‘ in viiqdulfl‘ulr , , ‘ 139%! I. I .‘ h » L ‘ '- . . ’. '-. _ {I “35144. 1'” . I w. .- . = 1.55;” . ' . vilsl c . ill- 1 fi’ ‘40.... x: Idfldn 1‘- . I: ll . . 'Vi' .BN‘L f| ..< p 4 s n ..t y. . Os 5 0‘ n I r ‘ ‘ D l .l A. .‘A I {’1‘ . v A ‘. on . - - {.anss.-«nn.9. Z . ... . . I. IOuQ’C‘. I‘l‘ ptylvvl‘ . ; t. ~ o w : .- z I vn\ul.ht TI... ‘1 .6 Q. 0 . A :-~«. c. . . i “u ... 1-14.. . - . .. -, t: . . -. . - , ...(7 ((10 . . .. . . o .. quuu. 1. A. «‘ n .v 2. .. ....- . . v. v v o u .. u ..n A . . ‘ u I ~ ~ tn V. A-IM w I“. IVI Nu..- t! . o .. . I I ~ ufw ‘ . . .. 03...... mt} . . . . Z a -. l?“ ‘ . . . n tyil... s v ... 4‘ O I A . 1%.-..qu alt . v .. _A .1. a v: _ 1.. -. .313. . .v ooooo . ‘ . ..-a . .n n . n1| |Ill I". ‘ I l's|. ‘Q‘l. V 0... I401 ‘ 7.32 V. ~ ~ I\ 3. .-. \Jfifinnm. :0 .v... ‘ .2 wmxflxmfifildwxzfl .h» ‘u 1 3‘.<3~|§|I9.v‘0\0“oflnfil J\ 0 c‘f‘l‘é‘l I . v! Q [It QC.1I.|& stir u... i. .u ....gkh n! .KHbohnlv...hl1. \: 'Jistl :94. J‘obl- lJIIn-h.."Iu‘ . I- cl Q 4- v, v n 1 ll :00 v .“||I\0|‘W| {« .Jll‘vs} \v 0.. o 0.: . I. 3 . - ain‘t ulnofbulIa-va .‘..‘1¢v\t|§1.y§ .. A .. n . Jo... ...»!!! \ J- n-v..‘¥l§.. I .laI‘I-zlo! cs; . . . s ‘ .... .m ... .1" o 1““. O‘.1Il.|1n3 \c... no .. an . t. -1 > . Infill. ml. I .u -v.‘ .3 \N‘qlal‘d C‘ll . o..vv~ c 1 . . L4 «v‘ v1.o.\ ....) h‘Vl 1 ..l “I“ t |.nOol N1. 3. I ‘.~ z . . k. in? .. v 4-¢AHN>PIHA..I“~|W¢M. . A v«~.io\1,“r“ 6%.1 o v v ., .‘1 Allyn/\‘filmmn I I '0. c‘tQ 9.1 c ‘ ..I.f..n.o ;. ll all . . O. Y5 I utt\lti pl)..|" . t..lr.l»‘.b:l 1.1"."U -.lo‘.,00 .d‘f‘? . 4!!le ‘1: »‘« ‘ 1. ..l 'iii‘o‘clz. I l.\l..t"‘kf""\'lr ll'4 :0 ,‘a$\1“-I.l‘n I 31.. .17., AI|IJ$IQII 19 zil: lyiflhfi wv -vl t A.'..0s'|n (. 1! (I‘. [Itls‘gv ‘l I}. . -‘vt‘... .9- .! .gh‘lb, l- 0 O ‘ 43:9.va100 x , oilild ‘ 9 .v ‘ ‘llltt - 1- :c D .l‘l'h‘r1 ‘ ‘ ‘ 31:33.- (It. 5 . . - . ...- . ll'...x...q..l..?..k|v- |||l..l.‘.. v I! ‘lit. 1 «IIII.J..Y { .I - 1.3.... . , F-I V? ,‘ Li /"" i a a l ...‘v ‘ ‘ .1 3:2" ...: l figs..- ~gra nggt 5 ’ «bull? e v.3;- . . .' . , ‘.‘ . '. ’ .gv, “r _ "- "" 1’ .4. ‘ . . & viii’ifl‘grfiaig‘jl a w - I‘-‘ "“x " - ‘ -4 «a. ..«, -, ,. ‘ t;-‘.>«_.%;:.--;:5n——— This is to certify that the dissertation entitled Gas Phase Transition Metal Ion Chemistry presented by Sunkwei Huang has been accepted towards fulfillment of the requirements for EA D degreein (l’fmifl/Vy 9%. flax/ire / Major professor MSU is an Affirmative Action/Equal Opportunity Institution 042771 MSU RETURNING MATERIALS: Place in book drop to remove this checkout from LIBRARIES 1—!!— your record. FINES will be Eharged if book is returned after the date stamped below. ~ :«m “r a” 2'31 am: h ' a: [9 3;? a "in. “an?“ 4-: In 5a.: *__ GAS PHASE TRANSITION METAL ION CHEMISTRY By Sunkwei Huang A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1983 ABSTRACT GAS PHASE TRANSITION METAL ION CHEMISTRY By Sunkwei Huang Catalysis attracts tremendously broad interest and con- stitutes an important research field for chemists. Although there are a plethora of papers being published to probe the catalytic behavior of surfaces and homogeneous organometallic complexes, the behavior of a single metal center is usually not well understood due to the complications of ligand electronic effects (for the former) and solvent effects (for the latter). Recently, Muetterties e£_al. have reported studies on metal clusters in an attempt to build a conceptual bridge between molecular and solid state chemistry. However, all of these fields have suffered from a lack of sufficient thermochemical and structural information and are still inevitably somewhat speculative in nature. Gas phase transi- tion metal ion/molecule reactions help to provide answers to the following questions: l. How do metal centers react with molecules in the absence of solvent molecules? 2. What factors control these reactions? 3. How can we apply these answers to obtain a better under- standing and control of macroscopic synthetic systems? In this dissertation, the chemistries of ions such as Fe+, Cr+, and Ni+ (containing multiple numbers of C0 ligands) with multifunctional organic molecules are report- ed. Mechanistic effects, ligand effects, the differences in chemistries of these metal ions and their reaction mechanisms with ethers and polyethers will be discussed. Also, a unique "double insertion, double 8-H shift" mechan- ism is proposed to explain the results of these metal ion reactions with cyclic polyethers. Since the reaction products we have observed in this ICR study are mass spectrometric peaks, the assignment of ion structures sometimes ambiguous, although some techniques such as double resonance, CID, and the use of labelled com- pounds can provide some useful information. Ab initio calculations can provide much information such as energy levels of various electronic states, orbital occupancies, etc. The structure of CrCHz+ is given as an example to elucidate the use of this technique. ACKNOWLEDGEMENTS I would like to thank my wife Chimiao Lieu for taking care of my daughter Catherine and all the hosework that enabled me to spend all my time doing research. Without her encouragement, I would not be where I am today. I also want to thank my parents and parents-in-law and all the members in our two families for giving me confidence and keeping the families in good shape so that I didn't need to worry about my father or father-in-law's sickness. Thank you, Dr. Allison. For four years, I have learned very much from you. I really appreciate your sincere direc- tion of my research work and also for giving me the courage to be a good scientist and teacher. I will always remember that you were the preceptor who put me in the right track for my later career and who helped me in many aspects.... I also want to express my appreciation to everyone in my group for sharing happiness, discussion and well being. Acknowledgements must also be made to Dr. Harrison for directing me to do the ab initio calculations. ii Page LIST OF TABLES.. .................... . ................. vi LIST OF FIGURES ....................................... viii LIST OF SCHEMES . ....... . ............................. ix A. THE TECHNIQUE - ION CYCLOTRON RESONANCE SPECTROMETRY 1. Introduction ................................... 1 2. The ICR Experiment ............................. l 3. ICR Facilities at Michigan State University.... 8 4. Interpretation of Experimental Data ............ 10 B. THE CHEMISTRY l. Purpose of Research In Gas Phase Organometallic Chemistry ....................... l5 2. Historical Review of Transition Metal Ion Chemistry.. ................ . ............... 20 3. Multifunctional Molecules: The Macrocyclic Effect ............ . ................ 43 4. The Gas Phase Chemistry of Iron, Nickel And Chromium Ions And Their CD Containing Ions With Linear Ethers, Cyclic Ethers, Cyclic Polyethers And Crown Ethers ............. 43 I. Fe(C0) T Reactions With Ethers And Polyethers ................................. 45 II. Cr(C0) + Reactions With Ethers And Polyet ers ................................. 95 III. Ni(CD) + Reactions With Ethers And Polyet ers ................................. l23 IV. Comparison of Fe(C0)x+, Cr(C0)x+ And TABLE OF CONTENTS Ni(c0)x+ in Their Reactions with Ethers ....l46 iii 5. Trends in First Row Transition Metal Ions In Gas Phase Reactions With Organic Page Molecules ...................................... 152 I. Reactions With Propane (C3H8) .............. l54 II. Reactions With Iodomethane (CH3I) .......... l59 III. Reactions With ISOpropylchloride (C3H7Cl).. 165 IV. Reactions Nith cis-2-Pentene (C5H1O) ....... l69 V. Reactions With l-Hexene (C6H12) ............ l75 VI. Reactions With 2-Pentanone (2-C5H100) ...... l80 VII. Reactions With sec-Butylamine (S-BUNHZ, C4H11N) .......................... l86 VIII. Conclusions ................................ l94 C. AB INITIO CALCULATION 1. Introduction ................................... 196 I. The Importance of Ab Initio Calculations... 196 II. Review of Ab Initio Calculations of Transition Metal Compounds ................. l98 III. Theory of Ab Initio Calculations ........... 200 2. Use of CrCH + As An Example of An Ab Initio Calculation ................................... 2]] 3. Discussion ..................................... 22l APPENDIX A. Schematic Diagram for Voltage Controls.... 228 APPENDIX B. Marginal Oscillator Setup in ICR Experiment ................................ 238 APPENDIX C. The Relationship Between Magnetic Field and Mass In The ICR Experiment ..... 239 APPENDIX D. Alternate CID Circuit for Conventional ICR ....................................... 24o iv APPENDIX E. Calculation Of The Collision Frequency For Co+ and \/0\/' ............ 243 APPENDIX F. Branching Ratios OF Fe+ Reactions With VOV .............................. 2'45 LIST OF REFERENCES ..................................... 249 Table 10 ll 12 l3 T4 15 16 17 LIST OF TABLES Page Collisional Parameters as Functions of Pressure for Co+ Reactions with Et20 ............ l2 Summary of Metal-Ligand Bond Dissociation Energies ...... . ....... . ......... ... ..... . ....... 20 Heats of Formation and Avera e Bond Ener ies of The Positive Ions From Fe%C0)5 and Ni?C0)4... 22 Fe(CO)5 Reactions with P- dioxane and P- dioxane- "d8 ..... ......... ....... .. ............. 52 Ion/Molecule Reactions of Fe+ with l2-crown-4 and l5-crown-5 ....... . ...... . ....... . ........... 68 The Reactions of Fe(C0)x+ with Ethers ........... 9l Cr(CO)5 Reactions with P-dioxane and P-dioxane- d8' .............. . ........... ... ................ TOO Cr+ Reactions with Cyclic Polyethers....... ..... ll2 Neutrals Lost In The Reactions of Cr(CO): With Ethers. .......... . ......... . ........... 120 Ni(C0)4 Reactions with P-dioxane and P-dioxane- d8 ..................................... . ........ 130 Ni+ Reactions With Cyclic Polyethers.. ....... ... l39 Neutrals Lost In The Reactions of Ni(CO)x+ With EtherSOOOOOO......OOCOOOOOOCOOOOCO ......... 145 Number of Reaction Products Observed Metal Centers In Various States of Coordination ....... l47 Reactions with Propane (C3H8)... ................ l58 Reactions with Iodomethane (CH3I) ............... 164 Reactions With Iodomethane (CH3I) ............... l68 Reactions with cis-Z-Pentene (C5H10) ............ l73 vi Table l8 l9 20 21 Page Reactions With l-Hexene (l-C6H12) .............. . l79 Reactions With 2-Pentanone (2-C5H100) .......... 184 Reactions With sec-Butylamine (S-BuNHz) ........ 192 Results of Ab Initio Calculations of CrCH; ..... 222 Figure 10 11 12 13 14 15 16 17 18 LIST OF FIGURES Page Block Diagram of ICR Spectrometer for Single Resonance Experiment .................... 2 The ICR Cell .................... . .............. 3 Modulat1on of VTrapping (frequency325 Hz) ...... 6 ICR Mass Spectra of Cr(C0)5 + p-Dioxane With Pressure Ratio l:l at Total Pressure 1.0 x l0-5 torr... .......... . ............ . ........... 7 Double Resonance Spectrum of m/e 140 in Figure 4 .................... . .................. 9 Energy Profile: CpNi+ and CH3CH0 .............. 27 Schematic Representation of The Potential Energy Surface For the Reaction Of A Transition-Metal Ion With An A1ky] Ha1ide00000000000000000 ...... 28 A Structure of Fe+-12-crown-4 After Fe+ Double Inserts Into Two C-O Bonds of lZ-crown-4 ....... 79 X-Ray Structure of l5-crown-5 in l5-crown-5 CUBrz.0.0.0..........OOOOOOOOOOOOOOO. .......... 80 6 6 Energy Levels of B1 and A2 States of CrCH2+.. 224 Energy Levels of 4B1 and 6A2 States of CrCH2+.. 225 Trapped Ion Cell Circuitry-Pulsing Section ..... 232 Trapped Ion Circuitry-Drift Section.... ........ 233 Trapped Ion Circuitry-Trapping Section ......... 234 Electron Filament Bias ....... ..... ............. 235 Timings 0f Trapped Ion Cell Circuitry in Figure 12. .......... .......... ...... . .......... 236 Experimental Setup of ICR ...................... 236 Schematic Circuit Diagram of CID Experiment For ICR... ......... .. .......................... 242 LIST OF SCHEMES Page Scheme I .............................................. 25 Scheme II ............................................. 29 Scheme III ............................................ 33 Scheme IV ............................................. 34 Scheme V .............................................. 35 Scheme VI ............................................. 37 Scheme VII ......................... . .................. 38 Scheme VIII .............................. . ............ 39 Scheme IX ..................... . ....................... 40 Scheme X .............................................. 56 Scheme XI.... ......................................... 64 Scheme XII... ......... ........... ..................... 74 Scheme XIII ........................................ ... 76 Scheme XIV. ............................ . .............. 77 Scheme XV ...................... . ...................... 81 Scheme XVI ............................................ 81 Scheme XVII ........................................... 82 Scheme XVIII .......................................... 82 Scheme XIX. ................................... . ....... 85 Scheme XX ............................................. 85 Scheme XXI...... ...................................... 103 Scheme XXII..... ...................................... 108 ix Scheme Scheme Scheme Scheme Scheme Scheme Scheme Scheme THE TECHNIQUE A. THE TECHNIQUE - ION CYCLOTRON RESONANCE SPECTROSCOPY 1. Introduction Ion Cyclotron Resonance (ICR) Spectrometry is a type of mass spectrometry especially designed for the study of ion-molecule reactions in the gas phase. ICR experiments provide a variety of chemical information such as acidity, basicity, heats of information, bond strengths, proton affinities and all types of information related to the chemical reaction]. There are quite a few reviews available on this techni- que1-9. ICR is based on the dynamics of charged particles in electric and magnetic fieldsB. The detection system con- sists of a marginal oscillator (M.0)10’11 12,13 or frequency sweep detector with phase sensitive detection14. The basic principle of ICR will be discussed. 2. The ICR Experiment A block diagram of an ICR spectrometer is shown in Figure 1. Figure 2 shows a detailed view of the three section ICR cell which is, in our laboratory, 0.88" x 0.88" x 6.25". The source is 2.00" long, analyzer region is 3.75" long and the collector is 0.50" long. The cell is placed between the poles of an electromagnet. It is housed in a stainless steel vacuum chamber which can 7 be evacuated to a pressure of 1.0 x 10' torr. DRIFT a TRAPPING VOLTAGES — mmnmu - cum AMPLIFIER OSCILLATOR — * .ETECTOR ' MODULATION [REFERENCE l ammo. $3333 OSCILLATOR ENERGY 5 SUPPLY — l , , EMISSION ' CURRENT - l I PHASE- [‘ CONTROL I SWEEP - ‘ SENSITIVE xx RECORDER CONTROL g DETECTOR _ g] l_2. Figure 1. Block Diagram Of ICR Spectrometer For Single Resonance Experiment The cell, situated in the magnetic field as shown in Figure 2, consists of an ion source, analyzer and collector. The electrons are emitted from a rhenium wire located outside of the cell by operating the electron energy and emission control as shown in Figure l. Emitted electrons follow magnetic field lines, forming a collimated beam which crosses the cell are collected by a collector located on the Opposite side of the cell, and are detected as emission current. The emission controller then regulates the emission current with a feedback circuit to provide a constant emission current. The electron current control can be used to adjust emission current (u A) and electron energy (0 to -100 eV bias). ISource IAmlyzer 'Conector r.f from 1 marginal oscillatorfi l Drift Plate Top Analyzer 1 Double Resonance r. f Oscillator _-L---- - --- .— I Drift Plate 1 Top Source ______1_ 1 1 1 , I I I 1 1 I Collector fl~ ' Source Drift Plate ’ — Drift Plate ._ _ _ Botto A Bottom Source In nalyzer a Figure 2. The ICR Cell Once ions are produced by electron impact, they will 'h move circularly in a uniform magnetic field B with angular frequency of _ gfi - w - mc (rad1ans/sec) (l) where m is the mass of the ion and c is the speed of light. Then, by applying a voltage E across the drivt plates of the ICR, the ions will be "drifted" down to the analyzer section where the M.0. detector is set to one specific fre- quency (for instance 153 KHz) with a drift velocity (va) of Ec Vd = if (2) One can then scan the magnetic field to bring ions hav- ing different angular frequencies due to different masses according to equation (1) to be at resonance with the mar- ginal oscillator. The ions can be detected as the power drop due to the power absorption from the LC resonant circuit of the M.o.‘°"‘. Data in this dissertation were obtained under normal drift-mode conditions using trapping voltage modulation and phase sensitive detection (See Appendix A for detailed hard- ware description). The marginal oscillator detector is based on the design of Warnick, Anders and Sharpl]. In this technique, ions are produced in the source by an electron beam with 70 eV electrons. In the presence of a magnetic field, ions move in a circular orbit in the XY plane. TO prevent ions from drifting to plates 2 and 4 in the Z direc- tion, a trapping potential is applied to these plates (V trap > 0 for positive ions, < O for negative ions), which creates a potential well near the center of the cell to trap the ions. By applying a voltage difference E across the top and bottom plates which are called drift plates (#l,3,5,6), the ions experience an E x E force, drifting toward the analyzer where the top and bottom plates (5, 6) form the capacitive element of the tank circuit of the marginal oscillator detector (See Appendix B). When ions move with a cyclotron frequency equal to the natural frequency of the tank circuit, they will absorb power from it and are detected. The S/N (signal to noise ratio) Of the detector output is greatly enhanced by use of a lock-in amplifier14. The reference wave for the lock-in amplifier is provided by a function generator, at a frequency of 25 Hz. To effectively improve S/N, the signal output of the marginal oscillator must be modulated at this frequency. This is done by using the same 25 Hz square wave to modulate the signal out of the cell, which can be easily accomplished by modulating the voltage on one trapping plate (e.g. plate 4) as shown in Figure 3. This modulates the presence of ions to be detected in the cell. If plate 2 is (+), and plate 4 is modulating, when 4 is (+), cations are trapped and when 4 is (-), all ions are swept out of the cell. As a result, the rf level of the marginal oscillator varies at the modulating frequency, and forms a 25 Hz modulated signal output to be processed by the lock-in amplifier. Usually, the marginal oscillator is operated at a frequency of g 153 KHz (wM.0) for convenience (see Appendix C). The magnetic field is varied (0-18 KG) so the cyclotron frequencies (we) of all ions present also vary. When w of c an ion matches wM 0 , power is absorbed and the ion is detected. +"T r 0 Ions Trapped In Cell Ions Are Not In Cell - VT .__ Figure 3 Modulation 0f VTrapping (frequency R 25 Hz) Figure 4 shows typical ICR spectra of Cr(C0)6, p-dio- xane and a 1:1 mixture (in pressure) of Cr(C0)6 and p- dioxane. 5 torr, a sufficient At high pressures, e.g. 1 x 10' number of ion-neutral collisions occur to produce ion-mole- cule reaction products. To umambiguously identify reaction sequences, an ion cyclotron double resonance (CDR) experi- ment is performed7’15. Consider the reaction sequence: + A + N -————+> 3* + M (3) B+ + N —————+ c+ + L (4) where B+ can be formed from many different precursors (A?) in addition to A+. Also, A+ can be a precursor forming different product ions (ET) in addition to 3*. If the magne- tic field is set to monitor B+ (B+ is in resonance with the M.0.) without scanning the magnetic field, another radio frequency signal can be introduced into the ICR cell so that A+ can gain power and be ejected. When this occurs, the N3 n13 x o...” gmmonm Horas 3 a . H Scum genome 5; oedxofivm +385 .8 $83 and. 6H .3 8.3a 8. 09 CV. \ . Ow. . .00. . . .O.m 00 0? ON ‘ I . n +805 52.8% Res. 0.85 o m U 528QO $2: $96.65 w o a +0 1.0.1 {1... 1‘11 - b in n h .. is. 1 2 Eda 3% are \ m ocOxofiia + «08.6 Lo 0535 E hem +6 e no intensity of 8+ decreases. By scanning the frequency we applied to the cell, we can eject all possible precursors A: forming B+ to get this "double resonance" response. Remember at constant B, wa = constant, therefore It is then predictable at which frequency ratio the inten- sity of B+ will decrease to indicate a unique reactant: product pair, since 3 8| 8 21> flea and wB = 153 KHz 8 Figure 5 shows a typical double resonance spectrum. In earlier ICDR experimentsg, the second oscillator was introduced in the analyzer region (on plates #5,6). This can lead to serious interference problems between this oscillator and the marginal oscillator which is also coupled to these plates. If applied in the source region, stronger second oscillating fields can be applied without affecting the mar- ginal oscillator. Appendix 0 describes other alternatives. 3. ICR Facilities at Michigan State University The ICR used in the experiments which will be discussed was built at MSU. .: gm 5 03 o}. no 5.5003 momma—omen 33.58 .m E on: was m w m .85 M. be noun. weaves fl . SR. w. +8 0 «HR. 1 10 The filament emission controller and plate voltage con- troller for the ICR cell were designed and constructed by Dr. M. Raab and Dr. J. Allison in the Department of Chemistry at MSU. The marginal Oscillator detector is based on the design of Warnick, Anders and Sharp]]. A Wavetek model 144 sweep generator is used as the secondary oscilla- tor in ICDR experiments. The ICR cell is housed in a stain- less steel vacuum system and is situated between the pole- caps of a Varian 12" electromagnet (1.5" gap). The electro- magnet is controlled by a Varian V-7800, 13 KW power supply and Fiedial Mark I magnetic field regulator. The instrument is pumped by a 4" diffusion pump with a liquid nitrogen cold trap, and an Ultek 20 l/s ion pump controlled by an Ultek 150 mA ion pump controller made by Perkin-Elmer. The lock-in amplifier used to enhance S/N is model 128 A, (0.5 Hz - 1000 KHz) from EG&G Princeton Applied Research. Samples are admitted from a dual inlet (separately pumped by a 2" diffusion pump and liquid nitrogen cold trap) by Varian 951-5106 precision leak valves. Approximate pressures are measured using a Veeco RG 1000 ionization guage. 4. Interpretation of Experimental Data Data were acquired in the following manner. High and 5 6 1<3w pressure (1 x 10' torr is. 1 x 10' torr) spectra of each compound were taken, and ion-molecule reaction products 11 in either "metal carbonyl" or "organic molecule only" experiments were determined. The "extra“ peaks observed in the mixture were taken as ion-molecule reaction products formed in 1:1 or 1:2 mixtures (in pressure) of metal car- bonyl to organic, at a total pressure of l x 10.5 torr as shown in Figure 4. All the product ions were then studied to unambiguously identify the precursor ions by the ICDR technique. The following points should be considered in the interpretation of ICR experiments. a. Ions are produced with thermal velocities, since the ions are trapped by small trapping voltages. Fast ions will not be trapped and thus do not react with neutral molecules in these experiments. The application of drift voltages only increases ion velocities by 10%. This contrasts with conventional mass spectrometers in which ion extraction from ion sources requires strong electric fields, accelerating ions to some keV energies. b. Ions have relatively long residence times in the analyzer region. Residence times of milliseconds or more enable the observation of gas phase ion-molecule reactions. This contrasts with microsecond intervals in conventional mass spectrometers. c. Low pressure experiments yield mass spectra with fragments experiencing no collisions. Table 1 shows the reVlationship between collisional parameters and pressure. 12 Appendix E shows how to calculate the Langevin collision rate, collision frequency and time between collisions. d. At higher pressure, gas phase ion-molecule reac- tions are observed. From these, (1) kinetic studies can be performed; (2) double resonance can be used to identify reaction sequences; (3) branching ratios can be calculated. (Appendix F) e. Various thermochemical quantities can be deduced from observed ion-molecule reactions. Upper and lower limits on heats of formation of charged and neutral species, and limits on bond strengths can be deduced. It is assumed that processes which are observed must be exothermic or thermoneutral. Table 1. Collisional Parameters As Functions of Pressure For Co+ Reaction With Et90 Time Between Collisions Pressure Number Density Neutral-Neutrala Ion-Moleculeb torr I molecules/cm3 sec. sec. 10‘2 3.24 x 1014 1.68 x 10'5 2.67 x 10'6 10'3 3.24 x 1013 1.68 x 10‘4 2.67 x 10'5 10'4 3.24 x 1012 1.68 x 10'3 2.67 x 10‘4 10'5 3.24 x 10H 1.68 x 10'2 2.67 x 10'3 10"6 3.24 x 1010 1.68 x 10“ 2.67 x 10'2 10'7 3.24 x 109 1.68 2.67 x 10'1 ——¥ a. Time between collisions]: ZR] /mEt / 4/?62p b. d(Et 0) = 8.78 x 10"24 cm3, K culeg, see Appendix E. L 1.154 x 10"9 cm3/mole- 13 For example, Fe+ + \/0\/ —————+ Fe(C2H60)+ + C2H4 . _ 18 s1nce C2H50C2H5-—————+ CZHSOH + CZH4 AH - 16.8 kcal/mole + . therefore D[Fe - (o ———-c2H5)] > 16.8 kcal/mole, H for the overall reaction to be exothermic. f. Since most ion-molecule reactions occur with a rate that is within an order of magnitude of their pre exponential factor, m l x 10'10 cm3 molecule-1 sec'], it is assumed that they occur with essentially no activation energy. This follows from the Arrhenius equation k(T) = A e(-Ea/RT) -1 10 where A is on the order of 10101 mol'lsec , namely 10' cmz‘lmoleculeqsec-1 for bimolecular gas phase reactions. 9. In the absence of solvent, all of the energy of Observed reactions must be accounted for by the products. Since we are assuming gas phase reactions are exothermic, the product ions formed must be more stable with respect to the reactants. In contrast, the energy can be dissipated through solvent molecules surrounding the product ions in solution. Because of this, there is rarely a single product In an ion molecule reaction (P+ + N -—— I+*). Bonds are +11: usually broken and 1+* fragments. At higher pressures, I can be stabilized by collisions, which is then called an 14 u . . n + +* + * addltlon product (P + 2N ——> I + N ——91 + N). h. For some endothermic reactions, the energy barrier can be overcome by raising the kinetic energy of a reactant‘g. THE CHEMISTRY B. THE CHEMISTRY 1. Purpose of Research In Gas Phase Organometallic Chemistry. The work which will be described in this text deals with the gas phase chemistry of a number Of metal- and metal-containing ions with various neutral organic sub- strates. The purposes of studying gas phase organometallic chemistry include the following: a. Gas phase results are used in kinetic theorieszo'2]. b. Gas phase reactions can be used to model condensed phase chemistry. Catalysis draws tremendously broad interests and consti- tutes an important research field for chemists. Although there are a plethora of papers being published to probe the catalytic behavior observed in surface chemistry and homo- geneous organometallic complex523, it is not well understood what occurs at a metal center due to the complication of ligand electronic effects31 for the former and solvent effects for the latter24. 0n the other hand, researchers 25 studies claimed that metal 26,27 in heterogeneous metal catalysis clusters play an important role in catalysts , which 28’29 and can be used as an interface between molecular and solid state chemistry27. agrees with ab initio studies Moreover metal-carbon double and triple bonds are frequently 30 proposed in mechanisms for such reactions However, all these fields have suffered from a lack Of sufficient 15 16 thermochemical and structural information and are still inevitably somewhat speculative in nature. Building blocks of this nature allows the facile correlation of vast amounts of chemical data. Gas phase metal-ion/molecule reactions can serve this purpose. Other processes in organometallic chemistry whose gas phase ionic chemistry parallels can be readily studied are: 1) Hot atom chemistry32 2) Atmospheric chemistry33 3) Matrix isolation studiesBz’34 4) Metal catalysis chemistry 5) Organometallic chemistry and the solution chemistry of metal complexes (e.g. ligand substitution reactions) c. The study of reactivity trends Gas phase ion-molecule reactions allow the investigator to study chemical dynamics and chemical events in the absence of solvent complications. At Operating pressure (10'5 m 10'6 torr) used in mass spectrometric techniques such as ICR and 35 experiments, single collision events (the basic ion beam unit of chemical reactions) can be studied, namely two species come together to react and separate as products. In this work, reactants are metal ions (M+),_metal-containing ions (ML:) and organic neutrals (A). By varying M, L, and the structure of A, we can study how metal ions react, and what factors affect their reactivity, and thereby 17 characterize the mechanisms of such reactions. d. Gas phase chemistry is suggestive of condensed phase processes The results from gas phase chemical studies are often suggestive of condensed phase experiments. Recently, 36 have demonstrated an intriguing Kametani and Fukumoto technique called "retro mass spectral synthesis". Fragmen- tation of a compound on electron impact in a mass spectro- meter is frequently very similar to chemical degradation reactions39. For example, cyclohexene decomposes to give butadiene and ethylene. This compound fragments on electron impact to give C4H6+ and C2H4. Cyclohexene can also be synthesized by butadiene and ethylene by a Diels-Alder reac- tion. Mass spectral fragmentations frequently, therefore, parallel synthetic pathways as well as degradation pathways. Kametani and Fukumoto have used this observation to develop new synthetic pathways for a number of natural product536. There are many ion-molecule reactions in the gas phase 37’38. Similarly, the nature having condensed phase analogs of the bonding of organic molecules on metals can be under- stood from this work, which can be used to examine what has been proposed, for example, in methathesisBo. e. Thermodynamics Although reaction rates can theoretically provide an understanding of the effects of structure on reactivity in organometallic reactions, most observed reactions can be 18 understood in terms of thermodynamics. In other words, bond- strengths can be reflected in the reactivity as a driving force to initiate the reaction. Limits on heats of forma- tion of various complexes, and limits on bondstrengths can also be obtained. Limits on D(ML:-A) where A is an alkyl group, hydrogen atom, halogen atom, oxygen atom, alkoxy group, or various n and n-donor bases can also be obtained. For example, consider the reactions: Cr(C0)2+ + \V,O\,r -——4 Cr(Et20)+ + 2C0 Cr(c0)3+ + \v/O\/r —4P4 Cr(Et20)+ + 3C0 we found that D(Cr+-2C0's) = 66.2 kcal/mole and D(Cr+-3C0's) = 87.7 kcal/mole, therefore, we conclude that 66.2 kcal/mole < D(Cr+-Et20) < 87.7 kcal/mole. f. The study of mechanisms of organometallic reactions Since organic molecules contain alkyl groups, the interaction of a metal center and an alkyl group is very important. The following mechanisms are found useful to explain reactions observed in the gas phase: 1. B-Elimination. This process involves the shift of a B-H atom from the alkyl group onto the metal4o'42. + + CH2 M-CHz-CHR == HM--N R\\‘JH CHR 43 The relative stabilities of dialkyl compounds , CH m pHCH2 m (CH3)3CCH2 >> N-C3H7, 3 l9 N-C4H9 > CZHS > t-C4H9 > i-C3H7, shows the parallel to carbonium ion stabilities. 2. Reductive Elimination. This is a pathway for metal- 40,44 carbon cleavage with B-atom migration e.g. D 1 + +1i/|--C6H5 ————-> MH + CGHSD H + + (CH3)3AuPPh3-—-——-—-9 CH3AuPPh3 + C2H6' Other mechanisms proposed in catalytic studiesao’45 have not been used in gas phase organometallic reactions yet. The organometallic literature since 1960 has shown that the thermodynamics of bonding in organometallic compounds is of interest, however relatively few actual bondstrengths have been measured. This can be done in gas phase reactions as described in b. Table 2 summarizes the metal-ligand dissociation energies obtained from ion beam studies by J.L. Beauchamp, which will be of great use in interpretation of results in this thesis. 9. Li+ use has been demonstrated as a mass chemical 197 reagent . It gives simple spectra, easy to inter- pret. 20 Table 2. Summary of Metal-Ligand Bond Dissociation Energiesa (kcal/mol) _B_ Cr+-R Mn+-R Fe+-R Co+-R Ni+-R H 35:4 53:3 58:5 52:4 43:2 CH3 37:7 > 48 68:4 61:4 48:5 CH2 65:7 94:7 96:5 85:7 86:6 0 77:5 57:3 68:3 65:3 45:4 a. Data were taken from ion bean studies by J.L. Beau- champ 94-97 2. Historical Review Of Transition Metal Ion Chemistry From its birth through the mid 19605, most of the appli- cations of mass spectrometry have involved organic compounds. In the past two decades, however, applications have been extended into inorganic chemistry. Mass spectrometry is now commonly used to provide molecular weights and formulae of inorganic and organometallic compounds. Many thermodynamic data were Obtained from this field. There are quite a few specific areas of interest in the general area of gas phase metal ion reactions: a. Fragmentation Studies. (Unimolecular reactions) Early work was devoted to investigations of the ionization 52, their fragmentation patterns, and how these are affected by ligand553. potentials of organometallic compounds Other 21 extensive studies of fragmentation patterns in main group organometallicsso’54 55 50 , metal carbonyls , polynuclear metal 56,57 carbonyls have been re- and coordination compounds ported. A typical fragmentation following electron impact is illustrated below58: + ~———————+ Cr + :C(0CH3)CH3 OCH Cr —— C -? Cr--CCH CH '———————+ Cr---OCH + H2C=CH. 3 Most of this early work was done on volatile organometallic 59 252 compounds, but, field desorption Cf 6O , and recently 61 62 plasma desorption , laser desorption and 63 , secondary ion fast atom bombardment techniques have made possible mass spectrometric studies on nonvolatile compounds and even ionic substances such as saltssg. In addition, chemical ionization mass spectrometry (CIMS) studies have also been performed with organometallic com- pounds. Examples are CIMS (using CH4) Of sandwich com- 64,65 65 , metal hexcarbonyls , cyclopentadienyl metal , and arene metal carbony1s64. pounds ha1ides65 b. Photochemical Studies. The photodecomposition path- ways of several metal carbonyl anions66 67 such Ni(CO)3' and CO(C0)4' and cations have yielded useful thermochemical data on metal carbonyl fragments, such as heats of formation and average bond energies as shown in Table III. Table 3 . Heats Of Formation 699.319.1369 80an Emerges Of The Positive ions From Fe(CO)5 And Ni(30)4_ photoionization AH; (2v) Average bond threshold(ev) energies Ee(00)5 D[Fe-(UO)§]=1.2510.0Bev Re(oo); 7.08: 0.01 0.37: 0.02 D[Fe-(tio)5]=1.23:0.o3ev Fe(OO)Z 8.77:0.1 23110.1 Fe(CO); 9.8710.1 4.55:0.1 Fe(00); 10.6810.1 16.51: 0.1 Fe 00* 11.53101 8.51101 Ice“ 14.03101 12.35 10.1 111(00)“ DENi-(m)4]=l.5310.03ev Ni(00); 83210.01 2.07:0.02 D[N1-(m)4]=1.3610.03ev 111(00); 8.77: 0.02 3.58:0.02 Ni(00): 10.10: 0.1 6.14: 0.1 NICO+ 11.65: 0.1 8.841 0.1 NI+ 13.75: 0.1 12.09: 0.1 4. a. The assumed processes are 111(00)x + hv——-)(1:(Co)x_n]+noo 22 23 c. Gas Phase Ion-Molecule Reaction Chemistry in Organometallic Systems. (i) Metal and Metal Containing Ion-Molecule Reaction with Organometallic Compounds. The formation of ions of the type M2(C0): from ion-molecule reactions in metal carbonyls have been reported for the group VI hexa- 55’68. Fe(C0)567"71 and the negative ion chemistry 72 carbonyls of Ni, Fe and Cr carbonyls (with accompanying loss of 1 or 2 C0 groups) as follows: ———4 Fe2(00)7+ + co Fe(c0); + Fe(C0)5 ——+— + ‘———+ Fe2(CO)6 + CO N1(CO)§ + N1'(CO)4 --—-—+ N12(CO)g + CO Fe(C0)3 + Fe(CO)5‘———————+ Fe2(C0)g + 2C0 Cr(CO):1 + Cr((:0)6 ————-> Cr2(CO);3 + 200 Cr(C0)3 + Cr(C0)6 ———————+ Cr(C0)g + 3C0 Further reactions to produce tri-iron carbonyl ions (Fe3(C0);), and the formation of ions up to Fe4(C0);2 were also observed. From these studies and ligand substitution 70 with a series of n- and n-donor bases, proton 70,72 studies of metal carbonyls and bond energies70 affinities (PA) can also be obtained. For example, P.A. (Fe(C0)5) = 204 i 3 kcal/mole; D [H-Fe(c0);] = 23 : 1o kcal/mole. 24 73 Also, the gas phase ion chemistry of ferrocene and 74 nickelocene have been reported. Predominant features are charge transfer processes and the formation of a bimetal- . + 11c complex M2(C5H5)3, e.g. Fe+ + Fe(C H ) ——> Fe + Fe(C H )" 5 5 2 5 5 2 + + + FeCSH + Fe(C5H + 5 5)2 "“‘* Fe2(CSH5)3 Also, "triple decker sandwich" complexes were reported 75,76 for these compounds From these studies, the proton affinity of Ni(C5H5)2 was reported to be 218.9 1 1.0 kcal/ mole74. MUller also reported the gas phase ion chemistry of other sandwich compounds such as dibenzene chromium68°77. He has also studied the chemistry of n- and n-donor bases with ions formed from electron impact on C5H5CrC6 6’ 78 C H6, CSHSVC7H7, and CSHSCY‘C7H7 . H MnC 5 5 6 In addition, the ion-molecule reactions in 88 C5H5V(C0)47, C5H5Mn(C0)379 and C5H5Cr(C0)2N0 have also been studied by MUller. The ionic chemistry in mixtures of C5H5Mn(C0)3 with PF3, AsF3, SbF3 and SF4 was reported by MUller and Fender179. The ion-molecule chemistry of CpNiN0(Cp = C5H5) and the chemistry of its ions with n- and n-donor bases were also reported80’81. From this study, a 4. series of CpNi-B relative bondstrengths were determined81. 25 (ii) Metal and Metal Containing Ion-Molecule Reactions with Organic Neutral Compounds. J. MUller has studied the chemistry of metal-containing ions with hydro- 77,80 carbons , in which H2 loss with simultaneous formation of a new complex is observed: CpNiN0+ +[::)H -—————+. CerE:::1 + N0 + H2 -H + CpCO+ > H2; 2 CpCo[:::] CH CpNi+ :[::::>-————9 CpNi---“1 + CSH10 CH CpNiNO + [::::> —————9 CpNi- 1£::::>¢—————et CpNi ----- m: H2 loss processes parallel the normal electron impact frag- mentation mode for similar organometallic compound558, e.g. wk 1 ‘* ‘ -2co (CO)3 Fe ———————9 (c0) Fe+ -———————> Fe+C0 3 1 -2e n: -H2 ! J.L. Beauchamp has studied the gas phase chemistry of 82 of CpNi+ with alkyl ha1ides and found that the reduction alkyl halides (RX) by CpNi+ to olefins and HX is similar to that observed for Li+83. 26 H x CpNi+ + H—e CpNi+---)( + HX Decarbonylation of aldehydes was also observed82: ___, CpNiC0+ + RH CpNi+ + RCHO ——1- ,____, CpNiRH+ + C0 The energy profile for this process is shown in Fig. 6. Deoxygenation of acetone by metal-containing ions have also been reported78: + _——-+ C6H6VO + C3H6 + C6H6V + 0 =’ C(CH2)2——' ' + -——3 V0(CH3)2 + C6H6 Recently, J. Allison and D.P. Ridge have studied the transition metal ions of iron, cobalt, and nickel with polar organic molecules and found that these transition metal ions 84,85 always insert into a polar bond and is followed by a B-H atom shift (Scheme I): Scheme I. K CD D ‘ Fe+ + CD CH I ———> CD CH -Fe+-I,-—i— ll 2"1J4 —-——> CH CD Fe++ DI 3 2 3 2 CH2 2 2 11 + CHD 1' + + Fe(CHDC02) + H1 4— H ---Fe-—-I : CHZDCDz-Fe-I CD2 ( J 27 2on5 Ba 28 u unmade 885 .m 993... + m ... N528 :0 U/ a... .9. mo 38 z m/ a / m I n a 1.0.126 5 1.:8 m 2 o o a 38 + $26 --~‘_----.o 30(an n: 1 N x4 casinos. b: H a 02095 + Mano 28 This process is proposed to follow the reaction dia- gram of intermediates shown in Figure 7. + MX + R+ M +RX + Mt u\ + HX X-M—< x4141- 1 Figure 7. Schematic Representation of The Potential Energy Surface For The Reaction Of a Transition-Metal Ion With An Alkyl Halide. However, this mechanism does not Operate for amines86: Co+ + >th —--x--» Co+lk + NH3 + .____, CO(C3H7N ) + H2 .____, CO(C2H5N )++ CH4 + + + —-—+ Co(C2H5N ) + H2 .——-a CO(CH3N )++ CH4 29 Ion-molecule reactions of iron with ketones and ethers87 follow a similar mechanism to that of in Scheme I and is shown in Scheme II. Scheme II. + 0 /Fe r T + H 11 e ~> /’ ‘\ z’ \\ CD3 0.020113 01120113 0 o §> [003— Fe— C-cnzcaa] [CD/c—Fe— (1:2083] 3 B-anqa shift [Ifluft + - obj—ne—OO CH ”CD 011%“ -00 °§ C0208 GD/C‘ :6 —- ll2 3 3 C32 13-8 shift 4102;7/ 1413301“) 0 CD CD '1' ‘1' 2 1 lL—-Fe Fb'—-" 11- Fe? 00 flan—,GHZ‘Fe Loo /c\ 0112 0112:. 0112 0112 0113 II CD CD CD30D= 082 3 21 EC :gzoo -li£L—-—9 D— Fe- -CD fldELFe-w I Shift | H H ‘/CD 30 Scheme II (cont’d) .1. g F 11/“ \1 K (11 C C + 0112/ \CD2 CD2/ \CD2 0112/ \Fe: Fe + 1 I -—-—+ | 1 -——9 ./ CD CHZN /CH2 CHZ\ /% CH2 / 032 L CH2 J L \Cfiz/ CI"2 13-8 shift 0 ‘ , 1 1 1 I a / \ / \ +/ CD + CD 2 Fe - 2 l I an??? / \>;?CD2 Cflz‘culciiycnz ““2 /CH K J 1 K CHZ J 1-..... /F°+\ 1° CD + CKCH_CH¢ 2 /Fe finz .-cn . 2 “awrfl/ I“ '3 1 Sift shift IE>~E;t-o\€] 1:)F—EJL-j1~ ] C3H70 -03H6 l -03H8 31 Scheme II (cont’d) + {e F3 + 0 ——) o, 0 + 0 0 “1 0 13.11 () ¥ shift \Fe+ // -H .fi-H H shift shif I 0 1+ ‘ (I T + H-—-Fe-f> [:;:Fe : ‘nflf 1-82. 1 H0 H I" \ / 38* /Fe;/ .| . - 0 JHFe-F 1'12 + F8 Hence, Fe+ inserts into a carbonyl-carbon bond or a C-O bond in its reaction with ketones and ethers respectively and form a metalcyclic intermediate in its reaction with cyclic ketones and ethers. This is followed by a B-H shift, sfinilar to that in Scheme I. 32 The study of the gas phase chemistry of titanium ions 88 89 with haloalkanes , and alkenes has also been reported, e.g. Ti+ + RX 1r————9'Tix+ + R x = Cl, Br, I _————a R'+ + TiCl R = CC13, CFClz, MeZCH, CHC12 . + + . T1Cl3 + RC1 -—————+ R + 1101 . + . . + . T1C13 + CH3T1C13-—————+ CH3T1C12 + T1Cl4 . + . + . CH3T1C12 + C2H4 —-———+ C3H5T1C12 + 11014 . + 85% . + CH3T1Cl2 + C204 -—————9 C3H5T1C12 + H0 CH T'C1+ + c H —————4 C H 1°c1+ + H 3 ‘ 2 3 6 4 7 ‘ 2 2 T'C1+ + C H -—-———e H 1'C1+ + HC1 ‘ 3 3 6 C3 5 I 2 Note that titanium has only four valance sites. TiClg reacts with C3H6 by eliminating HC1 to give C3H5Ti+C12, pro- bably being a resonance stabilized allyl-TiCl2 cation (I) as shown below: 33 CH31i+C12 reacts with C204, 85% of the reaction pro- ceeds by HD elimination. This suggests a mechanism (Scheme III), in which, after association with the Ti cation center, the C204 inserts into Ti-C bond of CH3TiC1; as would occur in the polymerization of ethylene. Scheme III. C1 C1 CH \ .+ In,“ .+ 3 /T1 — CH3 + C204 ——> (T1 002 ————> c1 c1 / C02 T H _H Cl CH3 ‘3‘ \C 'I + \ ”1,". . ...-.\\ / \ /C02 ———> 111:, C—0 + H0 Cl CD2 c1] D/C/ ”'D The resulting species undergoes unimolecular decomposi- tion because it is unable to dispose of its excess internal energy in this gas phase bimolecular process. Consequently, 1,2-elimination of HD across the B-and y-carbons gives the allyl-TiClz+ cation (I) and does not react further. To be effective, mediation of reactivity of CH3TiC12+ by solvent, the nearby presence of a counterion and another molecule such as AlMe3 as in Ziegler-Natta catalysts are apparently necessary. J. Allison and D.P. Ridge have also reported the chemi- stry of titanium containing ions with alkenes90 and oxygen- 91 containing organic compounds and reported the following 34 trends: 1. Ti+, TiCl+ eliminate nH2 with olefins (n 3 l). 2. TiClE, TiClg eliminate HC1 with olefins containing a carbon chain of g 5 carbon atoms. 3. TiC1+, TiCl; eliminate small olefins from olefins containing a carbon chain of 3 6 carbon atoms. Scheme IV shows these processes. Scheme IV. -2H2 Ti+ + ‘1[:f\-———9 Ti+---:1[:\1 >1111:::> \* ‘2” L + 2 +/ TiCl 1- -———> C1Ti --- ClTi 1 , \ NV \Jv ('5 —l —_{ u—lo Q . + T1Cl3 + . + T1C13 + //C C H H @/' O // Cl ’I L C\\ CH CH ‘71i* H //// H 2 1 2 Cl‘ 1" \\C -———5 Ti+ + c H H . H/ \C/ / 4 8 H CH3 C1 C1 35 Scheme IV (cont'd) H @ @ H \x // C———-C -~_ H/i@ @\\\\ //H 4//I:T\ ‘“1i+ H C1 | C1 // H C1 H H \\c<::;/’ / CH3 For oxygen-containing organic compounds, the following trends were reported: 1. Ti+, TiCl+ deoxygenate aldehydes and ketones. 2. 1101*, TiCl; eliminate HC1 with small aldehydes and ketones. 3. TiC12+, TiCl; eliminate small olefins from aldehydes and ketones containing 3 4-carbon chains. Scheme V shows reactions representative of these processes: Scheme V. + 0 .+ Ti + /fl\¢,\\ —————+ T1 0 + CSH10 0 + Tic1+ + ,JLV,,\‘-—————a TiClO + C5H10 36 Scheme V (cont'd) / \ c1 C1 \+ / +- " //Ti..Q + 0\ TiCl + RCH CR' -———> C1 ——-> C1 1i——,=C- R + HC1 3 2 5M 2 1 1' 61% \q/ \R' RC/ \ H 1 LR Cl C1 C1 . + ” l.+ 1 + I + T1C13 + RCHZCHZCR'-—9 CI-Tl -C1-—-+ CI-Tl -C1-—-fi CI-Tl -C1 0"] H orH 06/"+ /H\+ R RI/_‘|_\J\R R'+/\R H H (In ————> C1-Ti+-Cl + /\R 0 AK R' H Note that HC1 and Olefin eliminations proceed through a 6-membered ring intermediate with an a-hydrogen needed for the former process and a B-hydrogen for the latter. Ti+ reactions with alkanes have also been studied92’93. —-—->1i+/1L + H2 A _. e.g. 37 .+~\\ + -———>C1T1.u>/ + C2H6 + H2 TiCl + /~\/“\,/ ————- . + .L__9 T1C1C6H10 + 2H2 + .t.v>\ T1 +/\/ ————————9 T1; + 2H2 In contrast, transition metal ion chemistries with alkanesga'96 and alkenes97 by ion beam studies were reported by J.L. Beauchamp as shown in Scheme VI and VII. Scheme VI. H C0$:~$’:rt\H)\+ /CO +)1\——H—-—) Co +—)l\ \\N ‘:§\u\\c W+-/J______9 C°+'—/l1 CH 3-Co+ 1FNCH3 SN +/ + Co\\ ————9 Co C2H4 + C2H4 H CO + \\\\S H‘\\ + / 38 Scheme VII. C3H6 + CoC2H4+‘e—————r— ”.cot_lL + C2H4 + C0C3H6 €—————‘ Hence, Co+ can insert into either a C-H bond or a C-C bond followed by a B-H, B-methyl, allylic H, or allylic alkyl shift. Note that l-butene can be isomerized to cis- 2-butene and vice versa. Copper, on the other hand, does not insert into carbon- polar bonds in alkyl halidesgs, ketones and esters99 (similar to Li+ and Na+85) as is shown in Schemes VIII and IX. 39 Scheme VIII. Cu: T + C‘ H in“ H 9‘ cu + H ——» ———+ w k / \J / / \ W" + i + ---, + \__/ \_=_/ CuCl >7< HCl Cu + /_\ HCl +/_\ Halide transfer dehydrochlorination In order to determine if the two groups in an organic molecule behave as one "new group", a study of cobalt with bifunctional organic molecules is being performed in our 100 laboratory Some important results are summarized below: l. In the case of adjacent functional groups, it can- not be assumed that products indicative of each group will be observed in reactions with gaseous metal ions. 2. All of the metal-containing ions derived from Co(CO)3NO exhibit a rich chemistry with bifunctional mole- cules. Presumably, in cases where two groups can bond to the metal, much more energy is released in the intermediate complex than in the casetrfmonofunctional molecules. 3. In the case of allyl amine, we do see strong evi- -ence for insertion of Co+ into the C-N bond in contrast with regular amine586, presumably driven by the strong inter- action with allyl group. 40 Scheme IX. acnzccnzcnzn' 91+ + ('m“ ‘\ -—-—-% OH H. /'OH 31w )cg-jcnz“—i ll; 53. RCHZ/ \cncnzav R012 §CH RCH/1\cacnz :‘Tgcno flui- Cu, I on cmz / \ ROI-I 'v CHCHZ \HJ 0H / x... 1 .. ... (hf-- -|| Cum-H2 CHR (:32 Rallzfion' in“ E\ + ‘——‘—-> u c: R’ c“ E! __ _ ____9 _ nou’ \oa' g 3 g H R/ \H l R/ \H cJ—on' 41 H Co+ + 7/i\\v/’ + //z ‘——5 Co -—-NH2 / \"ii Fe(CH20)(C2H60) Fe(C H o) —— 2 2 .863 + --# Fe(CH30)(CzH60) Fe+ + \\J/O\\,/~—T9§;+ Fe(C2H4)+ + C2H60 .24, + r Fe(C2H4O) + C2H6 .41 + ‘-—+ Fe(C2H5OH) + C2H4 eiflie Fe(C4H80)+ + H 2 + Fe + o o -——9 Fe(C H 0)+ + c H o CH§/’\ A R ?\\CH3 3 6 5 12 3 Basically, all these reaction products can be explained in terms of Scheme X, i.e. Fe+ inserts into a C—O bond, 56 followed by a B-H shift. Scheme X. 0 0 + 0 + + / \ B-H Fe “—11 Fe + CJ/ \\CHI_—9.CH’/’Fe CH2 shift CH/'1 CH2 3 3 3 // 3H H8 + Fe(CHZO) + CH4 S1m11ar1Y9 CH3 CH \ I’O‘x h/’ 2\\ Fe+ + .~,o\v,. ——+ 6: Fe CH2 H1 H1 H2 2 58% H 13 o + shifts H Fe -—C2H5 i” ' + + H Fe (CZHGO) e———-CH3—CH2—o—Fe --H CH3 + i H C2H4 F +(C H ) + c H OH F +(C H 0) + C H e 2 4 2 5 e 2 4 2 6 o o o ,/o Fe+ + CH/ \_/\__/\__f\CH —-9 CH \Ctk 3 3 3 /, CH H1 2 /,o q\ 2 Fet—O-TH-CHZ \—J CH3 H shifts H1 H2 shifts _ CH EET\ fl 3\\0 CH 06 H + CH —o 5 ‘“ | ~. Ho 0 o 3 2 5 3 - a \\ o\ //CH CH/,Fe + \_j CH CH ll 3 2 CH 57 The successive reaction products, Fe(CH20)(C2H60)+ and Fe(CH30)(C2H6O)+ in the reaction of Fe+ with dimethyl ether can be explained as follows: CH CH'//¢ 3 3 CH3 1 CH 1 - Fe 3 \\0 + Fe -———9 \\0--~ + -——————+ / 1‘7 / I CH3 CH3 OCHZH o H \\ ./’ 8 CH2 CH 3 \‘o-~Fe+ + CH // 2 4 CH3 3 CH2: CH3 CH3/7 CH I 3 + \x + 0 + Fe ——_79 O ------ Fe + CH3. CH’// 1 //' 1 3 OCHZH CH3 OCH3 That is, the real structure of Fe(C2H60)+ is that in which Fe+ has inserted into the C-0 bond. When another molecule approaches, a B-H shifts to eliminate CH4. Alter- natively, the incoming dimethyl ether can undergo substitu- tion to replace the CH3 group. Note that the Fe+ reaction with TDE only produces one product ion, FeC3H60+. When the reactant neutral becomes a polyether, multiple metal-ligand interactions are expected, since these are observed in solution. While the number of 58 atoms in TDE with which the metal ion initially interacts may be > 1, the actual reactions occur involving only one site on the polyether. The initial multiple interaction may be important in directing the metal to a site of attack (insertion). The mechanism proposed above is typical in that, when the metal ion rearranges TDE into two smaller molecules, the smaller one is usually retained as a ligand, presumably being due to its effective electron donating ability of the smaller ligands. b. Reactions of FeCO+ FeCO+ reactions with dimethylether, dimethylether and TDE are summarized below: .137 + 4 Fe(C H 0 CH 0 + 0 + 1 2 6 )( 2 ) FeCO + /\ ———>Fe(CHO) —1 CH CH 2 5 863 + 3 3 '————-> Fe(C2H60)(CH30) Feco+ +vov "0; Fe(C2H40)+ + C2H6 + co (2) 43-8-4 Fe(C H 0H)+ + c H + co (3) 2 5 2 4 ..LUL; F + Leg—9 Fe(C4H]o)+ + C0 (5) + 91 + 0 0 0 -L——+ FeC H 0 + C H 0 + C0 5 12 3 CH/3\__/\_j\__/O\CH3 7 3 6 (7) . .09 + L——4 FeCOC4H802 + C4Hmo2 (8) FeCO+ + 59 The processes in (1)-(4) were observed for Fe+ alone and are presumed to occur by the same mechanism, however here we have concurrent cleavage of the M-CO bond. Reac- tion (5) is a commonly observed ligand substitution. Reac- tion (6) is similar to (2) where an alkoxy group replaces an alkyl group. FeC3H60+ in reaction (7) was also formed by Fe+ alone as explained in the last section. In reaction (8), since (a) C4H802 is not "extracted" from TDE by Fe+ alone, and (b) the C0 is retained, we interpret this as a MCO+ insertion into the center of the molecule, followed by a B-H shift. 0 FeCO+ + TDE ——-> I 1 + l—\ H Fe-——CH20 OCH3 .1 I C gFe + CH30 OCH3 l " \_/ CH30 pCH Note that, without incorporation of C0 in the inter- mediate, no B-H atoms would be available to shift. The pro- duct resembles a metal-ketene complex, which is frequently observed in such studie591. c. Reactions of other Fe(C0)x+ ions As indicated in the reaction product lists, as the num- ber of CO's present on the metal increases, ligand substitu- tion becomes the predominant process. Fe(C0)2+ and Fe(C0)3+ 60 will react with dimethylether to form FeCO(C2H60)+ with loss of up to 2 CO's. Diethylether can displace up to 3 CO's and so on, which is typical of other systems studied85- in fact most compounds which have been studied except alkyl 92’122 have been observed to displace CO from fluorides charged transition metal centers. Successive reactions also become important at higher pressures, for example: Fe(CO): +.v,o\,,——+ Fe(C4H100)+ +\V,0\//——9'Fe(C4H10); + xCO x = 1,2,3 + + + Fe(C4H100)2 + yCO y = 2,3,4 However, in these studies, no ion of composition Fe(TDE)+ was observed. The reactions of Fe(CO): with TDE are summarized below: Fe(co): q__—4 Fec3H802 +C5H1002 + xCO x=2,3 (9) ___—4 FeCOC3H80 + C5H1003 + (x ‘)C0 x=2,3 (10) .4 FeC0C3H502 + C5H1o°2 * "2 * (x"’C° x=4 (11) .___3 Fe(C0)2C 4 H1002+ + C4H8°2 + (x'2)CO x=2’3’1‘12) or Fe(C5H + C4 H 02 + (x-l)CO 1o 03) 8 61 All of the products for TDE are reactive rearrange- ments of the TDE; none are C0 displacement. This may reflect the ability of TDE to complex stepwise with the metal center i.e.- initial complexation may involve one oxygen of the ligand. As further oxygens interact, addi- tional energy is made available for metal-induced decomposi- tion of the polyether. Hence reactions (9), (10) and (11) can be explained H/U%O as follows: -CO Fe(CO)+ + TDE C + C0 2 :\)/00\\_JC H2 . shifts H3 shifts i“ Q /0— W” + 0 3 \ .fl\,fl\\ // x + Fe+ CH3 C Ee—— CO I 3 C C=t0 0 / H -C0 + 62 C0 .,C/0000>_/\__/\__/\H Fe(C0)2+ + TDE —;£9—+C H// H3 + C0 OL/F elk/H CO CH Ko-—Fe"—Co 4—— CH/0 Fle+ + o C2H5/ + 3 A/1 UL?\CH F_9£Q£3fl89_ H Note that the reactivity (which will be defined here as the number of products) increases as the neutral changes from an ether to a polyether. Also note that the CO ligands can be lost in a stepwise manner. The formation of Fe(CO)2C4H1002+ is very difficult to explain, however, the following mechanisms are possible. 0C C0 \..4 I \ ’ \ CH 3\0 Fe(CO)2+ + TDE -———9 {E::U :::;H3 -—————9 \F + 0 /°\/ \/°\CH Coj CH3 3 63 01" CH CH 3 3 \. //’ 0 + 1 0 Fe(CO)2 + TDE ———> + E j o o 0 \L_; 0 u eC-—4< H B-H shift + Fe(C5H1003) + C4H8O2 2. Cyclic Ethers a. Fe+ reactions with THF and THP. Fe+ reactions with tetrahydrofuran and tetrahydropyran are summarized below: Fe+ + THF '6Qe Fe(C3H6)+ + CHZO "9e Fe(C4H6)+ + H20 __;j!L, Fe(C4H80)+ Fe+ + THP —Hr—Ll§» Fe(C4H6)+ + CH4O -—4§3—> Fe(C4H8)+ + CHZO r—-L;§9 Fe(C3H60)+ + C2H4 ‘05> Fe(CusO)+ Again, all the products can be explained in terms of a metal insertion followed by a B-shift. The reaction mech- anism of Fe+ with THP is shown as an example in Scheme XI. (The reaction mechanisms for Fe+ with THF is shown in Scheme II) Scheme XI. Fe Fe + 0 -—-—-> o‘tfl—eo/ > + .2.“ Note that in the THF reactions, the formation of an intermediate containing C3H5 is a strong driving force. In THP, there is no favorable mechanism for forming this inter- mediate. Fe+ does react with THP to form Fe(C3H60)+, which is reasonable, being geometrically accessible (5-membered ring) and that Fe+ interacts with both oxygen and carbon equally well. (Table 2 p.20 ). b. Fe(CO): reactions with THF and THP Usually, the more ligands on the metal center, the less reactive these species will be except in the reaction with multifunctional organic compounds as explained in the last section. Hence in the reactions of Fe(CO): with THF and THP, 65 we only see the same products as Fe+ formed (by the same mechanism presumably) with concurrent cleavage of the M+-C0 bond. Ligand substitution and successive reactions also become important such as: Fe(CO): + THF —————$:Fe(CO):_a(THF) + aCO for x = 2 a = l x = 3 a = 2 x = 4 a = 2,1 x = 5 a = 2,1 Fe(CO); + THF —T———>»Fe(1HF)+ + co -—-—+ FeC0(THF)+ + C0 Fe(THF)+ + THF .5 Fe(THF); + -———> Fe(CH20)(THF) + 63116 + ~———9'Fe(C3H6)(THF) + CHZO 3. Cyclic Polyethers a. Fe+ reactions with 1,3 dioxolan, l,3 dioxane, p-dioxane-(dglé lZ-crown-4 and lS-crown-S Fe+ reactions with 1,3 dioxolan, 1,3 dioxane, p-dio- xane (d8), lZ-crown-4 and lS-crown-S are summarized below and in Table 5. 66 + Fe (C2 Fe+(CH + Fe (C2 Fe+(CH Fe+(C2 Fe+(C2 Fe+(C2 Fe+(CH Fe+(C2 4. Fe (C3 .1. Fe (C2 Fe+(CH H4) 20) H40) 202) H202) H402) H4) 20) H202) H60) H402) 2°) Fe+(C3H60) Fe+(C2 H402) Fe+(C2H4) Fe+(CD .1. Fe (C2 Fe+(C3 20) D4) 060) + + + + + + + + + + + + + + + + + + + Fe (C20402) + CHZO C2H4 CH CZH4O2 C3H60 c2H6 CH 0 CZH4 C3H60 CHZO C2H4 C H 0 2 4 2 C D O 3 6 C20402 C020 C2D4 (l3) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) 67 In its reactions iwth monofunctional ethers, both acyclic and cyclic ethers, it is energetically favorable for Fe+ to insert into the C-0 bond, followed by a B-H shift. If Fe+ interacts with two geometrically accessible oxygens which are in close proximity, such an interaction may involve more than twice the energy as the interaction 124 with only one oxygen due to the ligand effect This may result in a double insertion and double B-shift mechanism. This process can be stepwise145 , although it is convenient to use a concerted double insertion, followed by double B-H shifts (using a stepwise mechanism, it is hard to explain reactions (17) and (18)). Accordingly, the products in reactions (13)-(18) can be explained as follows: ,Fef H1 /o Fe+ + 0A0——> GOO ““19 . J>Fe+ H2 insert1on7 L—j Lj y l\0 H2 + + C2H4 + Fe(CHZOZ) H—Fe + Z) + Fe, + 0 x . ./ ./ ‘~ double , 4+ Fe + q/Ajp'—_9 (k‘\\,/*’P insertion" <:;:;EE\\J\\ o CHZO + Fe+(C2H4O) <——1— ...—pet---" 1 CH2 ——--0 + C2H40 + Fe (CHZO) e———I Table 5 Fe+ +[: . Ion/Molecule Reactions of Fe+ 68 And lS-crown-S. r—\ o 0:] QL_P .22 .13 .41, .02 + Fe C2 2 .1. Fe CZHZO Fe+C H o 2 4 + Fe C4H40 + C 2 2 + C6H14°2 + C H 6 1202 H o + 4 10 3 + Fe C4H60 + C4H1003 + Fe C4H80 + Fe C4H100+C4 + Fe C3H60 2 + c4H8°2 2 H603 + C5H 0 10 2 + Fe C4H803 + C4H80 + Fe C6H 1402 _. Fe+(C2H202) + C + C2H202 8“160 3 + + Fe (C2H6O) + C8H1404 + . Fe (C4H802) + C6H1203 + + . Fe (C4H803) + C6H1202 Fe+(C8H]403) + C2H60 3 2 H H 20 2 +H2 With lZ-crown-4 See Scheme XII XIII XIII XIV XIV XII XIV XIV XIII XV XV XV, XVI XVII XVIII XVIII XV 69 + /\‘ x’ = double ‘ Fe + QL_P 1 OrvL/X’P insertion ’ H1 \—0\ + + / 1 shift H 0 ’Azx' + + ”,Te———CH2-—é Fe(CZHZOZ) + CH4 \oH Note that there are only three possibilities for double insertion, each case gives the corresponding products. Reactions (23)-(26) were identified unambiguously by p-dioxane d8 reactions. The product ion in reaction (21) can be either Fe(C2H202)+ or Fe(C3H60)+. Fe(C3H60)+ is more reasonable based on the p-dioxane reactions (23)-(26) and can be easily explained by the proposed mechanism which follows. Note that the chair form of p-dioxane is the most stable conformation based on electron diffraction, sector 125,126 127-129 microphotometer studies , and complexation studies. It is shown thatZLCOC of p-dioxane is 112.45° 126. There- which is bigger than that of cyclohexane (108°) fore we assume that it is the most stable conformer in the gas phase. From X-ray studies of 2-chlrophenyl-l,3 70 130 131,132 dioxane and conformational analysis and modeling studiesl3], it is agreed that l,3 dioxane is more stable in the chair form with theL01C3O3 = 111°, which is puckered (dihedral angle, T = 63°) and LC C C = 108°, which is 4 5 6 flattened (dihedral angle, 1 = 55°), although the under- derivatized ligand has not been studied. Since Fe+ has al- most the same bonding energy to oxygen and to an alkyl group (Table 2) and because of the geometrical accessibili- ty, Fe+ is capable of interacting with both oxygens and one carbon as indicated by dashed lines in the following mechan- isms. However, it appears that Fe+ is even more likely to interact with one oxygen and-one carbon instead of two oxygens due to the geometrical restriction. The products in reactions (l9)-(26) can then be explained by the "double insertion, double B-H shift" mechanism, as shown below: ' + /,Fe-“\ d 1 O\\ fi u “0 oub e 5 L .../W W - If, 7 F6 -———9 -----Fe* insertion &£::ji::::;z \\__o ‘///EEZ//‘1‘\..O Fe+(C3H60) Fe+(CH20) ,Fe . 0 ,’ ‘ / 0 I ‘~ double A + --u ,+/’ Fe+(C2H4) Fe+(C2H402) 71 Fei ’x' “‘.0 /0 + N ——-x—+ Fe’ ——9 Fe (C2H40) (27) , 0 Similarly, ,oFei. o o 3W“ ...... . \..v H ..... ..4/ ZL/C“;-;7( insertion’ 4\\_:>7__9 CH + Fe(CHZO) Fe(C3H60) Fe+ /g\ M0 30. {01 E Fe“ 4 6 5 Ee+ ,'1/ 0 0 . 0 FTFTT“/// gaggliion$ £::¢::Fe+ ———:fl:;Fe:::o:> + + (28) ———)(-———> Oe<:> ——>)|;€+<:> (29) + + Fe(CHZOZ) Fe(C3H6) 72 Note that the only reason why reaction (27) doesn't occur is because Fe? can't interact with two oxygens and one -CH2- group at the same time. It is difficult for Fe+ to interact with two separated oxygens by breaking bonds as shown in (27). Instead, it prefers to interact with one oxygen and one methylene group. In the case of l,3-dioxane, Fe+ can either interact with two oxygens or one oxygen and one -CH2- group equally well. Since the 03-C4-C5 angle is flattened130 , the -CH2- at C5 can have a weak interaction with Fe+ unless Fe+ moves closer and weakens its interaction with another oxygen as in reaction (28). The reason why reaction (29) doesn't occur is probably because the 01-C2 and C2-03 bonds become weaker than other bonds, due to the strong interaction of Fe+ with two oxygens, which forms a strained four membered ring intermediate and results in an insufficient orbital overlap of 01-C2 and C2-03 bonds, in contrast with more flexibility of 01C6C5C403 ring on the other side. Therefore, 01-C2 and 02-03 bonds are easier to break and C6-01 and 03-C4 are more resistive to cleavage. Also note that all of the products in the reactions of Fe+ with 1,3 dioxane and 1,4-dioxane are from the metal ion "double insertion" only, because B-H's may not be geometric- ally accessible for shifting after double insertion by the Fe+, which might have a linear structure for the inter- mediate, RCH2 -Fe+ -0R. Linear structures for insertion products have been suggested by D.P. Ridge in the reactions 73 of Fe+ with alkane5194. In 12-crown-4, the chemistry is even richer. The reac- tion products together with their reaction mechanisms are indicated in Table 5. In some cases, Fe+ appears to break down the polyether into smaller cyclic ethers, which is also observed on elec- tron impact of crown ether5133. One product, FeC4H802+, appears to be the result of cleavage of 12-crown-4 into two 1,4-dioxane molecules which is shown in Scheme XII. The process is favored since 2 molecules of 1,4-dioxane are approximately 20 kcal/mole more stable than 1 molecule of 12-crown-4134. This arrangement of 12-crown-4 is consistent with a distinct secondary structure for the polyether which is shown in Scheme XII. This structure was derived from X-ray crystallographic studie5136. Thus, we assume that it may also be a favored configuration for the uncomplexed crown in the gas phase. (While the molecule is free to assume many secondary structures, intramolecular hydrogen bonding may 107’111.) Using the numbering favor this configuration system of Scheme XII, it can be seen that this chair-type structure has two oxygen atoms whose lone pairs are directed above the "plane" of the molecule (1 and 2) and two below (3 and 4). If Fe+ can complex with this ligand without extensive changes in the secondary structure, initial complexation with lone pairs on oxygen atom #1 and #2 will 74 +o~=~o£ «I..— floxJo ... +0¢...uu..._ 07x “ u ._ ...; «cameo + woozeu 6.. «......mo «We Imzo . “fl \\\ d a O . . O \UNI JUN—... o s .. .. .. 1 \on C . C x v «solo d «zolo... .’ V ’ n N are“ «Ifilmro; «Ivan «IO-I 10%.. 0:30. .0”: o, a .1111. Notwo ifw wax ... on / u z N / «as. .. OlNIU‘ mull. O +£\ “ SM .mzmxum 75 direct the metal towards C #5. These three atoms are in close proximity in this model. Scheme XII shows that such an initial interaction would lead to the FeC4H802+ product if Fe+ can bring 0#1 and C#5 in close proximity to form two rings (two p-dioxanes). On the other hand, if this inter- action brings C#5 and 0#2 in closer proximity, 9-crown-3 can be eliminated, leaving FeC2H4O+ which rapidly loses H2 to form FeC2H20+, which is also observed. Thus, the geometry of "free" 12-crown-4 suggests that, regardless of which side of the ring Fe+ attacks, there will be two 0 atoms and one CH2 unit (which is across the ring) in close proximity for initial interaction with the metal, similar to the cases of l,3-dioxane and 1,4-dioxane which have been discussed earlier. The geometry may facilitate the formation of new C-O bonds and smaller rings. Other products can be explained following initial metal complexation to the two oxygen atoms shown in Scheme XIII and XIV followed by metal ion double insertion and double B-H shifts as previously discussed. There are only two possibilities. Fe+ can interact with either two oxygen atoms next to each other or two oxygen atoms across the ring. Scheme XIII shows the former case to form FeC2H402+, FeCZHZOZ+ (similar to a metal-butadiene complex) and FeC H 6 14 with oxygen atoms #1 and #2. Following the double insertion, 02+. Note that it also follows initial interaction we see a situation where both the "left" and "right" groups 76 o. o. . “on. 1A1 m wan. Q + . , o. + I..\ L » \\ . 1 WI 01/. on... AAIIIIII fl om... .... o.\ o l\_._. «I; aim IaQ 25:00 0) AU. e9 1.... flow... 3.113.... “W ”EM MEMIUm 77 SCHEME XE b0 +H (:fFe 33—»OHQ: :Fe DOUBLE SHIFTS H (‘9? c4 limo/— \_ HHO . /‘°\ + H0 -—> O ,Fe+ o K,o m/z l60 H603 m/z I30 78 contain B-H atoms which could shift. Apparently, only those on the "left" do. Molecular models suggest that further interaction of the oxygens on the "right" group with the metal ion may move the "right" B-H atoms to positions spatially less accessible. These additional metal-oxygen interactions could thus, through an intermediate such as is shown in Fig. 8, control the availability of 8-H atoms for rearrangement. Scheme XIV predicts the remainder of the products through another double insertion, double B-H shift scheme, following an initial complexation of Fe+ with two oxygen atoms across ring. Note that if H's from the "left" shift, FeC4H100+ is formed, which corresponds to an ion of m/e 130. An alternative structure (FeC3H602+) for the same ion can be explained by a double insertion: flfi —-——)<—O\Fe /“\0 Fe+ +.4E§L—PH 0_—J/Z:/0\3> Here, Fe+ interacts preferentially with oxygen atom #l + -——9 Fe(C3H602) +C5H1002 and carbon atom #5 as indicated in Scheme XII and inserts into both the C-6 and the C-0 bonds similar to reaction (28). Nevertheless, CID or high resolution spectra are required to unambiguously distinguish FeC4H]OO+ from FeC3H602+ . In the reaction of Fe+ with lS-crown-S, one may expect to get more products, presumably due to increased interaction 79 + 4» Figure 8. A Structtn'e Of Fe-lZ-crown-JJ' After Fe Double Inserts Into 'IVto C-O Bonds 01‘ 1242:011th 80 of Fe+ with five oxygen atoms in lS-crown-S. However, Fe+ apparently can't react effectively with all 5 oxygens. (lS-crown-S has much bigger cavity than 12-crown-4 does (1.7 - 2.2 A vs, 1.2 - 1.5 A in diameter)). Although there are many reports on the complexed struc- 137 and also its free 1igand138, there 112,139,140 tures of 18-crown-6 are only a few reports on complexed lS-crown-S and the study of the free lS-crown-S molecule has not been reported yet. However, the structure of lS-crown-S was taken by X—ray and drawn in reference 139, in which the copper in CuBr2 interacts with only one oxygen atom of lS-crown-S, which makes the conformation of 15-crown-5 similar to that of a free ligand as shown in Fig. 9. Figure 9 X-Ray Structure Of lS-Crown-S In lS-Crown-S CuBrz. 81 Scheme XV Hb a 0 Pet 0 b a (\O */ Q ~ 3 ——+ . x 0d 0 (‘5: :0] ZA-H shift + Fe(CzHSOH) + [k0 :3) o 82 Scheme XVI l 3,1) J b a 0 < fig/j ——>Fe(C482+HO) 0\\/o/ \\/o 605 Scheme XVI I I (>5 6 [mi <1 . .403 I :3; . 3.; ’ + + bx“ 83 Hence, oxygen atom #4, #7 and #13 are up and #l, and 10 are down. Fe+ then can interact with [0#4 and #7], [0#4 and 0#13], or [0#7 and 0#13]. Since carbon atom #11 is with the metal up also, the interaction of both [0#13 and C#11] and [0#4 and C#ll] are also possible. Schemes XV, XVI, XVII, and XVIII are the proposed mechanisms leading to the observ- ed reaction products of Fe+ with 15-crown—5. b. FeCO reactions with 1,3 dioxolang, l,3-dioxane, p-dioxane, lZ-crown-4 and lS-Crown-S. Most products of FeCO+ reactions with l,3-dioxolan, l,3-dioxane and p-dioxane are seen in the Fe+ case with concurrent loss of CO except as shown in the following: /“\ Feco+ + °\__,° ———> Fe+(C3H602) + CO L————+ Fe(CH20)(C3H602)+ 4————--—> Fe(C2H4O)(C3H602)+ ‘1’ 4. Fe CO(C3H60) + CHZO FeC0+ + Q + 0 + FeCO + Ej—a Fe (C4H802) + co 0 . + , Fe (C4H802) + C0 + -—————9 Fe (CH20)(C4H802) (30) (31) (32) (33) (34) (36) (37) (38) 84 Reactions (30% (34) and (36) are typical of ligand substitution processes. In the l,3-dioxolane, l,3-dioxane, p-dioxane systems, the addition complexes are reactive enough to undergo successive reactions (31) (32) (35) (37) and (38). Note that the reaction products in (31) and (35) are also formed from Fe(CHZO)+, which is reactive too. Presumably, CHZO in reaction (33) carries away most of the reaction energy, leaving CO retained by Fe+. FeCO+ forms seven products with 12-crown-4; six of them were also formed by Fe+ alone. Thus, MCO+ insertion does not appear to predominate in the reactions with the cyclic polyether. The new product for FeCO+ with 12-crown-4 is m/e 204; FeC6H1204+, corresponding to loss of CO from the ionic reactant and C2H4 from the crown, presumably by a mechanism shown on Scheme XIX. On the other hand, FeCO+ forms nine products with lS-crown-S, which are shown below. FeCO+ + 1505 4411—» Fe+(CZH4O) + C8H1604 + c0 (39) + .14 1 Fe (C2H202) + C8H1803 + co 14 + .20 + _JEL_. Fe+(C H o ) + C H o + CO 4 8 2 6 12 3 .08 + _;___s F°+IC6H12°4) + C4H80 + CO (40) 85 Scheme XIX FeOO' + 12-crown-4 A °‘1°° .Fef’ j (,0 T (A E j o o [0 o) + 02H4+GO Scheme XX 86 + .04 ‘ + FeCO + 15C5 7 Fe (C8H1403) + CZHGOZ + C0 .06 + + Of which, six products were also formed by Fe+ alone. Again, MCO+ insertion does not seem to play a role in this larger cyclic polyether. The new product Fe(C2H40)+ in reaction (39) is presum- ably formed via the mechanism shown in Scheme XV, following double insertion as indicated by the pathway b. The C0 ligand plays an important role here. Apparently, available energy can be used to break a M+-CO bond instead of inducing B—H shifts. The product ion Fe(C6H1204)+ in reaction (40) is formed through interaction with oxygen atoms #7 and #13 as shown in Scheme XX. Note that we do not see this ion in the reaction with Fe+. Since oxygen atoms #7 and #13 are far apart, the CO ligand here may play an important role in bringing them closer as shown below: FEW .“ j/ ’:0 ‘Fe’ :) 0 0 \__/ 87 Once Fe+ double inserts into C-O bonds, it may further interact with oxygen atoms #1 and 4 to eliminate CO and C4H80. The retention of CO on the product ion in reaction (41) clearly indicates that no further stabilization energy is produced (product distribution ratio for this ion is 0.06 and only 0.05 in the reaction with Fe+). c. Fe(CO): reactions with l,3-dioxolan, l,3-dioxane,_p-dioxane, 12-Crown-4 and lS-crown-S. Higher CO-containing iron ion molecule reactions with 1,3-dioxolan, l,3-dioxane, p-dioxane and 12-crown-4 only undergo substitutions and successive reactions. Note that in the reactions with 12-crown-4, a new ion, FeC0C2H202+ is formed at m/e 142, presumably by the same mechanism as is shown in Scheme XIII. In the reactions of Fe(CO): with lS-crown-S, there are few new ions formed: 4 + + Fe(CO)2 + lS-crown-5-———+ Fe(C2H602) + C8Hl403 + 2C0 (42) + + Fe(CO)X + lS-crown-5-———+ FeCO(C2H602) + c8Hl403 + (x-l) CO x = 2,3,4 (43) + Fe(CO)2 + lS-crown-S ———9 Fe(C6H1204 + C4H80 + 2C0 (44) + + Fe(CO)2 + lS-crown-5 ———+ Fe(c8H1605) + C2H4 + C0 (45) + Fe(CO)3+ lS-crown-S ———9 Fe(CBH1605)+ + C2H4 + 200 (46) 88 The product ion, Fe(C6H1204)+ in reaction (44) can be formed by the mechanism shown in Scheme XIX. The loss of 2 CO's might be the result of strong interactions of Fe+ with oxygen atoms in lS-crown-5. Fe(C2H602)+ and FeCO(C2H602)+ in reactions (42) and (43) are formed by the mechanism shown in Scheme XV. Instead of forming the Fe(C8H1403)+, which is formed in the Fe+ and FeCO+ reactions, these two products are formed by retaining the smaller ligand (C2H602) with a concurrent loss of 1,2 or 3 CO's indicating the preference of iron ion to retain the smaller ligands (as noted by the product distribution ratio, .2 for the reaction (42) and only .06 and .05 for the formation of Fe(c8H1403)+ and FeC0(C8H1403)+ respectively). Finally, a similar mechanism to that shown in Scheme XIX can be used to explain the forma- tion of Fe(C8H1605)+ in reactions (45) and (46). Note that as in the 12-crown-4 case, C0 ligands appear to be important in forming this big metallocyclic product, since the energy which can be used for B-H shifts is instead used to break M+-CO bonds. 4. Thermodynamic Conclusions Table~ 6 lists all neutrals lost in the Fe(C0)x+ reactions with linear ethers, polyether, cyclic ethers and polyethers which we have just discussed in last sections and branching ratios. (Product distributions) Fe+ induces the rearrangement of dimethylether into CHZO and CH4, but FeCO+ does not. This can be readily 89 understood in terms of thermodynamics: (Dimethylether will displace one, but not two CO's from Fe+) Assumption Feco+ + CH OCH + 3 3-—————+ Fe (CZHGO) CO AH < O Y‘Xl'l + + Fe (C0)2 + CH30CH3 —*—— Fe (C2H60) + 2C0 Aern > O This implies that the initial Fe+-dimethylether inter- action energy is 27.17 kcal/mole < D(Fe+-— MeZO) < 73.17 kcal/mole. Note that there are two ways to interPret these 'ts results. (All AHf's are taken from ref. 142 and all un1 are in kcal/mole). A. In terms of bondstrengths Feco+-—————— Fe+ + co .°. AH = 27 kcal/mole 282 -26.42 + + . Fe(CO)2 —————— Fe + CO . . AH = 73 kcal/mole 282 -52.83 .'. 27 kcal/mole < D(Fe+-Me20) < 73 kcal/mole 90 B. In terms of heats of formation FeCo+ + CH OCH -———9 Fe(C2H60)+ + C0, 3 3 228 -43.99 -26.42 . + . AHf(Fe -Me20) < 2l0 Fe(CO): + CH OCH3 ———+ Fe(C2H60)+ + 2C0, 3 156 -43.99 -52.83 . + But, + + + Fe(CZHGO) ‘———9 Fe + C2H60 Aern = D(Fe ~Me20) D(Fe+-Me 0) = AH (Fe+) + AH (Me 0) - 2 f f 2 AHf(Fe+-Me20) = 282 + (-43.99) - [l65 < AHf(Fe+Me20) < 210] °. 28 kcal/mole < D(Fe+-Me20) < 73 kcal/mole If, however, the structure of Fe(CZHGO)+ is really CH -Fe+-0CH3, then 3 + + D(CHB-Fe -OCH3) = AHf(Fe ) + AHf(CH3) + AHf(0CH3) - + _ > 165 109 kcal/mole < D(CH -Fe+-0CH ) < 155 kcal/ 3 3 mole 91 8.: 8? 33.3 m .‘I 8 8“ 8" .2 tn; 8~ ~ fl . E. V RAHNMMIWM amen 8~ EV ax . ....J 8.. Q 8” a. a; was... AM. V 8Tr=koo Emu SN 92 Rive—3:00 8” 8A 8‘ 8— V _ . . 8" V 8~ ASH“ 8~ .noflw. ax 8~ 5 .Vlfifiofmflml 1 a 8N E a. BN V at. .luléil GN. ”NJ Sen. . AS. V 8~ . ....Nu _ V E.“ 8 .. my} 8~ V. E. V 82 frag 8" 5 V m? 8102....“ an a. _ 8.. as}: a. a. a. a. _ a. a. V 42V... 53-8 :23:- Iil 3 o TAB; Bataan—Mn FN V 83 aznu _+ $3 nomenn A8; 8 . Norm. . h $~.V 8 .53.an V 29V 8N . owzau _ :~.V 8 .nooaznu V Go V 8~. SNOMIMUS ARV 8‘ . .95.? 8 a V... u 3: R; V V V 3: 8~ V . 3.: Eu . saw so; Nam—.... Run 8" V 3.. V Bracing E: 89...]. 8V 8. 8. _ 8V _ 23 2.2.3300 _ 31V 8.68.4". 87:3. 8 V :V 8~.no..£mu 2: Sauna 8~ 8 an H 8~ V B WEB... . V 33 No.2”... V V. V _ V 25V 8 . ~%=~u V _ V _ V8; 8 . ......u V V 33 8....amxau , , V. . ~ . u _ . SUV 8 . omzau V V Go V 8 . oéd . _ V E: 8 . ox} V .3 V amass} EV 8 _ V V 1.1V 8 . “05:... So; 8 8~ V 83°..er 3.88. :6 _ V V REV 8.5046 3: 8 . m: 29V 8 7: V 8.3.3.} 3982.0 :13 8 , ARV 8 3: 8 V V EV 8 .62.} ARV 8.93.} 2.9V 8.3.6 ”8.83“: g V Examrmu A$.V8...=~u “2.8.415 , 2~.VR..Vm~u A8.V8.a=..u V . . o :5; woo—nap 81V 8.0.} Ao~.V8.onVo ASV 83°}? ARVBSHE 2.18.3} 2.98 238.38 V 8 8 .82. AS V ~00sz V V 2~ V N022$ is. V :5 V. V _ EV nan”? 2.... V _ Va; .4. EV a mu ......V ....m. 2.3 ....m. .3 Agnes... . d no. u 3.3 “:0 SNV WV... . GNV ”flux". AR. V are a V o=~u 3o. V cwau :3 axm ”at” ....uw . 3V 3 E V o _ a: mu REV 0:1... 3: wean”. Noawe 2 . we r. . n . . u . n . a V 39V ~ V o «3 V cmzu u< :3 ... .31.} 3 5 amino E V 0...». .8 V ed as w: at} a V ~39 E. V :1: V cacao V :3 ox? :3 and EV «a .... - 2) 0 . G ..O V O 44:53 m... E ‘ E.— O _ O /o\ El: 5:. “ABE. co 1.3%.! I: 5 an: 55.... Jljlc . an. 92 Also, the energies required to rearrange/fragment the organic molecule are 0.9 kcal/mole and 82.6 kcal/mole for the following two processes respectively CH30CH3 —77——<> CHZO + CH4 AH 0.9 kcal/mole > CH3 + OCH3 AH 82.6 kcal/mole Obviously, the first process (involves Fe+ insertion into C-O bond, followed by a B-shift) is a low energy pro- cess, which has to pass through an intermediate (insertion product) related to second process, in which the insertion product is formed. Although B-H shifts might not require energy (i.e. naturally occur), the extra energy after form- ing the insertion product will be removed by the Fe+-CO bond breaking (2 27.8 kcal/mole)142 to get no further reac- tion and stop at Fe(CH20)+. Since the initial complexation energy is smaller than the energy required to dissociate dimethylether into CH3 and 0CH3, the real mechanism might look like: [ Fe+ r Fe+ /0 + 5 x’ g ,+ Fe + /0\—) 5 —-> /0“\: ——-) /Fe CH3 ./°\J L CH3 CH3 (II) (III) (1v) 0 + + CH4 + Fe(CHZO) é——-CH4...Fe ...//JL\\ 93 Fe+ reacts with dimethylether to form (IV) with much energy released which is available for further reactions. However this energy will be taken away by a C0 ligand in the FeCO+ reaction. This explains why precursors of (IV) are composed of only 8.6% of Fe+ and 9l.4% of FeCO+ (see Appendix F). Analogously, diemthylether will displace up to three CO's from Fe+ implying that the initial Fe+-diethylether interaction energy is 96.6 kcal/mole < D(Fe+-Et20) < l33.3 kcal/mole. The energies required for Fe+ insertion into a C-O bond in diethylether, and for rearrangement procuts are: C2H50C2H5 —— > C2H5 + OCZH5 AH = 82.3 kcal/mole _———> C2H50H + C2H4 AH = l6.8 kcal/mole The second process is the lowest energy process among these three processes. A possible reason why the formation of Fe(CzHSOH)+ can displace up two CO's from Fe+ and only one CO for forming Fe(C2H4O)+ is that Fe+ may form a stronger bond with CZHSOH than that of C2H40. However, it is not understandable what makes so much difference in complexation energy of Fe+ with dimethylether from that of diethylether. Could it be from the polarizability difference? The rest of the reactions can be explained in a similar way. The data to date can be summarized as follows: 94 l. Whenever B-H atoms are available, they will shift to form stable products, following insertion into a C-O bond. If CO ligands are present, the breaking of M+-CO bonds may compete with this process. 2. A stable addition complex can only be formed with the concurrent loss of one or more CO ligand on the metal ion (to take away the energy). 143" is seen. In 3. A "mechanistic macrocyclic effect the case of lZ-crown-4, the Fe+ actually induces reactions to product 9 products, but only one product is observed in its linear analog (TDE). 4. In the case of l2-crown-4, we assume that the Fe+ actually induces reactions from "inside" the crown cavity. This is implied by the fact that Fe+ reacts with lZ-crown-4 to form 9 different products, FeC0+ gives 7 products, Fe(CO)2+ gives only 2 products, and as more CO's are added to the metal, no reactions are observed. Thus, ligands can prevent the metal from entering the crown cavity and induc- ing reactions‘43. However, this may not be true in the l5-crown-5 case, which has a larger crown cavity. Hence, Fe+ reacts with l5-crown-5 to form 7 products, FeCO+ gives 9 products, Fe(C0)2+ gives 6 products and both Fe(CO)3+ and Fe(C0)4+ give 2 products respectively. 5. Many products of reactions involving cyclic poly- ethers, lZ-crown-4 and l5-crown-5 can be explained using a double metal insertion, double B-H shifts process. In these 95 reactions, CO ligands act predominately as spectators. 6. When CO's are present on the metal center, they can act either as spectator ligands or as active groups (vis MCO+ insertion). An alternate interpretation is that, after M+-C bonds are formed, a CO ligand on the metal may insert into the M+-C bond. 7. The strength of a metal-ligand bond alone does not guarantee complexation in the species studied here. In the case of bulky ligands, the presence of 3C0's on a metal center may prohibit sufficiently close approach for inser- tion. + . II. Cr(C0)x Reactions With Ethers And Polyethers A. Results 1. Linear Ethers and Polyethers a. CrflCO)x reaction with dimethyletherLCzflfigl Ions formed as products of ion-molecue reactions in a mixture of Cr(CO)6 and dimethylether are listed below, with their precursors as identified by double resonance. m[e stoichiometry precursor(s) 98 Cr(C2H60)+ Cr+C0, Cr(C0)2+ + + + 126 CrCO(C2H60) Cr(C0)2 , Cr(C0)3 + + 144 Cr(C2H60)2+ CrC0+, Cr(C0)2 , Cr(C0)3 , Cr(C0)4+ 96 b. erC0)x+ reaction with diethylether L24fll 09.). Ions formed as products of ion-molecule reactions in a mixture of Cr(C0)6 and diethylether are listed below, with their precursors as identified by double resonance. mle stoichiometry precursor(s) + 96 Cr(C2H4O) CrT 11o Cr(C3H60)+ CrCO+ + + + + 126 Cr(C4H100) Cr , CrCO , Cr(CO)2 149 (C H O) H+ OCH T OC H T C H 0+ 4 1O 2 3 ’ 2 5 ’ 4 11 The ion at m/e 149 is (C4H100)2H+, the protonated diethylether's dimer is formed by C4HHO+ reaction with a neutral molecule of diethylether, which in turn is formed + by OCH + and OCZHS . 3 c. Cr(CO): reaction with triethylene glycol dimethylether (TDE, C851894l Ions formed as products of ion-molecule reactions in a mixture Of Cr(CO)6 and TDE are listed below, with their pre- cursors as identified by double resonance. mle stoichiometry precursor(s) 84 Cr(CH4O)+ CrC0+, Cr(CO)2+ + + 91 C4HHO2 C2H30 + + 96 Cr(C2H40) Cr .1.- 98 Cr(C2H60)+ CrCO+, Cr(CO)2+, Cr(CO)3 + + + 101 C5H902 C3H60 or C2H202 97 gig stoichiometry precursor(s) 11o Cr(C3H60)+ CrC0+, Cr(CO)2+ 112 CrCO(CH40)+ CrC0+, Cr(C0)2+ l26 Cr(C3H602)+ Cr+, CrC0+, Cr(CO)2+ 128 Cr(C3H802)+ Cr+, CrCO+, Cr(CO)2+ l38 CrC0C3H60+ Cr(CO)2+ 14o Cr(C0)ZCH4O+ Cr(CO)2+, Cr(CO)3+ 142 Cr(C4H]002)+ CrCO+, Cr(CO)2+ 156 CrCOC3H802+ CrCO+, Cr(CO)2+ l66 Cr(C0)2C3H60+ Cr(C0)2+ 168 CrCOC4H802+ CrCO+, Cr(CO)2+, CrCOC3H60+ 17O CrC0C4H1002+ ’ CrCO+, Cr(C0)2+, CrC0C3H60+ 2. Cyclic Ethers and Polyethers a. CrLCO)x+ reactions with tetrahydrofuran (THF, CQHBQ) Ions formed as products of ion-molecule reactions in a mixture of Cr(CO)6 and THF are listed below, with their precursors as identified by double resonance. ELE stoichiometry precursor(s) 124 Cr(C4H80)+ Cr+, CrC0+, Cr(CO)2+ + + + + 152 CrC0(C4H80) Cr(CO)2 , Cr(C0)3 , Cr(C04 + + + + l96 Cr(C4H80)2 CrCO , Cr(C0)2 , Cr(C0)3 , Cr(CO)4+, Cr(C4H80)+, CrC0(C4H80)+ 98 b. Cr(CO)x+ reactions with tetrahydropyran QTHFz Csfllog) Ions formed as products of ion-molecule reactions in mixture of Cr(C0)6 and THP are listed below, with their precursors as identified by double resonance. gig stoichiometry precursor(s) 11O Cr(c3H60)+ Cr+, CrCO+ 138 Cr(C5H100)+ Cr+, CrCO+, Cr(CO)2+ 166 CrCO(C5HwO)+ Cr(C0)2+, Cr(CO)3+ 224 Cr(CsH100)2+ Cr+, CrCO+, Cr(CO)2+, Cr(C5H100)+ c. Cr(CO)X+ reactions with 1,3 dioxolan 1&3fl5921 Ions formed as products of ion-molecule reactions in a mixture of Cr(C0)6 and l,3-dioxolan are listed below, with their precursors as identified by double resonance. m/e stoichiometry precursor(s) 82 Cr(CH20)+ Cr+ + + + 126 Cr(C3H602) Cr+, CrCO , Cr(C0)2 + + + 154 CrCO(C3H602) Cr(CO)2 , Cr(CO)3 + + + + 156 Cr(C3H602)(CH20) Cr , CrCO , Cr(C0)2 , .... Cr(C3H602) 99 d. Cr(CO)x+ reactions with l,3-dioxane Ions formed as products of ion-molecule reactions in a mixture of Cr(CO)6 and l,3-dioxane are listed below, with their precursors as identified by double resonance. gig 82 110 112 140 144 168 170 196 197 228 01" stoichiometry Cr(CH20)+ + Cr(C3H60) + Cr(CzH402) H O + C7 12 3 CrCO(C4H802)+ + Cr(C4H802)(CH20) + Cr(CzH402)(C3H60) Cr(CO)2(C4H802) Cr(CO)2C4H902)+ + Cr(C4H802)2 precursorLs) Cr+, CrCO+, Cr(CO)2+ + + C3H5 , 50 , C4H7O Cr(CO)2+, Cr(CO)3+ + + + Cr , CrCO , Cr(CzH402) + C2H 2 Cr(C0)4+ + + C3H70 or C2H302 Cr+, CrC0+, Cr(CO)2+, + + Cr(C4H802) . Cr(CO)3 , + + CrCO(C4H802) , Cr(CO)4 Cr(CQ)x+ reactions withgp-dioxane (C4H892) and its d8 isotope Ions formed as products of ion-molecule reactions in a mixture of Cr(CO)6 and l,4-dioxane and p-dioxane-d8 are listed in Table 7, with their precursors as identified by double resonance. 100 +Amomachnu mmomxaoVBuu . NMBVHV v.8 mAuomnaoVno EN . «mommaora . +o~8 £83088 :8 . +A8V8 p.88 mAmowzaoVB 9.8 mA8V8 .WABVHV “moan—8388 92 mABVB .wBVhV mmooxaoVBHV 8H w 8V8 w 8V8 noon. “.8 Mmomnao V8 mi ...88 n8 MmoszVanu 9: +8 +3~8V8 3m .8 ..AomfiVuo we QVuowusooHn 36.83303 “Va WVHomHVoam Faafidsn ME magi 23x31 $6188.68... 82 8888... 5.2V 8368.. m~83 . V. 638 101 f. Cr(CO)X+ reactions with l2-crown-4 55—8-51 694)— Ions formed as products of ion-molecule reactions in a mixture of Cr(C0)6 and lZ-crown-4 are listed below, with their precursors as identified by double resonance. gig stoichiometry precursor(s) 96 CrC2H4O+ Cr+ 112 CrC2H402+ Cr+ 128 CrC2H4O3+ Cr+ l56 CrC4H803+ Cr+, CrCO+, Cr(CO)2+ zoo CrCGH1204+ Cr+, CrCO+, Cr(CO)2+ 228 CrCBH1604+ Cr+, CrCO+, Cr(CO)2+ g. Cr(C0)x+ reactions with l5-crown-5 LC4052095)- Ions formed as products of ion-molecule reactions in a mixture of Cr(CO)6 and lS-crown-S are listed below, with their precursors as identified by double resonance. mle stoichiometry precursorjs) + + 110 Cr(CzH202) CrCO + p + + ll3 Cr(C2H502) CrCO , Cr(CO)2 + + 126 Cr(C4H]OO) Cr + + + 156 Cr(C4H803) Cr , CrCO + + + + 200 Cr(66H]204) Cr , CrCO , Cr(C0)2 102 B. Discussion l. Linear Ethers and Polyether 6. Reactions of Cr+ Cr+ does not react with dimethylether. Products of Cr+ reactions with diethylether and TDE are listed in the following. + cr + ‘~/0\v/_fi~ .64 Légé'Cr+(C4H100) + .66 Cr + ‘//O O O 0\\ Cr (CZH40) + C6H 140 3 (48) 3 15 Cr (CH 80 ) + C H O (49) 2 5 1O 2 .18 'fl Cr+ (CH 602) + C5H120 2(50) Reactions (47)-(50) can be explained again based on what was discussed in the case of Fe+, namely, metal ion insertion, followed by a B-hydrogen shift. This is shown below and in Scheme XXI. Cr + O ——-> Cr'+/- ---> ”1-7K \V/ \V/- \\/, H \v/ I H l + Cr(C2H4O ) + c2H6 Note that Cr+(C3H802) and Cr+(C3H602) in reactions (49) (50) are formed following insertion into the bond between skeletal atoms 5 and 6, which is in contrast to Fe+ CY+ (C2H4 0) + C2H6 (47) 103 «on . «an +0 z 65 . ..o 1 us _ as. V . \o ..o tot "/16 up. U ...tuox . I a 0 ...V \J . .20/Lena 1» us. «09:66 tw no ofu I o\ V _ A . .20 (Loo...~ once «.6 \l/ I +.. _ 220 0:0 _ v ‘ .10 @1161? :«uaooz _. .. V a .+ . . o x .V «:6 n w - :3 01:0\ +5 _ /I\ _ .... 0 /\.......ol\/ 6: o a: \J \J \J / x u Al All 36 o 0 022 + c 51 5101.: o/ 6.20 as. a MEMIum 104 which inserts into the bond between skeletal atoms 4 and 5 (p. 56). Attack of a more centrally located bond in the polyether may be evidence for a symmetric intermediate in- volving multiple metal-ligand interactions. This possibi- lity is shown as the first step in Scheme XXI. Also, the strong Cr+-0 interaction (D(Cr+-O) = 77 i 5 kcal/mole) may favor the formation of small ligands with more oxygen's than does Fe+ - explaining the difference in site of attack for Fe+ and Cr+. In Scheme VXI, both possible B-H atoms can shift to get corresponding products. How- ever, there is only one B-H shift which occurs in the case of Fe+. The probability of such a H-shift depneds on the stability of the final product. Both pathways in Scheme XXI form ligands which can donate 4 electrons to the elec- tron deficient metal. Thus, the option of which B-H's will shift is related in part to the ability of the final pro- duct tO be a good ligand (strong bond to Fe+). Cr+ also reacts with TDE to form CrC2H40+ (reaction (48)). A variety of mechanisms have been considered, how- ever none are consistent with other mechanisms and observa- tions. We can speculate that, to form a rearrangement product consisting of 2 carbons and one oxygen, an insertion into the bond between skeletal atoms 3 and 4 by Cr+ would be necessary. However, this intermediate, has no H atoms which are 8 - to the metal. The intermediate, written as follows, shows that a six-membered ring intermediate gould 105 form a 4-membered ring product CrC2H40+ with subsequent loss of CH3(OCZH4)ZOCH3. 2 / \ + 0 -——> 0 Cr + CH 0 0 OCH CH ——0 0 OCH 3 3 \CHz—CrV 2 \ /\ / 3 \CH2/ 1 cr+o o .03 <-—+ Cri"fiH CH3 Note that the explanation above is based on Cr+ inser- tion into a C-C bond instead of a C-O bond, which is not usual, because Cr+-alkyl bonds are weak. TDE, then, appears to react via the same mechanism as diethylether. Cr+ and Fe+ apparently choose to attack dif- ferent C-O bonds of this polyether. This choice may be in part forced by the structure of the initial metal-ligand complex, and further favored by the stabilities of the final products. b. Reactions of CrCO+ The CrCO+ reactions with dimethylether, diethylether and TDE are summarized on the next page: 106 + + L + Cr(C2H60)2 ;§14.Cr(C4H]0)+ + C0 + .005 + CrC0+CH00(‘OCHT——->Cr(CHO)+CH0+CO(53) 31.]L/ 3 007 2 6 5 12 3 ' x + .; 7 CFC3H602 + C5H1202 + C0 (54) .013 + .006 + .069 + _———+ CFCH30H + C7H1403 + C0 (57) .170 + .104 + .011 + .011 + .003 + CrCO+ reactions with dimethylether and diethylether only result in ligand substitution and successive reactions. The product in (52) is not formed by Cr+ alone. If the metal inserted into a C-C bond, there would be fl2_B-H shift and eliminate CH4. Thus, we conclude that the CO of CrCO+ is actively involved in the process and inserts in toto into 107 the C-0 ether bond. Based on past observations, the neutral product of (52) could be C2H40 or [CO and CH4]. 0 0 H + \\\ + CrCO + \\/1r\/,-———a \\\T”’ Cr__—.C H “" l H o H—t +Cr H l c H "—c e—a c:~(c3+150)+ I This mechanism leads to acetaldehyde as the neutral lost in (52). The products in (54)-(56) were formed by Cr+ alone, and thus have been discussed in Scheme XXI. Apparently CrCO+ extracts methanol from TDE. Cr+ alone does not do so, how- ever the CO is not actively involved in the mechanism, since both CrCH30H+ and CrCOCH3OH+ are products. Thus, we inter- pret this via M+ insertion process, in which the C0 acts as a spectator ligand as shown in Scheme XXII. Once the metal center has induced the elimination of metanol from TDE the Cr(CH30H)(CO)+ can use the remaining available energy to break the weaker metal ligand bond, Cr+-C0. The products in (60)-(62) can also be explained by CrCO+ insertion into the central skeletal bond of TDE as shown in Scheme XXIII. Note that both Crcoc4H802+ and O + have a small contribution from m/e l38, 4”10 2 CrCOC3H60+, as their precursor, however, it would not be CrCOC 108 Scheme XXII. + Croo+ mo 0 0 me \_/\_/\_I l mac—cg— on ”\2\ H‘-CH--(KL_/;) 9MB 00 CrCHBOfi+1F—-——-CHj--O--Cf<; 'f' Vt—-O p; 9MB . \ + CO (CHBOH)($'((D) Scheme XXIII CrGO i‘ PBQ\—J/jt_/;\L‘/DMB mMLJLZJOJmfimMKCIJ l K *0 "e o + '-. + cL‘HBo2 + orcocl‘ 0°26— C: Cr c He H 1'0 + + c53803+ arcanmo2 “(3511303) .... C43100 2 109 expected to be reactive enough to experience successive reactions with large molecule like TDE.* Also note that the formation of CrCOC4H1002+ illus- trated in Scheme XXIII occurs when CrCO+ inserts into a C-C bond, followed by a B-H shift, and the C0 is retained on the metal center. Products in (59) are formed by the same mechanism as in Fe+ case, where the metal ion inserts into a C—O bond (skeletal atoms 4 and 5), followed by a B-H shift. Since Cr+ does not do so, but both CrC3H602+ and CrC003H602+ are products, the CO on Cr+ might be a spectator as described above in the formation of CrCH30H+. The formation of Cr(CZH60)+ in reaction (53) is again explained by CrCO+ in- sertion into skeletal atoms 3 and 4, so that B-H is availa- ble for shifting. * This is based on the assumption that a large molecule such as TDE has a steric effect which could prevent CrCOC3H50+ from approaching TDE to do any insertion, B-H shift process 1.e. CH CH CH 3‘0.. 3\o‘ 0/ 3 ' “'-c;-oo+ ms: 'ciftm + oo+c o E; Fe... :1; .... 53123 CH2 “#3168 0113\0‘ L—x——-> (SING;- o/Cfl3 110 c. Reactions of CrLCOlX: Higher CO-containing chromium ion reactions with diemthylether, and diethylether only give substitution reactions. However, it is more striking to realize that all of the products for TDE are reactive rearrangement of TDE - none are simple C0 displacements. This may reflect the ability of TDE to complex stepwise with the metal center i.e. - initial complexation may involve one oxgyen of the ligands. As further oxygens interact, additional energy is made available for metal-induced decomposition of the polyether e.g. Cr 3 CH I l a a ”‘31.ij ”wag—””31? + 061-11202 + 61.638602 111 CO here also acts as an energy mediator to stabilize the system, so that Cr+ can interact with another oxygen without decomposing. 2. Cyclic Ethers a. Reactions of Cr+ Cr+ doesn't react with THF except by forming an addi- tion complex. It reacts with THP to form only one product, Cr(C3H6O)t which can be explained analogously to Fe+ 4. .Cf \ x’ x + 0; 7 ‘———A Cr(C3H60) + 02H4 3. Cyclic Polyethers Cr+ + THP a. Reactions of Cr+ Cr+ reactions iwth l,3-dioxolan, l,3-dioxane, p-dioxane (d8), lZ-crown-4 and l5-crown-5 are summarized in Table 8. Presumably, the products in (63), (64), (65), (66), (67) can be explained in terms of the mechanisms which have been proposed for Fe+ (p. 64) for the reaction of Cr+ with l,3-dioxolan, l,3-dioxane and p-dioxane. Note that Cr+ is even more "selective" compared with Fe+. Cr+ reactions with l,3-dioxolan and l,4-dioxane give only one product, Cr(CH20)t 0n the hand, the reaction of Cr+ with l,3-dioxane produces the same products as Fe+ except for Cr(C2H4)+. Two impor- tant factors may control these product distributions: (1) Cr+ forms strong bonds with oxygen (77 i 5 kcal/mole, Table 8. Cr+ /\ + o 0 Cr + \_J 0 Cr+ + [:;F 0 0 :1 0 112 Reactions With Cyclic Polyethers O} 01 w 01 4.. Cr (CHZO) + C2H40. 4. Cr (C3H602) l———> Cr(CH20)(C3H602)+ Cr+(CH20) + C3H60 Cr+(C3H60) + CHZO Cr+(C2H402) + C2H4 Cr+(C4H802) Cr+(CH20) + C3H60 Cr+(C4H802) Cr+C2H40 + C6H1203 Cr+C2H402 + C6H1202 Cr+CZH4O3 + C6H120 ... Cr C4H803 + C4H80 + .1. + Cr (C4H100) + C6H1004 4. Cr (C4H803) + 06H1202 4. Cr (C6H1204) + C4H80 (63) (64) (65) (66) (67) (68) (69) (70) (71) (72) (73) (74) (75) 113 O / w + b 0 shift it L) T O———Cr L.) + Cr(CHZO)+(—— C2H4O Ne 2w l 0 .T « ‘l h C O 2 H C r C + O O 6 H 3 a c O O /[I\ + + r C ,G-H -—————9 ShlfT +’0 H Cr + Cr + ___> 0 0 o cmcwof' <—— \ 242 ‘— ‘ + Cr(L2H4) 114 see (Table 2) compared with Cr+-C(H2)(R) (37 i 7 kcal/ mole), that is, it may prefer to produce products having a high O/C ratio and bonds to oxygen atom directly. (2) Stepwise metal ion double insertion, followed by B-H's shift is occuring so that chormium can insert into a bond which is in geometrical proximity, due to a weaker initial complexation energy as a driving force (see thermodynamic conclusion section) as shown on page 113. The chemistry of Cr+ with l2-crown-4 differs from that of Fe+, apparently due to the fact that, relative to Fe+, Cr+ interacts more strongly with oxygen and less strongly with carbon. The first product, CrC2H40+ in (68) could be explained using a mechanism similar to that in Scheme XII. Note that products in (69)-(72) have ligands with a higher O:C ratio than those formed in Table 5 with Fe+. Consider the pro- duct CrC2H402+. This could be considered as either Cr+ complexed to Eigflflg or as a metallocyclic 5-membered ring. The latter will be assumed because of the strong Cr+-0 bond. These products can then be explained if the Cr+ complexes in the crown cavity, and, because of its size, can only interact with two oxygens at a time. Apparently no H-shifts occur. The pathways leading to products (69), (7l), (72) are shown in Scheme XXIV. Note that all pairwise combina- tions of oxygens are sampled. The product in (70) has a very high O:C ratio in the 115 SCHEME m 0 O C J O O /—'\ \_/ \ o o C 3 0,. “O \__/ fl 0 O O O O C Cr: 1 C 96' j V C” 1 ‘~ / ’ “ ‘0 o o. ,0 9,4,0 , l F" ’ ‘ ’ ‘ \ + o o o o 0 Cr E j < > L/ 0 O \ +/ + Cr + 0 + /C'+\ ( 7 C2H4 O O 116 ionic product, and may be the result of the metal ion "puckering" up the molecule in an attempt to interact with 3 oxygens of the crown. 0 —— CH2 CH2/ \“ k \CH/0 0 I \Cr+ 2 I ”20,, \\\ +_ 0 CH2 ,I‘V¢\\ /,CH2 ‘———? l Cr-—-0 + 02H4 + \ I, “ /CH2 H C / 0 0 2 \0 \ / It is striking to notice that Cr+ forms an adduct ion (addition complex) with lZ-crown-4. Again, this could be the result of the small metal ion puckering up the molecule in an attempt to interact with 3 or 4 oxygens of the crown 117 as some alkali metal ions do However it is hard to interact all 1 oxygens because Cr+has a larger size than the cavity of Iz-crown-4, (or+ has a radius of 0.8l A146’147, 0 compared with the cavity of lZ-crown-4 which is l.2 A - 1.5 A in diameter), Cr+ must be to some extent "away" from the cavity center, which would result in a longer range interactions (ionic rather than covalent)* with all four oxygens and no bond breaking will occur. Fe+ has the same tendency for interacting with both oxygen atoms and carbon. * + 6 ‘ . . 5 Cr has a 5 ground state corresponding to an [Ar] 3d con- figuration157. The next exaited state is 6D at l.47 V, corresponding to [Ar] 45 3d . Despite the energy gap of 1.47 eV, it was reported that a long-lived metastable excited state of Cr+ exists in ICR on electron impact 02 Cr(C0) 14‘ 143. It is not surprising that Cr+( S) will inter- 3 act wigh oxygen atoms electrostatically through both intrin- sic and induced dipoles. To form covalent bonds, it has to use an sd hybrid and Cr(5D) will form covalent bonds with oxygens. 117 If they interact, energy is released and much chemistry occurs, since it can only form bonds covalently. Thus Cr+ exhibits quite a different chemistry with lZ-crown-4 than does Fe+. This can be accounted for by the greater pre- ference of Cr+ for bonding to oxygen, which appears to be the predominant driving force in the Cr+ reactions. The attempt to interact with g and 4 oxygens by Cr+ is also observed in its reaction with l5-crown-5. Products in reactions (74) and (75), namely Cr(C4H803)+ Cr(C6H]204)+ were formed by the mechanism shown in Scheme XXV, in which Cr+ need only make an effort to pucker oxygen atom #1, since oxygen atoms #7, 4, T3 are already on the same plane. The product in (73) can be explained by a mechanism similar to that of in Scheme XIV - after forming an inter- mediate (V) in Scheme XXV, Cr+ inserts into two C-O bonds, followed by double B-H shifts to form Cr(C4H100)+. Scheme XXV. :3 ..d;.oj bund;,d + o C. a. ..7 —-> «W2 Li J: LJ (V) C17) (‘o/W {—on :1 ——> (0‘02?) 03—) “<%’*3°3) + 8 \-—’° °\__Jo 118 b. Reactions of_Cr(C0)x: Cr(CO)x+ (x = 1-6) reactions with l,3-dioxolan, l,3- dioxane, p-dioxane, lZ-crown-4 only result in the addition complexes, substitution reactions and successive reactions. Cr(C0)x+ reactions With lS-crown-S yield two new products: CrCO+ + lS-crown-S ——> Cr(CZH202)+ + caumo3 + co (76) Cr(C0)x+ + lS-crown-S ——-a-Cr(C2H502)+ + C8H1503 + xCO (77) x = 1,2 Products in (76) and (77) can be explained by a mechan- ism similar to that shownin Scheme XIII. Formation of Cr(C2H502)+ may be an indication that Cr+ is sequentially inserting into C-O bonds, followed by B-H shifts to avoid an intermediate with an unusually high formal oxidation state for the metal. Further B-H shifts might not occur if they are geometrically inaccessible after one B-H shifts. Referring to Scheme XIII, no products corresponding to double or single B-H shifts from the larger ring side are observed. Since Cr+ has a strong bond energy with oxygen atoms, the formation of products in reactions (76) and (77) is accompanied by the loss of CO to carry away the extra energy. 119 4. Thermodynamic Conclusions Table 9 lists all neutral lost in the Cr(C0)x+ reac- tions with all the ethers discussed above and their branch- ing ratios (product distributions). Dimethylether can displace up to two CO's from Cr+, as can diethylether, implying that the initial Cr+-ether inter- action, 66.2 kcal/mole < D(Cr+-Me20), D(Cr+-Et20) < 87.7 kcal/mole, is less than that of Fe+, 96.6 kcal/mole < D(Fe+-Et20) < l33.3 kcal/mole, but larger than that of the dimethylether case, 27.l7 kcal/mole < D(Fe+-Me20) < 73.l7 kcal/mole. Cr+ can only react with diethylether by H-shift pathway (p. , Scheme X) to get Cr(C2H4O)+. The difference from that of Fe+ may be in part thermodynamics. Recall, L—-——e C2H50H + CZH4 AH l6.8 kcal/mole Cr+ can induce the rearrangement indicated in the first process, which requires very little energy, AH = 0.3 kcal/ mole. Fe+ induces both rearrangements. More energy is required for the second process, AH = l6.8 kcal/mole. Thus Fe+ forms "higher energy" products than does Cr+ However, one must be careful when thermodynamic conclu- sions are drawn for Cr+, since electron impact on Cr(CO)6 +l4l,l48 produces excited states of Cr Nevertheless, were SnQB floor} 3w. v8n+no~a=ou 8N n .335 120 2.“...V8N . omzau E . V8~ «of? AHOQ 8Nw 3H; 8~ 2.3” SN; 8~+om=aom soda-8 532%.- 938. o< c 8. SN. 338 + N lac E348 +~oon=nu So; non—2mg 938 +~o2=mu Amc.vao~ oudmzau a: near} ASJB oao~fl=no A8. v8~ .~oon=uu 2048.233} 3~¢8 «near? A$¢8~+no~fl=mu Go. v8~ .nomnzou AB.V8~+no£=~.u 8N -—--.-_-...———.__ were $38 . op} $.18 ~05.on Aow. V8 .nonxmu ANH. V8 4nomH=mu v -— —-———-—4 2o; 8 SA; 8 ...=~u I‘vil so; on; 63 Noafio EN; ..oofoo GN; 3 :3 as} 8 2.; an; 8+J:Nu_ 8 28; Nomzao 38; «our? 28¢8+~om=au A§.v8.no«£uu 82¢ no}? $38.93“? 38; «com—mu 88. V8 .Noormu A§.V8.~o~n=no Anoo¢8¢no§=ou 2.0; 8 an. Voéf 68.5 mm w< w... 8n; coxno Ann; 9. “no; oarnu 23; A8; a=~u U< U< 8a; ~o~n=ng $3 wooden". So; noise; 33 .2” cm; ozwm “IcaDHOIflH ox r u O. O O E £23 5; 495.5 no 23308.. 9:. 5 Illa-3 3.53.. .dllloSa. + _ \]o\/_ \ xo 121 although the dissociation energy of dimethylether into CHZO and CH4 is only 0.9 kcal/mole, we neither see Cr(CH20)+, nor addition complex from Cr+. The reason for this is not clear. It is of interest to note that no ion of composition Cr(TDE)+ or Fe(TDE)+ was observed. In the case of lZ-crown-4, no ion of composition Fe(lZ-crown-4)+ was observed. Cr(12-crown-4)+ is formed, with 3 precursors: Cr+ + lZ-crown-4 -———9 Cr(l2-crown-4)+ CrC01 + l2-crown-4 -———9 Cr(12-crown-4)+ + C0 Cr(c0)2+ + lZ-crown-4 -———> Cr(12-crown-4)+ + zco (78) Since Cr(CO)3+ is not a precursor to Cr(l2-crown-4)+, one may interpret this as an indication of the Cr+-crown bondstrengths: D(Cr+-CO) + D(CrCO+-CO) < D(Cr+-12-crown-4) < D(Cr+-C0) + D(CrC0+-CO) + D(Cr(C0)2+-CO) however, this may be an incorrect interpretation. The metal- polyether interactions should be very strong; complexation appears to release enough energy such that subsequent frag- mentation of the polyether always occurs, except for Cr+, which appears to form weaker bonds to ethers than Fe+. Rather, the reaction 122 /"‘\ O 0 Cr(C0)3+ + E: :j ———+ Cr(12-crown-4)+ + 3C0 may not occur because the carbonyl ligands prevent this large ligand from getting close enough to the metal for significant orbital overlap. This is verygraphically seen as we progress from diemthylether to diethylether to TDE to l2-crown-4. The reactions due to M(C0)n+ decrease as n increases. Thus, reactions ofM+ alone are presumably due to the metal in or very close to the crown-cavity. As CO's are added to the metal, reactions occur with the metal out of the cavity to the point where CO's prohibit metal-ligand interactions with this bulky ligand. Thus, when 3 or more 60's are present on the metal, sufficiently close approach of the bulky l2-crown-4 ligand is prevented and no reactions, not even simple idsplacement is observed. Thus, it is difficult to interpret the data here in terms of estimating M+-polyether bondstrengths. The Cr+- lZ-crown-4 bondstrength, based on reaction (78) must be > 66 kcal/mole. In conclusion, Cr+ has a weaker complexation energy with ethers than Fe+ does. Cr+ doesn't form an adduct ion with lS-crown-S, possibly due to the large cavity of lS-crown-S. As CO's are added to the metal, it only under- goes substitution and successive reactions in smaller ethers. 123 In the reactions with lZ-crown-4 and l5-crown-5, however, the interaction with these bulky ligands is prevented or confined to interact with only a few oxygens in the crowns. III. Ni(C0)x+ Reactions With Ethers and Polyethers A. Results 1. Linear Ethers and Polyethers a. Ni(C0)X+ reactions with dimethylether 1.92m). Ions formed as products of ion-molecule reactions in a mixture of Ni(C0)4 and dimethylether are listed below, with their precursors as identified by double resonance. ELE stoichiometry precursor(s) 88 Ni(CH20)+ Ni+, Nico+ 104 NI(C2H60)+ Ni+, Nico+ 132 Nic0(c2H60)+ Ni(C0)2+, Ni(c0)3+ 150 Ni(c2H60)2+ Ni(C0)2+, Ni(c0)3+ l60 Ni(CO)2(C2H60)+ Ni(CO)3+, Ni(c0)4+ 178 NiCO(C2H60)2+ Ni(C0)3+, Ni(c0)4+ 190 Ni2c0(c2H60)+ Ni+, NiCO+, Ni2(c0)2+ 218 Ni2(c0)2(c2H60)+ Ni+, Nico+ b. Ni(CO)X+ reactions with diethylether L£4fl1091 Ions formed as products of ion-molecule reactions in a mixture of Ni(CO)4 and diethylether are listed below, with 124 their precursors as identified by double resonance. gig stoichiometry precursor(s) 86 Ni(C2H4)+ Ni+ 102 Ni(c2H40)+ Ni+, NiC0+ 104 Ni(C2H50H)+ Ni+, Nico+ . + .+ . + . 132 N1(C4H100) N1 , NICO , N1(C0)2+ . + + 143 N1(CO)3H czns 145 Ni(00)2(ocn3)+ 01130+ . + + + 155 N1(C0)2C3H5) “ C3H5 , C3H7O 160 NiC0(C4H100)+ Ni(C0)2+, Ni(c0)3+ 188 Ni(CO)2(C4H]0)+ Ni(CO)3+, Ni(C0)4+ + + 191 (C4H100)2(C3H7) CH30 206 Ni(C4H]00)2+ Ni(C0)2+, Ni(C0)3+, . + N1C0(C4H100) 218 Ni2C0(C4H100)+ Ni+, NiC0+ , + . + . + 234 N1C0(C4H]00)2 Ni(c0)4 , Nl(C0)2(C4H]00) The ion of m/e l43 is a proton transfer product. The ions of m/e l45, 155, l91 are products from organic ion reactions with neutral Ni(CO)4 formed from 0CH3+ and C3H5+. c. Ni(C0)x+ reactions with triethylene glycol dimethylether (TDE, C8fl1894) Ions formed as products of ion-molecule reactions in a mixture of Ni(C0)4 and TDE are listed below, with their precursors as identified by double resonance. 102 103 116 118 130 133 134 144 148 160 162 163 188 204 221 236 125 stoichiometry Ni(C2H40)+ Niczhso+ Ni03H60+ Ni(03H80)+ Nic0(02H40)+ Ni(c3H702)+ Ni(C3H802)+ NiC0(C3H60) Ni(C4H]002)+ . + N1CO(C3H602) )+ + Ni(C5H]202 Ni(c4H903)+ Ni(00)2(c3H602)+ Ni200(02H40)+ Ni(c7HMo3)+ . + N1(C7H]504) Ni(10E)+ precursor(s) .+ N1 + 02H30+, C3H70 Nico+ Ni+, Nico+ Nico+ Ni+, Nico+ Ni+, Nico+ Ni(CO)2+, Ni(00)3+ Ni+, Nico+ NiCO+, Ni(C0)2+, Ni(00)3+ Ni+, Nic0*, Ni(00)2+ Ni+, Nico+ Ni(C2H40)+, Ni(C0)3+, Ni(00)4+ NiCO+, Ni(c0)2*, Ni(CO)3+, Ni(C0)4+ NiCO+, Ni(00)2+ Ni(C0)2+, Ni(C0)3+ Ni(CO)4+ chlic Ethers and Polyethers a. Ni(CO)x+ reactions with tetrahydrofuran THF Ions formed as products of ion—molecule reactions in a mixture of Ni(C0)4 and THF are listed below, with their pre- cursors as identified by double resonance. ELLE 100 102 112 127 130 155 158 186 202 216 230 244 Note that c H + stoichiometry Ni(c3H6)+ Ni(CzH4O)+ Ni(c4H6)+ Ni00(c3H5)+ Ni(THF)+ Ni(CO)2(C3H5)+ NiCO(THF)+ Ni(00)2(THF)+ Ni(THF)2+ NiZCO(THF)+ . + N12(THF)2 Ni2(00)2(THF)+ 3 5 precursor(s) Ni+, Nico+ Ni+, Nico+ Ni+, Nico+ + C3”5 NiC0+, Ni(00)2+ + C3”5 NiC0+, Ni(c0)2+ . + . + Nl(C0)3 , Ni(00)4 NiCO+, Ni(CO)2+, Ni(C0)3+, Nic0(TH1-')+ Ni+, NiCO+. Ni2(00)2+ . + . + N1(C0)2 , N1(C0)3 , Ni(CO)4+, Ni(CO)2(THF)+ Nico can displace up to 3 CO's from Ni(CO)4 to produce m/e 127, Ni00(c3H5)+ and m/e 155, Ni(C0)2(C3H5)+. 127 b. Ni(CO)x+ reactions with tetrahydropyran (THP, Csfllog) Ions formed as products of ion-molecule reactions in a mixture of Ni(CO)4 and THP are listed below, with their pre- cursors as identified by double resonance. mze stoichiometry precursor(s) . + .+ 112 N1(C4H6) N1 . + + 127 N1C0(C3H5) c3115 144 Ni(THP)+ NiCO+, Ni(00)2+ . + + 155 N1(C0)2(C3H5) c3115 172 Nicomip)+ Ni(C0)2+, Ni(00)3+ 205 Ni(C0)2(THP)+ NiC0+, Ni(C0)3+, Ni(00)4+ 230 Ni(THP)2+ NiC0+, Ni(C0)2+. Ni(00)3+ Nicomip)+ Note that c3H5+ reacts with Ni(00)4 to form m/e 127 and m/e l55 as was observed in the mixture containing THF. c. Ni(C0)x+ reactions with l,3-dioxolan I£3fl5921 Ions formed as products of ion-molecule reactions in a mixture of Ni(CO)4 and l,3-dioxolan are listed below, with their precursors as identified by double resonance. m[e stoichiometry precursor(s) 86 Ni(02H4)+ Ni+ 88 Ni(CH20)+ Ni+, Nico+ 128 gig stoichiometry precursor(s) 102 Ni(C2H4O)+ Ni+, NiCO+ 130 NiCO(C2H4O)+ c2H40+ 132 Ni(C3H602)+ NiCO+, Ni(c0)2+ 160 Ni00(03H602)+ Ni(CO)2+, Ni(c0)3+ 188 Ni(CO)2(C3H602)+ Ni(CO)3+, Ni(c0)4+ 206 Ni(C3H602)2+ Ni(CO)2+, Ni(CO)3+, Ni00(c3H602)+ 216 Ni(00)3(c3H602)+ Ni(00)3+ 234 NiCO(C3H602)2+ Ni(C0)3+, Ni(00)4+ . + N1(C0)2(C3H602) Here, the reactive organic species is CZH40+, which reacts with Ni(CO)4 to displace 3 CO's. d. Ni(CO)x+ reactions with l,3-dioxane L94fl8921 Ions formed as products of ion-molecule reactions in a mixture of Ni(C0)4 and l,3-dioxane are listed below, with their precursors as identified by double resonance. m[e stoichiometry . precursors(s) . + .+ 86 N1(C2H4) N1 88 Ni(CH20)+ Ni+, Nico+ . + + 115 N1(C3H50) C3H50 . + . + 118 N1(C2H402) N1CO . + . + 132 N1(C3H602) N1C0 129 mlg_ stoichiometry precursor(s) 145 Ni(C0)2(0CH3)+ . c2H +, OCH3+ 146 Ni(C4H802)+ Ni+, NiC0+, Ni(c0)2+ 160 Ni00(c3H602)+ Ni(C0)2+, Ni(c0)3+ 174 Nic0(c4H802)+ Ni(C0)2+, Ni(c0)3+ 188 Ni(CO)2(C3H602)+ Ni(CO)3+, Ni(c0)4+ 202 Ni(c0)2(c4H802)+ Ni(c0)4+ 220 Ni(C4H802)(C3H602)+ NiCO+, Ni(C0)2+, Ni(C4H802)+ 234 Ni(C4H802)+ NiCO+, Ni(C0)2+, Ni(00)3+ . + . + N1(C4H802 , N160(C4H802) Where m/e llS and l45 are products arising from the H 0+ and OCH + ion-molecule reactions with netural Ni(C0)4. c3 5 3 e. Ni(CO)x+ reactions with l,4-dioxane 12,889.22 Ions formed as products of ion-molecule reactions in a mixture of Ni(CO)4 and l,4-dioxane and p-dioxane-d8 are listed in Table 10, with their precursors as identified by double resonance. Note that m/e 87 was formed by H atom abstraction by Ni+. f. Ni(C0)x+ reactions with 12-crown-4 fl8fll 5941 Ions formed as products of ion-molecule reactions in a mixture of Ni(C0)4 and lZ-crown-4 are listed below, with their precursors as identified by double resonances. l 130 “Naming v82 .momzao .mABXz .wBXz wwomzsoxz +3 Amomooo V82 .momnso momma .mABXz @832 mfimomasozz RN £83: .mABVE ANomzsovaBXz 8m “#8:: .mABXz $093383: SN mamas". . + momnau .mABXz .wBVE .uwomESBE 8S .mABXz .w8zz «$933338: 8a m8: 8: n2 +A~om=suzz 9: +82 .+2 “$28“..qu NNH +82 ma ...Aosxwuxz «3 +82 ...“: Mosnmoxz 82 +82 “2 +845sz 8 3 +A 8N8: +82 w: +32 8 mo: pg +3883. co m: +3332 cm awquomnsoowm NM» 652230: fl. auogomu Egowngopw fl mm ocdxotflTA ocaommih mu 2338A 3 3:688 5:. 838.98: 88:... .3 33.. gig 100 102 103 118 146 157 174 178 202 204 206 214 216 232 234 131 stoichiometry Ni(C2H20)+ Ni(c2H40)+ Ni(002H5)+ Ni(c2H4oz)+ Ni(c4H802)+ Nic0(c2H30)+ Ni00(c4H802)+ . + N1(C5H]203) . + N12CO(C2H202) . + N1ZCO(C2H402) NiCO(CSH1203)+ . + N12(CO)2(C2H20) . + N12(C0)2(C2H40) Ni(08HMo4)+ , + N1(C8H1604) precursor(s) Ni+, NiCO+, Ni(c0)2+ Ni+, Nico+ + CZHSO Ni+, NiCO+, Ni(00)2+ NiCO+, Ni(CO)2+, Ni(00)3+ c2H30+ Ni(C0)2+, Ni(CO)3+, Ni(c0)4+ Ni(C0)2+, Ni(00)3+ Ni+, NiC0+, Ni2(c0)2+ Ni+, NiC0+, N12(C0)2+ Ni(C0)2+, Ni(C0)3+, Ni(00)4+ Ni+, NiCO+, Ni(C2H20)+ Ni+, NiC0+, Ni(C2H4O)+ Ni, Nico+ NiCO+, Ni(C0)2+, Ni(C0)3++ Ni(CO)4+ Note that 02H50+ is the most prominant ion in E1 spectra of all crowns from C2 Ions formed as products of ion-molecule reactions in a 5 H 0+ 149 The ions m/e l03, and C2H3O+ reactions with netural Ni(C0)4. g. Ni(C0)x+ reactions with lS-crown-S LC40520951 mixture of Ni(C0)4 and lS-crown-S are listed below, with 157 are formed 132 their precursors as identified by double resonance technique. gig stoichiometry precursor(s) 118 Ni(02H402)+ Ni+, Nico+ 152 Ni(c4H803)+ Ni+, Nico+ 190 Ni(CGH]203)+ Ni+, Nico+ 192 Ni(C6H1403)+ Nico+ 207 Ni(CGH]3O4)+ Ni+, Nico+ 278 Ni(C1OH2005)+ Ni+, NiC0+, Ni(CO)2+, Ni(CO)3+, Ni(c0)4+ B. Discussion l. Linear Ethers and Polyethers a. Reactions of Ni+ Ni+ reactions with dimethylether, diethylether and TDE are summarized below: Ni+ + /// \\ —794914 Ni(CH20)+ + CH4 (79) LgeiieiNi(c2H60)+ Ni+ + o —~9e5-1-)»Ni(czu4)+ + C2H60 (80) 9413—>Ni(c2H40)+ + C2H6 (81) 0.20 . + -————+ N1(C2H60) + C2H4 (82) 0.21 - + _—--9-N1(C4H10) 133 .+ 0,6] . + N1 Me%_JR_Ja_JPMe —v-——91N1(CZH4) + C6Hl403 (83) 0.10 - + 0.07 ———+ Ni(c3H702)+ + 05111102 (85) 0.10 . + «————> N1(C3H802) + 05111002 (86) 0.03 - + 0.02 . + -————7N1(CSH1202) + C3H602 (88) "9:925 Ni(c4H903)+ + 541190 (89) Products in reaction (79)-(82) are similar to those reported for Pet Note that Ni(CH20)+ doesn't undergo succes- sive reaction. Also, Ni+ can directly form an addition com- plex with diethylether. This may, perhaps, indicate that Ni+ has a weaker complexation energy with oxygen in ethers than Fe+ or Cr+. Products in (83), (85), (86) and (88) can be explained similarly to that of Cr+in (48)-(50) except that no B-H shifts in (85). It may shift but be retained on the metal center. The product in (84) can be explained in a similar way: 0 CH Ni+ + TDE-———a MeQ¥_/N1//J\V/O pMe-———9 Ni+---0<:i 3 + H CH3 0 0 OMe \\_/\_/ The product in (87) might be formed similarly to that 134 in (83): .... 2:: . + 0 3e/SH ;:}————9 N1(C4H]002) 1+ [::j LE4 The product in (89) is hard to explain. One possibility is that Ni+ inserts in a stepwide manner into c-o bonds between skeletal atoms (5,6) and (10,11): Me 0 Ni++Me0 o 0 0Me———>Me0 0 + \_J\_/\_/ ‘ \\ 0 Nt\v// 0 )+ + c 9 0\]———>Ni(C4H9 3 4H90 5. Reactions of Nico+ NiCO+ reactions with diemthylether and diethylether only result in substitution and successive reactions. NiCO+ reacts with TDE yielding some new products: NiCO+ + 105 -w———+ NiC3H60+ + c5H1203 + co (90) i———» NiCO(C2H4O)+ + €5H14°3 (91) -———9NiC0(03H602)+ + 05111202 (92) _———a Ni(c7HMo3)+ + CH40 + c0 (93) 135 )+ + CH + c0 (94) -———9 Ni(C7H]504 3 Products in (90 and (92) follow the same mechanism as that in (84) and (86) except that the B-H shift occurs from the "other" side after Ni+inserts into C-0 bonds. In both cases, CO acts as a spectator with and without con- current 1055 of C0 in (90) and (92) respectively. The product in (91) is 28 mass units above Ni(CzH40)+, however, the incorporation of C0 might be an indication of a different mechanism, because it makes a B-H shift availa- ble for shifting after NiCO+ inserts between skeletal atoms 3 and 4. Similarly, the product in (93) can be explained by Ni+ insertion into the c-o (skeletal atoms 10 and 11), followed by a B-H shift. The formation of Ni(C7H]504)+ is a high energy process, i.e. once N? inserts into the CH3-0 bond, it eliminates the «CH3 radical. Ni(C0)x+ reactions with diemthylether, diethylether and TDE only are substitution reactions. However, one important result has to be mentioned here: Ni(00)x+ + TDE-———? Ni(c0)x_2(105)+ + (x-2)CO x = 2,3,4 Hence, the formation of the Ni(TDE)+ adduct ion is different from what we have seen in the Fe+and Cr+ cases where the interactions between the metal center and oxygen 136 atoms (or -CH2-) are so strong that both Fe+ and Cr+ easily induce fragmentations. From Table 2, it is known that the bond energies of Ni+ to oxygen, a methyl group and hydrogen are small (D(Ni-O) 2 45 kcal/mole, D(Ni+-CH3) z 49 kcal/mole and D(Ni+-H) e 43 kcal/mole) so that it can form an addition complex and randomly insert into any C-O or C-C bond in TDE as shown in Scheme XXVI. Also, the odd mass products corresponding to either no B-H shift or H atom retention by the metal were observed in the reactions of Ni+ and TDE. The Ni+ and NiCO+ reactions with TDE can be summarized in Scheme XXVI (the numbers on the arrow bar are the skele- tal atoms of TDE and are used to indicate the bond into which Ni+ inserts to yield reaction products). Scheme XXVI. bonds bang: inserted reactions 111+ + CH3OE4jq5: (0,8 28% 3’“ 4(83). (91) 3 “'5 9(8‘1).(90) 5'6 Has). <86). <92) 6'7 e (87) 748 7‘ (88) 10,11 g) (93) 11,12 >(94) 137 2. chlic Ethers a. Ni+ reactions with THF and THP The Ni+ reactions with THF and THP are summarized below: Ni+ + THF '30; Ni(C4H6)+ + H20 (95) '27: Ni(c3H6)+ + CH20 (95) ——43§-> Ni(c2H40)+ + c2H4 (97) Ni+ + THP ; Ni(C4H6)+ + CH40 (98) Products in (95), (96) and (98) are similar to those observed for Fe+. However, the product in (97), Ni(C2H40)+, is a new product, and can be explained as follows: + 0 H 0\Ni+ Ni + ———+ ————_9 1 7 H H 0 0"‘Ni+"'|( ——-9Ni+--) + CZH4 :][: CH3 Note that the enol ligand might rearrange to keto from which is more stable. 138 b. Ni(CO)x+ reactions with THF and THP As the number of CO's present on the metal increases, ligand substitution becomes the predominant process. 3. chlic Polyethers .+ . a. N1 react1ons Ni+ reactions with l,3-dioxolane, 1,3-dioxane, p- dioxane (d8)’ lZ-crown-4 and 15-crown-5 are listed in Table 11. Again, most reaction products have been observed in the reactions of Fe+ with l,3-dioxolan, 1,3-dioxane, p-dioxane and lZ-crown-4 except that Ni(C2H40)+ is present in the reactions (106) and (109). Presumably, the formation of Ni(C2H4O)+ in reactions (101), (106) and (109) follows the same mechanism as was explained in Scheme XII. Note that Ni+ a1so reacts with lZ-crown-4 to give Ni(C8H]4O4)+ with an elimination of one molecule of H2 as shown in reaction (111). This result implies that Ni+ actually interacts with one or two oxygen atoms only, and then inserts into a C-O bond, followed by the B-H shift. Products in (112) and (113) are similar to that ob- served for Fe+ as shown in Scheme XV and XVII. Products in (114) and (115) can be explained in terms of Scheme XXVII and XXVIII. Table 11. N Ni+ + 6Ao \_J + N1 +6?) .+ 'l .29 ~37 k: .20 .10 .11 .21 L———> 139 , + N1(CH20) 1021140 + Ni(C2H,+) + 011202 111(c2114o)+ + 01120 + “(914) +C211402 11210311120)+ +C3H60 Ni( 0411802 )4’ ——->+Ni(C2H4) +028402 ._2§_, 111(01120) +c31150 L———+\+N1(021140) +c21140 N1“+ (cf-‘2) —=&-> 111(021120 )++ (161-11403 .o . —5—)N1.(C2H,+0)++ c611 12° 3 lanficzntpz) +c611120 2 L92——9Ni(08H140“) + H2 .20 -'——91~11(0211,+02)+ *°8"16°3 .12 ——-+N1(c,+H803)* + 0611120 2 0? ~N1(C6H1203): + 0411802 ieui(clouz:o5)* Reactions With Cyclic Polyethers (99) (100) (101) (102) (103) (104) (105) (106) (107) (108) (109) (110) (111) (IR) (113) (114) (115) (116) 140 Scheme XXVII H + 0 \N23 ——)Ni(C6l-11203) + (0:) H 1_1 " + 2 B -H CONN? + C4H602 L)" C2115 Scheme XXVIII The formation of the adduct ion, Ni(C10H2005)+ might indicate that Ni+ interacts with one or two oxygens in the crown and since their interactions are not so strong, frag- mentation does not occur. b. NiCO+ reactions Nico+ reactions with l,3-dioxolan, l,3-dioxane, p- dioxane, 12-crown-4 and 15-crown-5 are summarized below: /\ + 0 0 NiCO + \_j Nico+ + (:3) ——1 141 .7___.-14 N1(CH20)+ + c2H40 + c0 .67 . + _____, N1(C3H602) + co 413—» Ni(CzH40)+ + c1120 + co 499—) Ni(CH20)+ + C3H60 + co .23 . + .20 . + ___—’N1(C3H602) + CH2 + CO (118) 0 NiCO+ + [j —— 0 24.81., Ni(C4H802)+ + co r31.. Ni(CH20)+ + C3H60 + co Jae Ni(C2H40)+ + C2H40 + co 499—» Ni(C2H402)+ + c211 + co 4 [—1 NiCO++ ESL—PO) n 425—) Ni (C4H802)+ + co fig-e Ni(c21120)+ + C5H14°3 + co .05 - + ‘_—_9 N1(C2H40) + C6H1203 + C0 .36 - 4' -—_—>N1(CZH402) + C6H1202 + C0 014) ' + / N1(C4H802) + C4H802 + C0 (119) ll . . + ~———a N1(CBH1404) + H2 + C0 .02 . + —9N1(08H1604) + C0 (120) 142 0 + co o/W - + Ni(C5H]203)+ + (x-l) Co + C H 02 x = 2,3 4 6 Ni(C0)x+ + 12-crown-4 ———> NiCO(C5H]203)+ + (x-2)co + C H 0 x - 2,3,4 4 6 2 These two products can be understood in terms of the structure as shown in Scheme XII (p. 74), in which the Ni+ will interact with oxygen atom #1 and carbon atom #5 as shown below: -co 0\\’u+ \ ZB-H \ ° O/NL JO shift7 4/ c H 0 1‘ 9 < :.+ N1“\ + C4H602 0 ‘0 \ / \CH3 It is of interest to note that both 12-crown-4 and lS-crown-S can displace up to 4C0's from Ni+. If the reac- tions are induced from the cavity center, C0 ligands on Ni+ will prevent it from getting close to all of the crown's oxygen atoms. 144 4. Thermodynamic Conclusions Table 12 lists all neutrals lost in the Ni(CO)x+ reac- tions with all ethers discussed above and their branching ratios. Dimethylether can displace one CD from Ni+, and diethyl- ether can displace two CO's from Ni+, suggesting that 6.58 kcal/mole < D(Ni+-Me20) < 71.17 kcal/mole. Similarly, for diethylether, 71.17 kcal/mole < D(Ni+-Et20) < 101.74 kcal/ mole. In comparison with the complexation energy of Fe+, Ni+ has a smaller complexation energy. Thus, we did not observe Ni(C4H80)+ as a product in the reaction of Ni+ with diethyl- ether, since less energy is available. Such an analysis in the reactions with cyclic ethers and polyethers may not be useful, since orbital comparability, orientation in space and many other factors are involved. Both THF and THP can dis- place up two CO's from Fe+ and Ni+, but only one product is observed in the reaction of Ni+ with THP, (which produced three products in the case of Fe+). In contrast, Ni+ forms three products in its reaction with THF, (which also forms three with Fe+). The difference is that although both com- pounds can displace up to 2C0's from both Fe+ and Ni+, the bare Ni+ does not contribute to the formation of the addi- tion complex. This in turn relates to the bonding of ligands to the metal center and the orbitals used by the metal center, etc. 145 21:5: 30¢ 8... AR¢8n+o£nu :3 8~+~5 28.33 + 0:6 8.: A$.v§+~OD—_30 8N A0.v 8N 8N 8N 8N as. v8N¢NONH=h0 8N ”RSV“: “magmatic “WW.” . ...s 3M¢Snmmwunaw A84 830 9. A8; 8+~o§¢no Ao~.v8~.~om:..u 8” A8; SN 8H 8a 8a Ro¢8~mofiunu 8” Bo fin We”? E Q¢8~ NB 8~ 8~ 8~ €1v8~Mo~fno 8~ 8 mA8v= 3.38 + osuru Go; 8~ GH¢8~ + as} G”; 8~ + mo :38 . use...” :3 SN . of 69x3 2%.? 2K; 8 “no; 8~+~Oo=no A-¢8~+~ofi=ou 8 A8; 8 n5 8 8 8 28¢ 8+~ofiuno 8 8~ AS.V8~Mo£=ou 8“ AS; 8~ 8N 8N 8~ A8; 8+noflnnu 8N 8 mgr: 8a; 8 + ...u “so; 8 + can. 33 8 + oozau A8; 8 Keogh". 3o; ~o~fl=no 8.38 A8; 8 A8; 82%.? :58 + Raid; 8 w: RH; 8+~oon=m° 3:8 +~oo=¢2adv8 +~om=ao Tn“; 8 8.3 8 AR; 8 3o; Shear} SJB .~%=JAR.V8+~oN£ou “8; 8.3} 8: 836 Co; 8 $3 83$ 3”; noise... an; 8 A8.V8+~o~a=w%no.v8.no~n=ou “8.863%29; 8.33 A2; 8+o~6 A8; 83mg 88; 8+noormo 2a; 8+a=~u So; 8 A8.v8.now§3n.v8.nosroo f~¢8soonn2$v86¢=nu $fi86§~u 8 2b; 8:30 Go; 8+nofi=uo 8.3 8+oz~u an; 8.35 p82 33 lawman A8; Nowano in; 2 3.0.; 2 Re; «amuse 2.0; or} A8; m: 3H; fun 8“; ~oon=mu A3; 3 Co; ~om=cu 8.0.; «03.00 83 oazmu RN; 2 3...; ONE 33...} Cb; ~03} 8w; 3%... 3.3 No}? 20.10300 83 one: 33 as} 2.3 as? 33 i 33 “02.x". :3 new :3 3 SN; no}? 3“; no}? A8; 9.3.9.. AR; was} 58¢ Nora 0:5 2.83an So; no}? 2...; cons... AS; 30 +2. a n n ncxaonnn 41.59.84." 0 O Q Q G Bu. )0 \/ o \5 .933. 285» 5:. .88: 8 28:32. .5. 5 :3 $.53. + 146 In the case of polyethers, the thermodynamics conclu- sions are more difficult to make. The rough estimation of complexation energy for Ni+ with TDE, 12-crown-4 and 15- crown-S will be greater than 126.32 kcal/mole. It is unclear how many M+-O interactions this figure reflects. Obviously, many more experiments have to be conducted for an understanding of metal ion-polyether interactions. Metal ion reactions with multifunctional molecules are some what an interface between molecular and bulk interactions. Although the explanation of product ions is somewhat speculative, this is the first attemp to carry out this kind of study. Other attempts still have to be tried on smaller molecules to determine other factors controlling the forma- tion and distribution of products. Section 5 is an attempt to do this. + + . + x , Cr(CO)x And N1(C(fix In Their Reactions With Ethers. IV. Comparison Of Fe(CO) Table 13 summarizes the ether reactions for Cr+, Fe+ and Ni+, (excluding successive reactions and complexation reactions). By consulting Table 6, Table 9 and Table 12, it is readily seen that, as the number of 00's on the metal center increases, the "reactivity" decreases and only sub- stitution reactions are observed. In the reaction of Fe+, a clear macrocyclic effect is observed, namely the number of products changes dramatically 147 305603 cogmxoamsoo and 30:80am :33de gamuooofim 3305 won moon * erJix we no .2 mm .8 2 .....H .8 2 we no 2 mm no n+2 §§ 0 \V\ o § o \Q 0 “Q o WAS? & o o R o o I& H o & o o I\\A o o mA8V= o o o o o o n a o o o o o o 0 £8? . o o o o o o n n N o o o o o 0 was: 0 o o o o o w w 9 o H o o o o wsvz o o H m m o HH m 0H m n H H o 0 +82 H m H m m o H. H m m nL-...H1.iH 2H rm; +2 0 g a \/o \/ “no [ammo . :oH pagodmm *éowpmcflynoou mo monopm «:03; 5 «H350 Hot: no... ugsflo 388$ 833% mo nBssz .mH «Ea. Table 13. ( cont'd) 148 15-crown-5 c) §§§§S§:§. Mm COO 12 -cmwn-li NE§§S§\\V§ 3:\ CrF/e///N/iCrF/e//N/1/ s \\\\\\\ ~.\\\\\\\\)‘ .\\\\\\\\\ o s MT-Cr Fe N/iCrFe NICrF/e Ni Reactant Ion AA VV AAA VVV 149 from linear polyethers to cyclic polyethers. THF and THP do not produce as many products as diethylether, due to their geometrical constrants, possibly leading to insufficient orbital overlap. Hence, Fe(CO)2+ is still reactive in the reaction with diethylether and unreactive in its reaction with THF and THP. However, in its reactions with polycyclic ethers, Fe+ starts to interact with multifunctional atoms to induce more products from the cavity center, exhibiting macrocyclic effect with lZ-crown-4. When the ring size increases, this effect is decreased slightly and we expect that the reactions may be more like those observed linear polyethers in the reactions with lB-crown-6 and 21-crown-7 with Fe+ inducing products more randomly, presumably due to its incapability of interacting with all functional atoms at a time. .It is of interest to note that in the reactions with small polycyclic ethers, the Fe(CO)x+ ions (x 3,2) are unreactive, possibly implying that geometrical restrictions are present, although the initial interaction could be strong. Moreover, in the 12-crown-4 and lS-crown-S systems, the addition complexes were not observed. Also, no substitu- tion reactions occurred, suggesting that Fe+ interacts with all oxygen atoms in the former and 3 or 4 oxygen atoms in the latter and the interaction must be very strong to under- go prompt fragmentations. It also implies that Fe+ induces reactions from the cavity center or very close to it. Con- sequently, as more CO's are present on Fe+, it can no longer 150 get close to the cavity center of lZ-crown-4. Note that we also see some ligand effects, in which CO can act as either a spectator or a participator to give new products which are not seen in the ligand-free metal ion's case, eSpecial- ly in the TDE reaction. The above model can also be applied to Cr+ reactions with the knowledge that Cr+ prefers to retain bonds to oxygen. Note that the formation of Cr(12-crown-4)+ may not be the result of the interaction of Cr+ with 4 oxygen atoms from the cavity center, because it will result in fragmenta- tion, due to its strong interaction with oxygen atoms. Instead, it could be the result of interaction with two oxygen atoms as in p-dioxane. Again, we have no evidence that Cr+ could interact with all 5 oxygen atoms in lS-crown- 5. The failure of Cr+ to form an addition complex with lS-crown-S could be due to the insufficient oribtal over- laps. Also note that CO ligands play an important role in the reaction of this metal with TDE (see Table 9). Since Ni+ has relatively weak bond energies to oxygen and alkyl groups, the interaction between it and ether oxygen atoms is weak. Thus, we don't see the macrocyclic effect in this case. From Table 12, it is seen that NiCO+ produces more products than Ni+. Ni+ "randomly" inter- acts with the oxygen atoms in TDE to give many products. However, when the reactant cyclic polyether is lZ-crown-4, it only produces 3 small molecular products, with another 151 product having one H2 elimination. It appears that Ni+ only interacts with orbitals in close proximity, possibly implying that the smaller size d-orbitals are used. Forma- tion of addition complex of Ni+ with TDE, 12-crown-4 and lS-crown-S may indicate that Ni+ is unable to interact strongly with many oxygen atoms because of its low bonding energy and that Ni+ can't efficiently overlap with orbitals wich are far from it. Another way to look at the product distributions is from Staley's bond dissociation energy studies for two ligands in the gas phase158'162. Staley found that metal ions are softer (based on HSAB theory) across the Periodic Table in the following order: H+, Li+, A1+ > Mn+ > Fear+ > Co+ = CpNi+ > N0+ > Ni+ > Cu+ They also found that the bonding distance of interact- ing center of ligands to the metal ions is increasing in the following order: + + + + < NO+ < A1 < Ni < Mn < L1+ < CpNi+ H Hence, we expect that Cr+ prefers to retain harder acids such as oxygen acids with higher O/C ratios than Fe+ does. Ni+ prefers to retain softer acids with lower 0/C ratio or simply alkenes. On the other hand, the shorter bonding distance will reflect a larger tendency to retain larger ligands and show a greater substituent effect. 152 Unfortunately, the experimental data is not sufficient yet to be used here. 5. Trends in First Row Transition Metal Ions In Gas Phase Reactions With Organic Molecules In this work, Co(C)3NO, Cr(CO)6, Ni(CO)4) Mo(C0)6 and W(CO)6 were obtained from Alfa products. Fe(CO)5 and cis- 2-pentene were obtained from Aldrich Chemical Company. l-hexene, 2-pentanone and sec-butylamine were obtained from Chem Service. Iso-propyl chloride and propane gases were obtained from Matheson Gas Products Inc. Methyl iodide was obtained from MCB manufacturing Chemical Co., Inc. All compounds were used without further purification except cis-Z-pentene which was distilled for 8 hours before use. All compounds were subject to standard freeze-pump-thaw cycles before use. A number of papers have appeared in the literature on the chemistry of metal ions with organic molecules. Most of these papers have discussed one metal. No attempt has been made to discuss the reasons why different first row transition metals behave so differently with simple organic molecules. The decision was made to "target" some organic molecules whose reactivity may provide insights into the differences of metal ions. A number of factors which contribute to the chemistry observed in gas phase organomet- tallic reactions should include: 153 l. The number of available low lying empty orbitals on the metal ions available for reaction. (For example, Hg+ (51d10) can have two sp hydrid orbitals but Li(sz) cannot.) 2. Orientation and size of available orbitals of metal ions. 3. "Compatibility" of orbitals. For an insertion pro- cess to occur, bond lengths and bond angles in the initial metal-ligand complex must be compatible. 4. Thermodynamics (e.g., heats of formation, bond strength, promotion energy of metal ions (redistribution of electronic configuration)). 5. Orbital symmetry of the intermediates. 6. Electronic states of metal ions participating the reactions. 7. Kinetic factor (to be detected in ICR, the rate 9 '11 cm3/molecule/s). constant has to be in 10' - 10 The initial strategy in this work was as follows: a. Test if a metal ion inserts into the CH3-I bond. b. If a metal ion inserts, then test if there is B-H shift across the metal center by studying reactions with i-C3H7Cl. c. Test if 5 or 6 membered ring intermediates are preferred by investigating metal ion/molecule reactions with l-hexene, cis-Z—pentene and 2-pentanone. 154 d. Test if the metal ions react with nonpolar compounds such as propane, to estimate D(M+-alkyl). e. Test if the metal ions react with amines to determine the bond strength of the M+-NR2 bond. These experiments were performed with a number of 1;; row transition metal ions. 1. Reactions With Propane (C3fl8) A. Results 1. Fe(CO)x+ reactions With propane Ions formed as products of ion-molecule reactions in a mixture of Fe(CO)5 and C3H8 are listed below, with their precursors as identified by double resonance. gig stoichiometry preCursor(s) + + 84 Fe(C2H4) Fe 98 Fe(C3H5)+ Fe+, Feco+ + + 100 Fe(C3H8) FeCO 128 Fec0(c3H8)+ Fe(CO)2+ 153 Fe(c0)2c3H5+ c3115+ + + + + 169 Fe(CO)4H c2114 , c2115 , C3H8 and Fe(c0)3c2H5+ + + + + 181 Fe(CO)3C3H5 c2115 , C3H5 , C3H7 + + + 183 Fe(CO)3C3H7 c2115 , c3117 + + + 197 Fe(CO)5H CZHS , c3117 and Fe(CO)4C2H5+ 155 Note that CZHS+ , C3H5+ and C3H7 are reactive organic +168 species, especially C3H5 , which reacts with netural Fe(CO)5, displacing two or three CD'S. 2. Co(CO)x(NO)y+ reactions with propane. Ions formed as products of ion-molecule reactions in a mixture of Co(CO)3NO and C3H8 are listed below, with their precursors as identified by double resonance. mi; stoichiometry .precursor(s) 87 Co(C2H4)+ Co+ 101 Co(c3H6)+ 00+, Coco+ 103 Co(C3H8)+ Coco+ 118 CoNOC2H5 CZHS: 13o CoNOC3H5 c3115 131 Coc0(c3H8)+ Co(c0)2+ 133 CoN0(c3H8)+ Coho+ 145 Co(CO)2NOH+ C2H5: , Coco+ 158 CoCONOC3H5+ C3H5: 16O CoCONOC3H7+ C3H7: 174 Co(CO)3NOH+ c2”5: , Coco+ 188 Co(c012Noc3H7+ C3H7 3. Ni(CO)x+ reactionsfiwith propane Ions formed as products of ion-molecule reactions in a mixture of Ni(CO)4 and C3H8 are listed below, with their precursors as identified by double resonance. ng 86 100 102 115 127 130 143 145 155 157 159 186 156 stoichiometry . + N1(C2H4) . + N1(C3H6) . + N1C3H8) Ni(CO)2H+ + 5 N1c0(c3H8)+ Ni(c3H8)(c3H5)+ Ni(CO)3H+ Ni(c0)2c3H5+ N1(c0)2c3H7+ N1c0(c3H8(c2H5)+ N1(c0)3c3H8+ NiC0C3H _precursor(s) Ni+ Ni+, Nico+ Nico+ c2”5+ 3”s+ Ni(CO)2+, c3118+ c H * Ni+, Nico+ C 3 8 ’ CZH 1, N11, Nico+ + c3”5 + + . + c3115 , C3H , N1CO(C3H8) + . + C3H7 , N1CO(C3H8) N1+ The ion at m/e 100 has a very small peak intensity com- pared with that of the comparable product for Co+, but it is similar to that observed for Fe+. Cr(CO)x+ reactions with propane Ions formed as products of ion-molecule reactions in a mixture of Cr(C0)6 and C3H8 are listed below, with their precursors as identified by double resonance. mZe 80 96 124 stoichiometry Cr(c2H4)+ Cr(C3H8)+ Crc0(c3H8)+ precursor(s) Cr+, orco+ orco+ CrCO+ 157 5. Mo(CO)X+ reactions with propane. Ions formed as products of ion-molecule reactions in a mixture of Mo(CO)6 and C3H8 are listed below, with their precursors as identified by double resonance. gig stoichiometry (precursor(s) 162 Moc0(c3H6)+ MoCO+, Mo(00)2+ 19o Mo(CO)2C 3 H6+ Mo(CO)2+, Ho(c0)3+ 218 No(CO)3C 3 H6+ Mo(CO)3+ 8. Discussion Table 14 summarizes the reactions with propane for all five metal ions together with product distributions. All the product ions seen can be explained in terms of the mechanism pr0posed by Beauchamp94'96. M+ +,/”\\-—r———>11—1+—<<5-—————+ “\‘Mi—l ———%>M-— shift H//M U\ '————>H——M+ jiji///zH M = Fe, Co, Ni 5 shift - e shift "\Mt—H—ent-H :_//’ ""9 “‘3‘“ W / CH3 M Fe, Co, Ni, Cr 158 No 9“. a.” n. Hm. 3. Ha. 0H a." H m. oH “we 0;” “5. n. N H m H a w H ,N H "W .8 Hz 8 ea 8. 3. HN. o.H om. Hm. mu. .5 Hz 8 9m Ammmuvozou Tll m=mo + .38 + H: + wamovaoowzil H: + 8 + HmzanHiBwsfl 8w + AmznovmthSwzj 8+ Amenszflswfl 38: N: + momma: $8 + kamHuEUI mama + +2 Hanna 2888 5:. «5308: . 8H 838. 159 Thus, Fe+, Co+ and Ni+ exhibit very similar reactivity 164. It is of interest to note that Fe+, Ni+ with alkanes and Cr+ have stronger preference to inserting into C-C bonds to eliminate smaller alkanes. Mo+, never induced any reac- tions from propane unless it has ligands attached. CO ligands on Mo+ might affect the bonding abilities of other bonding orbitals. It is not surprising then since Mo+ is a second row transition metal, which has larger d orbitals (60% size of 55 orbital) to make the bonds using the 4 dz2 orbital, which in turn makes the second bond possible by use of other d orbitals (the promotion energy form 4dn to 55]- 4dn"l to make a 55 orbital available for bonding123 is about 72 kcal/mole). This contrasts with first row transi- tion metals which bond almost exclusively using the 4s orbital. The 3d orbitals are only 30% the size of the 4s orbital and are tightly bound to the nucleus. (It is pos- sible to mix 45 with 4p to make sp hybrid orbitals). II. Reactions With Iodomethane (CH3I) A. Results 1. Fe(CO)x+ reactions with CH3; Ions formed as products of ion-molecule reactions in a mixture of Fe(CO)5 and CH3I are listed below, with their precursors as identified by double resonance. 160 m/e stoichiometry precursor(s) 183 FeI+ Fe+ 198 FeCH3I+ Feco+ 226 FeCOCH3I+ Fe+, FeCO+, Fe(00)2+ + + + + 254 Fe(CO)2CH3I Fe(CO)2 , I , Fe(CO)3 , + + CH3I , Fe(c0)4 282 Fe(CO)3CH3I+ Fe+, Fe(CO)3+, CH3I+, Fe(CO)4+ 2. Co(CO)x(NO)y+ reactions with CH3; Ions formed as products of ion-molecule reactions in a mixture of Co(CO)3NO and CH3Iareelisted below, with their precursors as identified by double resonance. ELE stoichiometry precursor(s) 186 CoI+ co+ 201 CoCH3I+ Coco+ 229 CoCOCH3I+ CoCO+, Co(c0)2+ 231 CoNOCH3I+ CoCO+, CoCONO+, Co(CO)2NO+ 259 CoCONOCH31+ CH31+, Co(CO)2NO+, Co(c0)3Ho+ 287 Co(CO)2NOCH31+ Co(CO)3NO+ 3. Ni(CO)x+ reactions with CH3; Ions formed as products of ion-molecule reactions in a mixture of Ni(CO)4 and CH3I are listed below, with their precursors as identified by double resonance. 161 gig stoichiometry precursor(s) 185 Nil+ Ni+, Nico+ 200 Ni5H3I+ Ni+, NiCO+, Ni(c0)2+ 213 Nic01+ NiCO+, 1+ 228 NiCOCH3I+ NiCO+, Ni(CO)2+, Ni(50)3+ 255 Ni(CO)ZCH3I+ Ni(CO)2+, 1+, Ni(CO)3+, Ni(50)4+ 284 Hi(50)35H3I+ Ni(CO)3+, Ni(CO)4+ 4. Cr(CO)x+ reactions with CH3; Ions formed as products of ion-molecule reactions in a mixture of Cr(CO)6 and CH3I are listed below, with their pre- cursors as identified by double resonance. m/e stoichiometry, precursor(s) 179 CrI+ Cr+ 194 CrCH31+ CrCO+, Cr(CO)2+, 5H31+ 5. Mo(50)x+ reactions with CH3; Ions formed as products of ion-molecule reactions in a mixture of Mo(CO)6 and CH3I are listed below, with their precursors as identified by double resonance. 162 mle stoichiometry precursor(s) 219 MoI+ Ho+ 234 MoICH3+ Mo+, MoCO+, 1+, Mo(50)3+ 5. A1(5H3)_.,+ reactions with CH3; Ions formed as products of ion-molecule reactions in a mixture of A1(CH3)3 and CH3I are listed below, with their precursors as identified by double resonance. m/e stoichiometry precursor(s) + + + AlMe3 , CHBI * 199 AlIMe3 * Me denotes a methyl group 163 B. Discussion Table 15 summarizes all the reactions with iodomethane for all metal ions in this owrk, together with their product distributions. It appears that all metal ions insert into the polar bond (C-I in this case), followed by fragmentation to retain the iodine atom. Among them, Ni+ seems to have the strong- est interaction energy with CH3I to induce more reactions. For example, Ni+ is the only metal ion that produced MCOI+ and MI+ from NiCO+. Mo+ follows almost the same reaction pattern as that of Cr+. There are two possible mechanisms leading to the forma- tion of Mi+: (1)197,198 M+ + RI ——> R-M+-I ——>MI+ (2)85,98,169'170 X H M+ + IM'I. \ + e x H ———-—>MX + d J. Allison and D.P. Ridge197’198 reported a number of first row transition metal ions reactions with alkyl halides and found the following results: 164 $6 + MH£H 56.5 kcal/mole) than Mo+. Therefore, it seems that thermodynamics only can explain all the reactions presented to this point. The mechanism can be summarized below: H I M+ + >_c1 __, >_Mt_ c1 ——>J--M+——Cl HJH-H” + HCl l \\\\3 M+=Mo M+=Fe+,Co+,Ni+ + + M(C3H6) MCl Cr+,MO+ M+=Fe,Co,Ni,Cr M+=Cr ' ' The formation of MC3H7+ for Fe+, 00+, Ni+ and Cr+ implies that 0(M-53H7)+ > 82.03 kcal/mole. 168 o. H... .Hou :8. .. +2... ...»...HozHu... \HHS... o .. N “HquHuvnTHHquHSouxT way... 0.. 0.. 8.. ...N Essay M98... 0 .. H+HoH .. HN.:chH 8 .....Ho... 1.8... u UCO‘UUHUK U’dQOUUU-um HH .... +HH HHuHsHBHesHSczou ...... ..HH .HHHcHzHuzcsH $9.8 HHHquHo +HkuH=Huczou H... 8 + H8 HxHu.ozHux.8.o. HH. 8.. + 8 + HmzHSczHuchSoo mszcSou HH. ....H.. + MuszcuvzeJ H... e.H NH. 8N .. MH8H..HH.NL.H8.T ..H. .3. ..H. ..H. NH. NH. 8 + «HuHxHuvHuxH8.:1l .... HH. ..H. .... .... 8.. .+ 3.6.5821 HH. 8. NH. HN. HN. H... H..H H... ..H. 8 + 8.. .. «HzHuvHuwHouvkaHSw: N H .H H N H N H N H N H H: NH. monHurTl HH. H... HH. HH. .8 + H..H ......I ..N. H... HaHu + 86.11 HH. H... H... 8. H... 8.. +HH=Huwzil o: mum. aHz a a .II. .H :5 + 55118:. H": ++x AHqunu. ucHuoH:uHHmcumoo~ :NHa acoHuuuom .Ofl 0.23. 169 IV. Reactions w1th Cis-Z-Pentene (C5H10) A. Results + . . . l. Fe(CO)x react1ons w1th c1s-2-Csfl1O Ions formed as products of ion-molecule reactions in a mixture of Fe(CO)5 and (25H10 are listed below, with their precursors as identified by double resonance. ml£_ stoichiometry precursor(s) 95 Fe(C3H4)+ Fe+, Feco+ 98 Fe(c3H6)+ Fe+, FeCO+ 110 Fe(C4H6)+ Fe+, Feco+ 124 Fe(csHB)+ Fe+, Feco+ 125 Fe(c5Hm)+ I FeCO+, Fe(50)2+ 154 Fe50(55Hm)+ FeCO+, Fe(CO)2+, . Fe(CO)3+ 182 Fe(CO)2(C5H10)+ C3H6+, Fe, Fe(5013+ and Fe(50)353H6+ Fe(CO)4+ 210 Fe(CO)3(C5H10)+ Fe(CO)4+, Fe(CO)5+ 238 Fe(CO)2(C5H]0)+ Fe+, FeC0+, Fe2(CO)4+ Note that m/e 182 is a mixture of Fe(CO)2(CSH10)+ and + Fe(CO)3(C3H6) . + . . . 2. C0(CO)X(NO)y react1ons w1th c1s-2-Csfl10 Ions formed as products of ion-molecule reactions in a mixture of Co(C0)3NO and 05H10 are listed below, with their precursors as identified by double resonance. 170 m/e stoichiometry precursor(s) 99 Co(C3H4)+ 5o1, CSHIJ , Coco1 101 5o(c3H6)1 5o1, Coco1 113 C0(C4H6)+ 5o1, Coco1 127 C0(C5H8)+ 5o1, Coco1 129 Co(C5H]0)+ Co1, Coco1, 5o(50)21 130 CoN053H51 C3H5: , Coc01, Co(50)21 155 5o50N053H51 c3H31, c3HJ 157 CoCOCSHlJ Coco1, 5o(50)21 158 CoCON0C3H J c3HJ , Co(CO)2NO+ 159 C°N0C5H10+ 5o50N01, 5o(50)2N01 187 C0C0N0C5H10+ c3HJ , 05H101, 5o(50)2N01, Co(50)3No1 195 Co(C5H8)2+ Coco1, Co(50)21 199 Co(C5H]0)2+ Coco1, Co(50)21 215 Co(CO)2NOCSH10+ CSHIJ 215 C02CO(C5H10)+ Co1, Co250N01 218 5o250N0(c3H6)1 0o1, 5o(c3H6)1 229 CoNO(C5H]0)2+ Co1, C5H10+, Coco1, Co(50)2N01, Co(50)3H01 230 Co(c4H6)(5o50N0)1 5o1, Coco1, 5o(c4H5)1 244 Co250N0(55H8)1 Co1, Coco1, Co2(50)2N01 245 C02C0N0(C5H]0)+ 5o1, Coco1 Note that 03H5+, C3H3+ and C3H 6+ from organic fragmen- tations are reactive with neutral Co(CO)3NO to form m/e 130, l7l l56, 158 and m/e l87. 3. Ni(C0)x+ reactions with cis-2-pentene Ions formed as products of ion-molecule reactions in a mixture of Ni(C0)4 and C5H10 are listed below, with their precursors as identified by double resonance. m1; stoichiometry precursorLs) 98 Ni(c3H4)1 N11 100 N1(c3H6)1 N11, N1c01 112 N1(c4H6)1 N11, Nico1 l26 Ni(C5H8)+ N11, Nico1 128 Ni(C5H]0)+ N11, Nico1, Ni(c0)21 155 NiCO(C5H]0)+ Ni(C0)2+, N1(c0)31 17o Ni(c0)41 05H10+ 184 Ni(CO)2(C5H]0)+ C5H10+, Ni(C0)3, Ni(c0)41 198 Ni(C5H]0)2+ Ni(80)21, Ni(c0)31 212 N12C0(C5H8)+ N11, Nico1, N12(c0)21 Note that molecular ion of cis-Z-pentene reacts by charge exchange with neutral Ni(C0)4 to give m/e l70, N1(c0)41. 4. Cr(C0)x+ reactions with cis-2-pentene Ions formed as products of ion-molecule reactions in a mixture of Cr(CO)6 and CSH10 are listed below, with their precursors as identified by double resonance. 172 gig stoichiometry precursorfsl + + 120 Cr(C5H8) orco1, Cr(C0)2 + + + l22 Cr(C5H10) CrCO , Cr(C0)2 + + + 150 CrC0(C5H]O) Cr(C0)2 , Cr(CO)3 5. Mo(Cle+ reactions with cis-Z-pentene Ions formed as products of ion-moleCule reactions in a mixture of Mo(CO)6 and C5H10 are listed below, with their precursors as identified by double resonance. gig stoichiometry precursor(s) 160 Mo(csH8)1 M01, Moco1 l62 Mo(C5H]0)+ M01, Hoco1, Ho(c0)21 188 MoCO(c5H8)1 Mo(CO)2+ 19o MoC0(C5H10)+ Mo(c0)21 215 Mo(C0)2(C5H8)+ Ho(c0)31 B. Discussion Table 17 summarizes all the reactions of cis-Z-pentene with metal and metal-containing ions in this work. It is interesting to note that Cr+ does not react with cis-Z-pentene; Mo+ only eliminates H2. When CO ligand(s) are attached, an elimination of H2 is seen for both metal ions. It is also interesting to note that Cr+ reacts with C3H8 to eliminate CH4. There thus appears to be a chain l7l length effect or the double bond in alkenes prevents Cr+ from inserting into C-C bonds. From Table l4 and l8, it can 173 0.. +Ao£nuv8~8 Tll+28~8¢l| .8 8. NziymovBBNoo .88. 8 + AoannuvBBNoo 0.. +B~A8v~8 1.8 mgr. 8 .wonxnorTI.AomMoV§TlmA8r ~=++Aw=m38~z+ll MABVFHR! . tnowMSNABFiI “:85 +2 . 33303.» 253825 ma. oz" 3. wa. an. 8. no. no. MN. ea. c..." R. MN. 0...“ 5. ma . ma. ow. no. a. flu. ow. ON. 3. fi. no. 8m Janene VB?» 8! 8&3? V8148: U. 248 v. 82858.» 8:1 . 8 ++Aouznu Va...» 8! ... norms: 8:1 8n. +N: +M£MSN1HA8¥1 8+ n: ésmsfiiart 8 + :5 J33 V?» 8!... 8 Emu". Keane VTHAB Shim. ...AiuvauxABHU llwer aw. 3N. ma. agency 1 Nu ++Am=m8x+l so ++Aw=ao¥11 :MNU .o 50:95:14 can“. + .AaxnurtlorMoéé +.r. a: “3....an .=8=&.~130 5:5 98332“ :0de 174 be seen that Mo+ does not insert into a C-C bond in either alkanes or alkenes. The mechanism used to interpret the formation of all product ions can be summarized as follows (similar to those proposed by Beauchamp97): + + H— + all 110 H\ M —"M" _]———I—_; Hshift 3+7": sT_’1rtH /”> __’”(C 5118) Janync L141: Fe, oo.N1,Mo| fishiflt H \ _.._s -m \+/ + M? H ”275% m. /“ ———’“(C4“6) EM-Fe'“-"il 5741/» m3 /’ x + ”a H\+ _[——>M( can“) — -———>M( C5116)“ hE=Feu3n§§ H ‘—-) «03411,? lLr-F'BJOJTiI + | \M/l ( “H Note that these mechanisms also show a proposed pathway leading to the formation of M(C3H6)+ and M(C3H4)+, in which M+ induces the isomerization of cis-Z-pentene to l-pentene as a first step. Similar isomerization steps have been pre- viously reported91’97. 175 V. Reactions with l-Hexene (C5fl121 A. Results + .. . l. Fe(CO)x react1ons w1th l-Csfl12 Ions formed as products of ion-molecule reactions in a mixture of Fe(CO)5 and C6H12 are listed below, with their precursors as identifiedby double resonance. ELE stoichiometry precursor(sl + + 84 FeC2H4 C2H4 98 Fe(c3H6)1 Fe1, Feco1 11o Fe(c4H6)1 Fe1 + + + T40 Fe(C6H12) FeCO , Fe(CO)2 + + 153 (C3H5)Fe(CO)2 C3H5 + + l68 FeCO(C6H]2) Fe(CO)3 + + 181 Fe(CO)3(C3H5) C3H5 195 Fe(CO)2(C6H]2)+ Fe1, Feco1, Fe(c0)41 224 Fe(C0)3(C6H]2)+ Fe1, Feco1, Fe(c0)21, Fe(CO)5+ 250 Fe2(CO)2(C6H]0)+ Fe1, Feco1 252 Fe2(CO)2(CGH]2)+ Fe1, Feco1, Fe(c0)21 266 Fe2(CO)3(C5H]O)+ Fe1, Feco1 278 Fe2(CO)3(C6H]0)+ Feco1 + + + + 280 Fe2(C0)3(C6H12) FeCO , Fe(c0)2 , Fe(CO)3 Note that C3H5+ reacts with Fe(CO)5 neutral to displace two or three CO's to form m/e 18l and 153 respectively. The formation of m/e 266 is hard to explain. Yet, this peak is reproducible. 176 2. Co(C01x(N0)y+ reactions with l-C =0 l H. 5 12 Ions formed as products of ion-molecule reactions in a mixture of Co(C0)N0 and (26H12 are listed below, with their precursors as identified by doubTe resonance. m1; stoichiometry precursorLiL 101 Co(c3H6)1 Co1, Coco1 143 Co(C6H]2)+ Co1, Coco1, Co(c0)21 158 Coconoc3H51 c3H51 172 CoC0N0C4H7+ C4H7+ 173 Coc0N0(c4H8)1 C4H 1, C6H12+, + CoN0(C6H]2)+ Co(80)2N01 201 CoCON0(C6H12)+ Co(c0)3)No1 Again, organic fragments C3H5+ and C4H7+ reacts with Co(C0)3N0 neutral to form m/e l58 and 172 respectively. m/e 173 is the mixture of Cocouoc4H81, and CoN0(C6H]2)+. 3. Ni(Cle+ reactions with l-C5fl12 Ions formed as products of ion-molecule reactions in a mixture of Ni(C0)4 and C6H12 are listed below, with their precursors as identified by double resonance. 177 gig stoichiometry precursor(sl 86 Ni(c2H4)1 Ni1 100 Ni(c3H6)1 Ni1, Nico1 112 Ni(c4H6)1 Ni', Nico 127 Ni80(c3H5)1 c3H51 142 Ni(CGH]2)+ Ni1, Nico1, Hi(c0)21, Ni(c3H6)1 155 Ni(c0)2(c3H5)1 c3H51 17o Ni(CO)2(C4H8)+ C4H 1, CGHIZ, Ni(C0)2+, + NiC0(C6H]2)+ Ni(c0)31 198 Ni(C0)3(C4H8)+ c4H81, Ni(c0)31, Ni(c0)41 + Ni(CO)2(C6H]2)+ 4. Cr(C0)x+ reactions with l-Csfl12 Ions formed as products of ion-molecule reactions in a mixture of Cr(C0)6 and C6H12 are listed below, with their precursors as identified by double resonance. gig_ stoichiometry precursor(s) + + + + l34 Cr(CGH]0) Cr , CrCO , Cr(CO)2 + + + + l36 Cr(C6H]2) Cr , CrCO , Cr(C0)2 + + + 164 CrC0(CGH12) Cr(CO)2 , Cr(CO)3 +, . . ~ 5. Mo(CO)x react1ons w1th l-Csfl12 Ions formed as products of ion-molecule reactions in a mixture of Mo(CO)6 and C6H12 are listed below, with their precursors as identified by double resonance. 178 m/e stoichiometry precursoris) l72 Mo(C6H8)+ Mo+, MoC0+, Mo(C0)2+ + + 200 MoCO(C6H8) Mo(c0)2 . Ho(c0)31 204 MoCO(C6H]2)+ Mo(c0)21, Mo(c0)31 228 Mo(c0)2(c6H8)1 Ho(c0)31 + 232 Mo(C0)2(C6H Mo(C0)3 + 12) B. Discussion Table l8 summarizes the reactions of l-hexene with Fe+, Co+, Ni+, Cr+, and Mo+ together with their product distribu- tions. Basically, the reactions of l-hexene with these transi- tion metal ion can be explained similarly to that proposed 97 by Beauchamp , as follows: 141+ C6H12—>/@ ML,J +\\ -—_+H\;'q1/1._, l-E-Q alkfil V l + ethnic M(2 H Shift [i=é1831333m 3\@ B-aligl , H\+/\ + fl-H l{\144‘) F‘1n(czntt) Shift H/ \/ 1HH(04H6)1|Mf=Fe;N1| H\+ l79 m-fi$8~§ T1 Naawo fifiaxemel M «Axon E 8 WET. + we: “a. 8n +nm8v~211£8v§ NABVQ a: + «cameovuA8m81 8:91. ..A fixooxzihrwn. m2: . 8~ + “3883831 mos. . onnu + Moznuszfizoo + +2. 8n + “fivamgvfiilu 3&4 + MABETnABE +3.. . 23308." 233803 c.” o..." 8~+m-=ou§~148¥411mz£8v= 8. 3 8m +~=~ ++Ao=o3~1xA8rJ S. 9. .3 no. 8+~=~ lozouvnixABvxTi 8. . 8...: .woréjarii S. oA 8. no. on. 8. o4 o4 8N ...AmfoovfluABET .8. R. 2. a. 3. a. a. “N. n... a. Bramfisixaril 2. kNfizmovaBET. a. . 8 +83 $1.3?ng 3. mm. 3. 8 +o=no 4%:n8848vzilllxm8v: nNHnNH2n~a~H-~HHx o.“ ~=~ + oncurj 8. u: + Aonzouvfll .3. «A. 3. Aun=wo¥1|1 2. 8. mam. 182.6%] am. am. co. can". iwxnurill 2. 3...”. JazwsxileMzooA ++= £ .5 =— 8 o.— A «found v 885:4 5:. 338$ .... on See 180 Note that Cr+ and Mo+ prefers to insert into C-H bonds to eliminate H2. Mo+ can insert into C-H bonds to eliminate two hydrogen molecules. The intermediates having C3H5+ or four n electrons distributed on three carbons are reasonable, because they are good n-donor ligands and exist as the stable fragments in the mass spectrometer587’168. VI. Reactions with Z-pentanone (Z'Csfliogl A. Results + . . l. FeiCO)x react1ons w1th 2-C5fl1og Ions formed as formed as products of ion-molecule reac- tions in a mixture of Fe(CO)5 and C5H100 are listed below, with their precursors as identified by double resonance. gig stoichiometry grecursor(s) 84 Fec2H41 C2H4+ 114 Fe(03H60)1 Fe1, Feco1 14o Fe(c5H80)1 Fe1, Feco1 142 Fe(C5H100)+ Fe1, Feco1, Fe(80)21 17o FeCO(C5H100)+ Fe(00)21, Fe(00)31 181 Fe(c0)3c3H51 C3H 1, c3H71 l83 Fe(80)3c3H71 C3H7+ 197 Fe(c0)5H1 c3H71 198 Fe(C0)2(C5H]00)+ Fe(c0)31, Fe(c0)41 211 Fe(C0)4C3H7+ c3H71, C5H100+ + + + 225 Fe(C0)3(CSH]00) Fe(CO)4 , Fe(C0)5 + + + 181 m/e stoichiometry precursor(s) 0)1. 0)1 FeC0(C5H10 Fe(CO)2(C5H10 Note that C3H5+ and C3H7+ react with Fe(CO)5 neutral to yield m/e l8] and 183 and 2ll respectively. Also, C3H7+ can protonate Fe(CO)5 to produce Fe(CO)5H+. 2. ngtQ)_LNO) + reactions with Z-CCH O x y 0—10— Ions formed as products of ion-molecule reactions in a mixture of Co(CO)3NO and C5H100 are listed below, with their precursors as identified by double resonance. gig stoichiometry precursor(sl 117 Co(c3H60)1 Co1, Coco1 143 Co(c5H80)1 Co1, Coco1 + 145 Co(C5H]00)+ Co1, CoCO , Co(c0)21 + + l58 CoCONOC3H5 - C3H5 + + + l73 CoC0(C5H]00) CoCO , Co(C0)2 175 CoNO(C5H]00)+ Cocono1, Co(c0)2No1 + + + 203 CoC0N0(C5H]00) 05H100 , CoCONO , Co(c0)2Ho1, co(c0)3N01 + + 2l6 CoN0(C5H]0)(C3H5) C3H5 + + 128 CoNO(C5H]00)(C3H7) C3H7 + + + + 229 Co(CSH80)(CSH100) Co , coco , Co(C5H80) + + + + 23l Co(CSH]00)2 Co , CoCO ,~Co(C0)2 , Cocono1, Co(c0)2No1, co(c0)3N01, C5H100+ 182 Again, C3H5+ reacts with Co(CO)3NO netural to produce m/e l58 and m/e 2l6. C3H7+ reacts with Co(C0)3N0 to pro- duce m/e 2l8. 3. Ni(C0)x+ reactions with 2-C5fllog Ions formed as products of ion-molecule reactions in a mixture of Ni(C0)4 and C5H100 are listed below, with their precursors as identified by double resonance. gig stoichiometry Erecursor(s) 1oo Ni(c3H6)1 Ni1, Nico1 102 Hi(c2H40)1 Ni1, Nico1 ll6 Ni(c3H60)1 Ni1, Nico1 144 Ni(C5H]00)+ Ni1, Nico1, Ni(c0)21 155 Ni(c0)2c3H51 c3H51 157 Ni(c0)2c3H71 c3H71 172 NiC0(C5H]00)+ Ni(c0)21, Ni(c0)31 23o Ni2C0(C5H]OO)+ Hico1, Ni(c0)21, Ni(c0)31, Ni(C5H]OO)+, Hi(c0)41. NiC0(CSH100)+ . . + + . . S1milarly, CBHS and C3H7 react w1th N1(CO)4 neutral togive m/e l55 and m/e 157 respectively to displace 2 CD'S. 4. Cr(c0)x1 regctions with 2-c5glog Ions formed as products of ion-molecule reactions in a mixture of Cr(C0)6 and C5H100 are listed below, with their precursors as identified by double resonance. 183 gig stoichiometry grecursor(s) + + + l38 Cr(C5H]OO) Cr , CrCO + + + l66 CrCO(CSH100) Cr(CO)2 , Cr(C0)3 + + 224 Cr(C5H100)2 CrCO , Cr(C0)2+, or(c0)31 5. Mo(C0)x+ reactions with 2-csglog Ions formed as products of ion-molecule reactions in a mixture of Mo(CO)6 and CSHlDO are listed below, with their precursors as identified by double resonance. mie stoichiometry precursor(s) + + l74 Mo(C5H60) M0 175 Ho(c5H80)1 Ho1, Hoco1 + + + + 178 Mo(C5H100) Mo", MoCO , Mo(C0)2 204 Moc0(c5H80)1 Ho(c0)21, Mo(c0)31 + + + 232 M0(C0)2(C5H80) MO(C0)3 , M0(C0)4 B. Discussion Table 19 summarizes the Fe1, Co1, Ni1, Cr1 and Mo1 reactions with 2-pentanone, together with their product dis- tributions. 1 The formation of all reaction products can be explained 87 et. al. very similarly to the mechanism proposed by Freiser as shown in Scheme XXIX. Note that Fe+, Co+, Ni+, Cr+ and Mo+ can form stable addition products, M(C5H100)+. Hi1 and Cr1 failed to elimi- nate H2. However, Mo+ has extra energy to eliminate one “- u u +38 .2 3ofl88~£ll£8xz +moo~=n38§+|ooa=mu + NAB: .9 fi. Aoo~=n38~leooa=m8 + wh8v~=a|£8v= 38:. uoofmuxomchBTl oofizno dam—MES] can—Mu + 8 + ~23: . + 493850 . YNfievs . woofnovaxll comm”. + moorMSQoo .loofnu + .888 £85 . 0 800. Acommgootll corn". loormuvoofloouzmo + o .8 4 4 800 . ‘ . 133.qu 2; 388m 4 c; on. i. a.” 11 2.. B. B. o; oA oA me. an. no. 3. 00H “0. no. no. 2. mm. .HH. kw. o; 8. ma. 3. 4A oorMUVExABE 8w AoofmfiBWuABrMI 8 ¢ coranBAJBr +938 V: 8 Exam +moaz~37ufi8rol .AoornovraarT 8N +Moofm3~148rtl 8 cmooaxnuvaiuABIT 8~ .m: documoanABril 8 .N: éomsmovanuaerol 8 +a=~u JawznovfliuABrI 8 +oa=~u +vo=noVaunA8rT 3. nn. 3. no. mm. 3. no. an . ~a. wane ies—mugs] Carma 9 AmanoYT Aces—MOYI ~=~ zoo—Moral. u: .. Sonnet?! ammo o Aowanoroll common.“ + z ... o «I 8 p. 82.318 885a£-~ 5S, .8338 m. and. 185 fig .... +30%"; : Tam“. (2,18 a m mun-+4100 m fund : .85. E mam... n8 Aoznov z AoamNuV z fl\+zloo + + «2.00.9muz fl 8% al..“ m .M 3 z a A114: 18 \— Alum/3 ._1+:ti. /\/+: CODhMU 2.38 a: . :25 £48.27: e258- m. Aomznov z Aimwflli ....Hmnm TIQIQ 8:03."on /\/=\A|I/\J.\ + .w + +210 :\0 O ‘u x _ xx oszum 186 more Hz to form Mo(C5H60)+, implying that Mo+ has the strongest tendency to insert into a C-H bond. It is of interest also to note that Ni+ is the only metal which can insert into a carbonyl-carbon bond, producing Ni(C2H40)+ 87 also reported to see FeCO+ and and Ni(CBH6)+. Freiser Fe(C2H4)+ in a trace amount (< l%) in the reaction of Fe+ with 2-epntanone, which is not observed in our experiment. Another trace amount (6%) of the product ion, Fe(C4H8)+ reported by Freiser is not observed in our experiment. VII. Reactions with sec-Bugylamine (s-BUNHZ, c451rg1 A. Results l. Fe(001x+ reactions with sec-butylamine (s-BUNH2i_£4fl11gi Ions formed as products of ion-molecule reactions in a mixture of Fe(CO)5 and sBUNH2 are listed below, with their precursors as identified by double resonance. gig. stoichiometry grecursorisl_ 87 Fe(CHBNH2)+ Fe1, Feco1 98 Fe(c3H6)1 Feco1 11o Fe(c4H6)1 Fe1 113 Fe(c3H7N)1 Fe1, Feco1 114 Fe(C3H6NH2)+ Fe1, Feco1 125 Fe(c4H7N)1 Fe(c0)21, Fe(c0)31 127 Fe(c4H7NH2)1 Fe1, Feco1 128 FeCO(c2H6N)1 Feco1 Dis 129 157 185 202 stoichiometry + Fe(C4H9NH2) + FeC0(C4H9NH2) Fe(50)2(c4H9NH2)1 + Fe(C4H9NH2)2 187 ggecursor(s) Feco1, Fe(C0)2+ Fe(c0)21, Fe(c0)31, Fe(c0)41 Fe(c0)41, Fe(co)5+ Feco1, Fe(c0)21, Fe(c0)31, Fe(CO)4+ 2. Co(C0)x(NOly+ reactions with sec-bugylamine sBUNHz, c4511gl Ions formed as products of ion-molecule reactions in a mixture of Co(CO)3N0 and sBUNH2 are listed below, with their precursors as identified by double resonance. mie 90 100 102 104 ll3 116 128 130 132 160 162 stoichiometry Co(CH3NH2)1 Co(c3H5)1 Co(c2H5NH2)+ Co(c2H5NH2)1 Co(c4H6)1 Co(c3H7N)1 Co(c4H7N)1 Co(C4H7NH2)+ Co(c4H9NH2)1 Coc0(c4H9NH2)1 CoN0(c4H9NH2)1 precursorisl Co+ CoC0+ Co1, Coco1 Co1, Coco1 Co+ Co1, Coco1 Co+ Co1, Coco1 co1, Coco1, Co(c0)21 Coco1, Co(c0)21 Coc0N01, Co(80)2Ho1 Co(c0)3No1 188 mie stoichiometty precursor(s) + + 190 CoC0N0(C4H9NH2) Co(CO)3NO + + + + 205 C0(C4H9NH2)2 Co , COCO , C0(C0)2 + + 221 C02N0(C4H9NH2) Co + + 249 C02C0N0(C4H9NH2) COCO 3. Ni(C0)x+ reactions with sec-butylamine (sBuNHz. c4511gl Ions formed as products of ion-molecule reactions in a mixture of Ni(c0)4 and sBUNH2 are listed below, with their precursors as identified by double resonance. mie stoichiometry precursor(gi + .+ . + 72 C4H10N N1 , NlCO 89 Ni(CH3NH2)+ NiCO 103 Ni(c2NH2)1 Ni1, Hico1, Hi(c0)21 . . . + 115 Ni(c3H5NH2)1 N11, NlCO 117 Nic0(CH3NH2)1 Ni(c0)21, Ni(c0)31 . + .+ 127 N1(C4H7N) N1 . + + 128 N1C0(CH3CHNH2) CH3CHNH2 . + .+ l29 N1(C4H7NH2) N1 131 Hi(c4H9NH2)1 Nico1, Ni(c2H5NH2)1. Ni(c0)21 132 NiC0(C4H6)+ Ni(c0)21, Ni(c0)31 1 Hi(c0)41 145 Ni(C0)2(CH3NH2) 159 Nic0(c4H9NH2)1 Ni(co)21, Ni(c0)31, Ni(50)41 189 gig stoichiogetry precursor(gl . + . + . 204 N1(C4H9NH2)2 Ni(c0)2 , N1(C0)3+, Ni(c0)41, Ni(c4H9NH2)1. Nic0(c4H9NH2)1 . + .+ . + . + 215 N12CO(C4H7NH2) N1 , N1C0 , N12(C0)2 4. Cr(C0)x+ reactions with sec-bugylamine (sBUNHz, c4g11gi Ions formed as products of ion-molecule reactions in a mixture of Cr(C0)6 and sBUNH2 are listed below, with their precursors as identified by double resonance. gig stoichiometry grecursor(s) 112 CrCOCH3NH31 Crco1, Cr(c0)21 + 123 Cr(C4H7NH2) Cr1, Crco1, or(50)21 + 125 Cr(C4H9NH2) Cr1, Crco1, Cr(c0)21 + + + l53 CrC0(C4H9NH2) Cr(C0)2 , Cr(C0)3 , Cr(CO)4+ + + + l98 Cr(C4H9NH2)2 CrCO , Cr(C0)2 , + + Cr(C4H9NH2) , Cr(C0)3 , CrCo(C4H9NH2)+ 5. Mo(C0)x+ reactions with sec-butylamine stUNH2.§4flJ]fll Ions formed as products of ion-molecule reactions in a mixture of Mo(CO)6 and sBUNH2 are listed below, with their precursors as identified by double resonance. mie 151 163 165 191 217 219 190 stoichiometry Mo(c4H7N)1 Mo(c4H7NH2)1 Mo(c4H9NH2)1 MoC0(C4H7 Mo(C0)2(C4H7N)+ Mo(C0)2(C4H7NH2)+ + "”21 precursor(s) Mo+, MoCO+ Mo1, Moco1 Moco1, Mo(c0)21 + + M0(C0)2 , M0(C0)3 Mo(CO)3+ Mo(CO)3+ 6. WLC01x+ reactions with sec-butylamine (sBUNHz, cqgllgi N(C0)x+ reactions with sec-butylamine were also studied by use of the CEC 2l-llO B double focusing mass spectrometer without precursors identification. are listed below: mie 225 237 249 251 253 279 307 309 337 343 stoichiometry Hc(CH3NH2)1 w(c3H5N)1 HC(c3H5N)1 H(c4H7N)1 W(C4H7NH2)+ Hc0(c4H7N)1 W(CO)2(C4H7N)+ H(c0)2(c4H7NH2)1 w(c0)3(c4H7NH2)1 W(CD)3(C4H9NH2)+ The reaction products 191 B. Discussion Table 20 summarizes the Fe+, Co+, Ni+, Cr+ and Mo+ reactions with sec-butylamine and also their product distri- butions. Scheme XXX shows a proposed mechanism leading to the formation of reaction products. It is of interest to note that Ni+ undergoes a hydride abstraction from sec-butylamine to yield C4H10N+ and that Fe+ forms a high energy product, Fe(C3H8N)+ by losing a methyl group after inserting into the C-C bond of sec- butylamine. Table 20 also shows a trend that Fe+, Co+ and Ni+ exhibit much richer chemistries in their reactions with sec- butylamine. However, Cr+, Mo+, and N+ are mostly inducing hydrogen molecule(s) elimination. Note that NC+ is a reac- tive species, since it produces NC(CH3NH2)+. Also, w+ is the only species which can eliminate both CH4 and H2 to pro- duct W(C3H5N)+. In Scheme XXX, a process pathway leading to the formation of M(CH3NH2)+ is also shown. Note that this process needs an isomerization of sec-butylamine to isobutyl- amine. Similar isomerization steps have been previously reported91’97. Also shown in Scheme XXX are pathways leading to the formation of Fe(C3H6)+, Fe(C4H7N)+, Co(C3H5)+, Ni(CH3NH2)+ with a concurrent loss of the C0 ligand. n + ...-21*” i “9.1 192 mm 20 . huuou with soc-m2 (bit-l2) h b I1 u—ol(u,-z).+ Gals Mead-Y3 ski I. amazing-,1» c211. HKC‘IGYO .34 I2 “wai al., Hit-1,1131% a}, H‘ch‘ff1 a: Exam-1‘s I, «9.119.211 J37 .10 .03 .26 .23 .11 .19 .16 Knox—1 HKQ)‘.II‘Q gun. a grabbing-1,11. can‘t m ~Km)u(flj.z Y. can‘o m ~10).(m,~,)'9 ‘3': —.1(co)‘_1(a3-3y. can}. do ~9KQ)b1(P7I5I).o 0211... :2. w «991.19.»? en, Mews-5'12)“ can on MOLJm‘usnzii of“ 2m and. 1,4(95314 can). on “(mk‘ficjl‘fi 03.24. in «whom-.13 '3‘ I21 90 «mum-.33 If I.* w antitank-1i an an Mayday-3+ «1,. on *wlhfiwfi :29 m ammuwwlfo nzmn Mm),.,(q.s,li+ a, . do ~(m),_2(wf+ 2.112 mm ~Km),_,(c.l,lit again «away-1,11. “wk-I‘vv'z" ‘” H‘m)x-2(%-z).1 24D an),,,(c.I,-z)’* ID «90),; «mm-wqwii an E‘muflcu'gizl.’ 2m Km),-,n(c..l,-,)'i Jan .16 .15 .25 .19 .fi .‘6 .75 .61 .17 . 16 .19 .19 .16 . 35 .10 .18 .25 .60 .56 .50 .111 .36 T? .a a” 1.0 .56 1.0 a, .68 8 1'. action I a 1 v "1 W‘M-in-zF-Nwt ”((11); o hung-awa5-23—a No.51, )12 x e 3.4 0(0)}. muz—pwqugnzf—ouqnylzfz 230.1,! cm «9 «(0)3n—H032m119 o-m-zdbzflqlflz‘ no am. 4 Mm),p—omz(m)zflomz—mzm(6~l9flzy Nani; --wnz-+u(c,n,-I,1—ua(c.u,w’, x = o. 1 111(c11i);...nuint2 «1.5-2141(wlz’: W). “2.3.1. 94001; an. Q'io'lzf-‘mws'fl; 3.9.2). 331.2,) E n ......5... E ME 2.3” ......ae mazes! Aninfir manor :4. n A... . ./.\ new END .. 51w. . «a; uninzorfinflov immmll «sameneéinfi sf... .....5 \I \2 / \\ N...._ We... .— .a- Enos-gig? a 3mm“; ......MM . . fl 3 c. .3: flaw: n 1 _ E or w ...; m. ...... .... ...... a av... .2 .72. A N... no: fl ._ flawed mi EA .... fwd sum. ....n o .w . .32 may. nwiuuwe hmfikanfi CE. .2... .A n .— 75 55. fl ..- 43...; 32 2&5... : M233! .1 Q new n in.“ Sandy \_ __ 5w... new new. - NE - _ n x - “Mu—”HA 6 «SJ. 5&0 5 NEthhzmu nBVu .n new”... arena 2... Sat8§\_. gates; coated .3 one 0.0 : EVI/ Tulle... . xxx gunom 194 There are some new reaction products which are not seen in metals alone. The formation of CrCO(CH3NH3)+ can be explained similarly to the formation of M(CH3NH2)+ in Scheme XXX except that one more B-H shifts from the left side to give CrC0(C3H5)(CH3NH3)+ followed by a simple fragmentation. The formation of FeCO(C2H6N)+ can be explained by the FeCO+ insertion between skeletal atoms C2 and C3 with CO acting as a spectator (because CO is retained on Fe+). Note that the loss of C3H5-radical which is also seen in CID spectra of 166 Co+ reactions with alkanes is a high energy process. VIII. Conclusions Based on our original goals, we have the following results from this work: 1. All transition metal ions in this work insert into a C-I bond. 2. All transition metal ions in this work induce a 8-H shift. 3. It is hard to tell if 5 or 6 membered rings are preferred in the intermediates. It appears that 6-membered rings are the preferred intermediates in the reactions with 2-pentanone. Apparently, one cannot yet explicitly explain the differences of metal ions in terms of the controlling fac- tros which were assumed in the beginning of this section. However, from this study, it seems that thermodynamics only 195 dominates the reactions disregarding geometrical considera- tions. From this study, we do see a certain pattern of how different metals react in different ways, for instance, Fe+, Co+, and Ni+ are more reactive than Cr+ and Mo+ in that Fe+, Co+ and Ni+ have stronger interactions with organic sub- strates and thus can rearrange the organic molecules to different products. Cr+ and Mo+, on the other hand, seem to prefer to eliminate one or more molecules of H2, which is especially true for Mo+. In order to gain some insight into the differences in chemistry of metal ions, an ab initio calculation on CrCHz+ is being pursued and will be discussed in the next section. AB INITIO CALCULATION C. AB INITIO CALCULATIONS 1. Introduction I. The Importance of Ab Initio Calculations: In the last section (8.5), we have seen that Fe+, Co+, and Ni+ exhibit similar chemistries in their reac- tions with most of the organic compounds used, and Cr+, Mo+ and w+ fall into another group, which exhibit quite different chemistries in their reactions with the same organic com- pounds. However, we can not understand the bonding behavior of these metal ions in terms of their orbitals, orientation, size and symmetry as those listed at the beginning of sec- tion 8.5 from these studies. Note that the reaction products we have observed from ICR studies are mass peaks. Therefore the assignments of ion structures of these products are sometimes ambiguous, although the double resonance technique, the CID (collisionally induced dissociation) technique and the use of the labelled compounds can sometimes give helpful information. It is important to do the ab initio calcula- tions so that we can obtain insights into the nature of the bond between the metal and substrate. 60, corresponding to an [Ar] 157 6 Fe+, has a ground state, 1 6 157 45 3d configuration Cr+ has a ground state 8, corresponding to [Ar] 3d5. However, there is some evidence““’148 for the presence in the gas phase of its next excited state 6D (45 3d4) at l.47 eV. It is not surprising that Cr+(6S) will form an addition complex with l2-crown-4 196 197 electrostatically. To form covalent bonds with ligands, it presumably needs to use its 45 orbital. 0n the other hand, Cr+ (GD) might be able to form covalent bonds with ligands 2 4. by use of sp hybrid orbitals. Ni has a ground state D, 9 157 corresponding to the [Ar] 3d configuration and there is no evidence that its next excited state 4F at leV (corres- 8 ponding to [Ar] 451 3d configuration) participates in the gas phase chemistry. Thus Ni+ (20) will interact with ligands electrostatically and form the addition complexes with TDE, l2-crown-4 and l5-crown-5. It can also interact with ligands covalently by use of its 5 and d orbitals+163 or sp hybrid orbitals by promoting to 4F statel123’164’165. + Ni+ has a ground state of [Ar] 3d9. To form a covalent bond with a ligand, a d orbital has to be involved in bond- ing, or it may mix with a s orbital to form a sd hybrids, or Ni+ could be experiencing the configuration mixing of [Ar] 3d9 and [Ar] 45 3d9, as a result of positive charge154 on the metal center. CID spectra of some product ions in the reactions of Co+ with alkanes conducted in our laboratory166 also indicated that there are more than two different struc- tures at the same mass unit, which might be derived from the reactions of different electronic configurations of the metal ion. $ Note that Kunz concluded the 452 bonding to H by use of sp hybrids for all the first transition metals. Later, Walch et. al. correlated the s orbital with p orbitals, end- ing up with a much d character in ScH with '2+ as its ground state. However, its next excited state, 3A which is the ground state from Kunz's calculation is only 4.6 kcal/mole above 'Z+ ground state. 198 II. Review of Ab Initio Calculations of Transition Metal Compounds. From spectroscopic studies, P.R. Scott and 150 N.G. Richards concluded that three types of configura- tions are important in the monhydride, and they may be 2 n-2 loosely described as s , 5dr"1 and d". In ScH and TiH, 2 the s d"'2 is important. From TiH to FeH, the sdn'1 config- uration always gives rise to ground state; however, the d electrons are largely non-bonding in these compounds. How- ever, in NiH the s-orbital contribution is still not negligible. In the second transition series, the d-orbitals are relatively less stable, and PdH may be understood in terms of simple covalent bonding between 4d- and ls orbitals alone+. 151 However, recent ab initio calculation shows that ScH has a tremendous amount of do bonding in the ground state. 151 S.P. Halch incorporated atomic correlation terms leading 2 1 3d9 separation for the Ni atom into all electron MCSCF/CI calculations for the XZA state of to an accurate 4s 3d8- 4s NiH and found that although the bonding in NiH is predomin- 2 atly 4s1 3d9 like, 45 3d8 like configurations are found to be important in the small R region, and he expects that + 3d orbital is only 30% size of 4s orbital. Hence, there will be poor overlap of 3d with ligands; 45 on the other hand will form bonds easily. In contrast, 4d has 60% size of 55 orbital (electron density distribution), which then makes it possible to bond by use of a 4dz orbital. 199 2 l a mixture of 4s 3dn and 4s 3dn 1 to be important in l 6 most transition metal bonds except for Mn where 45 3d is 152 too high to be accessible for bonding From other calcu- lations on systems, such as NiH, NiCHz, NiCH3, Nico‘53, 2, CrH154, it was concluded that the s orbital is impor- CrCH tant in bonding with only a small contribution from d orbital or a weak d bonding. To form a n bond, the ligand has to get closer to the metal, and the strong repulsive interaction with 45 orbital in the 452 3dn configuration will avoid this energy lowering process, unless the 4s orbital mixes with 4p to become two 4s 4p hybrids as in NiCHz. Also, model studies of n-bonded organometallic sys- tmes Mn-CZH2 and Mn-CZH4 are calculated to have small bond- ing energy, < 10 kcal; Ni-CZH2 is bound by 2 20 kcal, and Steigerwald and Goddard calculated the CpZTi(C2H4) and ClzTi(C2H4) systems and found that metallocycle form is more stable than n-bonded form by C2H4. 0n the other hand, MnCH2+ and CrCHz+155 were calculated to have the very similar ground state molecular orbitals, they all use M+ 45 to bond the methylene 3a1 orbital although 1 5 ground configuration, while Cr+ is 3d5. 154 Mn+ has a 45 3d CrCH; seems to be similar Further, the ligand has a strong influence on the mtal center for its bonding to another ligand124’156. It turns out that the change of the electronic configuration of a metal center used for bonding might be crucial as we keep 200 putting ligands on it. In the solid state, if we keep add- ing electrons into ZnS sytem to become GaSe then CuPbS and then As system, we will see some bonds broken in the unit cell to become different structures. Therefore, the 123 strategy that Beauchamp used in relating promotion 1 3dn'], to bond energy between first energy, from 3dn to 45 transition metal ions and simple ligands has still to be scrutinized as ligands are changed to the multifunctional molecules, since the transition states or intermediates of the metal ion molecule reactions might involve many ligands to be bonded to the metal center. III. Theory_of Ab Initio Calculations Although there are a growing number of ab initio calculations published in the past two decades‘72, very few of them were devoted to the understanding of univalent metal ion-ligand species. There are some books available for discussing the theories in ab initio calcula- 173'177. However, a brief review will be discussed in tions the following. From quantum mechanics, a total wave function of an atom can sometimes be expressed by a Slater determinant. For example, the lithium atom can be written as ¢1s (1) ¢is (l) ¢2s (1) ¢1s (2) ins (2) ¢2s (2) ¢is (3) ¢1s (3) ¢2s (3) sill" 20l or in a simpler form of 4" ¢l$ 371—5- “’25 where o's are one-electron wave functions (spin orbitals). In most calculations the orbital m is expanded in terms of a set of basic functions, Our problem is to determine the coefficient Ci' The Xi can be chosen to form an orthonormal set, {xgdv=1 f X.X. dV 1 J 0 for i # j but in practice it is more convenient if they are normalized but not orthogonal, i.e. but Thus the wave function of the molecule W can be broken down as follows: W = and , these can be broken down into inte- grals involving the orbitals, whichiriturn reduce, or expand to atomic integrals involving the basic function Xi' In writing the electronic Hamitonian equations, there is considerable advantage if one works in atomic units, in which n (planck's constant devided by 2n), the electronic charge e, and the electron mass me are all unity. Also, the distance is given in bohr radii ao (bohrs), where l bohr O 8 cm or 0.52918 A. The unit of energy is is 0.529l8 x 10‘ the Hartree, equal to 27.2l ev or 627.5 kcal/mole and is equal to ez/ao. The rigorous mathematical expression of the molecular orbital model is the Hartree-Fock(HF) approximation. For closed-shell atoms and molecules the HF wave function is of the form we = A (n) ¢1(l)¢2(2)....¢n (n) (1.1) and is conveniently written as a Slater determinant: _1_ we = ,3. l ¢] (1) m2 (2) ...... ¢n and can be expressed as, N E (H 25 + J log loglo + °' = 2' £9) 9 where 513 is the energy of one electron in a molecule and 9 * is defined as flog (l) H? log (I) dV1 and J is a coulomb integral and is defined as (2) dV1 dV = 2 * ‘k 'l - ff] ——— Cg (1) log (2) r12 log (1) log 2 l 2 If log (1) -——- log (2) dV1 dV r12 2 (assuming the orbitals are real). More generally, the energy is a sum of one-electron, coulomb, and exchange terms E = 2e? + z m]? (l) —1—¢§ (2) dv1 av i i1. (1)- . 8scr J J m, (1) or, in an even more compact form we may write HSCF ¢l (1) = €§CF¢i (1) Now if ¢i = Z Cin Xn n SCF _ SCF then H i C1n Xn - 6i E C1n Xn By multiplying both sides of this equation by Xm and integrating over all space, we get i Cin ("ifiF ' iCF Smn) = 0 or .... Hggi - .§CF smn - o If HgfiF and Smn could be calculated the secular deter- minant could be solved directly for the eigenvalues, the SCF SCF orbital energies 8i However, both Hmn and Smn demand a knowledge of the wave functions we are trying to find and 206 the solution has to be iterative, which makes the use of a computer mandatory. 2 For example, in LiH, the structure for Li is ls 25 and H is ls, and NLiH; x 2 now In outline, our procedure might be: I. Guess some values of the 0'5 2. Calculate all the various atomic integrals and . SCF hence build up Hiojo and Sioj 3. Solve the determintal equation giving the possible values of e§CF 4. Substitute these in the secular equations giving new 0'5 5. Go back to stage I and repeat until the values of SCF 8i an arbitary threshold (m10'6) and then take the or the C's converge to steady values within values of the converged C's. The result of this will be to provide SCF orbitals, both occupied and unoccupied: that is to say the coeffi- cients in 178 For open - shell SCF methods, both Roothaan's and 207 179 Nesbet's method are used in an attempt to eliminate the SCF ii ted by means of unitary transformations. off-diagonal terms a which cannot be completely elimina- The secular equations may be simplified by the symmetry considerations. The role of symmetry enters the calcula- tions in the following way: l. It enters the integral calculation and the basic function transformation 188’189. 2. It enters SCF calculation. Linear combination of basic functions for different irrepresentation species will reflect the symmetry of the molecule. This will simplify the large matrices by reducing them to a series of blocks along the diagonal, one for each molecular orbital symmetry type. Also, the matrices can be simplified by the use of symmetry operator. 3. It enters MCSCF and CI calculations. By mixing configurations having the same symmetry, the matrices in both calculations can be simplified‘go. Before going to the next section, some important con- cepts used in ab initio calcualtions are summarized in the following: l. The solution of the HF equations including symmetry and equivalence restrictions yields the restricted Hartree- Fock (RHF) wave function. By removing the symmetry and equivalence restrictions placed on RHF wave functions, single-determinant wave functions of lower energy can fre- quently be obtained.. 208 2. Electron correlation. In the HF approximation, the motion of each electron is solved for in the presence of the average potential created by the remaining (n-l) electrons. The contribution to the total energy due to instaneous repulsions is called correlation energy180. The most frequently used method for approaching the electron correlation problem is configuration interaction (CI), which is just a linear combination of configurations with coefficients variationally determined. More precisely, the CI wave function is of the form, where the 6's are an orthonormal set of n electron configura- tions. The coefficients C1. are determined to minimize the energy fngewedz. Application of the variation principle leads to CI - where ”ii = f¢;He¢jdz and is the matrix elements between configurations. Note that matrix elements Hij between different configurations i and j are zero if i and j are of different symmetry. Therefore, the secular equation is greatly simplified by only considering configurations which have the total symmetry of the particular electronic state being investigated. 209 3. Multiconfiguration SCF (MCSCF). Solution of the eigenvalue problem in (1.3) will yield the optimun values of C1, C2, and C3. However, the wave function has not been variationally determined yet unless the spin-orbitals used in the configurations have also been varied to minimize the orbital energy. The MCSCF wave function is the best (lowest energy) wave function that can be obtained by simul- taneously varying both the orbitals ¢ and the CI coef- ficients C. This regards to MCSCF caluclations, which is again usually solved by the interative techniqueIe]. 4. Basis sets. Molecular calculations are generally carried out in terms of basis functions centered on each atom in the molecule. A primary consideration in evaluat- ing the reliability (i.e. in agreement with reality) of an electronic structure calculation is the basis set, in which both slater and gaussian functions are most frequently employed. Slater basis functions have the radial form A rn-l e-(r where A is a normalization factor, n is the principal quan- tum number, and E is the orbital exponent or screening parameter. Guassian basis functions have the radialy form, - 2 B r" e or In principle one would probably prefer to use slater functions in all molecular calculations. However, for 210 nonlinear polyatomic molecules the two-electron integrals (such as coulomb and exchange integrals) in terms of a slater basis set are extremely difficult to compute, while the same integrals in terms of gaussians may be evaluated comparatively simply and rapidly. There are many different basis sets used depending on the systems we are dealing with. a. The minimum basis set. It includes one function for each SCF-occupied atomic orbital with distinct n and t quantum numbers. The functions of a minimum basis set are usually slater functions. The atom-optimized minimum basis 182 may be used directly molecular set developed by Clementi calculations. b. Double Zeta and extended basis sets. Usually, the minimum basis sets yield SCF energies above the HF energies. For this reason electronic structure calculation are fre- quently carried out with larger basis set. A double zeta includes exactly twice as many functions as the minimum basis. Any basis set of slater functions larger than double zeta are referred to be an extended basis set. c. Contracted functions. As mentioned above, for big and nonlinear molecules, the use of gaussian functions are utilized to speed up computing the multicenter integrals, despite the fact that the number of gaussian functions is more than tiwce the number of slater functions. In order to save the computation time, the use of contracted 211 gaussina5183"185 , i.e, linear combinations of gaussians with fixed coefficients is suggested. In solving the SCF equations, then, only the coefficients in each SCF orbital of the contracted functions must be determined. A notation of (105 5p/35 2p) means a basis set of ten 5 and five p (for example six ls functions are used for ls orbital, two ls functions are used for 25 orbital and two ls functions are used for 35 orbital and three 2p functions are used for 2p atomic orbital and another two 2p functions are used for 3p orbital) has been contracted to three 5 and two p functions. d. Polarization functions. In order to approach the HF energy limit, functions with higher 2 values (d,f....) must be added to the basis when we are working on molecules. These functions with higher 2 values being added are called 186 polarization functions A basis including (4s 2p ld) on each first row atom and (25 lp) on each hydrogen is referred 187 to as double zeta plus polarization Note that the double zeta level for a first row atom is (4s 2p). The reliability of each basis set depends on the molecular pr0perties we are dealing with. Therefore, the choice of the basis set is often based on experience. 2. Use of CrCH2+ As An Example of Ab Initio Calculation In this section, a study of CrCH; will be given as an example of an ab initio calculation. 212 Assume Cr+ interacts with CH2 in a C2v symmetry in the following coordinates: y b H L L Cr+ ./ a Z 1 HR b b] R where the various orbitals can be associated with the irre- ducible representations of C2v character table 191, C2 ov(xz) o;(yz) 2v A1 1 1 ’ 1 z " x2,y2,22 -A2 l -l -l Rz xy 81 -l l -l x, Ry x2 82 -l -l l y, Rx yz I" o o 2 1 By using the formula, a. = %-EX(R)Xi(R) where h is the R group order, X(R) is the character of the matrix correspond- ing to operation R in the reducible representation, and Xi(R) is symmetry any irreducible representation, we can get the species for the bonding orbitals bL and bR. (2 + 2) l i.e., there is one A1 species n+a (2 + 2) l i.e., there is one 82 species A+a 213 To determine orbitals correspondong to A1 and 82, we use a projection operator‘gz. 3A1 HR = (1)1511R + (1) 62 HR + (1) 8v(xz)HR + (1) o;(yz)HR = HR + HL + HL + HR = 2 (HR + HL) A32 x A A A1 P HR = (l) E HR + (-l) 02HR + (-l) ov(xz) HR + ov(yz)HR = HR - HL - HL + HR = 2 (HR - HL) Hence, the sum of HR and HL belongs to the A1 irreduci- ble representation and the difference of HR and HL to the B2 irreducible representation. By consulting the character table of C2v symmetry, we have the following correspondence: lsc -ila1 2P5 A1 2sC —> 3:11 32 ——>lb2 81:2pg —1b1 2p: 1s2 - 15E . -92a1 ls? + lsé} Since two orbitals having almost the same energy and symme- try can form a bond, the 2p; may form a bonding orbital with H C lsL + ls: and 2p.y may form another bonding orbital with 153- lsE. Therefore the electronic configuration of CH2 in the 3 ground state 81 can be described as 2 2 2 3 la1 2a1 3a1 lbz lb1 ( 8]) 214 2 251 2p; 2p; 2p; configuration. 0n the 2 2 l l . . y sz conf1guration, we will have the electronic configuration of CH2 as, where carbon has ls other hand, if carbon uses ls 25 29 2 2 2 2 la1 2a1 3a1 lb2 which is the first excited state of 1A1 symmetry. Keeping this idea in mind, we will have the following species: species ground state first excited state or+ 1522522p63523p6, l522522p63523p6, 3d5(65) 4s‘3d4(5o) ‘ CH2 la$2a$lbg3a1lb1 la$2a$3a$lb§(]A1) 3 (3,) There are 23 electrons in Cr+, l8 electrons are in the Ar core and 5 electrons are valence. There are 8 electrons in CH2, 6 electrons are in the "core" and there are only 2 "valence" electrons. Therefore, the symmetry adapted core molecular orbitals can be described as "core" orbitals Cr+ contributions £52 contribution la1 ........ .7a1 ls,25,2pz,3s, 3pz lsc, bR + bL lb1 2b1 2px, 3px none and leave 7 valence electrons for bonding between Cr+ and CH2. 215 By symmetry, we have the following states when we leave Cr+ in C2v symmetry. states of Cr+ in an states of Cr+ in C2 symmetry isolated atom V 6 4 6 6 6 6 D (sd ) 2 A1 + 81 + 82 + A2 Note that there are five states in 02v symmetry when Cr+ is in its first excited state, because 0 has 2 = 2, and the number of states is given by 22 + l = 5. In other words, there are five different ways of distributing one s and d orbitals, each gives corresponding state. Now, if we leave Cr+ and CH2 in 02v symmetry at w separation, we will have even more possible states, which are summarized below: Cr+ CH2 (Cr - CH2)+ at w separation193 6 3 8 6 4 6s 1111 6111 l l l 2 2 6 l 6 6 6 6 0 A1 2 A1 + 81 + 82 + A2 Since the 1A1 state of CH2 has an electron pair located on the z axis, when CH2 approached to Cr+, in an effort to make a bond with it, these paired electrons might experience strong repulsion form the d electrons of Cr+. We think that this interaction will be repulsive and it is therefore 216 reasonable to disregard any states arising from the combina- 1 tion of A.' of CH2 with ground state Cr+. If we consider the first situation, 65 + 3 8B], 6B1 and 48], we will have 81 and the states which evolve from the no bond, one bond and possible double bond respectively. 881 - there are 7 unpaired electrons with the same spin, and it is unlikely to be bound, i.e. no bond will form. 81 - there are five unpaired electrons, i.e., there is one singlet couples pair. Hence, the pos- sible chemical bond at large separation must be due to the orbital overlap between 3 do of Cr+ and o orbital of CH2. B1 - there are three unpaired electrons, i.e., there are two singlet coupled pairs and hence a pos- sible double bond exists. Therefore, the calculation will focus on the 6B1 and 4 6 3 81 states. However, we have to remember that D + 1 B1 and A1 asymptotes also contribute states of these symme- tries. Nhile these are noninteracting at w separation, they may interact strongly as the distance R decreases to the equilibrium value. The 681 state arises from one bond formed by Cr (65) and CH2 (38]). Consider the valence electrons of both Cr+ and CH2 in this state. 217 Cr+ (core) 3d5 l—- 1 1 1 1 M31 b1 b2 a1 a2 1 1 CH2(core) 3— El 1 where do denotes dzz; d mx denotes dxz and dfly denotes d YZ’ dA+ denotes dx2_y2 and dA- denotes the dxy orbital with |m£|= 0,1,2 respectively. Also, the symmetry of each orbital can be found from the character table. Now, if a sigma bond is formed between do of Cr+ and po of CH2 (both have the same irreducible representation a1), then by tak- ing into account the core orbitals, the molecular orbitals of CrCH2+ at infinite separation can be described as: 2 core 2 + 8a](do + po) 3b](CH2) 9a](dA ) 1a2(dA') 4b1(dnx) 4b2(dfly) In order to carry out reliable ab initio calculation, good basis functions have to be chosen for each atom. The Cr+basis functions used in the CrCH; calculation are from Nachters, i.e., (l4 s, llp, 6d/ 55 4p 3d). The carbon basis (lls, 6p/3s 2p) is from Duijneveldt and the hydrogen basis (45/25) is from Huzenaga. In all 120 primitive gaus- sians are used to form 48 contracted orbitals over 4 centers. For convenience, each basis function is given by a number 218 so that they are easy to read. The basis functions used in CrCH2+ are the following. l 15 21 xz 4l yc 2 25 22 x2' 42 yé 3 35 23 xz" 43 2C 4 4s 24 yz 44 zé 5 55 25 yz' 45 sHR 6 x 26 yz" 46 SAR 7 x' 27 xx 47 sHL 8 x" 28 xx' 48 SfiL 9 x"' 29 xx" 10 y 30' yy 11 y' 31 yy' 12 y" 32 yy" 13 y"' 33 22 14 z 34 zz' 15 z' 35 22" 16 z" 36 5C 17 z“ 37 sé 18 xy 38 56 19- xy' 39 xC 20 xy" 40 xé where x, x', x" and x"' basis functions are used to stand for the px orbitals of chromium, and y, y', y", y"' are for the py orbitals of chromium, etc° xy, xy', xy", xz, xz', xz"--- 219 22" are used to build up the 3d orbitals of chromium. Number 36 to 44 are used for s, px, p and p2 wave functions y of carbon and SHR’ SHR are used for 5 wave functions of Hydrogen atom HR and SHL’ SAL are used for 5 wave functions of Hydrogen atom HL‘ Moreover, these basis functions can be catagorized to four irreducible representations in C2v character table according to their symmetry characters by the use of the same method used on page 212. Hence, we have the following arrangements 1 A1 Symmetry (22) g 82 Symmetry (11) 1 1 23 10 2 3 24 11 4 4 25 12 5 5 26 13 6 14 27 24 7 15 28 25 8 16 29 26 9 17 30 41 10 2 x 33 - 27 - 30 31 42 11 2 x 34 - 28 - 31 32 45 - 47 12 2 X 35 - 29 - 32 33 46 - 48 13 27 - 30 3 Bl Symmetry (9) 14 28 - 31 34 6 15 29 - 32 35 7 16 36 36 8 220 17 37 37 9 18 38 1 38 21 19 43 39 22 20 44 4o _23 21 45 + 47 41 39 22 46 + 48 42 4O 4 A2 symmetry (3) 43 18 44 19 45 20 Thus, the 48 basis functions are reduced to 45 symme- try orbitals after taking into account the symmetry in a C2v system. Note that the first and third column label the orbital numbering which is the computer readable form. For example, orbital 28 is acutally the basis function 25 (yz') and is 6b2 in C2v symmetry (Note that each symmetry species starts at 1). Similarly, orbital 21 is actually the sum of basis function 45 and 47, namely 5HR SHL’ and is 21 a1 in C2v symmetry. Therefore, the molecular electronic configuration of CrCHZ+ described on page 217 2 2 can be written in an orbital numbering form by referring to the symmetry orbitals given above as 221 (1 2 3 4 5 6 7 34 35 23 24 25)2 8 8 36 9 43 37 26 or in a ordered form: CORE: 1 2 3 4 5 6 7 23 24 25 34 35 VALENCE: 8 8 9 26 36 37 43 Table 21 lists the results of ab initio calculations of CrCH2+ at different electronic states. 3. Discussion Figure 10, and 11 summarize the energy levels of different states from the ab initio calculations. At this level of calculations none of these states seem to form a bond between Cr+ and CH2 as that reported by Beauchamp148 (65 i 7 kcal/mole). As we can see from these figures that 681 only forms a 12.5 kcal/mole bond at (2 x 2) SCF level, which includes the o bond correlation. 6 A2 forms a 22.0 kcal/ mole bond relative to excited state of Cr+(60) and ground state of CH2(381). It doesn't co-relate to ground states of separated species and is thus not the ground state. On the other hand, 481 gives a tremendously high energy at SCF level, reflecting the inadequacy of the SCF theory for this state. Correlating o and n configurations to allow elec- trons more freedom, the energy is lowered. In fact, the lowest energy of this sytem is formed at bond separation r = 1.85 A. However, its energy, —1082.02832 H shows no evidence of the presence of strong bond between ground 222 A 83238.8“. 8888 .....smmxisme ...s s. Nflmmmhfienafieoamefl 88o v + u wfivflofioov .68: m Hmmodmo? «wed «definaannfimmaflmmawmmfi 280 Z. 93mm 5.3030 5 m 588 . 831 «93 we Nearer.” {Hammafiarfl .88 T 833.58 98$ :3 a 988.821 «88 8m? x NV .3 m 88o . $31 «$5 Nflfanpmfieoafm wemNAanmnmmfifmfaf? 1 1 fig MI «0 833.58 83 5+5... x 359.831 «SJ 5m? x 3 Npefemflfoapn H83 8m x 038.831 «83 memfiwpmmpmmfifmafir? 1 1 Had 9:25.60 “Mono M 533950 ”mm“ mama .8 8033.38 2.35 2 no 39mg . Hm e38 223 m mmmmm.amoa1 «cm.a eumfin x mv 83.3968 82:. + m :8 a 838.831 «mm; 8m? x C A “hand 265 0.... 805033“ mo acumen.“ 23 3033 8838.48 83: + m + 5m x 88.831 «woo 8m? x 3 833488 883 Team x 39.831 «88 cam? x 3 N3 «.5 ado x «88.831 «8.4.. «Ammfmfampnmfimfifiem 1 1 1 44$ an: mamfmfnfieoaadmwma 88 Y. Namfafnffiaamwamfi .88 Y. «maflpmfipmammawamfi eumm1wm4 vcoesoo a m coapgmcoo N450 9.83 S .538. 4O 36 Figure 10. Energy Levels Of 631 And 224 CrCH§(SBI and 6.4.2) (Fixed oi Schoefer's Geometry) C '1089.98I66fi , 0*(601 +CH2(3B 1’ r- 1 \ I 1 1- \ 34.4 MHEZZchol I 1 : \ , 140329019; L ‘ ‘ _ 32.1 MH=24.311ca1 \ 1' \ : \ 11 .. mosaousoe 6A L'IOBQOIQYOfi 2 r Cr"(551+c1-12(38,1\ ; 108202550 : A : SCF \\ 19.5 MHEIZ.5kcal L 1- \ F \ t Ugagoaerr 63' E (2:12) SCF 1- 6.42 States Of “0112* 225 manage 38m wemv ...: an: 8 :33 9.8 .: paw: 38: go .5. ”a _ 59.58:: an .m. “8 Us Us _ m... . .on . m9. _Bxuaiz ..nn , .3. H 9. — “No ....mamzufiooro“ on 308.39. .8 1483105 (HWBV B. 222.8. mom mom 3.». ..{uzu ..aoiu ... _ _ . _ _ . ILILII l 3 2.938.-” _ _ _ _ ..mn.«:u+.€+o 111098.59- lllllllllllllLllllJ 0v qumxucu (HWEIV 226 state of Cr+(65) and CH2(3B]), which is in contrast with Beauchamp's suggestion that CrCHz+ has a double bond. Beauchamp did suggest that if metal ions promote one elec- tron into the s orbital, they will form a stronger bond with a ligand. This is borne out in this calcualtion since 6A2 forms a 22 kcal/mole bond with respect to Cr+(6D) and CH2(BB]). One should however include d electron correla- tions at the MCSCF level and then compare with the same MCSCF level energy of separated species, i.e. Cr+(65) and 3 CH2(uB]). 6 6 Neither the B], nor the A2 state is predicted to have comparable bond energy to the experimental value, although 6A2 has the most energy lowering relative to excited state of Cr+(60) and CH2 (38]), which somewhat matches Beauchamp's suggestion that essentially the bonding between the first row transition metal ions and ligands is via the metals 5 electron. However, he pointed out that CrCHZ+ doesn't fit in his plot of promotion energy (coordinate) versus bonding energy (abscissa). Perhaps the experiments he did do not refer to the bond energies of these high-spin electronic states of CrCH2+. However, since the calculation of 431 indicates a failure to form strong bonds, the experimental value is hard to refer to either ground state or the first excited state of Cr+ and ground state of CH2. This may sug- gest another possible structure of CrCH2+, for instance, + The H-Cr+-C-H instead of C v symmetry structure of CrCH2 2 227 possibility of forming bonds from excited state of CH2 has been overlooked, because when it gets closer to Cr+, the paired electrons on CH2 will experience a strong repulsive force. However, to check if it is possible to undergo a back bonding from d electrons of the Cr+, a 6 A1 state, derived from excited states of both Cr+(60) and CH2(1A]) is being calculated. Other states derived from Cr+(60) and CH2(3B]) have not been considered in this work. The reason being that other arrangements of four d electrons in five d orbitals of Cr+ would not make the bonding dramatically dif- ferent, neither would correlate the d electrons in MCSCF calculation, since the bonding is mostly contributed from 45 orbital of the Cr+ and 3a1 orbital of the CH2. Ab initio calculations provide considerable insight in- to the nature of the chemical bond since we can not only obtain the bond energies, but also orbital occupancies. Calculations of this sort might help us to understand the nature of transition metal insertion processes. For exam- ple, in order to understand how Cr+ can split up H2 mole- cule (D(H-H) = 104.2 kcal/mole) to yield a low energy pro- dUCt’ crH+(D(Cr+-H) = 35 i 4 kcal/mole)148 after putting the kinetic energy into Cr+. Ab initio calculations of CrH+ and CrH2+, are in progress. APPENDICES APPENDIX A. Schematic Diagram For ICR Voltages Controls There are two operating modes in ICR: normal drift mode and trapping mode. In the normal drift mode, we want all drift plates and trap plates of the cell to be +15 V to -l5 V (dc) adjustable. However, we need to pulse one trapping plate in both the source and the analyzer regions. Ions will have long residence times for reaction during positive pulse and are swept out of cell during the negative pulse. Also, the same pulsing will form a modulated signal out of the marginal oscillator, and then is input into the lock-in amplifier to be in phase with reference wave and consequently enhance the S/N ratio. This can be done by setting the mode switch to normal drift mode in Fig. l2. In this case, the monostable vibrator 74l2l doesn't get a triggering pulse and Q will output a +5 V dc. As a result, both 2N404A PNP and 2N3053 NPN are switched off and output a +l5 V do as a +GATE pulse. On the other hand, a will send out a -5v dc on a dar- lington pair of transistors on the other side, which will conduct both 2N404A PNP and 2N3053 NPN to output a -lS V dc as a -GATE pulse. Note that FETs need more than 10 Vs to drive, a regular TTL voltage cannot serve this purpose. 228 229 Hence, the FET at the bottom in Fig. 13 will always be ON and the drift potentials (adjustable) relative to the ground will appear on a drift plate after a voltage follower 741. Four of the same circuit will allow us to adjust four drift plates separately. On the other hand, the comparator Bll in Fig. l4 will send out a pulse train whose pulse width is determined by voltage inputs to trigger the 7412l so that output 6 will give a TTL pulse train with pulse width adjustable by external RC values. When the pulse is at +5 V, both PNP and NPN transistors don't conduct and will output a +l5 V at B, which will switch on N-channel FET (2N38l9). Conse- quently, a reverse setting will then appear on the pulsed trapping plate (remember we put a unit_gain of inverting amplifier in the front of 2N3819). When the pulse is at 0V, both PNP and NPN transistors will conduct to output a -l5 V at point B, which will then turn on p-channel FET (2N3820) so that the voltage setting will appear on the pulsed trapping plate. In this way, we can have a voltage put on the pulsed trapping plate. Meanwhile, the voltage will always appear on the constant trapping plate. 118, there are three pulsing In the trapping mode sequence: trapping, detection and quench. Quench sequence can be combined in the detection step. For the trapping and detection period, we need to keep the trapping plates posi- tively charged on both the source and analyzer. However, in 230 the trapping period, the source drift plates have to be at ground, and the analyzer drift plates have to be at negative potentials which can be done by adjusting the trapping poten- tial as shown in Fig. 13. In the detection period, opera- tion is set back to drift mode by putting a positive potential on the top drift plates and a negative potential on the bottom drift plates by adjusting the drift potential in Fig. 13. The necessary pulse timing for trapping and detection is provided again, by the circuit in Fig. 12. In order to be more versatile, input to pin 3 of the 311 com; parator can be either a slow ramp for varied trapping and detection times or a constant voltage as shown in Fig. 16. When pulse from Q output is at 0V, it will switch on both PNP and NPN transistors to give -15 V. When the pulse is at +5 V, it will switch off the PNP and NPN transistors to give + 15 V at +GATE output as shown in Fig. 16. When the +GATE reaches +15 V, it will switch on the bottom FET in Fig. 13 and at the same time the -GATE is at ~15 V to switch off the top FET. As a result, the drift potential will appear on drift plates for detecting ions. On the other hand, when +GATE is at -15 V, it will switch off the bottom FET. Meanwhile, -GATE will be at + 15 V to switch on the top gate and the trapping potentials will appear on the drift plates for trapping the ions. The trapping voltage in this mode can be set by switch- ing the timing circuit in Fig. 14 to ground, which will 231 switch on both PNP and NPN transitors and will send out -15 V dc to point B which will drive P-channel FET 2N 3820 only so that a constant voltage appears on the pulse trapping plate. 50 now we have the same voltage going to both trap- ping plates. To be flexible, we can make two of the same circuits for both source and analyzer. In the trapping mode, we also need a circuit to pulse the electron energy, this is done in Fig. 15. 0 output (pin 6) of a 74121 will send out a pulse train to switch on and off the 2N 5415 PNP transistor so that it can pulse the electron energy and form ions, producing a modulated signal out of detector for S/N enhancement as described earlier. As we can see from these cirucit diagrams, all the tim- ing pulses can be replaced by computer software programming using DACS. The computer setup for on line data acquisition is shown in Fig. 17. Where the setting cirucit on the left bottom is for the mass calibration and circuit after ADC 0817 is for amplifying the signal and tailoring the analog signal to the safe values (5 5.12 V for PET). The software has been developed in our laboratoryllg. 232 .eoflomm wfimglhflgoflo 58 8H 3&5 NH 25mg ..o. t. l a: nan AL“ What ‘8 .n £03.sz 0084 B; so- 233 -GATE IN + 15v +59A 1 -15v 0 (4:5)“) 10K C 2N38L9 74, . To Cell Plate 6 (One for each TRAPPING 'K ‘ drift Plate) POTENTIALS ”“3" 975 . 4 $50K +6ATE -lSV 3 l :pfio {-ISV +15v o 1" 459A ; 2~3319 (1;)? (OK (> u: j".— , L— J- I ~ ‘HSV To - ift plate -L tap some(UIS) Figure 13. Mapped Ion Cell Circuitry-m Drift Section. 234 . 338m gmafillggoflo Sumo :3 damage .3” gm?" 3mg M535 «539.501 I; B. 91014») x8 w («$2. : @ <33: 2 28a \ J ’1’.) IIPU‘1 b no. 0 o 1 ..I._ x8 :8. L w“. _ 5.1M“ as. .06 5.... )9. H3358 .. a _ a h... u H ...-P ...R 3.1 w o >h1. amfim 235 mean vamsuafim.:ouuooflm .ma onnwflh @fim >7 9.: who... 8.03 >nal ... an mmofiou no.5 owdvag no can: H >2. g. .n c: v . I“. j mszN 9 . . . ¢ ”7.! 1 - .. L r ..nm - . m8. ...... >3- ... To «...... . 9:05.38 on. 7 T6 I “a W H? mm H ... 152 no «339?... .8925 coupooam >n+ 9.8m and... 236 fast lamp input to pin 2 of 311 WWI/AV l/ 1/1/1/1/1/ A /1 {3.311.313.33... 5VIl‘llllllilllli'll.outputofjll V O, H n H n H H H H H output of 74121 Blow ramp output of 311 II” H H ' H H H Qoutputof74121 o_. .' - -0 Hsv drift(detection) ”i H i i '1‘ n ' -..--1__ _ __-_______4 +GA'IEoutput * Z. i‘ ppins Figure 16. Timings 0f tapped Ion Cell Circuitry In Figure 12 237 .25.“.6523 -o~ .28 . 6H .8 990m Beganfié .S 835 NMO. lv. >._.2 24m 5% xwm .24me $0092 >5 .giggafi 7.25 x 0%. OVON Emu . 241! Q- venom NOmm- 92.5 .35 .030 1 .OON hung 659:4 5.-..qu d 0N. mda m uoEcm>072 o 003$ 55> . ..bs. 3.8.8.... .25.... 3.3. 0022.9; . ooQ..> Egon—u 65> .N. €32.33. Appendix B. Marginal Oscillator Setup In ICR Experiment The tank circuit of a marginal oscillator is shown below: V R L C I —LICR ;J-Cell The principle of detecting ions is in references 10 and 11. 238 APPENDIX C. [he Realtionship Between Magnetic Field And Mass In The ICR Experiment -22-_- wc - mc 2nvc v an . _ c B - e At resonance, Vc = Vm.o for m = 100 and if vm.o = 153 kHz then B = 1.53 x 105 Hz x 100 amy x 1.67 x 10‘27 kg/amu x 2w 1.5 x 10“9 coul 1.0028 W/mz = 10028 c = 10.028 kG Similarily, for m = 350, 8 = 3.510 11m2 = 35.1 kG Therefore, by setting 0 = 153 kHz, at 10 k0 we can m.o get a peak corresponding to m/e 100 and at 35 k0, we can get a peak corresponding mass unit of 350 amu. 239 APPENDIX D. Alternate CID Circuit For Conventional ICR Double resonance techniques in ICR can be used to unambiguously identify the precursors of products. Yet, sometimes it is hard to determine the structure of product ions. For example, N'++CH ——-—>N°(CH)++H ‘ 410 ‘48 2 many possible structurescan be assigned to Ni(C4H8)+: N142 Max .0 11—~i+—H Substitution reactions can solve some of these questions. However, collisional-induced dissociation (CID)120’121 is a good technique for distinguishing one structure from another. In conventional ICR, one can design a circuit which will excite or eject an ion while scanning the magnetic field. If that ion we are exciting is the product ion, it will gain energy from a radiofrequency (r.f.) and will actively collide 240 241 with neutral molecules and then dissociate. We also can excite any reactant ions to determine how kinetic energy imparted to them affects the reaction. Fig. l8 shows this design which was constructed. Note that a ramp (0-l0v) from the magnet controller is fed into the op-amp to reduce the ramp to 0-5v, which is the useful range that can be fed into the VCG of a Wavetek frequency generator. In this way, the angular frequency of the ion will be linked with magne- tic field while scanning the magnetic field. A 10 turn pot is used for slope adjustment. The circuit in the upper section is for offset about 5v as shown below 10 V ramp from magnetic controller 0 < x; + offset{i Slope and offset adjustment - offset{: The fine adjustment offset value is done by use of a 500 k pot. 242 6 V Zener Diode Low Leahge 005 ”f 005 ”f ' l H k . m: l on” I *15" 10k ‘ 15v +lOv 51 ‘ VHS 5 1% 'Do ..‘v Uave’oek 902 LHoo'ro-oH *5" I»! K 1 O-lOv . from magnetic controller Figure 18. Schematic Circuit Diagram Of CID For Conventional ICR. Appendix E. Calculation Of The Collision Frequency For Co+ And .v/O\,, (l) polarizability16 e(ahc) = %(§TA)2A3 where N is total electrons of molecule a is polarizability TA is atomic hybrid components obtained from Table I in Ref. l6. [(1:C + 3TH + Tc + 21H)2 + TOJZ #43 N 0‘c H 4 100 [(1.294 + 3x0.314 + 1.294 + 2x0.314)2 + #4:- N 1.290]2 = %7 (92.275) = 8.7889 A3 expt'l = 8.73 A3 (2) Langevin collision rate17 KL = Zne (95)”2 where u is the reduced mass u u = ggfigg = 35.63758 -24 3 KL = 2x3.l4x4.8xl0'10x( 8.73xlO cm _24 )1/2 35.63758xl.67xlO g 243 244 l.lS45xl0'9cm3/molecules (3) collision frequency is KL.N But '. N at 23 % = P(t°rr)X5-°2x‘° = 3.24155x1019 molecules/l2 760x0.082x298 = Px3.24155x1016 mo1ecu1es/cm3 = 3.24155x1016 3 molecules/cm at 25°C P = 5xlO'6 torr for ethyl ether N = 5x10'5x3.24155x1016 = 15.20775x1010 molecules /cm3 '. collision frequency = KLN = 1.1545x10‘9x15.20775x 1010 = 187.11898s'1 1 1 3 (4) time between collisions ? = 8 .1 898 = 5.34419xl0 S Appendix F. Branching Ratios Of Fe+ Reactions with .~,0\,, Fe+ +xc,0\,r > Fe(CH20+) + CH The absolute intensity of each peak and the absolute power' 4 + + ~———+ Fe(C2H6O ) ~1F-+> (CH20)F9 (CZHGO) + drop in the double resonance spectra are arbitary but they have to be in the same scale. First, we have to take the peak height for each peak in question (m/e 86, l02, l32, and l33 in this case) gig_ 85 102 132 133 corrected ion peak height peak height Fe+(CH20) .80 9.30 Fe+(C2H60) 2.75 27.0 (CH20)Fe+(C2H60) .58 5.15 (CH30)Fe+(CzH60) 2.53 19.8 where corrected peak height = peak height/mass of the ion. + + _ Total Fe (CZHGO) from Fe - 1102+ + 1132+ + 1133+ 245 °. % of m/e 133 Double 246 resonance spectra of m/e l32 PUT30 55+ + '. % of m/z l32 .20 1.35\[ .2dffi \[1 15 84+ 102+ 112+ contribution from Fe(C2H60)+ = .07 (.30+l.§§+. Double 20+l.15) resonance spectra of m/3 l33+ _U:50 50+ + .60 T.50+4.2+.60+. 50V ADV 4.2\’ 84+ 102+ 112+ contribution from Fe+(C2H60) = 10) = 01] '. total contribution of m/3 132+, l33+ from m/e 102+ is: 1132+(.07)+ 1133+(.ll) '. branch ratio of m/e 132+ '. branch ratio of m/e l33+ = 5.l5 x .07 + 19.8 x .ll = 34 + 2.19 = 2.54 = .34/2.S4 = .l4 = 2.19/2.54 = .86 '. total contribution of Fe+(C2H60) from Fe+ = 1102+ + 1132+ + 1133+ = 27.0 + (.07) x 1132+ + ( 11) x I133+ = 27.0 + .07 x 5.15 + .ll x 19.8 247 = 29.5 The double resonance spectra of m/e 86 appears as 2%] follows: + 56 and m/e 102+: {HAT l.6 56+ 34+ .31 %contribution from Fe+ _ .15 _ 09 ’ .ls+l.5' ' ° m/e 132+, l33+ should be from successive reaction by m/e 102+ and m/e 85+ Fe+ + \,0\,.—r———9 Fe(CH20)+ + 0H4 m/e 86 4————a Fe(C2H60)+ —:lia»m/e 132 m/e 102 ‘86! m/e 133 . . + Total contr1but1on from Fe = 186+ + (.09) x 1102+ 9.302 + .09 x 29.5 = 9.30 + 2.54 11.8 '. branching ratio of m/e 85+ = 9.3/11.8 = .79 '. branching ratio of m/e 102+ = 2.54/11.8 = .21 However, if we assume all m/e 132+, 133+ are from successive reaction of m/e l02+, then; Total Fe+(C2H60) from Fe+ = I102+ + 1132+ + 1133+ = 27.0 + 5.15 + 19.8 = 51.9 248 ‘. % of m/e 132+ contribution from Fe+(C2H6O) = 5.15/(5.15 + 19.8) = .21 '. % of m/e 133+ contribution from Fe+(CzH60) 19.77/(5.15 + 19.8) = .79 i.e. 102+ '21; 132+ '79; 133+ Total contribution from Fe+ = 186* + (.09)I]02+ = 9.30 + (.09) x 51.9 = 13.8 '. branching ratio of 85+ = 9.30/13.8 = .58 ‘. branching ratio of 102+ = 4.46/13.8 = .32 .68 i.e. Fe+ +V0v Fe(CH20)+ + CH4 .32 + .21 Fe(C2H60) -~—— (CH20)Fe+(C2H60) .79 _—— (CH30)Fe+(C2H60) REFERENCES LIST OF REFERENCES 1. T.A. Lehman and M.M. Bursey, Ion Cyclotron Resonance Spectrometry, J. Wiley and Sons, New York (1976). 2. J.D. Baldeschwieler and 3.5. Woodgate, Acc. Chem. Res., .1. 114 (1971). 3. J.D. Baldeschwieler, Science, 1 9, 263 (1968). 4. H. Hartman, K.H. Lebert and K.P. Nanczek, Fortschr. Chem. Forch., 43, 57 (1973). 5. R.C. Burnier and 8.5. Freiser, J. Chem. Educ., 56, 687 (1979). 6. R. T. McIver, Jr., Sci. Amer., 243, 186 (1980). 7. J.L. Beauchamp, Ann. Rev. Phys. Chem., 22, 527 (1971). 8. J.M.S. Henis, Ion-Molecule Reactions, Vol. 2, J.L. Franklin, Ed., Plenum Press, New York (1972). 9. D. Hobschall, Rev. Sci. Instr., 36, 466 (1965). 10. R.T. McIver, Jr., Rev. Sci. Instr., 44, 1071 (1973). 11. A. Warnick, L.R. Anders and T.E. Sharp, Rev. Sci. Instr., 45, 929 (1974). 12. R.T. McIver, Jr., R.L. Hunter, E.B. Ledford, Jr., M.J. Locke and T.J. Francl, Int. J. Mass Spec. Ion Phys., 32, 65 (1981). 13. J. Wronka and D.P. Ridge, Rev. Sci. Instr., 5}, 491 (1982). 14. 1.0. O'Haver, 0. Chem. Educ., 49, A131 (1972); ibid, 19, A211 (1972). 249 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 250 . Goode, A.J. Ferrer-Correia and K.R. Jennings, . J. Mass. Spec. Ion Phys., 5, 229 (1970). --'7< H63 0 :0 dc; (+0 . Miller and J.A. Savchik, J. Amer. Chem. Soc., ___, 7206 (1979). G. Cioumousis and D.P. Stevenson, J. Chem. Phys., 29, 294 (1958). ‘— Heats of reaction can be calculated from the group values for Hf. See s.w. Benson, Thermochemical Kine- tics, 2nd ed., J. Wiley and Sons, New York(l976). There are three ways to product metal ion source: (1) electron impact on metal carbonyls, (2) thermo- ionic emission (see J.P. Blewett, Phy. Rev., 59, 464 (1936), and D.H. Smith and J.A. Cater, Int. J. Mass Spect. Ion Phys., 49, 211 (1981), (3) laser desorp- tion (see R.C. Burnier, J. Amer. Chem. Soc., 191, 7127 (1979), and R.B. Cody and R.C. Burnier, et al., Int. J. Mass. Spect. Ion Phys., 33, 37 (1980)). Some unstable ions can be detected by varying drift volt- ages on method (1) and (2). Laser ionization techni- que sometimes gives metal ions having higher kinetic energies, thereby giving more reactions in some cases. P.B. Armentrout and J.L. Beauchamp, J. Chem. Phys., 74, 2819 (1981). M.S. Foster and J.L. Beauchamp, J. Amer. Chem. Soc., 9], 4814 (1975). G.A. Somorjai, Chemistry in Two Dimensions: Surfaces, The George Fisher Baker Non-Resident Lectureship in Chemistry at Cornell University, Cornell. Univ. Press. 198] O R.F. Heck, Organotransition Metal Chemistry, Academic Press, New York (1974). H.F. Schaeffer III, Acc. Chem. Res., 19, 287 (1977). J.R. Anderson, Structure of Metallic Catalysts, Academic Press, London (1975). M.G. Thomas and B.F. Freiser, J. Amer. Chem. Soc., 98, 1296 (1976). E.L. Mutterties, Chem. Eng. News., Aug. 30, 28 (1982). w.A. Goddard, S.P. Walch, A.K. Rappe, T.H. Upton and C.F. Melius, J. Vac. Sci. Technol., 14, 416 (1977). 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 251 R.B. Brewington, C.F. Bender and H.F. Schaeffer, J. Chem. Phys., 66, 905 (1976). R.R. Schrock, Science, 212, 13 (1983). R. Burch, Acc. Chem. Res., 16, 24 (1982). G.A. 02in and A. Vander Voet, Acc. Chem. Res., 6, 313 (1973); ibid, 19, 21 (1977). Myron L. Corrin, J. Chem. Educ., 66, 210 (1978). W.E. Billups, Mark M. Kanarski, Robert H. Hauge and John L. Margrave, Tetra. Lett., 61, 3861 (1980). P.B. Armentrout and J.L. Beauchamp, Chem. Phys., 48, 315 (1980). ‘T T. Kametani and K. Fukumoto, Accts. Chem. Res.,‘g, 319 (1976). P.W. Tiedemann and J.M. Riveros, J. Amer. Chem. Soc., ‘26, 185 (1974). C.H. Depuy, J.J. Grabowski and V.M. Bierbaum, Science, 218. 955 (1982). S.E. Buttril1, Jr., J. Chem. Phys., §2, 6174 (1970). P.S. Brateman and R.J. Cross, JCS Dalton, 657 (1972). W. Mowat and G. Wilkinson, J. Organomet. Chem., 36, C35 (1972). W. Mowat and G. Wilkins, JCS Dalton, 1120 (1973). W. Mowat and Shortland, G. Yagupsky, and G. Wilkinson, JCS Dalton, 533 (1972). U. Klabunde and G.W. Parshall, J. Amer. Chem. Soc., 94, 9081 (1972). Y. Ishii and M. Tsutsu, Fundamental Research in Homo- geneous Catalysis, Plenum Press, New York (1978). C.T. Mortimer, Reaction Heats and Bond Strength: Pergamon Press, New York (1962). D. Ialage, S. Brown, J. Connor, and H.A. Skinner, J. Organometal. Chem., 61, 403 (1974). 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 252 K.W. Egger, J. Organometal. Chem., 65, 501 (1970). M. Brown and R. Puddephatt, J. Chem. Soc. Dalton, 1613 (1974). M.R. Litzow and T.R. Spalding, Mass Spectrometry of Inorganic and Organometallic Compounds. Elsevier Scientific Publishing Co., New York11973). J.L. Franklin, J.G. Dillard, H.M. Rosenstock, J.T. Herron, K. Draxl, and F.H. Field, Ionization Poten- tials and Heats of Formation of Gaseous Positive Ions, NSRDS-NBS, 26 (1969). R.E. Winters and R.W. Kiser, J. Phys. Chem., 66, 3198 (1965). A. Foffani, S. Pignataro, G. Distefane, and G. Innorta, J. Organometal. Chem., 1, 473 (1967). J. Charalambous, Ed., Mass Spectrometry of Metal Com- pounds, Butterworths, Boston (1975). C.S. Kraihanyel, J.J. Conville, and J.E. Sturm, Chem. Comm., 159 (1971). J.B. Westmore, Chem. Rev., 16, 695 (1976). S.M. Schildscrout, J. Phys. Chem., 66, 2834 (1976). J. Muller, Angew. Chem. Int. Ed. Eng., 11, 653 (1972). D.E. Games, A.H. Jackson, and K. Taylor, J. Organo- metal. Chem., 66, 345 (1975). R.D. Macfarlane and D.F. Torgerson, Science, 1_l, 920 (1976). M.A. Posthumus, P.G. Kistemaker, and H.L.C. Meuzelaar, Anal. Chem., 66, 985 (1978). J. Pierce, K.L. Busch, R.A. Walton and R.C. Cooks, J. Am. Chem. Soc., 196, 2583 (1981). F. Marcel and J.G. Roustan, Org. Mass. Spect., 11, 173 (1982). W.P. Anderson, N. Hsu, C.W. Stranger, and B. Munson, Organometal. Chem., 9, 249 (1974). . Hunt, J.W. Russel, and R.L. Torian, J. Organo- J D F. metal. Chem., 46, 175 (1972). 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 253 R. Dunbar and J. Hutchinson, J. Amer. Chem. Soc., 3816 (1974). G. Distefano, J. Res. -NBS-A. 74A, 233 (1970). J. MUller, and K. Fenderl, Chem. Ber., 165, 2199 (1971). 9_6_. M.S. Foster and J.L. Beauchamp, J. Amer. Chem. Soc., 93, 4924 (1971). M.S. Foster and J.L. Beauchamp, J. Amer. Chem. Soc 97, 4808 (1975). J. Wronka and D.P. Ridge, Int. J. Mass. Spect. Ion Phys., 66, 23 (1982). R.C. Dunbar, J.F. Ennever, and J.P. Fackler, Jr., Inorg. Chem., 16, 2734 (1973). S.M. Schildcrout, J. Amer. Chem. Soc., 66, 3846, (1973). R.R. Corderman and J.L. Beauchamp, Inorg. Chem., 1 665 (1976). M.S. Foster and J.L. Beauchamp, J. Amer. Chem. Soc 21, 4814 (1975). F. Schumacher and R. Taubenset, Helv. Chim. Acta., 1525 (1966). J. Muller, Advances In Mass Spectrometry, Vol. 6, _5. 11. A.R. West Ed., Appl. Science Publishers, Ltd. England, 823 (1974). J. MUller, W. Holzinger and W. Kalbfus, J. Organomet. Chem., 2], 213 (1975 J. MUller and K. Fenderl, Chem. Ber., 166, 2207 (l 971). 0. MUller and w. Goll, Chem. Ber., 129, 1129 (1973). R.R. Cordeman and J.L. Beauchamp, J. Amer. Chem. Soc., 96, 3998 (1975). R.R. Corderman and J.L. Beauchamp, J. Amer. Chem. 66, 5700 (1976). R.D. Wieting, R.H. Staley, and J.L. Beauchamp, J. Chem. Soc., 97, 924 (1975). Soc., Amer. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 254 J. Allison and D.P. Ridge, J. Amer. Chem. Soc., 66, 7445 (1976). J. Allison and D.P. Ridge, J. Amer. Chem. Soc., 1_l, 4998 (1979). 8.0. Radecki and J. Allison, submitted to J. Amer. Chem. Soc. R.C. Burnier, G.D. Byrd and B.S. Freiser, J. Amer. Chem. Soc., 166, 4360 (1981). J.S. Uppal and R.H. Staley, J. Amer. Chem. Soc., l_6, 4144 (1980). J.S. Uppal, D.E. Johnson and R.H. Staley, J. Amer. Chem. Soc., 166, 508 (1981). J. Allison and D.P. Ridge, J. Amer. Chem. Soc., 66, 35 (1977). J. Allison and D.P. Ridge, J. Amer. Chem. Soc., [_9, 163 (1978). J. Allison, Ph.D. Dissertation, University of Delaware, 1976. 6.0. Byrd, B.G. Burnier and B.S. Freiser, J. Amer. Chem. Soc., 165, 3565 (1982). P.B. Armentrout and J.L. Beauchamp, J. Amer. Chem. Soc., 166, 1736 (1980). P.B. Armentrout and J.L. Beauchamp, J. Amer. Chem. Soc., 166, 784 (1981). P.B. Armentrout and J.L. Beauchamp, J. Amer. Chem. Soc., 166, 6628 (1981). P.B. Armentrout, L.F. Halle and J.L. Beauchamp, J. Amer. Chem. Soc., 166, 6624 (1981). R.W. Jones and R.H. Staley, J. Amer. Chem. Soc., 1_g, 3794 (1980). R.C. Burnier, G.D. Byrd, and B.S. Freiser, Anal. Chem., 66, 1641 (1980). M. Lombarski and J. Allison, Int. J. Mass Spect. Ion Phys., 66, 281 (1983). 255 101. R.T. Morrison and R.N. Boyd, Organic Chemistry, 3rd Ed., Allyn and Bacon Inc., Boston, Massachusetts (1975). 102. R.A. Bartsch, Acc. Chem. Res., 6, 239 (1975). 103. G.A. Olah, Acc. Chem. Res., 6, 41 (1976). 104. F.A. Carey and R.J. Sundberg, Advanced Organic Chemistry, Plenum Press, New York (1977). 105. A. Pullman, Giessness-Prettre and Yu. V. Kruglyak, Chem. Phys. Lett., 66, 156 (1975). 106. G. Wipff, P. Weiner and P. Kollman, J. Amer. Chem. Soc., 196, 3249 (1982). 107. E.E. Astrup, Acta. Chem. Scand. 919, 85 (1980). 108. J.D. Dunitz and P. Seiler, Acta. Cryst., 696, 2739 (1974). 109. M. Dobler, J.D. Dunitz, and P. Seiler, Acta. Cryst., 696, 2741 (1974). ‘ 110. P. Groth, Acta. Chem. Scand., 666, 279 (1981). 111. J. Dale, Israd. J. Chem., 69, 3 (1980). 112. J. Dale, Tetrahedron 99, 1683 (1974). 113. D.K. Gabbiness and D.W. Margerum, J. Amer. Chem. Soc., 91, 6540 (1969). 114. D.A. Phipps, Metals and Metabolism, Oxford University Press, (1977). 115. J.L. Hall and-D.A. Baker, Cell Membranes and Ion Transport: Integrated Themes In Biology, Longmans, Harlow, U.K., (1977). 116. A.I. Popov, Pure & Appl. Chem” 61, 101 (1979). 117. N.S. Poonia and A.V. Bajai, Chem. Rev. 16, 389 (1979). 118. T.B. McMahon and J.L. Beauchamp, Rev. Sci. Inst., 96, 509 (1972). 119. S. McElvany, P. Hardebeck and J. Allison, 30th Annaul Conference on Mass Spectrometry and Allied Topics, p. 848 (1982). 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 256 R.B. Cody, R.C. Burnier and B.S. Freiser, Anal. Chem., 66, 96 (1982). R.B. Cody, R.C. Burnier, C.J. Cassady and B.S. Freiser, Anal. Chem., 69, 2225 (1982). A. Tsarbopoulos, J. Allison, unpublished results. L.F. Halle, P.B. Armentrout, J. Amer. Chem. Soc., 196, 6501 (1981). M.M. Kappes, R.W. Jones and R.H. Staley, J. Amer. Chem. Soc., 199, 888 (1982). W.B. Pearson, Structure Rgports, Vol. 28, p. 666, Int. Union of Crystallography (1963). M. Davis and 0. Hassel, Acta. Chem. Scand., 11, 1181 (1963). T. Bjorvatten, Act. Chem. Scan., g;, 1109 (1969). 0. Hassel and J. Hvoslef, Acta. Chem. Scan., 8, 1953 (1954). ’ 6. Hassel and J. Hvoslef, Acta. Chem. Scan., 6, 873 . Dekok and C. Romers, Rec. Trav. Chim., 66, 313 . Elie1 and Sr. M.C. Knoeber, J. Amer. Chem. Soc., 3444 (1968). E.L. Eliel, Acc. Chem. Res., 6, 1 (1970). R.R. Whitney and D.A. Jaeger, Org. Mass Spect., 15, 343 (1980). Based on the AHf (12-crown-4) = -150.8 kcal/mole as per, from ref. 136; and AHf(1,4-dioxane) = -85 kcal/mole as per, from ref. 135. All thermodynamical values, unless otherwise noted were obtained from H.M. Rosentock, K. Draxl, B.W. Steiner, J.T. Herron, J. Phys. Chem. Ref. Data, 6 (1977). K. Bystrom and M. Mansson, J. Chem. Soc. Perkin II, 565 (1982). 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 257 See for example, (1) P. Groth, Acta. Chem. Scan., §_A, 721 (1981), ibid, 35A, 541 (1981), ibid, 669, 109 (1982). (2) J.D. Dunitz and P. Seiler, Acta. Cryst., 999, 2733 (1974). J.D. Dunitz and P. Seiler, Acta. Cryst., 2739 (1974). P.E. Arte, J.F. Dupont, J.P. Declereq, Merrssche, Acta Cryst., 35B, 1215 (l 79 G.M. Van rc (1 R.B. Freas, D.P. Ridge, J. Amer. Chem. Soc., l_6, 7129 (1980). G. )0 P.J.F. Dupont, E. Arte, J.P. Decle G.G.M. Van Meerssche, Act. Cryst., 35B, 1217 ). 0.0 79 J.L. Franklin and J.G. Dillard, NSRDS-NBS, 26 (1969). S.K. Huang, Y.C. Lee, J. Allison and A.I. Popov, Spect. Leet., 16, 215 (1983). S.K. Huang, J. Allison, accepted for publication, Organometallics. Hydrogen shifts may be sequential, while we prefer to simultaneous double insertion. Sequential inser- tion would avoid intermediates with unusually high formal oxidation state for the metal. Based on the calculation of Pauling's formula. See B.E. Douglas and D.H. McDaniel, Concepts and Model; of Inorganic Chemistry, p. 109, Blaisdell Publ Co., Néw York (1965). L. Pauling, The Natureof the Chemical Bond, Cornell University Press, Ithaca, New York (1960). L.F. Halle, P.B. Armentrout, and J.L. Beauchamp, J. Amer. Chem. Soc., 199, 962 (1981). Y.C. Lee, A.I. Popov, and J. Allison, Accepted for publication, Int. J. Mass. Spect. Ion Phys. P.R. Scott and W.G. Richards, Molecular Spectroscopy. Vol. 4, Specialist Periodical Reports; the Chemical Society, Burlington House, London WIV OBN (1976). C.W. Bauschlicher Jr., and S.P. Walch, J. Chem. Phys., 16, 4560 (1982). S.P. Walch and W. Bauschlicher Jr., Chem. Phys. Lett., 86, 66 (1982). 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 258 R.V. Hodges, P.B. Armentrout and J.L. Beauchamp, Int. J. Mass Spect. Ion Phys., 66, 375 (1979). J.S. Uppal and R.H. Staley, J. Amer. Chem. Soc., 104, 1229 (1982). 7" T.H. Morton and J.L. Beauchamp, J. Amer. Chem. Soc., .6], 2355 (1975). W.G. Richards, P.R. Scott, E.A. Colbourn and A.F. Marchington, Bibliography of Ab Initio Molecular Wave Functions, Oxford University Press (1978). H.F. Schaeffer III, The Electronic Structure of Atoms and Molecules: A Survey of Rigorous Quantum Mechanil cal Results, Addison-Wesley Pub. Co. (1972). H.F. Schaeffer III, Applications of Electronic Struc- ture Theogy (Modern Theoretical Chemistry Vol. 4): PTenum Press, New York (1977). H.F. Schaeffer III., Methods of Electronic Theory (Modern Theoretical Chemistry Vol. 3)., Plenum Press, New York, (1976).~ W.G. Richards and J.A. Horsley, Ab Initio Molecular Orbital Calculations for Chemists, Oxford University Press (1970). D.B. Cook, Ab Initio Calculations in Chemistry, Butter- worths, Londfin (1974). C.C.J. Roothaan, Rev. Mod. Phys., 66, 179 (1960). R.K. Nesbet, PrOC. R. Soc., 230A, 312, 322 (1955). P.0. Lowdin, Adv. Chem. Phys., 63 207 (1959). J.H. Wilkinson, The Algebraic Eigenvalue Problem, Oxford University Press, London (1965). E. Clementi and A. Veillard, J. Chem. Phys., 69, 3050 (1966). E. Clementi and D.R. Davis, J. Comput. Phys., 63 223 (1967). J.L. Whitten, J. Chem. Phys., 66, 359 (1966). C. Salez and A. Veilland, Thev. Chim. Acta., 11, 441 (1968). 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 259 All the references can be found from : W.G. Richards, Bibliography of Ab Initio Molecular Wave Functions, Oxford University Press (1978). S.K. Huang, A. Lavarado, J. Allison and J. Harrison, 31st Annual Conference of Mass Spectrometry and Allied Topics, Boston (1983). M.A. Vincent, Y. Yoshloka and H.F. Schaeffer III, J. Phys. Chem., 66, 3905 (1982). W.C. Swope and H.F. Schaeffer III, M01. Phys., 69, 1037 (1977). C.E. Moore, Atomic Energy Levels: U.S. Government Printing Office, Washington, D.C. 1952, Vol. 2, NBS Circular 467. J.S. Uppal and R.H. Staley, J. Amer. Chem. Soc., 1 4, 1235 (1982). R.W. Jones and R.H. Staley, J. Amer. Chem. Soc., 196, 2296 (1982). M.M. 1913 as and R.H. Staley, J. Amer. Chem. Soc., 104, M.M. Kappes and R.H. Staley, J. Amer. Chem. Soc., 196, 1819 (1982) J.S. Uppal and R.H. Staley, J. Amer. Chem. Soc., 196, 1238 (1982). K ( S.P. Walch and W.A. Goddard III, J. Amer. Chem. Soc., 66, 7908 (1976). L.F. Halle, P.B. Armentrout and J.L. Beauchamp, Organo- metallics. l, 963 (1982). A.B. Kunz, M.P. Guse and R.J. Blint, J. Phys.,_66, 1358 (1975). J. Allison, et al., unpublished results. 0.8. Jacobson and B.S. Freiser, J. Amer. Chem. Soc., 196, 736 (1983). See for example, J.D. Lay Jr., and M.L. Gross, J.C.S. Chem. Comm., 970 (1982); R.D. Bowen, D.H. Williams, H. Schwarz and C. Wesdemiotis, J. Amer. Chem. Soc., 191, 4681 (1979). 260 186. R.K. Nesbet, Rev. Mod. Phys., 66, 272 (1960). 187. R.K. Nesbet, J. Chem. Phys., _6, 3619 (1964). 188. R.C. Raffenetti, J. Chem. Phys., 96, 4452 (1973). 189. R.C. Raffenetti, Chem. Phys. Leet., 69, 335 (1973). 190. K. Ruedenberg, L.M. Cheung, and S.T. Elbert, Int. J. Quan. Chem., 16, 1069 (1980). 191. For more complicated molecules, please refer to ref. 176, chapter 10. 192. F.A. Cotton, Chemical Applications of Grogp Theory, 2nd Ed., Wiley-Interscience, New York (1971). 193. Direct Product of 65 (6A in C2v Symmetry) and B1 will give B1 state, see ref. 192. 194. K.A. Kalmbach, D.P. Ridge, D. Peake and M.L. Gross, 31st Annual Conference on Mass Spectrometry and Allied Topics, Boston. (1983). 195. J. Allison and D.P. Ridge, J. Organomet. Chem., 66, C11 (1975). 196. J. Allison and D.P. Ridge, J. Amer. Chem. Soc., 96, 7445 (1976). 197. ?.V. 6odges and J.L. Beauchamp, Anal. Chem., 16, 825 1976 .