LIBRAR y Michigan State University This is to certify that the thesis entitled PREPARATION AND CHARACTERIZATION OF HYBRID PHASE CATALYSTS presented by Shiu—Chin H. Su has been accepted towards fulfillment of the requirements for Ph . D . degree in Organl c Cheml stry 4 / /[,/ 7‘ d7///\ 1/ I L, iVIajor professor Date 0-7639 I] . u -1 .lllllll IfffF...fl-.F1l’ltll .g.‘ Kit-.4311..L .... ...............t ....-._...... ... froéfirimi .. .H.....1L..- .H 5...... © 1978 SH! U-CHIN HUANG SU ALL RI GHTS RESERVED PREPARATION AND CHARACTERIZATION OF HYBRID PHASE CATALYSTS By Shiu—Chin H. Su A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1978 Lev/Q 4. 99;. ABSTRACT PREPARATION AND CHARACTERIZATION OF HYBRID PHASE CATALYSTS By Shiu-Chin H. Su Polymer-supported phosphines have been used as ligands in hydrogenationl, hydrosilylation2 or hydroformylation3 reactions involving various transition metals. In all cases, the insoluble catalyst could be filtered from the reaction mixture and reused without appreciable loss of activity. In most cases, styrene-divinylbenzene c0polymer "beads” have served as the polymer support. An easily-controlled and convenient route for functionalizing polystyrene polymers was deve10pedu. Styrene crosslinked with divinylbenezene (2 or 20%) was converted to polystyrene-lithium by treatment with a 1:1 complex of butyllithium-tetramethylethylenediamine in cyclohexane. The lithiated polymer was allowed to react with chlorodiphenylphosphine to produce polystyryldiphenyl— phosphine. p P autmrm cw¢2 © cyfloxone" THF ' Li 15/ \¢ Shiu-Chin H. Su Cumene was used as a low molecular weight model 1H and molecule for examining the position of metalation. 13C NMR revealed that only meta and para trimethylsilane substituted cumene derivatives were obtained. Glc analysis indicated that the ratio of meta/para was the statistical, 2:1. The radial distribution of phosphorus and the percent phosphorus substitution could be varied over a wide range by controlling the reaction time and temperature. Tris(triphenylphosphine)chlororhodium polymer—supported catalysts were prepared from these polymeric ligands and gave a rate of reduction of cyclohexene of 7.65 ml/min./mmol of Rh which is better than that observed with a similar batch of material prepared by the bromination routella Lithiated polymer was reacted with chlorotrimethyltin, tetrachlorotin, diethylaluminium chloride and ethylaluminum dichloride to give respectively polymer-supported trimethyl— tin(IV), polymer—supported trichlorotin(IV), polymer- supported diethylaluminium and polymer-supported ethylaluminium chloride. Phosphorus-31 NMR spectroscopy provides a useful tool for characterizing the polymeric organOphosphorus ligands and their transition metal catalysts. The 31P NMR measurements of phosphinated beads can be performed in a variety of swelling solvents, such as THF, benzene, toluene, xylene, chlorinated hydrocarbon, ethylacetate, acetone, diethyl ether, chlorobenzene and nitrobenzene. Since Shiu—Chin H. Su toluene has a wide temperature range for the liquid state, it served as the main swelling solvent in these studies. The chemical shift of polymeric triphenylphosphine (6 5.93 ppm vs. 85% HBPOQ) is essentially the same as that of triphenylphosphine. Polymeric triphenylphosphine oxide (6 -24.5 ppm vs. 85% H3P04) is very easily detected by 31? NMR. The extent of phosphination can be routinely measured by employing a sample of known phosphorus concentration and an external reference (tri—n-butylphosphine). The 31P-NlVlR peak of the metal complexed polystyryl— phosphine is too broad to be observed. Comparison of the relative peak areas of the external reference and the rhodium complexed copolymer revealed the concentration of uncomplexed polymeric phosphine. The maximum coordination number per rhodium atom by polymeric phosphines (Pj—PC/Rh) 31 was calculated from the P-NMR measurements. These Pj-PC/Rh values revealed multiple binding of polymeric phosphine to the rhodium(I). (Pj—Pfl2)3RhCl was prOposed if the catalyst was prepared by equilibration with RhCl(P¢3)3. However, in the case of [Rh(cyclooctene)201]2 equilibration, the presence of dimer was indicated. The Pj-PC/Rh values of samples prepared from [Rh(COE)2Cl]2 equilibration were increased to 3 or greater by treatment with P¢3 at high temperature. Both electron microprobe analysis and scanning electron microprobe data support the conclusion that the polymeric phosphine break the chlorine bridge to form a monomer. Shiu-Chin H. Su 31F spin-lattice relaxation times(Tl) of polymeric triphenylphosphine (1.9 seconds) were obtained by applying a 180°-r—90° pulse sequence. The reorientational correla— —10 seconds at tion time (T0) was estimated to be 5.7 x 10 room temperature. Temperature dependence of the 31? T1 of the polymeric PCB indicated that dipole—dipole interactions provide the dominant relaxation mechanism within the studied temperature range. In sharp contrast to Regen's5 results, the rotational correlation times (T0) of polymeric PCB calculated from the 31F T1 are not sensitive to the swelling properties of the solvents. The temperature dependence of 31P-NMR spectra of polymer—supported RhCl(CO)(P¢ prepared by equilibrating p2. polystyrylphosphine with RhCl(CO)(Pfl3)2 or [RhCl(CO)2]2, indicated that free polymeric PCB exchanges with complexed polystyrylphosphine. REFERENCES l.(a) R.H. Grubbs and L.C. Kroll, J. Am.Chem. Soc., 93, 3062 (1971). (b) E.M. Sweet, Ph.D. Dissertation, Michigan State University, 1976. 2. M. Capka, P. Svoboda, M. Creny and J. Hetfleje, Tetrahedron Letters, 4787 (I971) 3. W.O. Hagg and D.D. Whitehurst—Proceedings of the 5th International Congress on Catalysis, Amsterdam 1972. 4. Robert H. Grubbs and Shiu—Chin H. Su, J. Organometal. Chem., 122, 151 (1976). 5. S.L. Regen, J. Am. Chem. Soc., 2g, 5275 (1974). TO MY PARENTS, MY HUSBAND, MY BROTHERS and MY SISTER ii ACKNOWLEDGEMENTS I wish to express my deepest gratitude to my preceptor, Dr. R.H. Grubbs for his enthusiastic help, encouragement, suggestions and guidence throughout the course of this study. Thanks also extended to members of my committee, Dr. W.H. Reush, my second reader, Dr. M.T. Rogers and Dr. T.J. Pinnavaia. Financial support from U.S. Army, Eli Lilly, Dreyfus foundation and Department of Chemistry, Michigan State University has been gratefully acknowledged. My appreciation goes to Dr. E.M. Sweet for his fruitful discussion and his importing the inert atmosphere techniques which have made this work much less frustrating. In addition, I would like to thank Mr. V. Shull and Mr. S. Flegler for their help with the electron microprobe works. Dr. D. Gill and Dr. A. Miyashita for their moral support and discussion during this study. Thanks are also extended to the past and present members of "the grubbs group" for making my stay here a very pleasant and memorable experience. and Nb. Gwen Goretsas for her cordial concern during the preparation of the draft of this dissertation. Special thanks are given to my Host Family, Janice, Howard, Sally and Mike Cummings, for their help in many ways and their concern during the wedding. Finally, I would like to express my special appreciation to Biing-Ming, my husband, whose understanding wmdassistanoe with the NMR instruments made this dissertation possible. He suffered much loneliness, especially during the writing of this dissertation. I also wish to extend my appreciation to my parents and my brothers for their encouragement and support throughout the long course of formal education. iii TABLE OF CONTENTS PART I THE PREPARATION OF POLYSTYRENE-ATTACHED PHOSPHINE LIGANDS, ORGANOMETALLIC COMPOUNDS AND HYBRID PHASE CATALYSTS .................................. I. INTRODUCTION ................................. The Preparation of Polymeric Organophosphorus Ligands for Catalyst Attachment ............. II. RESULTS AND DISCUSSION ...................... The Preparation of Polymeric Organophosphorus Ligands for Catalyst Attachment ............. Studies Regarding the Position of Metalation. The Preparation of Phenyl—Supported Titanocene Dichloride ....................... The Preparation of Polymer-Supported Organometallic Compounds .................... III. EXPERIMENTAL ................................ Materials ................................... Purifications ............................... Preparations ................................ Biscyclooctenerhodium(I) Chloride Dimer, [RhCl(COE) 32 .................. u-Dichlorotetraethernedirhodium(I) ..... u—Dichlorotetracarbonyldirhodium(I) ..... Tris(triphenylphosphine)chlororhodium(I) 2—Cyclopentene—l-one .................... Cyclopentadienyltitanium Trichloride Preparation of Polymeric Organophosphorus Ligands for Catalyst Attachment ............. Lithiation .............................. Phosphination ......... . ......... . ....... Metalation .............................. Batch 1 ............................... Batch 10 .............................. Preparation of Polymer—Supported Titanocene Dichloride .................... . ........... Batch A ............................... Page TABLE OF CONTENTS--continued Batch B ................................ Preparation of Polymer-Supported Trimethyltin Trimethyltin(IV) ........................... Preparation of Polymer- Supported Chlorotin(IV) .............................. Preparation of Polymer-Supported Aluminum Compounds ......................... Preparation of Beads for MicrOprobe Analysis Determination of Elemental Radial distribution within a Bead .............................. Hydrogenation ............................... Supported Wilkinson's Catalyst Hydrogenation Polymer—Supported Titanocene Hydrogenation . Studies Regarding the Position of Metalation Metalation of Cumene ....................... Quenching of the Lithiated Cumene .......... PART II CHARACTERIZATION OF POLYSTYRE-RHODIUM(I) COMPLEXES BY PHOSPHORUS-31 NMR SPECTROSCOPY AND ELECTRON MICROPROBE ANALYSIS ............................... I. INTRODUCTION ................................. (A) Concepts in NMR .......................... Nuclear Magnetism .......................... Nuclear Precession ......................... Nuclear Magnetic Energy Level .............. Nuclear Magnetic Resonance ................. Relaxation and Saturation .................. Chemical Shift ............................. (B) Introduction to Fourier Transform NMR .. SE) General .................................. P- NMR of Transition Metal Phosphine Complexes ................................ NMR Study on Polymers, Exchange Resins and Gels ..................................... Analytical Problems in Solid— Phase Synthesis & Hybrid Phase Catalysts ............. .. Bridge— Splitting Reactions of Rhodium Carbonyl Chloride ........................ II. RESULTS AND DISCUSSION ...................... Phosphorus-31 NMR Spectra of Polystyrylphosphine ........................ Page 60 61 61 61 62 62 63 66 68 73 74 77 79 81 BL- 84 TABLE OF CONTENTS—-continued Page The Detection of Polymeric Phosphine Oxide by P— 31 NMR .. ............................ 88 Quantitative Analysis of the Content of the Polymeric Phosphines ....................... 88 The P-3l NMR Spectra of Polystyrylphosphine Metal Com lexes ............................ 91 RhC1(P ) Studies ....................... 96 [Rh(cyclooctene)2 C1] Studies ............ 96 Characterization by 2Elegtron Microprobe Analysis ................................... 106 The Effect of Triphenylphosphine on Polymer Bound Ligand Coordination .................. 110 P—31 Spin—Lattice Relaxation Study of Polymeric Triphenylphosphine ......... ...... 118 Phosphine Exchange in Carbonylchloro— bis(triphenylphosphine)rhodium(I) .......... 130 Phosphine Exchange in Polymer- Supported RhCi(co)(P¢3)2 ............................. 133 III. EXPERIMENTAL ................................. 147 Materials .................................... 147 Instrumentation ............................. 148 (I) DA— 60 NMR Spectrometer ............... 148 The External Lock .................... 150 Temperature Control and Measurement .. 150 Sampling and Referencing ............. 151 (II)Bruker HFX-9O NMR Spectrometer ....... 151 The Measurements of Spin-Lattice Relaxation Times . ...... . ..... . ..... 153 Preparation of Reference Sample .............. 156 General Preparation of the Samples for P-31 NMR Measurements ..... . ..................... 157 Preparation of Polymeric Triphenylphosphine for Chelation Studies ...................... 158 Preparation of Polystyrylphosphine Metal Complexes for Phosphorus— 31 Studies ......... 159 RhCl(P¢3)3 studies ....................... 159 [Rh(cos)201]2 studies-— Series II-V ...... 159 Phosphine Exchange in RhC1(CO)(P¢3 )2 ....... .. 160 Preparation of Polymer- Supported RhC1(CO)(P¢ )2 for Phosphine Exchange Studies ........ .3 164 Scanning Electron MicrOprobe Analysis ........ 165 The Degree of Swelling Values. q ... .......... 165 REFERENCES ........................................ 166 vi LIST OF TABLES TABLE . Effect ot reaction conditions on the phosphorus substitution and the radial distribution of phosphorus ......... ... ......... .... .......... .. Rates of hydro enation for samples of polymer— bound RhC1(P;Zf3 3 of various preparation ........ Chemical Shifts in 13C-NMR Spectra of Cumene and Trimethylsi1y1-substituted Cumene ..... ..... . Hydrogenation of Cyclohexene by Polymer— Supported Cp2T1C12 at room temperature......... . Hydrogenetion of Olefin by Polymer—Supported IO. 11 12. Cp2T1C12 as a function of Ti loading.... ...... . The Preparation of Polymer—Supported Aluminum Compounds ..... ........ ......................... The complexation of polymeric triphenylphosphine with rhodium(I).——RhC1(P03)3 Equilibration ..... The complexation of polymeric triphenylphosphine with rhodium(I) ——[:RhC1(C0E)2]2 Equilibration (1.237 mmole phosphine/gbeads) ................. The complexation of polymeric triphenylphosphine with rhodium(I) --[RhCl(COE)2]2 Equilibration (0.878 mmole phosphine/gbeads)................. The complexation of polymeric triphenylphosphine with rhodium(I) —-[RhC1(C0E)2]2 Equilibration (0.560 mmole phosphine/gbeads)...... ..... . ..... The complexation of polymeric triphenylphosphine with rhodium(I) ——[RhC1(COE) 12 Equilibration (0.379 mmole phosphine/gbeads)................. The P/Rh ratio calculated frOm the electron microprobe data... ...... ............... ........ vii Page 20 23 26 33 34 54 97 99 100 101 102 109 LIST OF TABLES-—continued TABLE 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. The effect of triphenylphosphine on polymer bound ligand coordination............ ......... The temperature dependence of the 31P spin- 1attice relaxation times(T ) of polymeric triphenylphosphine in toluene................. The 31P spin-lattice relaxation times(Tl) as a function of q(degree of swelling values).... The temperature de endence of 31P—NMR data of 0.05 M RhC1(CO)(P 3)2 and 0.2 M P03 in 01101 (complex/phosphine = l/4)..................... The degree of swelling values-q............... The preparation of samples IA—IF ............. . The preparation of samples IIA—IIG ............ The preparation of samples IIIA—IIIG ........ ... The preparation of samples IVA-IVG ............ The preparation of samplesVA—VG ............... viii Page 124 129 131 146 161 161 162 162 163 LIST OF FIGURES FIGURE 1. Phosphorus micr0probe spectrum of batch 3, 16 KV, 0.023uA ................... . ............ 2. Phosphorus microprobe spectrum of batch 4, 16 KV, 0.023uA ................................ 3. Phosphorus microprobe spectrum of batch 20, 16 KV, 0.023uA ................................ 4. (a) Phosphorus microprobe spectrum of batch 1, 15 KV. 0.023uA ............................ (b) Rhodium microprobe spectrum of batch 1, 15 KV, 0.023uA ............................ 5. (a) Phosphorus microprobe spectrum of batch 10, 15 KV, 0.023uA ............................ (b) Rhodium micr0probe spectrum of batch 10, 15 KV, 0.023 uA ........................... 6. 1H noise decoupled 13C—NMR spectrum of cumene. 7. 1H noise decoupled 13C—NMR spectrum of m-trimethylsilyl cumene ....................... 8. 1H noise decoupled 13C-NMR spectrum of p—trimethylsilyl cumene ....................... 9. Apparatus for filtering under an inert atmosphere .................... . ............... 10. Free induction decay ...... . ................... 11. 31P pulse FT spectra of triphenylphosphine and phosphinated beads .......... . ............. 12. 31? pulse FT spectrum of 11% phosphinated beads swollen with toluene .......................... 13. 31P NMR spectra of phosphinated beads and polystyrylphosphine metal complexes.. ......... 14. Plot of meq PC/g.beads vs. Rh/P ............... ix Page 14 15 16 17 17 18 18 27 28 29 40 71 85 89 92 103 LIST OF FIGURES-~continued FIGURE 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. P Micr0probe spectrum of batch IVC, Rh La, 15KV, 0.027uA ................... . ............ . 31P NMR specta of polystyrylphosphine rhodium complex. (a) spectrum of sample IID. (b) above sample after P03 treatment .......... P & Rh microprobe spectrum of batch IIF ....... MicrOprobe spectrum of batch IIF (after P03 treatment). P Kd, Rh La, 15 KV, 0.027uA ....... Phot0graphs of the scanning electron microprobe of sample IIF........ .............. Photographs of the scanning electron microprobe of sample IIG. ................. .... Use of the inversion- recovery method to obtain spin— —1attice relaxation time(Tl) for phosphorus —31 in phosphinated beads ....... ........ ...... Plot of In T vs. l/T for polymeric triphenyl- phosphine sw811en with toluene ................ Plot of log k versus 103/T(°K) ................ Variable temperature P- 31 NMR spectra of Polymer— supported RhC1(C0)( (P03) 2-—Samp1e th1. Variable temperature P- 31 NMR spectra of polymer- supported RhCl(C0)(P¢3) 2--th2 ........... . Variable temperature P— 31 NMR spectra of polymer- supported Rh01(00)(P¢3) 2-—th3.. ............ Variable temperature P- 31 NMR spectra of polymer— supported RhC1(C0)(P03) 2--th4.. ......... Variable temperature P- 31 NMR spectra of polymer— supported RhC1(C0)(P¢3) 2—-th5.. .... ..... . Variable temperature P-31 NMR spectra of polymer- supported RhC1(CO)(PCB)2--RhV5—2.. ............ age 107 111 116 117 119 121 123 126 132 134 135 136 137 139 140 LIST OF FIGURES--continued FIGURE Page 30. P & Rh electron microprobe spectra of RhV5-1 ........................................ 141 31. P & Rh electron microprobe spectra of RhV5-2, after second treatment with P03 ............... 142 32. The block diagram of DA-60 multi—nuclear NMR spectrometer .............................. 149 33. The sample and reference tube for P—31 measurements ...... ..... ........ . .............. 152 34. Representation of an inversion method for T1 measurement ........ . ....................... 154 xi COE P 03 TMEDA meq mmol Pj—P Pj—Pgrz — LIST OF ABBREVIATIONS cyclooctene, C8H14 phenyl, C6H5 triphenylphosphine, P(C6H5)3 N,N,N',N'-tetramethylethylenediamine mini equivalent mini mole polymer-supported triphenylphosphine coordinated to metal polymer-supported triphenylphosphine, G©P<: PART I THE PREPARATION OF POLYSTYRENE-ATTACHED PHOSPHINE LIGANDS, ORGANOMETALLIC COMPOUNDS AND HYBRID PHASE CATALYSTS I. INTRODUCTION Over the past decade, many highly active and selective catalysts have been derived from transition metal complexes. However, in the use of a soluble catalyst, a problem of practical importance is encountered : the separation of the catalyst from the reaction products often requires spectial treatment which usually results in deactivation. One way of overcoming this problem, while retaining the advantage of the transition metal complex catalyst, is to attach the complex to the surface of a solid support. In this case, the workup procedure for the isolation of the product is very much simplified, since the support may be removed from the reaction mixture by simple filtration. A general scheme whereby a polymeric catalyst would be used in a synthesis is presented in Scheme 11. In addition to the ease of separation from products, polymer—supported catalysts ( polymeric catalysts ) can have other advantages over their homogeneous counterparts, such as (A) enhanced size2 and positional selectivity, (B) ability to carry out sequential catalytic reactionsB, (C) the potential isolation of reactive catalytic species, with a resulting increase in reaction rate“. 2 ® UNFUNCTIONALIZD PARENT POLYMER FUNCTIONALIZED @_ C POLYMER Low MOLECULAR WEIGHT ® suarRAmS) I REUSEABLE DESIRED PRODUCT + @—C POLYMERIC m CATALYST FILTRATION CRUDE @ PRODCT SOLVENT EVAPORATION and PURIFICATION V PURE @ PRODCT SchemeI— Syntheses Using Polymeric Catalysts u, The improvement in reaction yields and the speed with which reactions may be completed by using polymer- supported catalysts has generated widespread interest in the application of polymer—bound metal complexes5-9. These polymer complexes can be considered as simple models for enzyme catalysts, and are expected to provide a key to understanding the nature of bio—catalysts. In one of the earliest studies of polymer—bound tris(triphenylphosphine)chlororhodium(l), it was demonstrated that the swollen polymer was an active hydrogenation catalyst that could be "reused" repeatedly with no loss of activity. After each use it was separated from the reaction products by simple filtration. Furthermore, this catalyst was capable of selectively hydrogenating small olefins while rejecting large ones. Evidently, large molecules were not able to permeate into the resin channels to reach the catalyst sites. The following results are typicalz. Olefin Beads rel rate Rh01(P¢3)3 rel rate Cyclohexene 1.0 1.0 Cyclododecene 1.0/4.45 1.0/1.5 Az-Cholestene 1.0/32 1.0/1.4 5 The Preparation of Polymeric Organophosphorus Ligands for Catalyst Attachment The most satisfactory way of supporting transition metal complex catalysts involves the formation of a chemical bond between the surface of the support, usually a poly- styrene 00polymer, and a ligand group involved in the metal complex. The major ligand used for homogeneous catalysts is triphenylphsophine. Thus, a variety of polymeric catalysts have been prepared by attaching the catalyst center to a phosphineZ’S—lz. Some examples are summarized in Scheme II. There are essentially two general methods of preparing polymer-bound ligands. This involve either chemically bonding ligand groups to a preformed polymer, in most cases a divinylbenzene-styrene copolymer, or polymerization of a ligand monomerlB. The first of these two methods has been the more widely used, and two approaches have been used to prepare polymer—supported phosphines. In the first route, polystyrene was functionalized by chloromethylationz, and then the chloromethylated polymer was treated with a tetrahydrofuran solution of lithio- diphenylphosphine. In the second route, resin-substituted triphenylphosphine was prepared by ring bromination (BrZ/CClu, FeCl or T1(OAc)3'l.5 H20) and phosphination of 3 the bromination product by lithiodiphenylphosphinell. Alternatively, a two stage procedure involving lithiation thmm“ (9—: Rh (CO) | e 15 “a RhCI(P¢3)3 I 1 +. ®—:);RhCI(P.¢)3-x RhC1—);MCO)4 ¢ ¢ RUCHCO)2 a .... @it RuC12(CO)2 Ni(CO) (PC! > I A ®-'I’)'2Ni“°’2 Ni(CO)2(P¢3)2 0P2 °p+RhC1a0n H ® a .A mangle :aufi n59. no na© A .IOIG t: IOIGAMU :23: 7 u ...!G Guam .. :31. a ¢_uu\«.a 3. E «5.5 a -u .A .« AT P. 5: :Mua .u :u A.“ e mo-u:u-_u UQNQ ... 5 8 of the polymer with n—butyllithium followed by reaction with chlorodiphenylphosphine12 could be used. These two routes are outlines in Scheme III. The first of these procedures is the simpler, requiring fewer steps and less Operation under inert atmosphere. However, it suffers from a number of dis- advantages. The resulting phosphine is a resin analog of benzyldiphenylphosphine. The major ligand used for homogeneous catalysts is triphenylphosphine. Thus, the resin-bound catalyst system prepared using this functionalization procedure are different from the homogeneous catalysts normally used. Mitchell and Whitehurstlu have shown that the preparation of phosphine- rhodium complexes attached to polymers leads, in some cases to quaternization as shown in Scheme IV. H2C CH HC CH2 1 2 2 \+/ - 0 fl Scheme IV 9 In view of these results, Mitchell and Whitehurst point out that the synthesis of polymer—bound catalysts may not lead to the expected product and, hence, that the behavior of such a catalyst does not necessarily parallel that of the monomeric, homogeneous catalyst. Another disadvantage of the first route is that chloromethyl ethers are carcenogenic. Several procedures for brominatin polymers have been triedlS, but a long work-up is required to remove the catalysts. Furthermore, phosphination by the reaction of lithium diphenylphosphine with brominated polymer results in a significant amount of residual bromine. This may be ascribed to the inherent difficulty in carrying out nucleophilic displacement on halogenated arenes. During the bromination of highly cross-linked polymers, the vinyl groups remaining after polymerization take up bromine. In order to overcome these problems, an easily— controlled and convenient route for functionalizing polystyrene copolymers has to be developedlé. It has been reported 17 that tertiary aliphatic amines influence markedly the reactivity of the organolithium component. The coordination of two ligands by lithium was indicated by the unusually high activity observed with bidentate ligands such as N,N,N',N'-tetramethylethylene- diamine (TMEDA) and sparteine. The greater stability of 10 these complexes is ascribed to the Chelation capability commonly associated with bidentate ligands. A number of reactions have been carried out with the n-butyl- lithium-TMEDA complexlB. For example, metalation of dimethyl sulfide to give methylthiomethyllithium; transmetalation 4—bromo—N,N-dimethylaniline to give p—dimethylaminOphenyllithium; metalation of olefins; and selective metalation of limonene, the latter being used for the synthesis of several bisabolane sesquiterpenes and monocyclicditerpene,ar—artemisene. Eberhardt's improtant discovery of the unusual reactivity of n—butyllithium in the presence of TMEDA made possible the direct metalation of the aromatic rings of polystyrene. There have been many reports of the direct lithiation of polystyrene on the aromatic rings by n—BuLi- 19-21. Broadus carries out a study of the metalation of alkylbenzenes with n—BuLi'TMEDA complexes22. TMEDA complex His results showed high yields of ring substitution, particularly in the meta and para positions. Evans reported20 a study of the kinetics and mechanism of the lithiation of polystyrene by n-BuLi'TMEDA. The infrared spectra of carboxyl and methyl derivatives of lithiated polystyrene indicate the reaction occured at both meta and para positions on the phenyl ring. Comparison of the carbon—13 NMR spectrum of 130 enriched methyl derivative of metalated polystyrene sample with that of the 11 characterized poly(d, ortho, meta and para—methylstyrenes) confirmed that metalation of polystyrene by n—BuLi-TMEDA occurs at the meta and para positions on the aromatic ring. No detectible metalation occurs either at the ortho position or at the d—carbonZl. To explore the possible application of the direct lithiation procedure to insoluble cross-linked polystyrene polymers. We carried out a series of reactions of n—BuLi'TMEDA with 2 and 20% divinylbenzene—styrene copolymers. During this study Leznoff reported 23 the direct lithiation of cross-linked polystyrene with n-BuLi-TMEDA as a means of attaching organic reagents to polymers. II. RESULTS AND DISCUSSION The Preparation of Polymeric Organophosphorus Ligands for Catalyst Attachment The synthesis of polymer—supported triphenylphosphine was acc0mplished in a single reaction vessel by the application of a direct lithiation procedurelg. Cross— linked divinylbenzene(DVB)—styrene copolymer were treated with n—butyllithium and TMEDA in purified cyclohexane. At the end of the reaction period, the dark red polymer, presumed to be lithiated on the phenyl ring, was separated and washed with more solvent. The phenyl—lithiated beads were then allowed to react with chlorodiphenylphosphine, as shown in Scheme V, and the pale-yellow product was analyzed for phosphorus. The distribution of phosphorus in the beads was determined by electron microprobe . 24 analySis . n-BuLi ~TMEDA _ 02W-l Cydoxune ,A THF ¢/\o Scheme V 12 13 A random polymer bead from each reaction was out in half with a razor blade under a low power microscope. A narrow ( 0.5 micron diameter) beam of high energy (10 to 25 KV) electrons was focused on the sample, causing X—rays characteristic of phosphorus (PKd) to be emited from a l to 10 micron diameter volume. The intensity of this emission was taken as a measure of the phosphorus density in that part of the bead. The bead cross—section was scanned by moving it across the electron beam at constant speed. X—ray intensities were plotted on the y axis of a stripchart recorder driven at constant speed. Thus distance on the chart was directly related to the distance across the bead, and the phosphorus density was determined as a function of distance from the edge of the bead. A radial distribution of phosphorus in the polymer bead could be determined from the intensity of the emission from each point in the cross sectional surface. Representative spectra are presented in Figures 1 through 5. Comparison of the spectra in Figure l and 2, shows that heating the beads in cyclohexane for an extended period before the reagents were added results in a more even distribution of the reagents throughout the polymer. AS shown in Figure 3, the reaction of 20% macroreticular DVB-styrene copolymer under similar conditions, gave phosphinated beads in which the phosphorus was evenly distributed throughout the bead. Figure 1. Emission (counts/min.) 1“ 160 lhO _ 80 _ 60 _ no— 20 _ l l l I l J O 96 192 288 384 480 576 Distance(micron) Phosphorus microprobe spectrum of batch 3, 16 KV, 0.023 uA. 15 T 180 — 160 .. 140 - 120 - 23 u 'g 100 L ' \ (I) .p S 80 _ O O 8 _H 60 — U) (D ~H (E *‘3 1+0 — 20 - 0 _ l I I l l g o 96 192 288 384 480 576 Distance (micron) Figure 2. Phosphorus microprobe spectrum of batch h, 16 KV, 0.023 uA. 140 r 120 r- 100 b 80 - 60 - Emission (counts/“fifh) no- 20r- l6 O l J l l J O 96 192 288 38“ #80 Distance (micron) Figure 3. Phosphorus micrOprobe spectrum of batch 20, 16 Kv, 0.023 uA. 17 iiihuIJWW 1% “Copowsv mosmpwan om: awn mmw was ea 1 u 4 J a .x ma . H nopmp mo szhpommw mpopmopOflE ezflconm ADV .x ma .H copes mo Edppommm wnoqu&ofiE mzuoznmonm Adv .: madman Azouoflav woCMPmHQ om: 2mm mmm «ma cm 0 J _ u a d Og Oh (‘utml/siunoo) uotsstmg Any Amv OfiI OZI OOI 08 09 Oh OZ ('unu/siunoo) uotsstmg 09I 08I 18 .x ma .oa nopmn mo sdhpommm wnonQOHOME esficocm ADV .vmma .oa :opdn mo Esppommw mnoumouoms mahogmmonm Amy ACOHOMEV wOCMPmHQ .m muzmfim Asouowsv moanewwa om: awn mwm NS 3 o om: 2mm m8 NS 8 o IIH J _ _ _ a o q _ _ . _ n lTL cl 0 _ :2 a 1 O m T: S S In Ir». 0 I O u .o; O _ 1.. m .. 0 1 S / m. 3 .oo, 3 - OCT 08 O9 Ofi OZ ('utm/Siunoo) uotsstmg OZI 19 A factor Ic is used to compare the distribution of phosphorus in samples from each reaction. The To is defined as the ratio of the intensity of radiation at the center of the bead divided by the maximum intensity. Since reactions involving polymeric beads progress from the surface into the center, the factor Ic varies between 0 and 1.0. A zero Ic value indicates that all the phosphorus is located at the outer edge of the bead and a value of l is characteristic of beads which have phosphorus evenly distributed throughout. In this work, the beads used were 425—500 H in diameter. Beads with a low Ic contained the majority of the phosphorus in the first 100— 150 u from the surface. The variation of the radial distribution of phosphorus and the percent phosphorus substitution has been examined as a function of a number of variables. The results of a study in which time and relative ratios of reactants were varied are presented in Table 1. As can be seen, this reaction provides a route to polymers having substi— tution percentages ranging from 0.50 to 15.6, and distri- butions of Ic=l to 0.1. Cyclohexane is a poor solvent for polystyrene and requires heating for maximum swelling to be reached. It was observed that heating the beads in cyclohexane for an extended period before the reagents were added increased the rate of the reaction. The table indicates that the percent substitution increases to a maximum and then falls off with extended 20 .ESQPoon mnohmohofle Esficogm Scam * .. mHH.O HN.H Ozoz OO ON OH w.H :H -- O3O.O Hm.O mcoz OO OH OH w.H mH Os.O OOOO.O ms.HH mcoz OO NH OH N.H NH OH.O NNsm.O Nm.m Ocoz OO H OH N.H HH *NOm.O OOOs.O H:.w HoOOvmhg :N Angmvoo Hemva ON N.H OH sm.O NONO.O ON.HH HoOOVmcg NH OO O o: N.H O NN.O mHNO.O NO.N HoOOthn :N HucvaO HemvNN Om N.H m OH.O mmsH.H NO.mH Ocoz OO ON OOH N.H s -- no:O.O BO.HH m:oz OO NH OOH N.H O ON.O NOmH.H NO.mH mcoz OO O OOH N.H m OH.O msOO.O ON.HH AoOOvmhg :N OO H OOH N.H : OH.O OOON.O NO.O meoz OO H OOH N.H m -- OOO.H ON.OH Ocoz HO : OOH N.H N *NsO.O NOON.O ss.OH .mcHe mH am we OOH N.H H wmm omemmpm\ems OQOHpspHpmnsm mwmwwwwmm wmmww wmmw Hammwm mwmmwwmw mmmmm .mSposamogm %o COHPSQHthHU HmHOMM map Ucm COHPSPHPmQSm monogamosm wzp Co mQOHPHOCOO Coapomcp mo Poo%wm .H wanna 21 .SPHmCoPQH ESEHNmE mzp SQ pwUH>HU Owen esp Mo HmPCmo asp Pm COHPmemH %o thmcmPQH one mo OHPmH one H OH .mHthmcm HMPQoEmHm 509% m R .OH x Adam.om\m RV u women .m\me o .mUMoD .m\vw8 n S .mm-OH x 2 + s.:OH\Hm.NNN x 2 -HVQ\HH-OH x 2V n .O:m s :OHpmsOO sons OOPMHOOHmO n .mpcmwmws one 90 COHwflwwm one oho%wp ocmxmgoaozo 39H; meson has %o PC®E¥M¢MP %o chapmpcmfiow cam oEHB m O.H .. .. HoOOvmcHs mH am we OOH ON ON .. :mmO.O :m.NH meoz Hm mm OOH ON OH O.H OBOB.O Ow.N HoOOvmng :N OO Os OOH ON NH -- OHNs.O O0.0 Ocoz HO : OOH ON NH HN.O ONN:.O sm.m HOoOOV.mcHe mH am we OH O.H OH OO.H mm:.O ON.m ocoz OO m: OH O.H mH 0H ompmmp.m\wo5 COHpSpHPmQSm mmcHaaozm Avov AHSV Hgsmnc mmmeHH copmm Q .QEwB wEHe & Immoho R HOmschcoov .H OHOMO 22 reaction time. This is apparently due to the slow decomposition of TMEDA by the alkyllithium reagents. Many batches of phosphinated beads have been prepared by ultilizing this route in this group. By experience, around ten percent substitution is very easy to achieve by using 30—40% lithiating reagent and heating the reaction at 60°C for about 12 hours. Two batches of the phosphinated polymer (batch l and 10) were equilibrated with tris—(triphenylphosphine)chloro- rhodium, and the attached catalysts were then used to reduce various olefins. Rhodium and phosphorus microprobe spectra of both batches are presented in figures 4 and 5. In batch 10, significant amounts of rhodium was present in the very center of the bead (Ic=0.36), while in batch 1, only very small amount of rhodium was present in the very center of the bead (Ic=0.07). The results of hydrogenation using these two batches are presented in Table 2. Relative rates are taken with‘ the average cyclohexene rate equal to 1. These two batches with a P/Rh ratio of 4.22 and 4.29 gave a rate of reduction of cyclohexene of 7.60 and 7.65 ml/min/mmol of rhodium. These rates are much better than that observed with a similar batch of material prepare by the bromination route. For example, one sample prepared by the bromination route with a P/Rh ratio of 4.2 reduced cyclohexene at a rate of 2.8 ml/min/mmol Rh. Table 2. bound RhCl(P¢3)3 of various preparations. 23 Rates of hydrogenation for samples of polymer— Batch P/Rh Olefin Initial rate Relative rate I (ml/min/meg Rh) —9 Cyclohexene 7.60 1.0 l—Hexene 15.42 2.70:0.20 l 4.22 Cyclooctene 5.90 0.59:0.04 0.072 l-Dodecene 6.92 0.92i0.10 B—Pinene 3.10 0.52:0.03 Cyclohexene 7.65 1.0 1—Hexene 6.65 0.95i0.06 10 4.29 Cyclooctene 5.37 0.82:0.07 0.362 l-Dodecene 15.19 1.80i0.20 B—Pinene 4.63 O.53i0.03 24 The unusually high rates of reduction of l—hexene for batch l and 1—dodecene for batch 10 seemed strange. The change in order of relative rate, from l-hexene greater than cyclohexene for batch l to cyclohexene greater than l—hexene for batch 10 was similar to what E.M. Sweet 15 observed, although it was puzzling . Studies Regarding the Position of Metalation The question of the position of metalation was considered. It was decided to use cumene as a low molecular weight model molecule. The purified cumene was treated with n—BuLi-TMEDA in cyclohexane at 65° to give a dark red solution. The lithiated cumene was then quenched with chlorotrimethyl— silane. The light tan milky looking mixture was then washed with water and aqueous hydrochloric acid, and dried over anhydrous calcium chloride followed by drying over molecular sieves. The two products were separated and purified by gas—liquid phase chromatography (glpc). These two pure samples were then examined by mass spectros— copy, proton magnetic resonance spectra and carbon—13 FT magnetic resonance spectra. The mass spectra for both fraction had a very strong parent peak at m/e = 192 (calculated for trimethylsilyl substituted cumene is 192.38). Major peaks also occured at m/e = 177, 18, 28 73 (decreasing magnitude). The 1H noise—decoupled 130 NMR spectra were recorded using a “9.....- , 25 Varian OFT-20 spectrometer. Assignment of the spectral lines is very straight forward. The chemical shifts are presented in Table 3. As Figures 6-8 show, all the lines are well-resolved. Thus, the treatment of cumene with n—BuLi-TMEDA followed by quenching with chlorotrimethylsilane gave meta and para derivatives. The ratio of meta/para derivatives was statistical, 2:1. No detectible metalation occurs either at the ortho position or at the d—carbon. The experiments by Evans et al21 on lithiation of soluble polystyrene also showed that both meta and para lithiation occurred, with a meta/para ratio of 2:1. It is likely that a considerable amount of lithiated 00polymer is actually in the meta position. Phenyl—Supported Titanocene Dichloride It has been found4 that titanocene dichloride was activated dramatically as a hydrogenation catalyst by coordinating an analog of TiCp2012 to the 20% DVB—styrene copolymer beads using a benzyl linkage to give,1,on the polymer of a phenyl linkage to give g’on the polymer. Wwfl ,1 2 ~ 26 Table 3. Chemical Shifts in l3c-NMR Spectra of Cumene,and Trimethylsi1yl-substituted Cumene. 1103ng 3 4 4 7 0 ° C 5 5 Sl( H3)3 Carbon Chemical Shift-ppm downfield from Interngl TMS. Atom Number Ia IIb IIIc 1 24.05 25.13 24.94 2 34.23 35.36 35.14 3 148.76 148.80 150.52 4 132.47 , ;126.41 126.99 5 141.15 . ‘}128.36 134.50 5 128.54 6 125.83 131.89 138.36 7 -— 1.00 1.06 8I = Purified Cumene bII = Si(CH3)3 group on C5(meta-) c _ - - III - Sl(CH3)3 group on C6(para ) 27 .oCmESO mo Exppommm mEZIUMH pmHQSOOmpnmeoc m . mpsmH H m 1H L WSE. n 0 .e . v c .md m .m e .v n e... 61/ 8: n: u men: 8”. AIIA "T.J./q L 28 .mCmESQ HSHHmHzgmeHHpIE we Esspoomm mzzlumH OmHmsooopumeoc : . mhst H s ...H mcfip A.Il« 3 n63 - ~—.—~.—*- L“ 7.: 29 I 66 ImmHoc m .mCmESo HhHHmHSLPoEHMPIQ mo ESHPoQO mzz Una mHQSO 6 . H .O anamHa o m use a N 0052» mxv 96 ms: 5 . o n n v Q 9.8 . .1. IUN Doom _ a \K 2.3 Iu — I :6 A.IIA 30 The removal of the benzylic methylene group from a position adjacent to the cyclopentadienyl ring activated the supported complex more, probably due to the elimination of the formation of,4 from 3. @645 We: 6 TIH2 [g 4 N The synthetic route for attaching titanocene dichloride to the polymer with a phenyl linkage also had been developed as shown in Scheme VI. However, the bromination reaction was a long work—up procedure and difficult to control to give reproducible results. It was decided to ultilize the direct metalation of the c0polymer on the phenyl rings by n-BuLi-TMEDA for attaching cyclo— pentadienyl groups to a polymer backbone. Several other functionalized copolymers were also prepared by this direct lithiation procedure. They are summarized in Scheme VII. 5“: 3'2/373© xs Buli ©‘H20 9 Scheme VI. 31 .COHpmNHHchHpoczm vmmm 1HH> msmnom 3.6 11."... N ”IN C. o 11 “261.. «5.5”. o .. _ I... 625 1 x-«—u:mx AA epucm n ‘1 325 I610 8.5.9.2 touaam 505:0.- " ® ocuxosofixu (Omsk. 336.. : 32 2% or 20% divinylbenzene—styrene copolymer was lithiated using the 1:1 complex of n-butyllithium and TMEDA in purified cyclohexane. Treatment of these dark-red polymer beads with chlorotrimethyltin, tetrachlorotin, diethylaluminium chloride, ethylaluminium dichloride, and 2—cyclopentene-1-one gave polymer bound trimethyltin(IV); polymer-bound trichlorotin(IV); polymer-bound diethyl- aluminium, ethylaluminium chloride; and polymer-supported cyclopentadiene respectively. This polymer—supported cyclopentadiene was used to prepare phenyl—supported dicyclopentadienyldichlorotitanium(IV) by the method developed by earlier workers. Samples containing 0.8935 meq of titanium per g. of beads and 0.075 meq of titanium per g. of beads were obtained. The ability of these two phenyl—supported titanocene to catalyze the hydrogenation of olefin was tested. These are listed in Table 4. The rate of hydrogen uptake was determined by the use of an electronic monitoring device, SAM15. In order to test the prOposal that the increase in activity for olefin hydrogenation with polymer—supported titanocene resulted from site isolation on the polymer, several different loading of catalysts were prepared. The loading was controlled by the amount of lithiating reagent used. Thus supported titanocene with loading values of 0.1649, 0.1441, 0.0835, 0.0543 mmole of titanium per g. of polymer was achieved, as indicated in Table 5. 33 mcmxw:0Homo e0 .HE om CH onxmonozo mo .HE a m m.m N.mm mmmo.o m sopmm NHONeOHe-Om:omppa 3.: 0.: mmam.o < nopem NHONQOHB-Omgoeepa mummnlw\4cHE\Az we HoEE\.cHquE meson .MNHBIUmE pmhaMpwo COHposnmh mm mo mvmm .mmHSpmnmmfimp 8009 pm NHQHBNQO popuoamsmupmezHom an osmxmzoaozo mo COHPMmeohpmm .2 manna 34 mmeozuH 40% mean COHPmamopph: one wH pmxomhn CH 63Hw> o .2mc wcfihopHCos oHc6ppome cm mo mm: was an omCHEpwpwu was mepms meopphn mo away one D mESHo> Hapop ac .HE om ans mcmx620Hoho CH mamanuH so mamwaOHozo Mo 2 m.o m .. .. mam0.0 ON.O >H -- -- mmN0.0 o:.O HHH oHao.OH HO.av on.O H O.va OO.OH mm.s :.O H m.am HasH.O OO.O HH OO.OH sO.a m.O H m.N: a:OH.o as.O H mpmmp WV.CHE\HE He HoEE\.CHENME women “OIWVHB mo HOSE we & nopmm QCoHpospmp N: mo mpmm wchmoq . mpzvmprva Book #8 NHQHBNQO m OmpnonmsmnpoezHom an CHHmHo mo COHPmcwwompzm .m oHnme 35 On the course of this study, a similar study has been accomplished by C.P. Lau25. It was decided not to continue this test. Obviously, the results in Table 5 are consistent with the site isolation approach25. III. EXPERIMENTAL Materials Compound Sourcea Treatmentb (N) (N) (A) (N) (B) (C) (C) (N) (N) (N) (D) (N) l-l n—Butyllithium Carbon monoxide Chlorodiphenylphosphine Chlorotrimethylsilane Cumene Cyclohexene Cyclooctene Cyclopentadienyltitaniumdichloride 2—Cyclopentene—l—one—ethyleneketyl Diethylaluminium chloride Divinylbenzene—styrene copolymer Ethylaluminium dichloride Ethylene l-Hexene Methyllithium Rhodium trichloride (RhClB‘ 3H20) Tetrachlorotin TMEDA Titanium butoxide Titanium cresylate (N) (N) (F) Titanium methoxide Trimethylchlorotin m F410 m 6314 m 0x14 u>~q m {rte H 4416 o.k4+4 H N: Triphenylphosphine a The following code is used for the various sources of materials. 36 37 l — Aldrich Chemical Company 2 — Alpha Inorganics Chemical Samples Company — Dow Chemical Company Engelhard Industries Incorporated — Matheson, Coleman & Bell Company — Matheson Gas Product CI) \1 0\ kn -(1‘ b.) l — Pressure Chenucal Company b (N) Chemical were used as obtained (A)—(F) Chemicals were purified prior to use as follows: Purifications (A) Chlorodiphenylphosphine, titanium butoxide and titanium cresylate were distilled under vacuum and stored in an inert atmosphere. (B) Cumene was purified by stirring with concentrated sulfuric acid, removing the acid layer. The procedure was repeated until darkening became slight. The purified cumene was washed with water, aqueous sodium hydroxide, water and saturated sodium chloride and dried over calcium chloride, 4A molecular sieves, and refluxed with sodium for 5 hours before distilling. (C) Cyclohexene, and cyclooctene were distilled from sodium under nitrogen. (D) all divinylbenzene—styrene copolymers were a gift of the Dow Chemical Company. The 1.8% crosslinked beads used for the experiment were 35-32 mesh (425-500u) fraction 38 which was obtained by the use of sieves to seperated the sample of 25—60 mesh 2% divinylbenzene-styrene 00polymer ( 1.8% crosslinked). This selected polymer was refluxed with a large excess of benzene for 24 hours, filtered, and then washed with 1:3, 3:1, 1:1, 1:3 and 0:1 benzene- methylene chloride mixtures. The purified polymer was dried for 24 hours at 60°C in vacuo before use. The 20% crosslinked macroreticular polystyrene—divinylbenzene c0polymer was washed with 10% hydrogen chloride solution, 10% sodium hydroxide solution, water, 3:1, 1:1, 1:3 water- methanol mixtures, 3:1, 1:1, 1:3, and 0:1 methanol— methylene chloride mixtures. The purified beads were dried 24 hours at 40°C in vacuo before use. (E) N,N,N',N'—tetramethylethylenediamine (TMEDA) was dried over molecular sieves and refluxed with sodium for at least 2 hours before distilling under nitrogen or argon. (F) Triphenylphosphine was recrystallized twice from 95% ethanol. The purified triphenylphosphine was stored under an inert atmosphere. All solvents were reagent grade. The aprotic solvents used were distilled under nitrogen from sodium or potassium, benzophenone ketyl. Protic solvents were dried over molecular sieves, then distilled under argon. Argon was passed through BASF-BTS catalyst heated to 140°C and 4A molecular sieves. Nitrogen was used as received. 39 Hydrogen was purified by passing through two columns (40 mm by 0.9m) of BASF-BTS catalyst heated to 140°C and two columns of 4A molecular sieves at room temperature. The purified hydrogen was then passed through two gas washes containing sodiumbenZOphenoneketyl in toluene. Preparations Biscyclooctenerhodium(1) Chloride Dimer [RhCl(COE)2]2 The title compound was prepared by a method similar to that reported by Porri et.a1.26. 2 RhCl3 + 4 08H14 + 2 CHBCHZOH ---» [RhC1(08Hlu)2]2 + 2 CHBCHO + 4 HCl Rhodium trichloride, RhC13- 3H20, (2.3329 g.) and freshly distilled cyclooctene (7m1) were dissolved in 55 ml. of degassed absolute ethanol. The solution was stirred under an inert atmosphere for seven days. The resulting precipitate was filtered from the solution of a frit under argon(as shown in Figure 9), washed four times with degassed ethanol, once with degassed diethyl ether, and dried under vacuum for twelve hours. A yellow microcrys- talline powder was obtained (2.3227 g., 70.0% yield). A second crOp of 0.54 g. was obtained after stirring for another seventeen days, giving a total yield of 2.8627 g. (86.3% yield based on rhodium trichloride). Note : This reaction is very slow, in order to get a good yield six to seven days of reaction time was required. 4O :33 o» .muozamoEPm Pmocfl cm pops: mchmpHH% pom mzpmpmmg< .m mpsmHm Joann—05 >35 09:...— Ootao v.00 tug. \ ...—.0 to... co: m a 0.. :02... \. A: / [ET 3; «00 :0:- 41 u—Dichlorotetraethylenedirhodium(I), [Rh(02Hu)201]2 u—Dichlorotetraethylenedirhodium was prepared by the method of R. Cramer27. //Cl\\ 2 RhC13' 3H20 + 6 02H4 --+ (CZH4)2Rh\\CI/’Rh(C2H4)2 + 4 HCl + 2 CHBCHO + 4 H20 Rhodium trichloride, RhC13' 3H20, (3.0092 g. was placed into a 250 ml. two side armed flask containing a Teflon— covered magnetic stirring bar, and one glass stoppers. Ten m1. of water was added to the round-bottomed flask and was stirred until the Rh013° 3H20 haddissolvedbefore 100 m1. of methanol was added. The flask was freed of oxygen by evacuation and pressuring with ethylene to 1 atmosphere three times. Ethylene was bubbled through the stirred methanolic solution at room temperature. After about 7 hours, the dark—rust—colored solid was collected by filtration under vacuum on a sintered—glass funnel. The solid was washed with methanol and ether, and dried in vacuo. The yield was 1.0604 g. (47.6%). A second crop was recovered by neutralizing the acid generated during synthesis. A solution of half a gram of sodium hydroxide in 2 ml. of water was added to the filtrate and washings from the first crop. The solution was treated with ethylene as before to recover 0.3088 g. of the title compound. This gave a combined yield of 1.3692 g. (61.6% 42 based on RhClB: 3H20). The red product was stored under an inert atmosphere at 0°C. u-Dichlorotetracarbonyldirhodium(I), [Rh(CO)ZCl]2 u—Dichlorotetracarbonyldirhodium(I) was prepared by the method of R. Cramer28. [RhCl(C2H4)2]2 + 4 CO —~ [RhC1(CO)2]2 + 4 02H4 Forty ml. of freshly distilled diethyl ether was added to 0.9113 g.(4.689 mmol) of u—dichlorotetraethylenedirhodium in a 250 ml. round-bottomed flask under argon. Then carbon monoxide was bubbled through the solution at a rate of about a bubble per second.(Caution: The reaction must be run in a well-ventilated hood.) After one and half hours, the orange-red solution was transfered to a 100 ml. flask with a syringe under argon leaving a solid which was saved for second crop. The solution was concentrated to about 10 m1. and chilled to 0°C. The yellow liquid was transfered into a 25 ml. Bantamware with a syringe under argon. The solid was vacuum dried giving 0.5723 g. (47.3% yield) of red—needle crystals. Evaporation of ether from the yellow solution left 0.1069 g. of less pure title compound. Thirty ml. of diethyl ether was added to the residue in a 250 m1. flask. Treatment of the residue solution as before gave 0.3628 g. of [RhCl(C0) The 232' 0 compound was stored in an inert atmosphere at O C. 43 Tris(triphenylphosphine)chlororhodium(I), [RhCl(P¢3)3] Tris(triphenylphosphine)chlororhodium(I) was prepared 29 by the procedure of O'Conner and Wikinson . L ———> RhCl3 + 1 P23 1111010323)3 + 012P¢3 012P¢3 + H20 ——~ OP¢3 + 2 H01 Rhodium trichloride trihydrate (1.3274 g., 5.222mmol) was placed into a 1 liter 3-neck round bottomed flask fitted with gas inlet tube, reflux condenser, and gas exit bubbler. This flask was then thoroughly purged with nitrogen before fifty ml. of oxygen free ethanol(95%) was added. A solution of 8 g. of triphenylphosphine (freshly recrystal— lized in ethanol) in 250 ml. of hot ethanol was then added. The flask was then closed except for the gas inlet system and the gas exit bubbler. The mixture was then stirred and refluxed for at least three hours. The mixture was cooled to room temperature, and then filter through a frit under nitrogen, as shown in Figure 9. The product was washed with four 50 m1. portions of oxygen— free ethanol and once with 30 ml. of dry diethyl ether. The solid was then dried under vacuum for 10 hours to yield 4.5332 g. ( 93.8% based on Rh) of deep—red micro- crystalline solid. 2—Cyclopetene—l-one The method od DePuy, et.a1.30 was used to prepare 2—cyclopentene-l—one via its ethylene glycol ketal. 44 2-Cyclopentene—1-one ethylene ketal 6.5 g. was added slowly to 100 ml. of 2N hydrochloric acid. The resulting solution was stirred at room temperature for 30 minutes before 20 ml. of methylene chloride was added. After cautious saturation of the aqueous layer with sodium bicarbonate, the mixture was filtered and the organic layer was separated. The aqueous layer was extracted with two 20 m1. portions of methylene chloride and twice with ether. The organic layers were combined, dried over anhydrous sodium sulfate, and distilled at 150-155°C before use. CyclOpentadienyltitanium Trichloride, [C5H5TiC13] The most convenient and efficient method for preparing cyclopentadienyltitanium trichloride appears to be the redistribution reaction between the sandwich compound, bis—(cyclopentadienyl)-titanium dichloride(II) and titanium 31 tetrachloride in xylene solution . (C5H5)2TiCl + TiClu ———————————— ~ 2 (05H5)T1C13 2 xylene Titanium tetrachloride (25.1 g., 0.136 mol) was added to 12.0 g. (0.048 mol) of biscyclopentadienyltitanium dichloride in 90 m1 of dry xylene contained in a three- necked flask under an inert atmosphere. The solution was .rV-wh.-- 45 stirred with a magnetic stirrer and heated at the boiling point (ca. 140°C) for 2% hours. The reaction mixture was then allowed to cool to room temperature. The resulting crystals were filtered under an inert atmosphere, washed with dry hexane, and dried briefly under nitrogen. The crude product was purified by recrystallization from hot benzene. For this a modified method was used, due to the low solubility of cyclOpentadienyltitanium trichloride in benzene. Charcoal was added to the crude product, which was then put into a thimble in a Soxhlet extractor equipped with a condenser, gas inlet tube. a side-armed flask and a magnetic stirrer. Dry benzene (300 ml.) was introduced to the flask by means of a syringe. The extraction was continued until the benzene solution in the Soxhlet extractor was very light in color. Removal of the benzene on a rotary evaporator left an yellow solid of 13.3875g. (Yield 65%). The solid was further purified by sublimation. Cyclopentadienyltitanium trichloride forms yellow to orange yellow crystals which are very sensitive to ‘ hydrolysis. Preparation of Polymeric Organophosphorus Ligands for Catalyst Attachment All Operations were performed under nitrogen with solvents prepared as previously described. 46 Lithiation Impurities such as water, air and carbon dioxide must be excluded since they rapidly react with organolithium compounds. In a typical preparation, 3g. of purified crosslinked polystyrene (2%DVB—styrene copolymer or 20% DVB-styrene cOpolymer) was suspended in 20 ml. of freshly distilled cyclohexane in a side—armed flask fitted with magnetic stir bar, condenser, gas inlet tube and gas outlet bubbler. The mixture was heated to 60°C for the required swelling period. To this mixture the required amount of purified TMEDA and n-butyllithium (1.6M in hexane) were added by means of a syringe. The reaction mixture rapidly turned red and was then heated with stirring for the required time period. After cooling, the liquid was removed by forcing it out under nitrogen pressure through a gas dispersion tube inserted through a rubber stOpper. A rubber tube leading to a flask was attached to the dispersion tube. The copolymer was then washed with 30 ml. portion of cyclohexane (four times) and three times with oxygen— free dry THF. The solvent was removed after each wash by the method described above. Phosphination The lithiated c0polymer (dark—red in color) from above was suspended in 50 ml. of dry THF under nitrogen. To this mixture, 6 m1. of Chlorodiphenylphosphine was added with a syringe. The copolymer rapidly decolorized during 47 this addition. On completion of addition, the flask was sealed and the slurry stirred for two days. The solution was then removed with a gas dispersion tube and the 00polymer was washed with 30 m1. portions of the following solvents or solvent mixtures (a) degassed THF (3 times), (b) 1:1 10% aqueous NHuCl (oxygen free): THF (3 times), (c) 1:1 THF : H 0 (3 times), 2 (d) THF, (e) 3:1, 1:1, 1:3, THF : benzene, and (f) benzene. Solvent removal was by the gas dispersion tube method described above. The beads were dried under vacuum for 24 hours at room temperature. They were analyzed for phosphorus by elemental analysis and microprobe analysis. Metalation All operations were conducted under nitrogen to exclude atmospheric moisture. All solvents were transfered by syringe. Batch 1 COpolymer (3g- of 2% DVB—styrene c0polymer) was phos- phinated by the method described above. To these phos- phinated beads, in a side—armed flask, 1.85 g. of tris- (triphenylphosphine)chlororhodium(1) (2 meq Rh) was added. The flask was evacuated and filled with nitrogen, lOOIml. of freshly distilled benzene added and the flask was sealed and the mixture stirred for seven days. The solution - was removed and the 00polymer was repeatedly washed with benzene until a wash remained clear for 24 hours and stored under nitrogen. 48 Analysis : 2.68% P (0.865 meq P/g.beads) 2.11% Rh(0.205 meq Rh/g.beads) Ic = 0.0723 (rhodium) Batch 10 The 2% DVB—styrene c0polymer (7.75 g.) was phosphinated as described above to contain 2.17% phosphorus (0.701 meq P/g.beads). The entire batch of phosphinated beads was equilibrated with 2.76 g. of tris(triphenylphosphine)- chlororhodium(I) in 120 ml. of benzene for 16 days under nitrogen. This solution was removed and the copolymer was washed with benzene until no coloration of the rinses was noted. The copolymer was then dried under vacuum for 24 hours. Analysis : C, 87.33%: H, 7.48%; P, 2.18% (0.704 meqP/g. beads); 01, 1.01%; Rh, 1.69% (0.167 meq Rh/g. beads). Ic = 0.362 (rhodium) Preparation of Polymer—Supported Titanocene Dichloride A 20 g. sample of washed 20% DVB—styrene copolymer beads were suspended in 100 ml. of freshly distilled cyclohexane in a side—armed flask equipped with a magnetic stir bar, condenser, gas inlet tube and gas outlet bubbler. To this mixture 28.5 ml of TMEDA (0.19 mol) and 100 m1. of 1.6 M n—butyllithium (0.20 mol) were added by a syringe. The mixture was heated to 50°C and stirred for 42 hours. 49 Excess methanol was added, while the slurry was chilled in an ice bath and the solution was removed by means of gas dispersion tube. The beads were then washed with methanol 1:1 aqueous HCl: THF; THF; 3:1, 1:1, 1:3 THF : CHZCIZ; CH2C12 and dried in vacuo for ten hours. All the following were carried out with oxygen—free dry solvents under a nitrogen atmosphere. These saturated beads were suspended in 100 ml. of cyclohexane in a 500 ml. side—armed flask. The mixture was treated with 5.7 m1 of TMEDA (0.04 mol) and 20 m1. of n—butyllithium (0.04 mol) and then refluxed for 38 hours. After removing the solution, the resulting tan beads were washed seven times with THF, then 40 ml. of THF was added to suspend the beads. While the mixture was stirred and cooled in an ice bath, 3.5 g. of freshly distilled 2-cyclopentene-l-one (0.042 eq) was added. The beads immediately decolorized, and turned to light tan after being stirred for 72 hours. The beads were washed five times with THF, six times with aqueous THF and twice more with dry THF and were then dried in a vacuum. Batch A A 7.5 g. sample of cyclopentadiene—substituted copolymer was weighed into a 250 ml. side-armed flask maintained under nitrogen. THF (25 ml.) was added to suspend the c0polymer and the mixture was treated with 10 ml. of 2.06 M methyllithium in diethyl ether for two days. They were washed three times with 50 ml. THF and twice with 50 50 ml. benzene. Cyclopentadienyltitanium trichloride (7.2 g., 0.032 mol) was added to the beads with 80 ml. of benzene. The mixture was stirred for 8 days, washed with benzene twice and extracted with benzene overnight in a soxhlet extractor. The vacuum dried beads were brownish pink in color. Analysis : 4.28% Ti (0.8935 meq Ti/g. beads) Batch B A suspension of the cyclopentadiene—substituted c0polymer in THF was treated with 14 ml of 2.06 M methyl- lithium in diethyl ether. After 5 days, the solvent was removed and the beads were washed three times with THF, twice with benzene. A portion of the resulting polymer was stirred with 6.30 g. of cyclopentadienyltitanium trichloride (0.028 mol) in 130 ml. of benzene for three days. The beads were treated with benzene in soxhlet extractor for 24 hours. vacuum dried, and analyzed for titanium content. Analysis : 0.36% Ti (0.0752 meq Ti/g.beads) Preparation of Polymer-supported Trimethyltin(IV) A sample of 2.04 g. of 1.8% crosslinked polystyrene beads was weighed into a 250 ml. side-armed round—bottomed flask equipped with a stirring bar, a condenser and a gas inlet tube. The system was evacuated and filled with argon. Purified cyclohexane (20 ml.) was introduced by a syringe, and the mixture was heated to 60°C. To the mixture 51 was added 4 m1. TMEDA followed by 16 ml. of 1.6 M n-butyl— lithium in hexane. After stirring for 5% hours at 60°C, the slurry was allowed to cool to room temperature. and the solution was removed by the use of a gas dispersion tube. The dark red beads were washed with cyclohexane five times and THF three times. The beads were then treated with a solution of 0.6 g. of chlorotrimethyltin in 15 ml. of freshly distilled THF. The color of the dark polymer faded in about 10 minutes. After three days of stirring under argon, the solution was removed and the copolymer was washed with THF; 1:1 oxygen-free 10% aqueous NHuCl THF: 1:1 oxygen-free H20 : THF; THF; 3:1, 1:1, 1:3, 0:1 THF benzene. The beads were vacuum dried. Analysis : 7.07% Sn (0.596 meq Sn/g.beads) Preparation of Polymer—supported chlorotin(IV) A 5.75 g. sample of 2% DVB—styrene copolymer was suspended in 35 m1. of purified cyclohexane inia 250 ml. sidearm round-bottomed flask under argon. The temperature was raised to 75°C and maintained at that temperature for 1 hour. TMEDA 8.5 ml and 35 m1. of 1.6M n—butyllithium were added and the mixture was stirred for 2 hours at 75°. After removing the solution, the c0p01ymer was washed with cyclohexane four times and THF three times. The dark red pure copolymer was then suspended in 45 m1. of freshly- distilled THF. Pure tetrachlorotin 2.2 ml. was added while the slurry was chilled in an ice bath. The mixture 52 was allowed to warm up to room temperature, and stirred for three days. After washing the copolymer with THF; 1:1 oxygen-free aqueous NHuCl : THF; 1:1 oxygen-free H20 : THF; THF; Benzene, the c0p01ymer was vacuum dried, and analyzed for the content of tin and chlorine. Analysis : 2.10% Cl (0.592 meq Cl/g.beads) 6.13% Sn (0.5165 meq Sn/g.beads) Preparation of Polymer—supported Aluminum Compound n—Butyllithium (1.6M, 19 ml.) and TMEDA (5 ml.) were used to lithiate 10.64 g. of 1.8% crosslinked DVB-styrene c0p01ymer suspended in 45 m1 of freshly distilled cyclo— hexane at 70°C for 10 hours. The red beads were washed with cyclohexane until no coloration in the rinses and twice with freshly distilled benzene. After suspending the beads in 20 ml. of benzene, diethylaluminum chloride (25% in hexane,8 ml.) was added to the slurry while they were cooled in an ice bath. The ice bath was then removed, and the mixture was stirred under argon for 31 hours. The milky looking solution was removed and the light red beads were washed six times with freshly distilled benzene. Benzene(50 ml.) and 25% diethylaluminlmi chloride in hexane (8 ml.) were added. The beads turned to pale yellow in one hour. The beads were stirred under argon for two days. After removing the solution. the beads were washed with benzene until the rinses were clear. The beads were then vacuum dried, analyzed and stored under argon. 53 Analysis : 2.52% A1 (0.934 meq Al/g.beads) Some other polymer-supported aluminum compounds were prepared similarly and they are summarized in Table 6. Analysis The percent of phosphorus, titanium, chlorine, rhodium, tin, and aluminum was determined by elemental analysis by Galbraith Laboratory or Schwarzkopf Microanalytical Labora— tory. MicrOprobe spectra were determined on an American Research Laboratories EMX-SM micr0probe. Preparation of Beads for MicrOprobe Analysis Beads were placed on the stage of a low power binocular microscope. They were held in place with small tweezers and cut in half with a razor blade. They were then attached to a graphite disk using the adhensive from freezer tape. The bead and disk were then coated with carbon from an arc and inserted into the microprobe. Determination of Elemental Radial Distribution within a Bead A selected bead was identified in the micrOprobe micro— scope. Using the secondary electron emission scan, at fast scan rate. the bead was aligned so that the X axis of the micrOprobe corresponded to a bead diameter. The X—ray detector was set to obtain a maximum reading on the required wavelength for the desired element. 54 oaN.O Ne.N N.mw NHOHsem Om s.NOH ON HH-ON-H< 66m.H O6.m H.:: NHOH. I Ml-9 In order to observe NMR, a sample containing nuclear spins (e.g. 1H, 31P) is placed in a static magnetic field Ho. An alternating field with radio-frequency V1 is applied perpendicularly to Ho. Usually, V1 is increased or decreased slowly and continuously during observation When v matches the Larmor frequency of the nucleus to be 1 observed. an absorption signal is recorded in the receiver of the NMR spectrometer. At equilibrium. the population of the nuclear magnetic energy levels will be governed by a Boltzman distribution with the parallel or lower energy state occupied by a 65 very slightly larger number of nuclei at 25°. Thus, in some cases the sensitivity of NMR can be slightly increased by lowering the sample temperature. However, care must be taken with temperature dependent spectra. Relaxation and Saturation If there were no processes by which nuclei in the upper state could return to the lower state, the continuous irradiation by the radio-frequency Hl would soon cause all nuclei to precess against HO. and no further absorption of energy would occur. This condition is referred to as saturation. In fact, energy absorption from radio— frequency fields due to NMR is observed for long periods if the rf power is not too high. The processes responsible are referred to as relaxation. These relaxation processes are radiationless mechanisms. There are two kinds of relaxation to be distinguished. The relaxation that establishes the equilibrium value of the nuclear magnetization along the direction of the external field is called longitudinal relaxation. It follows a kinetic law that is first order, in that the rate of relaxation depends on the first power of the excess (over the equilibrium value) of the number of nuclei in the upper state. -klt t/T N-Ne = (N-Ne)oe = (N-Ne)oe’ 1 The reciprocal of the rate constant k1 is called T1’ the longitudial relaxation time or spin—lattice relaxation time. It is due to the interaction of the nuclear spins 66 with various fluctuating local fields resulting from the motions of the electrons and of neighboring atoms. Addition of any paramagnetic substance to the solution can greatly shorten T1' The other kind of relaxation process is called transverse relaxation or spin—spin relaxation (characterized by a relaxation time T2). If the nuclei precessing about a field direction are in phase with one another, there will be a net component of magnetic moment in a plane XY normal to the axis Z of the magnetic field. Any disturbing field that tends to destroy this phase coherence will cause relaxation of the XY component of magnetic moment. One such process is spin-spin relaxation, in which a nucleus in a higher spin state transfers energy to a neighboring nucleus by exchanging spins with it. According to the Heisenberg uncertainty relationship the line width is Av% = l/(wTZ). The observed line width of an NMR signal depends additionally on the field inhomogeneity AHO. whose contribution to Av; arises from 2 376110 _ 1(yAHQ) AV%(inhom.) = 2n fl 2 Therefore, the observed line width Avl 1 L 1.411s _ ...L (TT)(T2 + 2 ) "' A : + . : V%(obs.) Ava Av%(1nhom.) Chemical Shift The environment surrounding the nucleus has a small but definitely measurable effect on the field sensed by the nucleus. The electrons surrounding a nucleus are acted upon by the external field to produce an induced 67 diamagnetism which partially shields the nucleus. The resulting change in resonance frequency is called chemical shift. In order to measure chemical shifts, the absorption signal of a reference compound R appearing at frequency VCR is assigned a shift of zero. The chemical shift of equivalent nuclei of a sample, having their signal at frequency V°S’ may then be measured as a frequency difference Avs in Hz. Because the frequency difference Avs is prOportional to the swept radio-frequency vl(in MHz). Chemical shifts Av obtained at different radio- S frequency v1 have to be adjusted to a common radio- frequency before comparison. In order to get chemical shift values which are independent of the frequency used, the 0 scale of chemical shifts is introduced. 0 values are obtained by dividing the frequency differences AvS(in Hz) by the frequency of v1 (in MHz = 106 Hz). v -vo _ v1 The shifts on the 0 scale are given in ppm(parts per 6>. A common reference compound used for calibrating lH million = units of 10- and 130 NMR spectra is tetramethylsilane(TMS), Si(CH3)4. For 31P NMR spectra, the reference substance is tradi- tionally 85% H3P04. The chemical shift with respect to H Hobs.‘HH3PO4 -6 5 = x 10 or H H3PO4 3P04 is defined 68 v —v obs. H PO -0 = v 3* 4 x 10—6 H3PO4 where, Hobs is the strength of the external field at the observed resonance and HH P04 is the strength of the field 3 at the resonance of H P04. So, positive chemical shifts 3 indicate greater nuclear shielding than in H P04, while 3 negative shifts indicate lesser shielding. Reference compounds can be added to the sample solution (internal reference) or kept separate from the sample in a sealed capillary (external reference). If an external reference is necessary, a correction term accounting for the difference between the bulk suscepti- bility of reference (XR) and sample solution (XS) must be added to the observed shift, éobs _ 231 _ écorrected — 6obs. + 3 (XR XS) (B) Introduction to Fourier Transform NMR In the conventional NMR experiment, the sample is irradiated with a slowly changing radio—frequency and the absorption of energy observed. The resulting continuous wave (CW) NMR spectrum is simply a plot of this absorption vs. frequency. This is known as a spectrum in the frequency domain. If the available sample size is very small, then the signal response will only be of the same order of magnitude as the "noise" inherently associated with the instrument. 69 The spectrum must then be scanned many times and the information from each scan stored in a digital computer capable of averaging the noise. This technique is known as the CAT method (Computer Averaged Transients). Noise. being of random nature, will be averaged out, while the signals associated with the sample will add up. If N is the number of sweeps, then the signal to noise ratio obtained is given by I signal enhancement 2 N : JN noise enhancement 3N The signal to noise is therefore a function of the square root of the number of sweeps. Although repeated scanning can lead to a good spectrum from a very small amount of material, it suffers from the significant drawback that the time required to performed the experiment is extremely long. This problem is more serious if we are dealing with a nucleus having low sensitivity due to low natural abundance or a small gyromagnetic ratio or both. For instance, the sensitivity for 130 is only about 10_LL of that observed for 1H, and the sensitivity for 31F is only 6.63% of that observed for 1H though the natural abundance for 31F is 100%. It would be more efficient if instead of irradiating at one frequency at a time, we could irradiate at all frequencies in the range of interest at once. Spectro— meters incorporating a large number (M) of transmitter frequencies each matched by a suitable receiver channel 70 can be built, but rapidly becomes uneconomical as M 33 increases. The Fourier transform technique accomplishes the same thing in a much more satisfactory manner. In a pulsed FT NMR experiment, the sample is irradiated by a short intense radio-frequency (rf) pulse. If A Hz is the entire range of chemical shifts to be recorded in the spectrum, the pulse length, (tp seconds) for a 900 pulse must be chosen such that (90°) << 1/44 31 13P Pulse lengths for P spectra may be of the order of microseconds. In conventional FT-NMR, the single 900 pulse is used to tip the magnetization of all the nuclei to be studied. Following this pulse, the system begin to return to equilibrium by the processes of spin—spin (T2) and spin- 1attice (Tl) relaxation. Thus, the signal detected in the spectrometer following the perturbing pulse tends to decay and is called a free induction decay (FID). The FID for a single resonance frequency, where the pulse is applied exactly on resonance, is (as shown in Figure 10b) an eXponentially decaying function with a *- time constant T2 related to the transverse relaxation time T2 and the magnetic field inhomogeneity AHo 1 _ 1 + :XAHO * - 2 T2 T2 If the spectrum consists of only a single frequency, but the rf pulse is applied somewhat off resonance, the FID 71 F‘(t) (a) tp + t PM)" (o) l 1 .c J (C) l 1 d <=:: <: <: '< ud °< < < Figure 10,Free induction decay (a) rf as input signal (b) output signal for rf at resonance (0) output signal for rf off resonance 72 observed in the receiver coil is a decaying sine wave (Figure 100) analogous to the wiggles of CW high resolution NMR spectra. This damped sine wave is a plot of signal intensity vs time and is thus called a time domain spectra. Now if there are several precession frequencies due to chemical shifts and spin-spin coupling, then the FID is a complex interference pattern. The time domain spectrum contains all the information about the frequencies and their relative intensities. but it is present in a form that is not directly interpretable. This information can, however, be reproduced by Fourier transformation of the FID in a digital computer, as a frequency spectrum equivalent to that obtained from a conventional technique. F(o) = f? f(t) ei°t dt where f(t) is the FID and F(w) is the resultant spectrum, w represents the difference between the frequency ml and the Igummn" frequency distribution mo+Aw, wo=ml—(wo+Am). The advantage of the FT method is that the entire process of excitation and detection of the FID occurs very rapidly. In FT-NMR experiments. pulses are usually applied to the sample repetitively, with coherent addition of the FID's. and the result is a dramaic improvement in signal/ noise for a given expenditure of time. For nuclei other than protons the FT method is especially valuable, since signals are generally very weak, and a large number of repetitive scans would be 73 needed. Furthermore, chemical shift ranges tend to be quite large, (e.g. the known 31F chemical shifts range from about —225 ppm to +460 ppm relative to 85% H3P04), so that extremely long times might be required for a conventional scan. With the pulse FT method the time needed for data acquisition is independent of the spectral range. (The acquisition time can be calculated by the equation 2 x SW x AQT = DP where SW is the spectral range, AQT the acquisition time and DP the number of data points in the FID.) With the advent of improved instrumentation and pulse and Fourier transform techniques, chemical applications of 31P NMR, 13C-NMR, and other nuclei (e.g. 113Cd. 109 207 Hg, Pb, 205Tl, ° ° ° etc.) grew rapidly. (C) General The nuclear magnetic resonance (NMR) phenomenon was first observed by Bloch et al34 in 1946. NMR has become one of the most widely applied physical methods, although it is only about twenty years since the first commercial instrument became available. Recent advances in NMR technology have improved the sensitivity, stability, and resolution of commercially available instruments to such a degree that nuclear magnetic resonance spectros00py has become a powerful tool for investigations of molecular structure, the structure of solids, and also of phenomena of molecular motion in liquids and solids. 74 The three elements to which high-resolution NMR is particularly adaptable are hydrogen, phosphorus. and fluorine, since for each an isotope exhibiting a nuclear spin of one—half is present at or near 100% natural abundance and these nuclei have large magnetic moments. By far, the greatest effort has been devoted to 1H NMR. However, among the three elements - hydrogen, phosphorus, and fluorine - only one, the phosphorus. is a central or backbone atom in molecular structures. Thus, nuclear magnetic resonance of the 31F nucleus is suitable not only for the structural elucidation of phosphorus compounds. but also for analytical and kinetic studies. By 1955, commercial instrumentation was in use and 31P NMR began to be of practical value as a tool for the study of molecular structure and analytical applications. Chemical shift data on several hundred phosphorus compounds a35:36. was soon publishe As instrumentation improved and became more generally accessible. chemical applications of 31P NMR grew rapidly and the literature on 31P chemical shift data for phosphorus compounds continue to accumulate in the literature and in unpublished private compilations 35-38. 31P-NMR of Transition Metal Phosphine Complexes The important role of transition metal phosphine complexes in homogeneous catalysis has prompted their study in recent years. The discovery that nuclear magnetic resonance "virtual coupling" in bisphosphines is apparently 75 related to the geometry of the complex has been important to solution structure work in this area. Among these, the most studied one is the solution of chlorotriSCtriphenyl- phosphine)rhodium(l), or so—called Wilkinson's catalyst. which is the most efficient hydrogenation catalyst yet discovered. RhCl(P03)3 is readily detected in solution using 31P NMR39. The phosphorus spectra supported a square-planar 390 structure (configuration 1). The spectra consist of four lines at high field (6P2 = -32.21), due to the mutually trans phosphorus (labeled P2) and six lines at lower field (6P = -48.94) due to cis phosphorus atom 1 (labeled Pl)’ where chemical shifts are in parts per million relative to external 80% H P0“, a negative value 3 being to low field. The coupling constant data are JPlP2 : ’37'5HZ’ JPth = -192 Hz and JPth = —l46 Hz. H] P. pa H2 CI pa a "‘ Ni» " ’2 Cl P2 H2 P The dissociation of RhC1(P¢3)3 in solution ( into RhCl(P¢3)2 and P03 ) has been a subject of considerable 76 interest and controversy, particularly in view of the indications of the importance of such dissociation to distinctive chemical properties of RhCl(P03)3, e.g.. as a hydrogenation catalyst and as a reagent for the decarbonylation of aldehydes. NMR measurementsLPO and chamical evidence indicate that the degree of dissociation in solvents such as benzene and chlorinated hydrocarbons is much smaller than suggested by early molecular weight measurementsul. The equilibrium constant, K for the l! dissociation of RhCl(P03)3 according to the reaction, K RhCl(P03)3 Ozzzzézzi Rhcl(P%3)2 + P03 1, has been determined42 spectrophotometrically to be 1.4x10— M in dilute benzene solution. By using the Fourier transform NMR technique, Tolman et al390 detected a weak free phosphine resonance in a solution of 0.05M complex in CH2C12 with an intensity about 3% of that of the principle species. It has also been demonstrated390 that in addition of molecular H to a solution of RhOl(P%3)3, the dihydride, 2 which contains three phosphorus ligands, was formed almost quantitatively. The structure was assigned as configu~ ration 2. The equilibrium may be reversed by passing N2 through the solution. Again no resonance for free triphenylphosphine was observed. 77 NMR Study on Polymers, Exchange Resins and Gels Nuclear magnetic resonance spectrosc0py has proved to be a method of considerable interest and importance for the study of physical and chemical properties of 43. Analysis 44-46 of the width and detailed polymers shape of absorption resonances and studies of magnetic relaxation times have yielded valuable information concerning the nature and frequency of molecular motions. such as the rotation of methyl groups and the onset of segmental motion of polymer chains. Results from such studies are affected to a great extent by the internal mobility of the polymer chains. It is well known 47 that the line width in NMR spectra of swollen crosslinked polymers is larger than the line width in NMR spectra of analogous linear soluble polymers at equal concentration and temperature. The increased line width in conventionally measured NMR spectra of crosslinked polymers is due to a number of effects (1). Insufficient averaging of dipolar interactions as a consequence of slow or anisotrOpic internal motions. (2). AnisotrOpy of magnetic susceptibility caused by simple physical heterogeneity of the sample. Both these effects can be suppressed by measurement of NMR spectra with rapid spinning of the sample about an axis inclined at the so called "magic angle" with respect to the direction of the stationary magic field (MAR—NMR). 78 Nuclear spin relaxation is a phenomenon which directly senses motional behavior of molecule. Doskocilova et alLL8 have used magic—angle sample rotation to investigate internal motions in crosslinked poly(methyl methacrylate) gels swollen in CHCl as a function of crosslinking agent. 3’ From the temperature dependence of the limiting line widths of the OMe. d—Me and skeletal CH groups, the activation 2 energy AE of the corresponding motions were determined. They found that AE increased slightly with. increasing crosslinked densities. Tao et all+9 noted that in resins composed of relatively large microparticles, about a micron in size, the internal water line is broadened by kinetic processes in the resin phase. In small particles (ca. 10 nm) inter—particle heterogeneity was shown by multiple peaks. In a bead copolymer of styrene and technical divinylbenzene two signals were observed with a chemical shift difference which was a function of the internal molality of H+. 0n spinning, the line width of the external water was less than that of the internal water, which differed in different types of resin. With the development of the pulse Fourier-transform nuclear magnetic resonance (FT—NMR) technique, relaxation time measurements have become feasible on biOpolymers. K. Akasaka50 measured proton and phosphorus—31 nuclear spin-lattice relaxation times (Tl) on single-stranded polyriboadenylic acid (poly(A)) in a neutral D20 solution 79 with the Fourier—transformed method. Analytical Problems in Solid-Phase Synthesis & Hybrid Phase Catalysts Analysis of hybrid phase catalysts. as well as polymer- supported reactants and products has suffered sever limitations. The most informative direct method employed now is infrared spectroscopy. The advantages of infrared are associated with the possibility of making unambiguous transmission mode measurements on insoluble polymers with approximately the same sensitivity as that attained with low molecular weight compounds. However, the complex spectrum of the resin tends to obscure the spectral features of the attached moiety. Fortunately, J.I. Crowley et al51 found that difference spectra, using purified unreacted co(polystyrene-2% divinylbenzene) in the reference bean, dramatically simplify the spectra of many polymer-supported compounds. They were able to routinely quantify the extent of chloromethylation by such a procedure, employing a standard curve prepared from resins of known percentages of chlorine. ESR has been used to analyze the mobility of a spin- labeled resin52 and to evaluate the paramagnetism of the polymer—supported titanocene catalyst36. Electron microprobe spectrosc0py has been used31 to investigate the structure of rhodium—phosphinated polystyrene. 53 54 Raman and 13C NMR spectrosc0py appears to be at, or 80 nearing, the level of development sufficient to encourage attempts at resin analysis. The alkyl and aromatic carbons of the co(polystyrene-2% divinylbenzene) can be easily indentified by natural abundance 130 NMR51. Use of specific enriched 130 functionality on the resin could permit observation of concermrations of the magnitude now being used in solid-phase reactions. However, 13C enrichment is not very economical and practical. In the polymer—supported homogeneous catalysts, phosphorus is an important linking ligand. Since phosphorus is a central atom in the functional group, nuclear magnetic resonance of 31P nucleus would provide a most economic and informative tool. The application of pulsed and Fourier- transform techniques and the improvements of modern instrumetation make 31P NMR of phosphinated beads (1.8% crosslinked) possible55. In the research area of polymer—supported homogeneous catalysts, one question which is still open concerns the exact structure of the catalytic site in the heterogeneous system as compared with the homogeneous one. Does the polymeric carrier inflict changes on the microenvironments of the catalytic site which might be determinal for catalytic activity ? The only study, which tries to find an answer to this question is one by Collman and coworkers 12. They showed that when phosphine containing polymers are exchanged with Rh/phosphine complexes. two phosphine ligands are released per Rh—ion taken up by the polymer. 81 In other words in the polymeric complex every Rh—ion seems to be surrounded by two ligands. The crosslinked polymer seems to be sufficiently mobile in that it can bring non—adjacent phosphine sites together. Bridge—Splitting Reactions of Rhodium Carbonyl Chloride It has been reported that reaction of [RhCl(CO)2]2 with excess triphenylphosphine gave the well—known trans— [RhCl(co)(P% The reaction of [Rh01(00) with 3),]. 212 triphenylphosphine (1:2 molar ratio) was first reported to give the complex tpepe-[RhC1(CO)2P¢3]56. However, T.A. Stephenson57 found evidence that the above pegge- '[Rh01(co)2PO3j' should be reformulated as the dimer complexes ppeQe-[RhCl(C0)P03]2. A mechanism for the reaction of [RhCl(C0)2]2 with P03 was prOposed (as shown in Scheme VIII) on the basis of detailed i.r. studies. L.D. Rollmann58 also reported that IR and NMR data provided structural evidence for the pie—Rh(CO)2(PR3)C1 made from stoichiometric amounts of phosphine monomers to CH2C12 solutions of the carbonyl chloride (Rh/P = 1) under C0. The polymer phosphine analog of eie-Rh(C0)2(PR3)Cl was prepared by the addition of PPBu2 phosphine resin to the carbonyl dimer solution. In the presence of high CO and H2 pressure, the trans isomer was proposed. Rollmann thought the crosslinking species. tpepe-Rh(CO)(PPBu2)2C1 was unlikely due to structural constraints in the resin. In contrast to this conclusion, G. Strathdee et a159 suggested (on the basis of the i.r. band) that on addition 82 (VCO ca. 1960 cm-1) Scheme VIII. 00 C1 00 \Rh/ \\R // OC/’ \.C//R h\\w l2P03 7 P93 OC\(/\/CO Rh /’R‘\ a/"\ 00 Cl 00 P03 (vCO = 2090, 2008 cm'l) (03° Cl 001 '03? C1 to; ‘\ /’ \\ x/ \x x/‘\\ // Rh Rhg and /Rh\~//Rh (_of/ \ln. P0}. L OC Cl \\00, (VCO 1980 cm- ) \\\\k(VCO = 2090, 2023 cm-1) isomerisation l in benzene +L(P03 etc.) 93? 011 CI /P031 \/ OC\l/ 2 Rh\ .‘_1 p/R h\C //\ LOd/ L J (93p |x . Proposed mechanism for reaction of [RhCl(CO)2]2 with P03. 83 of triphenylphosphine substituted polystyrene beads (ground to a powder ) to the benzene solution of dimer, [Rh(CO) Cl] 2 , most of the Rh(I) was present as trans Rh (CO)Cl(déP—)2 some might also be present as a dimer, perhaps as l/\P| %\$/ f“ M/\/\w Very recently60 it was reported that tri—tert—butyl- phosphine reacted with tetracarbonyl—u-dichlorodirhodium(I) to form carbonylchlorobis(tri-tert—butylphosphine)rhodium (I). ex1stlng in an equilibrium with dicarbonyl—u dichloro- bis(trl—tert-butylphosphine)rhodium in pentane solution The 31? spectrum (0H2012) of the yellow solid obtained from the reaction of Rh2012(C0)4 with P(0Me) 3 indicates the presence of Rh2C12[P(0Me) 3461 II. RESULTS AND DISCUSSION Phosphorus—31 NMR spectra of polystyrylphosphine A sample of 35—32 mesh (425—500 micron) 2% Divinyl- benzene(DVB)-styrene copolymer (zl.8% crosslinked) was lithiated with n—BuLi'TMEDA complex followed by treatment 16 The 31P NMR spectra of with Chlorodiphenylphosphine these phosphinated beads were then measured in swelling solvents. As Figure ll shows, spectra of good quality are obtained in a reasonable time (about 15 minutes) on the phosphinated beads sample swollen with toluene in a 10mm NMR tube. Disregarding the line broadening, which is due to restricted mobility in the bead, this spectrum is comparable with that for triphenylphosphine. The chemical shift (6:5.93 ppm) is essentially the same as that of a triphenylphosphine solution (6:5.83ppm). The tube interchange method was used for referencing. The chemical shift (6) with respect to 85% orthophosphoric acid is defined as H -H 5 - obs. HBPOu 6 — H x lO HBPOQ where Hobs is the strength of the external field at the observed resonance and H is the strength of the field H3 at the resonance of HBPOM' P04 84 Figure ll. 85 31F pulse FT spectra of triphenylphosphine and phosphinated beads. Chemical shifts (parts per million) are relative to 85% H P04. Exponential multiplication yielding 1.94 Hz line broadening was applied to the free— induction decays. (A) A 1.5 M solution of P913 in toluene in a 10 mm NMR tube. Result of one pulse. Acquisition time: 0.82 sec. Chemical shift: 5.83 ppm. Spectrum line width; 25.6 Hz. (B) 12% phosphinated DVB—Styrene copolymer in toluene in a 10 mm NMR tube. Result of 200 pulses with 3.5 seconds between pulses. Total accumulation time: 860 seconds. Chemical shift 5.93 ppm. Line width : 78 HZ. (C) Phosphinated beads which were oxidized. The low field signal is due to oxidized polymeric triphenylphosphine (polymeric phosphine oxide). Result of 200 pulses with 3.5 seconds between pulses. Total accumulation time: 860 seconds. Chemical shift: low field one, —26.44 ppm; high field one, +5.93ppm. Line width: low field one, 76 Hz; high field one, 82 Hz. Sample was in a 15 mm NMR tube. 86 (A) (c) 2OOHz Ho, -—-+> Figure ll. 87 The line width of the 31F signal of triphenylphosphine in solution is about 24 Hz. The unresolved spin-spin splitting by the B~hydrogen of the phenyl group is the cause of this peak broadening. Thus decoupling the lH-BlP spin interactions sharpens the 31F signal of a triphenyl- phosphine solution to a line width of about 7 Hz. The 31? NMR measurements of phosphinated beads can be performed in a variety of swelling solvents, such as tetrahydrofuran(THF), benzene, toluene, xylene, chlorinated hydrocarbon, ethylacetate, acetone, diethyl ether, chlorobenzene and nitrobenzene. The solvent effects on the chemical shifts are only about :0.5 ppm. In low swelling solvents, e.g. cyclohexane and nitromethane, elevated temperatures are necessary in order to observe the 31P NMR spectra of phosphinated beads. However, in solvents that shrink the pore sizes of the beads, such as ethanol, no 31? signal was observed even at temperatures as high as 70°C. The degree of swelling (q) of phosphinated beads in a variety of solvents are listed on page l46. A sample of 20% DVB-polystyrene copolymer was also phosphinated to give 0.78 meq P per g. beads. At room temperature, no phosphorus resonance signal was observed in toluene. However, very broad triphenylphosphine resonance was obtained at elevated temperature and the S/N ratio was improved by increasing the temperature. 88 The Detection of Polymeric Phosphine Oxide by 31P NMR As Figure 11 shows, the chemical shift of polymeric triphenylphosphine oxide is about -24.5 ppm from 85% phosphoric acid. In order to test the nature of polymeric triphenylphosphine, a sample of 11% phosphine-substituted polystyrene beads (1.8% crosslinked) was exposed to air. A small peak corresponding to triphenylphosphine oxide resonance was observed after 2 days. After refluxing the sample for 12 hours in toluene about 40% of the polymeric phosphine was oxidized. The results are summarized in Figure 12. Polymeric phosphine was very easily oxidized during phosphination, as well as in the presence of transition metal complexes, if oxygen is not excluded. The color of the phosphinated beads does not reflect their purity. Consequently, the synthesis of polymer—bound catalysts may not lead to the expected product, and the behavior of such catalysts does not necessarily parallel that of the monomeric homogeneous catalyst. Phosphorus-31 NMR spectrosc0py provides a useful tool to characterize the polymeric organOphosphorus ligands and their transition metal catalysts. Quantitative Analysis of the Content of the Polymeric Phosphines The extent of phosphination can be routinely measured by employing a sample of known phosphorus concentration. A sample of phosphinated beads was analyzed for the 89 .MMoQ ocflxo ocflsgmosm %o wcflsohm ozp maogw .mxooz 039 soapocm how new maze: NH Hog ocosaop CH woxzamop ccw Mam .oCoSHop Spa; Coaaogm meson copwcflsmmogm on whom m Mow Maw on op womomxm on mo mamemm m:.H“H Oflpms Noam op oomomxo ADV %o mamemm op comogxo Amv Mo mamewm womflcflxo oho%on Rafi mo ESMPoomm em omasg Apv on on Apv Amy mam .NH oosmum Amy h 90 content of phosphorus by elemental analysis. This standard was then swelled in toluene in a 15 mm NMR tube. The external reference (tri—n—butylphosphine), sealed in a 5 mm NMR tube, was inserted and this assembly was carefully lined up with a mark on the probe. The 31P NMR spectrum was then measured and the integration was recorded. The peak area ratio (I) of polystyrylphosphine to the tri—n-butylphosphine was then calculated. The same reference tube was used for the measurement of 31P-NIVIR spectrum of the sample of unknown phosphine content. The percentage of phosphorus was then calculated from the known percentage of phosphorus value and the relative peak areas. %P(standard) %P(unknown) = I(Standard) X I(unknown) Two to three measurements were made for each sample and the average value was used for the calculation of the percentage of phosphorus with an accuracy of about i5-lO%. The effect of the delay time (time between pulses) on the results was considered. The 31P-NMR measurements of a sample and a standard were carried out with a delay time of 3.5 seconds. The phosphorus content of the sample was calculated to be 60.8% that of the standard. The experi- ments were then repeated with a delay time of 100 seconds and the percentage of phosphorus of the sample was calculated to be 63.8% that of the standard. Thus the delay time parameter does not effect the obtained results (within error). In order to save time and get a better 91 signal to noise ratio, a 3.5 seconds delay time was used in this study unless otherwise mentioned. The 31P NMR Spectra of Polystyrylphosphine Metal Complexes The 1.8% crosslinked DVB-styrene copolymer was 16 The phosphinated phosphinated by the standard technique beads were then reacted with rhodium complexes, such as [Rh(COE)2Cl]2 (COE is the abbreviation for cyclooctene) and RhC1(P¢3)3, in benzene. After equilibration, the beads were washed with benzene until no coloration showed in the rinse solution. The beads were then vacuum dried. The metal complexed c0p01ymer was swelled in toluene, and 31P-NMR measurements of the polystyrylphosphine metal complexes and the phosphinated copolymer were then carried out under the same conditions. The 31P-NMR peak of the metal complexed polystyryl— phosphine is too broad to be observed due to the strong interaction of the rhodium electrons and the phosphine nuclei as well as the insufficient averaging of dipolar interactions. The polymeric phosphine oxide peak might have been mistaken for the metal complexed polystyryl- phosphine peak due to their similar chemical shifts. However, these signals were not hard to distinguish when a standard sample was employed. As can be seen from Figure 13, the uncomplexed polymeric phosphine peak slowly decreased as the ratio of rhodium complexes to the phos- phines was increased, until no peaks other than the Figure 13. 92 31P NMR spectra of phosphinated beads and polystyrylphosphine metal complexes. The high field signal is due to the reference sample, tri-n—butylphosphine. Spectrum from bottom to top is batch 4 phosphinated beads, complexes of IVB, IVD, IVE, IVF and IVG. 94 reference sample (P(n—Bu)3) were observed. The uncomplexed polymeric phosphine does not show a significant change in chemical shift or line width during this series of measurements. The amount of uncomplexed polymeric phos— phine in the metal complexed copolymer can be calculated by comparing the relative peak areas by the previously described method. T. Mizoraki et a162 observed a sharp absorption of 31P (32.4 ppm lower in field than free triphenylphosphine) for methylene chloride solution of soluble polystyryl nickel complexes (Pj-fl—Ni(P¢3)2Br) which were prepared from tetrakis(triphenylphosphine)nickel and brominated polystyrene. TheBlP chemical shift was independent of the nickel content of the polystyryl complex. An anaIOgous polystyryl platinum complex(P]—¢-Pt(P¢ I), prepared from 3)2 iodinated polystyrene and tetrakis(triphenylphosphine)- platinum, also shows one sharp absorption of 31F (26.7 ppm) (3065 Hz) in and a large coupling constant of JPt—P methylene dichloride. In hexane, however, neither of the above complexes showed absorption. The eXplanation suggested is that, when the polystyryl metal complex swells in a polar solvent such as methylene dichloride, the environment around the metal site is analogous to that of homogeneous bis(triphenylphosphine)—arylmetal(II)halide, which gives one sharp 31P-NIVIR absorption. In a non—polar solvent, such as n-hexane, the coordination sites of metal 95 atoms are blocked by either the halogen atoms or phenyl rings of the polystyrene, resulting in the 31P-NMR absorption becoming too broad to be detected. At the present time there is considerable controversy about the isolation of catalytic sites in the resin—bound chemistry. A resin-analog of tris(triphenylphosphine)- chlororhodium has been prepared by exchanging the phos— phine in the complex with polymeric phosphines (resin—bound phosphines). Originally, a single bond was assumed to be involved in the attachment of the polymer to each rhodiuml If this is the case, then the complex could be easily lost from the support. It has been found that use of low crosslinked polystyrene allows multiple binding to the polymerlz. When polymeric triphenylphosphine was treated with bis(tripheny1phosphine)carbonylchlororhodium, analysis of the filtrate from reactions using measured quntities of polymeric triphenylphosphin and the homogeneous complex showed that at least two triphenylphosphines were always released for each metal atom incorporated. The formation of new cross—links during repeated usage processing of the polymer-supported catalysts has also been ascribed to the multiple binding of bound catalysts. Two rhodium(I) complexes, RhCl. 91133: \Cl/ \COE 912}: \01/>ng P [ P 1 (IV) (V) ( Pj—P02)3Rh01 (VI) Scheme IX. 106 Samples of series III(IIIA—IIIG) were crushed. The increase of surface area resulted in the increase of polymeric phosphine concentration around the metal center. Consequently, the Pj—PC/Rh value is much higher than the other series. In addition, a low loading of monomeric complexed copolymer (P3—P02)3RhCl(VI) was also obtained. Characterization by Electron Micrpprobe Analysis Random beads were selected from batch II, III, IV, V, IIB, IIF, IIIA, IIIF, IIIG, IVA, IVC, IVD, IVF and IVG. The beads were then cut in half under a microscope. Each of the samples was subjected to electron microprobe analysis. The Ic values (as defined on page 19) for batch II, III, IV and V of phosphinated beads and IIIF, IIIG, IVG were calculated from the spectra as described in part I and they are as follows : Sample II III IV V IIIF IIIG IVG lg 0.22 0.28 0.41 0.46 0.32 0.33 0.24 Trace Element P P P P Rh Rh Rh When the attached catalysts were prepared by equili- brating with [Rh(COE)ZCl]2, using a deficiency of the complex (e.g. IIB, IIF, IIIA, IIIC, IIID, IVC) the metal was distributed locally in the first few micron of the bead with no metal in the center of the bead, as shown in Figure 15. The percentage of volume containing the metal was calculated and they are as follows : Emission (counts/min.) 107 60 r. so - 40 _ 30 b 20 - t 10 r O - 4 L .1 l 1 L 0 38 134 230 326 422 518 Distance(u) Figure 15. Micr0probe spectrum of batch IVC, Rh Ld, l5kV, 0.027uA. 108 Sample 118 I10 IIIA IVA IVC IVD IVF VRh/Vtotal 0.68 0.78 0.28 0.33 0.31 0.44 0.77 An attempt was made to obtain the elemental ratios of rhodium to phosphorus within the shell where rhodium was distributed. This is of interest, because the Pj-Pc/Rh 31 have been calculated from P NMR analysis. On the basis of phosphorus electron microprobe spectra of batch III and IV beads, the counts within the rhodium containing shell and the total counts in the whole bead were calculated. The ratios (CR) were obtained on several bead samples. The phosphorus content of the rhodium containing shell was obtained by multiplying the ratio CR by the elemental analysis data of the phosphorus content. The results are presented in Table 12. It is interesting to find that, the electron micrOprobe analysis and 31P NMR spectroscopy both gave a P/Rh value of about 1.6 on sample IVF. The electron micr0probe data clearly show that in the case of high metal loading, the content of phosphorus within the rhodium containing shell is less than twice the content of rhodium. Consequently, the Pj-PC/Rh ratio is smaller than two. In the case of low metal loading, due to the inaccuracy of edge position, it is quite possible to have overcalculated by 50% the content of phosphines within the rhodium containing shell. In spite of this uncertainty, the trend of increased P/Rh ratio as the loading decreases is apparent. 109 mvmw mflmzamsm mzz man one 809% woPMHSOHmo madcz NH 90% Oomm Pm mfim woe H.0 Spa; powwohp coop o>m£ mason NH 060 6066 ed memop.m\mam 66a 66.6 spa; coedoae o 0 mpwv onohmoHOflE map 800% 009MH50HMO o Haozm wQHQHMPsoo Esflcosh one Caspflg Q exop mom 6 6.H c.a Ham.6 Ham.6 6H6.6 m>H 6.a s.m 06H.6 m6m.6 0mm.6 0>H m.H m.m 060.0 mmm.0 000.0 0>H 0.6 666.6 6mm.6 066.6 <>H :.m H.6 000.0 0mm.0 Hmm.0 AQOPMMV AOMOMOQV AMQV A00000V o o A600 6 wemeemv 60\ 0-66 60\ 0-66 daae .as 6 66660.0\ 66 665 66666. \nm 008 0agswm .Coepwzflppooo Ucwwfla US$09 0086609 20 0QH£QmO£QH%C0:QHpP 60 #00660 058 .MH 06909 113 cooling the slurry to room temperature, the excess 31 triphenylphosphine was washed away before P analysis. Another of the four samples was used as a standard reference for comparison. The results are summarized as follows Sample Treatment Pj-PC/Rh VI-Rh~1 None 2.1 VI—Rh-2a add 0.48 meq P03, heat at 90°C 2.8 for 8% hours. VI-Rh-3 heated at 90°C for 9 hours 2.3 VI-Rh-4b add hydrogen 2.1 a The P—3l absorption was barely detectable. b No other changes was detected. It was found that the above four samples had been crushed to a powder during the equilibration with rhodium complex. This may have affected the coordination of polymer-supported triphenylphosphine, therefore another batch of beads was prepared by equilibration of 2.27 g. of phosphinated beads containing 0.94 meq of phosphine per g. of beads with 0.563 m eq of chlorobiscyclooctenerhodium dimer in benzene. After removing the excess dimer complex, these beads were subjected to 31p NMR analysis. The PJ-PC/ Rh value was 1.5, and it was found to be unchanged after heating at 90°C for 12 hours. However, the PJ-PC/Rh value was increased to 3.1 by adding 0.036 meq of triphenyl- phosphine per g. of beads and heating at 90°C for 48 hours. 114 31 P—NMR analysis was determined after removing the excess triphenylphosphine. As can be seen from Table 13, the Pj-PC/Rh value was increased to around three after the triphenylphosphine treatment. Interestingly, in the case of low rhodium loading, the final PJ—PC/Rh value is greater than 3 ( for example, Pj-PC/Rh of IIA was increased from 1.7 to 5.8). Added triphenylphosphine catalyzed the replacement of the cyclooctene ligand and the cleavage of the chlorine bridges by polymeric phosphines. //Cl \ /Rh Pfla, Toluene [COEJT p\Cl p/)|6 hen, +1®P¢213Rhcl l p This monomer complex, produced by cleavage of the chlorine bridges, then exchanges with nearby polymeric phosphines. In the case of high metal loading, for example sample IIF, the final Pj-PC/Rh value (2.2 ) was limited by the available polymeric phosphine within the bead. However, on addition of a large excess of triphenylphosphine, some homogeneous phosphines replaced the polymer—supported triphenylphosphine in the polymer complex and formed an orange—red complex, presumably [RhCl(P¢3)2]2, which precipitated on the bottom of the NMR tube. 0n the basis of elemental analysis of the rhodium content and the amount of complexed polymeric phosphine from 31P—NMR 115 measurement, the Pj—PC/Rh value of this polymer complex was calculated to be 3.1 (Table 13). If the beads were crushed during complex equilibration, the Pj—PC/Rh value was a little higher (2.1 for VI-Rh-l compare to 1.5 for the uncrushed sample). This is not unexpected, since an increase in surface area would make more polymeric phosphine available around the metal centers since the metal concentrates at the surface of the polymer beads. In addition, the Pj-PC/Rh value increased even further on heating although the PJ-PC/Rh value of uncrushed beads was found unchanged after heating at 90°C for 12 hours. Again, this may be ascribed to the increase in polymeric phosphine concentrations around the metal centers. The above conclusions were supported by an electron micr0pr0be analysis. Random beads were selected from sample IIF's (before and after the triphenylphosphine treatment). As we can see from Figure 17 and Figure 18, before the P03 treatment, electron micr0probe spectra of cross—section of a bead sliced in half shows that the metal was distributed mainly in the outer regions (ca. 70 micron from the edge), and no metal was detected near the center (Figure 17). However, after a second treatment with P03, the metal was distributed throughout the bead (Figure 18). The Ic value of the rhodium is comparable to that of phsophorus. Samples from IIG'S gave similar electron microprobe spectra. Emission (counts/min.) 300 400 500 600 700 200 100 Figure 17. 116 1?- fl 3' Rh U 96 £92 Distance(u) P & Rh micr0probe spectra (before P03 treatment) 140 120 100 80 60 40 20 288 384 of batch IIF. 117 1:: 0.30 1:: 0.40 ----- m. Rh- o o _ m } P o \n- N o o _. A N C u E \ m .p C o :3 UN" o .—1 0 $21 0 "-4 m m 0 "-1 o _ 5 .-+ o 1n)— o w 0 Figure 18. 1 J l 96 192 288 Di stance (u) Microprobe spectra of batch IIF(after P0 treatment). P Ka, Rh La, 15 KV, 0.027uA. 120 118 Photographs of the scanning electron micr0probe of sample IIF and IIG are presented in Figure 19—20. As can be seen from Figure 19(a), when sample IIF (before P03 treatment) was examined under a microscope, most of the beads were observed to be dimpled. However, after the treatment with P03, the dimples were shallower (Figure 19 (b) & 20(a)) and eventually perfect Spheres were observed. The results from electron microprobe analysis and scanning electron microprobe are consistent with the pathway (Scheme IX) we have prOposed. The cyclooctene rhodium dimer attaches to the polymer whenever one or two cyclooctene ligands are replaced by polymeric phosphines. The rhodium then tries to find other available polymeric phosphines to replace the cyclooctene ligands. Thus the bead becomes dimpled. In the presence of triphenylphosphine under high temperature, the free homogeneous P03 may replace polymeric P03 temporarily, thus allowing the metal center to rearrange to a lower energy conformation. Consequently, the beads return to their original. Spherical shape. 31 Polymeric Triphenylphosphine P Spin-Lattice Relaxation Study of 31P spin—lattice relaxation times (T1) of polymeric triphenylphosphine were obtained by applying a 180°—7-90° pulse sequence. The intensity A of peaks obtained from sets of such spectra (for example Figure 21) was used to determine Tl according to the expression Figure 19. 119 Photographs of the scanning electron micr0probe of sample IIF. Magnified 25 X (a) Before the P0 treatment, containing 0.364 meq Rh/g.beads, Pj—PC/Rh = 2.0 (b) After second treatment with P03. containing 0.2896 meq Rh/g.beads, Pj—PC/Rh = 3.1 ( see Page 112) Figure 20. 121 Photographs of the scanning electron microprobe of sample IIG. Magnified 25 x (a) After first treatment with P03. Pj—PC/Rh = 2.6 (b) After second treatment with P03. Pj—PC/Rh = 2.9 ( see also Page 112) 122 .m.060 006 00 006006585000 8000 0665000 03000 85060000 nomm .00009 006006000000 06 6mums0osmmonm 000 66900566 0060000600 006600610600 060690 00 000608 m00>0000|006m00>06 006 00 00: .6N 005060 123 6.60 66.0 66.6 60.0 66.6 66.0 66.6 60.6 60.0 60.6 66.0 66.6 251 66.0 60.6 it; 66.0 66.6 o. 3. s/\ j 124 Table 14. The temperature dependences of the 31P spin- lattice relaxation times(Tl) of polymeric triphenylphosphine in toluene. Temperature l/T x 103 Tl(sec.) 1n Tl <°K> (deg—l) 377.5 2.649 3.68 i 0.05 1.303 357 2.801 3.03 i 0.09 1.109 337.5 2.963 2.74 i 0.08 1.008 320 3.125 2.29 i 0.07 0.829 301.5 3.317 1.94 i 0.02 0.663 286 3.497 1.80 i 0.07 0.588 275 3.636 1.64 i 0.07 0.495 125 A = A0 [ 1-2 exp (-T/Tl)] where A30 is the equilibrium intensity of the Spectrum. The Tl's were calculated by the least—squares method. The temperature dependence of the measured phosphorus spin—lattice relaxation times (T1) of polymeric triphenyl- phosphine (batch I) swollen with toluene are presented in Table 14 and depicted in Figurezazas a plot of In T vs. 1 l/T. FigureZHBindicates that within the temperature region studied (2—106°c), the dipole—dipole interactions with the nearest protons provide the dominant relaxation mechanisms. This observation is consistent with the results of J.Jonas et al63 on the T study of triphenylphOSphine within the 1 temperature range of 90—160°C. Since dipolar interactions provide the main relaxation mechanisms, the reorientational correlation time, TC, can be estimated by assuming 2 2 1:6h2yPYHT T1 r6 C P-H where h is the Planck's constant divided by 20, YP is the gyromagnetic ratio for phosphorus, YH is the gyomagnetic ratio for hydrogen, rP—H is the interatomic distance of P and H, and TC is the reorientational correlation time in seconds. If PC = 1.83A, co = 1.393 and CH = 1.10 3, the calculated rP—H is 2.92A. 0n the basis of this interatomic distance of P and H, the reorientational correlation time, ln T1 0 a) Figure 126 slope = -8.29 x 102 103/T-°K 22. Plot of 1nTl vs. l/T (OK-l) for polymeric triphenylphosphine swollen with toluene. 127 10 T was calculated to be 5.7 x 10— seconds at room 0’ temperature. For comparison, the spin-lattice relaxation time (T1) of a solution of a 1.5 M triphenylphosphine in toluene was measured to be 11 seconds at room temperature. By applying the above bond lengths and equation, the approximate TC was calculated to be 10—10 seconds. In other words, the mobility of polymeric triphenylphosphine swollen with toluene is only one sixth that of 1.5 M triphenylphosphine in toluene. Jonas et al63 considered both the intramolecular and intermolecular dipolar 31 contributions to the P relaxation, and obtained a value 10 seconds for the reorientational correlation of 0.75 x 10— time of neat triphenylphosphine. By assuming l/Tl prOpotional to TC, the activation energy of motion was estimated to be 1.65 Kcal/mole. This is much lower than expected. In single-stranded polyribo— adenylic acid50 an activation enthalpy AH of 8.1 Kcal/mole was obtained for the molecular motion of the sugar— phosphate backbone, as deduced from the phosphorus relaxation data. The activation energy for triphenyl- phosphine was found63 to be 6.2 Kcal/mole from the 1H Tl plot. In contrast, by assuming 1/Tl proportional to TC, the activation energy we estimate from their 31P data is 2.3 Kcal/mole, which is much closer to our results. This seems to tell us that the motion hindrance of polymeric triphenylphosphine is comparable to that of neat triphenyl— phosphine. 128 A comparison of the temperature dependence of T1 of a triphenylphosphine solution with this work would probably give us more information, however, no extensive work was tried. An attempt was made to investigate the influence of solvent on the motion of polymeric triphenylphOSphine. This would be of interest , since S.L. Regen52 found that the degree of swelling of the polymer supports is an important factor in determing the mobility of the attached spin label (examined by electron paramagnetic resonance spectroscopy). As we can see from Table 15, in sharp contrast to Regen's results, the rotational correlation times (TC) of 31 polymeric triphenylphosphine calculated from P spin— lattice relaxation times (T1) are not sensitive to the swelling properties of the solvents. Interestingly, diethyl ether, which is a poor swelling solvent, had about the same effect on the motion as toluene which swelled the polymer lattice appreciably. No absorption was observed on the 31F FT—NMR of polymeric triphenylphosphine swollen in hexane at room temperature. At a higher temperature (55°C) a sharp absorption was obtained. 129 Table 15. The 31P spin—lattice relaxation times(Tl) as a function of q(degree of swelling values). Sample Solvent Tl(sec.) 7C(10-losec.)a qb Batch lC Toluene 1.94:0.02 5.7 2.95:0.08 EtZO 2.17i0.03 5.1 1.35i0.04 2.18i0.03 5.1 Acetone 2.05i0.03 5.4 1.68:0.02 2.01i0.02 5.5 Bd Toluene 2.09:0.03 5.3 Cumene 1.79i0.04 6.2 1.98:0.07 Batch 4C Toluene 1.90:0.05 5.8 Batch Se Toluene 1.72:0.07 6.4 THF 1.71i0.07 6.5 3.05i0.08 Benzene 1.76:0.07 6.3 3.06:0.04 Oxidizedf Toluene 1.80:0.05 6.2 S 1.3610.02 (P=O) 1.5M P03 Toluene 11.29:0.49 1.0 0.4M Toluene 9.84i0.85 P(n-Bu) 3 a 2 See page 146 Table 17. Calculated by assuming (Tl)-l = éfiZYH YPZTc/(rP-H) Phosphinated beads for multiple binding studies. to rhodium. ( [RhCl(00B)2]2) Equilibration. Swelled in toluene and exposed to air for days. 12% phosphinated beads, half of the phosphine complexed 14% phosphinated beads, ca. 5% phosphine was oxidized. 130 Phosphine Exchange in Carbonylchlorobis(triphenyl— phosphine)rhodium(I), RhCl(CO)(Pg:)2 The phosphorus—31 spectrum of trans—carbonylchloro— bis(tripheny1phosphine)rhodium(I) consists of a doublet 39b,6l,64 39b due to rhodium-phosphorus coupling The chemical shift was reported to be -29.1 ppm in chloroform solution vs. 85% H3PO4; -28.9 ppm61 in CH2C12 vs. 8 0 H 573 P00 and the coupling constant J was reported to be -124 Hz or 129.4 Hz respectively. Rh—P Since it has been found that the phosphine exchange in RhCl(P03)3 is a dissociative mechanism 39a, we decided to carry out a ligand exchange study of RhCl(CO)(P03)2 as a model for the analogous polymer complex. The 31P spectrum of a chloroform solution RhCl(CO)(P03)2 in a 15 mm NMR tube showed a chemical shift of —34.6 ppm vs. free triphenylphosphine ( or -28.9 ppm vs 85% HBPOQ) and a coupling constant of 129.4 Hz. Three samples of different complex to phosphine ratios were used in the temperature studies. Approximate methods65 were used to evalute the rate constants. As we can see from Table 16, the ligand exchange is very fast and only one broad peak was observed at temperatures as low as —44°C. The rate data are presented in the form of an Arrhenius plot in Figure 23 and may be described by the equation rate(T) = 106°62e—36OO/RT 131 The calculated activation parameters are B8 = 3.6 Kcal/mole, 60* = 12.0 Kcal/mole, # _ ¢ = _ . AH298 — 2.4 Kcal/mole, A8298 32.2/mole deg A plot of In k vs. ln [P03] gave a straight line with 0.23 as the slope. In other words, the exchange rate is prOportional to the concentration of triphenylphosphine to a power of 0.23. Table 16. The temperature dependence of 31P-NIVIR data of 0.05M RhC1(C0)(P03)2 and 0.2M P03 in 08013. (complex/phosphine = 1/4) Temperature half—width(Hz) k(sec.'l) x 10')"L c K 61 334 75.68 2.69 28 301 96.44 1.79 10 283 135.50 1.10 0 273 150.79 0.96 -10 263 193.37 0.70 -26 247 278.28 0.46 -34 239 380.14 0-32 -44 229 436.29 0.28 132 4.6 . 4.5 _ 4.4 _ slope = -0.781 x 103 4.3 . 4.2 - 4.1 P- 4.0 - 3-9 - x 3.8 1- 00 O H 3-7 - 3.6 t 3-5 ' 3.4 L 3'3 3.0 3R2 374 34.6 3.18 470 432 4.14 4.6 103/T (OK‘l) Figure 23. 3 0 Plot of log k versus 10 /T ( K). 133 Phosphine Exchange in Polymer—supported RhCl(C0)(P0'3)2 The temperature dependence of 31P NMR spectra of polymer—supported RhC1(C0)(P03)2 (RhV —RhV3) are 1 presented in Figure 24—26. These spectra indicate that an exchange phenomena is associated with this system, and that the exchange rate increases with the rhodium to. phosphine ratio. From previously study, it has been demonstrated that the 31P NMR spectrum of rhodium complexes of polymer— supported triphenylphosphine ligands is too broad to be observed. However, polymer—supported free phosphine may exchange with rhodium complexed monomeric phosphine, (Pl—P02)Rh01(co)(Pgl). In an effort to prove this, chlorodicarbonylrhodium(I) dimer was used to prepare the polymer—supported complex (RhV“). The temperature dependence of 31P NMR spectra of this polymer complex is presented in Figure 27. As can be seen from Figure 27(a)- (c), exchange did not occur or it is too slow to be detected by this method. Triphenylphosphine (0.037 meq per g of beads) was added to RhVu under argon and this slurry was heated at 90°C for 6 hours with occasional shaking. After washing away the triphenylphosphine with toluene, the 31P NMR was measured at room temperature and 94°C. Figure 27(d)&(e) clearly show that the ligands exchange. This result supports the belief that free polymer-supported triphenylphosphine exchanges with RT 35°C 69°C Figure 24. 134 “6\ P(n—Bu) 3 -2°c h¢\_ -20°c L00— —52°c t...» Variable temperature Phosphorus—31 NMR spectra of polymer—supported RhCl(C0)(P03)2--Sample Rth 135 <——P(n-Bu)3 85°C -35°C W 55°C V M LM 28°C Lvu’" L00» Figure 25. Variable temperature Phosphorus-31 NMR spectra of polymer-supported RhCl(C0)(P03)2—-Sample RhVZ. 136 P(n—Bu)3 RT 69°C o 108 C 1.0 Figure 26. Variable temperature phosphorus-31 NMR spectra of polymer—supported RhCl(Co)(P03)2-- Sample RhVB. 137 ((1) RT (a) 94°C P(n—Bu)3 000 (b) RT 4W (e) 94°C 0 (c) 5 C ,0) .... Figure 27. Variable temperature 31P NMR spectra of polymer—supported RhCl(CO)(P03)2--Sample RhVu. a,b,c—- before the P03 treatment d,e—- after the P03 treatment t. (II/C 138 coordinated monomeric phosphine. In order to obtain additional information about this reaction, another sample, RhV5, was also prepared by the chlorine bridge splitting reaction of [RhCl(C0)2]2 with phosphinated beads in benzene. After unreacted complex was removed, the copolymer was dried and then divided into two parts (RhV5—l and RhV5—2) before swelling with toluene. RhVS-l was retained as a reference sample. Triphenylphosphine (ca. 0.012 meq per g. beads) was added to RhV —2 and this 5 mixture was heated at 90°C for 5 hours. After removing excess triphenylphosphine, the 31P NMR spectra at room temperature and at 94°C were measured. RhVS-l was also heated at 90°C for 6% hours and a comparison of 31P NMR spectra at room temperature and 93°C is given in Figure 28. About 0.1 meq of triphenylphosphine per g. of beads was again added to RhV5—2 and this slurry was heated at 90°C for 24 hours. The triphenylphosphine was again 31P NMR spectra were removed and variable temperature measured. The results are presented in Figure 29. —1 and RhV Both RhV -2 were then subjected to electron 5 5 microprobe analysis of phosphorus and rhodium. The results are summarized in Figures 30 and 31. As can be seen from Figure 30, the rhodium in sample RhV5—l was distributed only in the outer regions (ca. 150 u from the edge), and no metal was detected near the center. However, after the second P03 treatment the metal was distributed throughout W W P(n—Bu12.; L» (s) Figure 28. 139 ( (e) P(n-Bu)3 / a) (b) LNVW“ (C) (d) Variable temperature phosphorus- 1 NMR spectra of polymer supported RhCl(C0)(P 3)2—-Sample RhV5 (a) phosphinated beads before metalation, at RT (b) RhV5-1 at RT (0) RhV5—l at 93°C(after heating at 90° for 6% hrs) (d) RhV5—2 at 93°C(after treating with 0.012 meq of triphenylphosphine at 90°C for 5 hrs) (e) RhV5-2 at 33°C(d sample heat 24 hrs, 90°C) (f) RhV5—l at 92°C(c sample heat 24 hrs, 90°C) (a) RhV5-2 at 92oC(e sample) 140 ‘——P(n—Bu)3 134°C 33°C 94°C -6°C W Figure 29. Variable temperature 31P-NMR spectra of polymer—supported RhCl(CO)(P03)2—- RhV5-2 (after treating 0.1 meq of P03 per g. beads at 90°C for 24 hours). 141 Emission (counts/min.) 5 0 O ... v 8 .z 1 __ ._ n- .c ,1. l m , I I 1 a. --—- ” o I N ‘ _ d o H n _. ..2 -g ...;— _—V "“ o 5 -~-—-'_-““ ° ..;m ' 0:Il:::::::i_.....;..._ a ‘ ‘—‘ _ %‘1 O 1 l l 1 l o o o o o o o O O O O n v n u .— Emission (counts/min.) Figure 30. P & Rh eleclectron microprobe spectra of RhV —1 Distance (0) Emission (counts/min.) 142 8 . ,2“ 470 I: = 0.34 ----P Rh r——D— 1c : 0.23 ——-—Rh 0 [0 30 av of 41 °"‘ ' ' ‘60 o Rh 0 _> 0" ‘ 450 n i 8" ..1 «40 ¢ .1 1 11, P 1 . 8 ' I ”r , «so 1 ., 1 1 0 '1 . O" 1 ‘ V .120 N 1 1 I ‘ 1 ‘ ‘ I1 1 Rh . I 1 'l '—> I I P 8- 1 t1 ' ‘— 40 o 1 - o 1 I L l 1 1 0 30 I26 222 318 414 Distance (0) Figure 31. P & Rh electron microprobe spectra of RhV5-2, after second treatment with P03. 143 the bead, as shown in Figure 31. Figure 31 also shows that the Ic value of the rhodium is comparable to that of the phosphorus. The rhodium microprobe spectra of RhV gave 1 an 10 value of 0.40. In the case of RhV -l the only phosphine source is 5 polymer-supported triphenylphosphine. However, Figure 28 indicates that the exchange is fast enough to be detected by 31P—NIVIR in RhVE-l. Therefore, this is some question about the prOposal that polymer-supported free phosphine exchanges with complexed monomeric phosphine. Also, the Pj—PC/Rh value, which was estimated from the decrease of free polymeric phosphine, is 1.85 for RhV2, 1.95 for RhVB, 2.3 for RhV“, and 2.2 for RhVS. These Pj—PC/Rh values revealed that most of the catalyst is present as (Pj-P02)2RhCl(Co). In the case of RhV RhV and RhV samples prepared 1' 2 3’ by ligand exchange reactions of RhCl(Co)(P0 with 3)2 phosphinated beads, the peak area ratio (1M) of the free polymer-supported phosphine and the complexed phosphine to the external reference sample (tri—n—butylphosphine) is about the same as the peak area ratio (Ip) of polystyryl- phosphine (prior to metalation) to the external reference sample. In the case of RhV“ and RhV5, samples prepared by the reaction of the phosphinated beads with [RhCl(CO)2]2 in benzene have an I much smaller than IP’ and the value M of I increased after the P03 treatment. In other words, M 144 the relative peak area of the low field resonance increased after treatment with triphenylphosphine. The above experimental results support the conclusion that the polymeric triphenylphosphines react with RhCl(00)(P0 or [RhCl(CO)2]2 to form (Pj-P02)2Rh01(co) 3)2 by ligand exchange and chlorine bridge splitting respectively. The complexed polymeric phosphine then exchanges with the free polymeric phosphine. However, in the case of [RhCl(CO)2]2 equilibration, the rhodium was distributed only in the outer region of the bead (Figure 30) due to the ease of ligand substitution of 00 by P03. The local crosslinking is dramatically increased. Thus, a very broad resonance for complexed phosphine was observed. Most of the free polymer-supported phosphine detectable by 31P NMR is in the center part of the bead. Consequently, the exchange rate is extremely limited by the mobility of free polymeric phosphine in the bead center. Since added P03 catalyzed redistribution of the metal throughout the whole bead (Figure 31), polymer-supported free phosphines must exist around the metal eomplexes and exchange with them. The increased relative area peak of the very low- field signal resonance might be due to increased mobility resulting from a reduction of local crosslinking. The coupling constant J in RhCl(CO)(P0’3)2 was Rh-P reported to be -124 Hz. In RhCl(P03)3, the coupling constants are JPl-Rh = —l92 Hz, JP -Rh = —146 Hz, where Pl 2 145 is the phosphine trans to chlorine. The strong interaction between phosphines and rhodium in RhCl(P03)3 may be the cause of the disappearance of the complexed polystyryl— phosphines 31P resonance. As we can see from Figure 24—26, ligand exchange in sample RhV occurs more slowly than in sample Rth. The 3 coalescence temperatures are 70° for Rth, 85° for RhV2 and 110° for RhV3. A homogeneous system shows that the exchange is associative and the rate proportional to the concentration of triphenylphosphine to 0.23 power. The difference of the exchange rate between Rth and RhVB, which have the same phosphine concentration but different rhodium loading, may be ascribed to this rate relationship. The Swelling Values—~g The degree of swelling values for phosphinated beads (1.8% crosslinked) were obtained by measuring the swelled volume and dry volume of the beads in a graduated cylinder The swelling value, q, is defined as the relative increase in volume of beads on adding solvent (q = V(swelled)/V(dry)). Some q values for 32 to 35 mesh phosphinated beads with 1.8% crosslinked are presented in Table 17. Some q values which were measured by E.M.Sweet, for unfunctionalized beads are also presented in this table for comparison. The smaller q values of the functionalized beads may be ascribed to the increase of crosslinking during functionalization. 146 Table 17. The degree of swelling values—q. Solvent qa(phosphinated beads) qa(unfunctionalized) Benzene 3.05:0.08 3.90:0.20 THF 3.0610.04 3.65i0.12 Toluene 2.95i0.08 3.45i0.l8 Dichloromethane 3.27:0.13 Cumene 1.98:0.07 Ethylacetate l.98i0.05 Acetone l.68i0.02 Diethylether 1.35i0.04 2.05:0.11 NitrobenzeneC 2.59:0.08 2.35:0.13 Cyclohexane 1.00 1.95:0.10 a q = V(swelled)/V(dry)- b Data of E.M. Sweet. 0 Beads float. III. EXPERIMENTAL Materials Solvents such as cyclohexane,benzene, toluene and xylene were reagent grade and were distilled under nitrogen from sodium or potassium benzophenone ketyl. Solvents for measuring the degree of swelling values were used as obtained, except for nitrobenzene. Acetone was dried over anhydrous CaSOu and distilled under argon. Dichloromethane was distilled from CaSOu and stored in the dark. Chloroform was washed with water to remove ethanol. After drying over CaCl2 it was distilled and stored under argon in the dark. Ethylacetate was purified by washing with aqueous 5% Na2003’ then with saturated CaCl2 and drying over CaSOu. The solvent was then further dried with molecular sieve before distillation. Nitrobenzene was purified by crystalization from absolute ethanol (by refrigeration). The ethanol was then removed at reduced pressure and the nitrobenzene was stored in the dark under inert atmosphere. It is very hygroscopic. Argon was purified by passing through BASF-BTS catalyst heated to 140°C and 4: molecular sieves. Biscyclooctenerhodium(I) chloride dimer and chloro- tris(triphenylphosphine)rhodium(I) were prepared as 147 148 described in part I. Bis(triphenylphosphine)carbonyl- rhodium(I)chloride was prepared by L.C. Kroll. The other reagents were obtained and treated as described in part I. Instrumentation (I) DA-60 NMR spectrometer A modified Varian DA—60 NMR Spectrometer was used to do all the phosphorus-31 studies except the studies of the 31P spin-lattice relaxation.times(Tl), for which a Bruker HFX—90 NMR spectrometer was used. The spectrometer consists of a modified specilities MP—1000 Spectrometer, a Nicolet 1083 computer with 12K of memory, and a home— built interfaceéé. The configuration is shown in the form of a block diagram in Figure 32. The frequency source of the RF unit is 56.44 MHz. The resonance frequency of the sample can be obtained by changing the synthesizer frequency, instead of changing the magnetic field. By utilizing the mixing network and tunable probe, this spectrometer can be used to observe NMR signals in the range of about 2 to 35 MHz. Before each experiment, the probe and image filter were well tuned to the phosphorus—31 resonance frequency, which is 24.289 MHz at this field. The homogeneity was also well adjusted, so that a 1.5 M solution of triphenyl— phosphine in toluene will give 29.3 Hz (or smaller) of half width. Exponetial multiplication yielding 3.89 Hz .006080060000 mzz 0006050166658 001_mUu~_ m fi12< UO " .IIIL. "I IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII L 1 _ ‘ u . _ 300.0 m .03: .622 0.0.0 , Swat—m ” 9 a m 50.001025 , . 06009 , 5600.2 _ M " 520002: 1 _ 5:: l .................. . . 00s. 39 1 fl . 0 m ._ . _ _ _ wan“ .22 610050002. m z . 5:532. 005.562. . 551 20¢; _ . - UD _ . . L m < _ w >000 m < a o» H 56.50 :56 “ -.. 230: > 000 u :2: E . 00.- 566:... 20050 n - 00> " l 5:320»: L ..OOV _Zw 150 line broadening was applied to the free-induction decays. The External Lock The field is locked by a home-built lock probe, which uses the DA—oo console to lock on the proton resonance of the upfield sideband of H20. A 3 mm O.D. quartz capillary was filled with water doped with CuClZ-ZHZO to give a linewidth of c.a. ”Hz. The lock probe slides into the back of the pulse probe. Since the lock probe was not thermally insulated, it tended to follow the temperature of the experimental sample. With the lock inserted to its fullest extent into the pulse probe, it was determined that the lock sample attained a temperature of O.l°C when the experimental sample was at approximately -l70°C. In order to go to lower temperatures while still using the lock, the lock probe must be partially withdrawn from the pulse probe to decrease the thermal contact. In this manner locked operation was possible down to the lowest attainable temperature. Temperature Control and Measurement The temperature was regulated by a heater-sensor and the Varian V-4343 variable temperature controller. The temperature were measured by a copper—constantan thermocouple. Thermal gradients in the probe were 66 measured and they were not found to be strongly dependent on the actual temperature. The magnitude of 151 these gradient was not greater than 2.20°C in 4.0 cm. The average gradient was about O.2°C across a 0.5 cm sample. Samples were allowed 20-30 minutes of equilibrium time between each experiment. During the experiments the temperature was very stable. Sampling and Referencing The probe in the DA-60 requires 5, 10 or 15 mm o.d. tube. In most of the studies, a 15 mm o.d. Wilmad NMR tube was employed. In order to keep the sample free from oxygen, the NMR tube was equipped with a 24/40 Knotes inner joint and a side arm with $2A glass stopcock. A coaxial type of reference tube was used. The design of the sample tube and the reference tube are shown in Figure 33. The reference sample (tri—n—butylphosphine in toluene) was put into the 5 mm o.d.NMR tube, and the whole reference tube was sealed off. The sample tubes and the reference tube were made by the glass shop in Michigan State University. The reference tube may be not exactly coaxial, therefore marks on the sample tube and reference tube were used. During these studies, the samples were not spun. (II) Bruker HFX:90 NMR spectrometer The phosphorus-31 spin—lattice relaxation times (T1) were measured by using a Bruker HFX-9O NMR spectrometer equipped with a Nicolet 1083 computer with 12K of memory, a Diablo disk memory unit, and a Nicolet 293 I/O controller. 152 24/40 Inner Joint ‘—/ i F W hook ' \ ' 3 2A glass stopcock :‘\ Mark to line up /l . 41—15mm NMR tube 3355‘— Solvent . ': Reference :6 1.. Sample :o '31-— Functionalixd Beads y}. 3: : 9.2.0. .0. 9.2 ' '37-?" - ~ . O 0' 5mm NMR tube —> 36000. .0000 .0... 0.. Figure 33. The sample and reference tube for P-Bl measurements. 153 The spectra were obtained at a frequency of 36.44 MHZ. Probe temperature were measured by inserting a COpper— constantan thermocouple directly into a solvent filled tube in the NMR probe and reading the temperature from a 25— Doric Trendicator 400 digital output. A thermocouple fastened near the bottom of the sample tube regulated the flow of cooled nitrogen gas. The Measurement of Spin—Lattice Relaxation Times (Tl) Various eXperiments have been designed to measure Spin—lattice relaxation times. The most common method of measuring T1 is the so—called inversion—recovery method or partially relaxed NMR spectroscopy. This method consists of the application of pulse sequence 180°, 1, 90°, where T is waiting period between the two pulses. The pulse sequence then, is abbreviated [180~T—9O—(sample)—T:]n where n is the number of FID to be summed and T is the delay time which must be at least 5 times the longest T1 in the system, to allow for complete recovery. As shown in Figure BU—ElJA and [2]A, the 180° pulse inverts the magnetization along the ~Z' axis. After a time T, spin—lattice relaxation has caused M to decrease (Figure BU—[ljB through zero if T is long enough to increase toward its positive equilibrium value (Figure 34—[2]B). The 90° pulse applied at time T samples the vaue of M at the instant by turning M to Y axis where it generates a FID. Fourier transformation of the FID gives a partially [2] r> T, In2 154 "IN ll Ho Figure 34. Representation of an inversion method for T1 measurement. (A) A 180° pulse is applied. (B) The system is allowed to relax for a period of T. (C) A 90° pulse is applied. The magnetization vector then returns to equilibrium. 155 relaxed Spectrum in which the intensity of the resonance line is proportional to the Z magnetization at the time of the 90° pulse (nonequilibrium state). The system is then allowed to relax back to equilibrium and the pulse sequence repeated with another value of T. Thus a series of Spectra was measured as a function of r and stacked in such a way as to creat the impression of a three dimentional plot of frequency and time. Accurate Tl's can be calculated from a plot of T versus the log of the difference between the intensity of each line and its equilibrium value. The phosphorus-31 Tl measurements were performed using the [l80°—r—90°—T]n sequence. In a typical experiment, eight tau values were utilized, with the initial and final 1 value being equal to T, which was chosen to be h-B times the value of T1. The spectra were measured over a 10000 Hz Spectral width. The pulse width required to flip the Spins 180° was determined by pulsing a sample of 85% H3P04 near the resonant frequency. The value of the pulse width was then varied until the experimental FID or FT Signal had minimum intensity. A 90° pulse is exactly half the value of a 180° pulse. About 200 pulses for each T value were necessary to achieve a sufficiently good signal to noise ratio to calculate Tl values. Since the signal to noise ratio of a frequency domain spectrum can be improved by the application of a 156 multiplication exponential with a negative exponent, the eight FID's acquired during each T expermient were 1 subjected to an exponential multiplication with —5 as the exponent and then Fourier transformed under identical conditions. The frequency domain spectra were plotted and the T1 was calculated by the T1 program which performed a least squares analysis to fit a straight line to the equation In (l-A/Aw) = -T/Tl + 1n 2 where A,ADo is the intensity of Spectra with T values equal to r, t»- Thus, a plot of ln(l-A/Aw) vs. 7 gives a line with SlOpe -l/Tl and a theoretical intercept of 0.693. Much has been published concerning the errors involved in the measurement of relaxation times by the various pulse procedures available. Errors arise from a variety of sources, principally inhomogeneities in HO and H1, incorrect pulse powers, incorrect timing, and inaccuracies in phasing. Preparation of Reference Sample A 0.4 M of tri—n—butylphosphine solution in toluene was used as a reference sample. A 25 ml. pear—shape flask with side arm was equipped with a gas inlet tube. Tri-n-butylphosphine (density 0.812 g./ml.) 1.15 ml was transfered into the flask after it has been purged with argon and mixed with 10.35 m1. of dry, oxygen-free toluene. 157 The concentration of this solution iS 0.0 M, if the mixture is additive. About 1 ml. of this solution was then placed into the reference tube under argon, and the tube was sealed off under vacuum. All the samples and solvent were transfered by means of syringes. General Preparation of the Samples for P—3l NMR Measurements A 0.9-1.2 g. sample of the beads was put into the sample tube under argon and sealed with a 24/40 outer type of stopper. The system was evacuated and flushed with argon three times. Then three to four ml. of the desired solvent was added from the tOp while the tube was under argon pressure. The tube was sealed off and the beads were allowed to swell for at least 2 hours before the P—31 Spectra were measured. However, in most cases the beads sample were allowed to equilibration for more than 20 hrs. In order to obtain the content of polymeric phosphine present in the beads, the reference tube was put into the sample tube and the marks were lined up. They were then lined up with a guide mark on the probe. During this studies, a 3.5 seconds of delay time (time between pulses) was used. The area ratio of a sample and a reference was changed with the delay time used due to the longer T1 of tri—n—butylphOSphine sample. Note : The height of the swelled beads in the NMR tube must be greater than the length of the coil of the probe. 158 Preparation of Polymeric Triphenylphosphine for Chelation Studies Alloperations were carried out under argon with solvents prepared as previously described. The standard method, described in part one was used to prepare the phosphine substituted polystyrene c0p01ymer. Five batches of different per cent substitution of phosphinated beads were prepared by varying the relative ratios of reactants and the reaction time. The amount of COpolymer used, relative amount of reactants and time for the lithiation are summarized as follows Temp. Time ml of %P Batch Beads(g.) %n—BuLi°TMEDA (00) (hr) 01P¢2 1 19.62 50 67 7 26 2.84 2 13.40* 20 65 6 no 65 12 23 3.83 14.93 20 74 6 17 2.72 4 10.02 10 7o 11 10 1.73 5 12.19 5 65 1% 6 1.27 The beads were treated with a solution of 20% lithiating reagents, washed with cyclohexane and then treated with a solution of 40% lithiating reagents. The above beads were used to prepare the following metal complexes for P—3l studies. 159 Preparation of Polystyrylphosphine Metal Complexes for PhOSphorus—3l Studies All solvents were dry and oxygen-free, and all reactions were carried out under argon. Solvent transfer was carried out by means of a syringe. RhCl(P03)3 studies A 1—2 g. sample of phOSphinated beads from batch 1 containing 0.95 meq phosphine per g. of beads was treated with RhCl(P03)3 in benzene for several days. The beads were then washed with benzene until the rinses were color- less. They were then vacuum dried and stored under argon. Sample of IA—IF were prepared as summarized in Table 18 [Rh(COE)2Cl]2 studieS—— Series II—V In a typical preparation, approximately 1 g. of a phosphinated beads sample were weighed into a 50 ml. Side- armed flask equipped with gas inlet tube. The required amount of biscyclooctenerhodium(I)chloride dimer was added. After evacuating and flushing with argon (three times), l5—30 m1. of benzene was added. The mixture was stirred under argon with a magnetic stirrer or with a shaker for a designated time. The 00polymer was then washed with benzene until no coloration shown in the rinses. The copolymer was then vacuum dried with or without heating and stored under argon. 160 Batch 2(1.237 mmole P/gbeads), 3(0.878 mmole P/gbeads), 4(0.560 mmole P/gbeads) and batch 5(0.379 mmole P/gbeads) of above phosphinated beads were equilibrated with [Rh(COE)2Cl]2 in benzene to provide sample of series II-V respectively. They are summarized in Table 19-22. The dried samples were placed into a 15 mm of NMR tube and swollen in toluene as previously described. The content of free phosphine was then analyzed by P-3l NMR Spectroscopy. Phosphine Exchange in RhCl(C0)(Pg'3)2 DA-60 was used for this study. RhCl(CO)(P¢ 0.2986 g.(0.4321 mmole) was placed 3)2. in a 25 ml. bantamware flask equipped with a gas inlet tube and a septem stOpper. The system was vacuumed and flushed with argon three times. Chloroform 5.4 ml. was added to make 0.08 M of yellow solution. Purified triphenylphosphine , 0.5985 g. in a 25 ml. bantamware flask equipped with a gas inlet tube was vacuumed and flushed with argon three times. Chloroform 3.8 ml. was then added to make a 0.6 M solution triphenylphosphine. The 15 mm NMR tube equipped with side—arm and stOpper was vacuumed and flushed with purified argon three or four times. 2.1 ml. of 0.08 M of Rh01(00)(P03)2 and 1.1 ml. of 0.6 M P03 were introduced by the means of syringes. The solution was well mixed and it contained 0.052 M of RhCl(CO)(P0 and 0.202 M of P03 (Rh/P03 = 3.9/1). 3)2 161 Table 18. The preparation of samples IA—IF. Sample Beads used RhCl(P03)3 Rh/P Equilibration color (mmol P) (mmol) time (day) (soln.) IA 1.79 0.137 0.076 9 faint yellow 1B 1.00 0.107 0.107 18 yellow 10 1.36 0.189 0.139 9 yellow ID 1.00 0.162 0.162 18 yellow IE 1.41 0.395 0.280 28 red IF* 1.52 0.579 0.385 8 dark red Reflux in benzene for 5% hours. Analysis IE— Rhodium 2.29% (0.2225 mmol/g.beads) IF- Rhodium 2.62% (0.2546 mmol/g.beads) Phosphorus 2.73% (0.8814 mmol/g.beads) Table 19. The preparation of samples IIA-IIG. Sample Beads used [RhCl(COE)2:I2 Rh/P color of soln. (mmol P) (meq Rh) before washing IIA 1.404 0.094 0.067 clear IIB 1.419 0.163 0.115 clear IIC 1.328 0.188 0.142 faint yellow IID 1.493 0.286 0.193 faint yellow IIE 1.423 0.506 0.356 yellow IIF 1.622 0.924 0.569 yellow 110 1.827 1.554 0.851 yellow Analysis IIF? Rhodium 2.98% (0.2896 mmol/gbeads) IIGl-Rhodium 4.27% (0.4149 mmol/gbeads) IIGZ-Rhodium 4.06% (0.3945 mmol/gbeads) Phosphorus 3.41% (1.101 mmol/gbeads) * Analysis performed after the second treatment with P03 Table 20. 162 The preparation of sampleS.IIIA-IIIG. Beads used [RhCl(COE)2]2 Rh/P color of soln. Sample (m mol P) (meq Rh) before washing IIIA 2.05 0.107 0.052 clear IIIB 1.492 0.121 0.081 clear 1110 1.278 0.112 0.088 clear IIID 1.386 0.1678 0.121 clear IIIE 1.259 0.2963 0.235 clear IIIF 2.375 1.013 0.427 faint yellow IIIG 0.822 0.822 1.000 yellow Analysis IIIF- Rhodium 3.25% (0.316 mmol/g.beads) Phosphorus 2.24% (0.723 mmol/g.beads) IIIG- Rhodium 4.54% (0.441 mmol/g.beadS) Phosphorus 1.78% (0.575 mmol/g.beadS) Table 21. The preparation of samples IVA-IVG. Sample Beads used [RhCl(COE)2]2 Rh/P color of soln. (mmol P) ( meq Rh) before washing IVA 0.668 0.047 0.070 clear IVB 0.696 0.055 0.079 clear IVC 0.660 0.083 0.125 clear IVD 0.893 0.172 0.192 clear IVE 0.844 0.279 0.330 faint yellow IVF 0.780 0.433 0.556 yellow IVG 0.962 0.837 0.870 yellow Analysis IVF- Rhodium 2.70% (0.2624 mmol/g.beads) Phosphorus 1.23% (0.3971 mmol/g.beads) IVG- Rhodium 3.94% (0.3829 mmol/g.beads) Phosphorus 1.23% (0.3971 mmol/g.beads) 163 Table 22. The preparation of samples VA-VG. Sample Beads used [RhCl(COE)2]2 Rh/P color of solution (mmol P) (meq Rh) before washing VA 0.622 0.034 0.054 clear VB 0.630 0.055 0.087 clear VC 0.579 0.088 0.151 clear VD 0.500 0.138 0.276 clear VE 0.425 0.152 0.357 faint yellow VF 0.447 0.257 0.576 yellow VG 0.397 0.338 0.851 yellow Analysis VG— Rhodium 2.94% (0.2857 mmol/g. beads) 164 Preparation of Polymer-Supported RhCl(CO)(P¢3)2 for Phosphine Exchange Studies Poly(styrene—divinylbenzene)copolymer(l.8% crosslinked), in bead form, was phosphinated by the previously described 16 technique . The functionalized beads were then equili- brated with RhCl(CO)(P0 or [RhCl(CO)2]2 in benzene. 3)2 The nonattached complexes were removed by washing with benzene until no coloration shown in the rinses. The preparation of sample Rth—RhV5 are listed in Table 23. The vacuum dried complexed c0p01ymer was swollen with toluene in a 15 mm NMR tube (as shown in Figure 33). The DA-60 multi-nuclei spectrometer was used for the 31P NMR measurements. A 3.5 seconds of delay time ( time between pulses ) was used for all the measurements. Table 23. The preparation of sample Rth—RhVE. Equilibration samplp meqP/gbeads meq Rh/gbeads Complex P/Rh RhVl 0.878 0.548 RhCl(00)(B23)2 1.60 th2 0.894 0.314 RhCl(00)(PQf3)2 2.85 RhV3 0.878 0.161 RhCl(co)(P03)2 5.45 th4 0.894 0.263 [RhCl(co)2]2 3.40 RhV5 0.894 0.234 [RhCl(CO)2]2 3.82 165 Scanninnglectron MicrOprobe Analysis The beads (about 150 mg.) were mounted on an aluminum stub using scotch double stick tape. The sample was then coated with 400 to 500 A of gold in a model EMS—41 sputler coater. The beads were then examined in a Super III scanning electron microprobe (International Scientific Instruments) operating with 10 KV accelerating voltage and zero degree specimen tilt. The Degree of Swelling Values. qp The degree of swelling values were measured by the method developed by E.M. SweetlS. About 2 ml. of phosphinated (ca. 13% substituted) beads were put into a 10 ml. graduated cylinder. After recording the volume of dry beads, the cylinder was filled with solvent. The volume of beads was taken periodically. The final volume was noted over a 24 hours period. Equilibrium was usually attained within a 2-hour period. The degree of swelling value, q, was then calculated. q = Swelled volume/dry volume REFERENCES 10. 11. 12. REFERENCES . N.K. Mathur and R.E. Williams, J. Macrol. Sci.—Rev. Macromol. Chem., 915, 117 (1976). (a) R.H. Grubbs and L.C. Kroll, J. Am. Chem. Soc., 23. 3062 (1971). (b) R.H. Grubbs, L.C. Kroll and E.M. Sweet, J. Macrol. Sci. Chem., A(Z), 1047 (1973). C.U. Pittman, Jr. and L.R. Smith, J. Am. Chem. Soc., 22. 1749 (1975). (a) R.H. Grubbs, C. Gibbons, L.C. Kroll. W.D. Bonds,Jr. and C.H. Brubaker, Jr., J. Am. Chem. Soc., 96, 2373 (1974). (b) W.D. Bonds, Jr., C.H. Brubaker, Jr., E.S. Chandrasokaram, C. Gibbons, R.H. Grubbs and L.C. Kroll, J. Am. Chem. Soc., 91, 2128 (1975). (a) J.C. Bailar, Jr., Catal. Rev., 19, 17 (1974). (b) C.U. Pittman, Jr. and 0.0. Evans. Chem. Technol., Sept. 1973, 560. (c) R.H. Grubbs, Chem. Technol., Aug. 1977, 512. .G. Allum, R.D. Hancock and I.V. Howell, Organometal. Chem., 81, 189 (1975). Nakamura and H. Hirai, Chem. Lett., 168 (1976). Dumont, J.C. Poulin, T.-P. Dang and H.B. Kagan, Am. Chem.Soc., 95, 8295 (1973). SD C—IE I-< CAPS Strukual, M. Bonivento, M. Grajiani, E. Cernia, and N. Palladino, Inorg. Chim. Acta, 12,15 (1975). C.U. Pittman, Jr., L.R. Smith, and R.M. Hanes. J. Am. Chem. Soc., 91, 1742 (1975). 0.0. Evans, C.U. Pittman. Jr., R. Mcmillan, R.T. Beach and R. Jones, J. Organometal. Chem., 51, 295 (1974). J. P. Collman et al. J. Am. Chem. Soc., 99, 1789 (1972). 166 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 167 T. Uematsu, F. Saito, M. Miura and H. Hashimoto, Chem. Lett.. 1977, 113. T.0. Mitchell and D.D. Whitehurst. Presented at the Third North American Conference of the Catalysis Society. San Francisco, February 1974. Edward M. Sweet, Michigan State University, Ph.D. Dissertation, 1976. R.H. Grubbs and Shin-Chin, H. Su, J. Organometal. Chem., 122. 151 (1976). G.G. Eberhardt and W.A. Butte. J. Org. Chem., _2__9_, 2928 (1964) Fieser and Fieser : "Reagents for Organic Synthesis". 2, 403; 3. 284; 4, 485-489. A.J. Chalk, J. Polym. Sci., Part B. 6, 649 (1968). D.C. Evans. M. H. George and J.A. Barrie. J. Polym. Sci., Polym. Chem. Ed., 1;, 247 (1974). D.C. Evans, L. Phillips. J.A. Barrie and M.H. George, J. Polym. Sci., Polym. Lett. Ed., lg. 199 (1974). C.D. Broaddus. J. Org. Chem., 35. 10 (1970). T.M. Fyles and 0.0. Leznoff, Can. J. Chem., 53. 935 (1976) R.H. Grubbs and E.M. Sweet, Macromol.. 8. 241 (1975). R.H. Grubbs, C.P. Lau. R. Cukier and C. Brubaker, Jr., J. Am. Chem. Soc., 99. 4518 (1977). L. Porri. A. Lionette, G. Allegra and A. Immirzi, Chem. Commun.. 1965. 336. R. Cramer. Inorg. Chem., 1, 722 (1962). R. Cramer,”Inorg. Synthesis” XV. 17 C. O'Conner and G. Wilkinson, J. Chem. Soc., 1968 A. 2665. C.H. DePuy. B.M. Ponder and J.D. Fitzpatrick, J. Org. Chem., 29. 3508 (1964). R.D. Gorsich, J. Am. Chem. Soc., 82, 4211 (1960). 32. 33- 34. 35- 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 168 J.L. Dye and V.0. Niceley. J. Chem. Edu.. 48, 443 (1971). (a) R.R. Ernst and W.A. Anderson, Rev. Sci. Instr.. 2. 93 (1966). (b) R.H. Ernst. ”Sensitivity Enhancement in Magnetic Resonance” in Advances in Magnetic Resonance. edited by J.S. Waugh, Vol. 2, Academic Press. New York, 1966. F. Bloch, W.W. Hansen and M. Packard. Phys. Rev., 62. 127 (1946). N. Muller and J. Goldenson, J. Am. Chem. Soc., 78. 3557 (1956). J.H. Van Wazer. C.F. Callis, J.N. Shoolery and R.C. Jonrd. J. Am. Chem. Soc.,28, 5715 (1956). R.A.Y. Jones and A.R. Katritzky, Angew. Chem. Intern. Ed. Engl.. 1. 32 (1962). M.M. Crutchfield, C.H. Dungan. J.H. Letcher. N. Mark. J.R. Van Wazer, "Topics in Phosphorus Chemistry, Volume—5, 31 Nuclear Magnetic Resonance ", John Willey & Sgns, New York, 1967. (a) D.R. Eaton and S.R. Suart, J. Am. Chem. Soc., 99, 4170 (1968). (b) T.H. Brown and P.J. Green, J. Am. Chem. Soc., 92, 2359 (1970). (c) P. Meakin, J.P. Jesson and C.A. Tolman, J. Am. Chem. Soc., 94, 3240 (1972). D.D. Lehman. D.F. Shriver and I. Wharf, Chem. Commun.. 1970. 1486. J.A. Osborn. F.H. Jardine. J.P. Young and G. Wilkinson, J. Chem. Soc., 1966,A. 1711. H. Arai and J. Halpern, Chem. Commun.. 1971, 1571. 1. Ya Slonim and A.N. Lynbimov, "The NMR of Polymers" Plenum Press. 1970. J. Schaefer. E.0. Stejskal and R. Buchoahl, Macromol., 8, 291 (1975). M.F. Froix, D.J. Williams and A.O. Goedde. Macromol., 9, 354 (1976). 46. 47. 48. 49. 50. 51. 52. 53- 54. 55- 56. 57. 58. 59. 60. 61. 62. 63. 169 Y. Inoue and T. Konno. Polymer J.. 8, 457 (1976). Kang-Jen Liu, and Burlant W., J. Polymer Sci., A‘lr it 1407 (1967). D. Doskocilova, B. Schneider and J. Trekoval, Collect. Czech. Chem. Commun.. 99, 2943 (1974). D.D. Tao, D. Doskocilova and J. Stamberg, Angew. Makromol. Chem., 38. 129 (1974). K. Akasaja, BiOpolymers..19, 2273 (1974). J.I. Crowley and H. RapOport, Acc. Chem. Res..9. 135 (1976). S.L. Regen. J. Am. Chem. Soc, 96, 5275 (1974). P.J. Hendra, Adv. Polym. Sci., 6, 151 (1969). (a) M.W. Duch and D.M. Grant. Macromol., 3, 165 (1970). (b) 0.0. Levey. Acc. Chem. ReS.. 6. 161 (1973). R.H. Grubbs, S.—C. H. Su and E.M. Sweet, 1977 Spring ACS meeting, Polymer Preprints. P. Uguagliati, G. Degamello, L. Busetto and U. Belluw. Inorg. Chem., 8. 1625 (1969). D.F. Steele and T.A. Stephenson, J. Chem. Soc. Dalton, 1972, 2161. L.D. Rollmann. Inorg; Chimi. Acta, 6, 137 (1972). G. Strathdee and R. Geven, Can. J. Chem., 52, 3000 (1974). H. Schumann, M. Heisler and J. Pickardt, Chem. Ber.. 110. 1020 (1977). P.E. Garrou and G.E. Hartwell, Inorg. Chem., 15, 646 (1976). T. Mizoraki. N. Kawata. S. Hinata. K. Maruya and A. Ozaki, Proceedings of the International Symposium on the Relations between Heterogeneous and Homogeneous Catalytic Phenomena. Held in Brussed. 1975. S.J. Seymour and J. Jonas. J. Chem. Phys.. 21. 487 (1971). 170 64. S.0. Grim and R.A. Ference. Inorg; Nucl. Chem. Lett.. 2. 205 (1966). 65. F.P. Gasparro and N.H. Kolodny, J. Chem. Edu.. 59, 258 (1977). 66. David A. Wreight. Michigan State University. Ph.D. Dissertation. 1974. M11111(111111111111115S