Ti precatalyst for C - N bond forming reactions. 2019 . ( 25 ) Majumder, S.; Gipson, K. R.; Odom, A. L., A Multicomponent Coupling Sequence for Direct A ccess to Substituted Quinolines. Organic Letters 2009, 11 (20), 4720 - 4723. 445 ( 26 ) McDaniel, T. J.; Lansdell, T. A.; Dissanayake, A. A.; Azevedo, L. M.; Claes, J.; Odom, A. L.; Tepe, J. J., Substituted quinolines as noncovalent proteasome inhibitors. Bio organic & Medicinal Chemistry 2016, 24 (11), 2441 - 2450. ( 27 ) Dissanayake, A. A.; Staples, R. J.; Odom, A. L., Titanium - Catalyzed, One - Pot Synthesis of 2 - Amino - 3 - cyano - pyridines. Advanced Synthesis & Catalysis 2014, 356 (8), 1811 - 1822. ( 28 ) Burés, J., A S imple Graphical Method to Determine the Order in Catalyst. 2016, 55 (6), 2028 - 2031. ( 29 ) Burés, J., Variable Time Normalization Analysis: General Graphical Elucidation of Reaction Orders from Concentration Profiles. 2016, 55 (52), 16084 - 16087. ( 30 ) DiFranc o, S. A.; Maciulis, N. A.; Staples, R. J.; Batrice, R. J.; Odom, A. L., Evaluation of donor and steric properties of anionic ligands on high valent transition metals. Inorganic chemistry 2012, 51 (2), 1187 - 200. ( 31 ) Bemowski, R. D.; Singh, A. K.; Bajo rek, B. J.; DePorre, Y.; Odom, A. L., Effective donor abilities of E - t - Bu and EPh (E = O, S, Se, Te) to a high valent transition metal. Dalton transactions 2014, 43 (32), 12299 - 12305. ( 32 ) Ugi, I. M., R.; Lipinski, M.; Bodesheim, F.; Rosendahl, F., Cycloh exyl Isocyanide. Org. Synth. 1961, 41 , 13. ( 33 ) Webster, R. L.; Noroozi, N.; Hatzikiriakos, S. G.; Thomson, J. A.; Schafer, L. L., Titanium pyridonates and amidates: novel catalysts for the synthesis of random copolymers. Chemical communications 2013, 49 (1), 57 - 59. ( 34 ) Lui, E. K. J.; Brandt, J. W. ; Schafer, L. L., Regio - and Stereoselective Hydroamination of Alkynes Using an Ammonia Surrogate: Synthesis of N - Silylenamines as Reactive Synthons. Journal of the American Chemical Society 2018, 140 (15), 4973 - 4976. ( 35 ) Möhrle, H.; Schake, D., Heterocyclische Spirocyclohexadienone aus substituierten Phenolen / Heterocyclic Spirocyclohexadienones from Substituted Phenols. In Zeitschrift für Naturforschung B , 1995; Vol. 50, p 1859. ( 36 ) Rieth, R. D.; Mankad, N. P.; Calima no, E.; Sadighi, J. P., Palladium - Catalyzed Cross - Coupling of Pyrrole Anions with Aryl Chlorides, Bromides, and Iodides. Organic Letters 2004, 6 (22), 3981 - 3983. ( 37 ) Mishra, S. J.; Ghosh, S.; Stothert, A. R.; Dickey, C. A.; Blagg, B. S. J., Transformat ion of the Non - Selective Aminocyclohexanol - Based Hsp90 Inhibitor into a Grp94 - Seletive Scaffold. ACS Chemical Biology 2017, 12 (1), 244 - 253. 446 CHAPTER 6. HOMOGENEOUS TITANIUM CATALYZED IMINOAMINATION AND CATALYST DISPROPORTIONATION PROCESSES 6.1 Introduction 7 , 8 Kinetic analysis of the heterogeneous catalyst, [Ti]700(X), provided insight that allowed for targeted enhancement of the observed catalytic rate for iminoamination (see Chapter 5). Parallel studies with the homogeneous titanium catalysts that we typica lly use to perform iminoamination reactions had never been undertaken. In part, this is because we foresaw several complications with performing kinetic analysis with these reactions, as typically in homogeneous iminoamination (3CC), substantial portions o f hydroamination and formamidine side products are noted. 1, 2 Even the four component co upling product, 4CC (aniline + alkyne + isonitrile + isonitrile), 2,3 - diaminopyrrole is often noted in small concentrations in the crude iminoamination reaction mixtures. 3 Some of these side products can be reduced by careful substrate selection, but generally speaking, the homogeneous iminoamination reactions are no t as clean as their heterogeneous counterparts. An illustrative example of this is shown in Fig. 6.1. 7 This project was picked up where a former group member, Dhwani Kansal, left investigations into Ti(IV) complex disproportionation and comproportionation re actions. She prepared and fully characterized several complexes and some of her data will be discussed in this chapter. 8 The work presented in this chapter is being prepared for submission to Faraday Discussions : Aldrich, K. E., Kansal, D., Odom, A. L., 2019 . 447 Figure 6 . 1 Crude GCMS analysis of the iminoamination of 3,5 - dimethylaniline, 1 - phenylpropyne, and CyNC catalyzed by 10 mol% Ti(dpm)(NMe 2 ) 2 ( left ) and 5 mol% [Ti]700(2,6 - dimethylphenylamidate) ( right ). Note that the reaction catalyzed by Ti(dpm)(NMe 2 ) 2 has a substantial peak at 8 min, which is the 3,5 - dimethylaniline starting material, as well as a large p eak at 18 min for hydroamination side product. The reaction catalyzed by [Ti]700 shows no other compounds in the GCMS trace (small peaks on baseline are polysiloxane column material from GC column). At the same time, a lot of work has been done in the Odo m group to demonstrate how detailed understanding and quantitative classification of ligand donor ability and steric profiles can be used to predict how ancillary ligand properties will affect the rate of high valent metal catalysts. 4 - 8 Highly specific studies correlating ancillary li gand properties with the rate of various Ti(IV) hydroamination catalysts, mentioned in previous chapters, for example, clearly indicate that electron - deficient, small ancillary ligands will produce faster catalysts. 4 In theory, using the same techniques we should be able to predict what types of ligands will make Ti - catalyzed iminoamination reactions faster. However, to effectively employ these same techniques, a much more thorough understanding of the iminoamination mechanism is needed to inform about what the slow step of the reaction is and how to speed it up. This would facilitate logical changes in ligand design that target this reaction step specifically. 448 A general mechanism for this reaction has previously been proposed. As discussed below, we think several pieces of this proposed reaction are accurat e. In addition to the reaction steps within the cycle, many of the concepts discussed in Chapter 5 (i.e. the off - cycle Ti species similar to those in the Doye Ti - catalyzed hydroamination mechanism and the [Ti]700(X) catalyzed iminoamination reaction) will be referenced here. 6.2 Kinetic Analysis of Iminoamination Catalyzed by Ti(dpm)(NMe 2 ) 2 With the end goal of using mechanistic insights to guide catalyst design, by correlating ligand properties and effects on reaction rate, we set out to determine the rate l aw that describes the iminoamination reaction. The general reaction studied kinetically is shown above Table 6.1. Due to the messy nature of these reactions, discussed above, we elected to observe reaction progress by GC rather than NMR. The yields listed in Table 1, which shows the conditions of each kinetic trial, are GCFID yields quantified by external calibration of the isolated iminoamination product and standardized against dodecane. This allowed for clean observation (and quantification) of the produ ct (both isomers), the hydroamination side product, and starting materials all in the crude reaction solution. Note that the substrates for the reaction were carefully selected. Initial screenings at the kinetic concentrations examined CyNC as an alternati ve isonitrile, as well as 1 - phenylpropyne and phenylacetylene as alternative alkynes. The smaller isonitrile led to the production of large quantities of formamidine (i.e. >30% overall yield of formamidine) and compromised the quality of the data that coul d be inferred about the iminoamination reaction. Switching to the larger t BuNC substrate reduced this problem substantially. In fact, in most of the kinetics runs shown in Table 1, the formamidine byproduct was noted in trace amounts by GCMS and showed neg ligible quantitation by GCFID. 449 The aromatic alkynes examined were originally considered because with the (dpm)Ti(NMe 2 ) 2 precatalyst, relatively high regioselectivities have been observed in the coupling of aromatic alkynes. 1, 2, 9 - 12 Thus, we would only have to consider a single regioisomer of the desired product. Despite higher selectivity in th e 3CC product yielded with these alkynes, the overall yields and conversions are much poorer. Additionally, substantial hydroamination is observed with these alkynes (up to 35% yield of hydroamination product with 1 - phenylpropyne) along with alkyne trimeri zation. When a terminal alkyl alkyne (1 - octyne) was examined instead, much lower quantities of hydroamination byproduct(s) were observed; alkyne trimerization processes were eliminated as well. While this does result in two regioisomers being produced in r oughly a 1:1 ratio, we were willing to compromise on selectivity to get cleaner reactions that should provide purer kinetic data describing iminoamination. Additionally, since there is essentially no regioselectivity with this alkyne, the assumption that t he rate of formation of either regioisomer is the same, is experimentally supported. (Also, provided that the [2 + 2] cycloaddition 450 Scheme 6 . 1 General iminoamination reaction and substrates examined to determine the effect of each substrate on the rate of the overall reaction. Table 6 . 1 Experimental c onditions examined for kinetic analysis of Ti(dpm)(NMe 2 ) 2 catalyzed iminoamination. The general parameters from the reaction scheme below were applied. Amounts listed in the table are given as cocnetrations (molar). Entry Ti(dpm)(NMe 2 ) 2 (mol%) H 2 NPh t BuNC 1 - octyne Total Conversion to 3CC (%) Regioisomer Ratio (A:B) Hydroamination Byproduct (%) c 1 0.02 (10%) 0.20 0.20 0.20 63 1.0:1 4 2 0.04 (20%) 0.20 0.20 0.20 57 1.2:1 8 3 0.01 (5%) 0.20 0.20 0.20 63 0.9:1 3 4 0.01 (5%) 0.20 0.20 0.40 65 0.9:1 3 5 0.01 (5%) 0.20 0.20 1.00 63 1.0:1 2 6 a 0.01 (5%) 0.20 0.40 0.20 61 0.9:1 3 7 0.01 (5%) 0.40 0.20 0.20 77 0.9:1 4 8 0.01 (5%) 1.00 0.20 0.20 80 0.8:1 4 9 b 0.01 (5%) 0.20 0.20 0.20 0 na 0 a This sample, with additional t BuNC relative to the other three component coupling reagents, shows a measurable amount of the 4CC (H 2 NPh+1 - octyne+2 t BuNC), a double insertion of the isonitrile on intermediate Ti - metallocycles which produces a 2,3 - diaminopyrrole. 3 b This run included 0.02 M 3CC product, added to the reaction mixture with the isonitrile and alkyne prior to heating samples. No product was detectable in the samples by GC analysis. c The hydroamination product generated appears to be during the initial heating period when the reaction comes to temperature. The amo unt of HA observed was generated between reaction initiation and the first sample analyzed, suggesting that after the induction period, the rate of HA is negligible compared to 3CC production. With substrate options evaluated and selected, the reaction con ditions listed in Table 6.1 were examined, sampling over a reaction period of 0 - 24 h. One issue with the catalyst that was immediately recognized is deactivation. Consider the reaction trace shown below for Entry 3. At 24 h, there is still excess of every starting material observed by GCFID, however, the yield for iminoamination product has essentially plateaued and the forward reaction progress has ceased. This deactivation could even be observed visually with many of the kinetics solutions, with a color change in the reaction solutions going from an opaque brown color early in the reaction to a transparent orange color by 24 h. This color change appears to correlate with deactivation of the 451 Ti catalyst (see below). As a consequence of this deactivation, most of the conditions in Table 6.1 show only about 60% total conversion to the iminoamination product. Given this complication, utilization of the graphical analysis method should be approached cautiously; results obtained via graphical analysis were compared to plots of the initial rates and generally gave good agreement for the assessed order. 13, 14 in associated figures are provided in the Experimental section. Figure 6 . 2 The reaction t races of two homogeneous Ti(dpm)(NMe 2 ) 2 catalyzed iminoamination reactions. Both reactions reach a maximum yield an after about 12 h. With the high catalyst loading, there is even what appears to be a decrease in concentration of the product from the maxim um measured concentration. Under the reaction conditions shown in Table 6.1 , Entry 3 was used as the baseline of comparison for determining the order of substrates as they affect the reaction rate. Analysis of the order in alkyne, comparing Entires 3 and 5, showed a zeroth order dependence. Similarly, isonitrile concentration dependence appears straightforward, as the reaction appears to be zeroth order in both reagents by comparison of Entries 3 and 6. However, there are three additional concentration s to consider in relation to their influence on the overall reaction rate: titanium concentration, aniline concentration, and the concentration of the product itself (3CC). The graphical method suggests that there is a fractional order in aniline concentra tion. The best visual agreement is observed when the concentration of aniline is raised to a power of ~ 0.6. 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0 50000 100000 150000 [3CC] (M) time (s) Reaction Trace: Entry 3, 5 mol% 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0 50000 100000 150000 [3cc] (M) time (s) Reaction Trace: Entry 2, 20 mol% 452 Graphically, this raises the question, is this value really ½ - order in aniline, or first order in aniline, n effective way of determining the error for order estimation. Assuming there is a fractional order in aniline, we can envision several mechanistic explanations that might lead to this dependence appearing in the rate law. Assessing the order - dependence b etween the reaction rate and product concentration, however, proved to be impossible to measure experimentally. As indicated by Entry 9, a kinetic run was attempted where the isolated 3CC k was introduced into the reaction mixture with the substrates. Howev er, upon addition of the solution containing t BuNC, 1 - octyne, and the 3CC k product to the solution containing the Ti(dpm)(NMe 2 ) 2 precatalyst and H 2 NPh, the solution rapidly went from the typical opaque brown color of the iminoamination reactions to transpa rent orange. Upon heating the reaction mixture, no catalytic product formation was observed, even after 24 h. Thus, it appears that inclusion of 3CC prior to heating the reaction mixture results in total deactivation of the catalyst. This observation supp orts what we had already suspected with regards to the catalyst instability and deactivation. Some form, or forms, of the catalyst can react with the product generated, and while this seems to take time when the solution is at reaction temperature, at room temperature, this reaction is rapid and irreversible in the presence of the iminoamination reagents. While, in understanding the kinetics data, experimental evidence of this deactivation pathway (catalyst + iminoamination product) was important to obtain, we were unable to experimentally determine the order of this deactivation under the conditions of the reaction. However, it is likely that an additional term (i.e. k obs [Ti][3CC]) should be included as a second term in the most accurate description of the rate law. 453 Finally, there is the relationship between rate and catalyst concentration. Upon initial assessment, it seemed like a first order dependence in catalyst concentration provided the best overlap for Entries 1 - 3. A plot that shows the results of th e graphical overlap produced by this order dependency is shown below in Fig. 6.3. However, this would provide a rate law that is first order in catalyst and fractional order in aniline concentration. While attempts were made to rationalize these results me chanistically, I was unconvinced that any of these explanations fully agreed with the experimental data, and after reading about some examples of reactions in which an off - cycle dimerization affects the order observed in catalyst concentration, I decided t o reexamine the graphical method data for Entries 1 - 3. 454 Figure 6 . 3 Graphical analysis of Entries 1 - 3 in Table 6.1. The plots are fitted with a fractional order ( top ) and first order ( bottom ) dependence. Si milar fits result from both analyses. Previously, it had been observed that a first - order dependence seemed more accurate than a half - order dependence, and the residual deviations between the points in the plot of the first order dependence were attribut ed to experimental error. This error could include simple errors associated with the measurements, as well as the small amount of hydroamination generated during the initial heating period; it also seems reasonable to anticipate error due to complicating f actors that could arise directly from catalyst deactivation. However, if the data is examined with a fractional order 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0 20 40 60 80 concentration 3CC (M) t[cat] 0.7 Fractional Order Catalyst Dependence 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0 10 20 30 40 50 60 concentration 3CC (M) t[cat] 1 First Order Catalyst Dependence 455 dependence, similar or arguably what looks like slightly better agreement is obtained graphically for Entries 1 - 3. Specifically, in the pl ot shown above ( Fig. 6.3 ), all three Entries were fitted with an exponent of ~ 0.7. These results are also presented below by examining the initial rates from the experimental kinetic runs (Entries 1 - 3), as well as the predicted k initial values for both 0.7 and first - order dependence on catalyst concentration. Figure 6 . 4 Examination of the initial rates for Entries 1 - 3 from Table 6.1. The results suggest that the catalyst concentration may not be simple firs t - order in these concentration ranges. Table 6 . 2 Initial rates predicted for entries 1 - 3 with different orders in catalyst concentration. The values calculated for a fractional order appear to agree better with the experimentally observed initial rates shown in Fig. 6.4. Entry k initial (0.7 order) k ini tial (first order) 3 0.00054 0.00054 1 0.00088 0.0011 2 0.0014 0.0022 The graphical analysis has generated a basic rate law of the form shown in Eq. 6.1 , for the range of reaction conditions examined so far. However, several big questions about iminoamination still remain. y = 0.0009x + 0.0043 R² = 0.926 y = 0.0014x + 0.0057 R² = 0.946 y = 0.0005x + 0.0007 R² = 0.9891 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 0 20 40 60 80 Concentration 3CC (M) Time (min) Plot of Estimated Initial Rates for Entries 1 - 3 entry 1 entry 2 entry 3 Linear (entry 1) Linear (entry 2) Linear (entry 3) 456 ( Eq. 6. 1) Additional studies are currently ongoing, but it is worth noting that the resulting fractional orders observed via the graphical order analysis are very similar to what is observ ed in a Heck coupling reaction described in the graphical method studies by Bures. 13, 15 This example is from a study by Blackmond and coworkers, in which the kinetic effects of the catalyst and substrate concentrations for the Pd - catalyzed Heck coupling of p - bromobenzaldehyde and butylacrylate were examined. In their system, an off - cycle dimerization of the active Pd species is observed, where the forward equilibrium constant for dimer formation is very large. This creates two extreme regim es in the catalytic system. One regime dominates when catalyst concentration is high, and the dimerization process is facilitated. Under these conditions, half - order dependence between metal complex concentration, [Pd], and the reaction rate is noted. The second regime dominates catalyst concentration is extremely low. Under these conditions, dimerization an inherently bimolecular process, which is second order in metal is reduced to the extent that the overall reaction rate then appears first - order in [Pd] . Between these two extremes, changes to the catalyst concentration can present orders from 0.5 - 1. In these Ti systems, other studies investigating the behaviors of related complexes in solution implicate off - cycle formation of a dimeric Ti species, brid ged by imide ligands, [Ti(dpm)( - NPh)] 2 . 16, 17 These Ti systems also indicate the potential for a bis(amide) speciation, as an equilibrium species that can form Ti( dpm)(=NPh) and the aniline substrate. A depiction of these potential processes is shown in Fig. 6.5. This process also takes the metal off - cycle. As was observed in the Blackmond studies, other studies of the Hec k reaction, and other homogeneous catalytic systems with these off - cycle dimerization pathways, I wondered if these competitive off - cycle equilibrium processes were also manifesting here as fractional orders in the rate law. 457 Figure 6 . 5 Potential forms of titanium catalyst likely present in the catalytic iminoamination reaction mixture. If these off - cycle processes and their equilibrium constants are causing fractional depe ndences in the rate law, it is possible that under dramatically different reaction conditions, we will observe a different experimental rate law. Specifically, if the dimerization process is exerting a dominant influence over the rate of the catalysis unde r the concentrations examined, [Ti] = 0.01 - 0.04 M (5 - 20 mol%), we would expect that at much lower concentrations, the propensity for dimerization will be reduced. This may result in a return to truly first - order behavior, as has been observed with these ty pes of mechanisms in the literature. 13, 15 Conversely, going to higher catalyst concentrations, we would expect the dependence on [Ti] to approach true half - order dependence. The issue with the kinetic experiments presented in Table 6.1 the n becomes that they were examined over too narrow a range of conditions. For example, in the Pd systems for Heck couplings, >2 orders of magnitude were spanned in catalyst concentrations. 458 Expansion of the catalytic conditions examined via these kinetic s tudies is currently underway. Of the two potentially fractional orders, the more challenging of the two to probe experimentally will likely be the concentration dependence for H 2 higher aniline concentration correlates with hig her conversion, seemingly preserving the catalyst by slowing its decomposition over time. This is consistent with off - cycle formation of the Ti(dpm)(NHPh) 2 equilibrium, provided that this species does not react to directly result in decomposition. Thus, we 2 NPh, because these are favorable in a practical sense to the generation of the iminoamination product. However, running experiments at low concentrations of H 2 NPh is likely to encourage deco mposition, as well as potentially encouraging side - product formation. Finally, from the rate equation that has been experimentally determined, and the suspected role of off - cycle pathways that appear to have a substantial effect on the rate, it is hard t o firmly rule out the possibility that under alternate conditions, the rate will be determined by whatever step within the catalytic cycle, represented by a single arrow above, is slowest. It is common for the rate law of a reaction, where competitive reac tions take some of the active metal off - cycle, to be altered by dramatic reaction conditions changes. 6.3 Investigations into Catalyst Deactivation During Ti - Catalyzed Iminoamination Primary among our unanswered questions at this point was, how does the catalyst deactivate? We have several pieces of evidence suggesting the involvement of the iminoamination product itself. This product, afterall, is a tautomer of the ubiquitous nacnac o r 1,3 - diketimine ligand class, so it is reasonable to assume it will have an affinity for binding to transition metals. This binding may then facilitate other reactivity, which leads to irreversible destruction of catalytically competent Ti species in solu tion. There are several potential forms of the titanium 459 catalyst, shown in Fig. 6.5, from which this decomposition may originate. We tried to investigate these possibilities in turn to get a better idea of what exactly occurs between point A, the active ca talyst, and point B, total deactivation Scheme 6 . 2 The Ti(dpm)(NMe 2 ) 2 complex reacts with the iminoamination product to yield an intractable mixture. Several stoichiometric in situ reactions, between various potential titanium species in solution and the iminoamination product, were examined. First, we considered whether th e precatalyst was stable toward the product. Ti(dpm)(NMe 2 ) 2 and the iminoamination product were combined in a ratio of 1:1.5 in C 6 D 6 . The sample was heated in a J - young tube for 16 h, at which time the solution was examined by 1 H NMR. The heated solution i s a mess, showing an unidentifiable mixture of products with Ti(dpm)(NMe 2 ) 2 not clearly identifiable after heating . precatalyst is not intact after 16 h of heati ng at 80 ° C with 3CC (Scheme 6. 2 ) . This was a start, but due to the order of addition of the reagents when setting up an iminoamination reaction, the titanium species in solution are likely a combination of the [Ti(dpm)( - NPh)] 2 and the Ti(NHPh) 2 (dpm) com plex. There may also be the proposed active species, Ti(dpm)(NPh). Of course, datively bound H 2 NPh and t BuNC could also be present interacting with these species, as well. At any rate, the relevance of titanium species reacting with the iminoamination prod uct is greater when we start closer to the species present in the authentic iminoamination reaction mixture. To that end, the dimeric bridging imido species was synthesized using H 2 Ntolyl and Ti(dpm)(NMe 2 ) 2 . The single crystal X - ray structure of this speci es is shown below in Fig. 6. 6 . Note that the molecule has a center of inversion, such that only half the molecule 460 appears in the asymmetric unit, with the other half of the molecule generated by symmetry in the solid - state structure. Figure 6 . 6 - Ntolyl)] 2 complex. The single crystal X - ray structure is shown with ellipsoids at the 50% probability and H atoms omitted for clarity. Another interesting feature of this species is that within each dpm ligand, one pyrrole ligand binds 1 while the other binds 5 . Rapid interconversion of the two pyrrole rings, a haptotropic shift or exchange, causes the 1 H NMR to appear broad at room temperature, as a result of the fluxionality. Even down to - 75 ° C, some broadness in the aromatic and aliphatic signals (CH 3 signals from the dpm linkers) is observed. However, at this low temperature, the spectrum can be assigned (see Experimental, Fig. 6. 3 3 - 3 5 ). Attempts to characterize the complex in situ using 1 H DOSY NMR provided a calibrated molecular weight of 463 ( ± 51) g/mol. This molecular weight is intermediate to the monomer and the dimer molecular weights, and it is unclear whether this is a result of an exchange between the monomer and dimer in solution or if it is a result of inaccurate calibration due to the broadness of the peaks that we are trying to observe. Additionally, as the peaks remain broad, it is hard to rule out the presence of other species at room temperature (i.e. Ti(dpm)(NHPh) 2 ). However, these 461 omeric species in solution at room temperature. Scheme 6 . 3 - Ntolyl)] 2 . When the isolated dimeric titanium species is combined with an excess of the iminoamination product and heated in C 6 D 6 , no reaction is observed over the course of several hours (Scheme 6. 3 ) . This suggests that one of the other possible Ti species in solution is res ponsible for reacting with the iminoamination product. Given that the precatalyst reacts with the iminoamination product, we suspected that the Ti(NHPh) 2 (dpm) complex was a likely candidate for reaction with the iminoamination product. Additionally, the p roposed active species which in theory is the most reactive form of the titanium in the reaction mixture may also undergo these undesired reactions. Thus, we kept looking, trying to mimic any titanium species that may exist in the catalytic reaction soluti ons. The addition of roughly 3 equivalents of t BuNC to isolated [Ti(dpm)( - Ntolyl)] 2 renders an interesting spectroscopic change. The peaks on the dpm ligand sharpen considerably and are clearly observed at room temperature. This likely indicates coordinat ion of the isonitrile to the Ti center, which is supported by the 1 H NMR, as well. Only 4 aromatic peaks appear, which indicates that each pyrrole ring on the dpm ligands is equivalent (and one is coincident with a doublet peak from the tolyl group); likew ise, the CH 3 groups on the dpm linker are also equivalent. 462 This compound, [Ti(dpm)(CN t Bu)( - Ntolyl)] 2 , reacts rapidly with iminoamination product in solution at room temperature. Upon adding the iminoamination product to an NMR sample of the isonitrile ad duct, the compounds have reacted in the time it takes to walk a sample to the NMR instrument (Scheme 6. 4 ) . Scheme 6 . 4 Proposed decomposition pathway observed when both t BuNC and iminioamination product are added to - Ntolyl)] 2 . - Ntolyl)] 2 when heated with an excess of a different aniline in solution for 1 h at 80 ° C quantitatively exchanges t he H 2 Ntolyl with the new aniline. This was observed with 3,5 - dimethylaniline and 2 - fluoro - 5 - methylaniline. When the dimer was heated with an excess of H 2 Ntolyl for 48 h at 80 ° C, the dimeric complex even appears to undergo exchange with the dpm ligand. The se observations suggest that, even if the dimer is present in relatively high concentrations in the iminoamination mixture, relative to other titanium species, it can still undergo rapid exchange (Scheme 6. 5 ) . Before concluding anything specific, we can s ay that generally, there appears to be a competition for the Ti(dpm) species formed in the iminoamination solution between the product being generated versus deactivation processes. This fundamentally requires a faster or more robust catalyst in order to p rovide full conversion to products. Otherwise, the catalyst will be deactivated well before all the starting materials are consumed. 463 Scheme 6 . 5 Conversion of the - - NAr species upon the addition of an excess of H 2 NAr. With a coordinated H 2 NAr, there are several different pathways conceivable by which the anilides are exchanged by quick proton transfer steps. Based on these initial investigatio ns into the nature of the deactivation reactions, there are several suspected pathways from which these titanium complexes can likely lose their activity and - Ntolyl)] 2 ct directly with the iminoamination product, all of the following species appear to lead to reactivity with the iminoamination product: Ti(NHPh) 2 (dpm), Ti(dpm)(NPh), and [Ti(dpm)(CN t - Ntolyl)] 2 . As is evident from the reaction traces from the kinetic r uns, these deactivation processes have a profound effect on the rate observed in these reactions and the final conversion. Subsequently they cannot be ignored when considering catalyst design in these systems. A summary of the suspected pathways of decompo sition that we have observed through these stoichiometric NMR - scale experiments is shown in Fig. 6. 7 . 464 Figure 6 . 7 Various titanium species found to (or suspected of) react dir ectly with the iminoamination product. This reactivity provides plausible means of titanium complex deactivation throughout the iminoamination reaction. 6.4 Modification of the Bis - Chelating Ancillary Ligand With the information gleaned from kinetic studies ( Sections 6.2 and 6.3), we suspected that a modification to the ancillary ligand could be beneficial for several reasons. One of these reasons relates to catalyst stability. As discussed in the previous sections, the Ti(dpm)(NMe 2 ) 2 catalyst reacts with the iminioamination product, and potentially other species in the reaction mixture, leading to deactivation of the titanium catalyst. We thought that a more donating ancillary ligand with a lower pK a might be less susceptible to substitution o r protonation by other species in the reaction mixture. 465 We also reasoned that, regardless of the slow step of the reaction sequence, shown above in Fig. 6.5, Ti species other than the active catalyst exist in solution. Namely, there is the dimeric Ti - spe cies and the Ti(Bis(amide)), which are both off - cycle species that reduce the amount of catalytically active metal in solution. Moving to an ancillary ligand that is more electron - rich should serve to push equilibria with the off - cycle species (where the L ewis acidic Ti, picks up other ligands in solution to increase its electron density) toward the catalytically active terminal imido complex, resulting in faster rates of iminoamination. To that end, we considered some of the other precatalysts that had b een examined by the group previously for other applications. One bischelating ancillary ligand that seemed promising, which performed poorly as a hydroamination catalyst due to its strong donor ability and steric bulk, was 2,2' - methylenebis(6 - tert - butyl - 4 - methylphenoxide) (OArCH 2 ArO). 4 le iminoamination experiment (1 mmol H 2 NPh/1.5 mmol t BuNC/ 1.5 mmol 1 - octyne, 2 mL toluene, 5 mol% catalyst (0.025 M)) was performed using Ti(OArCH 2 ArO)(NMe 2 ) 2 . GC analysis of the crude reaction mixture revealed 72% yield of the 3CC products, 10% yield of hydroamination, with the remainder going to the 2,3 - diaminopyrrole (4CC). While this result shows a substantial amount of byproduct, it more importantly shows consumption of the limiting reagent. Thus, despite the generation of some side products, this cat alyst demonstrates higher relative activity, as no leftover H 2 NPh is observed in the reaction solution. Running the reaction again under the significantly more dilute conditions for Entry 3 in Table 6.1 (0.2 M in each reagent, 0.01M Ti(dpm)(NMe 2 ) 2 ), and m onitoring reaction progress by GC, a very similar reaction trace was noted for the Ti(OArCH 2 ArO)(NMe 2 ) 2 catalyst relative to Ti(dpm)(NMe 2 ) 2 , with the yield maxing out at about 60%. Overall, the rates are very similar, too (see experimental). This suggests that the Ti - catalyst is still suffering from deactivation, regardless 466 of what we predicted would be a more stable ancillary ligand in the presence of the iminoamination product among other acidic protic species. One additional benefit of the ( - OArCH 2 ArO - ) ligand, over dpm, is that it has readily identified NMR signals in an uncrowded region of the NMR spectrum. The bisphenoxide ligand, when bound to 4 - coordinate Ti, results in distinct 1 H NMR shifts for the two hydrogens on the - CH 2 - linker . The chelate f orms an 8 - membered ring with the metal, such that two unique environments are created for the two protons (diastereotopic). This is shown in the figure below ( Fig. 6. 8 ). 18, 19 Figure 6 . 8 ( left ) 1 H NMR spectrum and ( right ) single crystal X - ray structure 4 of Ti(NMe 2 ) 2 (OA rCH 2 ArO). The two protons on the methylene linker have unique positions due to the conformation of the 8 - membered ring formed by the ligand with Ti, which is readily observed by the distinct doublets in the 1 H NMR spectrum. This trait applies to all of the complexes of the general form Ti(X) 2 (OArCH 2 ArO) and makes them easy to observe and distinguish between by 1 H NMR. Taking advantage of these inherent ligand properties, an NMR scale iminoamination reaction using 20 mol% Ti(OArCH 2 ArO)(NMe 2 ) 2 was examined in tol - d 8, to try to directly determine the fate of the catalyst in solution. Unlike in the Ti(dpm)(NMe 2 ) 2 , where free H 2 dpm ligand is observed in solution after heating the catalysts with 3CC and isonitrile, with Ti(OArCH 2 ArO)(NMe 2 ) 2 no dissociated or free chelating ligand is observed in the reaction mixture. Rather, after 16 h of heating, 45% yield of the iminoamination product was observed, 467 along with 10 mol% of Ti(OArCH 2 ArO) 2 (i.e. ½ of the original Ti added to the solution). The other 10 mol% of the cata lyst could not be directly identified but presumably has picked up a variety of N - bound ligand species in solution from the reactants and products of the catalysis (See Experimental). This observation demonstrates two consequences of switching from the (d pm) to the (OArCH 2 ArO) ligand: 1) while the ligand is stable with respect to maintaining coordination to Ti (as opposed to dissociation to the free ligand) the properties of the ligand have also opened up disproportionation pathways, and 2) with full catal yst deactivation observed on the order of 16 h of reaction time, the enhanced reactivity of this catalyst in the normal scale reaction is even more impressive in terms of the rate of the reaction when the catalyst is still active. The latter observation re ally suggests that further ligand modification could result in a much faster catalyst, provided the stability issues with the catalyst can be overcome. Scheme 6 . 6 Iminoamination reaction catalyzed by Ti(OArCH 2 ArO)(NMe 2 ) 2 following standard reaction conditions. These experiments support our kinetically - derived hypothesis that more electron - rich bischelating ancillary ligands can potent ially improve the performance of homogeneous Ti(IV) iminoamination catalysts. However, the marked difference in the means of catalyst deactivation upon transitioning from the dpm to (OArCH 2 ArO) bischelating ancillary ligand raised several additional questi ons with regards to ligand design. Is disproportionation a decomposition pathway 468 that is related to the electronics of the ligand? If this phenomenon is general, can we predict what ligands will be susceptible to these decomposition pathways? Are there oth er ligand design strategies that can prevent these ligand - exchange reactions from compromising the catalysts while maintaining catalyst stability toward the reaction products? In order to effectively search for promising ancillary ligand candidates to impr ove the iminoamination reaction, we need answers to some of these questions. 6.5 Predicting What Ligands Lead to Stable Catalysts Versus Disproportionation Considering the precedence in the literature and our own experiences with Ti(IV) species, the results d escribed in the previous section were not surprising. Ti(IV) complexes have been noted, with a variety of different X ¯ ligands, to undergo ligand dis - and comproportionation reactions. For example, mixing Ti(NMe 2 ) 4 with TiCl 4 in equal proportions results i n quantitative generation of the heteroleptic Ti(NMe 2 ) 2 Cl 2 species in minutes at room temperature. 20 Likewise, these comproportionation reactions can even facilitate formation of the heteroleptic complexes with chelating bis(amide) ligands and TiCl 4 ; 21 the same observation can be made with Ti(OR) 4 and TiX 4 (R = Cy or i Pr, X = Cl or Br), and can even be utilized as a synthetic methodology to yield heteroletpic species of the general form Ti(A) 4 - n (X) n . 22 - 24 Another example which highlights the same phenomenon if the mixed ligand, heteroleptic compound being preferred, is Ti(dpm)(NMe 2 ) 2 . We note that no ligand disproportionation occurs to produce the homoleptic species, even when the compound is heated in C 6 D 6 for several weeks at 85 °C ( Scheme 6.7 summarizes these ligand exchange processes). Ultimately, understanding these ligand processes is critical to improving our ability to select and design better ancillary ligands for both improvements in catalyst rate and stability. While these processes have been observed and acknowledged for decades, and early transition metal 469 chemists are all too familiar with these complications in synthesizing complexes, no systematic study to show correlation or causation has been undertaken. 25, 26 Scheme 6 . 7 General equilibria observed for severa l Ti(IV) complexes that have been noted in the literature and through our observations. Table 6 . 3 T he ligand combinations listed in the table above, the heteroleptic Ti(X) 2 A 2 complexes are noted quantitativel y. This necessitates that K eq is very large. Ligand Sets where K eq >> 1,000 LDP (kcal/mol) A ¯ X ¯ NMe 2 9.34 NMe 2 Cl O i Pr 10.56 O i Pr Cl Cl 14.97 OCy Cl Br 15.45 O i Pr Br Pyr (dpm) 13.64 NMe 2 dpm OCy - With these motivations in mind, we set out to determine what ligand properties lead to ligand exchanges in the Ti(IV) complexes of interest. Looking at the complexes mentioned above, we noticed a common theme among the ligand combinations that demonstrate rapid comproportionation to produce exclusivel y the heteroleptic Ti(X) 2 (A) 2 complexes. With Ti(NMe 2 ) 4 and TiCl 4 , for instance, the NMe 2 and Cl ligands have very different donor abilities with LDP values of 9.34 and 14.97 kcal/mol respectively. 5, 27 With these highly different ligands bound to each Ti(IV) species, the ligands rapidly exchange to form the Ti(NMe 2 ) 2 Cl 2 complex, where each Ti has two electron rich NMe 2 ligands and two electron - poor Cl ligands. This results in Ti centers with the same electron density than existed in the two individual starting materials. Again, approaching this problem from the other direction, we can consider Ti(dpm)(NMe 2 ) 2 . Each half of the dpm chelate donates similar electron density to a pyrrole ligand, which has an LDP value of 13.64 kcal/mol (bound 1 to Cr), while again, the NMe 2 ligands have an LDP value of 9.34 kcal/mol. This is still a very large difference in donor abilities between the 470 X ¯ and A ¯ ligands in the Ti complex, and so the species does not demonstrate any reversion to the Ti(dpm) 2 or Ti(NMe 2 ) 4 parent homoleptic complexes, which would relegate the two Ti centers to experience highly discrepant electronic environments. Thus, in reactions utilizing Ti(dpm)(NMe 2 ) 2 as precatalyst, no disproportionation to make Ti(N) 4 and Ti(dpm) 2 has been observed (N = any nitrogen - bound ligand in reaction mixture). A tendency for metal complexes to undergo these ligand exchange process to produce more equivalent electronic environments at any given metal atom in an entire system may have a very simple thermodynamic explanation. Ionic bonds are stronger than covalent bonds. In a system like TiCl 4 and Ti(NMe 2 ) 4 the Ti should form more ionic bonds wi th the more electronegative Cl ligands. However, as the effective oxidation state or charge of Ti is increased, the Ti becomes more electronegative, and the ionicity of the bonds is reduced. By contrast, the Ti in Ti(NMe 2 ) 4 , where the ligands are relativel y donating, should exhibit a lower effective oxidation state. By mixing the two ligand types and producing Ti(NMe 2 ) 2 Cl 2 the overall bond enthalpies could then be maximized, as the Ti Cl bonds in the heteroleptic species maintain a higher degree of ionicity . This is one simple explanation for why the heteroleptic species may be favored over the homoleptic ones when the ligand donor abilities are so discrepant. With other ligands, we could invoke overlap discrepancies, Lewis acidity, etc. If we consider th e available donors to occupy the two protolytically exchangeable coordination sites to Ti in the iminoamination reaction mixture with the Ti(OArCH 2 ArO)(NMe 2 ) 2 precatalyst, we can imagine that the donor abilities of the N - donors in solution are comparable t o the bis(phenoxide). The NMe 2 sites will exchange with NHPh (anilide) ligands, bridging imides ( - NPh), or even the iminoamination product (with donor properties presumably somewhere between an amine and an amide). The donor ability of an electron - rich ph enol is much closer to 471 that of an anilide or nacnac - type interaction than the comparisons above (i.e. dpm or Cl ¯ ), where LDP was on the order of 4 or more kcal/mol. Specifically, a typical phenol LDP value is 11.8 - 12 kcal/mol, while anilides are typically around 10 kcal/mol. Something like the iminoamination product, if considered independent of sterics, could demonstrate a similar LDP value to an anilide ligand (likely higher due to resonance delocalization). In fact, it seems reasonable that the electronic donor ability would be similar to that of phenoxide. With these similar donor abilities, disproportionation is observed, rapidly generating the homoleptic complexes from initial heteroleptic Ti(IV) species. Thus, when the donor abilities of X ¯ and A ¯ ligands available for Ti are similar, it seems like there is no clear driving force for formation of the heteroleptic complex exclusively, and rather a mixture of the homoleptic complexes, the heteroleptic complex, or species somewhere in between, c an easily be formed. These observations are consistent with work performed in the group by Dhwani Kansal. In the Ti(OArCH 2 ArO) 2 di(bischelate) complex in s olution. This scenario is outlined in Fig. 6. 9 , below. As mentioned above, we suspected that the equilibrium constants for these ligand redistribution reactions are related to the difference in the donor abilities of the two different ligands. By experimen tally determining the K eq values for different X ¯ ligands with the (OArCH 2 ArO) bischelating ligand and comparing these values to the differences in the LDP values between the chelate and the variable ligand, X ¯ , we hoped that a correlation could be establi shed. Establishing this sort of relationship, between donor abilities of the ligands and the tendency of a compound to undergo ligand redistribution reactions, would facilitate more educated ancillary ligand selection for Ti(IV) catalysts. 472 The ligand sets used for these initial studies were carefully matched, as use of a bischelate in combination with a given TiX 4 species means that only 3 different Ti complexes are possible in solution, provided bridging interactions are not observed with the chelate. The se complexes are TiX 4 , Ti(OArCH 2 ArO) 2 , and Ti(OArCH 2 ArO)X 2 . Also a careful consideration in these systems is the steric protection applied to the aryloxide ligands; when aryloxide or alkoxide ligands are put on Ti, bridging interactions often occur which l ead to multinuclear Ti species (i.e. dimers and oligomers). The inclusion of a tert - butyl group in the 2 - position of the aryloxides ensures that the complexes are monomeric. Keeping the substitution of the aryloxide ligand constant in the 2 and 6 positions of the ring also makes these ligands isosteric. Thus, any changes observed in the equilibrium behavior of the aryloxide ligands can be assigned purely to electronic differences. Additionally, since the determinations of K eq were carried out using 1 H NMR s pectroscopy to probe the concentrations of each species in solution, the benefits of the (OArCH 2 ArO) species, described above, were highly useful in quantification of each species. Dhwani had examined the equilibrium constants for comproportionation (Fi g. 6.1 0 ) with the tetra(2 - tert - butyl - 4 - R - phenoxide)Ti(IV) complexes where R = CF 3 , H, Me, and Br. Her preliminary results showed that, with these similar A ¯ (OArCH 2 ArO) and X ¯ (2 - tert - butyl - 4 - R - phenoxide) ligands (in terms of their donor abilities) all 3 s pecies (Fig. 6. 9 ) coexist in solution, in relatively similar concentrations. This indicates that when the A ¯ and X ¯ ligands have similar donor abilities, there is no significant driving force to favor the heteroleptic species, and so a mixture of products is observed. Her results also suggested that as the differences in the donor abilities of the aryloxide and the chelating (OArCH 2 ArO) were enhanced (i.e. 2 - tert - butyl - 4 - CF 3 - phenol = aryloxide), the equilibrium shifted toward the heteroleptic complex. Unfor tunately, Dhwani was not able to finish these studies prior to leaving the group. As the applications of these studies had 473 become highly relevant to the targeted iminoamination catalyst systems, we decided to expand and finish the studies that Dhwani had b egun examining. Figure 6 . 9 The comproportionation reaction monitored by determination of the equilibrium constant K eq . The 3 possible Ti species in equilibrium in these solutions are shown, including both starting materials and the only possible product. We reexamined the values, for R = H, Br, Me, and CF 3 finding similar K eq values to t Bu, an d OMe. Additional ligands with dramatically different electronic and steric properties were also pursued, specifically, with X = NMe 2 , O i Pr, I, and Cl. With these values for K eq established, we approached building a model. Several possible relationships b etween the donor ability (LDP), the size (%V bur ), and the observed K eq value were considered. Strong graphical evidence suggested that a dependence on the electronic difference between the bischelate and the X ¯ ligand was second order, given the general pa rabolic appearance of the plot shown in Fig. 6.1 0 . The values used to perform the data modeling are listed in Table 6. 4 . 474 The correlation between K eq 2 bur bur )) 2 was evaluated by fitting equations of the general form shown below, in Eq. 6.2, with a least squares approach. Goodness of fit was evaluated by examining plots of the model - predicted vs. the experimental K eq values. Perfect agreement between these two sets of values would be represented by R 2 = 1 and a linear equati on of y = x, so the closer the plot of model - predicted vs. experimental K eq gets to these qualifications, the better the model. Figure 6 . 10 Plot of K eq vs. LDP of the X ¯ ligands in the Ti(OArCH 2 ArO)(X) 2 . ( Eq. 6. 2) To establish the impact of each term in Eq. 6.2, several iterations of the least squares fit, were performed. In turns, coefficients c , d , and e were each set to zero, and the least squares fit reapplied (see Experimental for more details). Primarily, these modeling exercises highlight several important aspects of the relationship between K eq and the steric and electronic properties of the X ¯ ligands. 0 500 1000 1500 2000 8 10 12 14 16 18 K eq LDP (kcal/mol) K eq vs. LDP 475 First, the K eq value is essentially independent of sterics, with the fitted coefficients for bur bur ) 2 weighted very small in magnitude relative to a , b , and c . Note, when every variable in Eq. 6.2 longer make chemical sense and several negative K eq values are calcualted. Thus, the coefficients for d and e were fitted independently in turns. Essentially the same quali ty fit was obtained when bur bur ) 2 were used as the steric parameter; both provided an approximately 0.002 improvement in R 2 , which is over 0.99 with just electronic effects modeled . Collectively these observations suggest that sterics d o not have a measurable effect on K eq , and the minute improvement in R 2 observed is simply due to adding another parameter to the least squares fit. In a system with a larger, more rigid bischelating ancillary ligand, or with a dramatic increase in the siz es of the X ¯ ligands under study, steric influence could reasonably affect the K eq observed. With the ligand selection under study, however, only about 85% of the first coordination sphere is occupied by ligands in any given Ti(IV) complex and the size ran ge spans only a 5% range in %V bur . Table 6 . 4 Values used to model the relationship of sterics and electronics to the equilibrium constant for the interconversion of the homoleptic and heteroleptic. Ti(OArCH2Ar O)X 2 LDP LDP 2 %V bur %V bur ( %V bur ) 2 K eq exp. Error K eq modeled X NMe 2 9.34 - 2.52 6.35 21.9 0.7 0.49 1120 118 1133 OiPr 10.33 - 1.53 2.34 17.4 - 3.8 14.44 495 80 491 I 15.8 3.94 15.52 19.2 - 2 4 1830 190 1742 Cl 14.97 3.11 9.67 16.8 - 4.4 19.36 898 179 1029 2 - tert - butyl - 4 - R - phenoxide t Bu 12.01 0.15 0.022 21.4 0.2 0.04 14 4 10 H 11.98 0.12 0.014 21.7 0.5 0.25 21 5 12 Me 11.82 - 0.04 0.0016 21.2 0 0 8 4 26 Br 12.18 0.32 0.102 21.3 0.1 0.01 38 6 4 F 11.99 0.13 0.016 21.9 0.7 0.49 14 4 11 OMe 11.71 - 0.15 0.022 21.6 0.4 0.16 14 3 40 CF 3 12.55 0.69 0.4761 21.2 0 0 71 11 19 476 In fact, when both d and e are set equal to zero we obtained the following fit, dependent exclusively on electronic properties of the X ligand: ( Eq. 6. 3) This fit is plotted with the experimental data in Fig. 6.1 2 , as well as the model - determin ed values for K eq . If we scale the variables fitted in this equation, we can determine the relative magnitudes of the two coefficients despite the dramatically different ranges for the parameters, LDP and LDP 2 . This indicates the importance of the two te rms in determining the K eq . The scaled coefficients are presented in Eq. 6.4. Figure 6 . 11 A plot showing model - predicted versus experimental data relating the donor ability of a given X¯ ligand to the equilibrium constant observed for formation of the heteroleptic species, Ti(X) 2 (OArCH 2 ArO) (Fig. 6.15). ( Eq. 6. 4) From the scaled coefficients, we can see that the coefficient of the squared term is much larger in magn second - order polynomial equation, with the squared term dominating the K eq y = 135.8x 2 - 98.571x + 21.959 R² = 0.9916 -50 450 950 1450 1950 2450 -4 -2 0 2 4 6 Keq LDP (kcal/mol) Equilibrium Constant vs. LDP Model Predicted Experimental Poly. (Experimental) 477 and the linear term dominating K eq e the aryloxide the second order term exceeds the magnitude of the linear term and will begin to dominate the electronic effects on K eq . We think that, in theo ry, this general relationship applies to the equilibrium process of ligand disproportionation with any metal, but that the coefficients for these electronic terms are ionation and comproportionation processes the coefficients modifying the effects of LDP and ( LDP) 2 would then be very large (approaching - and comproportionation properties to Ti(IV), i.e. high va lent U, Zr, or Hf, the general form of Eq. 2 with similar coefficients to those found with Ti(IV) may effectively describe the ligand exchange processes. In this way, we think the correlations discovered here with Ti are likely observed with other metals, as well, and demonstrate the fundamental properties that control these ligand exchange processes. This realization is interesting, and further studies examining these relationships in other metal systems are a continuing interest in the group. For the im inoamination reaction under study and the applications of this study to catalyst development the relationships shown by the model suggest that there is a minimum difference in donor ability for the ancillary ligand and the species that occupy the protolyti cally exchangeable sites on Ti(IV) during the iminoamination reaction which will maintain catalyst stability in regards to ligand processes. Relative to the (OArCH 2 ArO) ligand, we should select a ligand that is less electron rich. Considering the model, so mething that is > 0.3 kcal/mol higher or lower than the LDP value of the chelated iminoamination product would perhaps offer the greater donor ability sought to increase the proportion of the active Ti - imide species while preventing the ligand 478 exchange dea ctivation pathway. Alternatively, something substantially more donating may also improve results. For example, a chelating bis(amide) ligand with a high degree of conjugation to reduce basicity, that is sterically protected, may provide the same benefits b ut with an even faster rate of iminoamination (again via enhanced donor ability facilitating generation of the active species or faster product protonation). These results provide a direction for future ligand - screening progress. Some candidate ligands of interest are shown in Fig. 6.12 , below. Figure 6 . 12 Potential ligands of interest that could avoid deactivation via ligand disproportionation as their donor abilities (LDP values) are predicted to be more and less donating than a chelated iminoamination product, potentially disfavoring ligand exchange reactions. 6.6 Conclusions Obviously, a simple clean rate law with simple first - order dependences are easier to understand and implement than those with complex reaction orders, like the one we have uncovered that describes the process of homogeneous titanium - catalyzed iminoamination. Despite the fractional order in aniline, potentially fractional order in catalyst, the decomposition caused by product, and the conversion - limiting deactivation processes observed throughout kinetic trials, the experimental results from these studies have been highly informative. Because of the complicated rate law, we saw improving cat alyst stability and minimizing the potential contribution of catalyst resting states, which persist due to the electron deficiency of the titanium metal center, as the most viable options for improving catalyst performance. Promising results have been obse rved by pursuing a very easy ancillary ligand switch, indicating the merits of this approach in improving 479 the practical application of these catalysts to iminoamination as a first step in the production of several functionalized heterocycles. The observa tions made with the new catalyst, Ti(OArCH 2 ArO)(NMe 2 ) 2 , forced us to consider the aspects of ligand properties and design that might lead to the most stable catalysts for this reaction in solution. Systematically approaching the characterization of catalys t stability via ligand exchange reactions correlated to donor ability has allowed for the development of a model system. This model can now guide subsequent adjustments to the ancillary ligands selected to perform the iminoamination reaction. Afterall, in these systems, the primary motivator for achieving a faster rate is to achieve a higher yield. However, if we make a more stable catalyst, higher yields can be achieved regardless of the rate relative to our existing catalyst systems. While shorter reactio n times are one of the ultimate goals for targeted catalyst design for iminoamination, a 1 - pot - 2 - step reaction that take 2 days to yield a complex heterocycle is still faster than traditional organic methodology that might take 10 steps to generate the sam e heterocycle. 6.7 Experimental General Considerations Synthesis Considerations All syntheses and handling of materials were carried out under an inert N 2 atmosphere, either in an MBraun glovebox or by standard Schlenck technique unless otherwise specified. G enerally, this includes reaction set - up for catalytic reactions and the preparation of the Ti species, as well as preparations of NMR samples that contain Ti complexes. Column chromatography, GC sample preparation, and the characterization of organic produ cts were preformed in air, on the benchtop with solvents handled and stored in air. 480 Solvents including toluene, diethyl ether, and pentanes, were purchased commercially. These solvents were dried and deoxygenated by sparging with N 2 and passage over an ac tivated alumina column before use. The NMR solvent CDCl 3 was purchased from Sigma - Aldrich and used as received (for routine organic compound characterization) or dried over P 2 O 5 and distilled under N 2 prior to use (for titanium complexes). The NMR solvent C 6 D 6 was purchased from Sigma - Aldrich and dried over CaH 2 ; it was then distilled under N 2 prior to use. All alkynes were purchased commercially (Alfa Aesar) and dried over Na 2 SO 4 and distilled under N 2 prior to use. Aniline (and any derivatives) was dried over CaH 2 and distilled under vacuum prior to use. Tert - butyl isonitrile ( t BuNC) was prepared according to literature procedures. 28 Dodecane was sparged with N 2 prior to use. Materials used for colum n chromatography, including hexanes, diethyl ether, and triethylamine (TEA), were purchased commercially and used as received. The Ti(dpm)(NMe 2 ) 2 complex was prepared as previously reported and matched literature 1 H and 13 C NMR. 2 Instrumentation NMR Routine characterization spectra were obtained using an Agilent DDR2 500 MHz NMR spectrometer equipped with a 5 mm PFG OneProbe operating at 499.84 MHz ( 1 H) and 125.73 MHz ( 13 C). 1 H NMR chemical shifts were referenced to residual CHCl 3 in CDCl 3 as 7.26 ppm, or residual C 6 HD 5 in C 6 D 6 as 7.16 ppm. 13 C NMR chemica l shifts are reported relative to 13 CDCl 3 as 77.16 ppm, or ( 13 C)C 5 D 6 as 128.06 ppm. The Varian Dbppste_cc (DOSY bipolar pulse pair simulated spin echo convection corrected) pulse sequence was utilized for all experiments where DOSY NMR was used. All spect ra were multiplied by a weighted exponential of 10 Hz and baseline corrected before applying DOSY processing. Standard DOSY processing, as supplied by the vendor, was used based on peak 481 heights and with compensation for non - uniform gradients. For notes on NMR - based determination of K eq , see below. GC GCMS data was collected on an Agilent 5973 MSD with a 6890N series GC. GCFID data was collected on a Hewlett Packard 6890 series GC system and standardized against dodecane as an internal standard. 3CC product s were quantified in situ by utilizing GCFID standardized calibration curves generated by quantification of the authentic iminoamination product, isolated from a catalytic reaction mixture. Full characterization data is given below. The hydroamination side product was quantified in a similar manner from previously isolated amine derivatives of the genuine imine product (see Chpt 4). The 4CC product was quantified analogously to the 3CC product. X - ray All single crystal X - ray structures were collected at the MSU Center for Crystallographic Research. The data was collected on Bruker diffractometers running Cu - K radiation. The collection data and information about the unit cell, etc. for these structures is provided below. Synthesis of Iminoa mination Product from aniline, 1 - octyne, and t BuNC: A 15 mL pressure tube was charged with 62 mg of Ti(dpm)(NMe 2 ) 2 , 1 mL of toluene, and a stir bar. To the stirred solution was added a 1 mL solution containing 186 mg of aniline (2 mmol, 1 equiv) in toluen e. This mixture was stirred for 10 min at room temperature over which time the solution went from a transparent bright orange color to an opaque reddish - brown. Then a 1 mL solution containing 184 mg of t BuNC (2 mmol, 1 equiv), and 220 mg of 1 - octyne (2 mmo l, 1 equiv) in toluene, was added to the solution in the pressure tube. The tube was sealed and transferred from the glovebox to a 110 °C oil bath. The tube was heated and stirred for 24 h. The 482 tube was removed from the bath and allowed to cool ambiently. The volatiles were removed by rotary evaporation, and the resulting crude, dark brown oil was separated by column chromatography (Al 2 O 3 , Hexanes(1%TEA), gradient Et 2 O from 0 to 25%). The isolated product was obtained as an orange oil (310 mg, 54%), which p roved to be a mixture of regioisomers A and B, shown above. Standard column conditions could not be found that effectively separate the two regioisomers. 1 H NMR (500 MHz, chloroform - d ) ( A ) 9.97 (s, 1H 4.72 (d, J = 8.0 Hz, 1H), 1.27 (s, 9H), 0.90 (m, 3H); ( B ) 10.83 (s, 1H), 7.77 (d, J = 2.8 Hz, 1H), 7.11 (d, J = 2.8 Hz, 1H), 1.32 (s, 9H), 0.84 (t, 3H); ( A/B ) 7.33 7.26 (m, 4H), 7.04 (d, J = 7.7 Hz, 2H), 7.03 6.97 (m, 1H), 6.85 6.78 (m, 3H), 2.37 (s, 1H), 2.22 2.09 (m, 3H), 1.48 (m, 5H), 1.18 (m, 8H ). 13 C NMR (126 MHz, chloroform - d ) 171.01, 153.80, 151.55, 150.11, 146.87, 142.28, 129.06, 128.43, 122.66, 121.83, 121.38, 119.00, 103.60, 91.79, 52.60, 51.03, 33.68, 33.21, 32.91, 31.93, 31.80, 31.46, 30.36 (d, J = 1.9 Hz), 30.31, 29.47, 29.32, 28.74, 28 .67, 27.73, 23.69, 22.72, 22.47, 14.15, 14.05. HRMS: 19 H 31 N 2 : 287.2487; found: 287.2484. EA 19 H 30 N 2 : C, 79.66; H, 10.56; N, 9.78. Found: C, 79.88; H, 10.44; N, 9.44. Synthesis of 2,3 - diaminopyrrole from aniline, 1 - octyne, and 2 t BuNC: On several occasions, the product mass that corresponds to the coupling of 1 equiv aniline, 1 equiv alkyne, and 2 equiv of isonitrile was observed by GC/MS in reactions catalyzed by homog eneous Ti - catalysts. Typically, the amount of this product was relatively small. However, under certain conditions when reactions were carried out on large enough scales, substantial masses of the 4CC product were noted in various column fractions when iso lating the 3CC products by column chromatography. 483 On one such occasion, a very clean fraction of the 4CC product was isolated from a 2 mmol scale reaction (H 2 NPh, t BuNC, 1 - octyne) with 5 mol% Ti(dpm)(NMe 2 ) 2 as precatalyst. The 4CC product was the first co mpound eluted from an alumina column, basified with 2% TEA in Hexane. Note from the same column, the 3CC product was also isolated, but as a later fraction with the addition of Et 2 O on a gradient from 0 - 30%. The 4CC product was characterized by GCMS, HRMS, 1 H NMR, 13 C NMR, and a few additional 2D NMR techniques. The following structural assignment seems to most closely match the characterization data for this product, in agreement with previous studies by our group. 3 Scheme 6 . 8 Productio n of 4CC product from iminoamination reaction mixture. 1 H NMR (500 MHz, benzene - d 6 ) 7.43 (d, J = 7.1 Hz, 2H), 7.13 (t, J = 7.8 Hz, 2H), 6.97 (t, J = 7.5 Hz, 1H), 6.53 (t, J = 0.9 Hz, 1H), 2.99 (s, 1H), 2.68 2.54 (m, 2H), 2.34 (s, 1H), 1.75 (pentet, J = 7.7 Hz, 2H), 1.52 1.42 (m, 3H), 1.40 1.29 (m, 06H), 1.23 (s, 12H), 0.92 (s, 15H). 13 C NMR (126 MHz, benzene - d 6 ) 142.18, 128.80, 125.41, 125.16, 123.74, 122.06, 121.84, 1 13.63, 55.46, 54.53, 32.33, 30.66, 30.60, 30.54, 30.07, 26.37, 23.16, 14.41. HRMS: QTOF EI (positive 24 H 40 N 3 + : 370.3222; found: 370.3218. EA 24 H 39 N 3 : C, 77.99; H, 10.69; N, 11.37. Found: C, 78.09; H, 10.85; N, 11.11. Synthesis of [Ti( - Ntolyl)(dpm)] 2 : A solution of 50 mg of Ti(dpm)(NMe 2 ) 2 (1 equiv) in 1.5 mL of C 6 D 6 was prepared and stirred at room temperature. To this solution was added a solution of 34 mg (2 equiv) H 2 Ntolyl in 0.5 mL of C 6 D 6 . The resulting solution immediatel y began to 484 darken from light yellow to dark brown. The solution was stirred for 10 minutes at room temperature and was then examined by 1 H NMR. The spectrum shows that one equivalent of H 2 Ntolyl has reacted with the Ti(dpm)(NMe 2 ) 2 while one equivalent rem ains free in solution. The peak shifts and integral values for the new species approximately matches the formula which contains [Ti(dpm)(Ntolyl)] in those ratios. There is fluxtionality with the Ti complex at room temperature, evidenced by the broadness of the peaks for the new species in the 1 H NMR. X - ray quality crystals were grown from a concentrated toluene solution layered with n - hexane and stored at - 35 °C for 24 h. The compound can also be purified by precipitation from a concentrated Et 2 O/n - hexane solution stored at - 35 °C for 2 d (yield: 30 mg, 58%). The purified compound still presents broad signals by 1 H NMR at room temperature due to rapid haptotropic shifting of the 1 / 5 pyrrole rings of the dpm ligands. The peaks begin to resolve a round - 75 ° C. 1 H NMR (500 MHz, tol - d 8 , - 75 ° C) 7.63 (s, 2H), 6.76 (d, J = 8.5 Hz, 2H), 6.53 (s, 1H), 6.44 (s, 2H), 6.36 (m, 6H), 6.06 (d, J = 6.2 Hz, 2H), 5.82 (s, 2H), 5.65 (s, 1H), 1.97 (d, J = 17.6 Hz, 0H), 1.93 (s, 0H), 1.58 (d, J = 17.4 Hz, 0H). 13 C NMR (126 MHz, tol - d 8 , - 75 ° C) 173.05, 170.55, 158.50, 157.34, 127.15, 126.84, 125.90, 124.54, 122.32, 113.80, 108.45, 45.38, 39.86, 28.70, 28.45. Note, repeated attempts to obtain passing elemental analysis failed to yield adequate results. Synthesis of [ Ti( - Ntolyl)(OArCH 2 ArO)] 2 · HNMe 2 : A scintillation vial was charged with 100 mg (0.21 mmol, 1 equiv) Ti(bisphenoxide)(NMe 2 ) 2 , a stir bar, and 2 mL benzene. A separate solution of 23 mg (0.21 mmol, 1 equiv) of H 2 Ntolyl was prepared in 1 mL benzene. The H 2 Ntol yl solution was added dropwise to the stirred Ti solution, which resulted in a color change from yellowish orange to dark brown. The solution was stirred for 1 h at room temperature, and the volatiles removed under reduced pressure. This provided powdery d ark brown residue, which was 485 dissolved in a minimal amount of n - hexane. The concentrated n - hexane solution was chilled at - 35 °C for 2 d to yield X - ray quality crystals of [Ti(OArCH 2 ArO)(µ - Ntolyl)] 2 NHMe 2 (41 mg, 39 %). 1 H NMR (500 MHz, benzene - d 6 ) 7.06 6.98 (m, 6H), 6.80 (d, J = 7.8 Hz, 2H), 3.80 (d, J = 14.2 Hz, 1H), 3.36 (d, J = 14.3 Hz, 1H), 2.15 (s, 6H), 2.01 (s, 3H), 1.79 (d, J = 4.8 Hz, 1.5H), 1.66 (s, 18H). 13 C NMR (126 MHz, benzene - d 6 ) 161.02, 136.29, 132.93 (d, J = 47.4 Hz), 130.26, 129.65, 129.37, 126.11, 121.40, 40.30, 31.30, 21.19, 20.81. EA 62 H 81 O 4 N 3 Ti 2 : C, 77.39; H, 9.74; N, 0.0. Found: C, 76.97; H, 9.46; N, 0.10. 486 Additional Titanium Complexes: The following Ti complexes were synthesized and fully characterized by Dhwani Kansal: tetrakis(2 - tert - butyl - 4 - methyl - phenoxide)Ti(IV), tetrakis(2 - tert - butyl - 4 - methoxy - phenoxide)Ti(IV), tetrakis(2 - tert - butyl - 4 - bromo - phenoxide)Ti(IV), tetrakis(2 - tert - butyl - phenoxide)Ti(IV), tetrakis(2 - tert - butyl - 4 - fluoro - pheno xide)Ti(IV), and tetrakis(2 - tert - butyl - 4 - trifluoromethyl - phenoxide)Ti(IV). Additionally, the following complexes were prepared according to literature procedures: Ti(Cl) 2 (OArCH 2 ArO), Ti(I) 2 (OArCH 2 ArO), Ti(O i Pr) 2 (OArCH 2 ArO), and Ti(OArCH 2 ArO) 2 . 18 The recorded 1 H and 13 C NMR data matched previous reports, however, the latter 3 complexes had never been structurally characterized. Single crystal X - ray structures were determined with these 3 complexes. Additionally single crystals of the Ti(Cl) 2 (OArCH 2 ArO) were examined by X - ray diffraction, and a matching unit cell to previous structural reports was determined. The X = O i Pr complex is dimeric in the solid state, but monomeric in solution, according to in situ molecular weight calibrations with D OSY NMR. For completeness, Ti(I) 2 (OArCH 2 ArO) was also examined by DOSY NMR, and similarly determined to be monomeric in solution. This complex was also observed as a monomer in the solid state. These results are presented below. Ti(NMe 2 ) 2 (OArCH 2 ArO) was p repared from modification of literature reports. 4 The new procedure for preparation of this complex is described below. The 1 H and 13 C NMR data for this complex matches previous reports. Synthesis of tetrakis(2,4 - di - tert - butyl - phenoxide)Ti(IV): A scintillation vial was charged with 200 mg Ti(NMe 2 ) 4 (0.89 mmo l, 1 equiv), a stir bar, and 5 mL n - hexane. The vial was chilled in a coldwell cooled with liquid N 2 for 20 min, until the hexane solution was frozen. The vial was warmed ambiently, with stirring, until the solution was just thawed, and a 2 mL toluene solu tion 487 of 732 mg 2,4 - di - tert - butylphenol was added dropwise (3.6 mmol, 4 equiv). The pale yellow solution rapidly turned bright orange. This solution was stirred for 2 h at room temperature, and the volatiles removed in vacuo to yield a powdery orange residu e. The residue was rinsed with hexane and dried once more. The residue was dissolved in a minimal amount of toluene and the concentrated solution was stored at - 35 °C overnight to yield X - ray quality crystals of Ti(OPh 2,4 - ditBu ) 4 (589 mg, 76 %). 1 H NMR (5 00 MHz, benzene - d 6 J = 2.4 Hz, 4H), 7.42 (d, J = 8.3 Hz, 4H), 6.95 (dd, J = 8.3, 2.4 Hz, 4H), 1.60 (s, 36H), 1.21 (s, 36H). 13 C NMR (126 MHz, benzene - d 6 145.41, 135.96, 124.70, 123.67, 122.98, 35.43, 34.63, 31.70, 30.57 . EA 56 H 84 O 4 Ti: C, 77.39; H, 9.74; N, 0.0. Found: C, 76.97; H, 9.46; N, 0.10. Modified Synthesis of Ti(OArCH 2 ArO)(NMe 2 ) 2 : A scintillation vial was charged with 280 mg of Ti(NMe 2 ) 4 (1.25 mmol, 1 equiv), 8 mL of Et 2 O, and a stir bar. This solution wa s chilled in an N 2 (l) coldwell for 10 min. Separately, a solution of 425 mg (1.25 mmol, 1 equiv) of the H 2 (bisphenoxide) ligand was prepared in 2 mL of Et 2 O. The chilled Ti solution was stirred, and the ligand solution was added dropwise to it over the cou rse of a few minutes. The solution changed from pale yellow to intense yellow - orange upon addition. The stirred solution was allowed to come to room temperature and stirred for 4 h. The volatiles were removed under reduced pressure to yield a sticky orange oil. The orange oil was dissolve in 2 mL pentane and the volatiles removed once more to give a sparkly orange foam. This foam was rinsed with cold pentane ( - 35 °C) to give an orange pentane extract and a yellow powder. The pentane extract can be chilled t o yield yellow powder (product). Additionally, the yellow powder can be purified by recrystallization from pentane or hexane ( - 35 °C) to yield 402 mg (68%) of the purified compound. Characterization of the complex matches the previous report. 4 488 Examination of Decomposition Pathways with Iminoamination Product and Ti(dpm)(X) 2 and a different Ti(OArCH 2 ArO)(NMe 2 ) 2 Precatalyst: To supplement the kinetics - of the catalyst deactivation event(s), stoichiometric attempts to probe the reactivity of various Ti species that may exist in the catalytic solution were carried out (6.3). These attempts primarily served to elucidate some of the places where catalyst degradation and noninnocence of the product toward the catalyst may arise. The results of these studies were informative and confirm several possible means by which catalyst deactivation may occur. 489 Figure 6 . 13 1 H NMR of Ti(dpm)(NMe 2 ) 2 and 3CC heated at 80 °C, 40 h in C 6 D 6 . 490 Figure 6 . 14 1 H NMR of [Ti(dpm)(Ntolyl)] 2 , t BuNC (xs), and 3CC in C 6 D 6 , 80 °C at 6 h. 491 Figure 6 . 15 1 - Ntolyl)(dpm)] 2 with t BuNC in situ in C 6 D 6 . 1 H NMR (500 MHz, Benzene - d 6 ) 7.50 (s, 4H), 6.70 (d, J = 8.0 Hz, 4H), 6.43 (s, 8H), 6.21 (s, 4H), 2.00 (s, 6H), 1.86 (s, 12H). 492 Stability of [Ti(OArArO)(Ntolyl)]· 1 / 2 (NHMe 2 ) in - situ: A J. Young tube was charged with 10 mg of [Ti(OArArO)(µ 2 - Ntolyl)] 2 NHMe 2 (which appears to exist mostly as the monomer in soluti on by 1 H DOSY Molecular Weight Calibration), and 1.0 mL of C 6 D 6 . The solution was mixed and the tube sealed with a Teflon stopper. The tube was transferred from the glovebox to an 85 °C oil bath and was heated for 16 h. The contents of the tube were examin ed by 1 H NMR. Several sets of peaks were easily distinguished after heating. The characteristic doublets for the CH 2 linker protons in each of the following species were observed: Ti(OArArO)(Ntolyl) · 1 / 2 (HNMe 2 ), Ti(OArArO) 2 , and Ti(OArArO)(NMe 2 ) 2 . This als o suggests there is at least one more complex in solution, which have distributions of N - based ligands, and lack an (OArArO) fragment as well, i.e. [Ti(Ntolyl) X - 4 (NMe 2 ) 4 - X ] n . This experiment, suggests that disproportionation likely changes the catalyst loa ding in the 3CC reactions over the course of the reaction period in agreement with 1 H NMR observations of a genuine 3CC reaction with the Ti(OArCH 2 ArO)(NMe 2 ) 2 catalyst. 493 Figure 6 . 16 1 H NMR of [Ti(OArCH 2 ArO)(µ - Ntolyl)] 2 ·NHMe 2 after heation 16 h at 80 ° C. 494 In - situ Reactivity of [Ti(OArArO)(Ntolyl)] 2 with 3CC: A scintillation vial was charged with 20 mg of the Ti(Ntolyl)(bisphenoxide) species and 1.5 mL of C 6 D 6 . The solution was stirred to ensure complete dissolution of the Ti species. Then, 20 m g of the isomeric mixture of 3CC k was dissolved in C 6 D 6 and added to the Ti solution. The mixture was allowed to sit at room temperature for 10 minutes, and a basel ine NMR spectrum was taken, This spectrum is shown in Fig. 6. 16 and shows the two isomers of 3CC, as well as the Ti(Ntolyl)(OArCH 2 ArO). The solution was then heated in an oil bath in a J. Young tube with a Teflon stopper for 3 h at 85 °C. Another 1 H NMR spectrum was taken and is shown in Fig. 6. 17 . The starting materials are all still present, however, many of the peaks have decreased in intensity, and a few new peaks are beginning to grow into the baseline of the spectrum. The sample was returned t o the oil bath and heating continued for a total of 48 h at 85 °C. The spectrum after 48 h is shown in Fig. 6 18 . We can see that one of the 3CC k isomers (B) has decreased in overall peak intensity. Additionally, a new set of peaks that correspond to the Ti (OArCH 2 ArO) 2 disproportionation product are evident in solution. It is also relevant to note that there is no evidence of free H 2 Ntolyl in this spectrum. This suggests that the (Ntolyl) fragments are bound to the other ½ equivalent of Ti still in solution . This is also likely where the consumed 3CC product is. However, clear peaks that correlate to this species cannot be deciphered from the crowded aromatic and aliphatic regions of this 1 H NMR spectrum. This experiment is another piece of evidence that one possible pathway for decomposition of the catalyst is via ligand disproportionation facilitated directly by interaction with the 3CC product. 495 Figure 6 . 17 Proposed decomposi tion pathway and final products observed (top) and proposed (bottom) for the Ti - imide species upon heating with the iminoamination product in C 6 D 6 . 496 Figure 6 . 18 1 H NMR of 3CC(A/B) + [Ti(OArCH 2 ArO)(Ntolyl] 2 ·HNMe 2 heated for 0 h at 85 °C showing no Ti(OArCH 2 ArO) 2 . 497 Figure 6 . 19 1 H NMR of 3CC(A/B) + [Ti(OArCH 2 ArO)(Ntolyl] 2 ·HNMe 2 heated for 3 h at 8 5 °C showing a small amount of Ti(OArCH 2 ArO) 2 . 498 Figure 6 . 20 1 H NMR of 3CC(A/B) + [Ti(OArCH 2 ArO)(Ntolyl] 2 ·HNMe 2 heated for 48 h at 85 °C showing only Ti(OArCH 2 ArO) 2 as identifiable Ti species. 499 Iminoamination reaction catalyzed by Ti(OArCH 2 ArO)(NMe 2 ) 2 : A 15 mL pressure tube was charged with 47 mg of Ti(OArCH 2 ArO)(NMe 2 ) 2 (0.1 equiv), a stir bar, and 1 mL of toluene. To this solution was added 1 equiv H 2 NPh and the solution was stirred for 5 min at room temperature. Then a 1 mL solution of 1 equiv t BuNC, 1 equiv 1 - octyne, and dodecane (0.0001 mol, 17 mg) in toluene was added to the pressure tube solution. The tube was sealed and transferred from the glovebox to a preheated oil bath (110 ° C) and was heated with stirring for 16 h. The reaction solution was analyzed by GC - MS and the amount of iminoamination and other reaction products quantified by GC - FID analysis . In situ Iminoamination Reaction Catalyzed by Ti(OArCH 2 ArO)(NMe 2 ) 2 : A solution containing ferrocene (46.5 mg, 0.05 M internal standard), H 2 NPh (93 mg, 0.2 M), 1 - octyne (110 mg, 0.2 M), and t BuNC (82 mg, 0.2 M) in tol - d 8 (diluted to 5.0 mL) was prepared volumetrically. To 19 mg of the Ti catalyst (0.04 M) 1.0 mL of the prepared solution was added. After complete dissolution of the Ti complex, the solution was loaded into a J - young NMR tube and sealed with a Teflon cap. The tube was heated at 110 ° C for 1 6. The solution was then examined by 1 H NMR (shown below). 500 Figure 6 . 21 1 H NMR of the iminoamination reaction catalyzed by 20 mol% Ti(OArCH 2 ArO)(NMe 2 ) 2 in tol - d 8 . Peaks at 11.2 and 10.4 ppm are for the two different regioisomers of the 3CC product. The large singlet at 3.97 ppm is Fc as internal standard. The peak at 3.35 ppm belongs to the Ti(OArCH 2 ArO) 2 disproportionation species. 501 NMR and GCMS Spectra Figure 6 . 22 1 H NMR of an isomeric mixture of 3CC(A) and (B) in CDCl 3 . 502 Figure 6 . 23 13 C NMR of an isomeric mixture of 3CC k (A) and (B) in CDCl 3 . 503 Figure 6 . 24 GCMS of 3CC isomers A and B; fragmentation pattern for A isomer. 504 Figure 6 . 25 GCMS of 3CC isomers A and B; fragmentation pattern for B isomer. 505 Figure 6 . 26 HRMS for isomeric mixture of 3CC. 506 Figure 6 . 27 1 H NMR of the 4CC product in CDCl 3 . 507 Figure 6 . 28 13 C NMR of the 4CC product in CDCl 3 . 508 Figure 6 . 29 GCMS of the 4CC product and MS fragmentation pattern. 509 Figure 6 . 30 HRMS of the 4CC product. 510 Figure 6 . 31 gCOSY NMR of the 4CC product in CDCl 3 . 511 Figure 6 . 32 HMBC NMR of 4CC in CDCl 3 . 512 Figure 6 . 33 1 - Ntolyl)(dpm)] 2 in tol - d 8 .(room temperature, high vac grease and hexane impurities) 513 Figure 6 . 34 1 H NMR of - Ntolyl)(dpm)] 2 in tol - d 8 .( - 75 °C, high vac grease and hexane impurities) 514 Figure 6 . 35 13 - Ntolyl)(dpm)] 2 in tol - d 8 .(room temperature, high vac grease and hexane impurities) 515 Figure 6 . 36 1 H NMR of [Ti(OArCH 2 - Ntolyl)]·HNMe 2 in C 6 D 6 . 516 Figure 6 . 37 13 C NMR of [Ti(OArCH 2 - Ntolyl)]·HNMe 2 in C 6 D 6 . 517 Figure 6 . 38 1 H NMR of Ti(2,4 - di - tert - butyl - phenoxide) 4 in C 6 D 6 . 518 Figure 6 . 39 13 C NMR of Ti(2,4 - di - tert - butyl - phenoxide) 4 in C 6 D 6 . 519 X - ray Structures - Ntolyl)] 2 (KA_10Dec2018) Figure 6 . 40 C rystal data and structure refinement for earlyy2. Identification code earlyy2 Empirical formula C 18 H 19 N 3 Ti Formula weight 325.26 Temperature/K 173.0 Crystal system triclinic Space group P - 1 a/Å 7.7583(3) b/Å 9.5204(4) 520 c/Å 11.4421(4) 70.673(2) 88.342(3) 84.629(3) Volume/Å 3 794.01(5) Z 2 calc g/cm 3 1.360 - 1 4.546 F(000) 340.0 Crystal size/mm 3 0.381 × 0.152 × 0.123 Radiation 8.188 to 136.746 Index ranges - - - Reflections collected 6753 Independent reflections 2787 [R int = 0.0877, R sigma = 0.0758] Data/restraints/parameters 2787/0/202 Goodness - of - fit on F 2 1.015 R 1 = 0. 0613, wR 2 = 0.1527 Final R indexes [all data] R 1 = 0.0734, wR 2 = 0.1604 Largest diff. peak/hole / e Å - 3 0.98/ - 0.68 521 [Ti(OArCH 2 ArO)( - Ntolyl)] 2 · HNMe 2 (KA_04Dec2018) Figure 6 . 41 Crystal data and structure refinement for c2c_early_a. Identification code c2c_early_a Empirical formula C 66.5 H 91.5 N 3 O 4 Ti 2 Formula weight 1092.72 Temperature/K 173.0 Crystal system monoclinic Space group C2/c a/Å 34.0256(18) b/Å 17.3283(7) 522 c/Å 25.5196(12) 90 121.054(2) 90 Volume/Å 3 12890.1(11) Z 8 calc g/cm 3 1.126 - 1 2.456 F(000) 4700.0 Crystal size/mm 3 0.223 × 0.205 × 0.16 Radiation 5.934 to 136.488 Index ranges - - - Reflections collected 83770 Independent reflections 11770 [R int = 0.1037, R sigma = 0.0513] Data/restraints/parameters 11770/14/686 Goodness - of - fit on F 2 1.030 R 1 = 0.0570, wR 2 = 0.14 88 Final R indexes [all data] R 1 = 0.0881, wR 2 = 0.1693 Largest diff. peak/hole / e Å - 3 0.84/ - 0.33 523 Ti(OArCH 2 ArO) 2 (KA_06Dec2018CU) Figure 6 . 42 Crystal data and structure refinement for p21_c_a. Identification code p21_c_a Empirical formula C 35.67 H 52 O 2.67 Ti 0.67 Formula weight 555.37 Temperature/K 173.0 Crystal system monoclinic Space group P2 1 /c a/Å 12.9534(4) b/Å 21.2798(5) 524 c/Å 19.3799(4) 90 98.805(2) 90 Volume/Å 3 5279.0(2) Z 6 calc g/cm 3 1.048 - 1 1.665 F(000) 1812.0 Crystal size/mm 3 0.193 × 0.154 × 0.121 Radiation 6.208 to 136.882 Index ranges - - - Reflections collected 29162 Independent reflections 9481 [R int = 0.1607, R sigma = 0.1945] Data/restraints/parameters 9481/0/520 Goodness - of - fit on F 2 0.896 R 1 = 0.0796, wR 2 = 0.1959 Final R indexes [all data] R 1 = 0.1801, wR 2 = 0.2407 Largest diff. peak/hole / e Å - 3 0.83/ - 0.32 525 [Ti(OArCH 2 ArO)(O i Pr)( - O i Pr)] 2 (KA_22Dec218) Figure 6 . 43 Crystal data and structure refinement for rjs. Identification code rjs Empirical formula C 30 H 45 Cl 3 O 4 Ti Formula weight 623.91 Temperature/K 173.01 Crystal system triclinic Space group P - 1 a/Å 11.57250(10) b/Å 12.38990(10) 526 c/Å 13.0612(2) 67.6580(10) 80.5460(10) 66.6990(10) Volume/Å 3 1590.67(3) Z 2 calc g/cm 3 1.303 - 1 4.849 F(000) 660.0 Crystal size/mm 3 0.21 × 0.141 × 0.107 Radiation 7.318 to 136.352 Index ranges - - - Reflections collected 20754 Independent reflections 5629 [R int = 0.0697, R sigma = 0.0519] Data/restraints/parameters 5629/6/367 Goodness - of - fit on F 2 1.065 R 1 = 0.0509, wR 2 = 0.1339 Final R indexes [all data] R 1 = 0.0738, wR 2 = 0.1479 Largest diff. peak/hole / e Å - 3 0.41/ - 0.54 527 Ti(2,4 - di - tert - butyl - phenoxide) 4 (KA_14Dec2018) Figure 6 . 44 Crystal data and structure refinement for uc_a. Identification code uc_a Empirical formula C 56 H 84 O 4 Ti Formula weight 869.13 Temperature/K 172.99 Crystal system tetragonal Space group P - 42 1 c a/Å 12.3120(3) b/Å 12.3120(3) c/Å 17.7572(8) 528 90 90 90 Volume/Å 3 2691.73(18) Z 2 calc g/cm 3 1.072 - 1 1.650 F(000) 948.0 Crystal size/mm 3 0.315 × 0.21 × 0.2 Radiation 8.74 to 136.512 Index ranges - - - Reflections collected 9325 Independent reflections 2462 [R int = 0.0433, R sigma = 0.0539] Data/restraints/parameters 2462/0/144 Goodness - of - fit on F 2 1.022 R 1 = 0.0350, wR 2 = 0.0846 Final R indexes [all data] R 1 = 0.0397, wR 2 = 0.0869 Largest diff. peak/hole / e Å - 3 0.13/ - 0.39 Flack parameter 0.019(6) 529 Ti(OArCH 2 ArO)I 2 (KA_25Jan2019) Figure 6 . 45 Crystal data and structure refinement for early_a. Identification code early_a Empirical formula C 26 H 40 O 2 TiI 2 Formula weight 640.17 Temperature/K 173.0 Crystal system triclinic Space group P - 1 a/Å 9.6900(10) b/Å 9.8844(13) 530 c/Å 13.151(2) 90.518(7) 92.411(5) 95.161(4) Volume/Å 3 1253.3(3) Z 2 calc g/cm 3 1.6963 - 1 22.342 F(000) 625.6 Crystal size/mm 3 0.161 × 0.157 × 0.064 Radiation 8.98 to 136.3 Index ranges - - - Reflections collected 15520 Independent reflections 4410 [R int = 0.1065, R sigma = 0.0869] Data/restraints/parameters 4410/0/261 Goodness - of - fit on F 2 1.028 R 1 = 0.0521, wR 2 = 0.1228 Final R indexes [all data] R 1 = 0.0767, wR 2 = 0.1366 Largest diff. peak/hole / e Å - 3 1.38/ - 1.39 531 Note: an X - ray crystal structure of the Ti(Cl) 2 (OArCH 2 ArO) has been previously reported. Single crystals were grown and examined by X - ray diffraction, providing a matching unit cell to the previous report. The species is monomeric in the solid state. 532 1 H DOSY NMR Molecular Wei ght Calibrations Figure 6 . 46 DOSY MW determination for [Ti(OArCH 2 - Ntolyl)] x ·1/2 NHMe 2 in C 6 D 6 . Table 6 . 5 Experimentally determined diffusion coefficients (D) and calculated MW for Ti species for [Ti(OArCH 2 - Ntolyl)] x ·1/2 NHMe 2 in C 6 D 6 . Compound MW (g/mol) Log(MW) D Log(D) THF 72 1.86 27.2 1.45 Benzene 83 1.92 26.2 1.41 TMS 4 Si 321 2.51 14.5 1.12 [Ti(Ntolyl)(OArCH 2 ArO)] x 551 ( ±62 ) 2.74 10.5 1.02 y = - 0.4262x + 2.2315 R² = 0.9991 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 0 1 2 3 4 5 log(D) log(MW) MW Determination of [Ti(OArCH 2 ArO)(Ntolyl)] 2 533 Figure 6 . 47 - Ntolyl)] x in C 6 D 6 . Table 6 . 6 Experimental determination of diffusion coefficients (D) and calculated MW for Ti species for - Ntolyl)] x in C 6 D 6 . Compound MW (g/mol) Log(MW) D log(D) bnz 72 1.86 27.9 1.45 Fc 186 2.27 16.7 1.22 TMS 4 Si 321 2.51 12.7 1.10 [Ti(dpm)(Ntolyl)] n 465 ( ± 51) 2.67 10.8 1.03 y = - 0.5281x + 2.4251 R² = 0.9996 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 0 1 2 3 4 5 log(D) log(MW) MW Determination of [Ti(Ntolyl)(dpm)] 2 534 Figure 6 . 48 DOSY MW determination of Ti(OArCH 2 ArO)(I) 2 in C 6 D 6 . Table 6 . 7 Experimentally determined diffusion coeffieicnts (D) and calculated MW for the Ti species for Ti(OArCH 2 ArO)(I) 2 in C 6 D 6 . compound MW log(MW) D log(D) Error Fc 186.04 2.269606 17.61 1.245759 0.25 C 6 D 5 H 78 1.892095 26.29 1.419791 0.58 Ti(OArArO) 2 724.85 2.860248 8.95 0.951823 0.81 Ti(OArARrO)(I) 2 563 ( ± 94) 2.750929 10.16 1.006894 0.2 Real monomer weight 638 y = - 0.4846x + 2.3401 R² = 0.9996 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 1.5 2 2.5 3 log(D) log(MW) MW Determination of Ti(OArCH 2 ArO)I 2 535 Figure 6 . 49 DOSY MW determination of Ti(OArCH 2 ArO)(O i Pr) 2 complex in C 6 D 6 . Table 6 . 8 Experimentally determined diffusion coefficients (D) and calculated MW of the Ti species for Ti(OArCH 2 ArO)(O i Pr) 2 complex in C 6 D 6 . compound MW log(MW) D log (D) Error Fc 186.04 2.269606 20.21 1.305566 0.76 C 6 D 5 H 78 1.892095 24.6 1.390935 0.76 Ti(OArCH 2 ArO) 2 724.85 2.860248 9.62 0.983175 1.4 Ti(OArCH 2 ArO)(OiPr) 2 522.62 ( ± 74) 2.718186 11.57 1.063333 0.33 Real monomer weight 504 y = - 0.4322x + 2.2381 R² = 0.9615 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 1.5 2 2.5 3 log(D) log(MW) MW Determination of Ti(OArCH 2 ArO)(O i Pr) 2 536 Kinetic Analysis for Homogeneous Ti(dpm)(NMe 2 ) 2 Catalyzed Rate Law: For each set of kinetics conditions examined to determine the rate law, the following general procedure was applied. The specific conditions for each run are listed in Table 1 above. General Procedure: The fo llowing reagents were measured separately by mass: (1) Tidpm(NMe 2 ) 2 (78 - 312 mg, 5 - 20 mol%), (2) dodecane (212 mg, 1.25 mmol, 0.05 M), (3) H 2 NPh (465 mg - 2.32 g, 5 - 25 mmol), (4) t BuNC (415 - 830 mg, 5 - 10 mmol), and (5) 1 - octyne (550 mg - 2.75 g, 5 - 25 mmol). In a scintillation vial, the Tidpm(NMe 2 ) 2 was dissolved in 5 mL toluene, and the dodecane and H 2 NPh were added to this solution, causing the solution to change colors from bright orange to dark reddish brown. This solution was stirred at room temperature for 5 - 10 min and transferred to a 25.0 mL volumetric flask. The t BuNC and 1 - octyne were added to the flask, and the solution was diluted to 25.0 mL with toluene. This solution was thoroughly mixed and transferred in 1 mL aliquots to sample tubes (generally 10 - 1 2 per entry). The tubes were sealed and transferred from the glovebox to a preheated oil bath. The elapsed time from the start of the reaction was recorded each time a sample was removed for GC analysis, ranging from 30 min to 28 h. The samples were analy zed by GC - MS to look for reaction products and detection of unwanted side products. GC - FID was used to quantify the amounts of 3CC, HA, and FA or 4CC production in each sample based on external calibrations standardized with internal dodecane (0.05 M) from the authentic isolated products, obtained by separation from the organic reaction mixtures. The concentrations of products were used in the graphical analysis of the order of each reactant. 13, 14 537 Scheme 6 . 9 Iminoamination reaction examined under different reaction conditions to probe the rate law and suggest optimal reaction conditions. 538 Figure 6 . 50 The graphical determination of reaction rate dependence on alkyne concentration. Purple spheres = 0.2 M (Entry 3); Red spheres = 1.0 M (Entry 4); Grey spheres = 0.4 M (En try 5). 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0 500 1000 1500 2000 [3CC] (M) t*[alk] 0 Determination of Order in Alkyne 539 Figure 6 . 51 The graphical determination of reaction rate dependence on isonitrile concentration. Purple spheres = 0.2 M (Entry 3); Orange spheres = 0.4 M (Entry 6) . 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0 200 400 600 800 1000 1200 1400 1600 1800 [3CC] (M) t*[ t BuNC] 0 Determination of Order in Isonitrile 540 Figure 6 . 52 The graphical determination of reaction rate dependence on amine concentration. Purple spheres = 0.2 M (Entry 3); light blue spheres = 0.4 M (Entry 7); Green spheres = 1.0 M (Entry 8). 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0 200 400 600 800 1000 1200 1400 [3CC] (M) t*[amine] 0.6 Determination of Order in Amine 541 Kinetic analysis o f Ti(OArCH 2 ArO)(NMe 2 ) 2 catalyzed Iminoamination: The general kinetic analysis procedure was applied, using the reaction conditions described for Entry 3 above (Table 1). The figures below show a side - by - side comparison of the reaction results with Ti(dpm)( NMe 2 ) 2 catalyst, Figure 6 . 53 Reaction progress of two identical kinetics trials run with Ti(dpm)(NMe 2 ) 2 and Ti(OArCH 2 ArO)(NMe 2 ) 2 . Similar results were obtained with both catalysts under the conditions u sed for kinetics despite better performance of the Ti(OArCH 2 ArO)(NMe 2 ) 2 under normal conditions applied to a typical iminoamination reaction. 0 0.05 0.1 0.15 0.2 0.25 0 1000 2000 Concentration (M) of Aniline Time (min) Kinetic Analysis of Ti catalyzed 3CC Ti(Bis- aryloxide) Ti(dpm) 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0 500 1000 1500 2000 Concentration (M) of 3CC product A+B time (min) Kinetic Analysis of Ti(bisphenoxide) catalyzed 3cc Ti(Bis- Aryloxide) Ti(dpm) 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0 500 1000 1500 2000 concentration (M) Time (min) Total Concnetration of All Reaction Products (M) vs Time (min Ti(OArArO) Ti(dpm) 542 K eq determination by 1 H NMR: The determination of the K eq values for the ligand exchange reactions, of the general form shown in Fig. 6. 54 below, were performed by monitoring the concentrations of the 3 species in solution by 1 H NMR. Solutions were prepared in C 6 D 6 on the order of 0.025 - 0.05 M (in titanium). Fer rocene was included as an internal standard. Due to the long T 1 of ferrocene ( ~ 30 s) in deoxygenated NMR solvents, the NMR experiments were performed differently from the standard 1 H NMR experiments with d1=150 and gain = 30. The solutions were examined ev ery few days until the integral values of the species in solution had stopped changing. Once the solution concentrations of the species of interest had leveled off, 3 spectra were taken and averaged. From these triplicate measurements, an error could be a ssigned to the K eq values calculated, as they will be affected by the error of manual integration in the 1 H NMR. For the following X ligands, the equilibrium processes were initiated from the heteroleptic Ti(X) 2 (OArCH 2 ArO) species: I ¯ , Cl ¯ , ( O i Pr) ¯ , and (N Me 2 ) ¯ . The Ti(OAr) 2 (OArCH 2 ArO) species typically begin the ligand exchange processes during isolation from the crude reaction mixture and therefore cannot be isolated free of impurities. For the ligand exchange reactions with the various 2 - tert - butyl - 4 - R - phenoxide ligands, the equilibrium exchange process was initiated from the two homoleptic species Ti(OAr) 4 and Ti(OArCH 2 ArO) 2 added in equivalent molar amounts to the initial solution. 543 Figure 6 . 54 Equilibrium ligand exchange reaction used to determine K eq experimentally. For reactions where K eq is small, starting materials 1 and 2 were used. For reactions where K eq was large, 3 could be prepared and isolated cleanl y and was utilized in these experiments. 544 Figure 6 . 55 1 H NMR of Ti(OArCH 2 ArO) 2 in C 6 D 6 . 545 Figure 6 . 56 13 C NMR of Ti(OArCH 2 ArO) 2 in C 6 D 6 . 546 Figure 6 . 57 1 H NMR of the equilibrium mixture of Ti(NMe 2 ) 4 , Ti(NMe 2 ) 2 (OArCH 2 ArO), and Ti(OArCH 2 ArO) 2 . 547 Figure 6 . 58 1 H NMR of the equilibrium mixture of Ti(OAr 4 - tert - butyl ) 4 , Ti(OAr 4 - tert - butyl ) 2 (OArCH 2 ArO), and Ti(OArCH 2 ArO) 2 . 548 Modeling of K eq as a function of sterics and electronics: Excel was used to perform standard ordinary least squares fits on the desired parameter arrays following the basic matrix equation shown below: Here each array is treated using standard matrix formulas in excel to give final values for the coefficients for the K eq equation in the output array, C. The matrix [A] contains the parameters fitted, in this case LDP and LDP values. The matrix [B] contains the property cor relating to the parameters, in this case K eq . The following combinations of variables were considered and the best overall fit resulted from simple electronic treatment of the K eq data. Table 6 . 9 Combinations of parameters examined for fitting the dependence of K eq bur . Least Squares Fit Trial Parameters 1 LDP, ( LDP) 2 , %V bur 2 2 , bur ) 2 3 2 , bur ) 2 bur 4 2 bur 5 2 We noted that while a very slight improvement in the R 2 value for the fitted parameters was noted in Trial 3, several of the calculated K eq values were small negative numbers. This appears to impart the data. Additionally, comparing trials 1 and 2 to trial 4, almost no improvement to the fit is made 549 Figure 6 . 59 Least Squares fit result for predicting K eq from LDP. y = x - 2E - 13 R² = 0.9916 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 K eq experimental K eq model - predicted Least Squares Fit: Trial 5 550 REFERENCES 551 REFERENCES (1) Cao, C.; Shi, Y.; Odom, A. L., A Titanium - Catalyzed Three - Component Coupling To - - Iminoamines. Journal of the American Chemical Society 2003, 125 (10), 2880 - 2881. (2) Odom, A. L.; McDaniel, T. J., Titanium - Catalyzed Multicomponent Couplings: Efficient One - Pot Syntheses of Nitrogen Heterocycles. Acc. Chem. Res. 2015, 48 (11), 2822 - 2833. (3) Barnea, E.; Majumder, S.; Staples, R. J.; Odom, A. L., One - Step Route to 2,3 - Diaminopyrroles Using a Titanium - Catalyzed Four - Component Coupling. Organometallics 2009, 28 (13), 3876 - 3881. (4) Billow, B. S.; McDaniel, T. J.; Odom, A. L., Quantifying ligan d effects in high - oxidation - state metal catalysis. Nature Chemistry 2017, 9 , 837. (5) DiFranco, S. A.; Maciulis, N. A.; Staples, R. J.; Batrice, R. J.; Odom, A. L., Evaluation of donor and steric properties of anionic ligands on high valent transition m etals. Inorganic chemistry 2012, 51 (2), 1187 - 200. (6) Bemowski, R. D.; Singh, A. K.; Bajorek, B. J.; DePorre, Y.; Odom, A. L., Effective donor abilities of E - t - Bu and EPh (E = O, S, Se, Te) to a high valent transition metal. Dalton transactions 2014, 4 3 (32), 12299 - 12305. (7) Aldrich, K. E.; Billow, B. S.; Holmes, D.; Bemowski, R. D.; Odom, A. L., Weakly Coordinating yet Ion Paired: Anion Effects on an Internal Rearrangement. Organometallics 2017, 36 (7), 1227 - 1237. (8) Aldrich, K. E.; Billow, B. S. ; Staples, R. J.; Odom, A. L., Phosphine interactions with high oxidation state metals. Polyhedron 2019, 159 , 284 - 297. (9) McDaniel, T. J.; Lansdell, T. A.; Dissanayake, A. A.; Azevedo, L. M.; Claes, J.; Odom, A. L.; Tepe, J. J., Substituted quinolin es as noncovalent proteasome inhibitors. Bioorganic & Medicinal Chemistry 2016, 24 (11), 2441 - 2450. ( 10 ) Majumder, S.; Gipson, K. R.; Odom, A. L., A Multicomponent Coupling Sequence for Direct Access to Substituted Quinolines. Organic Letters 2009, 11 (20 ), 4720 - 4723. ( 11 ) Majumder, S.; Odom, A. L., Titanium catalyzed one - pot multicomponent coupling reactions for direct access to substituted pyrimidines. Tetrahedron 2010, 66 (17), 3152 - 3158. ( 12 ) Dissanayake, A. A.; Staples, R. J.; Odom, A. L., Titanium - Catalyzed, One - Pot Synthesis of 2 - Amino - 3 - cyano - pyridines. Advanced Synthesis & Catalysis 2014, 356 (8), 1811 - 1822. ( 13 ) Burés, J., A Simple Graphical Method to Determine the Order in Catalyst. 20 16, 55 (6), 2028 - 2031. 552 ( 14 ) Burés, J., Variable Time Normalization Analysis: General Graphical Elucidation of Reaction Orders from Concentration Profiles. 2016, 55 (52), 16084 - 16087. ( 15 ) Rosner, T.; Le Bars, J.; Pfaltz, A.; Blackmond, D. G., Kinetic Stu dies of Heck Coupling Reactions Using Palladacycle Catalysts: Experimental and Kinetic Modeling of the Role of Dimer Species. J. Am. Chem. Soc. 2001, 123 (9), 1848 - 1855. ( 16 ) Hy droamination of Alkynes. Angew. Chem. Int. Ed. 2001, 40 (12), 2305 - 2308. ( 17 ) Walsh, P. J.; Baranger, A. M.; Bergman, R. G., Stoichiometric and catalytic hydroamination of alkynes and allene by zirconium bisamides Cp2Zr(NHR)2. Journal of the American Chem ical Society 1992, 114 (5), 1708 - 1719. ( 18 ) Okuda, J.; Fokken, S.; Kang, H. - C.; Massa, W., Synthesis and Characterization of - Methylene - bis(6 - tert - butyl - 4 - methylphenol). 199 5, 128 (3), 221 - 227. ( 19 ) Floriani, C.; Corazza, F.; Lesueur, W.; Chiesi - Villa, A.; Guastini, C., Eine empfindliche Sonde für Veränderungen in der Koordinationssphäre von Titan: Achtgliedrige Dioxatitanacyclen und ihre metallorganischen Derivate. 1989, 101 (1), 93 - 94. ( 20 ) Sun, L., Devore, D. D. (Dow Global Technologies, LLC), A process for preparing functional polymers through addition of amino and polymeryl groups to aldehyde moieties. United States Patent 2018, (PCT/US2016/054190). ( 21 ) Tinkler, S.; Deeth, R. J.; Dunca lf, D. J.; McCamley, A., Polymerisation of ethene by the novel titanium complex [Ti(Me3SiNCH2CH2NSiMe3)Cl2]; a metallocene analogue. Chemical communications 1996, (23), 2623 - 2624. ( 22 ) Mikami, K.; Terada, M.; Nakai, T., Catalytic asymmetric glyoxylate - en e reaction: a practical access to .alpha. - hydroxy esters in high enantiomeric purities. Journal of the American Chemical Society 1990, 112 (10), 3949 - 3954. ( 23 ) Mikami, K.; Matsumoto, Y.; Xu, L., Modification of alkoxo ligands of BINOL Ti ladder: Isolatio n and X - ray crystallographic analysis. Inorganica Chimica Acta 2006, 359 (13), 4159 - 4167. ( 24 ) Mikami, K., T., Masahiro, N. Takeshi, S., Noboru, K., Hidenori (Takasago International Corporation), Process for producing optically active alpha - hydroxycarboxyl ates. United States Patent 1990, (US4965398). ( 25 ) van Leeuwen, P. W. N. M., Decomposition pathways of homogeneous catalysts. Applied Catalysis A: General 2001, 212 (1), 61 - 81. ( 26 ) Lehn, J. - S. M.; Hoffman, D. M., Synthesis and Structures of Zirconium Ami Complexes. Inorganic chemistry 2002, 41 (15), 4063 - 4067. 553 ( 27 ) Group, O., Ligand Donor Parameters. Odom Group Wiki Page 2012 - 2019 . ( 28 ) Gokel, G. W. W., R. P., Weber, W. P., Phase - transfer Hofmann Carbylamine Reaction: tert - Butyl Isocyanide. Organic Syntheses 1976, 55 , 96. 554 CHAPTER 7. AN EXPLORATION OF THE SYNTHESIS AND ELECTRONIC PROPERTIES OF RUTHENIUM AND IRON IMIDO COMPLEXES 7.1 Introduction 9 , 10 Imide ligands and metal - nitrogen multiple bonds have been the topic of countless studies over the last 50 years. 1,2 The group 8 metals ha ve been no exception and their potential for metal nitrogen multiple bond formation has been particularly interesting for several reasons. 1) any) genuine examples of terminal oxo complexes exist and can be readily synthesized, isolated, and structurally characterized, specifically with octahedral coordination and a d - electron count >5. 3 Even today, only a handful of terminal oxo complexes in group 9 have been remotely characterized, some only in situ , and all with coordination numbers of 4 or 5. 4 - 6 Last year, the Anderson group published a well - characterized terminal Co(III) - oxo species with basal C 3v symmetry. The Co O bond in this structure, interestingly, is bent, similar to the electronic structure exhibited by many of the Group 8, +2, d 6 imido complexes discussed below with similar electronic structure. 7 This example highlights, one of the big (indirect) reasons why Fe, Ru, and Os multiple bond character is of inter est. 9 The work in this chapter is an expansion from initial investigations of Ru imido chemistry started by Dr. Amrendra Singh while he was a postdoc in the Odom group. His efforts are what first demonstrated the ligand - based - radical character upon oxidation of these species from Ru(II) to Ru(III). 10 These results have been submitted for publication to Inorganic Chemistry as a research article. The manuscript is currently in revision, but was accepted with minor revisions: Kelly E. Al drich, B. Scott Fales, Amrendra K. Singh, Richard J. Staples, Benjamin G. Levine, John McCracken, Milton R. Smith III, and Aaron L. Odom , Electronic and Structural Comparisons Between Iron(II/III) and Ruthenium(II/III) Imide Analogs 2019 , accepted with minor revisions. 555 Learning about these bonding interactions with group 8 metals stands to provide more information from which analysis of late, low valent metal - ligand multiple bonds, relevant to highly reactive metal - imido, metal - alkylidene, or metal - oxo species in ca talytic processes, can be applied. 2) Despite the much greater number and diversity of metal - imido compounds that have been synthesized and studied in the last several decades, several big questions about the nature of metal imide bonds remain. There are som e general trends known. For example, when a metal is substantially less electronegative than N, the - bonding orbitals of the M N bond is primarily N centered. When the metal becomes more electronegative, this orbital shifts its distribution toward the met al. 8 The resultant change in electronic structure from specific changes to the property of a metal, and the ability to predict what these changes will do to the reactivity of a metal imido complex have not been fully developed. One of the simplest trends, which has not been investig ated in a systematic way, is what happens to a metal imido bond transcending a group (i.e. Fe, Ru, and Os), and how differences in the metal down a series changed the reactivity. More targeted efforts to establish basic trends stand to provide more rapid i nformation about these types of interactions than random syntheses of various M=NR complexes. 3) Fe, Ru, and Os multiple bonds to nitrogen have been repeatedly found in highly active biological systems (i.e. nitrogenase enzyme active site 9 and cytochrome P450 10,11 ). In many of these systems, highly reactive Fe N mu ltiple bonds are suspected to be important intermediates within catalytic cycles. 12 - 14 However, exact intermediates and mechanism of action in many of these systems is not fully understood. Further study is needed to 556 completely establish how these enzymes catalyze reactions like N 2 reduct ion or aminations. Along the same lines, the Haber Bosch process utilizes solid Fe, Ru, or Os catalysts to reduce N 2 to NH 3 in heterogeneous reactors that operate at high pressures and temperatures. 15,16 Several mechanisms have been p roposed for this reduction process, including the adsorption of both N 2 and H 2 to solid catalysts. Many steps in the process, including intermediates, have become more fully understood over the past several years as advanced spectroscopy techniques have be en developed (i.e. Auger electron spectroscopy and SEM). So, while progress has been made to understand these catalysts, much room to improve these catalyst systems fur ther. 15 Practically, we need ammonia for food production, and currently there is no way around the highly energy intensive conditions required to reduce N 2 to NH 3 . Chemically, there should be a way to connect the dots between nitrogenase and the solid catalysts used in the Haber - Bosch process. Thorough investigation of these M N multiple bonds within discrete organometallic complexes offers a way to gain insight into these immensely important catalytic processes and reduce energy usage to produce ammonia catalyt ically. For these reasons, among others, an astounding amount of work has gone into the synthesis and characterization of Fe - imido complexes in the last few decades. 17 - 22 Using careful ligand design, Fe - imido complexes have been synthesized in the +1 to +6 oxidation state; as cationic, neutral, and anionic species; and demonstr ating a wide variety of electronic structures. In recent years, two groups have even successfully published the synthesis of three - coordinate Fe - bis(imido) complexes in the +4 and +5 oxidation states. 23,24 557 Figure 7 . 1 Examples of terminal mono - and bis - imido Fe complexes in the literature. Note the prevalence of both bulky and chelating ligands, which stabilize these complexes. 20 - 28 There are also several terminal Os - imide complexes known, again, with a variety of oxidation states and ligands. Specifically, Schrock and coworkers have prepared and studied the electronic properties of several bis - and tris(imide) comple xes, including Os(NAr) 3 , Os(NAr) 2 O 2 , Os(NAr) 2 (PMe 3 ) 2 , and Os(NAr) 2 ( 2 - alkyne) species, in addition to work by Sharpless. 29 - 33 Typically, the preparation of these complexes starts with OsO 4 , which is (relatively for Os) cheap and commercially available. Unfortunately, RuO 4 and FeO 4 materials are not, and so achieving analogous comple xes with the congeners is not feasible via the same synthetic routes. As a result, Os - imide complexes have the most well developed high - valent chemistry. Osmium - imide chemistry will be discussed more thoroughly in the following chapter (see Chapter 8). 8 Lagging behind both Fe and Os imido chemistry is Ru. Very few examples of Ru - imides exist in the literature. Both Schrock and Steedman have examples of Ru(II) mono(imido) complexes; monomeric, termi nal imido. 34 - 36 In 2013, the Odom group published a terminal Ru(II) imido complex, Ru(NAr)(PMe 3 ) 3 . 37 This complex, while interesting due to its unique geometry, reacts 558 much as one would predict an imido complex in a low oxidation state. The d 6 Ru(II) metal center nable enough vacant - orbitals of proper symmetry and orientation to localize the N - and bonding interactions. This results in a heavily N - centered HOMO, which effectively behaves as a lone pair of electron den sity on the imide N. The imide moiety, therefore, reacts with strong nucleophilic character. This electronic structure is also what causes the odd geometry of the complex, where the imide ligand tips toward one of the PMe 3 ligands preferentially, breaking the C 3 v symmetry of the molecule. The metal orbital participating in the HOMO is the d z 2 orbital. According to calculations, an 18 kcal/mol stabilization of the HOMO is achieved when the out - of - phase interaction between the Ru d z 2 orbital and the N orbital minimize their overlap (See Experimental, Fig. 7.1 7 ). This minimum occurs when the N orbital shifts its orientation, falling into the node of the d z 2 orbital. The distortion from C 3v symmetry is shown in the crystal structure in Fig. 7.2, below. 37 559 Figure 7 . 2 (top) X - omitted for clarity (N = blue, Ru = teal, P = pink). (bottom) Table 7 . 1 Table of relevant bond lengths and angles for Ru1 . Bond Angle ( ° ) Bond Distance ( Å ) Ru1 N1 C1 174.86 Ru1 N1 1.811 N1 Ru1 P1 113.70 Ru1 P1 2.224 N1 Ru1 P2 128.41 Ru1 P2 2.254 N1 Ru1 P3 122.63 Ru1 P3 2.239 Again, while this complex demonstrates an interesting electronic structure, which can be supported and rationalized with experimental and theoretical arguments, it quite possibly raises more questions than it answers. Do Fe and Os analogues impose the same geometric distortion, or do the differences in orbital overlap between metals have an effect? Would a less symmetric basal phosphine set make this more or less likely to occur? Does this electronic structure compromise stability upon oxidation? We lack th e fundamental knowledge about the character of M - imido bonds to make these sorts or predictions and assess what affects these changes would have on the reactivity. A lot of knowledge stands to be gained by synthesizing and characterizing the electronic st ructure of even very basic Ru - imide complexes. A direct comparison with Fe analogues serves 560 as a tool to simultaneously examine the bonding differences going down a series. Additionally, given the comparative wealth of knowledge that has been gained by the larger number of terminal Fe imides published in the past two decades, examining Fe and Ru analogues simultaneously facilitates comparisons with other known complexes. Thus, we set about preparing Fe and Ru analogues of various imide complexes to compare their reactivity, bonding properties, and overall electronics. By employing direct comparison between these congeners, we have gained some insight into the relative instability of Ru - imide complexes, which have helped explain synthetic difficulties associa ted with targeting these types of molecules. 7.2 Synthesis and Oxidation of Terminal Ru Imido Complexes The Ru(NAr)(PMe 3 ) 3 ( Ru1 ) complex mentioned above presented an ideal place to start when we first looked at expanding the synthetic chemistry of Ru imides. Because oxidation state effects on the character of the Ru N multiple bond was one direction that we wanted to explore with this project, direct chemical oxidation of Ru1 was attempted. However, this produced an unstable compound, leading to intractable mixtures that were not amenable to purification or additional characterization. Thus, we began looking for ways to stabilize t he Ru imide fragment prior to oxidation. phosphines. However, under mild conditions, bidentate ligands can replace the monodentate PMe 3 ligands. When the closest chelating ele ctronic surrogate to PMe 3 was employed for the ligand exchange, dimethylphosphinoethane (dmpe), two equivalents of dmpe add to Ru. This results in the formation of the stable, 5 - coordinate, 18 - electron Ru(NAr)(dmpe) 2 ( Ru2 ) complex shown in Fig. 7.3 . This c omplex is highly insoluble, and produces two enantiomers, which complicated 561 attempts at further study. We were also interested in retaining the unique geometry of the 4 - coordinate Ru - imide, so a different bis - phosphine was selected. Diphenylphosphinoethan e (dppe), when added to Ru(NAr)(PMe 3 ) 3 , is sterically demanding enough to limit addition of the chelate to Ru to a single equivalent, yielding Ru(NAr)dppe(PMe 3 ) ( Ru3 ) as the only product. This complex was characterized structurally and is shown in Fig. 7.3 . The structural properties of Ru3 are very similar to Ru1 , with the smallest N1 Ru1 P1 angle measuring 106.52 ° . 562 Figure 7 . 3 (top) Synthetic schemes for the synthesis of Ru2 and Ru3 . ( bottom) X - ray crystal structure for Ru3 (left) (Ru(NAr)dppe(PMe 3 )) and Ru2 (right) ; ellipsoids shown at 50% probability, hydrogens omitted for clarity. For Ru2 , the two enantiomers co - crystalize an d are disordered across the axis coincident with the P1 Ru1 P3 bond. Select bond lengths and angles are shown. 563 Table 7 . 2 Select bond lengths and angles from the single crystal X - ray structures for Ru3 and Ru2 , shown in Fig. 7.3. Bond Angle ( ° ) Bond Distance ( Å ) Ru3 Ru1 N1 C1 166.65 Ru1 N1 1.808 N1 Ru1 P1 106.52 Ru1 P1 2.237 N1 Ru1 P2 136.68 Ru1 P2 2.277 N1 Ru1 P3 127.03 Ru1 P3 2.239 Ru2 C1 N1 Ru1 163.82 Ru1 N1 1.921 N1 Ru1 P1 90.32 Ru1 P1 2.301 N1 Ru1 P2 132.34 Ru1 P2 2.317 N1 Ru1 P3 88.70 Ru1 P3 2.323 N1 Ru1 P4 131.30 Ru1 P4 2.317 With Ru3 isolated and bearing a similar geometry and electronic structure to Ru1 , attempts to oxidize Ru3 were undertaken. Amrendra discovered that when Ru3 and a silver salt (AgSbF 6 or AgBArF 24 ) are combined in a mixed solvent system containing DME and MeCN, oxid ation of the Ru3 occurs. However, this oxidation is accompanied by dimerization of the Ru species via radical para - coupling of the Ar fragment, as shown in Fig. 7. 4 , resulting in Ru4 . Figure 7 . 4 ( top ) Synthetic scheme for the production of Ru4 from Ru3 via oxidation with AgSbF 6 (AgBArF 24 can also be used). ( bottom ) X - ray crystal structure of the dimeric species with ellipsoids shown at 50% probabi lity; H atoms and disordered counter anion omitted for clarity. 564 When the same oxidation was performed in the absence of acetonitrile (i.e. only DME is used), a color change occurred, going from dark red to a pinkish - brown color, and the precipitation of A g 0 was evident. When the Ag 0 was removed by filtration and the filtrate was worked up, an unstable, powdery brown residue was obtained. While the material is not crystalline, and cannot be structurally characterized, several properties have led us to propo se the structure shown in Fig. 7. 5 , [Ru(NAr)(PMe 3 )dppe][SbF 6 ] ( Ru5 ) . Whereas the Ru3 precursor is soluble in highly nonpolar organic solvents, the new material is insoluble in most organic solvents; it is sparingly soluble in THF and DME and soluble but reactive with halogenated solvents. The material is also paramagnetic and shows both m etal - and ligand - based radical signal by EPR spectroscopy (vide infra). However, thorough EPR analysis of this species was not possible due to the instability of the species, even frozen as a glass in 2 - methyltetrahydrofuran. Figure 7 . 5 Proposed resonance contributors for Ru5 with radical distribution across the ortho and para positions of the imide aryl group. Given the reactivity demonstrated by the Ru species upon oxidati on (Fig. 7.4 ), and the presence of ligand - centered radical signal in the preliminary EPR data, delocalization of the unpaired electron density induced by oxidation appears to spread across the imide ligand fragment. Specifically, resonance localization sho uld put radical character on the metal, nitrogen, and ortho - and para - carbons of the Ar group. While the ortho positions are protected by the isopropyl substituents, the unprotected para carbon position appears to destabilize the oxidized complex. 565 There is precedence in the literature of similar ligand radical character in copper - imide and iron - imide systems, where similar reactivity has been demonstrated. 20,38,39 With this experimental evidence for ligand - radical based destabilization, I proposed a switch in the aryl derivative we were using for the imido fragment, adding an additional substituent to the para position. Similar efforts had shown significant improvement in the stability of the Fe - imide system published by Betley, et. al. 39 where their phenylimido was exchanged for 4 - t Bu - phenylimide and stability in the complex dramatically improved. The tri - substituted aniline, 2,4,6 - triisopropylaniline, can be readily prepared from 1,3,5 - triisopropylbenzene in two steps. 40 The procedure to attach the more highly substituted aniline to Ru as an imido ligand is then identical to the existing procedure to make Ru1 . The crystal structure of the resulting Ru(NAr*)(PMe 3 ) 3 ( Ru1* ) is shown in Fig. 7.6. Again, analogous to the pathway to convert Ru1 to Ru3 , addition of 1 equiv of dppe to Ru1* yields Ru(NAr*)(PMe 3 )dppe ( Ru3* ). Note, due to the greatly reduced solubility of Ru3* relativ e to Ru1* , Ru1* was not typically isolated. Rather, the synthesis was carried through to Ru3* , which made purification and removal of the residual equivalent of H 2 NAr* easier. 566 Figure 7 . 6 Synthesis procedure for Ru1* and Ru3* from cis - RuCl 2 (PMe 3 ) 4 starting material. The crystal structure of Ru1* is shown with e llipsoids are shown at 50% probability; H atoms omitted for clarity. With the potentially more stable Ru3* isolated, oxidation of Ru3* with AgSbF 6 was performed using DME as the solvent. This reaction resulted in the precipitation of Ag 0 and a similar color change as was previously noted with oxidation of Ru3 , going from re d to brown. The same structure is proposed for the oxidation product, [Ru(NAr*)(PMe 3 )dppe][SbF 6 ] Ru5* , as we previously proposed for Ru5 . Similar properties were noted between the two, with Ru5* also demonstrating paramagnetism consistent with one unpaired electron and similar solubility properties. Furthermore, Ru5* proved stable enough for more thorough investigation by EPR spectroscopy and was probed at length using this technique. These results are discussed in sections 7.5 and 7.6, below. 567 7.3 Synthesis and Oxidation of Fe Imide Analogues In tandem with the development of a more stable Ru - Imido cation, the extension of this chemistry to Fe congeners was also pursued. Starting from FeCl 2 , the addition of excess PMe 3 (6 equiv) leads to the generation of FeCl 2 (PMe 3 ) 4 in situ . This reaction has been previously reported and, like previous reports, we observe that this species is unstable when isolation is attempted. When the crude solution is exposed to reduced pressure, the solution changes color from pale gree n to brown, and after full removal of volatiles, appears to yield FeCl 2 as a pale orange residue. 41 No NMR signals are observed for the transient species in situ due to its high magnetic moment (high spin Fe(II)). When 2.1 equiv of LiNHAr is added to the in situ generated FeCl 2 (PMe 3 ) 4 , a rapid color change is noted, from transparent pale green to opaque dark orange. Upon stirring at room temperature for 24 h, the dark orange color gradually turns to a dull brownish green. This dark green compound is likel y Fe(NAr)(PMe 3 ) 3 ( Fe1 ) based on subsequent reactivity. The terminal Fe(II) - imido was trapped and isolated with chelating phosphines. Fe1 was also characterized in situ by 31 P and 14 N NMR of the crude reaction mixture, showing singlet shifts in the expected ranges based on observed shifts with Ru1 . However, like its precursor, it is also unstable under reduced pressure. When we removed the volatiles under reduced pressure, the reactio n mixture turned black, and the 31 P and 14 N NMR signals were no longer observed in the resulting residue. Upon addition of dppe to the crude Fe(NAr)(PMe 3 ) 3 reaction, no color change is noted. However, upon removal of the volatiles, a dark residue remains , which, when washed with hexane and recrystallized from toluene to yield X - ray quality crystals of Fe(NAr)(PMe 3 )dppe ( Fe3 ). However, the yield of Fe3 from this reaction is not very high (39%). This is, in part, because there is a second product yielded fr om this reaction which is a red, powdery substance that provides a 568 magnetic moment consistent with a high - spin Fe(II) complex ( µ eff = 5.42) ( Fe2 ). With the Fe2 isolated from the crude reaction mixture, X - ray quality crystals could not be grown. However, fr om additional experiments, including targeted synthesis of the complex we suspected to be Fe2 , it has been identified as Fe(NHAr) 2 dppe. (vide infra, Fig. 7.7 ) Attempts to oxidize Fe3 with AgSbF 6 , much like those to oxidize Ru3 , outwardly appeared to resul t in oxidation. However, the oxidized product Fe4 , was unstable and challenging to purify. Given the improvement that switching to NAr* offered for the Ru synthesis, we thought it could also improve our synthetic attempts with Fe. Following the same synthe tic protocols to go from FeCl 2 to Fe1 to Fe3 , we were able to synthesize Fe1* and Fe3* . As was observed for the synthesis of Fe3 , in the synthesis of Fe3* a second major product is also made. The product is also a dark red, paramagnetic complex, however, it is notably more crystalline than Fe2 , and X - ray quality single crystals were grown directly from the byproduct isolated from the crude reaction mixture. Thus Fe2* was identified as Fe(NHAr*) 2 dppe from isolation of the single crystals. Structures and sy nthetic schemes outlining the syntheses of these Fe analogues of the Ru complexes are shown below in Fig. 7.7. 569 Figure 7 . 7 (top) Synthesis of Fe analogues of Ru - imido complexes, Fe1 , Fe1* , Fe3 , and Fe3* . The side product, Fe2/2* , also results from this synthetic route, and is separated from Fe3/3* by several extractions and recrystallizations. (bottom) Crystal structures of Fe2* and Fe3* ; ellipsoids are shown at 50% probability with H atoms omitted for clarity. Note, Fe2* crystallizes in the C2/c spacegroup, with the crystallographic 2 - fold axis bisecting the N1 Fe1 N1 and P1 Fe1 P1 angles. Thus, half of the molecule is symmetry gen erated. 570 Table 7 . 3 Select bond distances and angles from the single crystal X - ray structures of Fe3* and Fe2* shown in Figure 7.7. Fe3* Fe2* Bond Angle ( ° ) Bond Distance ( Å ) Bond Angle ( ° ) Bond Distance ( Å ) N1 Fe1 P1 111.74 N1 Fe1 1.654 C1 N1 Fe1 138.00 N1 Fe1 1.928 N1 Fe1 P2 132.17 P1 Fe1 2.166 N1 Fe1 P1 105.21 P1 Fe1 2.446 N1 Fe1 P3 122.79 P2 Fe1 2.158 N1 Fe1 N1 135.81 - - C1 N1 Fe1 172.09 P3 Fe1 2.161 P1 Fe1 P1 83.21 - - With Fe3* isolated, oxidation via AgSbF 6 was performed, and while high - spin Fe(III) impurities are sometimes detected by EPR spectroscopy, its enhanced stability allowed for more thorough study of the electronic structure of Fe4* . The results of these electronic structure investigations are discussed in detail below (7.5 and 7.6). 7.4 Exploration of a Trischelating Phosphine Ligand Platform for Fe and Ru Imides As shown in the introduction to this chapter, trischelates are a particularly common ligand moti f for the synthesis and stabilization of terminal Fe - imides. We thought this type of ligand platform could lead to interesting electronic behavior by inducing more rigid C 3v symmetry on the imide complexes. Electronically similar phosphine ligands to thos e in Fe3 and Ru3 can be employed for the purpose of reducing differences in the metal - ligand interactions; this focuses analysis on the electronic structure effects due to the 3 - fold symmetric chelation. With Fe, Amrendra had previously shown that this is a straightforward and relatively stable complex to reach, Fe(NAr) t P 3 ( Fe5 , t P 3 = (Me 2 PCH 2 ) 3 Si(C(CH 3 ) 3 ) or triphos). However, his original synthesis of this complex (see experimental) was not reproducible, even in his own hands. I modified his original syn thesis of this complex, such that it is analogous to the synthesis of Fe3/3* , utilizing the chelate to trap in situ Fe1 . This synthesis has proven to be a reproducible and reliable method to yield Fe5 . With the trischelate, no high - spin Fe(II) bis(amide) s pecies were evident in the reaction mixture. 571 Oxidation of Fe5 with 1 equiv of AgSbF 6 results in a color change from dark purple to bright blue, accompanied by the precipitation of Ag ° . X - ray quality crystals of [Fe(NAr) t P 3 ][SbF 6 ] ( Fe6 ) can be grown from D ME layered with hexane at - 35 ° C. It is interesting to note that Fe5 shows almost no distortion of the N1 Fe1 P angles, which are close to equivalent. However, upon oxidation, Fe6 shows the smallest N1 Fe1 P angle (most contracted) noted for any of the str ucturally characterized Ru or Fe imide complexes that we have structurally characterized. This feature is highlighted in Fig. 7.8. Table 7 . 4 Select bond distances and angles for the single crystal X - ray structures of Fe5 and Fe6 . Note that Fe6 has two unique molecules in the asymmetric unit; several of the bond lengths and angles show statistical differences, so both measurements are shown. Images of thes e structures are shown in Fig. 7.8. Bond Angle/Distance ( ° / Å ) Fe5 C1 N1 Fe1 178.98 N1 Fe1 P1 121.19 N1 Fe1 P2 121.79 N1 Fe1 P3 124.47 N1 Fe1 1.667 P1 Fe1 2.136 P2 Fe1 2.144 P3 Fe1 2.148 Fe6 C1 N1 Fe1 167.82/172.51 N1 Fe1 P1 101.46/105.56 N1 Fe1 P2 127.96/130.78 N1 Fe1 P3 133.53/129.47 N1 Fe1 1.653/1.643 P1 Fe1 2.192/2.195 P2 Fe1 2.232/2.232 P3 Fe1 2.235/2.229 P2 Fe1 P3 94.36/93.93 Angles between Fe1, N1, P2, P3 355.85/354.18 572 While several attempts have been made by both Amrendra and myself to replicate a triphos - imido complex with Ru, successful synthesis has never been achieved. These results never Figure 7 . 8 Synthesis of Fe5 and Fe6 by trapping the unstable Fe1 with the trischelating (PMe 2 CH 2 ) 3 Si t Bu ligand. X - ray crystal structures of Fe5 and Fe6 are shown for comparison; ellipsoids are shown at 50% probability with H atoms and counteranion ( Fe6 ) omitted for clarity. 573 resulted in the clean synthesis of a monomeric, terminal imido complex. Generally, addition of the triphos ligand to a Ru complex (i.e. Ru1 or cis - RuCl 2 (PMe 3 ) 4 ) results in dozens of new 31 P signals in the NMR of the crude reaction solution without a clear major product. I suspect this is due to a when fully coord inated to a metal. This could lead to incomplete chelation between a single t P 3 ligand and a single metal center, giving rise to multiple oligomeric species. 7.5 Characterization of Cationic Ru5*, Fe4*, and Fe6 by EPR Spectroscopy (7.5) 11 Figure 7 . 9 M(III) cationic complexes examined by EPR spectroscopy. Fe6 is reasonably crystalline and has been characterized by single crystal X - ray crystallography. Ru4* and Fe4* are amorphous. The become oily when exposed to polar ethereal solvents, and attempts at crystallization generally produce powder solids. EPR spect roscopy is an incredibly useful experimental technique that provides details about the electronic structure of a paramagnetic complex by examining the behavior of the unpaired electron(s). We thought this could provide invaluable insight in characterizing the radical cations Fe4* , Fe6* , and Ru5* , which are all S = ½ systems. While trying to synthesize and study these compounds, delocalization of the radical was noted in the Ru5 system (see above, Fig. 7.4 and 7.5 ). The Fe complexes were also unstable, howev er, they did not exhibit the same type of radical reactivity as the Ru analogue; in the sense of their radical character, there seemed to be a substantial difference due to the identity of the metal and the basal ligand set. Going into these 11 EPR data was collected and interpreted by Professor John McCracken. He has also provided figures representing the experimental data and simulated spectra for the studies discussed in this section. 574 experiments, w e suspected that, in the Ru systems, substantial delocalization of the radical was - system, while the Fe systems were likely simple Fe(III) complexes with metal - localized radicals. To test this hypothesis, radical complexe s were examined by X - band cw - EPR spectroscopy at several temperatures and with several preparations of each compound. These studies proved challenging due to the instability of the species of interest. Even with the addition of a para - isopropyl substituent Fe4* and Ru5* ), these complexes decompose within a few days stored at - 35 ° C as solids. This, coupled with their lack of crystallinity, precluded effective purification. Consequently, some variation was note d from spectrum to spectrum among the different samples prepared for Fe4* and Ru5* . Even small differences, such as the purity of the Ag - salt used in the oxidation or the particular batch of PMe 3 used to generate the starting material, appear to affect the character of Fe4* and Ru5* after oxidation. By contrast, Fe6 , which is crystalline and could be reproducibly purified following the oxidation of Fe5 , provided very consistent EPR spectra from sample to sample. The spectrum of Fe4* originally looked like a composite of low - spin Fe(III) and low - spin Fe(III) - coupled ligand radical. However, the spectrum proved to be simpler than we initially thought. The species of interest in Fig. 7.1 0 a i s a bona - fide low - spin Fe(III), S = ½ paramagnet. By studying addition al spectra from several preparations of Fe4* , we were able to determine what was really happening in these samples. Even when the spectrum appears to be a single species ( Fig. 7.1 0 a), a very small impurity of a high - spin Fe(III) - coupled contaminant with p aired ligand radical (Fig. 7.1 0 , inset) skews the fit of the first derivative of the absorption spectrum. This contaminant seems to be an impurity, or possibly even a decomposition product from the preparation of Fe4* . Since the synthesis of Fe3* 575 is accompanied by the formation of at lea st one other Fe species (high spin Fe(II), Fe2 * , Fig. 7.7 ), co - produced in this one - pot - three - step synthesis, either decomposition of Fe3*/Fe4* or contamination from the oxidation product of Fe2* seem like possible sources for this impurity. The sample tha t provided the EPR spectrum highlighted in the inset of Fig. 7.10 came from an oxidation of Fe3* to Fe4* in which the reaction solution quickly went from the dark green color characteristic of Fe3* and Fe4* to a dark brownish - grey upon standing. Thus, no F e4* appeared to be present in the sample The radical contaminant appears to be primarily organic, with a g value of 1.982 and 14 N hyperfine coupling (A iso = 13.3 MHz; A dip = 80.4 MHz), consistent with a - based radical. 42 It is important to note that this g - value is lower than what is expected for an isolated organic radical (2.0023), thus the paramagnetic character of this species appears to be affected by coupling with a metal ion (Fe(II/III)). Fitti ng this impurity, whether a side product or decomposition product, accurately and by itself, clarified the nature of subsequent spectra by facilitating improved accuracy of species separation in the spectral simulations. Consequently, the spectra for the t wo Fe4* samples shown in Fig. 7.1 0 a and 7.1 0 b could be fitted with roughly the same results, despite outward differences. 576 Figure 7 . 10 EPR spectra (black) of 2 different preparations of Fe4* (a and b), utilizing identical synthetic preparations. The insert shows a mixed Fe(III) radical species that seems to form as an impurity (or decomposition product) upon oxidation of Fe3* to Fe4*. The spectrum shown in a is relatively pure, while b shows the Fe(III) mixed radical impurity superimposed on the spectrum of Fe4*. Red traces represent simulated spectra. For more simulation details, see Experimental. Spectra were recorded at 10 K. In both s amples of Fe4* , similar g - values were calculated which agree with the assignment of Fe4* as a low spin Fe(III) - centered paramagnetic species. The g - values were determined as 2.49, 2.10, and 1.96 and 2.47, 2.10, and 1.97, respectively, for the samples in Fi g . 7.1 0 a and 7.1 0 b. The major difference between the two spectra is that, in 10 b, there is roughly 18% impurity of the 577 ligand radical species superimposed on the Fe4* spectrum. The radical impurity was fitted with g = 1.982 and 14 N hyperfine couplings of A iso = 6.4 MHz and A dip = 78.8 MHz; these parameters agree quite well with those found in the sample of the radical impurity. These studies of the Fe(III) terminal imide - species, Fe4* suggest that the oxidation of the complex results in unpaired electron lo calization on the metal center. The ligands remain intact, and while the species is unstable, the instability does not likely manifest itself in the type of reactivity exhibited by Ru5 to produce Ru4 . In contrast to Fe4* , Fe6 , which is easily recrystallized, does not present impurities in the EPR spectrum and has consistently given a composite spectrum with both metal - and ligand - based radical signal. In both spectra in Fig. 7.1 1 a and 7.1 1 b, there are two sets of features whic h show low - spin Fe(III) character and organic - imido based radical character with 14 N hyperfine coupling with a g - value substantially shifted from an isolated organic radical. The spectrum of Fe6 , overall covers a narrower magnetic field than Fe4* . The Fe(I II) paramagnetic center was assigned g - values of 2.35, 1.99, and 1.97 in Fig. 7.11a , and 2.36, 2.00, and 1.99 in 10b. Respectively, these Fe(III) - centered contributions to the spectra in the two different samples are 71% and 75% of the total signal observ ed. The imido - centered radical in Fig. 7.11b provided a g - value of 2.014 and 14 N hyperfine couplings of A iso = 8.4 MHz and A dip =76.5 MHz. In the spectrum, the 14 N coupling is not resolved due to large linewidths, which is markedly different from the organ ic - ligand radical 578 signals observed in the Fe4* system (or Ru5* , below). Also note that the spectrum in Fig. 11a was recorded with Fe4* in toluene while the spectrum in Fig. 7.11b was in 2 - methylTHF. While the formality of fitting the EPR spectra might define the spectra for Fe6 shown in Fig. 7.1 1 a and 7.1 1 b most properly as two different species with separate S = ½ paramagnetic Figure 7 . 11 EPR spectra of two different preparations of Fe6 (black) and their simulated spectra (red), utilizing identical synthetic methods. Both samples show very similar characteristics, with about 75% Fe(III) - centered character and 25% ligand - centered radical with 14N hyperfine coupling. Two different lineshapes are noted in toluene (top) and 2 - methylTHF (bottom), but the radical character has very similar properties in both spectra. Spectra were recorded at 10K 579 spectra are recorded for the molecular species frozen in a solid matrix of 2 - methylTHF or toluene, and thus the populations observed should be proportional to the solution state populations in terms of where the radical character is delocalized in the molecule. Another way to describe this is that are two electronic ground state resonance contributors of the Fe6 molecule. 580 Figure 7 . 12 EPR spectra of two different samples of Ru5* (black) and their simulated spectra (red), utilizing identical synthetic methods. The samples are treated as a composite of two distinct paramagnetic centers one centered on Ru(III) and one centered on the imide fragment. The two spectra show dramatically different proportions of each radical center in the sample. Spectrum (a) is about 90% Ru(III) and 10% ligand radical, whereas spectrum (b) shows about 30% Ru(III) character and 70% ligand radical. These spectra were taken from different batches of Ru5* , prepared following the same synthetic procedures. The relative distribution of radical character seems to be affected by synthesis and sample preparation of the compounds. The param agnetic behavior of Fe6 is similar to that exhibited by Ru5* , however, the relative contributions of the ligand - centered and Ru(III) - centered paramagnets vary substantially from sample to sample. The spectra shown in Fig. 7.1 2 a and 7.1 2 b are, again, from separate preparations of Ru5* . The two spectra show the same two paramagnetic centers giving rise to the 581 same sets of signals, but in different proportions. The spectra in Fig. 7.1 2 a is composed of about 90% Ru(III) signal with g - values o f 2.06, 2.01, and 1.97, with the remaining 10% originating from a radical contributor with g = 2.007 and 14 N hyperfine coupling of A iso = 4.0 MHz and A dip = 80.6 MHz. By contrast, the spectra in 11b is only about 30% Ru(III) - centered where g = 2.05, 2.04, and 1.96 and 70% ligand - based radical with g = 2.0042 (A iso = 15.5 MHz and A dip = 56.6 MHz). The ligand - - value is closer to free electron g - value than those noted in Fe6 and Fe4* , demonstrating different orbital angular momentum coupling w ith the metal. While few literature examples of EPR characterization for low - spin Ru(III) are available, the values determined for Ru(III) are similar to literature reports of Ru(III) paramagnetic centers in zeolites. 43 The results of these EPR analyses are summarized in Fig. 7.1 3 , showi ng the relative distribution of radical character observed in each species. With these analyses in hand, we had additional experimental support for the induction of ligand - radical character in the oxidized Ru5* (and Ru5 ) species. We had also observed that with the same ligand set on Fe (P 3 = PMe 3 /dppe), the radical character stays localized on the metal. This suggests a fundamental difference in the electronic structure in these two systems based on the identity of the metal. Alternatively, with the more ri gid and roughly C 3 v symmetric triphos ligand system (P 3 = (Me 3 Si)C(CH 2 PMe 2 ) 3 ), some ligand - centered radical character is induced in the Fe(III) system. This suggests that changing the ligand set can induce similar behavior between Fe and Ru(III) metal cent ers toward their ligands, restoring their electronic structure similarities between the two metals. The question still remains, what is the most accurate way to describe radical character on the metal versus the ligand in these systems? In both the Ru5* and Fe6 samples, the ligand - centered paramagnetic centers contribute a substantial proportion of the paramagnetic character observed in these compounds. Is the delocalization a reversible delocalization that represents 582 proportional population of a ground a nd excited state that lie close in energy? If this is the case, the two paramagnetic centers may still be most aptly discussed as a single species, in which EPR spectroscopy of a frozen solution allows us to view the population ratio of the two distinct el ectronic states when the solution was prepared. Is the phenomenon of radical delocalization between the metal and aryl imido ligand an equilibrium process? Alternatively, is delocalization of the radical in the aryl imido system an irreversible process, resulting in different proportions in different samples that have otherwise indistinguishable characteristics? For example, if the unpaired electron density is shifted to the imide ligand, a geometric distortion of the complex could then pose a barrier to reverse this ligand, external factors like temperature and the age of the sample may result in different populations of the metal - vs. ligand - centered radi cal states arising from differences in the handling of the complex. Regardless of the formalism with which we label the behavior of these complexes, the trend was observed that in these species, Ru has the ability to delocalize a great deal of radical cha racter onto the aryl imide ligand (up to 70% was observed). By comparison, the Fe analogue to Ru5* , Fe4* , localizes the radical on the metal. On the other hand, in Fe6 , the unpaired electron is, again, consistently delocalized across both the metal and the imide ligand. This characterization by EPR supports the reactivity observed with Ru, which led to the isolation of Ru4 . Although the same reactivity was not observed in the Fe systems, these experiments also demonstrate that substantial radical delocalization also occurs in Fe6 . Thus both the metal and the basal ligand set play significant roles in the electronic behavior exhibited by unpaired electron density in these complexes. 583 Figure 7 . 13 The figure shows the radical localization in the Fe4* , Ru5* , and Fe6 cations. Note that with Fe4* , th e cationic species is most accurately described by a single paramagnetic center. However, the Ru5* and Fe6 spectra could only be successfully modeled as a composite of two paramagnetic species, where both a metal - and ligand - centered radical contribute to the entire spectrum. 7.6 Computational Analysis Comparing Fe and Ru Analogues 12 The crystallographic and EPR data for complexes Fe4* , Fe6 , and Ru5* clearly indicate a wide variance in the amount of radical localization versus delocalization observed in the M(I II) cationic imide radicals discussed above. Moller - Plesset second - order perturbation theory (MP2) and multistate complete active space second - order perturbation theory (CASPT2) 44,45 were applied to investigate the observed differences between the iron and ruthenium complexes. For 12 Calculations were performed by Profe ssor Ben Levine and his former graduate student, Dr. B. Scott Fales. The results presented here are a summary of their findings based on experimental work in the Odom group. 584 computational convenience, truncated versions of Fe1 (Fe(NAr)(PMe 3 ) 3 ) and Ru1 (Ru(NAr)(PMe 3 ) 3 ) were approximated with model compounds Fe7 and Ru6 : Ph N=M(PH 3 ) 3 . Computational details are presented in the Experimental section, along with the optimized coordinates. Table 7 . 5 Data from the X - ray crystal structures of several Fe and Ru - imide compounds for comparison to the computed optimized structures of Fe7 and Ru6 . Compound Bond (Å)/ Angle (°) Ru1 Ru1* Ru3 Fe3 Fe3* Fe5 Fe6(1) b Fe6(2) b M - N1 1.811(2) 1.817(4) 1.808(6) 1.657(4) 1.653(9) 1.667(3) 1.643(4) 1.653(4) M - P1 2.224(1) 2.224(1) 2.240(2) 2.156(2) 2.166(4) 2.136(1) 2.232(2) 2.232(2) M - P2 2.239(1) 2.240(2) 2.240(2) 2.155(2) 2.157(3) 2.144(1) 2.229(2) 2.235(2) M - P3 2.254(1) 2.253(1) 2.275(2) 2.175(2) 2.162(3) 2.148(1) 2.195(2) 2.192(2) N1 - C1 1.372(4) 1.364(6) 1.388(9) 1.381(5) 1.387(13) 1.372(5) 1.386(6) 1.382(6) M - N1 - C1 174.9(2) 179.9(3) 166.5(5) 171.0(3) 172.1(8) 178.9(4) 172.5(4) 167.8(4) N1 - M - P1 113.69(8) 119.3(1) 106.1(2) 107.2(1) 111.7(3) 121.2(1) 105.6(2) 101.5(2) N1 - M - P2 122.63(8) 121.9(1) 127.7(2) 125.0(1) 122.8(3) 121.8(1) 129.4(2) 128.0(2) N1 - M - P3 128.40(8) 124.5(1) 136.4(2) 130.3(1) 132.2(3) 124.4(2) 130.8(2) 133.5(2) P1 - M - P2 96.47(3) 94.70(6) 97.38(8) 100.77(6) 96.4(1) 93.97(5) 93.93(6) 93.23(6) P1 - M - P3 95.17(3) 95.36(5) 99.20(8) 102.35(6) 102.0(1) 93.88(5) 91.95(6) 93.16(6) P2 - M - P3 93.27(3) 93.99(5) 82.07(8) 86.10(6) 85.4(1) 93.70(5) 93.05(6) 94.36(6) The experimentally determined structure of Ru3 (Ru(NAr)dppe(PMe 3 )) deviates farther from C 3v symmetry than that of Fe3 , as indicated by a difference in the range spanned by the three N1 M P angles (30.3 ° for Ru3 vs. 23.3 ° for Fe3 ), as shown in Table 7. 5 above . A similar trend is noted in the MP2 - optimized structures of Ru6 and Fe7 , which have N1 M P angles spanning 585 ranges of 9.5 ° and 3.4 ° , respectively ( Table 7. 6 ). The smaller ranges in these computed compounds relative to Ru3 and Fe3 can be attributed to the d ifference in the ligands. In both Ru3 and Fe3 the phosphine ligands are not identical, which likely encourages further deviation from C 3v symmetry. For example, consider the ranges observed in the X - ray structures of Ru1 and Fe5 , which are 14.7 ° and 3.2 ° r espectively, where three identical PR 3 ligands are bound to the metal; these values are much more similar to the ranges computed for Ru6 and Fe7 . Table 7 . 6 N1 - M - P angles (°) as optimized at the MP2 level of theory. The range spanned by each set of angles is also given. Compounds Angle Ru6 Fe7 N1 - M - P1 122.1 121.2 N1 - M - P2 122.3 121.6 N1 - M - P3 131.6 124.6 Range 9.5 3.4 As described in the introduction, it was previously proposed that the deviation of Ru1 from C 3v symmetry is due to the nature of the HOMO, which is composed of the Ru d z 2 orbital and a - interaction, a deformation of the complex, which shifts the imide ligand off the z - axis and breaks the C 3v symmetry of the molecule, stabilizes the HOMO of the complex. 46 By comparison, the optimized structure of Fe7 is much more symmetric. Thus, in this configuration, the HOMO (Figure 7.14, top - left ins et) is purely antibonding the aryl - imide fragment remains relatively aligned with the z - axis of the molecule. As the imido ligand moves farther from the C 3 axis (z - axis), as in Ru6 , the nitrogen atom shifts closer to the node of the metal d z 2 orbital. This generates a mixed HOMO with both bonding and antibonding character ( Figure 7.1 4 , top - center inset). The partial bonding character is represented by the blue lobe, marked by a green arrow. We can attribute the stronger distortion in Ru6 , rela tive to Fe7, to the stronger bonding/antibonding interactions associated with second - row transition 586 metals compared to first - row transition metals in the same group, consistent with increased overlap in second row metals (due to the primogenic effect). Figure 7 . 14 The orbital energies (defined as the negative ionization potential, as described in the text) of Fe7 , Ru6 , and Ru6_mod computed at the CASPT2 level of theory. Insets show the HOMO and HOMO - 1 orbitals (SONOs of the cations, as described in the text). A green arrow indicates the bonding lobe of the HOMO of Ru6 . These results, therefore, agree with conventional theories and experimental evidence, which supports the accuracy of the orbital description provided by these calculations. With the ground state structures for M(II) compounds established, the balance between M(III) and ligand - centered radical character in Ru6 and Fe7 were examined . Experimental EPR results (above) indicate that Fe4* is Fe(III) in character, while Ru5* demonstrates mixed M(III)/ligand radical character. CASPT2 calculations also suggest that Ru6 would more likely exhibit ligand - radical character than Fe7 ( Fig. 7.1 4 ) and provide a physical explanation for this trend. In Fe7 , the metal - centered ( d z 2 ) HOMO is 0.85 eV above the HOMO - 1 orbital, which has significant population on 587 the ligand. This large energy gap prevents mixing of the HOMO and HOMO - 1 in [ Fe7 ] + , resulting in an iron - centered radical. In contrast, the HOMO and HOMO - 1 of Ru6 are split by only 0.04 eV. Similar to the Fe7 case, in Ru6 , the HOMO - 1 orbital also has a large orbital contribution from the imide ligand. Unlike in Fe7 , the closeness of th e energies of the HOMO and HOMO - 1 will facilitate mixing of the two orbitals, leading directly to ligand radical character. These comparisons line up well with the experimental characterization of Ru5* and Fe4* , but the intermediate behavior of Fe6 remaine d unexplained. To investigate the relationship between the basal PR 3 ligand set and the HOMO and HOMO - 1 energies, a hybrid model complex was generated. Ru6_mod was developed by taking the optimized structure of Fe7 and replacing the central Fe atom with Ru . The bond distances were then altered by rigidly stretching the metal ligand bonds to match the optimized Ru6 bond lengths (ligand - metal - ligand bond angles and ligand - internal coordinates remain frozen at their Fe7 - optimized values). This preserved the co ordination geometry while preserving realistic orbital overlaps on changing the metal identity. The results of this exercise are consistent with other calculations. The closer the geometry of the complex to true C 3v symmetry, the bigger the energy gap is between the HOMO and the HOMO - 1. This energy gap between the HOMO and HOMO - 1 is 0.48 eV in the more symmetric Ru6_mod , versus 0.04 eV in Ru6 . The only structural difference between these two models is the ligand orientation, more or less C 3 symmetric. This energy difference is created by a destabilization - bonding HOMO - 1 interaction between Ru and N by 0.16 eV. Thus, the orientation of the P 3 - basal set seems to be crucia l in determining the localization of the unpaired electron upon oxidation of these complexes. What is perhaps most interesting about this finding is that in the Fe system where the P ligands themselves are symmetric, with triphos in Fe5/Fe6 , the oxidized complex demonstrates 588 radical character more consistent with distortion from C 3v symmetry. This is supported by the heavily distorted solid - state X - ray structure obtained for this complex. With the inherently less symmetric P 3 basal set of PMe 3 and dppe, in Fe3*/Fe4* , a higher degree of C 3 symmetry is demonstrated upon oxidation. This conclusion initially seems counterintuitive, as we would expect the triphos ligand to enforce more rigid C 3v symmetry. The symmetry of Fe5 is higher than that of Fe3 or Fe3* , for example. However, due to the specific orientations enforced by the chelator which in the Fe(II) complex maintains symmetric P coordination the ligand is less adaptable to electronic changes at the metal. When the Fe(II) is oxidized to Fe(III), no subtl e shifts in the positions of the P 3 ligands, to stabilize the complex upon oxidation, is possible. As a result, a dramatic distortion of the N1 Fe1 Px bond angles occurs. This results in a more symmetric ligand set producing a less symmetric radical cation , which facilitates ligand radical character. Such a result demonstrates a need for detailed electronic structure studies with M E multiple bond complexes, as results are often counter to predictions. Based on the small sample size of complexes in this stu dy, it also seems that generalizations about the electronic structures require a far larger sampling of complexes. As the small changes in ligand choice here led to dramatic changes in the character of the complexes examined, it is clear that no single fac tor dominates the electronic character of these complexes. 7.7 Reactivity Studies with Fe5 Due to the instability of Fe1 , and our inability to isolate it from the crude reaction mixture, which contains H 2 NAr/Ar*, LiCl, and Fe2/2* , reactivity studies similar t o those undertaken previously with Ru1 , could not be assessed with the exact Fe analogue. Fe1 and Fe5 are likely to possess similar electronic structure, given their 3 - fold symmetric tri(alkyl) phosphine basal sets, so reactivity studies were instead pursu ed with Fe5 . 589 If we consider a simple molecular orbital model for this system, applying idealized C 3v symmetry, the frontier orbitals are likely an e set of * orbitals representing antibonding interaction between the imide N and the Fe d xz and d yz orbita ls. The - bonding orbitals should be more heavily N - centered, making these antibonding orbitals more Fe - centered. This is consistent with the picture for Ru analogues, and in the Fe system, could be even more pronounced; this would be consistent with the periodic t rend of weaker orbital overlap among ligands and first row transition nucleophilic character to both the Fe and N in the metal - imide bond. This makes reactivity fro m either site possible, with specific Fe5 interactions with substrates influenced by both steric and kinetic factors. A series of reactants with an electrophilic center were combined with Fe5 to probe its reactivity and begin looking for trends. Carbon d isulfide (CS 2 ) is an interesting substrate for several reasons. Several plausible possibilities from reaction of this substrate with Fe5 can be imagined, from a [2+2] - cyclometallated product to a terminal sulfide species. Upon addition of 1 equivalent or a large excess (neat) of CS 2 , dark purple solutions of Fe5 turn dark red and a maroon colored, microcrystalline solid precipitates. Analysis of the crude solution by GCMS and 14 N NMR appears to show production of 2,6 - diisopropylthioisocyanate. We propose th at the minimally soluble, maroon colored microcrystalline solid is a dimeric or oligomeric Fe(II) triphos species with bridging sulfides. While we were unable to confirm this structurally, the byproduct characterization heavily supports this stoichiometry. It also suggests that both the imide N and the Fe center are directly involved in the reactivity of Fe5 with CS 2 . This could occur either via a concerted [2+2] cycloaddition process, followed by elimination of SCNAr or potentially by nucleophilic attack o f the imide N on the CS 2 C, followed by insertion of S into the N Fe bond. 590 This result appears consistent with the results published by Deng, et. al. for the reactivity noted between CS 2 and their Co(NAr)(PMe 3 ) 3 complex. They report that addition of CS 2 w ith their Co(NAr)(PMe 3 ) 3 as a [2+2] cycloaddition. Upon exposure of the metallocyclic compound to wet hexane, SCNAr is liberated in 90% yield. 47 The accompanying Co - S complex, however, was not isolated or structurally characterized. It would be interesting to determine if Fe5 can perform a similar reaction with CO 2 , or even a carbodiimide, which could demonstrate metathesis with the Fe N imide fragment. Figure 7 . 15 Addition of CS 2 to the terminal Fe(II) imide, Fe5 , results in the production of 2,6 - diisopropylphenylthioisocyanate and an insoluble red species proposed to be [Fe( t P 3 - S)] 2 . Another interesting substrate with which Fe5 reacts is benzaldehyde. Initially, the reaction was performed by combining 1 equivalent of benzaldehyde with 1 equiv of Fe5 . The reaction solution went from dark purple to bluish green upon addition, and from the reaction solution, crystals of Fe5 were obt ained. The reaction was performed again, by combining 2 equivalent of benzaldehyde with 1 equivalent of Fe5 . This resulted in a high yield of the chiral, metallocyclic Fe complex, Fe8 , shown in Fig. 7.16 . This reactivity differs from that observed by Deng, et. al. for the Co(NAr)(PMe 3 ) 3 which simply adds one equivalent of benzaldehyde to the complex to form a 2 - amidate ligand on Co. 591 Based on the structural data for Fe8 , the 7 - membered metallocycle is best represented as drawn in Fig. 7.16. The bond lengt hs indicated double - bond character between N1 and C1, as well as between O1 and C1. The compound is diamagnetic, and lacks a counter ion of any kind, and the structural characteristics of O2 suggest it is a monoanionic ligand. This suggests that O1 is also formally an RO ¯ type ligand and that resonance exists with the adjacent imine nitrogen (N1). This assessment is also supported by the planar orientation of N1, C1 and O1, as well as the observation that this protonation state yields the lowest R1(%) upon refinement of the structure. Addition of H atoms to different portions of the structure reduces the quality of the crystallographic refinement parameters. Also, no clear evidence of a terminal Fe H is observed by 1 H NMR. This description of the ligand is c onsistent with a loss of H 2 and suggests that upon interaction of the electrophilic carbon in the benzaldehyde with the imide moiety, a highly reactive Fe species is generated, capable of inducing C C bond formation (coupling of the benzaldehydes). Figure 7 . 16 Reaction of 2 equivalents of benzaldehyde with Fe5 produces a metalacyclic species which is chiral, and appears to be diamagnetic, low spin Fe(II). Based on bond lengths and angles in the crystal structure, the two oxygens coordinated to Fe appear anionic, with N1 best described as an imine. This means th at loss of H 2 has occurred during the reaction. The crystal structure ( right - hexane in the lattice were omitted for clarity. (Fe1 O1 = 1.985 Å, Fe1 O2 = 1.858 Å, O1 C1 = 1.291 Å, C1 N1 = 1.297 Å, O2 C9 = 1.416 Å). One additional type of reaction examined with Fe5 is imide group exchange with protic H 2 NR species. This process is slow but does appear to occur when an excess of H 2 NPh is added 592 (i.e. 10 equiv. of H 2 NPh gives about 50% conver sion of Fe5 over the course of 5 days estimated by NMR integrations). However, isolation of the converted product from a solution containing a huge excess of H 2 NR was not productive. The imide group has been observed to exchange with 1,1, - dimethylhydrazine at a similar rate in solution, but again, isolation of the resulting hydrazido complex was not achieved from the reaction solution. Specifically, with the hydrazine exchange, it appears to be an equilibrium process, as complete conversion is not noted aft er extended periods of time or with a large excess of H 2 NNMe 2 ; this, coupled with the relative lability of H 2 NNMe 2 and H 2 NAr, may prevent separation of the new hydrazido species from the crude reaction solution. However, these reactions demonstrate that th ese pathways could provide interesting opportunities to generate and study new complexes in solution. Lastly, we also discovered that 10 mol% Fe5 , when combined with 1 equivalent of phenylacetylene catalyzes the dimerization of the alkyne into an enyne c ompound. However, alkyne dimerization and polymerization are not new or particularly difficult reactions to perform, so further optimization or expansion of this reaction was not pursued. 7.8 Conclusions Examination of a series of Ru(II) and Fe(II) imido c omplexes has illustrated important geometric and electronic structure differences that result from changes in the metal and phosphine ligand set. In these Ru(NAr)P 3 and Fe(NAr)P 3 complexes, Ru appears to produce more bonding character in the HOMO and resul ts in the geometric distortion of the N1 M P bond angles. With Fe, the HOMO appears to be more strictly antibonding in nature which prevents the distortion observed with Ru. The basic properties of the metal in these analogues triggers a chain of electroni c differences that lead directly to differences in reactivity. 593 These differences are exacerbated upon oxidation, as highlighted by the behavior of the complexes, Ru5* , Fe4* , and Fe6 . The distortion observed in the Ru structures facilitates mixing of the HOMO and HOMO - 1 orbitals, which provides a means of radical deloc alization across the metal center and the arylimide ligand. By contrast, in the Fe system ( Fe4* ), the HOMO and HOMO - 1 have more discrepant energies, and therefore, upon oxidation, unpaired electron density is isolated on the Fe. Intermediate to these two c ases is Fe6 , which shows some amount of ligand - radical character, but not to the extent that was observed with Ru ( Ru5* ). In these examples, both changes in the metal (i.e. going from Fe to Ru), and changes that restrict the flexibility of the P 3 basal set elicit a similar distortion of the N1 M P angles upon oxidation. Through this structural effect, the HOMO and HOMO - 1 energies are shifted and mixing between the two orbitals can occur. In both Fe6 and Ru5* , this results in partial oxidation of the aryl im ide ligand and delocalized radical character. While the resultant electronic structures are similar, the initial cause of the structural distortion in both molecules is very different, and was not something that we predicted when synthetic efforts began. These results highlight the need for more fundamental studies that probe the electronic structure of M=E complexes, where the character and subsequent reactivity associated with the M=E - and * - orbitals are directly impacted by the properties of the met al changing. Only through systematic studies such as this will understanding of these types of bonds in transition metal complexes develop to the point of providing predictive guidance in designing and synthesizing new complexes. 7.9 Experimental General Con siderations . All manipulations were carried out under inert N 2 atmosphere, either in an Mbraun glovebox or under standard Schlenk techniques. The solvents acetonitrile, 594 toluene, dimethoxyethane (DME), pentane, and diethyl ether were sparged with nitrogen and passed over an activated alumina column prior to use. The s olvents benzene, tetrahydrofuran, and n - hexane were dried over sodium - benzophenone ketal radical, refluxed, and distilled under nitrogen prior to use. All deuterated NMR solvents were purchased from Cambridge Isotope Labs. Benzene - d 6 was dried over CaH 2 an d distilled under N 2 . The solvents CDCl 3 and CD 2 Cl 2 were dried over P 2 O 5 and distilled under N 2 . Tetrahydrofuran - d 8 and 2 - methyltetrahydrofuran (used in the EPR experiments) were dried over Na and distilled under nitrogen. All solvents were stored over 3 Å molecular sieves in a glovebox after purification. The triphos - ligand ( t P 3 ) was prepared according to literature procedures. 41,48 Trimethylphosphine was purchased from Strem Chemicals, Inc. and used as received. Anhydrous FeCl 2 was purchased from Sigma - Aldrich, Inc. and used as received. 2,6 - Diisopropylaniline was purchased from Sigma - Aldrich, Inc., distilled under N 2 from CaH 2 , and was stored in the glovebox after purification. 2,4,6 - Triisopropylaniline was prepared following literature procedures 49 and dried by azeotropic removal of water in a Dean - Stark apparatus using benzene. FeC l 2 (dppe) was prepared as described in the literature. 50 AgSbF 6 was purchased form Sigma - Aldrich, Inc. and used as received. (Note, the AgSbF 6 was dissolved in THF, filtered over Celite, and preci pitated with n - hexane prior to use if it was substantially darkened in color). LiNHAr and LiNHAr* were prepared by addition of 1 equiv of 2.5 M n BuLi in hexanes to a cold ( 78 °C) solution of the respective amine in hexane; after stirring for 2 h and warm ing to room temperature, the salts were collected by filtration, washed with hexane, and used without further purification. 595 CS 2 was dried over Na 2 SO 4 and distilled under N 2 prior to use. Benzaldehyde, phenyl acetylene, and 1,1 - dimethylhydrazine were disti lled under N 2 prior to use. 1,1 - dimethylhydrazine was stored over 3 Å molecular sieves after distillation. EPR spectroscopy EPR measurements were made on a Bruker E - 680X spectrometer at X - band using a 4122 SHQE - W1 resonator. Cryogenic sample temperatures were achieved using an Oxford ESR - 900 cryostat together with an ITC - 503 temperature controller. EPR data were simulated usin g EasySpin 5.2.11 running in the MATLAB 2017b environment. 51 Electronic structure calculations CASPT2 is an inherently many - body theory, making interpretation of its results in terms of intuitive concepts such as orbitals and orbital energies s omewhat ambiguous. Here, we define the HOMO of the neutrally - charged species as the singly - occupied natural orbital (SONO) of the ground state of the cation of the same species at the same geometry. The negative of the computed vertical ionization potentia l can be interpreted as the orbitals and orbital energies for lower occupied orbitals (HOMO - 1 corresponds to the SONO of the first excited state of the cation, and so on). Although one can imagine ambiguities arising in this analysis, in the present work all of the discussed orbitals and orbital energies were completely unambiguous. Neutral Ru6 and Fe7 were optimized at the MP2 level of theory. Vertical ionization p otentials are computed as the difference between the MP2 energies of the neutral and the CASPT2 energies of the cations at the neutral - optimized structures. An active space of 7 electrons in 4 orbitals and a state average over 4 states was used for CASPT2 calculations of Ru6 and Fe7 . All CASPT2 and MP2 calculations were performed using the cc - pVTZ basis for Fe, the cc - pVTZ - PP basis and effective core potentials for Ru, and the cc - pVDZ basis for all other atoms. 52 - 55 The 596 multistate variant of CASPT2 56 was used for all calculations. All calculations were performed with the MolPro software package. 57 - 61 Orbital pictures were created with VMD. 62 Figure 7 . 17 Results demonstrating the electronic basis of the distortion of Ru1 away from C 3v symmetry. 46 Instrumentation NMR Rogers NMR Facility. These include a UNITYplus 500 spectrometer equipped with a 5 mm switchable broadband probe operating at 36.12 MHz ( 14 N); a Varian Inova 500 spectrometer equipped with a 5mm pulse - field - gradient (PFG) switchable broadband probe operating at 499.84 MHz ( 1 H); a Varian Inova 600 spectrometer equipped with a 5 mm PFG switchable broadband probe operating at 599.89 MHz ( 1 H); and an Agilent DDR2 500 M Hz NMR spectrometers equipped with a 5 mm PFG switchable broadband probe operating at 499.84 MHz ( 1 H), 125.73 MHz ( 13 C), and 202.35 ( 31 P). 1 H NMR chemical shifts are reported relative to residual C 6 HD 5 in C 6 D 6 as 7.16 ppm. 13 C NMR chemical shifts are repor ted relative to 13 C 6 D 6 as 128.06 ppm. 14 N NMR shifts are referenced to the internal peak for dissolved N 2 in NMR solvent (309.6 ppm vs. external nitromethane as 0 ppm). 597 X - ray Crystallography All crystallographic data was collected at the Michigan State University Center for X - ray Crystallography. All structures were collected on Bruker AXS instruments operating with either copper or molybdenum radiation sources. Data was collected at 173 K. St ructure solutions were typically found using XT Intrinsic Phasing and refined by least squares using Olex software. For further information please see the .cif files provided as supporting information. Cyclic Voltammetry All electrochemical experiments (C Instruments Electrochemical Workstation. The standard conditions were to prepare a 5.0 mL solution of 0.2 M TBAPF 6 (387 mg) in THF with 2 mmol of the complex under investigation. The experiments were run in an N 2 atmosphere M Braun glovebox, using a 3 - electrode setup. This involved a Pt disc working electrode, Ag/AgNO 3 0.1M reference electrode in MeCN, and a Pt wire counter electrode. All compounds were internally referenced to the ferrocene/ferrocenium couple as 0 V. Reversibi lity of events was assigned by observing linear correlation in a plot of (current) 2 vs. scan rate for a given redox couple. Notably, the Ru(III) species seemed to react with electrolyte upon dissolution (evidenced by color change and lack of redox waves), thus adequate characterization by cyclic voltammetry could not be obtained. Uv - vis UV - Vis spectra were collected using an Ocean Optics DH - mini UV - Vis - NIR spectrophotometer in an N 2 glovebox. Experiments were performed in dry THF using a quartz cell. The raw data were fit with OriginPro 9.0 software to obtain accurate peak separation and assignment of maxima assuming gaussian peak shapes. 598 Improved synthesis of cis - RuCl 2 (PMe 3 ) 4 A 35 mL pressure tube was charged with Ru(COD)Cl 2 (1.00 g, 1 equiv), a stir bar, toluene (10 mL), and PMe 3 (1.60 g, 6 equiv). The tube was sealed inside the glovebox and transferred to a 110 °C oil bath. The solution was stirred for 12 h. Over the reaction time, the solution changed from opaque brown to transparent yellow. The pressure tube was removed from heat and transferred to the glovebox. The reaction solution was concentrated to about 2 mL in vacuo to yield large blocky yellow crystals of cis - RuCl 2 (PMe 3 ) 4. The remaining reaction solution was decant ed from the crystals, and chilled to yield additional product. NMR of the material match published spectra. 46 Yield: 1.4 g (83%). Synthesis of RuNAr(PMe 3 ) 3 (Ru1 ) Ru1 was prepared as previously reported, 46 using cis - RuCl 2 (PMe 3 ) 4 as prepared above. Elemental analysis, 31 P/ 1 H/ 13 C NMR, and the structure from X - ray diffraction were published previously. 46 UV - vis absorption (THF, 21 °C): 465 nm (3039 cm 1 M 1 ), 290 (5532 cm - 1 M - 1 ), 232 nm (14065 cm - 1 M - 1 ). 14 N NMR (benzene - d 6 , 36 MHz, 25 °C): 326.3 ppm. Synthesis of Ru(NAr)(dmpe) 2 (Ru2) A 20 mL scintillation vial was charged with Ru1 (106 mg, 0.2 mmol, 1 equiv), 5 mL THF, and a magnetic stir bar. To this red - orange solution was added a solution of dmpe (65 mg, 0.4 mmol, 2 equiv) in 2 mL THF, dropwise, at room temperature. The solution was stirred for 4 h, over which time the solution became brownish - yellow in color. The volatiles were removed in vacuo, and the residue was rinsed with several small aliquots of n - hexane. The solids were then dissolved in a minimum amount of THF and layered with n - hexane. The layered solution was stored a t 35 °C overnight to yield plate - like green - brown crystals of Ru2 (52 mg, 45%). NMR: 1 H NMR (500 MHz, benzene - d 6 J = 7.3 Hz, 2H), 6.51 (t, J = 7.3 Hz, 1H), 2.87 (septet, J = 7.0 Hz, 2H), 1.51 - 1.38 (m, 4H), 1.34 (dd, J = 10.6, 6.9 Hz, 12H), 1. 29 - 1.19 (m, 4H), 1.17 (d, J = 8.2 Hz, 6H), 1.10 (t, J = 3.2 Hz, 6H), 0.99 (t, J = 2.2 Hz, 6H), 0.82 599 (d, J = 6.3 Hz, 6H). 13 C NMR (126 MHz, THF - d 8 30.44, 30.10 (m), 27.78, 25.73, 24.81, 23.73 22.89 (m), 22.39 (d, J = 12.2 Hz), 21.06 20.21 (m), 18.97, 13.92. 31 P NMR (202 MHz, benzene - d 6 J = 14.7 Hz), 30.76 (t, J = 14.8 Hz). 14 N NMR (36 MHz, Synthesis of Ru(PMe 3 )(dppe)(NAr) (Ru3) To a stirred solution of Ru1 (134 mg, 0.398 mmol, 1 equiv) in THF was added a solution of dppe (164 mg, 0.402 mmol, 1 equiv) in 2 mL of THF. The reaction mixture was stirred for 1 h at room temperature. The volatiles were removed in vacuo, and the dark red solid residue was rinsed with cold hexane (3 x 1 mL). The volatiles were once again removed in vacuo, and the solids were dissolved in a minimal amount of toluene. The concentrated toluene solution was stored at 35 °C overnight to yield flaky, red - orange crystals of Ru3 (202 mg, 87%). M.p.: 110 °C (dec). NMR: 1 H NMR (500 MHz, benzene - d 6 ): 8.08 (t, J = 9.9 Hz, 4H), 7.58 7.51 (m, 4H), 7.13 (s, 3H), 7.12 6.81 (m, 12H), 4.90 (septet, J = 7.0 Hz, 2H), 2.02 1.89 (m, 2H), 1.85 1.71 (m, 2H), 1.41 (d, J = 6.9 Hz, 12H), 0.60 (d, J = 9.8 Hz, 9H). 13 C[ 1 H] NMR (126 MHz, benzene - d 6 ): 133.90 (t), 130.9 4 (t), 128.97(s), 128.14(s), 122.73(s), 122.31 (d), 120.40 (d), 118.51(s), 30.23 (t), 27.77(s), 26.67(s), 23.65(s), 22.20 (d), 21.97(s). 31 P NMR (202 MHz, benzene - d6): 99.31 (d, J = 21.3 Hz), 23.65 (t, J = 21.2 Hz). 14 N NMR (36 MHz, THF): 337.6 (s). UV - vis absorption (THF, 21 °C): 472 nm (3610 cm 1 M 1 ), 314 (6576 cm - 1 M - 1 ), 272 nm (16709 cm - 1 M - 1 ). 41 H 50 NP 3 Ru: C, 65.59; H, 6.71; N, 1.87. Found: C, 65.07; H, 6.80; N, 1.73. Synthesis of [Ru(dppe)(PMe 3 )(NAr)(NCCH 3 )] 2 [BAr F 24 ] 2 (Ru4) A solution of Ru3 was prepared (60 mg, 0.080 mmol) in 6 mL of a 1:1 (volume: volume) mixture of MeCN and DME. This red - orange solution was stirred at room temperature. A separate solution of AgSbF 6 (30 mg, 0.087 mmol) in 2 mL of MeCN/DME, was added dropwise to the solution of Ru3 . After addition, 600 the reaction was stirred for 24 h at room temperature, over which time the reaction solution turned bright purple. The reaction mixture was filtered using Celite as a filtering agent to remove Ag 0 , and the filtrate was concentrated in vacuo. The concentrated filtrate was layered with hexane and stored at 35 °C for 3 days to get dark purple crystals of Ru4 (62.1 mg, 75.6%). M.p.: 140 °C (dec). NMR: 1 H NMR (500 MHz, CD 2 Cl 2 ): 7.92 (s, 8H), 7.64 (t, J = 8.2 Hz, 8H), 7.38 (t, J = 7.4 Hz, 8H), 7.33 (dd, J = 8.2, 5.2 Hz, 12H), 7.27 7.20 (m, 12H), 6.93 (d, J = 7.4 Hz, 4H), 6.91 - 6.86 (m, 2H), 4.55 (septet, J = 6.8 Hz, 4H), 2.42 - 2.25 (m, 4H), 2.09 (m, 2H), 2.07 - 1.95 (m, 4H), 1.18 (d, J = 6.9 Hz, 24H), 0.67 (d, J = 9.9 Hz, 18H). 19 F NMR (470 MHz, CD 2 Cl 2 ): 62.88 (s). 31 P NMR (202 MHz, CD 2 Cl 2 ): 99.24 (d, J = 22.5 Hz), 24.43 (t, J = 22.5 Hz). (Note: 13 C and 14 N could not be obtained due to compound instability in solvents in which it was soluble enough to see NMR signal). Synthesis of Fe(NAr)(PMe 3 ) 3 (Fe1) A 20 mL scintillation vial was charged with FeCl 2 (50 mg, 0.394 mmol, 1 equiv), a magnetic stir bar, and 8 mL of THF. To this off - white suspension, trimethylphosphine (0.25 mL, 2.37 mmol, 6 equiv) was added at room temperature. The mix ture was stirred for 1 h, over which time the FeCl 2 dissolved, and the solution changed color from pink to pale aquamarine. After this color change, the solution was chilled ( 78 °C), while a separate solution of LiNHAr (152 mg, 0.827 mmol, 2.1 equiv) in 2 mL THF was prepared. The chilled iron - containing mixture was stirred and to it was added the LiNHAr solution dropwise. Upon addition, the solution turned orange. The solution was allowed to warm to room temperature, and stirring was continued for 18 h, at which point the solution had turned dark green. Attempts to isolate Fe1 led to decomposition, but the complex is stable in the reaction solution for a few days with a slight excess of PMe 3 present. 31 P NMR (127 MHz, THF, 20 °C): 38.37 (s). 14 N NMR (36 M Hz, THF, 20 °C): 312.1 (s). 601 Synthesis of Fe(dppe)(NHAr) 2 (Fe2) (Method A) A clean sample of X - ray quality single crystals could not be obtained from the reaction mixture to synthesize Fe3 . However, the presence of a red, paramagnetic impurity was noted i n this reaction. (Method B) A 20 mL scintillation vial was charged with FeCl 2 (dppe) (150 mg, 0.29 mmol, 1 equiv), 4 mL THF, and a stir bar. This mixture was chilled at 78 °C. Separately, a solution of LiNHAr (105 mg, 0.58 mmol, 2 equiv) was prepared in 2 mL of room temperature THF. The chilled suspension of FeCl 2 (dppe) was stirred, and the LiNHAr solution was added dropwise. Upon complete addition, the solution was opaque and red in color. The mixture was stirred for 4 h, warming to room temperature. The v olatiles were removed in vacuo, resulting in a dark red residue. This residue was extracted with diethyl ether and filtered using Celite as a filtering agent. The filtrate was concentrated in vacuo, and the concentrated solution was stored at 35 °C to yie ld large, red crystals of Fe2 (37 mg, 16%). eff (benzene - d 6 B . Synthesis of Fe(NAr)(PMe 3 )(dppe) (Fe3) An in situ generated solution of Fe1 (30 mg scale FeCl 2 ) was stirred at room temperature. To this mixture was added dppe (84 mg, 0.21 mmol, 1 equiv) as a solution in 2 mL of THF. The resulting mixture was stirred for 1 h at room temperature. The volatiles were then removed in vacuo, and the resulting dark brow n residue was rinsed with hexane (4 mL, discarded). The residue was then extracted with diethyl ether to give a bright green solution. This ether solution was filtered using Celite as a filtering agent, concentrated in vacuo, and stored at 35 °C overnight . This yielded flaky, dark green X - ray quality crystals of Fe3 (64 mg, 39%). M.p.: 134 - 135 °C. 1 H NMR (500 MHz, benzene - d 6 ): 8.05 (t, J = 8.7 Hz, 4H), 7.55 (t, J = 8.2 Hz, 4H), 7.34 (t, J = 7.6 Hz, 1H), 7.13 (t, J = 7.4 Hz, 4H), 7.10 7.00 (m, 9H), 6.97 ( t, 1H), 4.78 (septet, J = 6.7 Hz, 2H), 1.89 (m, 4H), 1.33 (d, J = 6.8 Hz, 12H), 0.65 (d, J = 8.2 Hz, 9H). 13 C[ 1 H] NMR (126 MHz, benzene - d 6 ): 156.53 (s), 141.58 (s), 140.18 (s), 133.84 (d), 130.81 (s), 602 128.72 (s), 120.94 (s), 120.46 (s), 35.13 (s), 27.00 (s), 23.90 (s), 23.56 (s), 22.32 (s), 21.73 (d). 31 P NMR (202 MHz, benzene - d 6 ): 115.16 (d, J = 6.5 Hz), 35.11 (t, J = 6.4 Hz). 14 N NMR (36 M Hz, THF): 325.6 (s). UV - vis absorption (THF, 21 °C): 591 nm (2009 cm 1 M 1 ), 388 (4471 cm - 1 M - 1 ), 298 nm (13460 cm - 1 M - 1 41 H 50 NP 3 Fe: C, 69.79; H, 7.14; N, 1.99. Found: C, 69.40; H, 6.77; N, 1.77. Synthesis of Fe(NAr) t P 3 (Fe5) An in situ solution of Fe1 (prepared using 100 mg FeCl 2 , 0.79 mmol, 1 equiv) was stirred at room temperature. To the mixture was added t P 3 (247 mg, 0.79 mmol, 1 equiv) as a solution in 2 mL THF. The reaction solution rapidly changed color from dark green to dark purple upon addition. The resulting solution was stirred for 1 h at room temperature, and the volatiles were removed in vacuo. The resulting black residue was extracted with hexane and filtered using Celite as a filtering agent until the extracts were colorless. The filtrate was then concentrated in vac uo to ~3 mL and stored at 35 °C overnight to yield blocky, dark purple X - ray quality crystals of Fe5 (142 mg, 34 %). M.p.: 196 °C (dec). 1 H NMR (500 MHz, benzene - d 6 ): 7.15 (dd, J = 12.4, 4.7 Hz, 1H), 6.97 (d, J = 7.5 Hz, 2H), 4.31 (septet, J = 6.9 Hz, 2H), 1.66 (m, 18H), 1.38 (d, J = 7.0 Hz, 12H), 0.67 (s, 9H), 0.27 0.24 (m, 6H). 13 C[ 1 H] NMR (126 MHz, benzene - d 6 ): 159.73, 139.51, 122.35, 119.98, 28.26 27.65 (m), 27.41, 26.43, 23.44, 16.76 (q, J = 5.9 Hz), 10.74. 31 P NMR (202 MHz, benzene - d 6 ): 47.38. 14 N NMR (36 MHz, THF): 315.2 (s). UV - vis absorption (THF, 21 °C): 632 nm (2809 cm 1 M 1 ), 549 (3327 cm - 1 M - 1 ), 383 nm (7015 cm - 1 M - 1 ). 25 H 50 FeNP 3 Si: C, 55.45; H, 9.31; N, 2.59. Found: C, 55.45; H, 9.54; N, 2.61. E 1/ 2, + ): (Fe +2/+3 ) 0.91 V (rev.). Synthesis of [Fe(NAr) t P 3 ]SbF 6 (Fe6) A 20 mL scintillation vial was charged with Fe5 (40 mg, 0.0727 mmol, 1 equiv), a magnetic stir bar, and 4 mL of DME. This solution was stirred at room temperature. Separately, a solution of AgSbF 6 (25 mg, 0.0727 mmol, 1 equiv) was prepared 603 in 2 mL DME, which was added dropwise to the stirred solution of Fe5 . The resulting mixture was stirred for 12 h at room temperature, during which time the reaction so lution went from dark purple to bright blue and solid Ag 0 precipitated. The mixture was filtered using Celite as a filtering agent, concentrated to ~3 mL total volume, and layered with 4 mL of hexane. This layered solution was stored at 35 °C overnight to yield small, blue X - ray quality crystals of Fe6 (43 mg, 76%). M.p.: 121 °C (dec). UV - vis absorption (THF, 21 °C): 624 nm (3704 cm 1 M 1 ), 353 (5463.6 cm - 1 M - 1 ), 289 nm (13031 cm - 1 M - 1 ). 25 H 50 F 6 FeNP 3 SbSi: C, 38.63; H, 6.48; N, 1.80. Found: C, 38.60; H, 7.03; N, 1.73. eff (THF - d 8 , B . E 1/2, Fc/Fc + ): (Fe +2/+3 ) 0.89 V (rev.). Synthesis of Ru(NAr*)(PMe 3 ) 3 (Ru1*) To an Erlenmeyer flask was added a stir bar, 200 mg of cis - RuCl 2 (PMe 3 ) 4 (0.410 mmol, 1 equiv), and 20 mL THF to give a pale - yellow solution. This solution was chilled to 78 °C and to this cold solution was added a room temperature solution of LiNHAr* (196 mg, 0.875 mmol, 2.1 equiv) in 3 mL THF dropwise. Upon complete addition , the solution had turned light orange. The reaction mixture was stirred overnight at room temperature to give a viscous dark red solution. The volatiles were removed in vacuo, and the residue was extracted with hexane until the filtrate was colorless. Thi s extract was filtered using Celite as a filtering agent and concentrated in vacuo. The concentrated solution was stored in the freezer for 2 - 3 d at 35 °C to yield blocky, red - orange X - ray quality crystals of Ru1* (162 mg, 70%). Note: Due to the NH 2 Ar* ge nerated upon imido production and the high solubility of Ru1* in aliphatic solvents, analytically pure compound was obtained by repeated recrystallization from hexamethyldisiloxane. This resulted in a substantially reduced yield of approximately 12%. CHN a nalysis was taken of these single crystals. For the purposes of utilizing the complex in further reactions, specifically in the subsequent reaction to make Ru3* , samples of such high purity were 604 not necessary, and a small amount of the 2,4,6 - triisopropylan iline was tolerable in samples of Ru1* . M.p.: 112.7 - 114 °C. NMR: 1 H NMR (500 MHz, benzene - d 6 ): 7.04 (s, 2H), 4.42 (septet, J = 7.0 Hz, 2H), 3.03 2.45 (septet, J = 6.9 Hz, 1H), 1.43 (d, J = 7.0 Hz, 12H), 1.33 (d, J = 7.0 Hz, 6H), 1.29 (m, 26H). 13 C[ 1 H] NM R (126 MHz, benzene - d 6 ): 139.50 (d), 138.95 (s), 138.32 (s), 132.03 (s), 120.46 (s), 119.62 (s), 34.89 (s), 34.14 (s), 27.97 (s), 26.59 (s), 26.32 25.40 (m), 24.50 (s), 24.20 (s), 23.22 (s), 22.33 (s). 31 P NMR (202 MHz, benzene - d 6 ): 19.47 (s). 14 N NMR (36.5 MHz, benzene - d 6 ): 32 8.0 (s). UV - vis absorption (THF, 21 °C): 459 nm (4808 cm 1 M 1 ), 338 (5642 cm - 1 M - 1 ), 270 nm (14016 cm - 1 M - 1 24 H 50 NP 3 Ru: C, 52.73; H, 9.22; N, 2.56. Found: C, 52.79; H, 8.70; N, 2.57. Synthesis of Ru(NAr*)(dmpe) 2 (Ru2*) This complex was synthesized following the same procedure as Ru2 . A 20 mL scintillation vial was charged with Ru1* (150 mg, 0.27 mmol, 1 equiv), 5 mL THF, and a magnetic stir bar. To this red - orange solution was added a solution of dmpe (85 mg, 0.5 4 mmol, 2 equiv) in 2 mL THF, dropwise, at room temperature. The solution was stirred for 4 h, over which time the solution became brownish - yellow in color. The volatiles were removed in vacuo, and the residue was rinsed with several small aliquots of n - he xane. The solids were then dissolved in a minimum amount of THF and layered with n - hexane. The layered solution was stored at 35 °C overnight to yield plate - like green - brown crystals of Ru2* (42 mg, 25%). 1 H NMR (500 MHz, Benzene - (s, 1H), 3.19 (septet, J = 7.1 Hz, 1H), 2.91 (septet, J = 7.1 Hz, 2H), 1.61 (d, J = 6.9 Hz, 6H), 1.38 (d, J = 6.8 Hz, 12H), 1.35 1.31 (m, 2H), 1.21 (m, 12H), 1.03 (s, 6H), 0.98 0.90 (m, 3H), 0.86 (d, J = 5.5 Hz, 6H), 0.84 (t, J = 1.6 Hz, 3H). 13 C[ 1 H] NMR (126 MHz, Benzene - d 6 J = 8.0 Hz), 34.26, 33.26 (d, J = 17.1 Hz), 30.28, 27.99, 26.56, 25.79 (d, J = 7.5 Hz), 23.92 (dd, J = 13.9, 7.5 Hz), 22.94 (d, J = 8.6 Hz), 22.36, 21.07 (t, J = 7.6 Hz), 19.27, 13.03 (t, J = 13.5 Hz). 31 P NMR 605 (202 MHz, benzene - d 6 J = 14.5 Hz), 30.24 (t, J = 14.7 Hz). 14 N NMR (36 MHz, benzene - d 6 27 H 55 RuNP 4 : C, 52.42; H, 8.96; N, 2.26. Found C, 53.24; H, 9.1 1; N, 2.27. HRMS(ESI + ) Synthesis of Ru(NAr*)(PMe 3 )(dppe) (Ru3*) This compound was prepared similarly to Ru3 starting with a reaction solution of Ru1* (0.30 mmol, 1 equiv) and adding dppe (120 mg, 0.30 mmol, 1 equiv). The crude reaction was stirred for ~1 h at room temperature, and then volatiles were removed in vacuo. The sticky red residue was washed with hexamethyldisiloxane (3 × 2 mL), and once agai n dried in vacuo. The remaining solids were extracted with Et 2 O and filtered using Celite as a filtering agent until the extracts were colorless. The Et 2 O solution was then concentrated and stored at 35 °C to yield flaky, red crystals of Ru3* (98 mg, 41%) . Unfortunately, these crystals demonstrate severe full - molecule disorder by X - ray diffraction, and an adequate solution for the data could not be found. M.p.: 188.2 - 189.6 °C. NMR: 1 H NMR (500 MHz, benzene - d 6 ): 8.18 (t, J = 7.9 Hz, 4H), 7.64 (m, 4H), 7.23 (s, 2H), 7.14 (dt, J = 19.7, 7.3 Hz, 8H), 7.09 6.96 (m, 4H), 5.01 (septet, J = 6.7 Hz, 2H), 2.97 (septet, J = 6.8 Hz, 1H), 2.04 (m, 2H), 1.87 (m, 2H), 1.53 (d, J = 6.5 Hz, 12H), 1.43 (d, J = 6.8 Hz, 6H), 0.70 (d, J = 9.6 Hz, 9H). 13 C[ 1 H] NMR (126 MHz, benzene - d 6 ): 133.96 (d), 132.74 (t), 130.95 (s), 128.91 (s), 128.67 128.21 (m), 128.08 (s), 127.94 (s), 120.47 (s), 120.16 (d), 34.98 (s), 34.13 (s), 30.29 (t), 27.98 (s), 26.79 (s), 24.50 (s), 24.28 (s), 24.12 (s), 23.73 ( s), 22.50 21.88 (m). 31 P NMR (202 MHz, benzene - d 6 ): 99.84 (d, J = 21.1 Hz, 2H), 23.02 (t, J = 21.2 Hz, 1H). 14 N NMR (36 MHz, THF): 337.2 (s ). UV - vis absorption (THF, 21 °C): 465 nm (3053 cm 1 M 1 ), 330 (8035 cm - 1 M - 1 ), 280 nm (10045 cm - 1 M - 1 ). Elemental 44 H 56 NP 3 Ru: C, 66.65; H, 7.12; N, 1.77. Found: C, 66.46; H, 7.04; N, 1.73. + ): E 1/2 (Ru +2/+3 ) = 0.58 V (rev.). 606 Synthesis of [Ru(NAr*)(PMe 3 )(dppe]SbF 6 (Ru5*) A 20 mL scintillation vial was charged with a stir bar, Ru2* (50 mg, 0.063 mmol, 1 equiv), and 4 mL DME. This red - orange solution was stirred at room temperature. Separately, a solution of AgSbF 6 (22 mg, 0.063 mmol, 1 equiv) was prepared in 2 mL of DME. The AgSbF 6 was then added dropwise to the solution of Ru3* wit h vigorous stirring. The solution gradually changed from bright red to pinkish - brown, and a precipitate formed on the sides of the vial. The reaction was stirred at room temperature for 24 h, and then filtered using Celite as a filtering agent to remove Ag 0 . The filtrate was concentrated in vacuo to give a brown oil. The oily residue was washed with hexane until the filtrate was colorless. The residue was once again dried in vacuo and was then dissolved in a minimum amount of DME. The DME solution was layer ed with hexane and stored at 35 °C overnight to yield a fine pinkish - brown powder (34 mg, 53%). The mother liquor was decanted, and the powder was dried in vacuo. The pink - brown powder was assigned as complex Ru5* as identified by EPR spectroscopy. M.p.: 147 ºC (color change 80 °C). eff (THF - d 8 B . UV - vis absorption (THF, 21 °C): 612 nm (912 cm 1 M 1 ), 475 (1297 cm - 1 M - 1 ), 353 nm (2937 cm - 1 M - 1 ), 280 nm (6575 cm - 1 M - 1 ). Note: Satisfactory elemental analysis was not obtained for Ru5* after se veral attempts, presumably due to its high sensitivity. Sample masses were noted to change rapidly when sample holders when taken out of the inert atmosphere glovebox and into air, despite our best attempts. Synthesis of Fe(NAr*)(PMe 3 ) 3 (Fe1*) This compoun d was prepared similarly to compound Fe1 using FeCl 2 (50 mg, 1 equiv), PMe 3 (180 mg, 6 equiv), and LiNHAr* (185 mg, 2.1 equiv). Similarly, it was found to decompose during attempted isolation and could not be separated from the reaction mixture, but the complex is stable for a few days in the reaction solution in the presence of excess PMe 3 . 31 P NMR (127 MHz, THF): 43.06 (s). 14 N NMR (36 MHz, THF): 320.6 (s). 607 Synthesis of Fe(dppe)(NHAr*) 2 (Fe2*) (Method A) Fe2* was recovered as a byproduct from the synthesis of Fe2* (vide supra). This was accomplished by extracting the remain ing solid residue, after pentane extraction, with diethyl ether, resulting in a red solution. This solution was filtered using Celite as a filtering agent, concentrated in vacuo to ~1 mL, and stored at 35 °C for 4 d to yield large, red X - ray quality cryst als of Fe2* (26 mg, 7%). (Method B) A 20 mL scintillation vial was charged with FeCl 2 (dppe) (100 mg, 0.19 mmol, 1 equiv), 4 mL THF, and a stir bar. This mixture was chilled at 78 °C. Separately, a solution of LiNHAr* (69 mg, 0.38 mmol, 2 equiv) was prepar ed in 2 mL of room temperature THF. The chilled suspension of FeCl 2 (dppe) was stirred, and the LiNHAr* solution was added dropwise. Upon complete addition, the solution was opaque with a pink color. The mixture was stirred for 4 h, warming to room temperat ure. The volatiles were removed in vacuo, resulting in a dark red residue. This residue was extracted with diethyl ether and filtered using Celite as a filtering agent. The filtrate was concentrated in vacuo, and the concentrated solution was stored at 35 °C to yield large, red crystals of Fe2* (47 mg, 28 %). M.p.: 144.4 - eff (benzene - d 6 B . Synthesis of Fe(NAr*)(PMe 3 )(dppe) (Fe3* ) This compound was prepared similarly to Fe3 . The reaction mixture of Fe1* (50 mg, 1 equiv) and dppe (145 mg, 0.95 equiv) was used. After the reaction volatiles were removed in vacuo, the crude residue was extracted with pentane and filtered using Celite as a filtering agent. The filtrate was concentrated in vacuo and stored at 35 °C to yield dark green crystals of Fe3* (200 mg, 66%). These crystals were not X - ray quality but were suitably pure for NMR and elemental analysis. X - ray quality crystals were grown from a very dilute solution in n - hexane at 35 °C. M.p.: 148.9 - 150.2 ° C. NMR: 1 H NMR (500 MHz, benzene - d 6 ): 8.02 (t, J = 8.6 Hz, 4H), 7.52 (t, J = 8.1 Hz, 4H), 7.09 (d, J = 7.3 Hz, 3H), 7.06 6.96 (m, 608 8H), 6.93 (t, J = 7.2 Hz, 2H), 4.76 (septet, J = 6.9 Hz, 2H), 2.81 2.68 (septet, J = 6.8 Hz, 1H), 1.84 (m, J = 36.7, 21.9, 1 3.8 Hz, 4H), 1.33 (d, J = 6.9 Hz, 12H), 1.27 (d, J = 6.8 Hz, 6H), 0.62 (d, J = 8.3 Hz, 9H). 13 C[ 1 H] NMR (126 MHz, benzene - d 6 ): 156.53 (s), 141.58 (s), 140.18 (s), 133.79 (s), 130.81 (s), 128.72 (s), 120.94 (s), 120.46 (s), 35.13 (s), 27.97 (s), 27.00 (s) , 24.49 (s), 23.90 (s), 23.56 (s), 22.32 (s), 21.73 (d). 31 P NMR (202 MHz, benzene - d 6 ): 115.35 (d, J = 5.2 Hz), 34.93 (t, J = 5.2 Hz). 14 N NMR (36 MHz, benzene - d 6 ): 324.8. UV - vis absorption (THF, 21 °C): 589 nm (1754 cm 1 M 1 ), 404 (4114 cm - 1 M - 1 ), 301 nm (11596 cm - 1 M - 1 ). Elemental Analysis 44 H 56 FeNP 3 : C, 70.68; H, 7.55; N, 1.87. Found C, 71.48; H, 7.57; N, 1.75. Cyclic + ): E 1/2 (Fe +2/+3 ) = 0.75 V (rev.). Synthesis of [Fe(NAr*)(PMe 3 )(dppe)]SbF 6 (Fe4*) A 20 mL scintillation vial was charged with Fe2* (50 mg, 0.067 mmol, 1 equiv), a magnetic stir bar, and 4 mL of DME. This solution was stirred at room temperature. Separately, a solution of AgSbF 6 (23 mg, 0.067 mmol, 1 equiv) was prepar ed in 2 mL DME. The silver solution was added dropwise to the solution of Fe3* . The resulting mixture was stirred for 12 h at room temperature, during which time the reaction solution remained dark green and solid Ag 0 precipitated. The mixture was filtered using Celite as a filtering agent, and the solvent was removed in vacuo. This yielded an oily green residue, which was rinsed with 5 mL pentane and dried in vacuo to obtain a powdery green solid of Fe4* (32 mg, 49%). M.p.: 110 °C (dec). UV - vis absorption (THF, 21 °C): 584 nm (1245 cm 1 M 1 ), 387 (2566 cm - 1 M - 1 ), 293 nm (5233 cm - 1 M - 1 ). eff (THF - d 8 : 1.97 B . = Fc/Fc + ): E 1/2 (Fe +2/+3 ) = 0.77 V (rev.). 609 NMR Spectra Figure 7 . 18 14 N NMR of Ru1 in C 6 D 6 . 610 Figure 7 . 19 1 H NMR of Ru2 in C 6 D 6 . 611 Figure 7 . 20 13 C NMR of Ru2 in d 8 - THF. NH 2 Ar dmpe THF - d 8 TMS 612 Figure 7 . 21 31 P NMR of Ru2 in d 8 - THF. 613 Figure 7 . 22 14 N NMR of Ru2 in C 6 D 6 . 614 Figure 7 . 23 1 H NMR of Ru4 in CD 2 Cl 2 . 615 Figure 7 . 24 31 P NMR of Ru4 in CD 2 Cl 2 . 616 Figure 7 . 25 19 F NMR of Ru4 in CD 2 Cl 2 . Note: This complex does not exhibit sufficient solubility in most NMR solvents. While soluble in CD 2 Cl 2 , it does slowly react with the solvent. This prevented accurate acquisition of 13 C and 14 N NMR data for this complex. 617 Figure 7 . 26 1 H NMR of Ru3 in C 6 D 6 . 618 Figure 7 . 27 13 C NMR of Ru3 in C 6 D 6 . 619 Figure 7 . 28 31 P NMR of Ru3 in C 6 D 6 . 620 Figure 7 . 29 14 N NMR of Ru3 in C 6 D 6 . 621 Figure 7 . 30 1 H NMR of Ru1* in C 6 D 6 . 622 Figure 7 . 31 13 C NMR of Ru1* in C 6 D 6 . 623 Figure 7 . 32 31 P NMR of Ru1* in C 6 D 6 . 624 Figure 7 . 33 14 N NMR of Ru1* in C 6 D 6 . 625 Figure 7 . 34 1 H NMR of Ru2* in C 6 D 6 . 626 Figure 7 . 35 13 C NMR of Ru2* in C 6 D 6 . 627 Figure 7 . 36 31 P NMR of Ru2* in C 6 D 6 . 628 Figure 7 . 37 14 N NMR of Ru2* in C 6 D 6 . 629 Figure 7 . 38 1 H NMR of Ru3* in C 6 D 6 . 630 Figure 7 . 39 13 C NMR of Ru3* in C 6 D 6 . 631 Figure 7 . 40 31 P NMR of Ru3* in C 6 D 6 . 632 Figure 7 . 41 14 N NMR of Ru3* in C 6 D 6 . 633 Figure 7 . 42 31 P NMR of Fe1 in THF ( in situ ). 634 Figure 7 . 43 14 N NMR of Fe1 in THF ( in situ ). 635 Figure 7 . 44 1 H NMR of Fe3 in C 6 D 6 . 636 Figure 7 . 45 13 C NMR of Fe3 in C 6 D 6 . 637 Figure 7 . 46 31 P NMR of Fe3 in C 6 D 6 . 638 Figure 7 . 47 14 N NMR of Fe3 in C 6 D 6 . 639 Figure 7 . 48 31 P NMR of Fe1* in C 6 D 6 . 640 Figure 7 . 49 14 N NMR of Fe1* in C 6 D 6 . 641 Figure 7 . 50 1 H NMR of Fe3* in C 6 D 6 . O = NH 2 Ar*, * = Et 2 O, ** = n - hexane 642 Figure 7 . 51 13 C NMR of Fe3* in C 6 D 6 . 643 Figure 7 . 52 31 P NMR of Fe3* in C 6 D 6 . 644 Figure 7 . 53 14 N NMR of Fe3* in C 6 D 6 . 645 Figure 7 . 54 1 H NMR of Fe8 in C 6 D 6 . Note: This is a crude NMR and the compound is chiral, so full assignment of the spectrum was not attempted. 646 Figure 7 . 55 31 P NMR of Fe8 in C 6 D 6 . 647 Figure 7 . 56 1 H NMR of the C 6 D 6 soluble extracts from the reaction of Fe5 and CS 2 . The 1 H NMR shows H 2 NAr and SCNAr. 648 Figure 7 . 57 31 P NMR of Fe5 + 1,1 - dimethylhydrazine in C 6 D 6 . The peak at 47 ppm is the starting Fe5 . The peak at 52 ppm is a new compound. ( - 49 is free triphos ligand) 649 UV - Vis Characterization Figure 7 . 58 0.000109 M, THF 0 2000 4000 6000 8000 10000 12000 12000 17000 22000 27000 32000 (M cm) Wavenumber (cm) Extinction Coefficient vs. Wavenumber Ru1 650 Figure 7 . 59 0.00022 M THF 0 2000 4000 6000 8000 10000 12000 12000 17000 22000 27000 32000 (M - 1 cm - 1 ) Wavenumber (cm - 1 ) Extinction Coefficient Vs. Wavenumber Ru3 651 Figure 7 . 60 0.00031 M, THF 0 5000 10000 15000 20000 25000 30000 35000 40000 12000 17000 22000 27000 32000 37000 (M - 1 cm - 1 ) Wavenumber (cm - 1 ) Extinction Coefficient Vs. Wavenumber Fe3 652 Figure 7 . 61 0.00031 M, THF 0 1000 2000 3000 4000 5000 6000 7000 8000 12000 17000 22000 27000 32000 (M - 1 cm - 1 ) Wavenumber (cm - 1 ) Extinction Coefficient Vs. Wavelength Ru1* (Vis) 653 Figure 7 . 62 0.00010 M, THF 0 2000 4000 6000 8000 10000 12000 14000 16000 13000 18000 23000 28000 (M - 1 cm - 1 ) Wavenumber (cm - 1 ) Extinction Coefficient Vs. Wavenumber Ru3* 654 Figure 7 . 63 0.00044 M, THF 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 13000 18000 23000 28000 Ru5 Wavenumber (cm - 1 ) Extinction Coefficient Vs. Wavenumber Ru5* 655 Figure 7 . 64 0.000068 M, THF 0 2000 4000 6000 8000 10000 12000 14000 15000 20000 25000 30000 (M - 1 cm - 1 ) Wavenumber (cm - 1 ) Extinction Coefficient Vs. Wavenumber Fe3* 656 Figure 7 . 65 0.000071 M, THF Note: additional features in the UV range of the spectrum for Fe3* and Fe4* could not be 0 2000 4000 6000 8000 10000 12000 14000 15000 20000 25000 30000 (M - 1 cm - 1 ) Wavenumber (cm - 1 ) Extinction Coefficient Vs. Wavenumber Fe4* 657 Figure 7 . 66 0.000238 M, THF 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 12000 17000 22000 27000 32000 (M - 1 cm - 1 ) Wavenumber (cm - 1 ) Extinction Coefficient Vs. Wavenumber Fe5 658 Figure 7 . 67 0.00016 M, THF 0 5000 10000 15000 20000 25000 12500 17500 22500 27500 Fe6 Wavenumber (cm - 1 ) Extinction Coefficient Vs. Wavenumber Fe6 659 X - Ray Crystallography The following complexes have been structurally characterized by single crystal X - ray diffraction - 1856961 ( Ru2 , Ru2* , Ru3 , Ru3* , Ru4 , Fe2 , Fe2* , Fe3 , Fe3* , Fe5 , and Fe6 ). One additional crystal structure was collected for structure elucidation: Fe8 . This strucure has not been submitted to the CCDC, but the .cif file has been added to the MSU structural database managed by Dr. Richard Staples. Basic structural data, including the unit cell and diffraction data for the structure, is provided below. 660 Xray - data for Fe5: spacegroup, Pbca Figure 7 . 68 Crystal data and structure refinement for Pbca. Identification code Pbca Empirical formula C 30 H 49.33 Fe 0.67 N 0.67 O 1.33 P 2 Si 0.67 Formula weight 621.42 Temperature/K 173.15 Crystal system orthorhombic Space group Pbca 661 a/Å 18.909(2) b/Å 17.097(2) c/Å 29.757(3) 90 90 90 Volume/Å 3 9620.1(19) Z 12 calc g/cm 3 1.287 - 1 0.682 F(000) 3864.0 Crystal size/mm 3 0.364 × 0.276 × 0.262 Radiation 3.484 to 50.744 Index ranges - - - Reflections collected 61350 Independent reflections 8820 [R int = 0.1333, R sigma = 0.0843] Data/restraints/parameters 8820/0/493 Goodness - of - fit on F 2 0.923 R 1 = 0.0569, wR 2 = 0.1400 Final R indexes [all data] R 1 = 0.0910, wR 2 = 0.1592 Largest diff. peak/hole / e Å - 3 0.75/ - 0.32 662 Electrochemical Characterization of M(NAr)(PMe 3 )dppe Complexes and Oxidized Species Fe5 presents a reversible Fe(II/III) couple which appears around 900 mV versus the Fc/Fc + couple. Additionally, there is also a reversible couple, assigned as the Fe(III/IV) redox wave centered at 160 mV. Upon scanning to more highly reducing potentials, a small irreversible anodic wave is noted, which causes the shoulder located on the catho dic wave of the Fe(II/III) couple to grow upon reversal of the current. This is likely an irreversible Fe(I/II) couple. This species is perhaps related to the small redox event centered at roughly 450 mV. Upon oxidation of Fe5 to Fe6 a slight shift is no ted in the Fe(II/III) redox wave, however, this process is still electrochemically reversible. The Fe(III/IV) couple, however, has shifted to much higher potentials and is irreversible when starting from the chemically oxidized species, Fe6 . This suggests that the electronics of Fe6 (in the +3 state) may be different from the electrochemically oxidized Fe5 + . Relative to Fe5 and Fe6, the Fe3* and Fe4* couples are shifted to slightly less reducing potentials. The reversible Fe(II/III) couple is 750 mV relati ve to Fc/Fc + . Further oxidation (Fe(III/IV)) is irreversible, with the onset of a large cathodic wave around 0 V. The oxidized species shows very similar features, with a reversible Fe(II/III) couple at 770 mV vs. Fc and a large irreversible anodic wave, with an onset slightly before 0 V (Fe(III/IV)). This suggests that the species generated by chemical oxidation is similar electronically to that generated by electrochemical oxidation. These species display potentials within agreeable ranges to similar der ivatives in the literature. 1 The ruthenium derivative ( Ru3* ) demonstrates two reversible couples. The Ru(II/III) couple, is assigned as 580 mV vs Fc. It is interesting to note that this potential is more oxidative than either of the Fe(II/III) redox couples. The Ru(II) is actually more difficult to oxidize than t he 663 lighter congers. This difference in the redox potentials between Ru and Fe could be interpreted as a resistance of Ru(II) to oxidize to Ru(III), or more specifically, that the electron removed from the Ru species originates from an orbital of different make - up than the electrons removed when oxidizing the Fe derivatives. This observation agrees with the experimentally observed ability of It might be expected in simple ligand fields, that the heavi er congeners are easier to oxidize relative to lighter ones. For instance, the redox potential to go from Ru(II) to Ru(III) in an aqueous system is reported as 0.24 V (vs. SCE), while the Fe(II) to Fe(III) couple is reported as 0.77 V. 2 Upon introducing mo re complicated ligands, however, this trend has been noticeably disrupted. For instance, the series of M(bpy) 3 +2/+3 couples for M = Fe, Ru, and Os are reported (vs. SHE) as 1.06 V, 1.27 V, and 0.84 V, respectively. 2 This same trend is noted comparing hexac yano complexes and their reduction potentials looking down Group VIII, as well. With these traditionally - accepting ligands, it seems Ru often proves harder to oxidize than Fe. Comparing the Ru(III/IV) redox potential vs. the Fe(III/IV) potentials, we see that it is intermediate to the two Fe species. While still more positive in potential than the reversible couple for Fe5 , it is slightly more negative than the onset of the irreversible couple noted for Fe3* . There is another wave, that appears reversible , but with about half the current density for the Ru(II/III) and (III/IV) couples following the Ru(III/IV) couple. This wave has not been assigned, but could take place on the ligands. Additionally, Ru5* appears to react with the electrolyte, thus electroc hemical characterization could not be compiled for comparison. While the rest of the species examined were stable enough over the time frame necessary to compile the CV measurements, the solutions often changed color after 12 to 24 hours stored at 35 °C. 664 Figure 7 . 69 CV for Fe3* in THF with TBAPF 6 electrolyte . -1.20E-04 -1.00E-04 -8.00E-05 -6.00E-05 -4.00E-05 -2.00E-05 0.00E+00 2.00E-05 4.00E-05 -1.25 -0.75 -0.25 0.25 0.75 Current (A) Potential (V Vs. Fc/Fc+) CV Fe(dppe)PMe 3 NAr* Fe3* 665 Figure 7 . 70 CV for Fe4* in THF with TBAPF6 electrolyte. -2.20E-05 -1.70E-05 -1.20E-05 -7.00E-06 -2.00E-06 3.00E-06 -1.15 -0.65 -0.15 0.35 Current (A) Potential (V vs. Fc/Fc+) CV [Fe(dppe)PMe 3 NAr]SbF 6 Fe4* 666 Figure 7 . 71 CV for Fe5 in THF with TBAPF6 electrolyte. -6.00E-05 -5.00E-05 -4.00E-05 -3.00E-05 -2.00E-05 -1.00E-05 0.00E+00 1.00E-05 2.00E-05 -1.3 -0.8 -0.3 0.2 Current (A) Potential (V Vs. Fc/Fc+) Cyclic Voltamogram Fe t P 3 NAr Fe5 667 Figure 7 . 72 CV for Fe6 in THF with TBAPF6 electrolyte. -1.00E-04 -8.00E-05 -6.00E-05 -4.00E-05 -2.00E-05 0.00E+00 2.00E-05 4.00E-05 -1.2 -0.7 -0.2 0.3 0.8 Fe5 Potential (V Vs. Fc/Fc+) CV [Fe t P 3 NAr]SbF 6 Fe6 668 Figure 7 . 73 CV for Ru3* in THF with TBAPF6 electrolyte. -8.00E-05 -6.00E-05 -4.00E-05 -2.00E-05 0.00E+00 2.00E-05 4.00E-05 -1.3 -0.8 -0.3 0.2 0.7 Current (A) Potential (V vs. Fc/Fc+) CV Ru(dppe)PMe 3 NAr* Ru3* 669 Optimized Geometries for Calculated Model Complexes RuNPh(PH 3 ) 3 (Ru6) 25 MP2/CC - PVDZ,RU=CC - PVTZ - PP ENERGY= - 1407.58451778 C 0.1333360925 3.3187269145 7.6940896461 C 1.0945904142 3.4428915742 6.6664653368 C 1.2155783789 4.6492344203 5.9420744321 C 0.3737222832 5.7287664983 6.2480304854 C - 0.5841795137 5.6051420981 7.2717808070 C - 0.7046040282 4.4027803033 7.9939087184 N 1.9287940576 2.3700077093 6.3622262914 Ru 2.8658620169 0.9040458595 6.1562339740 P 4.0165815330 0.0118280810 7.8041101187 P 4.52 27374948 0.3953028798 4.7667986457 P 1.9324430492 - 1.0347768015 5.6992149233 H 4.7166128800 - 1.2375641985 7.6776872573 H 5.1051203810 0.7125249397 8.4113773163 H 3.3619280325 - 0.3251187642 9.0290989892 H 5.1551494186 - 0.8966007301 4.7459173907 H 4.3321299963 0.4717225459 3.3524722093 H 5.7432078795 1.1389309804 4.7856710839 H 2.6867185834 - 2.2547449440 5.5991363692 H 0.9311228973 - 1.5671271113 6.5691037801 H 1.1939593112 - 1.2351195462 4.4913987396 H 1.9699166487 4.7153560325 5.1511410708 H 0.4635209516 6.665938 1805 5.6893309016 H - 1.4513206566 4.3117603633 8.7893463376 H 0.0648367387 2.3698921329 8.2366226796 H - 1.2399348407 6.4495085821 7.5079874961 FeNPh(PH 3 ) 3 (Fe7) 25 MP2/CC - PVDZ,FE=CC - PVTZ ENERGY= - 2576.09296549 N 0.1744675944 3.2025124108 7.7043270825 Fe 1.2474484024 3.2243995144 6.6244549027 P 1.6084432519 4.7961179621 5.4847817695 P 1.228820216 6 1.9780924698 5.1041275069 P 3.1493705724 2.9327928327 7.0287976816 H 4.1640559345 2.9487028288 6.0027379655 H 3.8833613192 3.7901276860 7.9200903533 H 3.6218627878 1.7158018645 7.6308413911 H 2.6297277134 4.7888878525 4.4654107503 H 0.6024863291 5.3634395462 4.6289181868 H 2.0121438641 6.0607124595 6.0362951427 H 2.2479088317 1.99817 56449 4.0825386548 H 1.2848233809 0.5531834685 5.2865755059 670 H 0.1381534655 1.9275529229 4.1681458281 C - 0.8147099140 3.1630425515 8.7073630985 C - 1.6018977523 4.3046711429 8.9242841282 C - 2.5869252091 4.2651739181 9.9240898017 C - 2.7615073764 3.0918982270 10.6800917287 C - 1.9659809420 1.9540748546 10.4529620705 C - 0.9790043131 1.9863211240 9.4548 126361 H - 1.4315494176 5.1957838372 8.3127131456 H - 3.2131148749 5.1418319659 10.1122799995 H - 3.5299972676 3.0635961108 11.4589723303 H - 2.1137969697 1.0492010329 11.0491336039 H - 0.3378296272 1.1250547715 9.2442527352 RuNPh(PH 3 ) 3 (Ru6_mod) 25 N 0.7392 - 0.0191 0.0013 Ru - 1.0127 - 0.0024 0.0003 P - 2.2513 - 1.8465 0.0066 P - 2.1744 0.9312 - 1.6167 P - 2.1788 0. 9451 1.6070 H - 3.6219 0.9622 1.6013 H - 2.0811 0.5202 2.9774 H - 2.0542 2.3437 1.9151 H - 3. 6906 - 1.7518 0.0455 H - 2.2327 - 2.8063 - 1.0633 H - 2.1778 - 2.8365 1.0463 H - 3.6160 0.9147 - 1.5539 H - 2.0949 2.3296 - 1.9405 H - 2.1207 0.4966 - 2.9865 C 2.1485 - 0.0126 0.0007 C 2.8291 - 1.2397 0.0022 C 4.2332 - 1.2327 0.0016 C 4.9208 - 0.0056 - 0.0005 C 4.2269 1.2181 - 0.0020 C 2.8231 1.2180 - 0.0007 H 2.2537 - 2.1703 0.0039 H 4.7866 - 2.1760 0.0028 H 6.0154 - 0.002 7 - 0.0009 H 4.7753 2.1643 - 0.0041 H 2.2432 2.1459 - 0.0009 671 REFERENCES 672 REFERENCES (1) Cundari, T. R. Transition metal imido complexes. J. Am. Chem. Soc. 1992, 114 (20), 7879. (2) Berry, J. F. TERMINAL NITRIDO AND IMIDO COMPLEXES OF THE LATE TRANSITION METALS. Comments on Inorganic Chemistry 2009, 30 (1 - 2), 28. (3) .; Ando, N. S.; Schultz, A. J.; Koetzle, T. F.; Piccoli, P. M. B.; Hedman, B.; Hodgson, K. O.; Bobyr, E.; Kirk, M. L.et al. Revisiting the Polyoxometalate - Based Late - Transition - Metal - Inorg. Chem. 2012, 51 (13), 7025. ( 4) Hay - Motherwell, R. S.; Wilkinson, G.; Hussain - Bates, B.; Hursthouse, M. B. Synthesis and X - ray crystal structure of oxotrimesityliridium(V). Polyhedron 1993, 12 (16), 2009. (5) Wang, B.; Lee, Y. - M.; Tcho, W. - Y.; Tussupbayev, S.; Kim, S. - T.; Kim, Y.; Seo, M. S.; Cho, K. - B.; Dede, Y.; Keegan, B. C.et al. Synthesis and reactivity of a mononuclear non - haem cobalt(IV) - oxo complex. Nature Communications 2017, 8 , 14839. (6) Hong, S.; Pf aff, F. F.; Kwon, E.; Wang, Y.; Seo, M. - S.; Bill, E.; Ray, K.; Nam, W. Spectroscopic Capture and Reactivity of a Low - Spin Cobalt(IV) - Oxo Complex Stabilized by Binding Redox - Inactive Metal Ions. 2014, 53 (39), 10403. (7) Goetz, M. K.; Hill, E. A.; Filatov, A. S.; Anderson, J. S. Isolation of a Terminal Co(III) - Oxo Complex. Journal of the American Chemical Society 2018, 140 (41), 13176. (8) Nugent, W. A., Mayer, J. M. Metal - Ligand Multiple Bonds. Wiley Interscience 1987 . (9) Peters, J. W.; Szilagyi, R. K. Exp loring new frontiers of nitrogenase structure and mechanism. Current Opinion in Chemical Biology 2006, 10 (2), 101. (10) Svastits, E. W.; Dawson, J. H.; Breslow, R.; Gellman, S. H. Functionalized nitrogen atom transfer catalyzed by cytochrome P - 450. Journa l of the American Chemical Society 1985, 107 (22), 6427. (11) Singh, R.; Bordeaux, M.; Fasan, R. P450 - Catalyzed Intramolecular sp3 C H Amination with Arylsulfonyl Azide Substrates. ACS Catalysis 2014, 4 (2), 546. (12) Hoffman, B. M.; Lukoyanov, D.; Yang, Z . - Y.; Dean, D. R.; Seefeldt, L. C. Mechanism of Nitrogen Fixation by Nitrogenase: The Next Stage. Chemical Reviews 2014, 114 (8), 4041. (13) Barney, B. M.; Lee, H. - I.; Dos Santos, P. C.; Hoffman, B. M.; Dean, D. R.; Seefeldt, L. C. Breaking the N2 triple b ond: insights into the nitrogenase mechanism. Dalton Transactions 2006 , DOI:10.1039/B517633F 10.1039/B517633F(19), 2277. 673 (14) Hohenberger, J.; Ray, K.; Meyer, K. The biology and chemistry of high - valent iron oxo and iron nitrido complexes. Nature Communica tions 2012, 3 , 720. (15) Modak, J. M. Haber process for ammonia synthesis. Resonance 2002, 7 (9), 69. (16) In Ullmann's Encyclopedia of Industrial Chemistry , DOI:doi:10.1002/14356007.a02_143.pub2 doi:10.1002/14356007.a02_143.pub2. (17) Mehn, M. P.; Peters, J. C. Mid - to high - valent imido and nitrido complexes of iron. J. Inorg. Biochem. 2006, 100 (4), 634. (18) Moret, M. - E.; Peters, J. C. Terminal Iron Dinitrogen and Iron Imide Complexes Supported by a Tris(phosphino)borane Ligand. Angew . Chem. Int. Ed. 2011, 50 (9), 2063. (19) Thomas, C. M.; Mankad, N. P.; Peters, J. C. Characterization of the Terminal Iron(IV) NAd}+. J. Am. Chem. Soc. 2006, 128 (15), 4956. (20) Iovan, D. A.; Betley, T. A. Characterization of Iron - Imido Species Relevant for N - Group Transfer Chemistry. J. Am. Chem. Soc. 2016, 138 (6), 1983. (21) - Spin Iron Imido Complexes. J. Am. Chem. Soc. 2011, 133 (13), 4917. (2 2) Cowley, R. E.; DeYonker, N. J.; Eckert, N. A.; Cundari, T. R.; DeBeer, S.; Bill, E.; Ottenwaelder, X.; Flaschenriem, C.; Holland, P. L. Three - Coordinate Terminal Imidoiron(III) Complexes: Structure, Spectroscopy, and Mechanism of Formation. Inorg. Chem. 2010, 49 (13), 6172. (23) Wang, L.; Hu, L.; Zhang, H.; Chen, H.; Deng, L. Three - Coordinate Iron(IV) Bisimido Complexes with Aminocarbene Ligation: Synthesis, Structure, and Reactivity. J. Am. Chem. Soc. 2015, 137 (44), 14196. (24) Ni, C.; Fettinger, J. C. ; Long, G. J.; Brynda, M.; Power, P. P. Reaction of a sterically encumbered iron(i) aryl/arene with organoazides: formation of an iron(v) bis(imide). Chem. Commun. 2008 , DOI:10.1039/B810941A 10.1039/B810941A(45), 6045. (25) J.T. Wilding, M.; A. Iovan, D.; A. Betley, T. High - Spin Iron Imido Complexes Competent for C - H Bond Amination , 2017. (26) Brown, S. D.; Peters, J. C. Ground - State Singlet L3Fe - - N) - FeL3 and L3Fe(NR) Complexes Featuring Pseudotetrahedral Fe(II) Centers. J. Am. Chem. Soc. 2005, 127 (6), 1913. (27) Mehn, M. P.; Brown, S. D.; Paine, T. K.; Brennessel, W. W.; Cramer, C. J.; Peter s, J. C.; Que, J. L. High - spin and low - spin iron(ii) complexes with facially - coordinated borohydride ligands. Dalton Transactions 2006 , DOI:10.1039/B509580H 10.1039/B509580H(10), 1347. 674 (28) Bucinsky, L.; Breza, M.; Lee, W. - T.; Hickey, A. K.; Dickie, D. A.; Nieto, I.; DeGayner, J. A.; Harris, T. D.; Meyer, K.; Krzystek, J.et al. Spectroscopic and Computational Studies of Spin States of Iron(IV) Nitrido and Imido Complexes. Inorg. Chem. 2017, 56 (8), 4751. (29) Schofield, M. H.; Kee, T. P.; Anhaus, J. T.; Schrock, R. R.; Johnson, K. H.; Davis, W. M. Osmium imido complexes: synthesis, reactivity, and SCF - X.alpha. - SW electronic structure. Inorg. Chem. 1991, 30 (19), 3595. (30) Anhaus, J. T.; Kee, T. P.; Schofield, M. H.; Schrock, R. R. Planar "20 - electron" osmium imido complexes. A linear imido ligand does not necessarily donate its lone pair of electrons to the metal. J. Am. Chem. Soc. 1990, 112 (4), 1642. (31) Wolf, J. R.; Bazan, G. C.; Schrock, R. R. Exchange of oxo ligands in osmium tetroxide with imido ligands in bis(arylimido)bis(tert - butoxo)molybdenum complexes, Mo(NAr)2(O - tert - Bu)2. A facile route to Os(NAr)2O2 and Os(NAr)3O and osmium(IV) complexes of the type Os(NAr)2L2 (NAr = N - 2,6 - C6H3 - iso - Pr2; L = a phos phine). Inorg. Chem. 1993, 32 (19), 4155. (32) Herranz, E.; Sharpless, K. B. Improvements in the osmium - catalyzed oxyamination of olefins by chloramine - T. The Journal of Organic Chemistry 1978, 43 (12), 2544. (33) Patrick, D. W.; Truesdale, L. K.; Biller, S. A.; Sharpless, K. B. Stereospecific vicinal oxyamination of olefins by alkylimidoosmium compounds. The Journal of Organic Chemistry 1978, 43 (13), 2628. (34) Kee, T. P.; Park, L. Y.; Robbins, J.; Schrock, R. R. Synthesis of the ruthenium imido complexes , [Ru([small eta] - C6H6)(N - 2,6 - R2C6H3)]2(R = Pri or Me), and the crystal structure of [Ru([small eta] - C6H6)(N - 2,6 - Pri2C6H3)]2. J. Chem. Soc., Chem. Commun. 1991 , DOI:10.1039/C39910000121 10.1039/C39910000121(2), 121. (35) Burred, A. K.; Steedman, A. J. ([sm all eta]6 - p - Cymene)Ru[triple bond, length half m - dash]N(2,4,6 - tri - tert - butylphenyl): a monomeric ruthenium(II) complex containing a terminal imido ligand. J. Chem. Soc., Chem. Commun. 1995 , DOI:10.1039/C39950002109 10.1039/C39950002109(20), 2109. (36) Burr ell, A. K.; Steedman, A. J. Synthesis, Structure, and Reactivity of Ruthenium(II) Terminal Imido Complexes. Organometallics 1997, 16 (6), 1203. (37) Singh, A. K.; Levine, B. G.; Staples, R. J.; Odom, A. L. A 4 - coordinate Ru(II) imido: unusual geometry, syn thesis, and reactivity. Chem Commun (Camb) 2013, 49 (92), 10799. (38) Bakhoda, A.; Jiang, Q.; Bertke, J. A.; Cundari, T. R.; Warren, T. H. Elusive Terminal Copper Arylnitrene Intermediates. Angew. Chem. Int. Ed. 2017, 56 (23), 6426. (39) King, E. R.; Henne ssy, E. T.; Betley, T. A. Catalytic C - H bond amination from high - spin iron imido complexes. J. Am. Chem. Soc. 2011, 133 (13), 4917. (40) Liu, J. - Y.; Zheng, Y.; Li, Y. - G.; Pan, L.; Li, Y. - S.; Hu, N. - H. Fe(II) and Co(II) pyridinebisimine complexes bearing di fferent substituents on ortho - and para - position of imines: 675 synthesis, characterization and behavior of ethylene polymerization. J. Organomet. Chem. 2005, 690 (5), 1233. (41) Karsch, H. H.; Appelt, A. Functional trimethylphosphane derivatives. 18. Methyl(phosphinomethyl)silanes and methyl(phosphinomethyl)stannanes. Z. Naturforsch., B: Chem. Sci. 1983, 38 (11), 1399. (42) Henning, J. C. M. 14N Hyperfine Structure in ESR Spectra of Heterocyclic Anions. 1966, 44 (5), 2139. (43) Oliver, S. W.; Smith, T. D.; Pilbrow, J. R.; Harvey, T. G.; Matheson, T. W.; Pratt, K. C. An ESR study of zeolite - supported ruthenium hydrodenitrogenation catalyst. Inorganica Chimica Acta 1986, 117 (1), L9. (44) Andersson, K.; Malmqvist, P. A.; Roos, B. O. 2nd - order Perturbation - Theory with a Complete Active Space Self - Consistent Field Reference Function. J. Chem. Phys. 1992, 96 , 1218. (45) Celani, P.; Werner, H. - J. Multireference perturbation theory for lar ge restricted and selected active space reference wave functions. J. Chem. Phys. 2000, 112 , 5546. (46) Singh, A. K.; Levine, B. G.; Staples, R. J.; Odom, A. L. A 4 - coordinate Ru(II) imido: unusual geometry, synthesis, and reactivity. Chem. Commun. 2013, 49 (92), 10799. (47) Liu, Y.; Du, J.; Deng, L. Synthesis, Structure, and Reactivity of Low - Spin Cobalt(II) Imido Complexes [(Me3P)3Co(NAr)]. Inorg. Chem. 2017, 56 (14), 8278. (48) Thoreson, K. A.; Follett, A. D.; McNeill, K. Synthesis and Characterization of Pentaphosphino Zero - Valent Iron Complexes and Their Corresponding Iron(II) - Chloride and - Hydride Complexes. Inorganic Chemistry 2010, 49 (8), 3942. (49) Liu, J. Y.; Zheng, Y.; Li, Y. G.; Pan, L.; Li, Y. S.; Hu, N. H. Fe(II) and Co(II) pyridinebisimine com plexes bearing different substituents on ortho - and para - position of imines: synthesis, characterization and behavior of ethylene polymerization. Journal of Organometallic Chemistry 2005, 690 (5), 1233. (50) Evans, D. J.; Hitchcock, P. B.; Leigh, G. J.; Ni cholson, B. K.; Niedwieski, A. C.; Nunes, F. S.; Soares, J. F. The synthesis of triangulo - trimetal complexes containing both iron(II) and vanadium(II). Inorg. Chim. Acta 2001, 319 (1 - 2), 147. (51) Stoll, S.; Schweiger, A. EasySpin, a comprehensive software package for spectral simulation and analysis in EPR. Journal of Magnetic Resonance 2006, 178 (1), 42. (52) Dunning, T. H. Gaussian - basis sets for use in correlated molecular calculations. 1. The atoms boron through neon and hydrogen. J. Chem. Phys. 1989, 90 (2), 1007. (53) Woon, D. E.; Dunning, T. H. Gaussian - basis sets for use in correlated molecular calculations. 3. The atoms aluminum through argon. J. Chem. Phys. 1993, 98 (2), 1358. 676 (54) Balabanov, N. B.; Peterson, K. A. Systematically convergent basis sets for transition metals. I. All - electron correlation consistent basis sets for the 3d elements Sc - Zn. J. Chem. Phys. 2005, 123 (6), 15. (55) Peterson, K. A.; Figgen, D.; Dolg, M.; Stoll, H. Energy - consistent relativistic pseudopotentials and correlation consistent basis sets for the 4d elements Y - Pd. J. Chem. Phys. 2007, 126 (12), 12. (56) Finley, J.; Malmqvist, P. A.; Roos, B. O.; Serrano - Andres, L. The multi - state CASPT2 method. Chem. Ph ys. Lett. 1998, 288 (2 - 4), 299. (57) Celani, P.; Werner, H. J. Multireference perturbation theory for large restricted and selected active space reference wave functions. J. Chem. Phys. 2000, 112 (13), 5546. (58) Knowles, P. J.; Werner, H. J. An efficient 2nd - order MC SCF method for long configuration expansions. Chem. Phys. Lett. 1985, 115 (3), 259. (59) Werner, H. - J.; Knowles, P. J. A second order multiconfiguration SCF procedure with optimum convergence. J. Chem. Phys. 1985, 82 (11), 5053. (60) Werner, H . J. Third - order multireference perturbation theory - The CASPT3 method. Mol. Phys. 1996, 89 (2), 645. (61) Werner, H. J.; Knowles, P. J.; Knizia, G.; Manby, F. R.; Schutz, M. Molpro: a general - purpose quantum chemistry program package. Wires Comput Mol Sc i 2012, 2 (2), 242. (62) Humphrey, W.; Dalke, A.; Schulten, K. VMD: Visual molecular dynamics. J. Mol. Graph. 1996, 14 (1), 33. 677 CHAPTER 8. PURSUIT OF RUTHENIUM BIS(IMIDO) COMPLEXES AND HIGHER OXIDATION STATES 8.1 Introduction 13 In 1992 Wilkinson and coworkers published the synthesis of a unique, Ru(IV)(NAr) 2 (PMe 3 ) 2 ( Ru1 ), square planar, d 6 complex (Ar = 2,6 - diisopropylphenyl). 1 In their original report, they sta rt from trans - RuCl 2 (PMe 3 ) 4 and add a significant excess of LiNHAr (3.1 equiv), along with excess PMe 3 . The reaction mixture is refluxed for 1 week and produces a red intermediate, which is carried on without isolation. This complex is postulated to be a Ru(II) solution turns greenish - blue and the Ru(NAr) 2 (PMe 3 ) 2 product was isolated in 16% total yield. Given the ambiguity of the oxidation step in this reaction, th e unidentified intermediate product, and the low overall yield of the Ru bis(imido) species, it seems like several other products are likely formed as a result of this synthesis; in our hands, these results have not been reproduced on any scale. Figure 8 . 1 The synthetic procedure presented by Wilkinson and coworkers in 1992 which lead to the square planar d 4 Ru species on the right. For ruthenium, this is a highly unusual complex for several reasons. As men tioned in the previous chapter, there are sparing examples of terminal Ru imido complexes in the literature. 13 Portions of this work have been published in the following article: Aldrich, K. A Photochemical Route to a Square Planar, Ruthenium(IV) - bis(Imide) Chem Commun , 2019 , 55(30) , 4403 - 4406. 678 These include the Ru(II) (NPh 2,4,6 - tri - tert - butyl )( 6 - cymene) complex published by Steedman 2,3 , as well as the Ru(NAr/Ar*)(PMe 3 ) 3 , Ru(NAr/Ar*)(dmpe) 2 , Ru(NAr/Ar*)dppe(PMe 3 ) complexes that 4,5 Outside of these complexes, which all have similar ligand electronics, there is Ru1 , 1 and a few different derivatives of Ru(VI) porphyrin bis(imido) complexes. 6 - 9 It is notable that these reported Ru(VI) bis(imido) complexes are fleetingly stable and highly reactive, sometimes invoked as reactive intermediates. 10 In this way, the chemistry of Ru imides is highly i ncomplete, marked with small groupings of electronically similar species with large gaps in the oxidation states, coordination environment, and general diversity observed with other metals that form imido complexes. Bis(imido)Os analogues with porphyrin ligands can be isolated and studied to the extent of full structural and electronic characterization. 6,11,12 Mono - imido analogues with Fe have also been reported, and much like the Os examples, are markedly more stable relative to the Ru complexes. 13,14 Additional examples with Fe - porphyrin cores abound if we expand consideration to nitrides and oxos, due to their biological relevance, and this motivation, as a whole has greatly enriched the chemistry of iron ligand multiple bonds. 15 - 17 Access to mid - to high - valen t Os imido complexes has been synthetically facilitated from the OsO 4 , which is commercially available. The first Os - imido compound synthesized, in fact, was Os(O) 3 (N t Bu), which can be generated from addition of H 2 N t Bu to OsO 4 . 18,19 aching high valent Os - imido species. 20 - 24 Alternatively starting from a low valent Os source can produce terminal imido species as well. 14 Thus, synthetically, a fairly wide variety of these types of species have 14 Analogus reaction pathway to that presented in Chapter 7 for Ru(NAr)(PMe 3 ) 3 can be applied to Os(NAr)(PMe 3 ) 3 . However, due to c hallenging starting material pathways and similarity to the Ru species, full electronic structure studies were not pursued with these derivatives. 679 been easy to access with Os. Complementary routes for Ru and Fe howe diverse. Some clever routes have been devised to achieve midvalent bis(imido) species with Fe, however these syntheses start from low valent Fe sources and leverage the reducing ability of Fe(I) and Fe 0 species to transform organ ic azides into N 2 and imido ligands. Two examples in the last decade have shown that this method, coupled with bulky ancillary ligands, can yield isolable Fe(IV) or Fe(V) bis(imido) complexes which have been fully characterized. 25,26 These complexe s are shown in Fig. 8.2. Figure 8 . 2 Examples of various Group 8 mono - and bis(imido) compounds. The history of Group 8 imido complexes seems to demonstrate that part of the reason why Ru - the metal - examples is due to synthetic challenges. Certainly, from our attempts to access mid - valent Ru - imido species, this observation seems accurate (Chapter 7). These difficulties seemed even more significant after our cedure to make Ru1 failed. Recognizing the lack of 680 synthetic techniques to access mid - to high - valent Ru - imido species, we set out to explore alternate synthetic pathways to access these complexes. 8.2 Synthesis of Ru and Os 2 - diphenylhydrazido Complexes fro m Azobenzene We formulated a method of producing higher valent Ru imido complexes that we thought might lead to potentially less reactive intermediates along the pathway to oxidize the metal. As we had previously learned, direct chemical oxidation of an e xisting low - valent, terminal imido complex tends to produce radical species that are highly unstable (see Chapter 7). As an alternative, we sought to use ligands that contain N N bonds that we could harness by attaching them to a low valent Ru complex, a nd then breaking the N N bonds in the ligand to generate imide ligands. Breaking N N bonds in proximity to the metal center could induce an oxidative addition, without requiring the addition of external oxidants or reductants. Rather, heat or light might p rovide enough energy to induce such reactivity. Similar reactivity has previously been reported with an Fe(CO) 3 (1,4 - diphenyltetrazene) species, which upon exposure to light, results in the elimination of N 2 and the formation of a dimeric Fe species [Fe(CO) 3 ] 2 ( - PhNNPh). 27 Similarly, this type of reactivity has also been shown using uranium and external reductant. 28 This general strategy is shown in Fig. 8.3. In a way, this is similar to the methods often employed, for example, to produce the Fe bis(imide) complexes shown in Fig. 8.2. 25,26 In these reactions, a spontaneous redox reaction between the metal and an aryl azide results in oxidation of the metal and formation of imide ligands accompanied by N 2 elimination upon reduction of the azide. Many other metal - imide complexes ha ve also been synthesized using the same technique. However, it requires an extremely reducing metal, often M - 2 to M +1 , or the addition of a strong external reductant such as KC 8 . 14,29,30 681 Figure 8 . 3 Illustration of the synthetic route proposed to access Ru - imido species in mid - to high oxidation states. As a starting point, we chose to target a 2 - hydrazido moiety. Literature procedures have shown t hat upon addition of excess (4 equivalents) Li to azobenzene in THF, a di - lithiated, 2 - electron - reduced species can be generated in situ . Addition of this solution of reduced azobenzene to cis - RuCl 2 (PMe 3 ) 4 results in the production of the ( 2 - diphenylhydra zido)Ru(PMe 3 ) 4 ( Ru2 ) complex shown in Fig. 8. 4 . Figure 8 . 4 ( top) The X - ray crystal structure of Ru2 with ellipsoids shown at 50% probability; H atoms and solvent molecule omitted for clarity. 2 - diphenylhydrazido)Ru(PMe 3 ) 4 from in situ reduced azobenzene and cis - RuCl 2 (PMe 3 ) 4 . The same procedure can be utilized to produce the Os analogue of this compound. 682 8.3 Thermal and Photochem ical Reactivity of ( 2 - diphenylhydrazido)Ru(PMe 3 ) 4 As is evident from the isolation of this complex, spontaneous cleavage of the N N bond upon addition to Ru does not occur. However, with Ru2 in hand, we began exploring methods to induce the oxidative addition of the electron density in the N N bond to the metal. Initially, we examined what happens when the complex is simply heated. Following the reaction by 31 P NMR, the tightly spaced doublet of triplets, characteristic of Ru2 , transforms quantitatively into a doublet of doublets ( - 9.0 ppm, 2P) and a doubl et of triplets ( - 13.9 ppm, 1P), with the loss of 1 PMe 3 ligand. When the volatiles are removed from the reaction solution, and the resulting dark brown residue is recrystallized, X - ray quality crystals of the ortho - C - H activated azobenzene adduct are iso lated ( Ru3 ). This complex shows a distinct Ru H resonance by 1 H NMR at - 11.39 (triplet of doublets), which presumably occupies the vacant 6 th coordination site observed in the single - crystal structure of the complex. Evident from the structure is the drama tic shortening of the N N bond relative to that distance in Ru2 , indicating the reformation of a N N double bond. The Ru1 N1 is also consistent with a dative interaction. Additionally, there is a slight distortion from perfect octahedral geometry, but this is likely a result of the equatorial ligands shifting toward the hydride where the first coordination sphere is less congested. A summary of this reaction and the crystal structure of Ru3 are shown in Fig 8.5. 683 Figure 8 . 5 The Ru(II) terminal hydride (Ru3) species produced upon heating Ru2. The single crystal X - ray structure is shown with ellipsoids at 50% probability; H atoms and solvent omitted for clarity. The same reaction sequence shown above with Ru to produce Ru2 and Ru3 was also undertaken with Os. The same reactivity is noted, giving highly similar structural analogues. However, due to scarcity of the cis - OsCl 2 (PMe 3 ) 4 and no reliable synthesis found by which to remake the starting material from available Os complexes in the laboratory, further reactivity with Os was not pursued. Since heating the complex did not yield the desired result, of breaking the N N bond, we shifted our attention to light - driv en reactivity. There are a few examples of this type of photolysis in the literature, 27 and we had previously observed that dilute (i.e. NMR) samples of Ru2 change color from bright orange to dark green when exposed to ambient light. This initial color change was followed by complete decomposition of the sample (i.e. an intractable mixture with several new 31 P resonances observed). Intentional photolysis of Ru2 agrees with qualitative observations. When a solution of Ru2 in benzene (C 6 D 6 or C 6 H 6 ) or THF is exposed to an intense UV - Vis light source (800 W Hg Arc lamp), the solution rapidly goes from orange to green. The green species ( Ru4 ) can be identified by new 31 P resonances, and increases in the size of these new peaks is accompanied by diminishing 684 peaks for the starting material resonances ( Ru2 ). This green species is transient, however, as it rapidly de composes to Ru3 at room temperature. Therefore, efforts to characterize Ru4 have been limited to in situ experiments that can be done rapidly or are amenable to low - temperatures, as reduced temperatures appear to dramatically slow the conversion of Ru4 to Ru3 . We know that Ru4 is diamagnetic and, like Ru3 , loses one PMe 3 ligand upon formation. The species does not appear to be an alternative C - H activation product, as characteristic 1 H or 13 C resonances for Ru H or Ru C bonds are not observed in photolys is samples in C 6 D 6 . Interestingly, Ru4 also lacks a 14 N NMR resonance. This negative response is very inconclusive, however, as we have noted several other species that contain M N bonds are often 14 N NMR - silent. The quadrupolar relaxation of the 14 N nucle us likely broadens these signals to an extent that our instrumentation cannot detect them; this potential broadening could be exacerbated with coupling (i.e. several 31 P nuclei or 14 N in proximity in the molecule). 19,31,32 As mentioned, by 31 P NMR two new resonances are observed at 11.0 ppm (td, 1P) and - 7.8 ppm (dd, 2P). The integration ratio of 1:2 indicates the presence of a C 2 - axis or mirror plane within Ru4 , and along with the observation that 1 equiv of unbound PMe 3 ( - 62.6 ppm), confirms that 3 PMe 3 ligands remain coordinated to Ru. Aside from this spectral data, 1 H and 13 C NMR of the in situ photolysis solution of Ru4 agree that there are 3 PMe 3 ligands on Ru, and that the hydrazido fragment is still bound to Ru. However, no clear indica tion of the geometry or binding properties of the hydrazido fragment are evident. UV - Vis spectra of the complex, generated in situ in THF show distinct new absorbances, accounting for the dramatic color change. A plot of the electronic absorption spectrum is shown in Fig. 8.6 for both Ru2 and Ru4 . The differences in these spectra are readily observed, including the absorption feature at ~ 15,000 cm - 1 , which likely results in the green color of the complex in 685 solution. While the product, Ru4 , demonstrates st ronger absorbance characteristics over most of the range observed by UV - Vis spectroscopy, relative to the starting material, this does not appear to impact the conversion efficiency of Ru2 to Ru4 . In situ 31 P NMR demonstrates quantitative conversion in rel atively short photolytic exposure times. This suggests that the quantum efficiency of photon absorption resulting in the chemical conversion is very high, or that the incident radiation responsible for the chemical transformation is outside of the range of wavelengths examined. Attempts to identify what excitation event may be leading to reactivity using time - dependent DFT did not provide useful insight. Experimentally, this could be examined by repeating the photolysis with band pass filters, to narrow in on the energy of the transformation. However, since full conversion was achieved and product stability precluded our ability to fully structurally characterize Ru4 , this study was not pursued (see below). Figure 8 . 6 Ru2 and Ru4 . Although the photolysis product absorbs more strongly than the starting material across most of the spectrum, full conversion is still achie ved in these photolytic conversions. (Note that the sharp feature at ~15,000 cm - 1 is due to the light source change in the UV - lamp). 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 12500 17500 22500 27500 32500 (M - 1 cm - 1 ) Wavenumber (cm - 1 ) Electronic Absorption Spectra for Ru2 and Ru4 Ru2 Ru4 686 Figure 8 . 7 ( left ) Schematic showing the interconversion of Ru2 , Ru4 , and Ru3 . ( right ) 31 P NMR of photolysis solution to generate Ru4 . The inset shows the new 31 P resonances, while the sharp singlet at - 62 ppm is free PMe 3 . To aide in experimental deductions of the possib le structure of Ru4 , we turned to density functional theory. The structures (geometries) of Ru2 , Ru3 , and several candidate molecules proposed as Ru4 were optimized utilizing B3LYP as the functional, LANL2DZ(Ru)/6 - 311g+(d,p) (C,H,N,P) basis sets, and the C PCM THF solvent model. The ground state single point energies for these optimized complexes were determined, with the same basis set assignments, solvent model, and the B3LYP functional. Using these calculated energies, the G 0 and H 0 of several possible reactions from Ru2 to Ru4 were estimated. The results of these calculations are shown in Fig. 8.8. 687 Figure 8 . 8 Several possible products that were consid ered in identifying Ru4 Ru2 to the product number listed on the x - axis. Each complex is numbered and shown structurally on the right. Combined with experimental data, two main possibilities emerge from the options that we have considered thus far. Structure 3 appears to be close to thermal neutral in terms of free energy. It also demonstrates a structure with 31 P environments consistent with the NMR spectrum for Ru4 , making it a probable candidate. The 31 P NMR data makes option 5 unlikely, as this structure should present a single 31 P NMR resonance. Complex 7 is difficult to predict, as each PMe 3 is unique relative to the positions of the imides; I would predict that this complex would have a single broad resonance and some degree of PMe 3 exchange, or that all 3 positions would have different shifts by 31 P NMR. However, it is not obvious that the splitting pattern would match what was expe rimentally observed. Also note that when a square pyramidal geometry is provided for optimization, with the imido groups either cis - or trans - they distort until they optimize in alternate geometries. Specifically, as shown for 4 , attempts to optimize a tr ans - Ru(NPh) 2 (PMe 3 ) 3 complex distorted into the same geometry as 3 , which has two equatorial (=NPh) ligands and a trigonal 688 bipyramidal geometry. The geometry of 7 , similarly, was started as cis - Ru(NPh) 2 (PMe 3 ) 3 but distorted to a geometry intermediate betwee n square pyramidal and TBP geometry. The other possible structure, aside from 3 , that seems likely from the list above, is 6. In 6 , the N N bond of the 2 - hydrazido ligand remains intact, but rather the photolysis may dissociate a PMe 3 ligand. This would leave a vacant coordination site on the Ru, generating what would likely be a highly reactive species. The geometry for this structure could not be successfully optimized by DFT, however, so an assessment of the energy change to generat e this species from Ru2 also be a high - energy intermediate; for example, this species is likely somewhere along the reaction coordinate for the conversion of Ru2 to Ru4 or Ru4 to Ru3 . Based on the calculations, however, it cannot be directly ruled out. Collectively, these results do not provide the same strength in structural identity for Ru4 that we would typically acquire through X - ray crystallography or detaile d reactivity studies. The most complete case can be made for Ru4 3 , based on the NMR spectra, UV - Vis properties (see below), and calculations. This result certainly encouraged further exploration of this synthetic strategy to produce mid - valent Ru imido complexes. Several efforts were undertaken by which ligand modification was attempted, to try to produce a more stable derivative of Ru4 . - tetramethylazobenzene, however, the addition of the reduc ed azobenzene species to the Ru metal center results in a mixture of products (many peaks by 31 P NMR of crude product). Conventional separation techniques did not lead to isolation of the desired product. Similarly, when we tried to add bulk to the phosphi ne ligands (i.e. switching to PMe 2 Ph), a similar result was observed. 689 The starting molecule for the photolysis, Ru2 , appears to be very nearly sterically saturated. The space filling model of the single crystal X - ray structure, shown in Fig. 8.9 below illu strates this property well. It seems likely that by adding bulk to the substituents close to the Thus, given the fleeting nature of the complex and the ine vitable conversion to the terminal Ru(II) hydride species ( Ru3 ), as well as our inability to alter the ligand substitution for enhanced stability, we sought a slightly more tamable ligand as the nitrogen source. Figure 8 . 9 The single crystal X - ray structure of Ru2 shown ( left ) with thermal ellipsoids and ( right ) as the spacefilling model with van der Walls radii on all atoms. Notice that essentially none of the central Ru atom is visible from the spacefilling perspective, demonstrating the steric crowding in this molecule. 8.4 Synthesis of Ru(II)(1,4 - diaryltetrazene)tris(trimethylphosphine) Complexes As mentioned in the Introduction (8.1), addition of organic azides to low - valent me tals has been shown to lead to spontaneous reduction of the azide and oxidation of the metal. Sometimes this can be achieved by adding an external reductant or simply by heating the reaction. 14,29,30 When an aryl azide is added to the terminal Ru(NAr)(PMe 3 ) 3 complex, a spontaneous redox reaction does not occur. Rather, the azide adds cleanly to the existing Ru N double bond to form a Ru(II)(1,4 - 690 diaryltetrazene)tris(trimethylphosphine) ( Ru5 ) complex. Similar complexes with Ru and Ir, with 6 - arene or Cp ligand s respectively, have been published by Wilkinson and Hursthouse in addition to examples with other metals. 33 - 35 This reaction, and subsequent isolation of the tetrazene species, has been achieved with N 3 Ar, N 3 Mes, and N 3 Ar to Ru(NAr*)(PMe 3 ) 3 ( Ru7 ) to yield the symmetric Ru(1,4 - (2,6 - diisopropylphenyl)tetrazene)(PMe 3 ) 3 ( Ru5 ), as well as the asymmetric Ru(1 - mesityl - 4 - (2,6 - diisopropylphenyl)tetrazene)(PMe 3 ) 3 ( Ru6 ) and Ru(1 - (2,4,6 - triisopropylphenyl) - 4 - (2,6 - diisopropylphenyl)tetrazenen)(PMe 3 ) 3 species ( Ru7 ), respectively. This scheme is outlined in Fig. 8.10. Even N 3 TMS undergoes the same initial reaction, and a preliminary crystal structure of the Ru(1 - (2,6 - diisopropylphenyl) - 4 - trimethylsilyl - tetrazene)tris(trimethylphosphine) ( Ru8 ) complex has been obtained. However, the geo metry of this species appears fluxional in solution (broad 1 H and 31 P NMR), and the complex readily decomposes as a solid at room temperature or from exposure to ambient light. It is also worth mentioning that the crystal structure is preliminary because t he crystals decompose upon irradiation with X - rays. For these reasons, further study of The X - ray crystal structure of Ru5 is shown in Fig. 8. 10 , along with an outline showing the general synthesis. The structural characterist ics of all 3 complexes, Ru5 - 7 , are similar. The base geometry is close to square pyramidal, with the plane defined by N1, N4, P1, P2, and Ru1. The angles residing in the plane sum to a total of 352 °. The plane is slightly distorted, as the 4 ligand atoms in the square base flex slightly below the plane; the axial P3 is also tilted slightly complex = 1 is trigonal 691 bipyramidal it is found to be 0.02. Again, it seems like the distortion from square planar in this structure are primarily a result of steric congestion. This is demonstrated by the spacefilling model of Ru5. Figure 8 . 10 (top) Single crystal X - ray structure of Ru5. Thermal ellipsoids are shown at 50% probability and H atoms and solvent were omitted for clarity and s pacefilling mode l of Ru5 . (bottom) General synthetic scheme for making Ru(II) tetrazene complexes. 692 Table 8 . 1 Selected bond distances and angles from the single crystal structure of Ru5 in Fig. 8.10. Bond Distance (Å) Angle (°) Ru1 N1 2.043 N1 Ru1 N4 73.29 N1 N2 1.374 N1 Ru1 P2 95.85 N2 N3 1.274 N4 Ru1 P1 92.82 N3 N4 1.366 P1 Ru1 P2 89.92 Ru1 N4 2.043 P1 Ru1 P3 95.94 Ru1 P1 2.326 P2 Ru1 P3 93.47 Ru1 P2 2.309 N1 Ru1 P2 105.63 Ru1 P3 2.245 N4 Ru1 P1 109.83 8.5 Reactivity of Ru(II)(1,4 - diaryltetrazene)tris(trimethylphosphine) Complexes The goal of reactions with Ru5 - 7 is to achieve cleavage of the tetrazene ring accompanied by the elimination of N 2 . Initial studies with Ru5 showed that chemical oxidants and reductants are relatively unreactive with the tetrazene species, even with an excess of oxidant or strong reducta nt (KC 8 ) or mild heating (< 50 ° C). At higher temperatures, decomposition is noted by 1 H and 31 P NMR, however, even when alone in solution, Ru5 decomposes at temperatures over 65 ° C. This also demonstrates that Ru5 W ith the failures of reductants, oxidants, and thermally driven reactivity, we turned to light - induced reactivity. A 0.002 M solution of Ru5 in THF was prepared in a Schlenk tube. The reaction was then transferred to a jacketed chiller fitted with a quartz window and the sample was irradiated using a mercury arc lamp. After 8 hours of irradiation at 800 W, the transparent orange solution had begun to darken, after 24 h the solution was opaque and murky green, and after48 h the solution was dark blue. Crude 3 1 P NMR shows that about 25% of the Ru5 was converted to Ru1 , Ru(NAr) 2 (PMe 3 ) 2 . Upon work - up, a yield of 21% of Ru1 crystals were obtained, with a small impurity of the phosphine - imine (ArN)PMe 3 . 693 A single crystal X - ray structure of Ru1 was collected and provided the same unit cell and parameters as were previously reported by Wilkinson, et. al. Note that the crystal structure, as shown in Fig. 8.11 shows the full molecule, but the Ru sits on a crystallographic inversion center. As a result, half of the molecule is symmetry generated. Repeated recrystallizations of impure Ru1 appear to remove most of the impurity, however, satisfactory elemental analysis of the material was not obtained. In part, this is due to the limiting masses that can be carried through the photochemical synthesis and the inherent conversion limit. This conversion limit seems to be a direct result of the absorption properties of the product ( Ru1 ) and the re actant ( Ru5 ). Both complexes absorb strongly across the UV - Vis spectrum. At several points, Ru1 even absorbs more strongly than Ru5 . Thus, when a substantial concentration of Ru1 has accumulated in solution, further conversion of Ru5 is halted because of t he limited penetration depth of the incoming radiation. When the concentration of the photolysis solution is reduced, more decomposition is noted, and the desired yield increase is not achieved. We suspect that this is likely the result of adventitious wat er in the solvent or the effect of a small Figure 8 . 11 Single crystal X - ray structure of Ru(NAr)2(PMe3)2 with ellipsoids shown at 50% probability and H atoms omitted for clarity. 694 explanation for the additional decomposition observed, lower conversion at lower concentrations could also indicate that a bimolecular mechanism is responsible for the conversion, where the rate would be decreased with a decrease in concentration of Ru5 . However, given the increased answer our practical problems. Similar results were achieved with Ru7 , with the photolysis solution going from orange to dark green over 24 h and a 31 P signal growing in at - 22 ppm. Howev er, isolation of the more heavily - substituted Ru(NAr)(NAr*)(PMe 3 ) 2 ( Ru1* ) derivative from the starting material, phosphine - imine byproduct, and small amounts of both anilines, was not productive. Ru6 , on the other hand, exclusively shows decomposition upon irradiation. A transparent orange solution of Ru6 , irradiated for 8 h at 800 W, turns pale yellow. A dark precipitate is formed and H 2 NMes and -0.2 0.3 0.8 1.3 1.8 2.3 200 300 400 500 600 700 800 Abs Concentration (M) Abs. Vs Wavelength (nm) for Ru5 (orange) and Ru1 (blue) Figure 8 . 12 Absorption spectra for 0.002 M solutions of Ru5 (orange trace) and Ru1 (blue trace) in THF. The strong absorbance of the product ( Ru1 ) across the UV - Vis spectrum likely contributes to the conversion limit of 25% in solution. 695 H 2 NAr are observed in the crude reaction residue by 1 H NMR. This result highlights a point that will be further illustrated below that Ru1 (and proposed Ru1* ) is an anomaly, stabilized by extreme steric protection imparted by the 4 isopropyl groups which point toward the metal. Steric bulk on the aryl imides is necessary to prevent undesired reactivity and even the reduction of an i Pr to a Me group in the ortho positions can lead to decomposition. This makes efforts to further improve this reaction pathway challenging. Significant alteration of the ligand electronics is the most straightforward way to change the absorption properties of the product and reactants, which is necessary to overcome the c onversion limit. Dramatic stereoelectronic changes, however, are also likely to destabilize the Ru(IV) bis(imido) formation. The same problem we faced with the Ru2 complex, where ligand manipulations led to undesired reactivity, was encountered again in th is system with the tetrazene ligands. Thus , these synthetic challenges prevented the synthesis and isolation of the targeted derivatives for Ru1 . 8.6 Synthetic Alterations of the Ru Platform: A Larger Phosphine Ligand Based on the photochemical decomposition of Ru6 , it seemed like alteration of the aryl group on the imido was detrimental to stability. Instead, we sought to alter the phosphine ligand, going from PMe 3 to the slightly larger PPhMe 2 (TCA of 118 ° and 122 °, respectively). Replication of the synthe ses to yield cis - RuCl 2 (PPhMe 2 ) 4 and Ru(NAr)(PPhMe 2 ) 3 ( Ru8 ) was straightforward and identical to the procedures used to make the PMe 3 derivatives. 696 Figure 8 . 13 Reaction scheme to produce a Ru(II) imide species with bulkier phosphine ligands, PPhMe 2 , and subsequent lack of reactivity upon addition of aryl azide. Upon addition of aryl azide to Ru8 , however, no reaction occurred. The two species simply coexist in solution without yielding the desired tetrazene species. We suspected that the increase in the size and rigidity of the phosphine ligands lead to steric inhibition that prevented reaction of the Ru species and the azide. The size of the aryl group on the imide was reduced in order to increase the reactivity of the Ru(II) imido. However, this increased the reactivity of the complex too much. Upon addition of 2.1 equiv of LiNHMes to cis - RuCl 2 (PPhMe 2 ) 4 , the clear yellow solution rapidly turns dark red. After 2 h at room temperature, the red solut ion turns pink. After work - up, the product recovered is a Ru(II)( 2 - NH(6,4 - Me - 2 - CH 2 - Phenyl)((PPhMe 2 ) 4 ( Ru9 ), where C H activation of one of the ortho ( - CH 3 ) groups on the mesityl imide fragment has occurred. 697 While inte resting, these results prevented our attempt to generate a tetrazene species with larger phosphine ligands. Based on these efforts, it seems like producing Ru(NR) 2 (PR 3 ) 2 complexes with variable substitution will require very careful balancing of the R groups on the imido and phosphine ligands at every step in the synthesis. The promising results that we observed with both the Ru5 and Ru7 derivatives certainly suggest that expansion of this synthetic method deserves further exploration. While it may be possible that Ru1 the wider variation in ligand stability observed with related Os - imido complexes suggests that alternate substituent com binations should be accessible. 8.7 A Basic Reactivity Study with Ru1 and Comparison to Known Os Analogues As mentioned above in the introduction, Schrock and coworkers published several Os bis - and tris(imido) species. Of particular relevance to this work is the synthetic procedure that transforms OsO 4 into Os(NAr) 2 O 2 and finally Os(NAr) 2 (PMe 3 ) 2 , which is outlined below in Fig. 8.15. The Os(NAr) 2 (L) 2 derivatives where L = PPhMe 2 , PPh 2 Me, and PPh 3 were also synthesized in a similar fashion. Figure 8 . 14 The formation of a 6 - coordinate, C H activated mesityl anilide species, resulting from an attempt to generate a terminal Ru(=NMes)(PPhMe 2 ) 3 . 698 Figure 8 . 15 2 - diphenylacetylene)Os(NAr) 2 ( Os10 interaction of the unsaturated C C bond participants with the Os N multiple bond. The same synthetic method was applied to produce Ru10 . In addition to electronic structure analysis with these complexes, these studies also explored several basic reaction pathways. One that particularly interested us is the addition of an alkyne to M(NAr) 2 (PMe 3 ) 2 , as shown in the figure. We sought to observe whether thi s reactivity 2 - binding of the alkyne to the metal would be observed with the Ru analogue, as well. Electronically, this is an interesting result. If we think about the nature of the M N multiple bond, the nature of this bond should be impacted by t he oxidation state of the metal. A metal - imide bond in a low oxidation state metal tends to be nucleophilic, with residual electron density on the N, and Conve rsely, as the d - orbital manifold is emptied and the metal gains electronegativity, as occurs orbitals. This tends to saturate more electron density into the M N bond, making the N less nucleophilic. This facilitates alternate reactivity, such as sigmatropic rearrangements (i.e. cycloaddition of another unsaturated species). In complexes with similar ligands, but where the Os has maximal valency, like Os(NAr) 3 O unsaturated C C bonds can add to the imide nitrogens to form two new C 699 Alternatively, as we have noted with complexes such as Ru(NAr)(PMe 3 ) 3 , the imide nitrogen is a strong nucleophile, attacking Lewis acidic or electrophilic atoms, such as the Cu in CuI or the central C atom in phenylisocyanate (OCNPh). 36 Primarily this seems to differentiate the two diverging reaction pathways, where categorical differences arise depending on the valency of the metal and its relation to the HOMO and LUMO character. Reactivity exhibited by high valent systems, therefore, s eems to involve a delocalized orbital spanning the metal and the N of the imide ligand. In low valent systems, on the other hand, the HOMO tends to be nonbonding to antibonding in nature and centered on the N. therefore, reactivity in these complexes is dr iven by nucleophilic attack of the N atom. Given these two avenues of M N double bond reactivity, one might expect that a mid - valent metal - imide complex of this type could react in either manner, depending on the substrate, sterics in the system, etc. Bas ed on the reaction shown in Fig. 8.15, however, neither reactivity is exhibited. In fact, the M N bonds demonstrate inertness. In this same reaction with Ru1 , addition of a phosphine scavenger CuI) and an alkyne to Ru1 , addition of the alkyne results in a color change in the solution. However, by NMR no reaction is evident. Only upon addition of the phosphine - scavenger (CuI) does the alkyne interact with the remaining Ru compound in solution. Because only small amounts of Ru1 could be isolated, we conducted these investigations in situ , examining the reaction by NMR. Both 14 N and 1 H NMR present similar spectra to those of the Os 2 - PhCCPh)Ru(NAr) 2 complex. The 14 N shifts no ted for Os(NAr) 2 (PMe 3 ) 2 , Os(NAr) 2 2 - PhCCPh), and Os(NAr) 2 (O) 2 were observed at 283.0, 365.6, and 390.8 ppm, respectively. Across this series, the 14 N shifts for the imido moieties increases as the Os becomes more formally oxidized or less electron rich an d the 700 Os N bonds become, presumably, more covalent. We can see a similar trend with the Ru values, where Ru(NAr) 2 (PMe 3 ) 2 has a shift of 303.3 ppm and Ru(NAr) 2 2 - diphenylacetylene) has a shift of 387.1 ppm. These values follow the trends we would predict b ased on the electronic factors affecting 14 N NMR shifts. 31 2 - diphenylacetylene imide shifts for Ru10 and Os10 fall closer to those in the Os(NAr) 2 (O) 2 complex than Ru1 and Os1 . This observation 2 - alkyne complexes have M N bonds more similar in character to those in the Os(VIII) complex than the Os(IV) or Ru(IV) complexes, ( Os1/Ru1 ) respectivel y. This change in the 14 N chemical shift suggests a higher bond order or more electron density donation from the imide groups likely occurs in Ru/Os10 than in Ru/Os1 worth considering Ru/Os10 as M(VI) species. H owever, this may also be due to the proposed geometry change, whereby the two imide groups may no longer be trans to one another. Electronic structure calculations probing charge distribution in the system and bond order between the metal 2 - PhCCPh to assess these differences would be interesting to pursue. 701 Figure 8 . 16 14 N NMR of Ru10 (387.1 ppm) and Os10 (365.6 ppm). 8.8 Conclusions In terms of reactivity, it would be very interesting to observe whether or not the inertness observed for Ru(NAr) 2 (PMe 3 ) 2 ( Ru1 ) is from some electronic effect of its middling oxidation state (i.e. midvalency of the metal spreads the HOMO and LUMO across both bond participants, making the metal - imide bond neithe r nucleophilic or electrophilic), or if this effect is a manifestation of 702 the sterics needed to prevent C H activation reactions of the ligands. To that end, continued expansion of synthetic techniques to yield Ru(III) - (V) imides with varied ligand platfor ms is needed. As the studies presented here demonstrate, the process of generating midvalent Ru imidos is a delicate synthetic challenge. We have observed C H activation, total decomposition, and a complete lack or reactivity all as results of very subtle changes to the ligand R groups. While the synthetic strategies that we have discovered here are promising and may be expandable to generate similar complexes (i.e. using slightly different ligand scaffolds with photochemical methods), initial efforts high light the challenges that face generalization of these processes. Considering analogous chemistry with Fe and Os, it seems that the Ru syntheses are somewhat more sensitive to ligand alteration in terms of finding side reactions and their subsequently high propensity for off - target products. 8.9 Experimental General Considerations Synthetic Considerations All manipulations were carried out under inert atmosphere, either in an N 2 atmosphere MBraun glovebox or using standard Schlenk techniques. The solvents THF and n - hexane were dried over Na and distilled under N 2 prior to use. The solvents toluene, Et 2 O, and pentane were dried by passage over activated alumina and sparged with N 2 prior to use. The NMR solvents C 6 D 6 and tol - d 8 were dried over CaH 2 and distilled under N 2 prior to used. The solvent hexamethyldisiloxane was dried over CaH 2 and distilled under N 2 prior to use. The azide starting materials N 3 Ar, and N 3 Mes, were synthesi zed according to literature procedures. 37 However, the purification of the organic azide by silica gel chromatography was omitted, as the crude product 703 was found to be pure by 1 H and 13 C NMR after removal of volatiles. N 3 TMS was purchased from Alfa Aeasar and used as received. The H 2 NAr and H 2 NMes were purchased from Oakwood and distilled under vacuum from CaH 2 prior to use. The cis - RuCl 2 (PMe 3 ) 4 starting material, Ru(NAr)(PMe 3 ) 3 , and Ru(NAr*)(PMe 3 ) 3 were prepared according to published procedures. 5 4 LiNHR salts (where R = 2, - diisopropylphenyl or mesityl) were prepar ed by adding 2.5 M n BuLi (in hexanes, 1 equiv) to a chilled soluition (liquid N 2 coldwell) of the respective H 2 NR (1 equiv) in hexane. The reaction was warmed to room temperature, while stirring, which resulted in the precipitation of a white to pale yello w powder. This powder was collected by filtration, rinsed several times with n - hexane, and dried under reduced pressure. The powder was used without further purification. It was stored in the glovebox freezer ( - 35 °C) to reduce exposure to light. The phosp hines PMe 3 and PPhMe 2 were purchased commercially (Strem Chemical Co.) and used as received. They were stored in sealed containers in an N 2 glovebox. Instrumentation NMR All NMR data was collected at the Max T. Rogers NMR facility. Routine characterizatio n spectra were obtained using an Agilent DDR2 500 MHz NMR spectrometer equipped with a 5 mm PFG OneProbe operating at 499.84 MHz ( 1 H), 125.73 MHz ( 13 C), and 202.35 MHz ( 31 P). Additional experiments, including 14 N and variable temperature NMR measurements w ere mad using the following instrumentation: a UNITYplus 500 spectrometer equipped with a 5 mm switchable broadband probe operating at 36.12 MHz ( 14 N); a Varian Inova 500 spectrometer equipped with a 5mm pulse - field - gradient (PFG) switchable broadband prob e operating at 499.84 MHz ( 1 H); a Varian Inova 600 spectrometer equipped with a 5 mm PFG switchable broadband probe operating at 599.89 MHz ( 1 H). 1 H NMR chemical shifts are reported relative to residual C 6 HD 5 in C 6 D 6 as 7.16 ppm. 13 C NMR chemical shifts are reported relative to 704 ( 13 C)C 5 D 6 as 128.06 ppm. 14 N NMR shifts are referenced to the internal peak for dissolved N 2 in NMR solvent (309.6 ppm vs. external nitromethane as 381.6 ppm, which sets NH 3 to 0 ppm). X - Ray Crystallography A ll crystallographic data was collected at the Michigan State University Center for X - ray Crystallography. All structures were collected on Bruker AXS instruments operating with either copper or molybdenum radiation sources. Data was collected at 173 K. Str ucture solutions were typically found using XT Intrinsic Phasing and refined by least squares using Olex software. For further information please see the .cif files provided as supporting information. Photochemical Reactions Photolysis experiments were ca rried out using an Oriel Instruments Mercury Arc Lamp (Model # 66921) operating between 450 - 2 O column was used as an IR filter between the lamp and the jacketed chiller, fitted with a quartz window. This reduced the temperature at the q uartz window between incoming light and the reaction vessel, which was important for some experiments. A picture of the photolysis set - up is shown below. UV - Vis Spectroscopy UV - Vis spectra were collected using an Ocean Optics DH - mini UV - Vis - NIR spectropho tometer in an N 2 glovebox. Experiments were performed in dry THF using a quartz cell. The raw data was fitted with OriginPro 9.0 software to obtain accurate peak separation and assignment of maxima. Synthetic Procedures In - situ Reduction of PhNNPh A scin tillation vial was charged with 176 mg of azobenzene (1 equiv, 1 mmol), a stir bar, and 6 mL of THF. The solution was stirred at room temperature, and 30 mg (4 equiv, 4 mmol) of Li pellets was added to the solution. The solution went from bright, 705 transpare nt orange to dark green rapidly. The solution was stirred for 24 h, over which time it turned pale, transparent yellow. Note, solid Li was still present after this time. This solution was stable at room temperature in the presence of excess Li. Full conver sion was assumed in calculating stoichiometries. The same procedure as listed above can be performed with Na in place of Li metal, however, results in a dark brown solution after 24 h, and seems to lead to some reduction when utilized in the subsequent rea ction step. Synthesis of Ru(PhNNPh)(PMe 3 ) 4 · PhMe (Ru2) A 35 mL pressure tube was charged with 480 mg (1 equiv, 1 mmol) of cis - RuCl 2 (PMe 3 ) 4 , a stir bar, and 5 mL of THF. To this solution was added a 6 mL solution in THF containing 190 mg (1 equiv, 1 mmol) of reduced azobenzene (Li 2 [PhNNPh]). The tube was sealed and removed from the glovebox and heated at 60 °C in an oil bath for 8 h. The pressure tube was cooled and returned to the glovebox, where the THF was removed under vacuum. The dark brown resid ue was extracted with toluene until the extract was colorless. The toluene extracts were filtered over Celite, concentrated, and n - hexane was layered into the toluene solution. This layered solution was stored at - 35 °C for 48 h to yield 415 mg (61 %) of c rystals of Ru(PhNNPh)(PMe 3 ) 4 · PhMe. A similar procedure can be used with Na 2 [PhNNPh], with slight modifications. A scintillation vial was charged with 200 mg (1 equiv, 0.42 mmol) of RuCl 2 (PMe 3 ) 4 , a stir bar, and 5 mL of THF. This solution was stirred at room temperature. To the stirred solution, a solution containing reduced azobenzene Na 2 [PhNNPh] (originally 75 mg of azobenzene (1 equiv, 0.42 mmol) and 40 mg Na (4 equiv, 1.68 mmol)) was added dropwise; note, the excess sodium in this solution was not tra nsferred to the Ru solution. The reaction solution rapidly changed color from pale yellow to dark brown upon addition of the reduced azobenzene species. The reaction mixture 706 was stirred for 8 h at room temperature, and the volatiles removed under reduced p ressure. The dark brown residue was extracted with toluene until the extracts came off colorless. The toluene extract was then filtered over Celite, and concentrated. The concentrated toluene filtrate was layered with hexane and stored at - 35 °C to yield 1 33 mg (47 %) of crystals of Ru(PhNNPh)(PMe 3 ) 4 · PhMe. This material had identical properties to that obtained using Li as the reductant above. 1 H NMR (500 MHz, benzene - d 6 J = 8.1 Hz, 2H), 7.20 (t, J = 7.7 Hz, 2H), 7.06 (d, J = 7.5 Hz, 2H), 6.6 8 (t, J = 7.1 Hz, 2H), 6.46 (d, J = 8.4 Hz, 2H), 1.08 (d, J = 6.4 Hz, 18H), 0.93 (t, J = 2.6 Hz, 18H). 13 C[ 1 H] NMR (126 MHz, benzene - d 6 25.02 (d, J = 21.2 Hz), 20.35 (t, J = 13.5 Hz). 31 P NMR (202 MHz, benzene - d 6 - 0.92 - 6.11 (m). Elemental Analysis calculated for RuC 31 H 54 P 4 N 2 : C, 54.78, H, 8.01, N, 4.12; found, C, 54.30, H, 7.95, N, 4.13. UV - max 1 - cm - 1 1 - cm - 1 5029 M 1 - cm - 1 ). Synthesis of Ru - H (Ru3) A 35 mL p ressure tube was charged with 100 mg of RuN 2 P 4 , a stir bar, and 10 mL of THF. The pressure tube was sealed and transferred from the glovebox to an oil bath and heated at 60 °C for 12 h. The pressure tube was cooled to room temperature and returned to the g lovebox, where the THF was removed under reduced pressure, leaving a dark brown residue. The residue was extracted with n - hexane (6 mL), and the extract filtered over Celite. The filtrate was concentrated and stored at - 35 °C for 5 days to yield 42 mg (57% ) of crystals of RuH . 1 H NMR (500 MHz, benzene - d 6 8.61 (m, 1H), 8.45 8.40 (m, 1H), 7.47 7.40 (m, 2H), 7.24 (tdd, J = 7.0, 1.7, 1.0 Hz, 2H), 7.16 7.14 (m, 3H), 7.11 7.05 (m, 1H), 0.99 (d, J = 5.2 Hz, 9H), 0.90 0.81 (m, 18H), - 11.39 (td, J = 33.1, 1 5.4 Hz, 1H). 13 C[ 1 H] NMR (126 MHz, 707 benzene - d 6 123.11, 120.98, 23.72 (d, J = 17.0 Hz), 22.19 21.57 (td, J = 13.8, 3.0 Hz). 31 P NMR (202 MHz, Benzene - d 6 - 8.98 (dd, J = 29.9, 13.2 Hz, 2P), - 13.86 (td, J = 29.6, 29.2, 5.5 Hz, 1P). 14 N NMR (36 MHz, benzene - d 6 2 P 3 H 37 C 21 : C, 49.31, H, 7.29, N, 5.48; found, C, 49.10, H, 7.69, N, 4.82. Synthesis of TBP cis - R u(NPh) 2 (PMe 3 ) 3 (Ru4) ( in situ ) A J.Young tube was loaded with a solution containing 20 mg of RuN 2 P 4 in 1 mL C 6 D 6 . The tube was sealed with a Teflon stopper and transferred out of the glovebox to the photolysis apparatus. The tube was irradiated with a merc ury arc lamp at 800 W in an ice bath (0 °C) for 2 hours. The tube and its contents were then taken for in situ measurements. Care was taken to keep the solution cold ( - 30 °C) for long measurements and while storing the solution to prevent thermal conversio n to Ru 3 . 1 H NMR (500 MHz, benzene - d 6 J = 7.8 Hz, 1H), 8.43 (d, J = 7.4 Hz, 1H), 7.36 7.30 (m, 2H), 7.25 (td, J = 7.9, 7.4, 1.5 Hz, 1H), 7.22 7.17 (m, 3H), 1.36 (d, J = 7.2 Hz, 9H), 0.70 (t, J = 2.9 Hz, 18H). 13 C[ 1 H] NMR (126 MHz, benzene - d 6 128.67, 128.17, 120.42, 24.24 (dd, J = 21.0, 3.5 Hz), 21.21 (t, J = 13.2 Hz). 31 P NMR (202 MHz, benzene - d 6 J = 42.9, 42.4, 6.0 Hz), - 7.83 (dd, J = 42.6, 7.7 Hz), - 62.61. UV - vis: 636 1 - cm - 1 1 - cm - 1 1 - cm - 1 M 1 - cm - 1 ). Synthesis of Ru(1,4 - bis(2,6 - diisopropylphenyl)tetrazene)(PMe 3 ) 3 (Ru5) A scintillation vial was charged with 150 mg (1 equiv, 0.3 mmol) of Ru(NAr)(PMe 3 ) 3 , a stir bar, and 5 mL of Et 2 O. To this stirred solution, a separate solution containing 60 mg (1 equiv, 0.3 mmol) N 3 Ar in 2 mL of Et 2 O was added dropwise. After 10 min of s tirring at room temperate, a fine orange precipitate had started to form. The reaction was stirred another 2 h at room temperature, at which time the 708 fine orange powder was collected by filtration and rinsed with n - hexane. The powder was dried under vacuum and found to be analytically pure, giving 120 mg (57 %) of RuN 4 P 3 . Chilling the original filtrate ( - 35 °C) resulted in the precipitation of an additional 52 mg (25 %) of RuN 4 P 3 . X - ray quality crystals were grown from concentrated THF layered with hexane a t - 35 °C. 1 H NMR (500 MHz, benzene - d 6 J = 8.2, 7.1 Hz, 2H), 7.30 (d, J = 7.6 Hz, 4H), 3.15 (p, J = 6.8 Hz, 4H), 1.39 (d, J = 6.6 Hz, 12H), 1.31 (d, J = 6.9 Hz, 12H), 1.01 0.81 (m, 26H). 13 C[ 1 H] NMR (126 MHz, benzene - d 6 23.66 (m), 21.83. 31 P NMR (202 MHz, benzene - d 6 33 H 61 P 3 N 4 : C, 55.99, H, 8.69, N, 7.92; found, C, 55.46, H, 8.47, N, 7.83. UV - max - 1 cm - 1 - 1 cm - 1 ). Synthesis of Ru(1 - (2,6 - diisopropylphenyl) - 4 - mesityl - tetrazene)(PMe 3 ) 3 (Ru6) The same procedure used for RuN 4 P 3 above was us ed for the preparation of RuN 4 Ar/Mes P 3 , utilizing 76 mg (1 equiv, 0.15 mmol) Ru(NAr)(PMe 3 ) 3 and 32 mg N 3 Mes (1 equiv, 0.15 mmol). This yielded 79 mg (73 %) of the crude powder. X - ray quality crystals were grown from concentrated THF layered with hexane at - 35 °C. 1 H NMR (500 MHz, benzene - d 6 7.36 (m, 1H), 7.30 (d, J = 7.6 Hz, 2H), 6.95 (s, 2H), 3.20 (p, J = 6.8 Hz, 2H), 2.33 (s, 3H), 2.20 (s, 6H), 1.40 (d, J = 6.7 Hz, 6H), 1.30 (d, J = 7.0 Hz, 6H), 0.93 (dd, J = 4.6, 2.5 Hz, 28H). 13 C[ 1 H] NMR (126 MHz, benzene - d 6 145.65, 135.19, 133.79, 128.64, 125.87, 122.51, 27.55, 26.85, 24.28 23.58 (m), 22.03, 20.82, 20.16. 31 P NMR (202 MHz, benzene - d 6 30 H 55 P 3 N 4 : C, 54.12, H, 8.33, N, 8.42; found, C, 54.10, H, 8.64, N, 8.19. UV - max - 1 cm - 1 - 1 cm - 1 ). 709 Synthesis of Ru(1 - (2,6 - diisopropylphenyl) - 4 - (2,4,6 - triisopropylphenyl) - tetrazene)(PMe 3 ) 3 (Ru7) A scintillation vial was charged with 70 mg Ru(NAr*)(PMe 3 ) 3 (0.128 mmol, 1 equiv), 5 mL THF, and a stir bar. This solution was stirred at room temperature, and a solution of N 3 Ar (29 mg, 0.128 mmol, 1 equiv) in THF (1 mL) was added to it dropwise. The solution was stirred for 4 h at room temperature, at which point an orangish - red, powdery precipitate had formed. The solution was decanted from the precipitate, and the precipitate was rinsed with hexane and dried. This yielded 68 mg of the crude powder, which was recrystallized from HMDSO t o give 24 mg X - ray quality crystals. Additional powder was precipitated from the mother liquor of the reaction by chilling the solution to - 35 °C. This provided a total yield of 53 mg (55%). 1 H NMR (500 MHz, benzene - d 6 J = 8.2, 7.0 Hz, 1H), 7 .33 7.28 (m, 2H), 7.25 (s, 2H), 3.16 (hept, J = 6.7 Hz, 4H), 3.02 (hept, J = 7.0 Hz, 1H), 1.46 1.38 (m, 18H), 1.36 (d, J = 6.9 Hz, 6H), 1.32 (d, J = 6.9 Hz, 6H), 0.97 0.85 (m, 27H). 31 P NMR (202 MHz, benzene - d 6 6.51. 13 C NMR (126 MHz, benzene - d 6 155.02, 152.88, 146.37, 146.17, 145.77, 126.31, 122.80, 120.39, 34.85, 28.05 (d), 27.59 (d), 24.91, 24.68, 24.60 (d), 24.54, 24.47, 22.31 (d), 2.10. 36 H 67 P 3 N 4 : C, 57.66, H, 9.01, N, 7.47; found, C, 57.33, H, 9.04, N, 6.80. UV - max - 1 cm - 1 - 1 cm - 1 ). Synthesis of Ru(NAr) 2 (PMe 3 ) 2 (Ru1) A 500 mL Schlenck tube, fitted with a Teflon stopper, was charged with 75 mg of RuN 4 Ar P 3 (0.11 mmol), a stir bar, and 200 mL of THF. This provided a 5.5 mM s olution. The Schlenck tube was sealed and transferred from the glovebox to the photolysis apparatus. The Schlenck tube was submerged in a water - jacketed chiller which maintained a temperature of about 16 °C during the photolysis process. The jacketed chil ler was placed on a stir plate, and in the path of the Hg - Arc lamp, with the quartz window of the jacketed chiller aligned with the lamp. The lamp was then run at 800 W, 710 irradiating the stirred solution in the Schlenk flask for 48 h. After the photolysis p eriod, the exterior of the Schlenk tube was dried and the vessel was returned to the glovebox. The reaction solution was transferred to a side - arm flask, and the volatiles removed under reduced pressure to provide a tacky, dark brown residue. This residue was extracted with cold n - hexane until the extracts came off colorless. The extracted n - hexane solution was filtered over celite and the filtrate concentrated to about 1 mL. This solution was stored at - 35 °C for several days to provide 38 mg of crystals o f Ru(NAr) 2 (PMe 3 ) 2 , mixed with crystals of RuN 4 Ar P 3 and P(NAr)Me 3 . To obtain a purer sample of Ru(NAr) 2 (PMe 3 ) 2 , 2 more recrystallizations from n - Hex were preformed successively. This sample was utilized to examine the UV - Vis spectrum of the complex and HRM S. Attempts to obtain elemental analysis were unsuccessful. 1 H NMR (500 MHz, benzene - d 6 6.90 (m, 4H), 6.85 6.81 (m, 2H), 4.45 (hept, J = 7.0 Hz, 4H), 1.28 (d, J = 6.9 Hz, 26H), 1.23 (t, J = 2.6 Hz, 19H). 31 P NMR (202 MHz, benzene - d 6 - 20.90. 14 N NMR (36 MHz, benzene - d 6 - max - 1 cm - 1 ), - 1 cm - 1 - 1 cm - 1 - 1 cm - 1 ). Repeated attempts to obtain passing elemental analysis failed. Given the relatively small sc ales on which this reaction can be performed, and the product compound isolated, HRMS was instead attempted. A peak for the species shown below was observed by QTOF - HRMS running in 2 C 18 H 36 + : 429.1296; found: 429.1096. Figure 8 . 17 Fragment for Ru1 observed by HRMS. 711 Scheme 8 . 1 Photochemical conversion to yield Ru1 from Ru5 . Synthesis of Ru(NAr) 2 2 - Diphenylacetylene) (Ru10) (in situ) A scintillation vial was charged with 12 mg Ru1 (1 equiv, 0.021 mmol), 5.5 mg diphenylacetylene (1.5 equiv, 0.031 mmol), a stir bar, and 2 mL of C 6 D 6 . The mixture was stirred at room temperature. To the stirred solution was added 8 mg (2 equiv, 0.042 mmol) of CuI, portionwise. The reaction solution was stirred vigorously for 4 h, over which time it went from deep blue to reddish - purple in color. The reaction mixture was filtered over a pad of Celite, and the filtrate transferred to an NMR tube for in situ analysis. 1 H NMR (600 MHz, benzene - d 6 d, J = 7.4 Hz, 4H), 7.22 (t, J = 8.6 Hz, 6H), 6.92 6.86 (m, 7H), 3.73 (p, J = 6.9 Hz, 4H), 1.14 (d, 28H). 13 C[ 1 H] NMR (151 MHz, benzene - d 6 200.41. (Note: the full 13 C NMR spectrum could not be assigned in situ due to the presence of excess diphenylacet ylene, in addition to other impurities. However, this new resonance noted at >200 ppm is close to the observed 13 C resonance for the acetylenic carbon in the Os derivative of this molecule previously reported). 14 N NMR (36 MHz, THF - d 8 mpts to grow X - ray quality crystals from this reaction solution did not result in the isolation of a new complex.) Synthesis of Me 3 PNAr A scintillation vial was charged with 38 mg of PMe 3 (0.5 mmol, 1.2 equiv), 3 mL of THF, and a stir bar. At room temperat ure, 78 mg (0.4 mmol, 1 equiv) of N 3 Ar was added dropwise, as a solution in 1 mL THF, to the stirred PMe 3 solution. The reactions immediately began to produce bubbles. The reaction was allowed to stir for 4 h, and the volatiles removed in vacuuo, resulting in 90 mg (94%) of a powdery, pale yellow residue. The crude product was used without further purification. X - ray quality crystals can be grown from n - hex at - 35 °C. 712 1 H NMR (500 MHz, benzene - d 6 J = 7.6, 1.5 Hz, 2H), 7.08 (td, J = 7.5, 2.8 Hz, 1H), 3.58 (septet, J = 6.9 Hz, 2H), 1.34 (d, J = 6.9 Hz, 12H), 0.93 (d, J = 12.0 Hz, 9H). 13 C[ 1 H] NMR (126 MHz, benzene - d 6 31 P NMR (202 MHz, benzene - d 6 - 15 H 26 PN: C, 71.68, H, 10.43, N, 5.57; found, C, 71.25, H, 9.78, N, 5.61. Synthesis of cis - RuCl 2 (PPhMe 2 ) 4 A 35 mL pressure tube was charged with 0.506 g [RuCODCl 2 ] x (1.9 mmol, 1 equiv), a stir bar, and 3 mL toluene. The solution was stirred and to it was added 1.4 g PPhMe 2 (9.5 mmol, 5 equiv). The pressure tube was sealed and transferred from the glovebox to a 120 °C oil bath. The pressure tube was heated, with stirring, for 16 h, over which time the opaque, brown suspension turned an orangish - yellow. The pressure tube was removed from heat and cooled ambiently, resulting in the precipitation of copious amounts of yell ow solids, which was an assortment of various sized crystals. (Note, from this precipitate, X - ray quality crystals were obtained). When the pressure tube was cooled, it was returned to the glovebox. The mother liquor was decanted, and the solids dried unde r reduced pressure to yield 1.1 g (80%) of cis - RuCl 2 (PPhMe 2 ) 4 . Once the title compound is precipitated from toluene, it has poor solubility in NMR solvents, as well as most organic solvents. The complex demonstrates marked color changes when dissolved in DMSO - d 6 or MeOD, which seem to correspond to solvent reactivity. Consequently, adequate NMR spectra could not be obtained. The collected precipitate was found to be analytically pure by elemental analysis, without further treatment, and was used in subsequ ent reactions. The crystals obtained upon cooling were also structurally characterized. Elemental 32 H 44 P 4 Cl 2 : C, 53.05, H, 6.12, N, 0.00; found, C, 52.72, H, 6.29, N, 0.06. 713 Ru(NH - 2,4 - dimethyl - 6 - CH 2 - phenyl)(PPhMe 2 ) 4 (Ru9) A scintillation vial was loaded with 124 mg of cis - RuCl 2 (PPhMe 2 ) 4 (0.16 mmol, 1 equiv), a stir bar, and 5 mL of THF. In a separate vial, 48 mg of LiNHMes (0.34 mmol, 2.1 equiv) was dissolved in THF. This solution was added dropwise to the first solution, with stirring, at room temperature. Upon addition, the solution went from a pale yellow suspension to an intense red solution. The reaction was stirred for 8 h and the volatiles removed under reduced pressure. The residue was extracted with hexane and filt ered over celite until the filtrate came out colorless. The reddish - pink filtrate solution was concentrated under reduced pressure and stored at - 35 °C for 24 h to yield 46 mg (42 %) X - ray quality crystals. 1 H NMR (500 MHz, benzene - d 6 0 (t, J = 7.2 Hz, 8H), 7.04 (t, J = 7.6 Hz, 11H), 6.97 (s, 1H), 4.11 (s, 1H), 2.88 (s, 2H), 2.56 (s, 3H), 2.23 (s, 3H), 1.19 (d, J = 4.8 Hz, 24H). 13 C[ 1 H] NMR (126 MHz, benzene - d 6 J = 26.8 Hz), 130.64 (d, J = 10.4 Hz), 128.35, 127.98 (d, J = 3.5 Hz), 126.41, 120.84, 119.66, 31.98, 21.57, 18.85, 18.13. 31 P NMR (202 MHz, benzene - d 6 41 H 55 P 4 N: C, 62.58, H, 7.05, N, 1.78; found, C, 62.36, H, 7.16, N, 1.85. Ru(NAr)(PPhMe 2 ) 3 (Ru8) A scintillation vial was charged with 194 mg (1equiv, 0.25 mmol) of cis - RuCl 2 (PPhMe 2 ) 4 , a stir bar, and 8 mL of THF. To this stirred solution, a solution of 100 mg (2.1 equiv, 0.52 mmol) LiNHAr in 2 mL THF, was added dropwise at room temperature. The solutio n rapidly changed color from pale yellow to bright red. The solution was stirred for 16 h at room temperature and the volatiles were removed under vacuum. This yielded a dark red residue which was extracted with n - hexane and filtered over celite until the filtrate came off colorless. The filtrate was concentrated to 2 mL and chilled to - 35 °C to give 124 mg (67%) of amorphous crystals, which were not X - ray quality. 714 1 H NMR (500 MHz, benzene - d 6 J = 8.8, 4.5, 1.9 Hz, 6H), 7.22 (s, 3H), 7.08 (t, J = 7.5 Hz, 6H), 7.04 6.93 (m, 3H), 4.62 (septet, J = 6.9 Hz, 2H), 1.40 (d, J = 7.0 Hz, 12H), 1.39 1.36 (m, 18H). 13 C{ NMR (126 MHz, benzene - d 6 143.90 (m), 140.54 (q, J = 6.2 Hz), 130.59 (d d, J = 7.6, 3.7 Hz), 128.34, 128.14 (dd, J = 6.1, 3.2 Hz), 122.78, 120.49, 26.68, 23.99, 23.72 22.74 (m). 31 P NMR (202 MHz, benzene - d 6 36 H 50 P 3 N: C, 62.59, H, 7.30, N, 2.03; found, C, 62.13, H, 7.49, N, 2.11. Os2 · toluene The procedure outlined for the synthesis of Ru2 was applied to produce analogous Os2, utilizing 200 mg cis - OsCl 2 (PMe 3 ) 4 ( 0.34 mmol, 1 equiv), 62 mg azobenzene ( 0.34 mmol, 1 equiv) reduced with 4 equiv Li ( 10 mg, 1.4 mmol), and 5 mL THF. The reaction solution was dried under reduced pressure to give a brown residue, which was extracted with toluene and filtered over Celite until the filtrate was colorless. The toluene solution was concentrated to 2 mL and layered with n - hexane. Storing the layered solution at - 35 ° C yielded 110 mg (42%) of X - ray quality orange cryst als. Os3 A pressure tube was charged with 50 mg of Os2, 5 mL of THF and a stir bar. The tube was sealed and heated at 65 ° C for 12 h. After this time, the p ressure tube was retunred to the glovebox and the reaction solution dried under reduced pressure. This yielded a dark brown residue. The residue was dissolved in hexane and recrystallized at - 35 ° C to yield 21 mg of X - ray qualtiy brown - orange crystals of O s3. 1 H NMR (500 MHz, benzene - 8.39 (d, J = 5.7 Hz, 1H), 8.00 (d, J = 7.9 Hz, 1H), 7.37 (d, J = 7.6 Hz, 4H), 7.23 (dd, J = 11.0, 7.4 Hz, 2H), 7.05 (dd, J = 16.9, 6.8 Hz, 3H), 1.15 (d, J = 6.0 Hz, 9H), 0.98 (t, J = 3.6 Hz, 16H) . 31 P NMR (202 MHz, benzene - d 6 - 49.30 (d, J = 20.5 Hz), - 55.72 (t, J = 20.7 Hz). (xtal structure below). 715 Os(NAr) 2 (L) 2 Several compounds originally reported by Schrock, et. al. were synthesized following literature reports for the purpose of examiing the 14 N NMR spectra of the complexes. These data provided a useful comparative tool when analyzing the Ru analogues presente d here. These compounds include Os(NAr) 2 (O) 2 , Os(NAr) 2 (PMe 3 ) 2 2 - diphenylacetylene)Os(NAr) 2 . 22,23 These complexes matched reported 1 H, 13 C, 31 P and X - ray diffraction unit cell parameters provided in the literature. Computational Analysis All calculations were performed using the MSU HPCC facilities. DFT optimizations and single - point energy calculations were performed using Gaussian09, and dat a handling was done with GaussView software. For the claculations presented in section 8.3, the following parameters were used in order to optimize structures: initial optimization using PBEPBE functional and a split basis set (cc - pvtz - PP on Ru, 6 - 311g+(d, 2p) on all other atoms, CPCM solvent model using THF polarization). The optimized structures were then reoptimized with the same basis set assignment with the b3lyp functional. From the structures optimized with b3lyp, single - point energy calculations were for the transformation to each proposed species. The coordinates for each optimized, theoretical structure are provided below. Complex 1 N 1.376900 0.495500 0.677100 N 1.398000 - 0.5 00300 - 0.358600 C 2.182600 1.624300 0.528600 C 2.053900 - 2.691300 - 1.149800 H 1.155000 - 2.717100 - 1.750800 C 4.377200 - 2.614200 0.390100 H 5.278400 - 2.575900 0.991900 C 4.142300 - 3.720200 - 0 .436800 H 4.849500 - 4.537500 - 0.485200 C 2.965700 - 3.740600 - 1.205600 H 2.760700 - 4.585200 - 1.853500 C 2.810600 3.844700 1.354900 716 H 2.633700 4.664600 2.041200 C 1.991600 2.721700 1.410500 H 1.190700 2.666300 2.134500 C 3.255700 1.714000 - 0.396300 H 3.438700 0.885500 - 1.063500 C 3.472900 - 1.552800 0.459700 H 3.673800 - 0.708000 1.101900 C 2.276900 - 1.559400 - 0.314700 C 3.861900 3.926500 0.425400 H 4.498100 4.800700 0.386600 C 4.072100 2.848000 - 0.440800 H 4.882600 2.885300 - 1.159700 Ru - 0.607000 - 0.031200 0.004400 P - 2.031600 - 1.659800 - 1.106000 C - 3.931700 - 1.553600 - 1.091300 H - 4.286500 - 1.590700 - 0.064000 H - 4.355800 - 2.388600 - 1.648300 H - 4.251500 - 0.619000 - 1.546100 C - 1.778000 - 1.968600 - 2.968500 H - 2.048100 - 1.088100 - 3.541800 H - 2.406500 - 2.802100 - 3.280100 H - 0.736400 - 2.216700 - 3.159000 C - 1.848800 - 3.465500 - 0.536600 H - 0.797600 - 3.742700 - 0.523800 H - 2.386200 - 4.116000 - 1.225500 H - 2.266300 - 3.581500 0.458200 P - 0.550000 - 1.25380 0 2.147100 C - 0.147400 - 0.136500 3.622400 H 0.740000 0.436000 3.371700 H 0.034000 - 0.747400 4.506100 H - 0.974400 0.541000 3.821400 C 0.767400 - 2.586500 2.408400 H 0.673800 - 3.352100 1.642700 H 0.632400 - 3.030700 3.394400 H 1.752500 - 2.140000 2.330700 C - 2.039900 - 2.212600 2.853000 H - 2.978700 - 1.777600 2.528700 H - 1.990700 - 2.197800 3.941100 H - 1.991900 - 3.246100 2.519200 P - 2.383800 1.432800 0.850300 C - 3.733400 0.841700 2.058900 H - 4.306700 0.035700 1.607600 H - 4.400200 1.673900 2.281100 H - 3.287000 0.491000 2.985100 C - 3.538100 2.369200 - 0.341200 H - 2.9 57400 2.970500 - 1.033900 717 H - 4.189500 3.023000 0.237300 H - 4.146400 1.664300 - 0.901800 C - 1.770700 2.934700 1.846100 H - 1.246000 2.598100 2.736200 H - 2.620600 3.550300 2.139000 H - 1.087700 3 .522400 1.238100 P - 0.397500 1.342300 - 2.038500 C - 1.864500 1.565900 - 3.248000 H - 2.780300 1.172300 - 2.821000 H - 1.646700 1.042900 - 4.176600 H - 2.000200 2.623300 - 3.468300 C 0.085800 3.164100 - 1.809200 H - 0.709800 3.696100 - 1.291800 H 0.245200 3.620000 - 2.785900 H 0.995100 3.223600 - 1.220000 C 0.934100 0.806700 - 3.270400 H 1.892500 0.796200 - 2.765200 H 0.951100 1.499200 - 4.111600 H 0.711500 - 0.196500 - 3.624800 Complex 2 Ru 0.349600 0.437100 0.451000 N - 0.215700 - 0.897200 - 1.134200 C 3.498300 0.127200 0.015400 H 3.685600 0.853200 0.795700 C 2.158400 - 0.204000 - 0.310600 N 0.734300 - 1.549000 - 1.761000 C - 1.531100 - 1.272200 - 1.588600 C - 4.100300 - 1.953300 - 2.479800 H - 5.092800 - 2.215600 - 2.819800 C - 2.439100 - 0.270000 - 1.947100 H - 2.127500 0.761000 - 1.881800 C 2.016100 - 1.186800 - 1.340600 C - 3.716700 - 0.609900 - 2.397200 H - 4.408200 0.170700 - 2.684900 C 4.404300 - 1.438700 - 1.625500 H 5.257600 - 1.895300 - 2.109100 C 4.590400 - 0.467000 - 0.621600 H 5 .596900 - 0.182400 - 0.337200 C - 1.905300 - 2.619500 - 1.690600 H - 1.192700 - 3.386400 - 1.426100 C 3.112800 - 1.798900 - 1.984500 H 2.925100 - 2.538000 - 2.753300 C - 3.189900 - 2.955000 - 2.125200 718 H - 3.477700 - 3.995900 - 2.189200 P - 1.945700 1.049100 1.396500 C - 3.174700 2.267000 0.592200 H - 2.751600 3.267200 0.598100 H - 4.101100 2.266400 1.166200 H - 3.388900 1.971300 - 0.430600 C - 1.954800 1.846800 3.123500 H - 1.418200 1.224400 3.831900 H - 2.979600 1.991000 3.466000 H - 1.454200 2.811100 3.056800 C - 3.168200 - 0.386900 1.665700 H - 3.460300 - 0.775300 0.693100 H - 4.051500 - 0.035500 2.198700 H - 2.697900 - 1.186000 2.232200 P 0.658200 - 1.317200 2.137400 C 2.442300 - 1.652900 2.669600 H 3.031800 - 1.959400 1.812100 H 2.450800 - 2.435700 3.427200 H 2.866400 - 0.739500 3.079300 C 0.125200 - 3.071600 1.653100 H - 0.949900 - 3.094700 1.489300 H 0.390200 - 3.777800 2.439600 H 0.630000 - 3.348000 0.729900 C - 0.104000 - 1.189400 3.870600 H 0.271700 - 0.289200 4.351800 H 0.190800 - 2.060000 4.455100 H - 1.187500 - 1.144200 3.821300 P 0.681600 2.412900 - 0.934100 C - 0.536500 3.864600 - 1.012200 H - 1.489000 3.528100 - 1.411400 H - 0.130100 4.644200 - 1.656000 H - 0.684900 4.263400 - 0.011200 C 0.912300 2.082400 - 2.785700 H 1.730100 1.378100 - 2.916400 H 1.136800 3.011000 - 3.309600 H 0.002100 1.646000 - 3.190800 C 2.264100 3.376800 - 0.559700 H 2.2 08700 3.763500 0.455100 H 2.373300 4.201300 - 1.263500 H 3.116000 2.708800 - 0.640400 H 1.033900 1.385800 1.629800 Complex 3 Ru 0.002200 0.490300 - 0.024600 N - 1.632000 - 0.033900 - 0.773800 719 C - 2.5 55800 - 1.053500 - 0.697100 C - 3.780400 - 0.879700 0.015100 C - 2.412600 - 2.263400 - 1.439600 C - 3.421300 - 3.224600 - 1.459000 H - 3.275400 - 4.131400 - 2.034200 C - 4.784300 - 1.846200 - 0.024500 H - 5.702800 - 1.676400 0.525100 C - 4.619100 - 3.031600 - 0.755000 H - 5.400200 - 3.779400 - 0.778700 N 1.707500 0.123900 - 0.707000 C 2.710300 - 0.814900 - 0.606900 C 2.673200 - 2.043700 - 1.330500 C 3.909100 - 0.529900 0.113400 C 4.988600 - 1.411600 0.098400 H 5.884700 - 1.158500 0.653100 C 3.757800 - 2.918700 - 1.326500 H 3.692700 - 3.842900 - 1.888800 C 4.928700 - 2.617300 - 0.615300 H 5.768800 - 3.29840 0 - 0.620900 H 3.971500 0.401500 0.662000 H 1.783400 - 2.276600 - 1.901500 H - 1.502200 - 2.415000 - 2.005500 H - 3.924000 0.034100 0.577500 P - 0.036200 2.082500 - 1.978900 C - 0.090100 1.054100 - 3.557 600 H - 0.090600 1.699800 - 4.434900 H - 0.990600 0.447200 - 3.537300 H 0.781100 0.404800 - 3.573600 C 1.445100 3.202400 - 2.328200 H 2.345900 2.598800 - 2.250800 H 1.492000 4.008000 - 1.599600 H 1.368700 3.625900 - 3.328700 C - 1.503600 3.243600 - 2.251400 H - 1.504900 4.033600 - 1.503000 H - 2.418300 2.663200 - 2.157200 H - 1.453800 3.690200 - 3.243600 P 0.054900 - 1.350300 1.576600 C 1.520900 - 1.437200 2.769800 H 1.430700 - 2.330700 3.386300 H 2.440500 - 1.480300 2.192200 H 1.543700 - 0.560100 3.408800 C - 1.412300 - 1.495100 2.761800 H - 1.496000 - 0.600300 3.373000 H - 2.322200 - 1.610000 2.178500 H - 1.275300 - 2.360900 3.408300 C 0.101900 - 3.109800 0.889900 720 H - 0.775400 - 3.281600 0.273600 H 0.998400 - 3.232600 0.289000 H 0.114600 - 3.815800 1.719600 P - 0.189700 2.228500 1.724700 C 0.721400 3.841400 1.355500 H 0.573100 4.536200 2.180600 H 1.781900 3.633900 1.236900 H 0.337100 4.283300 0.441200 C - 1.971900 2.814900 1.943200 H - 2.576400 1.988900 2.308600 H - 2.001700 3.638100 2.655400 H - 2.360600 3.141200 0.982800 C 0.349800 1.961900 3.516200 H - 0.179400 1.122100 3.955600 H 1.420000 1.774900 3.545600 H 0.126100 2.861600 4.087400 Complex 5 Ru - 0.000000 0.000000 0.000000 N - 1.778300 0.053400 - 0.000300 C - 3.146300 0.078900 - 0.000200 C - 3.879100 0.081000 - 1.218000 C - 3.878500 0.103200 1.217700 C - 5.271200 0.128000 1.209900 H - 5.809000 0.146500 2.149200 C - 5.271700 0.105800 - 1.209900 H - 5.810000 0.107000 - 2.149100 C - 5.979400 0.129200 0.000100 H - 7.060600 0.148100 0.000200 N 1.778300 - 0.053400 0.000300 C 3.146300 - 0.078900 0.000200 C 3.879200 - 0.081000 1.218000 C 3.878500 - 0.103200 - 1.217800 C 5.271100 - 0.128000 - 1.210000 H 5.808900 - 0.146500 - 2.149300 C 5.271700 - 0.105800 1.209800 H 5.8 10100 - 0.106900 2.149000 C 5.979400 - 0.129200 - 0.000200 H 7.060600 - 0.148100 - 0.000300 H - 3.330500 0.063700 - 2.150200 H - 3.329500 0.103000 2.149800 H 3.329400 - 0.103100 - 2.149800 H 3.330600 - 0.063700 2.150200 P - 0.140000 - 2.473300 0.005700 721 C - 1.161000 - 3.223900 1.404900 H - 1.214100 - 4.306500 1.296300 H - 2.163300 - 2.803300 1.376200 H - 0.699200 - 2.973500 2.357300 C 1.472500 - 3.442200 0.117300 H 1.980400 - 3.185900 1.043900 H 2.112200 - 3.173400 - 0.719600 H 1.267700 - 4.511500 0.094700 C - 0.960500 - 3.231100 - 1.517100 H - 0.371700 - 2.985300 - 2.398100 H - 1.956900 - 2.809700 - 1.627700 H - 1.028700 - 4.313300 - 1.412000 P 0.140000 2.473300 - 0.005700 C 1.161000 3.223900 - 1.404900 H 1.214100 4.306500 - 1.296200 H 2.163300 2.803300 - 1.376200 H 0.699200 2.973500 - 2.357300 C - 1.472500 3.442200 - 0.117300 H - 1.980400 3.185900 - 1.043800 H - 2.112200 3.173400 0.719600 H - 1.267700 4.511500 - 0.094600 C 0.960500 3.231100 1.517200 H 0.371800 2.985300 2.3981 00 H 1.956900 2.809700 1.627700 H 1.028800 4.313300 1.412100 Complex 7 1 Ru1 0.1356 - 0.2803 - 0.2874 Ru 2 P2 1.1889 - 2.5286 - 1.1188 P 3 N3 1.8559 0.3107 - 0.8835 N 4 C4 2.5006 1.5038 - 0.6044 C 5 C5 3.9228 1.5678 - 0.7427 C 6 C6 1.8285 2.7263 - 0.2967 C 7 C7 2.5354 3.9223 - 0.1484 C 8 H8 1.9937 4.8394 0.0709 H 9 C9 4.6187 2.7619 - 0.5695 C 10 H10 5.700 8 2.7736 - 0.6736 H 11 C11 3.9322 3.9531 - 0.2721 C 12 H12 4.4764 4.8841 - 0.1453 H 13 C13 1.2924 - 2.3011 - 2.9942 C 14 H14 1.8559 - 3.1180 - 3.4572 H 15 H15 1.7853 - 1.3464 - 3.1915 H 16 H16 0.2814 - 2.2699 - 3.4114 H 17 C17 0.5212 - 4.3057 - 1.0275 C 722 18 H18 - 0.5199 - 4.3393 - 1.3570 H 19 H19 0.5874 - 4.6760 - 0.0007 H 20 H20 1.1196 - 4.9556 - 1.6752 H 21 C21 3.0015 - 2.8552 - 0.6867 C 22 H22 3.07 26 - 3.2605 0.3264 H 23 H23 3.5403 - 1.9074 - 0.7391 H 24 H24 3.4385 - 3.5704 - 1.3916 H 25 H25 4.4524 0.6529 - 0.9948 H 26 H26 0.7493 2.7101 - 0.2028 H 27 N27 - 1.0173 1.0163 - 0.9037 N 28 P28 - 2.0987 - 1.6684 0.3500 P 29 P29 0.7992 - 0.3290 2.0578 P 30 C30 - 2.1422 1.7837 - 0.7437 C 31 C31 - 3.1119 1.8830 - 1.7859 C 32 C32 - 2.3547 2.5654 0.4308 C 33 C33 - 4.2282 2.7069 - 1.6476 C 34 H34 - 2.9 572 1.3072 - 2.6936 H 35 C35 - 3.4686 3.3977 0.5449 C 36 H36 - 1.6208 2.5104 1.2288 H 37 C37 - 4.4186 3.4739 - 0.4857 C 38 H38 - 4.9539 2.7604 - 2.4549 H 39 H39 - 3.6016 3.9894 1.4470 H 40 H40 - 5.2855 4.1198 - 0.3886 H 41 C41 - 0.6002 - 0.0775 3.3005 C 42 H42 - 1.2726 - 0.9374 3.2942 H 43 H43 - 0.1753 0.0375 4.3030 H 44 H44 - 1.1621 0.8215 3.0372 H 45 C45 1.6143 - 1.8971 2.7326 C 46 H46 0.9817 - 2.7626 2.5207 H 47 H47 2.5859 - 2.0319 2.2508 H 48 H48 1.7581 - 1.8062 3.8145 H 49 C49 2.0216 0.9760 2.6607 C 50 H50 2.1951 0.8241 3.7309 H 51 H51 2.9658 0.8877 2.1208 H 52 H52 1.6119 1.97 38 2.4931 H 53 C53 - 3.5822 - 0.8016 1.1556 C 54 H54 - 3.2988 - 0.4110 2.1357 H 55 H55 - 3.9087 0.0286 0.5267 H 56 H56 - 4.4076 - 1.5122 1.2730 H 57 C57 - 2.9292 - 2.2912 - 1.2354 C 58 H58 - 3.8338 - 2.8627 - 1.0017 H 59 H59 - 3.1945 - 1.4289 - 1.8536 H 60 H60 - 2.2369 - 2.9222 - 1.7986 H 61 C61 - 2.0701 - 3.2549 1.3937 C 62 H62 - 1.2952 - 3.9339 1.0344 H 63 H63 - 1.8651 - 3.0044 2.4385 H 723 64 H64 - 3.0416 - 3.7 577 1.3358 H PMe 3 P 0.000500 - 0.000700 - 0.594000 C 1.553900 - 0.537500 0.275800 H 1.802100 - 1.556900 - 0.021400 H 2.378500 0.110500 - 0.022700 H 1.447700 - 0.499800 1.362100 C - 0.311100 1.613900 0.275100 H 0.447100 2.338800 - 0.022300 H - 1.285100 2.004000 - 0.021600 H - 0.289300 1.502800 1.361500 C - 1.243400 - 1.075600 0.275300 H - 2.249400 - 0.780300 - 0.024300 H - 1.095200 - 2.114600 - 0.02020 0 H - 1.160000 - 0.999700 1.361700 724 Spectral Data for Complexes Figure 8 . 18 1 H NMR spectrum of Ru(PhNNPh)(PMe 3 ) 4 · PhMe ( Ru2 ) in C 6 D 6. 725 Figure 8 . 19 13 C NMR spectrum of Ru(PhNNPh)(PMe 3 ) 4 · PhMe ( Ru2 ) in C 6 D 6 . 726 Figure 8 . 20 31 P NMR spectrum of Ru(PhNNPh)(PMe 3 ) 4 · PhMe ( Ru2 ) in C 6 D 6. 727 Figure 8 . 21 1 H NMR spectrum of Ru4 (in situ) in C 6 D 6 . 728 Figure 8 . 22 13 C NMR spectrum of Ru4 (in situ) in C 6 D 6 . 729 Figure 8 . 23 31 P NMR spectrum of Ru4 (in situ) in C 6 D 6 . 730 Figure 8 . 24 1 H NMR spectrum of Ru3 in C 6 D 6 . 731 Figure 8 . 25 13 C NMR spectrum of Ru3 in C 6 D 6 . 732 Figure 8 . 26 31 P NMR spectrum of Ru3 in C 6 D 6 . 733 Figure 8 . 27 1 H NMR of Ru5 in C 6 D 6 . 734 Figure 8 . 28 13 C NMR of Ru5 in C 6 D 6 . 735 Figure 8 . 29 31 P NMR of Ru5 in C 6 D 6 . 736 Figure 8 . 30 1 H NMR of Ru6 in C 6 D 6 . 737 Figure 8 . 31 13 C NMR of Ru6 in C 6 D 6 . 738 Figure 8 . 32 31 P NMR of Ru6 in C 6 D 6 . 739 Figure 8 . 33 1 H NMR of Ru7 in C 6 D 6 . 740 Figure 8 . 34 13 C NMR of Ru7 in C 6 D 6 . 741 Figure 8 . 35 31 P NMR of Ru7 in C 6 D 6 . 742 Figure 8 . 36 1 H NMR of photolysis reaction containing a mixture of Ru1 , Ru5 (starting material), and H 2 NAr (decomposition byproduct). RuN 4 P 3 (starting material) Ru(NAr) 2 (PMe 3 ) 2 NH 2 Ar (decomposition) 743 Figure 8 . 37 31 P NMR spectrum of Ru1 (after extraction and repeated recrystallization) in C 6 D 6. RuN 4 P 3 starting material Staudinger Product 744 Figure 8 . 38 13 C NMR spectrum of Ru1 in C 6 D 6 . 745 Figure 8 . 39 14 N NMR of Ru1 in C 6 D 6. 14 N NMR peak for Ru(NAr) 2 (PMe 3 ) 2 N 2 = *, ref (309.6 ppm) 746 Figure 8 . 40 QTOF - HRMS fragmentation patterns (top) calculated and (bottom) experimental for Ru1 . 747 Figure 8 . 41 1 H NMR spectrum of Ru(NAr) 2 2 - diphenylacetylene) containing diphenylacetylene (excess) and H 2 NAr impuri ties. Note: in the above spectrum, peaks assigned with multiplet values and denoted by green triangles correspond to the title complex formed in situ . The peaks noted match very closely the splitting pattern and chemical shifts of those reported for the osmium analogue of this complex by Schrock, et. al. 23 748 Figure 8 . 42 14 N NMR spectrum of Ru(NAr) 2 2 - diphenylacetylene) in C 6 D 6 (in situ). 749 Figure 8 . 43 1 H NMR of Ru8 in C 6 D 6 . 750 Figure 8 . 44 13 C NMR of Ru8 in C 6 D 6 . 751 Figure 8 . 45 31 P NMR of Ru8 in C 6 D 6 . 752 Figure 8 . 46 1 H NMR of Ru9 in C 6 D 6 . 753 Figure 8 . 47 13 C NMR of Ru9 in C 6 D 6 . 754 Figure 8 . 48 31 P NMR of Ru9 in C 6 D 6 . 755 Figure 8 . 49 1 H NMR of ArNPMe 3 in C 6 D 6 . 756 Figure 8 . 50 13 C NMR of ArNPMe 3 in C 6 D 6 . 757 Figure 8 . 51 31 P NMR of ArNPMe 3 in C 6 D 6 . 758 Figure 8 . 52 1 H NMR of Os3 in C 6 D 6 . 759 Figure 8 . 53 31 P NMR of Os3 in C 6 D 6 . 760 Figure 8 . 54 14 N NMR of Os(NAr) 2 (O) 2 . 761 Figure 8 . 55 14 N NMR of Os(NAr) 2 (PMe 3 ) 2 . 762 Figure 8 . 56 14 N NMR of Os(NAr) 2 2 - diphenylacetylene). 763 UV - Vis Spectroscopy of Several Ru Complexes Figure 8 . 57 Ru1 (0.000188 M in THF). 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 12000 17000 22000 27000 32000 37000 Axis Title Wavenumber (cm - 1 ) Ru1 UV - Vis: M - 1 cm - 1 ) Vs. Wavenumber (cm - 1 ) 764 Figure 8 . 58 Ru2 (0.00023 M in THF). 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 13000 15000 17000 19000 21000 23000 25000 27000 29000 31000 Wavenumber (cm - 1 ) Ru2 UV - Vis: M - 1 cm - 1 ) Vs. Wavenumber (cm - 1 ) 765 Figure 8 . 59 Ru4 (0.00019 M in THF). 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 13000 15000 17000 19000 21000 23000 25000 27000 29000 31000 Wavenumber (cm - 1 ) Ru4 UV - Vis: M - 1 cm - 1 ) Vs. Wavenumber (cm - 1 ) 766 Figure 8 . 60 Ru5 (0.000203 M in THF). -100 900 1900 2900 3900 4900 5900 6900 7900 12000 17000 22000 27000 32000 Wavenumber (cm - 1 ) Ru5 UV - Vis: (M - 1 cm - 1 ) Vs. Wavenumber (cm - 1 ) 767 Figure 8 . 61 Ru6 (0.00030 M in THF). 0 1000 2000 3000 4000 5000 6000 12500 14500 16500 18500 20500 22500 24500 26500 28500 30500 32500 Axis Title Wavenumbers (cm - 1 ) Ru6 UV - Vis: (M - 1 cm - 1 ) Vs. Wavenumber (cm - 1 ) 768 Figure 8 . 62 Ru7 (0.00031 M in THF). 0 1000 2000 3000 4000 5000 6000 13000 15000 17000 19000 21000 23000 25000 27000 29000 31000 Axis Title Wavenumbers (cm - 1 ) Ru7 UV - Vis: (M - 1 cm - 1 ) Vs. Wavenumber (cm - 1 ) 769 Figure 8 . 63 Photochemical irradiation setup utilizing a mercury arc lamp. 770 Crystallographic Data The following molecules have been characterized by single crystal X - ray crystallography, and their structures deposited in the Cambridge Structural Database: Ru1 , Ru5 - 7 , Ru9 , P(NAr)Me 3 and cis - RuCl 2 (PPhMe 2 ) 4 (CCDC 1895000 - 05 and 1895272). Additional molec ular structures have been collected for several other compounds included above, but their structures not deposited in the Cambridge Structural Database. Information about these structures is given below. The .cif files have been added to the MSU Structural Database managed by Dr. Staples. 771 Ru2 Figure 8 . 64 Crystal data and structure refinement for p21c. Identification code p21c Empirical formula C 31 H 54 N 2 P 4 Ru Formula weight 679.71 Temperature/K 173.15 Crystal system monoclinic Space group P2 1 /c a/Å 10.1433(17) b/Å 26.868(4) 772 c/Å 12.584(2) 90 98.261(3) 90 Volume/Å 3 3393.8(9) Z 4 calc g/cm 3 1.330 - 1 0.673 F(000) 1432.0 Crystal size/mm 3 0.174 × 0.083 × 0.05 Radiation 3.604 to 50.784 Index ranges - - - Reflections collected 28322 Independent reflections 6224 [R int = 0.1249, R sigma = 0.1130] Data/restraints/parameters 6224/0/367 Goodness - of - fit on F 2 1.026 R 1 = 0.0602, wR 2 = 0.1260 Final R indexes [all data] R 1 = 0.1153, wR 2 = 0.1520 Largest diff. peak/hole / e Å - 3 0.93/ - 0.48 773 Ru3 Figure 8 . 65 Crystal data and structure refinement for early_a. Identification code early_a Empirical formula C 10.5 H 18.5 NP 1.5 Ru 0.5 Formula weight 255.76 Temperature/K 173.0 Crystal system triclinic Space group P - 1 a/Å 9.1159(8) b/Å 11.8325(11) 774 c/Å 12.4414(10) 77.950(5) 79.242(5) 70.487(6) Volume/Å 3 1227.04(19) Z 4 calc g/cm 3 1.3844 - 1 7.074 F(000) 534.9 Crystal size/mm 3 0.189 × 0.12 × 0.089 Radiation range for data collection/° 7.32 to 144 Index ranges - - - Reflections collected 13254 Independent reflections 4445 [R int = 0.0550, R sigma = 0.0539] Data/restraints/parameters 4445/1/257 Goodness - of - fit on F 2 1.026 R 1 = 0.0460, wR 2 = 0.1123 Final R indexes [all data] R 1 = 0.0571, wR 2 = 0.1201 Largest diff. peak/hole / e Å - 3 1.70/ - 0.57 775 Os2 Figure 8 . 66 Crystal data and structure refinement for KA_OsBisImido. Identification code KA_OsBisImido Empirical formula C 31 H 54 N 2 OsP 4 Formula weight 768.91 Temperature/K 173.15 Crystal system monoclinic Space group P2 1 /c a/Å 10.1721(10) b/Å 26.872(3) 776 c/Å 12.6014(13) 90 98.3670(10) 90 Volume/Å 3 3407.8(6) Z 4 calc g/cm 3 1.4985 - 1 3.954 F(000) 1558.0 Crystal size/mm 3 N/A × N/A × N/A Radiation collection/° 3.04 to 50.74 Index ranges - - - Reflections collected 25327 Independent reflections 6213 [R int = 0.0618, R sigma = 0.0564] Data/restraints/parameters 6213/0/351 Goodness - of - fit on F 2 0.887 Final R R 1 = 0.0296, wR 2 = 0.0637 Final R indexes [all data] R 1 = 0.0425, wR 2 = 0.0727 Largest diff. peak/hole / e Å - 3 1.96/ - 0.85 777 Os3 Figure 8 . 67 Crystal data and structure refinement for smalltwin5. Identification code smalltwin5 Empirical formula C 21 H 37 N 2 OsP 3 Formula weight 600.63 Temperature/K 173.15 Crystal system triclinic Space group P - 1 a/Å 9.2309(12) b/Å 11.8060(16) 778 c/Å 11.8841(16) 78.4871(16) 74.9609(16) 81.8801(17) Volume/Å 3 1220.1(3) Z 2 calc g/cm 3 1.635 - 1 5.431 F(000) 596.0 Crystal size/mm 3 0.225 × 0.209 × 0.132 Radiation 3.536 to 50.956 Index ranges - - Reflections collected 5185 Independent reflections 5185 [R int = ?, R sigma = 0.0515] Data/restraints/parameters 5185/258/267 Goodness - of - fit on F 2 1.086 R 1 = 0.0440, wR 2 = 0.0938 Final R indexes [all data] R 1 = 0.0544, wR 2 = 0.0983 Largest diff. peak/hole / e Å - 3 1.87/ - 1.75 779 REFERENCES 780 REFERENCES (1) Danopoulos, A. A.; Wilkinson, G.; Hussain - Bates, B.; Hursthouse, M. B. Synthesis and X - ray crystal structure of trans - bis(2,6 - diisopropylphenylimido) bis - (trimethylphosphino)ruthenium(IV): The first structural determination of a terminal imido ruthenium co mpound. Polyhedron 1992, 11 (22), 2961. (2) Burred, A. K.; Steedman, A. J. ([small eta]6 - p - Cymene)Ru[triple bond, length half m - dash]N(2,4,6 - tri - tert - butylphenyl): a monomeric ruthenium(II) complex containing a terminal imido ligand. J. Chem. Soc., Chem. C ommun. 1995 , DOI:10.1039/C39950002109 10.1039/C39950002109(20), 2109. (3) Burrell, A. K.; Steedman, A. J. Synthesis, Structure, and Reactivity of Ruthenium(II) Terminal Imido Complexes. Organometallics 1997, 16 (6), 1203. (4) Singh, A. K.; Levine, B. G.; S taples, R. J.; Odom, A. L. A 4 - coordinate Ru(II) imido: unusual geometry, synthesis, and reactivity. Chem Commun (Camb) 2013, 49 (92), 10799. (5) Aldrich, K. E. F., B. S.; Singh, A. K.; Staples, R. J.; McCracken, J.; Levine, B.; Smith, M. R.; Odom, A. L. . Inorganic chemistry 2019 . (6) Smieja, J. A.; Shirzad, K.; Roy, M.; Kittilstved, K.; Twamley, B. Unusual Bis(imido)metalloporphyrins. The synthesi s of M(4 - R - TPP)(NC6H4NO2)2 (M=Ru, R=H; M=Os, R=Cl) and structural study of the osmium derivative. Inorganica Chimica Acta 2002, 335 , 141. (7) Law, S. - M.; Chen, D.; Chan, S. L. - F.; Guan, X.; Tsui, W. - M.; Huang, J. - S.; Zhu, N.; Che, C. - M. Ruthenium Porphyrin - Conjugated Arylamide and Arylimide Ligands. 2014, 20 (35), 11035. (8) Au, S. - M.; Huang, J. - S.; Yu, W. - Y.; Fung, W. - H.; Che, C. - M. Aziridination of Alkenes and Amidation of Alkanes by Bis(tosylimido)ruthenium(VI) Porphyrins. A Mechanistic Stu dy. Journal of the American Chemical Society 1999, 121 (39), 9120. (9) Leung, S. K. - Y.; Tsui, W. - M.; Huang, J. - S.; Che, C. - M.; Liang, J. - L.; Zhu, N. Imido Bond Dissociation Energies, and RuVI/V Reduction Potentials. Journal of the American Chemical Society 2005, 127 (47), 16629. (10) Zardi, P.; Pozzoli, A.; Ferretti, F.; Manca, G.; Mealli, C.; Gallo, E. A mechanistic investigation of the ruthenium porphyrin catalysed aziridination of olefins by aryl azides. Dalton Transactions 2015, 44 (22), 10479. (11) Au, S. - M.; Fung, W. - H.; Huang, J. - S.; Cheung, K. - K.; Che, C. - M. Synthesis, Spectroscopic Properties, and Reactivities of Bis(tosylimido)osmium(VI) Porphyrin C omplexes. 781 X - ray Crystal Structure of [OsVI(TPP)(NSO2C6H4 - pCH3)2] (TPP = Tetraphenylporphyrinato). Inorganic chemistry 1998, 37 (26), 6564. (12) Leung, S. K. - Y.; Huang, J. - S.; Zhu, N.; Che, C. - M. Reactivity of Dioxoosmium(VI) Porphyrins toward Arylhydrazine . Isolation of Hydrazidoosmium and Amidoosmium Porphyrins. Inorganic chemistry 2003, 42 (22), 7266. (13) Exploration. Journal of the American Chemical Society 2002, 124 (13), 3 206. (14) Cramer, S. A.; Hernandez Sanchez, R.; Brakhage, D. F.; Jenkins, D. M. Probing the role of an FeIV tetrazene in catalytic aziridination. Chem. Commun. 2014, 50 (90), 13967. (15) Hohenberger, J.; Ray, K.; Meyer, K. The biology and chemistry of high - valent iron oxo and iron nitrido complexes. Nature Communications 2012, 3 , 720. (16) Mehn, M. P.; Peters, J. C. Mid - to high - valent imido and nitrido complexes of iron. J. Inorg. Biochem. 2006, 100 (4), 634. (17) Mahy, J. P.; Battioni, P.; Mansuy, D. Formation of an iron(III) - porphyrin complex with a nitrene moiety inserted into an iron - nitrogen bond during alkene aziridination by (tosylimidoiodo)benzene catalyzed by iron(III) porphyrins. Journal of the American Chemical Society 1986, 108 (5), 1079. (18) Clifford, A. F., Kobayashi, C. S. 130th National Meeting of the American Chemical Society 1956 , 50R. (19) Nugent, W. A., Mayer, J. M. Metal - Ligand Multiple Bonds. Wiley Interscience 1987 . ( 20) Patrick, D. W.; Truesdale, L. K.; Biller, S. A.; Sharpless, K. B. Stereospecific vicinal oxyamination of olefins by alkylimidoosmium compounds. The Journal of Organic Chemistry 1978, 43 (13), 2628. (21) Herranz, E.; Sharpless, K. B. Improvements in the osmium - catalyzed oxyamination of olefins by chloramine - T. The Journal of Organic Chemistry 1978, 43 (12), 2544. (22) Wolf, J. R.; Bazan, G. C.; Schrock, R. R. Exchange of oxo ligands in osmium tetroxide with imido ligands in bis(arylimido)bis(tert - butoxo) molybdenum complexes, Mo(NAr)2(O - tert - Bu)2. A facile route to Os(NAr)2O2 and Os(NAr)3O and osmium(IV) complexes of the type Os(NAr)2L2 (NAr = N - 2,6 - C6H3 - iso - Pr2; L = a phosphine). Inorg. Chem. 1993, 32 (19), 4155. (23) Schofield, M. H.; Kee, T. P.; Anhaus, J. T.; Schrock, R. R.; Johnson, K. H.; Davis, W. M. Osmium imido complexes: synthesis, reactivity, and SCF - X.alpha. - SW electronic structure. Inorg. Chem. 1991, 30 (19), 3595. (24) LaPointe, A. M.; Schrock, R. R.; Davis, W. M. Imido/Oxo Exchange between Os mium and Tantalum as a Route to Os(NAr)2R2 and OsO(NAr)R2 Complexes (NAr = N - 2,6 - i - 782 Pr2C6H3; R = CH2CMe3, CH2CMe2Ph, CH2SiMe3) and Attempts To Induce .alpha. - Hydrogen Abstraction To Give Alkylidene Complexes. Organometallics 1995, 14 (6), 2699. (25) Wang, L .; Hu, L.; Zhang, H.; Chen, H.; Deng, L. Three - Coordinate Iron(IV) Bisimido Complexes with Aminocarbene Ligation: Synthesis, Structure, and Reactivity. J. Am. Chem. Soc. 2015, 137 (44), 14196. (26) Ni, C.; Fettinger, J. C.; Long, G. J.; Brynda, M.; Power, P. P. Reaction of a sterically encumbered iron(i) aryl/arene with organoazides: formation of an iron(v) bis(imide). Chem. Commun. 2008 , DOI:10.1039/B810941A 10.1039/B810941A(45), 6045. (27) Hansert, B.; Vahrenkamp, H. Preparation, structure, and some react ions of the evasive - Ph2N2)]. Journal of Organometallic Chemistry 1993, 459 (1), 265. (28) Kiernicki, J. J.; Higgins, R. F.; Kraft, S. J.; Zeller, M.; Shores, M. P.; Bart, S. C. Elucidating the Mechanism of Uranium Mediated Di Inorg. Chem. 2016, 55 (22), 11854. (29) Lin, K. - M.; Wang, P. - Y.; Shieh, Y. - J.; Chen, H. - Z.; Kuo, T. - S.; Tsai, Y. - C. Reductive N - diketiminate. New Journal of Chemistry 2010, 34 (8), 1737. (30) Gehrmann, T.; Lloret Fillol, J.; Wadepohl, H.; Gade, L. H. Synthesis, Characterization, and Thermal Rearrangement of Zirconium Tetraazadienyl and Pentaazadienyl Complexes. Organometallics 2012, 31 (12), 4504. (31) Mason, J. Multinuclear NMR ; Plenum Press: New York [etc.], 1987. (32) Mason, J. Nitrogen nuclear magnetic resonance spectroscopy in inorganic, organometallic, and bioinorganic chemistry. Chem. Rev. 1981, 81 (3), 205. (33) Danopoulos, A. A.; Wilkinson, G.; Sweet, T. K. N.; Hursthouse, M. B. Reactio ns of imido complexes of iridium, rhodium and ruthenium. J. Chem. Soc., Dalton Trans. 1996 , DOI:10.1039/DT9960003771 10.1039/DT9960003771(19), 3771. (34) Heyduk, A. F.; Blackmore, K. J.; Ketterer, N. A.; Ziller, J. W. Azide Addition To Give a Tetra - azazirc onacycle Complex. Inorg. Chem. 2005, 44 (3), 468. (35) Cowley, R. E.; Bill, E.; Neese, F.; Brennessel, W. W.; Holland, P. L. Iron(II) Complexes with Redox - Active Tetrazene (RNNNNR) Ligands. Inorg. Chem. 2009, 48 (11), 4828. (36) Singh, A. K.; Levine, B. G. ; Staples, R. J.; Odom, A. L. A 4 - coordinate Ru(ii) imido: unusual geometry, synthesis, and reactivity. Chem. Commun. 2013, 49 (92), 10799. (37) Park, J. Y.; Kim, Y.; Bae, D. Y.; Rhee, Y. H.; Park, J. Ruthenium Bisammine Complex and Its Reaction with Aryl Azides. Organometallics 2017, 36 (18), 3471.