Early Transition Metal Organometallic Reactivity. Catalytic Acetylene Hydrogenation and Dimerization. Carbene Migratory Insertion Reactions Citation Threlkel, Richard S. (1980) Early Transition Metal Organometallic Reactivity. Catalytic Acetylene Hydrogenation and Dimerization. Carbene Migratory Insertion Reactions. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/ay0x-zs17. https://resolver.caltech.edu/CaltechTHESIS:10312023-222717548 Abstract [Not transcribed] Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Chemistry) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Bercaw, John E. Thesis Committee: Unknown, Unknown Defense Date: 4 February 1980 Record Number: CaltechTHESIS:10312023-222717548 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:10312023-222717548 DOI: 10.7907/ay0x-zs17 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 16230 Collection: CaltechTHESIS Deposited By: Kathy Johnson Deposited On: 31 Oct 2023 22:45 Last Modified: 31 Oct 2023 22:45 Thesis Files PDF - Final Version See Usage Policy. 31MB Repository Staff Only: item control page

Early Transition Metal Organometallic Reactivity. Catalytic Acetylene Hydrogenation and Dimerization~ Carbene Migratory Insertion Reactions Thesis by Richard Threlkel In Partial Fulfillment of the Requirements For the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California 1980 (Submitted February 4, 1980) ii ACKNOWLEDGMENT Grateful acknowledgment is made to all the members of the Bercaw group who have constantly reminded me how full of it I am during the last four years. Juan Manriguez, Don McAlister, Bob Bell, Bob Sanner, John Thi ch, Dave Erwin, Pete Wolczanski, Steve Cohen, Mike Flysuk, Buzz Curtis, Ayusmen Sen, Paul Barger, and Jimmy Mayer have all taught me something about chemistry. Some who tried harder than others might bear special mention except that I'll probably never par the eighth hole at Brookside or attempt a hopelessly disordered crystal structure in the near future. I would like to especially thank John Bercaw. In addition to his good advice in lab, I have appreciated his chemical curiosity to find out what the hell is really going on. I must also thank the Fannie and John Hertz Foundation for their financial support for the last three years. Finally, the expert typing job by Henriette Wymar is greatly appreciated. iii ABSTRACT The synthesis of several bis-pentamethylcyclopentadienyl titanium and zirconium mono-acetylene complexes from the j dinitrogen complexes, [(C 5 Me 5 ) 2 Ti] 2 N2 and [(C 5 Me 5 ) 2 ZrN 2 2 N2 , is reported. The steric bulk of the pentamethylcyclopentadiene ligands prevents metallocyclopentadiene formation commonly observed in normal titanocene acetylene reactions. These mono-acetylene complexes are efficient hydrogenation catalysts for acetylenes. A mechanistic inverstigation indicates that (C 5 Me 5 ) 2 TiH 2 or (C 5 Me 5 ) 2 ZrH2 are the active hydrogenating reagents. (C 5 Me 5 \ Ti also catalyzes the dimerization of terminal acetylenes such as propyne to form CH 3C=C-C( CH 3 )= CH 2 • A series of substituted niobocene zirconoxy carbenes, (C 5 H 5 ) 2 Nb(R)=CHOZr(H)(C 5 Me 5 ) 2 , are prepared by reaction of (C 5Me 5 ) 2 ZrH 2 with (C 5H 5 ) 2 Nb(R)CO ( R = H, CH 3 , CH 2 ¢, and¢). These oxycarbenes readily undergo migratory insertion into the niobocene substituent bond. The product of hydride migration can be trapped with CO to form (C 5H5 \Nb(CO)CH 2 OZr(H)(C 5Me 5 ) 2 • The alkyl migration products /3-eliminate to yield zirconoxy enolates, (C 5Me 5 \Zr(H)OCH=CH 2 and (C 5Me 5 \Zr(H)OCH=CH¢. Phenyl does not migrate at a significant rate at 70 °C. Kinetics of the alkyl migrations indicate the rate limiting step is the carbene-alkyl migratory insertion. The carbenehydride migration which is reversable is studied using saturation transfer nmr techniques. The relative rates for migratory insertion are determined to be H >> CH 3 > CH2 ¢ » <t>. This series of migratory aptidudes is discussed in terms of the electrophilic nature of migratory insertions. iv The syntheses of several substituted (C 5 Me 5 ) 2 Nb derivatives is described. (C 5Me 5 ) 2 Nb BH 4 is a useful precursor for (C 5Me 5 ) 2 NbH 3 , (C 5Me 5 ) 2 Nb(H)CO, (C 5Me 5 ) 2 Nb(CH 3 )CO, and (C 5Me 5 ) 2 Nb C¾. Low valent niobocene derivatives such as (C 5Me 5 ) 2 NbCO and (C 5Me 5 \Nb seem to be viable synthetic goals. V ABBREVIATIONS cyclopentadiene Cp CsHs Cp* Cs (CH3)s pentamethylcyclopentadiene Zr Cp*2 Zr Unless otherwise indicated, Ti Cp*2 Ti zirconium and titanium are always bound to two Cp* ligands. Nb Cp2 Nb Unless otherwise indicated, Ta Cp2 Ta w Cp2 W two Cp ligands. cp C6Hs phenyl tolane cp·C=C·cp diphenyl acetylene L A two electron donating ligand R A one electron donating substituent FT-nmr Fourrier transform nuclear magnetic resonance niobium, tantalum, and tungsten are always bound to vi TABLE OF CONTENTS Chapter I. Page Acetylene Reactions with Bis-Pentamethylcyclopentadienyl Titanium and Zirconium. II. Introduction 2 Results 3 Discussion 7 Experimental 12 The Kinetics and Mechanism of the Zirconoxy Carbene Migratory Insertion Reaction III. Introduction 19 Results 24 Discussion 48 Experimental 56 Synthesis of Bis-pentamethylcyclopentadienyl Niobocene Compounds Introduction 73 Results and Discussion 78 Experimental 82 1 Chapter 1 Acetylene Reactions with Bis-Pentamethylcyclopentadienyl Titanium and Zirconium. 2 INTRODUCTION ~ The search for new polymerization and hydrogenation catalysts has stimulated interest in the reactions of bis-cyclopentadienyl titanium complexes with acetylenes. Cp 2 Ti(CO) 2 catalytically hydrogenates acetylenes to alkenes and alkanes. 1 Because the only isolable titanium products of these reactions are unreactive metallocyclopentadienes, an unsaturated titanium monoacetylene complex was proposed as the active catalyst. This hypothesis was consistent with 2 the higher activity of Cp 2 Ti ~Cc?cf> in acetylene hydrogenations. f:.' . A mono-acetylene complex <I> of titanium was not isolable due to the facile metallocycle formation. The steric crowding observed for bis-cyclopentadienyl complexes 3 suggested that increased steric bulk around the metal would prevent coordination of a second acetylene. The perm ethyl titanocene system, in addition to yielding titanium II complexes cleanly, seemed to fulfill this requirement. 4 , 5 The synthesis of several group IV metallocene acetylene complexes is reported. Their behavior as hydrogenation catalysts enables further insight into the hydrogenation mechanism. In addition, the lackof metallocycle formation enables the titanium complex to catalytically dimerize terminal acetylenes. 3 RESULTS ~ a brown crystalline compound which nmr, ir, elemental analysis, and molecular weight studies indicate is a monomeric monoacetylene complex. The analogous reaction with 2-butyne yields a mono-acetylene R I * C Cp2 Ti - Ill C * and CH 3 , .,..._2 diphenyl acetylene per zirconium. However, two equivalents of 2-butyne react per zirconium to yield a zirconocyclopentadiene. An I </> * C Cp 2 Zr - Ill C 3 \ </> 4 .,..._ increase in steric crowding is apparent between the normal titanocene acetylene derivatives and the permethylated titanocene and zirconocene 4 ~ R / ~ M R zdC= C) Ti CH 3 Ti C5H5 1683cm- 1 1647cm- 1 Zr C6H5 < 1600cm- 1 8l.73(5H) 8 1.60 ( I H) 8=3 8=2 8= I TMS 5 complexes. Whereas metallocyclopentadienes are the most stable 1 acetylene products with normal titanocene, only the larger permethyl zirconocene can accomodate two small acetylenes. Otherwise the permethyl metallocenes are too crowded to allow metaH.ocycle formation. These mono-acetylene complexes are efficient acetylene hydrogenation catalysts. The general features of the reaction are that unactivated acetylenes such as 2-butyne and propyne are reduced to olefins much faster than activated acetylenes such as phenyl acetylene. The hydrogenation of unactivated acetylenes yields olefins whereas activated acetylenes become fully saturated. This double hydrogenation is stepwise. Stilbene is the initial product which is reduced further once a significant concentration is formed. For similar acetylenes, titanium is a more efficient catalyst than zirconium. All acetylenes cannot be compared due to metallocycle formation with 2-butyne and zirconium. Unlike the normal ring metallocycles, 4 -'"'- does function as a catalyst at elevated temperatures. These hydrogenation catalysts behave similarly to the normal ring titanocene system. The main difference is the relative reaction rates. Reasonable rates using Cp 2 Ti(CO) 2 required fifty atmospheres of hydrogen at 55-60° C. The reduction of 2-butyne to cis-butene by cp;Ti- has a turnover rate of gre,ater than 0. 75 equivalents of butyne per titanium per minute at -21 ° C. Even at -21 ° C, the reaction rate is stir-rate dependent at maximum stirring speeds. By preventing metallocycle formation, permethyl titanocene is able to catalytically dimerize terminal acetylenes. The reaction of 6 propyne and .,.__ 1 followed by nmr shows two multiplets grow in the vinyl region as the acetylenic hydrogen signal loses intensity. On a preparative scale, CH 3 C=C-C(=CH 2 )CH 3 is isolated as the only product in as high as 50 to one turnovers of propyne to titanium. 6 A similar reaction is observed with phenyl acetylene. 7 Unfortunately, it appears that both acetylenes which are coupled must be terminal. Even in neat 2-butyne with a small amount of propyne present, only the propyne dimer is formed. 2 R-C=C-H 1 .,.__ 7 DISCUSSION ~ The infrared stretching frequencies of the triple bond of the acetylene complexes and some similar complexes is shown in Figure 1. C=C Stretch (reference) C=C stretch(reference) ,</> ,</> C ( P c/> 3 )2 Ni - Ill C\ 1800 cm -1 C (9) Cp2 Mo -rn ,</> 1 740, 1768 cm -1 'cp ,</> C 1647 cm -1 * CP2 Zr - UI C '¢ f;H3 C (9) Cp 2 Mo -II 1830 cm -l (11) C\ CH 3 1 CH 3 C Cp2 W-111 C CH 3 1750 cm -1 \ ,</> C (10) 'cp C cP2* Ti - Ill C '¢ -1 C </> ( P c/>3 l2 Ft - Ill C 1771 cm ,CH3 < 1600 cm -1 C Cp2* Ti - Ill C 1683 cm tH 3 The difference between the free acetylene and the bound acetylene stretching frequencies, is usually rationalized on the basis of orbital overlap of the 7T and 1r* acetylene orbitals with the metal ct-orbitals. For a series of structurally similar complexes, the size of this shift -6.v(C=C) has been interpreted as a gauge of metal acetylene bond strength. 8 In general, Av(C=C) increases going down a particular -1 (11) 8 column (Ni - Pt or Mo - W). The ir spectrum of .,.,_ 3, otherwise indentical to .!., shows no obvious C=C stretch between 1600 and 2000 cm - i . A likely explanation is that relative to titanium, the lower energy C=C stretch of i has shifted into the phenyl C=C stretch region at 1585 cm- 1 . 3 indicate that .,.,_ The acetylene stretching vibrations of .,.,_ 1, .,.,_ 2, and of the two resonant structures, Sb the Iretallocyclo- propene is a better representation of the group IV metal acetylene complexes. ,R C -<- - > M - ,~ C 'R The metal acetylene bond strength is inversely proportional to the rates of acetylene hydrogenation. Activated acetylenes are hydrogenated slower than unactivated ones and the zirconium system is slower than the titanium catalyst. This behavior implies that dissociation of the acetylene from the metal is a necessary step of the catalytic cycle. Because perm ethyl titanocene oxidatively adds hydrogen very rapidly at -21 ° C to form Cp!TiH , 12 dissociative 2 substitution of the acetylene with hydrogen may generate the active hydrogenating agent. In fact, cp;TiH 2 is observed by nmr to react * Cp2 TiH2 + 2 CH 3 C=CCH 3 -80 °C fast + 9 with 2-butyne at -80° C to produce one equivalent of cis-butene and .?_. At -80° C, the 2-butyne complex 2 only slowly hydrogenates 2-butyne. '"' These data are consistent with the mechanism shown in Scheme I. Step one is rate limiting because both steps two and three are Scheme I slow ~H3 * C ---') Cp2 Ti - Il + cis- butane C CH 3 \ observed to be faster than the overall hydrogenation rate. In addition, cp;Ti reacts at -80° C with hydrogen and 2-butyne to consume only one equivalent of hydrogen and form 2 and cis-butene. This reaction '"' confirms the slowness of step one relative to steps two and three. 13 For every equivalent of titanocene formed, only one equivalent of butyne is hydrogenated. Unfortunately, the most interesting transformation, step three, is obscured by the rate-limiting acetylene dissociation. Scheme II depicts a possible scenario for step three. The acetylene inserts into 10 Scheme II ~H3 * C Cp2 Ti - Ill + cis-butene C CH3 the titanium hydride forming a vinyl hydride. Reductive elimination is promoted by another equivalent of acetylene. Note that either unassisted reductive elimination or hydrogen-assisted reductive elimination of the vinyl hydride violates the requirement of one hydrogenation per titanocene formation. A possible mechanism for the dimerization of terminal acetylenes is shown in Scheme III. The reaction is formally the addition of an acetylene hydrogen bond across another acetylene triple bond. Somewhat analogous to the acetylene hydrogenations, cp;Ti oxidatively adds across the acetylene hydrogen bond. Although any acetylene might react with the acetylene hydride in step two, only a terminal acetylene can coordinate and insert to form the en-yne hydride. The resulting reductive elimination 11 yields the cis-acetylene hydrogen addition product. Scheme ill R _J.! H * F c' Cp Ti - Ill 2 ,..R * _,..H 4) Cp2 Tl ,..,9 p 'c = C' H / R \ H C \ R 12 EXPERIMENTAL ~ All manipulations were performed under rigorously oxygen- and moisture-free conditions using either vacuum line techniques or a Vacuum Atmospheres glove box. All solvents were purified by distillation from titanocene immediately prior to use. 2-Butyne and phenyl acetylene were purified by distillation from molecular sieves. Propyne was freeze-pump-thawed and distilled at -78 ° C. Diphenyl acetylene (Aldrich) was used without purification. The nitrogen complexes of titanium and zirconium as well as the dihydrides were prepared by literature methods. 5, 14 , 15 Proton nmr were obtained from either a Varian A-60 or 220 MHz spectrometer. Infrared spectra were recorded on a Beckman IR-12. Molecular weights were determined cryoscopically in freezing benzene using a standardized molality-freezing point curve with a ferrocene standard. Hydrogenation product ratios were measured by peak weights of gas chromatograph output from a Varian 9 20 gas aerograph equipped with a 5% SE-30 on chromosorb column. Preparation of cp;TiC 2 (CJI 5 ) 2 , 2 400 Mgs of (Cpif i) 2 N2 and 200 mgs of tolane was dissolved in fifteen mls of toluene in a 50-ml flask attached to a swivel frit. The reaction proceeded at 0 ° C and the solution changed color from blue to brown. The toluene was pumped away and pentane distilled onto the reaction mixture. Pentane removal left microcrystalline product which was recrystallized from octane. Anal. Calcd for C34 H40 Ti: C, 82.23; H, 8.12; Ti, 9.65. Found, C, 81.93; H, 8.02; Ti, 9.90. 13 1 H nmr (d 6 benzene) o 1.92 (singlet, 3H), o 6.9 (complex, lH). Ir (nujol) 1647 cm- 1 (vC=C). Molecular weight 463± 40. Preparation of Cp;TiC 2 (CH 3 ) 2 , _?. 500 Mgs of (Cp;,ri) 2 N2 was dissolved in fifteen mls of toluene in a 50 mls flask attached to a swivel frit. An excess of 2-butyne was distilled onto the solution at -78 ° C and warmed to 25 ° C. 1 H nmr of the green solid shows pure_?., (d8 toluene) o 1. 89 (singlet, 5H), o 1. 74 (singlet, lH). Ir (nujol) 1683 (vC=C). Preparation of Cp;ZrC 2 (C 6H 5 ), 3 The same procedure used for !. substituting (Cp; ZrN2 ) 2 N2 6 yielded a dark-green crystalline compound, Anal. Calcd. for C34H10 Zr: C, 75.60; H, 7.41; Zr, 16.90. Found C, 75.39; H, 7.60; Zr, 16.81. 1 H nmrs (d6 benzene) o 1.80 (singlet, 3H), o 7.5 (complex, lH). Preparation of cp; ZrCiCH 3 ) 4 , 4 400 Mgs of (Cp;zrN2 ) 2 N2 was dissolved in fifteen mls of toluene in a 50-ml flask attached to a swivel frit. An excess of 2-butyne was distilled onto the solution. This mixture reacted at -20° C yielding an orange solution. After removing the toluene, seven mls of pentane was added. Slow cooling to -78 ° C yielded microcrystalline .,... 4. Anal. Calcd. for C 28 H45 Zr: C, 71. 53; H, 8. 94; Zr, 19. 45. H, 8. 98; Zr, 19. 62. 1 Found C, 71. 25; H nmr o 1. 80 (singlet, 5H), o 1. 79 (singlet, lH); o 1. 58 (singlet, lH). Yields in the syntheses of compounds .,...1 through 4 were all in excess of 90 percent. -" Hydrogenations The hydrogenations were performed in a 50-ml flask with a septum adapter. A typical run used between 0. 1 to 0. 25 mmoles of 14 Cp2* TiH 2 in eleven mls of toluene. 33 Mmoles of 2-butyne was frozen into solution at -78 ° C. One atmosphere of hydrogen was put over the solution and the reaction immersed in a -21 ° C temperature bath. Immediately on warming, the orange solution of Cp:TiH 2 turned green. The solution remained green until all 2-butyne was gone then returned to orange. The reaction was monitored by removing 30 µl aliquots periodically throughout the reaction. The relative ratios of alkyne versus alkene were determined by G. C. peak weights. A kinetic study was impossible because the reaction was stir-rate dependent. An average turnover of 2-butyne was 0. 75 equivalents per minute per titanium. Hydrogenations of tolane though much slower were complicated by further reduction to dibenzyl. The reaction of Cp:TiH 2 with 2-butyne under hydrogen at -50° C by nmr was very rapid; the reaction was 75% reacted to form ~ and cis-butene in five minutes. Once the dihydride had completely reacted, further hydrogenation continued slowly until the hydrogen in the nmr tube was consumed. The reaction of 0. 5 mmoles of cp;Ti with hydrogen and 2-butyne was performed by adding 5 mmoles of 2-butyne at -196° C and 1. 70 mmoles of hydrogen. stirring the reaction for three minutes at -78° C and then collecting all volatiles through -196° C traps by toepler pump yielded 1. 22 mmoles of hydrogen. Gas chromatography of the remaining hydrocarbons showed a ratio of nine to one 2-butyne to 2-butene. Dimerization of Terminal Acetylenes The catalytic dimerization of propyne was performed by 15 disgolving 0.12 mmoles of 1 in toluene and adding five mmoles of " propyne at -78 ° C. Warming to 25 ° C resulted in a red solution which was stirred for 24 hours. The solution slowly turned brown. All volatile products were pumped through a -78° C trap. The involatile product at -78° C proved to be 2-methyl-pentene-3-yne. Their and nmr matched those of an authentic sample. 16 16 REFERENCES ~ 1. K. Sonogashira, N. Hagihara, Bull. Chem. Soc. Jap., ~' 1178-1182 ( 966). 2. G. Fachinetti, C. Floriani, J. Chem. Soc. Chem. Comm., 1974, p. 66. 3. A. Nakamura, M. Aotake, S. Otsuka, J. Am. Chem. Soc., ~' 3456 (1974). 4. (a) J. Bercaw, H. H. Brintzinger, J. Am. Chem. Soc., 93, .,..._,..._ 2045 (1971); (b) J. Bercaw, R. Marvich, L. Bell, H. H. Brintzinger, J. Am. Chem. Soc., .,..._,..._ 95, 1219 (1972) . 5. R. Sanner, J. Manriquez, R. Marsh, J. Bercaw, J. Am. Chem. Soc., ~' 8351 (1976). 6. L. S. Meriwether, E. C. Colthup, G. W. Kennedy, J. Org. Chem., 26, .,..._,..._ 5163 (1961) . 7. R. J. Kern, J. Chem. Soc. Chem. Comm., 1968, p. 706. 8. (a) J . Chatt, L . A. Duncanson, R. G. Guy, D. J. Thomson, J. Chem. Soc., 5170(1963). (b) J. Chatt, G. A. Rowe, A. A. Williams, Proc. Chem. Soc., 208 (1957). 9. E. 0. Greaves, C. J. L. Lock, P. M. Mathis, Can. J. Chem., 46, .,..._,..._ 3879 (1968) . 10. A. Nakamura, S. Otsuka, J. Am. Chem. Soc., ~' 1886 (1972). 11. J. L. Thomas, J. Am. Chem. Soc., .,..._,..._ 95, 1838 (1973) . 12. J. E. Bercaw, J. Am. Chem. Soc., .,..._,..._ 96, 5087 (1974) . 17 REFERENCES (continued) ~ 13. If step one is reversible then less than one equivalent of hydrogen would have been consumed. 14. J. M. Manriquez, J. E. Bercaw, J. Am. Chem. Soc., .,...._,..._ 96, 6229 (1974). 15. J.M. Manriquez, D. R. McAllister, R. D. Sanner, J. E. Bercaw, J. Am. Chem. Soc., "'-'"' 98, 6733 (1976). 16. Prepared by reacting CH 2 =C(CH 3 )-C=C8, 2-propenyl acetylide anion with CH 3 1. 18 Chapter 2 The Kinetics and Mechanism of the Zirconoxy Carbene Migratory Insertion Reaction 19 INTRODUCTION ~ Growing interest in the selective conversion of carbon monoxide and hydrogen to hydrocarbons has focused attention on the requisite reduction of CO. The fundamental process of forming a hydrogen bond to CO mediated by transition metal catalysts has been the center of considerable research. l- 3 Migratory insertion of a metal carbonyl hydride to produce a metal formyl has been postulated as the initial step in Fischer-Tropsch hydrocarbon syntheses. To date however, only the reverse reaction has been directly observed. 2 - observable hypothetical ':,, L M - C=O n I H The stoichiometric reduction of CO by Z rH 2 may be an exception. ZrH2 coordinates one equivalent of CO at low temperature producing Zr(H)2 CO. Above -30°C, Zr(H)2 CO decomposes to form transHZrOCH=CHOZrH and HZrOCH 3 • The relative amount of each product depends on the amount of excess Z rH 2 in solution during the decomposition of Zr(H)2 CO. Only the trans-enolate is formed when no excess ZrH 2 is present. HZrOCH 3 isthe major product when four equivalents of ZrH 2 are in solution with Zr(H)2 CO. Both the transenolate and the methoxy-hydride result from CO hydrogenation 1 20 H / 0-C Zr '\ "H H H "/ Zr 95 % C -0 / H / Zr - C=0 '- H 56 % mediated by Z rH 2 • The most straightforward mechanism of this reaction involves migratory insertion of CO into a zirconium hydride bond of Zr(H\C0 to yield a formyl intermediate. Subsequent dimerization of the formyl intermediate with accompanied rearrangement formally yields the trans-enolate. ZrH2 can trap the formyl yielding HZr0CH 2 ZrH which reductively eliminates HZr0CH 3 under hydrogen. / 0 H Zr - C=0 " H 0 II ➔ iH 0-C I \ Zr,_....C __ H ~ Zr+-;C--H ~ zl \ \ if H l ZrH2 /0-CH 2 , Zr '\ H Zr I H H H \ g C-O \ Zr 21 Although intramolecular insertion into a zirconium hydride bond is consistent with ZrH 2 reactions with free CO, the nearly quantitative yield of cis-enolate, HZrOCH=HCOZrH, from the reaction of ZrH 2 with Zr(CO)2 implies a direct reaction of ZrH 2 with Zr(CO) 2 • 3 This observation also raises questions concerning the molecularity of the Zr(H)2 CO decomposition. The mechanistic implications of inter- H I Zr (CO\ + ZrH 2 - - - - I Zr c-o' ~ Zr 'c~ H,, H Zr-C=O + \ H molecular CO reduction in the Zr(H)2 CO system are too broad to discuss in detail here. In general, intermolecular CO reduction by Z rH 2 is viewed as hydride transfer from zirconium to carbon analogous to borohydride reductions of metal carbonyls. 2 Unlike borohydride M = C - 0 - Zr I H I H reductions, the zirconium oxygen bond remains intact. A zirconoxy 22 carbene results instead of a formyl anion. The zirconium zirconoxy carbene, (CO)Zr=CH-O-ZrH, has never bee.n observed, probably due to its inherent instability. In addition to the steric crowding resulting from the proximity of two Cp;zr moieties, zirconium II has only two d electrons to stabilize the two 1T acidic ligands, CO and oxycarbene. Models for this hydride transfer product were sought which would yield stable carbene complexes. Electron-rich mono-carbonyls were the first choice to avoid the stability problems previously discussed. The reactions of ZrH 2 with group VI metallocene carbonyls yield the zirconoxy carbenes, Cp2 M=CH-O-ZrH, (M = W, Mo, and Cr). 3 Hydride transfer is rapid at -80 ° C and the molybdenum and tungsten products are stable above 100 ° C. The closely related niobocene carbenes, RNb=CH-O-ZrH, are perhaps more interesting chemically. Several niobocene substituents readily migrate to the carbene. The mechanism of this rarely observed migratory insertion 4 is investigated herein. The syntheses of several substituted niobocene carbenes are described. 13\ 0 13/ HH '\ 176 2JCH': 6 1 JcH : 'J CC' -- 98 3 JHH' = 11 Hz -- I - 'I Ii1: Ii I: ,1 ! 1 ~ Zr-H - -0 13CH = 13CH0- Zr C=C, /Zr 'H ~ 0 / toluene-Q 7 .1 'I - - I - -0 13CH = 13CHO- Zr-H . '~ 0 13CH 3 Zr/ - JcH'z 20 2 IJCH z 177 'Jee'= 90 3 JHH' = 3.5 Hz ,,.,,,o"' 13 ,,,,,o, Zr 13C=C ?-r 'H~ 'H~ ~ c,., 24 RESULTS ~ 1) Reactions of ZrH 2 with metal carbonyls: Preparation of substituted niobocene zirconoxy carbenes. ZrH 2 reacts with Zr(CO) 2 underhydrogen at 25 ° C to form the cis1 enolate, HZrOCH=CHOZrH in two hours. The H nmr of the cis-enolate, has an olefinic proton upfield from the zirconium hydride peak in con13 trast to the trans-enolate. 1 The C substituted cis-enolate displays the classic ten-line AA'XX' H nmr pattern 5 for the olefinic proton 1 JHH' is three hertz for the cis compared to eleven for the trans-enolate. ZrH 2 reacts rapidly with other metallocene carbonyls. At -78° C, ZrH 2 reacts as soon as it dissolves with Nb(H)CO to form 1 HNb=CH -O-ZrH. The H nmr of HNb=CH -O-ZrH is similar to the nearly isostructural W=CH-O-ZrH. 3 The carbene hydrogen resonance is found at low field at o 11.63. The Cp protons are a singlet at 25°C but separate into the two expected equal intensity peaks about 0. 01 ppm apart below -35 ° C. Bonding in the metallocene equatorial plane to the carbene p-orbital directs oxygen toward one Cp and hydrogen toward the other Cp. 6 The zirconium hydride and niobium hydride signals are at o 5.70 and o -3.14, respectively. HNb=CH-O-ZrH also exhibits a temperature-dependent 13 C nmr spectrum. At 30° C the carbene carbon peak is a broad doublet at o265. 5. The resonance coalesces to a broad lump at -30° C. At -90° C, the signal resolves into two doublets at o272. 8 and li 215. 1 of seven to one relative intensity. Apparently a process not involving any C-H bond making or breaking occurs rapidly which equilibrates two isomers 25 of HNb=CH-O-ZrH. Rotation about the niobium carbon double bond cannot be the dynamic process because the two ppm separated Cp resonances remain constant over the entire temperature range. Of the remaining bonds in the molecule, rotation around the carbon oxygen bond is impossible. Models suggest that two rotomers of the zirconium oxygen bond ought to be stable conformations. In one conformation, the zirconium hydride points away from the carbene-hydrogen bond as in W=CH-O-ZrH. The zirconium hydride is directed away from the carbene hydrogen in the other conformer. ,,--·· \ ,.......H 0 - -Zr I Nb ------ \/ ~ C H H Reactions of ZrD2 with Nb(H)CO show zirconium dihydride reacts as a hydride transfer reagent to CO. Additionally ZrH 2 does not 0 ~ ,,,c Nb 'H + -80 °C Nb= C - 0 - Zr I I I H D D promote the migratory insertion of CO into the niobium substitutent bond in the hydride transfers to niobocene alkyl car bony ls. In all cases the niobocene substituent remains on niobium during hydride transfer. A proton substituted carbene is formed characterized by a I 12 ------~~-----.----"-, . I 11 l 10 9 t I 8 Nb'lC-o_.;:zr I 'H H I H -' I 7 I\ I 6 I 5 .A..J~ • I 4 I 3 I 2 •• ., -' _J,.. M I -1 I 0 e I __ . . . , _ • I I -2 -3 -4 8 I ,....-._ __ N m 27 1 low field resonance in the H nmr at o11-12. Nb= C - 0 - Zr ~•1 I ~ RH H + ~~ NbI = C - 0 - Zr H I I R H The reaction of ZrH 2 with Nb(CH 2 ¢)CO produces ¢CH 2 Nb=CH -O-ZrH isolated in greater than 90% yield. Similar to HNb=CH-O-ZrH, the Cp ligands appear equivalent at higher tempera1 . tures by H nmr. Below 40° C the Cp signals separate by 0. 02 ppm. The Cp* methylresonances are also inequivalent. Similar to (ZrN2 ) 2 N2 , 7 the Cp and Cp * ligands must stagger. Because niobium has two ligands in the equatorial plane, each Cp * is in a different environment. This effect is observed for all niobocene carbenes except the hydride. 28 H I Nb1/C-o-zr I I 2 H ~ OCH 3 13 12 11 10 9 8 7 6 5 4 3 2 0 8 29 A variety of substituted niobocene carbenes have been prepared in this manner, including ire thy 1, ethyl, vinyl, phenyl, benzyl, and zirconoxymethylene. A limitation to this procedure is Nb(Cl)CO. In addition to hydride transfer to the carbonyl, ZrH 2 also exchanges with chloride to form Nb(H)CO and ZrHCl. A mixture is obtained from the Nb - C=O + I ----~),., Cl Nb = C - 0 - Zr I Cl I H I H Nb= C - 0 - Zr I I I H H Cl ,f Nb + C Nb= C - 0 - Zr I 'H H ----------- I H I H ------- Nb'e;P ' --------- Nb= C - 0 - Zr I --,. Cl I I H Cl Cl 1 ZrH 2 reaction with Nb(Cl)CO whose H nmr is consistent with the above scheme. That ZrHCl will transfer hydride is verified independently by the formation of HNb=CH -O-ZrCl from ZrHCl and Nb(H)CO. Zr HF reacts similarly. However, ZrHCl or ZrHF will not transfer hydride to Nb(CH 2 ¢)CO. Apparently the halide reduces the hydricity of the zirconium hydride enough that hydride transfer is unfavorable to the more sterically crowded Nb(CH 2 ¢)CO. 30 2) Synthesis of substituted niobocene carbonyls Most of the substituted niobocene carbonyls used in the carbene syntheses have not been reported. In the case of Nb(CH 3 )CO, the literature preparation is unreproduceable. 8 Methyl lithium and Nb(Cl)CO produce Nb(CH 3)CO in less than one percent yield under a variety of conditions. Nb(CH 3 )CO is prepared in 40 percent yield by slow addition of methyl grignard to Nb(Cl)CO. Similar reactions of both phenyl grignard and benzyl grignard with Nb(Cl)CO failed to yield a significant amount of the desired carbonyl. Nb(<P)CO is prepared by the reduction of Nb<f>Cl with Na/Hg under CO in 40% yield. One equivalent of phenyl grignard reacts with NbC¼ + ,,.C6Hs C 6 H5 MgBr -..-. ► Nb, Cl NbC¾ to produce Nb<PCI. The same synthetic method fails to yield Nb(CH 2 <P)CO. Nb(CH 2 <P)Cl appears to be unstable with respect to Nb(CH 2 <P) 2 and NbC¾. One equivalent of benzyl grignard with NbC½ forms only half an equivalent of Nb(CH 2 ¢) 2 • One equivalent of HCl reacts with Nb(CH 2 <P) 2 yielding half NbC¾ and half unreacted Nb(CH 2 q>) 2 • A scheme similar to the preparation of W(CH <P) 9 from WH also fails. 2 2 2 The deprotonation of Nb(H)CO by LDA followed by reaction with benzyl bromide yields no Nb(CH 2 ¢)CO. Nb(CH 2 cp)CO is foz:med in 60% yield by the reduction of Nb(CH 2 <P) 2 followed by protonation under CO. The cyclic voltamogram of 31 Nb(CH 2 q>) 2 shows the seventeen-electron complex has reversible redox potentials of equal intensity at -2.08V and -0.62V relative to a Ag/Ag+ reference. Coulometry of the most cathodic process proves the reduction is a one-electron transfer to form the expected anion Nb(CH 2 ¢)~ The slow protonation with CF 3 CH 2 OH of dilute solutions of Nb(CH 2 ¢)~ under CO yields Nb(CH 2 ¢)CO. The dilute conditions required and electron transfer Nb(CH 2 ¢) 2 the major by-product indicate a likely side reaction is the protonation of the intermediate hydride to form Nb(CH 2 <P)r and hydrogen. NbH(CH 2 ¢) 2 , similar to NbH 3 , ought to be hydridic enough to reduce protons. Subsequent electron transfer from Nb(CH 2 ¢)~ to the cation Nb(CH 2 ¢)~ is favorable by 1. 46 V based on the cyclic voltamogram. 32 ......-------------;~- ---- > IO. 0 (t) '>E q> I q (f) - ~ 0 0 > (/) > LL :t: ~ ·-C > 0 • C\J I > <D • C\J I -.5 SCAN RANGE: IOOOG FIELD SET: 3292.75G 0 MJDULATION AMPLITUDE: 1.25G MODULATION FREQUENCY : IOOk Hz +.5 MICROWAVE FREQUENCY: 9.2515 GHz ~ Nb~OCH3 ~,.---<§>- OCH3 c., c., 34 ro 0 C\J LO co 35 3) Migratory insertions of the zirconoxy-carbene into the niobocene hydride bond The tautomerization of the eighteen-electron HNb=CH-O-ZrH, formally the product of a-elimination by Nb-CH 2 -O-ZrH, occurs readily at 25 ° C. The sixteen-electron tautomer ought to be unstable Nb= C - 0 - Zr I I I Nb - C H2 - 0 - Zr H H H H I analogous to NbH, lO NbCH 2 CH 3 , ll and NbCL 12 Solutions of the carbene-hydride darken slowly at 25 ° C and decomposition is much faster at higher temperatures. Added ligand can trap the unsaturated niobium III species. Carbon monoxide reacts with the carbene-hydride to form (CO)Nb-CH 2 -O-ZrH. Hydrogen oxidatively adds to form the transient niobium V alkyl dihydride. The resultant alkyl-hydride reduction elimination and further reaction with H2 yields NbH 3 and Zr(H)OCH 3 • 36 The CO reaction with HNb=CH -0- ZrH is conveniently followed by visible spectroscopy as the orange carbene hydride converts to the green (CO)Nb-CH 2 -O-ZrH. Under pseudo-first-order conditions in CO, the rate is first-order in carbene hydride concentration. Firstorder dependence on CO pressure is demonstrated up to five atmospheres. The CO concentration at these pressures is two to forty mmolar 13 compared to the 16 mmolar initial carbene hydride concentration. The rate of dissolving CO must be faster than the CO trapping of Nb-CH 2 -O-ZrH or else the observed rate is only a measure of the CO salvation rate. A limiting CO diffusion rate is unlikely because the reaction rate was insensitive to the stirring speed. The rate expression for the CO mediated hydride migration is: Rate k obs k obs rl [Nb = C - 0 - Z I I I H H H = p co = k1 ~ [HNb = CH - 0 - ZrH] Pco k i + ~ p co sec -1 atm-1 Apparently a-elimination is much faster than trapping by CO. The unlikelihood of an associative process by CO with HNb=CH-O-ZrH implies a rapid pre-equilibrium of carbene insertiona-elimination. The niobium hydride and the carbene hydrogen ought to exchange during this pre-equilibrium. This process is observed directly by nmr at low temperature in the spectrum of HNb=CD-O-ZrD. At -80° C, HNb=CD-O-ZrD displays the characteristic carbene 37 Nb=C-O-Zr + CO_.Nb-CH 2 -O-Zr I I I I I H H H CO H 500nm 600nm 700nm 38 spectrum lacking the zirconium hydride and carbene-hydrogen peaks. Nb= C - 0 - Zr I I H D I D Nb= C - 0 - Zr I I I D H D Even at -80° C the niobium hydride signal slowly loses intensity concommitant with the appearance of the carbene-hydrogen resonance. At room temperature this exchange occurs very rapidly. The resonance for the carbene-hydrogen and the niobium hydride do not average on the nmr time scale. This places an upper limit on the 3 -1 14 exchange rate of 2. 7 x 10 sec . Another direct measurement of this equilibrium is the saturation transfer technique using nmr. 15 If the exchange process between the carbene hydrogen and the niobium hydride occurs on the order of their proton relaxation times, then saturating one resonance will reduce the intensity of the other. This saturation transfer experiment succeeds for HNb=CH -O-ZrH. A more quant iative understanding of the exchange rate is gained by relating the intensities of the saturated peak to the exchanging peak as a function of delay time from the moment of complete saturation. A typical experiment involves saturating one signal, then waiting a specified time before recording the spectrum. The correlation of the difference in intensity of the two peaks as a function of delay time with the magnetization equation o(MA - MB) at (MA - MB) T1 = - ---- - 2kex(M A -MB) 16 yields the exchange rate. MA and MB are the saturated resonance and the exchanging peak magni- 39 tudes, respectively. T 1 is the average relaxation of these two protons. K ex equals the exchange rate. Figure I shows the series of spectra for the saturation transfer of magnetization from the carbene hydrogen to the niobium hydride. Intuitively, saturation appears to have transferred at the time when the niobium-hydride signal has no intensity and the carbene hydrogen has minimum intensity. The reciprocal of the delay time of this spectrum 0 ought to be an estimate of the rate of exchange. At 32. 5 C, the minimum spectrum is near 0. 036 sec. A guess for the rate is therefore ln2(1+0. 036) = 19 sec- 1 . 17 The non-linear least-squares fit of the series of spectra to the magnetization equation yields a rate of 15. 7 sec - i . The rate is identical when either the niobium hydride or the carbene hydride resonance is saturated. Since either methylene proton can a-eliminate from Nb-CH 2 -O-ZrH, the insertion rate equals twice this exchange rate: 31. 4 sec - i . The rate invariance over a fourfold concentration range of carbene indicates the hydride migration is unimolecular. These results, in conjunction with the CO kinetics, indicate a-elimination is roughly 200 times faster than CO trapping at one atmosphere. Above 10 ° C the exchange rate satisfies the condition of being faster than the average proton relaxation time. Exchange rates were measured from 10 ° C to near the decomposition temperature of HNb=CH -O-ZrH at 40 ° C. An Arrhenius rate plot at four temperatures yields the activation parameters of ~Hf = 13. 06 kcal/mole and ~sf = -9. 6 e. u. (Figure 2). ~st is not significantly negative to raise questions of the molecularity. I H , " !j - J #W - \. FROM \. '\ \. 40 .,,,,Q- I 3.0sec 40msec 32 Omsec ec J _A__ ~ I -C, H H Nb- ,....0-Zr H Nb = C to Nb - H 'H IOVIT 7 20msec IOmsec sec 2msec delay time TRANSFER Nb =C_,.,O-ZrH SATURATION ~ 0 3.16 -4.00 -c: -3.00 .::.:. ..__... ' -r- -2.00 3.20 -1.00 40°c 3.24 3.28 3.32 30°c I 200mM Nb-H .67sec 14.5 ± 0.9 sec- 1 30 3 (I /T)xl0 3.36 3.40 3.60 5ImM Nb=C~ti .835sec 15.7 ± 0.8 sec- 1 30 3.52 3.56 51mM Nb-H .835sec 15.2 -+ 0.9sec -I 51mM Nb-H .935sec 40.5±2.4sec 1 30 3.44 3.48 Nb-H .625sec 7.6 ± 0.4 sec- 1 5ImM 20 42 Nb-H ,525sec 3.3 ± 0 .4sec -I EXCHG. RATES Temperature 5ImM CONCENTRAAVEIIAIE TION SATURATED PROTON HNb=CH-OZrH PEAK RELAXATION TIMES I0°C 10 TEMP. 20°c ~ ~ 42 4) Migratory insertions of the zirconoxy carbene into niobocene alkyl bonds The alkyl substituted niobocene carbenes react w th added ligands to form Nb(H)L and Zr(H)OCH =CHR. R equals hydrogen, phenyl, L Nb= C - 0 - Zr I I I RCH 2 H Nb L "-H H + p-methoxy phenyl and -OZrH. In all cases, the enolate has trans geometry. The ligands include carbon monoxide, hydrogen, and tolane. The yield is greater than 95% by nmr when the promoting ligands are hydrogen or tolane. A secondary reaction occurs when CO is the promoting ligand. The initial products, Nb(H)CO and Zr(H)OCH =CH cp react forming a new compound with an AB pattern for the olefin (still trans) and a new Cp resonance. Zr(H)OCH=CH¢ probably transfers hydride to Nb(H)CO, a reaction similar to ZrHCl hydride transfer to Nb(H)CO. The subsequent carbene insertion yields (CO) Nb-CH 2 -0- Zr-OCH =CH cp. ,H 0 o/ N;: 'tt 0-C + z( H ~ -R If" The best kinetic data are obtained from the tolane reactions. The products, Nb-H and Zr(H)OCH=CHR are stable to reaction conditions. I ¢-=-¢ ' 43 Table I Rate ( 50 °C ) Compound Nb= C - 0 - Zr I I CH 3 H I H 7. 71 x 10 Notes -4 sec -1 Only 8 7% Nb----~ was _L. -'+' formed. ¢-= The remaining 13% was not identified. Nb= C - 0 - Zr I I I CH 2 H I -4 -1 2. 09 x 10 sec H ¢ Nb= C - 0 - Zr I I I CH2 D I 2.65x10 -4 sec -1 D ¢ formed. Nb= C - 0 - Zr I I I CD2 H H I 1. 79 x 10 -4 sec -1 cp is formed. 3.37xl0 Nb= C - 0 - Zr I I CH 2 H Only I H 3. 33 x 10 -4 -4 sec sec -1 -1 d - ZrH Nb= C - 0 - Zr I I I CH2 H I ~ 5.07x10 -5 sec -1 Only 50% Nb..--/~ was formed . itf-~ The rate is aisappearance of carbene Cp resonance. 44 8 7 6 5 4 3 2 08 45 A convenient rate is obtained at 50 ° C. By either visible spectroscopy or FT-nmr, the reaction rate is first order in alkyl-carbene for three to five half-lives. The rate is independent of tolane concentration. The rate constants for several alkyl migratory insertions are listed in Table I. The isotope effect for the deuterium-substituted benzylic carbene is 1. 1 7. An inverse isotope effect is observed in the deuterium labeled carbene: :KH/l\ 0 = O. 79. A p value for the benzylic migrations is -1.77. The insertion rates for ¢CH 2 Nb=CH-O-ZrH and CH 3 Nb=CH-O-ZrH were monitored at several temperatures between 32. 5 ° C and 80 ° C. The Arrhenius rate plots are shown in Figure 3. The activation parameters for the benzyl migration are ~Hf = 25. 3 kcal/mole and ~sf = 2. 7 e. u. For the methyl migration, ~Hf = 23. 5 kcal/mole and ~sf = 0.2e.u. C - I.IO 1.50 1.90 ~ 2.30 ..c (1) ~ 2.70 ~ (.) 0 3.10 ..c -- 3.50 ~ (1) 3.90 4.70 4.50 4.30 0 ~ 2 H I 20 0CH 3 I 40 60 H I Nb=C-0-Zr 80 120 minutes 100 140 @ 160 180 I H I I CH 2 H Nb =C-0-Zr 200 220 240 ~ C7) C: k <P cp Nb=C-0-Zr- Nb-CH-0-Zr I I I I I CH 2 H H CH 2 H I I Fig,ure 3 ( 1/T) X 10 3 3.16 3.20 3.24 _ _ _ _ _ . _ _ . . - - L , . _ - - L ,_ k 3.28 ___,___, Nb=C-0-Zr- Nb-CH-0-Zr I I I I I CH 3 H H CH 3 H -11.0 .____.___....___ _.___.....____.......___ ____.____ ___.____ __.__ 2.80 2.84 2.88 2.92 2.96 3.00 3.04 3.08 3.12 -16.0 -15.0 ......... ~ -14.0 -.._ -13.0 -12.0 -11.0 ~ -1 48 DISCUSSION ~ The facile formation of several substituted niobocene carbenes has enabled a systematic study of the carbene insertion reaction. In the most direct observation of the carbene insertion, HNb=CD-0-ZrD isomerizes to DNb=CH-0-ZrD at low temperature. Carbon monoxide traps the intermediate tautomer to form (CO)Nb-CH 2 -0-ZrH. At higher temperatures, carbene insertion into the niobocene alkyl bond yields an intermediate which ,8-eliminates faster than ligand trapping. Only zirconoxy enolates are produced along with substituted niobocene hydrides. Saturation transfer experiments verify the rapid rate of hydride migration relative to alkyl migration. In the only case where Nb= C - 0 - Zr I H I H I H co_,,_;,. Nb - CH - 0 - Zr Nb - CH2 - 0 - Zr _ _ 2 I I I C H H Ill 0 Nb= C - 0 - Zr - - ~ Nb - CH - 0 - Zr - L I I RCH 2 H I H Nb= C - 0 - Zr I cf> I H I H I RCH 2 I cf> L + 'tt H -CH-0-Zr t I ► Nb H L 0-C., z(, H \\C-R H'" ? H ,:3-hydrogens would not be present on the intermediate insertion product, no migration of phenyl to carbene is observed. The order of migratory aptitudes of niobocene substituents is hydride >> methyl > benzyl » phenyl. 49 Considering the relative stabilities of the species involved in hydride migration there must be at least a two kcal energy difference between the tautomers. Nb-CH 2 -O-ZrH is never observed by nmr although the tautomerization occurs radily at 25 ° C. There could be as Scheme I Nb= C - 0 - Zr I H I H I H ].t hydride [ migration Nb,.:-.C-O'Zr ~::;;;:.=-;:,-=,~~ ~ &-elimination 8 ' }J Nb - CH 2 - 0 - Zr I H much as a sixteen kcal difference between tautomers if the a-elimination has no activation energy. The hydride migration appears to be a onestep process, formally breaking the niobium-carbene 1r bond and the niobium hydride bond to form a carbon-hydrogen bond. The estimated bond strengths for the niobium carbene TT bond and the niobium hydride bond are considerably more than the measured activation energy of thirteen kcals. A reasonable carbene 1T bond strength is eighteen kcals based on activation energies for carbene rotation in Ta(CH<p)(CH 2 <P) and Nb(CHCMe 3 )Cl. lS A lower limit for theniobium hydride bond is sixty kcals based on some first-row metal hydride bond strengths. 19 The new carbon hydrogen bond of the methylene tautomer must be formed to a significant extent in the transition state to lower the energy requirement from 78 to sixteen kcals. 50 The mechanism of alkyl migration is masked because the migration product cannot be trapped with added ligands. Additional steps produce zirconoxy enolates. Scheme Il shows the simplest Scheme II Nb= C _ O-Zr I I <PCH 2 H I H migration slow Nb-CH-O-Zr ~elimination L I fast I </>CH 2 H fast mechanism for the reaction . Ligand trapping of NbH must be faster than the rate limiting step because the rate law is independent of ligand concentration. ,8-Elimination cannot be rate determining since no intermediates are observed by nmr. A fast pre-equilibrium prior to rate limiting ,a-elimination is unlikely because this requires repeated carbon-carbon bond breakage and formation. By process of elimination alkyl migration is the slow step for the mechanism of scheme II. The observed inverse isotope effect for the deuterium substituted carbene insertion is consistent with rate limiting migratory-insertion. An inverse isotope effect indicates the carbene p-hybridization increases in the transition state. 20 Neither {3-elimination nor ligand trapping satisfy this criterion. Another mechanism which satisfies some of the above conditions is the addition of benzyl radical to the carbene followed by niobium benzyl bond cleavage. This alternative seems possible considering the 51 + Nb= C - 0 - Zr I I ¢CH 2 H I - H Nb - CH - 0 - Zr I I--+ I ¢CH 2 CH 2 ¢ H Nb - CH - 0 - Zr + ¢CH 2 • I ¢CH 2 I H 25 kcal activation energy relative to a cobalt benzyl bond strength of 19 kcals. 21 However, a crossover experiment indicates no niobium benzyl bond homolysis occurs during the reaction. A mixture of ¢CD2 Nb=CH-O-ZrH and ¢CH 2 Nb=CD-O-ZrD produce no Zr(H)OCH =CH ¢ . The primary thermodynamic difference between the alkyl migration and the hydride migration is the relative strengths of the reacting bonds. The bond energy difference for hydride migration Nb=C-O-Zr I I I H H H Nb= C-O-Zr I I I CH 3H H --:> ► Nb-CH-O-Zr I I Niobium hyjride bond broken Nb-CH-O-Zr Niobium alkyl bond broken CH 3 Carbon carbon bond formed H I H I H Carbon hydrogen bond formed equals -35 kcal: sixty kcals for the niobium hydride bond broken minus 95 kcals for the carbon hydrogen bond formed. The bond energy difference for alkyl migration is -45 kcal: 35 kcal for the niobium 52 alkyl bond broken minus 80 kcals for the carbon-carbon bond formed. First-row transition metal alkyls are roughly 25 kcals weaker than the corresponding hydrides. Though these bonds may be stronger for second-row transition metals, the differences are probably the same. These bond energy differences must be revised to obtain heats of reaction. Other contributions to the overall energy include carbene 7T bonds broken during migration as well as energy changes of carbon oxygen and carbon hydrogen bonds as a result of rehybridization. 22 However, these factors are probably equal for each migration so that the net difference is ten kcals. This reacting bond energy difference implies alkyl migration is less endothermic than hydride migration. These ground state thermodynamic effects fail to explain the series of migratory aptitudes: hydrogen » methyl > benzyl » phenyl. The most endothermic migration is the fastest. If the mechanisms are similar, the relative migratory aptitudes must be a reflection of some ligand property which affects the three-center transition state. Indeed, alkyl migrations appear to be concerted three-center reactions analogous to the hydride migration. The 53 kcal sum of the niobium n bond plus the niobium alkyl bond is much more than the measured 25 kcal activation energy. The trend of migratory aptitudes is significantly different for carbene insertions compared to carbonyl insertions. Hydride migration to CO has never been directly observed, 23 although the relative rates of methyl versus benzyl migration are nearly the same for CO and 24 oxycarbenes. From what little is known about isonitrile migratory insertions, hydrides and alkyls migrate at more comparable rates. 25 53 The presence of substituents on the carbene (zirconoxy) and isonitrile (methyl of phenyl) increase the electrophilic character of the inserting groups and enhance their rate of migratory insertion. Nb(CH 3 )CO does not undergo migratory insertion, 26 whereas the carbene-methyl, (CH 3 )Nb=CH-O-ZrH, readily inserts. A similar effect rationalizes the lewis acid promotion of CO insertions. 27 The failure of a variety of nucleophilic carbenes, Ta(CHR)CH 2 R, to insert 28 supports the notion of the electrophilic nature of carbene migratory insertions. The electrophilic reactivity of an oxycarbene could explain the trend of migratory aptitudes. The resonance structures of oxycarbenes display interactions stabilizing the positively charged carbene carbon. 8 Nb= C - 0 - Zr I R I H I H ~ EB 8 Nb - C - 0 - Zr I R I H I H ~ EB Nb - C = 0 - Zr I R I H I H If the migrating group moves with its pair of electrons to the electrophilic carbene analogous to CO insertions, 29 electron donating groups should migrate easier. The negative p value measured for benzyl migrations is consistent with this supposition. Electronegativity is perhaps a more general measure of the electron donating ability of a migrating group. Less electronegative groups are more electron donating. The order of electronegativity increases in the order: hydrogen, methyl, benzyl, phenyl. 30 The trend of migratory aptitudes to the zirconoxy carbene correlates with this series of group electronegativities. 54 The carbene migratory insertion is structurally similar to several organic rearrangements. In the Wagner-Meerwien and Pinacol rearrangements, a migrating group moves to an adjacent positively charged carbon. 31 Migratory aptitudes in these reactions are usually hydrogen > aryl > alkyl, 32 The degree of s-character in the bond of the migrating atom correlates with the general order of migrating groups. This correlation is usually rationalized on the basis of the orbital overlap of the migrating group to the molecular framework of the transition state. In general, hydrogen bonds are stronger than aryl bonds which are stronger than alkyl bonds. Presumably, the transition state is stabilized the most by the stronger hydrogen bond. Electronegativity and hybridization seem to be interrelated factors determining the stability of the transition state. Electron donating groups cannot stabilize the transition state if there is no overlap with the migrating group. Good overlap of a strongly electronwithdrawing group to the electrophilic carbene destabilizes the transition state. These considerations probably apply to most electrophilic insertion reactions. In this vein, one expects similar trends for carbonyl insertions across the transition metals compared to carbene insertions. Carbonyl insertion into transition metal hydrides are unprecedented, although ZrH 2 CO appears to react by carbonyl insertion into the hydride. 33 The most hydridic of the transition metal hydrides, zirconium hydride is most likely to migrate to CO. Presumably, the hydridic niobium hydride 34 also shows a greater tendency to migrate to carbenes than later transition metal hydrides. For the acidic hydrides 55 of group VIII, one would expect slower carbene-hydride migratory insertions, at least relative to alkyl migrations. Hopefully, the rate correlation of migratory insertion to the electronegativity of the migrating group will be clarified by studies of some iron oxycarbene insertions. 56 EXPERIMENTAL ~ All manipulations were performed in an inert atmosphere using a glove box or vacuum techniques. Ether solvents such as glyme and THF were distilled from sodium benzophenone ketyl. All other solvents including nmr solvents were purified by distillation from titanocene. 35 Hydrogen, argon and nitrogen were passed through MnO on vermiculite 36 and 4 .A molecular sieves. Carbon monoxide (MCB) was used without purification. Tolane was recrystallized from octane. ZrH and NbC½ 37 were prepared by literature methods. ZrHF was 2 prepared from ZrH 2 and ZrF 2 analogous to ZrHCl. 38 1 H nmr spectra were obtained using a Varian EM 390 and a JEOL FX90Q spectrometers. 13 C nmr were also measured on the JEOL. Infrared spectra were recorded on a Beckman 4240 spectrophotometer. Esr spectra were obtained on an X-band Varian E-line Century Series spectrometer. Visible spectra were observed using a Cary 14 spectrophotometer. Cyclic voltamograms were conducted on a PAR 174 A Polarographic analyzer. Coulometry was analyzed on a PAR Potentiostat 173. 1) Preparation of Nb(CH 3)CO 2. 0 Grams of NbClCO (7. 0 mmoles) was suspended in 140 mls of warm toluene. 2. 0 Mls of 3. 2 M methyl magnesium bromide (6. 4 mm was added in the following manner. Ten percent was added at once followed by the slow addition over a period of fifteen minutes. The toluene was then immediately removed by vacuum transfer. Several pentane extractions followed by treatment with methyl iodide yielded a 57 green solid after solvent removal. The green residue was extracted with pentane to separate Nb(CH 3 )CO from NbICO. Sublimation yielded 0. 8 gram of Nb(CH 3 )CO, 43% yield. 2) Preparation of Nb(p)CO. 2. 0 Grams of NbC12 (6. 8 mmoles) was reacted with 2.1 mls of 3. 2 M phenyl magnesium bromide (6. 7 mmoles) in THF at 25 ° C. After eight hours, all the NbC4 had dissolved and the brown solution had considerable white precipitate. All solvent was removed by vacuum transfer and 70 mls of toluene added. A warm filtration followed by solvent removal left green-brown crystals of presumably Nb(¢ )Cl. After redissolving in toluene and cooling to -80° C, four mls of one percent Na/Hg was added via syringe. All gas was removed and the reaction mixture was stirred at 25 ° C under one atmosphere of CO for eight hours. The resulting green solution was decanted from the amalgam and filtered prior to solvent removal. The green solid was sublimed at 120° C to yield 0. 9 gram of Nb(¢)CO, 40 percent based on NbC4. 3) Preparation and Characterization of Nb(CH 2 CJf 4X) 2 , X = H and OCH 3 2. 0 Grams of NbC½ (6. 8 mmoles) was suspended in 100 mls of THF. 17 Mls of 0.92 M benzyl magnesium chloride (15.6 mmoles) was added and the reaction stirred for six hours. The resultant purple solution was quenched of excess grignard with one ml of deoxygenated methanol. All solvent was immediately removed by vacuum transfer. 100 Mls of benzene was added to the mixture and warmed to 40° to dissolve the product. The solution was filtered and the benzene volume reduced to 25 mls. The resultant purple crystals were redissolved by 58 warming before cooling to yield 2. 4 grams of Nb(CH 2 cp) 2 • A further recrystallization from toluene separated more impurity from the compound. Final yield was 2. 0 grams of dibenzyl niobocene. The same methodology yielded Nb(CH 2 C 6H4 OCH 3 ) 2 in a similar yield. The esr of these compounds displayed a ten-line spectra with g values of 2. 01 for both Nb(CH 2 C 6H5 ) and Nb(CH 2 C6 H4 OCH 3 ) 2 • Cyclic voltamograms of the dibenzyl compounds were conducted in THF with 0.1 M tetrabutyl ammonium tetrafluoroborate. The reference potential was Ag/ Ag+ (0. 01 M in CH 3 CN). Peak to peak separations were 90 mV at 200 mV / sec scan rate. These were considered reversible based on the identical peak to peak separation of added ferrocene at +0. 2 V. 39 The redox potentials for Nb(CH 2 C6 H 5 ) 2 were -2. 02 volts for reduction and -0. 62 volt for oxidation relative to Ag / Ag+. For Nb(CH 2 C~ 4 OCH 3 ) 2 the potentials were -2.12 V reducing and -0. 79 V oxidizing. Coulometry consumed 1. 01 electrons per equivalent for the reaction of Nb(CH 2 C 6H5 ) 2 • 4) Preparation of Nb(CH 2 cp)CO 810 Mgs of Nb(CH 2 cp) 2 (2 mmoles) are dissolved in 200 mls of THF. Four mls of one percent Na/Hg was added and stirred for two hours when the solution was dark brown. The solution of dibenzyl anion was separated from the amalgam by decantation and filtration. A solution of 0.14 mls of CF 3 CH 2 OH in twenty mls of THF was added slowly over a period of twelve hours to the anion solution under one atmosphere of CO. The solvent was removed. The benzyl carbonyl was extracted fromthe brown residue with 70 mls of warm octane. A greP.n powder was isolated which was washed with cold pentane. 59 375 Mgs of Nb(CH 2 ¢)CO were obtained from the sublimation of green powder at 90° C; 5 percent yield. 5) Preparation of RNb=CH-O-ZrH; General Syntheses 1. 0 Grams of NbHCO (4 mmoles) and 1. 6 grams of ZrH 2 were reacted in thirty mis of toluene. This orange solution was filtered and toluene removed by vacuum transfer. The orange residue was suspended in pentane and filtered to yield 2. 2 grams of HNb=CH-O-ZrH: 90 percent yield. 5) Preparation of RNb=CH-O-ZrH; General Syntheses 1. 0 Grams of NbHCO (4 mmoles) and 1. 6 grams of ZrH 2 were reacted in thirty mis of toluene. This orange solution was filtered and toluene removed by vacuum transfer. The orange residue was suspended in pentane and filtered to yie ld 2. 2 grams of HNb=CH-O-ZrH: 90 percent yield. All carbenes were prepared in this manner and stored at -20 ° C. 6) Preparation of (CO)Nb-CH 2 -O-ZrH 1. 0 Gr (1. 62 mmole) of NHb=CH-O-ZrH was reacted in 35 mis of toluene with one atmosphere of CO for eight hours at 45 ° C. After cooling and filtration, the volume was reduced to about eight mis. The green precipitate was recrystallized and washed with pentane. 825 Mgs of (CO)Nb-CH 2 -O-ZrH was isolated for a 80 percent yield. 7) Preparation of Nb-H I <P-=::.-<P 300 Mgs of NbH 3 34 (1. 33 mmoles) and 230 mgs of tolane (1. 3 mmole) were heated to 65° C in toluene for two hours. The reaction was cooled to 35 ° C, then filtered. Solvent removal resulted in a 60 brown oil. A yellow powder results from the repeated addition and removal of pentane by vacuum transfer. This powder was recrystallized from 40 mls of octane, yielding 400 mgs of Nb-H in 75% yield. cf>-==- -cp 1 The H nmr spectrum is identidcal to the niobium product of the alkyl carbene migratory insertions. 8) Preparation of Zr(H)OCH=CHp(trans) 700 Mgs of cpCH 2 Nb=CH-O-ZrH (1 mmole) and 175 mgs of tolane (1 mmole) were reacted in toluene at 25 ° C for two days. Toluene was removed by vacuum transfer. Several pentane addition-removal cycles yielded a tractable powder. Most of the zirconium enolate was extracted by filtration of a 15 ml pentane solution at 0° C. The resultant yellow solution was reduced to three mls and cooled to -80 ° C to form colorless crystalline product. Cold filtration yielded 300 mgs of Zr(H)OCH =CH cp in 62 percent yield. A mixture of the cis and trans-zirconium phenyl enolates was prepared by reacting NaOCH=CHcp with ZrHCl. 80 Mgs of NaH was suspended in 80 mls of THF. 0. 35 Mls of phenyl acetaldehyde was added at -30 ° C and reacted for eight hours at 25 ° C. This solution was filtered onto 800 mgs of ZrHCl and stirred four hours. The THF was removed and replaced with pentane. A yellow powder was isolated after filtration whose nmr showed both the cis and trans-enolate hydrides. 9) Kinetics of the hydride migration mediated by CO A typical experiment involved 40 mgs of HNb=CH-O-ZrH dissolved in 40 mls of benzene (16. 2 mmolar) in a ten mm quartz cell with a 50 ml total volume. The reaction was run at 22° C under a 61 measured CO pressure. The visible spectrum was recorded periodically as the reaction proceeded. The reaction was stirred between recordings but little rate deviation occurred without stirring. Plots of ln(A 00 -A) at 600 nm versus time were linear for three half lives. Rate plots of CO pressure indicated first order CO dependence. 10) Kinetics of hydrogen exchange using saturation transfer The pulsing technique employed is described by Freeman. 40 The relaxation times for the protons of HNb=CH-O-ZrH were measured using the inversion recovery program of the JEOL FX90Q. 41 Saturation transfer spectra were obtained using a pulse sequence program written by Perkins. 42 It involves the timing for saturating a specific resonance and delaying the observation pulse a specified interval. A three parameter non-linear least squares fit 43 of the data to the magnetization equation was performed on a Commodore Pet computer. 11) Kinetic measurements of alkyl migrations The rates of alkyl migrations were followed by FT-nmr. Less than 45 degree observation pulses were used. Repetition rates were greater than twice the relaxation times of the niobium resonances. Reaction temperatures were maintained by the probe temperature controller and observed to be constant to within one degree by the peak separations of ethylene glycol. 44 A typical reaction involved 30 mgs of cpCH 2 Nb=CH-O-ZrH (42. 4 µmoles) and 30 mgs of tolane (168 µmoles) dissolved in 0. 45 mls of d6 benzene. As the reaction proceeded, the niobium carbene Cp resonance lost intensity as the Nb-H Cp signal appeared. The plot of the carbene Cp decay I ¢-==-¢ 62 showed first-order behavior for better than three half lives. The rate for benzyl migration was independent of tolane concentration over a fifty-fold range. Most migration rates were studied with only a threefold excess of tolane yet showed first-order carbene decay kinetics for at least three half lives. 12) Benzyl migration crossover experiment 25 Mgs each of ¢CD2 Nb=CH-O-ZrH and ¢CH 2 Nb=CD-O-ZrD were reacted with 30 mgs of tolane in d6 benzene. The reaction was monitored by nmr. The olefinic proton at 5. 73 o appears as a singlet for Zr(H)OCH =CD¢. No doublet is observed for Zr(H)OCH =CH¢. Anal. Calcd. %C, 64.65; %H, 7.00; %Zr, 12.92 ; %Nb, 13.16; 4.96 s d C5ff~ Nb-CH _z -<P 2.73 4.98 s C5ff5 1. 97 s C5(Cff 3)5 d 6 benzene 1. 95 s C5(Cff3)5 (C 5H 5)2Nb( CH 2 <P) =CH -0- Zr(H) (C 5Me 5)2 5.90 s 11. 72 s =CH-OZr-H ? s Found: %C, 60. 85; %H, 7 .10; %Zr, 14. 21. 5.01 s C5!!~ Nb-Cff 3 4.97 1.99 s C5(C!f3\ s 5.70 s ZrH C5ff5 11. 63 s =CH-O- Anal. Calcd. %C, 61.02; %H, 7.20; %Zr, 14.48; (C 5H5)2 Nb(CH 3)=CH-O-Zr(H)(C 5Me 5)2 Found, %C, 58.76;%H, 6.80 d 6 benzene -3.14 s Nb-H - 5.02 s C5!!_,; Anal. Calcd. %C, 60.46;%H, 7.04 5.01 s C5!!_,; d6 benzene (C 5H 5)2 Nb(H) =CH -O-Zr(H)( C 5Me 5)2 JHH' = 11 Hz Chemical Shift Group Solvent Compound O') c..., 7.70 3.45 6.02 1.93, 2.02 d s s s H' !! OC!:! 3 Zr-H C5(C!!3)5 JHH = 8 Hz 6.87 d Found:%C, 61.53;%H, 6.81 ; Jt)x,- JHH = 11 Hz 11.76 7.28 m C6!!5 s 1. 99 d Nb-CH!!-¢ JHH = 11 Hz Nb=CH-O- 2.73 d Nb-C!!H-cp 2.68 4.96 s C5!i; d 4.98 s C5!!5 Nb-CH _2 - 7.28 m c~5 Anal. Calcd: %C, 63. 65; %H, 6. 98: 2.03 s C5(C!!3)5 4.99 1. 96 s C5(C!!3)5 s 6.05 s Zr-H - C5!!5 11. 87 s =CH-O- d6 benzene Chemical Shift Group Solvent (C 5H5)2 Nb(CH 2 C 6H4OCH 3)=CH-O-Zr(H)(C 5Me 5)2 (C 5H 5)2 Nb(CH 2 cp )=CD-O-Zr(D)( C 5(CD3)5)2 Found: %C, 61. 64; %H, 6. 56; %Zr, 12. 48; %Nb, 12.85. Compound 0) ~ 7.2 12.23 6.90 2.01 5.12 5.33 5.93 6.10 6.65 1. 94 1. 99 2.04 2.10 m s s s s s s s s s s s s Nb-CJ:! 5 Nb=CH-0Zr-H Cs(C!!s)s (C 5H 5 ) 2 Nb(CH 2 0Zr(H)(C 5Me 5 ) 2 )=CH-0Zr(H)(C 5Me 5 ) 2 d 6 benzene C 5!! 6 Nb-CH _2 -0 ZrH ZrH' Cs(C!!s\ Anal. Calcd: %C, 61. 99; %H, 7. 50; %Zr, 18.11; Found: %C, 61. 85; %H, 7. 39; %Zr, 18. 33. 5.23 s Cs!!; Cs!!; 5.14 s d 6 benzene C 5!! 5 Chemical Shift (C 5H 5 ) 2 Nb( C 6H 5 ) =CH -0- Zr(H)( C 5Me 5 ) 2 Group Solvent Compound C}l O') 2.65 7.23 11.30 1. 90 1. 97 4.39 1. 82 7.23 4.40 1. 84 6.88 7.17 3.45 d m s s s s s m s s s -CH 2 -Ca!! 5 Nb=CH-OC5(C!!3)5 C5(C!!3)5 CH 2 -C~ 4 -O d d Nb-CH _2 - (C 5H 5)2 Nb(CH 2 Cafl 4OCH 3)CO Found:%C, 63.04;%H, 5.15;%Nb, 27.00. Anal. Calcd: %C, 63.17; %H, 5. 01; %Nb, 27 .15; OCJ:! 3 Nb-CH _2 - d 6 benzene C 5!! 5 CH 2 -Ca!! 5 Nb-CH _2 - d 6 benzene C 5!! 5 5.06 s C5!!5 (C 5H5) 2 Nb(CH 2 Cafl 5)CO 4.97 s d 6 benzene Cl! 5 JHH = 8 Hz JHH = 10 Hz Chemical Shift (C 5H5) 2 Nb( CH 2 C 6H 5) =CH -0- Z r(I)( C 5M e 5)2 Group Solvent Compound en en Found: %C, 70.80; %H, 7.55. Anal. Calcd: %C, 69.80; %H, 7.95; trans 6.23 5.73 7.52 7.08 s d d m ZrH -O-CH= =C!!,P -CJ:I5 7.5 m 1.93 -0.28 s 4.85 s d6 benzene C 5(C!! 3) 5 -CJ! _5 Found: %C, 71.67; %H, 5.37, %Nb, 23.05. (C 5Me 5) 2 Zr(H)OCH=CHcf> Nb-H s 2.00 s C5(C!!3)5 d6 benzene C 5!! 5 5.20 5.20 4.68 s s s ZrH Nb-CH _2 -0- Anal. Calcd: %C, 71. 64; %H, 5. 26; %Nb, 23. 09; (C 5H5)2 Nb(H)(CJI 5)2 C2 Found:%C, 59.36;%H, 6.94;%Zr, 14.33 Anal. Cald: %C, 59.70; %H, 6.58; %Zr, 14.17 d6 benzene C 5!! 5 7.05, 7.70 m CJ:!5 (C 5H5)2 Nb(CO)CH 2 OZr(H)(C 5Me 5)2 4.53 s d6 benzene CH 5_5 Chemical Shift (C 5H5)2 Nb(CJI 5)CO Group Solvent Compound ~ ~ (C 5Me 5\ Zr(H)OCH = CH 2 (C 5Me 5)2 Zr(H)OCH =CH</> I Compound cis d6 benzene d6 benzene Solvent 6.43 6.69 5.10 7.08 s d d m ZrH -OCH= =C!!c/> -Ca!:!5 d =CH 2 c JHH~ = 5 Hz d =CH 2 t 3.89 dd JHH' = 17 Hz 6.77 s ZrH 0-CH= JHH' = 7 Hz 4.10 6.10 s C5(C!!3)5 1. 92 1.88 s Chemical Shift C5(C!!3)5 Group co C1) 69 References 1. J.M.Manriquez, D.R.McAlister, R. D.Sanner, J.E.Bercaw, J. Amer. Chem. Soc., 100, 2716(1978) ~ 2a. C.P.Casey, M.A.Andrews, D.R.McAlister, J.Amer.Chem.Soc., 101, 33 71 (1979) """"'-"b. C.P.Casey,S.M.Neuman, J.Amer.Chem.Soc., !QQ_, 2554(1978) c. W.Tam, W.K.Wong, J.A.Gladysz, J.Amer.Chem.Soc., .,.._,...,_,.._ 101, 1589 (19 79) 3. P. T. Wolczanski, R. S. Threlkel, J.E. Bercaw, J. Amer . Chem. Soc. , 101, 218(1979) ~ 4a. F. D. Mango, I. Dvoretzky, J. Amer. Chem. Soc., 88, 1654(1966) """ b. H. D. Empsall, E. M. Hyde, M. Markham, W. S. McDonald, M. C. Norton B. L. Shaw, B. Weeks, J. Chem. Soc. Chem. Comm., 589(1977) 5. J. W. Emsley, J. Feeney, L. H. Sutcliffe, "High Resolution Resonance Spectroscopy", Vol. l, Pergamon Press, Oxford, 1965 p. 396 6. J. L. Lauher, R. Hoffmann, J. Amer. Chem. Soc., 98, 1729(1976) ~ 7. J.M.Manriquez, D.R.McAlister, E.Rosenburg, A.M.Shiller K. L . Williamson, S. Chan, J.E. Bercaw, J. Amer. Chem. Soc., 100 , 3078(1978) ..,.,.._,._ 8. S.J.Fredericks, J.L.Thomas, J.Amer.Chem.Soc., .,._,._,,,_ 100, 350(1978) 9. B. R. Francis, M. L. H. Green, T. Luong-thi, G. A. Moser, J.Chem.Soc.Dalton, 1976 1339 ~ 10. F.N.Tebbe, G.W.Parshall, J.Amer.Chem.Soc., ~' 3793(1971) 11. L.J.Guggenburger, P.Meakin, F.N.Tebbe, J.Amer.Chem.Soc., 96, 5420(1974) ~ 12. A.N.Nesmeyanov,D.A.Lemenovskii, V.P.Fedin, E.G.Perevalova, Dokl. Akad. Nauk. SSSR, .,..._,...,.... 245, 609(1979) 13. International Critical Tables, Volume 3 p. 265 70 14. D. E. Leyden and R.H. Cox, "Analytic Applications of NMR", Wiley and Sons 1977 p. 283 15a. S. Forsen, R. A. Hoffmann, J. Chem. Phys., iQ, 1189(1964) b. F. W.Dahlquist, K.J.Longmuir, R.B.DuVernet, J.Magn.Res., 1 7 , 406(1975) -'""' . 16. D. A. Kooistra, J. H. Richards, S. H. Smallcombe, Org. Mag. Res., 1979, 000 ~ 1 7. The rate is actually ln 2 times the reciprocal time of the zero intensity spectrum. T. C. Farrar, E. D. Becker, "Pulse and Fourrier Transform NMR", Academic Press, New York 1971 p. 20 18. R.R.Schrock, L.W.Messerle, C.D.Wood, L.J.Guggenberger, J. Amer. Chem. Soc., 100, 3793(1978) ~ 19. J. W. Rathke, H. M. Feder, J. Amer. Chem. Soc., 100, 3625(1978) ~ 20a. R. W. Alder, R. Baker, J.M. Brown, ''Mechanism in Organic Chemistry", Wiley Interscience New York 1971 p.19 b. P. Brown, R. C. Cookson, Tetrahedron, -'""' 21, 199'3(1965) 21. J. Halpern, F. T. T. Ng, G. L. Rempel, J. Amer. Chem. Soc., 101, 7124(1979) .,.._,..,_,.__ 22. S. W. Benson, ''Thermochemical Kinetics", John Wiley and Sons, New York 1976 23.J.P.Collman, S.R.Winter, J.Amer.Chem.Soc., .,..._,._ 95, 4089(197'3) 24a. A. Wojciki, Adv.Organomet.Chem., .,..._,._ 11, 88(1973) b. F. Calderazzo, Angew. Chem. Int. Ed. Eng., 229(1979) 25. P. T. Wolczanski, J.E. Bercaw, J. Amer. Chem. Soc., .,.._,.__,.._ 101, 6450(1979) 26. E.E.H.Otto, H.H.Brintzinger, J.Organomet.Chem., 1979170(3) 373 ~ 71 27. S. B. Butts, E. M. Holt, S. H. Strauss, W.W. Alcock, R. E. Stimson, D. F.Shriver, J.Amer.Chem.Soc., 101, 5864(1979) ~ 28a. R.R.Schrock, P.R.Sharp, J.Amer.Chem.Soc., .,..,_,..,,,... 100, 2389(1978) b. The only exception is : P. R. Sharp, R.R. Schrock, J. Organomet. Chem. , 19 79 .,..,_,..,,,... 1 71 (1) 43 29. A. Davison, N. Martinez, J. Organomet. Chem. , 1974 7i, Cl 7 30. J.E. Huheey, "Inorganic Chemistry: Principles of Structure and Reactivity", Harper and Row, New York 1972 p. 167 31. Y. Pocker, "Molecular Rearrangements", Ed. by P. de Mayo, Interscience Publishers, New York 1963 32. J. March, "Advanced Organic Chemistry", McGraw-Hill New York 1968 p. 788 33. J.M. Manriquez, D.R. McAlister, R. D. Sanner, J.E. Bercaw, J. Amer. Chem. Soc., .,..__,.._ 98, 6733(1976) 34. J.A.Labinger, K.S.Wong, J.Organomet.Chem., 1979170(3) 373 .,..,_,..,,,... 35. R.H.Marvich, H.H.Brintzinger, J.Amer.Chem.Soc.,E, 2046(1971) 36. T. L. Brown, D. W. Dickerhoof, D. A. Bafus, G. L. Morgan, Rev. Sci. lnstrum., .,..__,.._ 33, 491(1962) 37. C.R. Lucas, Inorg. Syn., Vol. 16 p. 107 38. D.R. McAlister Private communication 39. R.Alexander, A.J.Parker, J.H.Sharp, W.E.Waghorne, J.Amer.Chem.Soc., ~ ' 1148(1972) 40. G. A. Morris, R. Freeman, J. Magn. Resn., .,..__,.._ 29, 433(1978) 41. JEOL FX90Q Operation manual Chapter 6 42. T. Perkins Private communication 43. W. E. Hull, Ph.D. Thesis Harvard Uni'rersity 1975 44. Varian EM 390 Users manual p. 21 72 Chapter 3 Synthesis of Bis-Pentamethylcyclopentadienyl Niobocene Compounds 73 INTRODUCTION ~ The organometallic chemistry of niobium has only recently been investigated. For reasons of stability and availability, most research has been directed toward an understanding of the niobocene moiety. These niobium complexes have displayed rich and diverse chemistry. l- 3 , 5 - 9 Niobocene trihydride, Cp NbH , catalyzes exchange 2 3 of molecular hydrogen with d6 -benzene. 1 The absence of pairwise exchange led Parshall to propose a mechanism involving reversible Cp2 NbH 3 Cp2 NbH -HD oxidative addition of an aryl hydrogen bond to coordinatively unsaturated niobocene hydride. 2 The very reactive niobocene hydride is generated by reductive elimination of H2 from the trihydride. In the absence of added ligands, this compound appears to react with its own carbon hydrogen bonds to form niobocene dimer, 3 somewhat analogous to titanocene chemistry. 4 With added ligands, niobocene hydride is trapped and several interesting products have been isolated by this 74 procedure. Carbon monoxide traps niobocene hydride to form Cp Nb(H)CO. 2 Both acetylenes and olefins react with Cp NbH to yield 2 2 3 acetylene 5 and olefin 6 hydride complexes. These rarely observed complexes readily undergo migratory insertion to form vinyl and alkyl complexes. Similar chemistry has been observed for the tantalum analogs of these compounds. The trihydrides are also hydrogenation catalysts for olefins. In view of the observed sequence of events leading from Cp2 NbH 3 to Cp2 Nb(L)C 2 H5 in the reaction with ethylene, the mechanism shown below was proposed for hydrogenation. 75 Another possibility was proposed by Kochi who observed the rapid low temperature reaction of Cp NbH with olefins. 7 Cp NbH was prepared 2 2 2 2 by hydrogen atom abstraction from Cp2 NbH 3 with t-butyl peroxide. 8 Esr signals consistent with Cp 2 NbH 2 , Cp2 Nb(H)C 2 H5 and Cp2 Nb were observed consecutively suggesting a hydrogenation mechanism involving I ;- H Cp2 Nb-H CpzNb, ~ - - + Cp2 Nb + ethane H paramagnetic species. In fact, there appears to be an equilibrium involving Cp NbH and Cp NbH • 9 (Vide infra.) 2 3 2 2 76 Cp2 NbH 3 It is clear that oxidation states of two through five are common in niobocene chemistry. Several of the intermediates of the above reactions are interesting molecules which could not be isolated due to irreversible reactions involving the carbon hydrogen bonds of the cyclopentadienyl ligands. It was hoped that substitution of pentamethyl cyclopentadiene would alleviate this problem and facilitate isolation of coordinative ly unsaturated complexes. 1O Additionally, we hoped niobium would mimic some of the carbon monoxide chemistry observed with zirconium. 11 Perhaps the expected weaker niobium oxygen bonds would permit reductive elimination of methanol and enable catalytic carbon monoxide reduction with hydrogen. 0 \\ + Ha ,,,,c ..."H ~ Cn~ Nb Cn~ Nb_cH2 i-.i \ -..H - i-.i H '\H o, OCH 3 Cp2N(-CH2 --+- Cp2NbOCH3 _+_H_2-~ Cp2N~-H H 0 I '--- ~ CP2NbH + CH 3 OH 77 By and large, these goals have not been achieved. For the CO reaction, the failure of CO to insert, even into a niobium methyl bond, precludes any possibility of CO reduction. The synthesis of several bispentamethyl cyclopentadienyl niobium complexes which bear on this question are reported. More promising results are the syntheses of several paramagnetic niobocene derivatives. The impetus for studying paramagnetic niobocene chemistry originated from the similar behavior of titanocene 10 and niobocene dialkyls. 12 Both are thermally sensitive and explosively release methane under certain conditions. The use of pentamethyl cyclopentadiene ligands in the titanium system stabilized the chemistry enough to understand the reaction. 13 cp;Ti(CD ) loses d methane at 3 2 4 90 ° C to yield the fulvene complex. This reaction is the first step of the permethyl titanocene synthesis. lO We hope to observe similar reactivity with niobium to obtain low valent niobium compounds. The starting point for the investigation of niobocene and singly substituted niobocene derivatives is presented. 78 RESULTS AND DISCUSSION The entry into all permethyl niobocene chemistry is cp;NbBH 4 • It is prepared by reaction of niobium pentachloride with lithium permethyl cyclopentadiene and NaBH 4 in dimethoxyethane. Yields as high as 45% are obtained which depend critically on the solvent purity as well as the purity of all starting materials. The nmr of the green borohydride shows a singlet for the Cp* methyl hydrogens. In addition to broadening due to boron quadrupole coupling, the possibility for fluxionality renders a complex analysis for the borohydride proton resonances. 10 The borohydride complex reacts with pyridine under hydrogen to produce cp;NbH 3 in 80% yield. Presumably, pyridine reacts with Cp*2 NbH Cp*2 Nb H3 borohydride complexes similar to the phosphine BH 3 abstractions. 15 The trihydride is unstable in solution in the absence of hydrogen -, analogous to the normal ring niobocene trihydride. Heating speeds the hydrogen reductive elimination. Attempts to isolate cp;NbH resulted in intractable oils. 79 cp;Nb(H)CO is produced quantitatively from the CO reaction with the trihydride. Nmr and ir spectra are similar to the normal ring niobocene hydride carbonyl. Cp;Nb(H)CO is unreactive toward several atmospheres of hydrogen at elevated temperatures. The reaction of Cp:Nb(H)CO with HCl yields Cp;Nb(Cl)CO. At low temperature, the initial product appears to Cp;NbH 2 CO. EB This yellow cation is isoelectronic with Cp; ZrH 2 CO. It is amusing that this d0 dihydride carbonyl only undergoes reductive elimination on warming to yield hydrogen and cp;Nb(Cl)CO. Unfortunately, cp;NbH 2 COEB is not sufficiently long-lived to allow measurement of the ir stretch of the d0 carbonyl. The use of non-coordinating acids such as CF 3 CO2 H and HC1O4 to produce CptNbH 2 COEB did not prevent facile reductive elimination above 0°C. Cp:Nb(Cl)CO reacts with methyl lithium to form Cp2 Nb(CH 3 )CO in high yield. Under high pressures of CO or several equivalents of trimethyl phosphine, methyl migration to form an acyl is not observed. An experiment yet to be done is the protonation reaction to form cp;Nb(H)(CH 3 )CO. EB In light of the facile CO insertion into the alkyl bond of isoelectronic cp;zr(CH 2 CH(CH 3 ) 2 )(H)CO, one expects acyl formation from the niobium methyl hydride carbonyl cation. cp;NbBH 4 reacts with aqueous HCl to yield cp;NbClz in 70 percent yield. The esr of Cp;NbClz displays a ten-line spectra with a g value of 2. 01 characteristic of niobium IV. Magnetic susceptability indicates the compound has a magnetic moment of 1. 91 BM. Reduction of the dichloride with Na/Hg under CO yields a brown complex which esr and ir evidence indicates is a niobium II carbonyl. A similar 80 reduction under hydrogen produced a red crystalline compound whose esr displayed a ten-line pattern with proton hyperfine indicative of two protons per metal. However, the purity is ambiguous because the same compound shows an nmr spectrum chracteristic of Cp;NbH 3 • Apparently an equilibrium exists under hydrogen between the dihydride and the trihydride. Low-temperature preparation of cp;NbH 2 may circumvent this problem. Oxidative addition of H2 to niobocene is an attractive synthesis, presuming that niobocene is synthetically attainable. In this vein, attempts to duplicate the permethyl titanocene preparation are underway. Additionally, reductive elimination of an alkyl hydride may + be an attractive route to niobocene at low temperature. The reaction of methyl lithium produces cp;Nb(CH 3 )Cl. Reactions with suitable hydride reagents should yield the methyl hydride. 81 CONCLUSION ~ The main barrier to studying the permethylcyclopentadienyl chemistry of niobocene has been surmounted. Low valent permethyl niobocene complexes can be easily studied now that cp;Nbc½ is readily available. In addition to providing theoretically interesting molecules such as niobocene, these compounds have potential for interesting ligand transformations such as carbon hydrogen bond activation, olefin oligomerization and migratory insertion reactions. 82 EXPERIMENTAL ~ All manipulations were performed in an inert atmosphere using a glove box or vacuum line techniques. Ether solvents such as glyme and THF were distilled from sodium benzophenone ketyl. Halogenated solvents were distilled in vacuo from 4 A molecular sieves. Toluene, 1 benzene, and pentane were purified by distillation from titanocene. 7 Hydrogen, argon, and nitrogen were passed through MnO on vermiculite 18 and 4 A molecular sieves. Carbon monoxide (MCB) was used without purification. Pyridine was refluxed over CaH 2 for a day and stored in vacuo over 4 'A molecular sieves. Lithium pentamethylcyclopentadienide was prepared by literature methods. 19 NbC1s was carefully purified by two sublimations. The initial sublimation at 110° C separated NbC15 from most of the oxychloride impurities. A second sublimation at 100° C yielded pure NbC1 5 which showed no niobium oxide stretches in the ir. 1 H nmr spectra were obtained using a Varian EM 390 spectrometer. Infrared spectra were recorded on aBeckman 4240 spectrophotometer. Esr spectra were measured on an X-band Varian E-line Century Series spectrometer. Magnetic susceptabilities were determined using a Cahn/Ventron susceptability balance. Preparation of cp;NbBH 4 18.15 Grams of pentamethylcyclopentadiene and 55. 5 mls of 2. 4 M BuLi(Alfa/Ventron) were added at -78 ° C to 300 mls of glyme. This reaction was stirred at 25 ° C for six hours before nine grams of NaBH 4 was added. The solution was cooled to -78° C and 16 grams of 83 freshly sublimed NbC15 was slowly added with vigorous stirring. After the addition is complete, the solution is warmed to 25 ° C, then heated at reflux for three days. The glyme is then removed by vacuum transfer. The resulting purple residue is extracted four times with pentane. The pentane volume is reduced to yield green crystals from the maroon solution. Recrystallization from octane as well as subliming the purple residue yield 8. 5 grams of cp;NbBH 4 : 45% based on NbC1 5 • Nmr (d6 benzene) o 1.67 (s, 30H); o 5.17 (d, 2H); o -18.2 (s, 2H). Anal. Calcd. %C; H, 9. 06 ; Nb, 24. 55; B, 2. 86; Found: %C, 63. 36; H, 8. 9 6; Nb, 25 . 04; B, 2. 9 2. Preparation of cp;NbH 3 4. 1 Grams of Cp!NbBH 4 and ten mls of pyridine were added to 40 mls of toluene at -78 ° C. The mixture was warmed to 25 ° C under one atmosphere of hydrogen and stirred eight hours. All volatile materials are removed by vacuum transfer and the orange residue was extracted with 40 mls of diethyl ether. The solvent volume was reduced to ten mls and the resulting yellow crystals were collected by filtration. The extraction process was repeated twice and the crude trihydride recrystallized from octane under hydrogen. 3. 3 Grams of pale-yellow Cp:NbH 3 were obtained in 82% yield. Nmr (d6 benzene) ol.90 (s, 30H); o -1.20 (brd, lH); o -2.30 (d, 2H). Preparation of Cp;Nb(H)CO 1. 2 Grams of Cp*2 NbH 3 , dissolved in 25 mls of toluene, was stirred under one atmosphere of CO at 95 ° C for sixteen hours. The toluene was removed from the red solution by vacuum transfer and the residue was extracted with 25 mls of octane at -30° C. The octane volume was 84 reduced to seven mls and the red precipitate was recrystallized to yield 1.1 grams of Cp!Nb(H)CO: 90%. Nmr (d 6 benzene) o 1. 80 (s, 30H); o -5. 6 (brd, lH). Ir (nujol) vc=o 1880 cm -1 , vNb-H 1700 cm -1 . Preparation of CptNb(Cl)CO 300 Mgs of CptNb(H)CO were dissolved in 30 mls of toluene. Two equivalents of HCl gas were frozen into the reaction pot at -196° C. The reaction was warmed to -60° C to melt all the toluene, then recooled to -78° C. The red solution is then stirred for several minutes until the solution is yellow. All excess HCl is then pumped off. Toepler experiments indicate that no hydrogen was evolved. Warming above -10 ° C resulted in a significant rate of gas evolution which was complete after 30 minutes. One equivalent of hydrogen was formed to produce green cp;Nb(Cl)CO. The compound was washed with pentane. Nmr (d6 benzene) B 1. 60 singlet. Ir (nujol) vC=:O 1890 cm -l. Preparation of CptNb(CH 3 )CO 210 Mgs of Cp!Nb(Cl)CO was dissolved in diethyl ether along with 0. 4 mls of 1. 6 M methyl lithium (three-fold excess). The reaction was stirred for a day with no color change. The diethyl ether was removed and the green residue extracted with toluene. The toluene was removed by vacuum transfer and the resulting green crystals suspended in cold pentane and filtered. 190 Mgs were collected in 90 percent yield. Nmr (d6 benzene) o 1. 53 (s, 30H); o -0. 93 (s, 3H). Preparation of Cp!NbC½ 5. 3 Grams of cp;NbBH 4 dissolved in 100 mls of benzene was reacted with 15 mls of de oxygenated 3M HCl at O° C. The solution evolves gas vigorously and turns brown with yellow precipitate. 85 After stirring for 20 minutes, all volatiles were pumped away. The brown residue was soxelut extracted with benzene. The brown microcrystalline product was dissolved in 120 mls of CH 2 C¼. This solution was added to 500 mgs of degassed grade III alumina. Filtration separated cp;Nbc¼ from all hydroxy niobium compounds. Recrystallization from toluene yielded 4. 1 grams of cp;NbC12 : 67%. Esr (benzene) ten line spectrum, g = 2. 01. Magnetic susceptability = 1. 91 BM. Anal. Calcd. %C, 55.31; H, 6.96; Nb, 21.39; Cl, 16.33; Found, %C, 55.41; H, 7.03; Nb, 21.52; Cl, 16.29. Reduction of cp;NbC¼ under hydrogen 970 Mgs of Cp;NbClz was reduced in benzene with Na/Hg under one atmosphere of hydrogen for one day. The resulting red solution was decanted from the amalgam and filtered. Benzene was replaced with twenty mls of octane by vacuum transfer. The mixture was heated under hydrogen (mistake) to dissolve the red crystals, then filtered. Slow cooling to -40° C yielded 600 mgs of red crystalline compound. Esr (benzene) ten-line spectrum, g = 2. 01. Nmr (d 6 benzene) identical to cp:NbH3, 86 REFERENCES 1. U. Klabunde, G. W. Parshall, J. Am. Chem. Soc., ,..__,..__ 94, 9081 (1972). 2. F. N. Tebbe, G. W. Parshall, J. Am. Chem. Soc., .,.,_,..._ 93, 3793 (1971) . 3. L. J. Guggenburger, F. N. Tebbe, J. Am. Chem. Soc., .,.,_,..._ 93, 5924 (1971). 4. H. H. Brintzinger, J.E. Bercaw, J. Am. Chem. Soc., .,.,_,..._ 92, 6182 (1970). 5. J. A. Labinger, J. Schwartz, J. Am. Chem. Soc., .,.._,,..,_ 97, 1596 (1975) . 6. L. J. Guggenberger, P. Meakin, F. Tebbe, J. Am. Chem. Soc., 96, 5420 (1974) . .,._r,.. 7. I. H. Elson, J. K. Kochi, J. Am. Chem. Soc., .,.,_,..._ 97, 1264 (1975) . 8. I. H. Elson, J. K. Kochi, U. Klabunde, L. E. Manzer, G. W. Parshall, F. N. Tebbe, J. Am. Chem. Soc., 96, 7374 (1974). -'"'-" 9. J. A. Labinger, K. S. Wong, J. Organomet. Chem.,~ 373 (1979). 10. J. E. Bercaw, J. Am. Chem. Soc., .,.,_,..._ 96, 5087 (1974) . 11. J. M. Manriguez, D. R. McAllister, R. D. Sanner, I. E. Bercaw, J. Am. Chem. Soc., 98, 6733 (1976) . .,._r,.. 12. L. E. Manzer, Inorg. Chem., !.§_, 525 (1977). 13. R. Schmitt, J. E. Bercaw, manuscript in preparation. 14. S. A. Cohen, J. E. Bercaw, manuscript in preparation. 15. C.R. Lucas, M. L. H. Green, J. Chem. Soc. Chem. Commun., 1972, 1005. ~ 16. J.M. Manriquez, D. R. McAlister, R. D. Sanner, J. E. Bercaw, J. Am. Chem. Soc., 100, 2716 (1978). ~ 17. R.H. Marvich, H. H. Brintzinger, J. Am. Chem. Soc., -'"'-" 93, 2046 (1971). 87 RE FERENC ES (continued) ~ 18. T. L. Brown, D. W. Dickerhoof, D. A. Bafus, G. L. Morgan, Rev. Sci. Instrum., "33, 491 (1962). "19. R. S. Threlkel, J.E. Bercaw, J. Organomet. Chem., 177(1), 1 (1977). 88 PROPOSITIONS 89 PROPOSITION I ~ Abstract: A series of electrochemical measurements is proposed for several quinone derivatives complexed to transition metals. It is hoped that a four-electron redox process will occur when the twoelectron quinone potential couples with the two-electron metal redox potential. By varying substituents on both the metal and the quinone, the net four-electron potential might be adjusted to significantly positive values to efficiently reduce oxygen to water. 90 Attempts to efficiently electrochemically reduce oxygen have generated considerable interest in the requisite four-electron transfer process. Several complications have clouded the understanding of this deceptively simple process: 0 2 + 4H+ + 4e- --+ 2H 2O. The fourelectron reduction primarily occurs in some poorly understood enzyme reactions. 1 Simple molecules have thus far shown no propensity toward four-electron redox processes so their relevant molecular properties are undefined. Oxygen reduction has been studied with metal complexes known to be unstable to oxygen. Unfortunately the reducing agents form hydrogen peroxide instead of water. 2 With respect to fuel cell efficiency, this two-electron reduction at 0. 68 V represents a major loss in energy relative to the 0 2 /H 20 couple at 3 1. 23 V. A modification currently under investigation involves bimolecular catalysis by face to face porphyrins. 4 Geometrical modifications will hopefully enhance the known bimetallic oxygen reduction by iron porphyrins 5 such that peroxide formation is suppressed. Some success has been achieved in this vein. 4 No peroxide was formed when oxygen is reduced by a Cobalt-Cobalt porphyrin complex. However, the 650 mV reduction potential of the cobalt complex still represents a substantial overpotential. In addition, the electrochemistry appears to be a series of one-electron steps which somehow net an overall four-electron process. Modifications which could improve the catalyst seem obscure and synthetically formidable. A system which has two coupled two-electron processes might allow more insight into the fourelectron process. Also a simpler system would offer more flexibility 91 in adjusting the overall potential to approach 1. 23 V. The quinone functionality is one of the most extensively studied two-electron redox processes. 6 A wide variety of quinones are commercially available with substituents that predictably alter the quinone/ hydroquinone potential. Of special relevance to the oxygen electrode problem is the commercial importance of anthroquinone catalysts for the production of H O • 7 The reduced form reduces 2 2 0 ~~ ~ ~ ~ + H1. 02. 0 oxygen, albeit to H2O 2 • To efficiently use the oxygen reduction energy of 1. 23 V, peroxide formation must be avoided. Although the twoelectron oxygen reduction to H2 O2 is 0. 68 V exothermic, this process does not use the additional 0. 5 volts of energy electrochemically available from complete oxygen reduction. In the quinone system, e H 0 92 complete reduction will result by cleaving the oxygen-oxygen bond of the peroxide intermediate before H2O2 extrusion. In fact, this bond is the weakest in the molecule. The proximity of a sufficiently strong reducing agent such as a metal complex could reductively cleave the oxygen-oxygen bond. stability of the reduced and oxidized catalyst and a large metalquinone interaction is necesssary to electrocatalytically reduce oxygen. Chemical reversibility may not be fortuitous due to the expected ligand reorganizations involved with the two-electron metal couple. strong interaction of the metal and the quinone is needed for fast electron transfer to break the oxygen-oxygen bond. When the meta1+2 /metal and quinone/hydroquinone potentials are nearly the same, the metal quinone interaction will facilitate a simultaneous four-electron reduction. Once this potential is adjusted more positive than 0. 68 V, peroxide formation becomes endothermic. A model which may satisfy these requirements is: OH Other metals can be Pd+ 2 and Ir +l 0 Other quinones can be ~ ~ 0 X can be an amine or a hydroxy group OH Chemical reversibility between the plus two and plus four oxidation 93 states should be enhanced by the macrocyclic ligand. 8 The proximity of the metal to the quinone ought to facilitate rapid electron transfer from the metal to the intermediate peroxide. Considerable modifications are possible to match the metal and quinone reduction potentials. Substituted quinones have reduction potentials ranging from 0. 0 V to 1. 1 V. 6 Platinum amine reduction potentials range from 0. 5 V to 0. 9 V. 9 By varying the nature of the macrocycle as well as substituents on the quinone, a voltage significantly more positive than 0. 7 volts seems attainable. 94 REFERENCES ~ 1. (a) J. A. Fee, Coord. Chem. Rev., .,.._.,.._ 23, 1 (1975); (b) R. A. Holwerda, S. Wherla.nd, H. B. Gray, Ann. Rev. Biophys. Bioeng., ~' 363 (1976). 2. S. Bhaduri, L. Casella, R. Ugo, P. R. Raithby, C. Zuccaro, M. B. Hursthouse, J. Chem. Soc. Dalton, p. 1624 (1979). 3. J. 0. M. Bockris and S. Srinivasan, "Fuel Cells: Their Electrochemistry," McGraw-Hill, New York (1969). 4. I. P. Collman, M. Marrocco, P. Denisevich, C. Koval, F. C. Anson, J. Electroanal. Chem., 101, 117 (1979). 5. N. Sutin, "Oxidases and Related Redox Systems," (Ed. T. E. King, S. H. Mason, and M. Morrison), p. 37, John Wiley and Sons, New York, N. Y., 1965. 6. (a) A. J. Bard, "Encyclopedia of Electrochemistry of the Elements," Vol. 12, Dekker, 1979; (b) W. M. Clark, "Oxidation-Reduction Potentials of Organic Systems," Williams and Wilkens, 1960. 7. R. Powell, "Hydrogen Peroxide Manufacture," Noyes Devel. Corp. Park Ridge, New Jersey, 1968. 8. (a) R. C. Johnson, F. Basolo, J. Inorg. Nucl. Chem., .!1_, 36 (1960); (b) F. Basolo, M. L. Morris, R. G. Pearson, Disc. Faraday Soc. , 29, .,.._.,.._ 80 (1960) . 9. A. J Bard, "Encyclopedia of Electrochemistry of the Elements," Vol. 6, Dekker, 1979. 95 PROPOSITION II ~ Abstract: An electrochemical study of a series of bimetallic complexes is proposed. The two-electron nature of metal-metal bonds will hopefully cause two-electron redox electrochemistry instead of one-electron transfer observed for most transition metal complexes. The metal-metal bond formed by electron transfer ought to substantially affect the potential. 96 The paucity of two-electron redox processes of transition metal complexes 1 is probably related to the small energy difference between M+n and M+n+l oxidation states. Compared to p-orbital energies, most dn states are closer in energy to dn-l than pn states are to pn-l. 2 Thus one-electron metal oxidation is stabilized by relatively minor ligand reorganizations. In p-orbital redox processes, ligand reorganization may not be sufficient to stabilize a different oxidation state. More drastic changes can result such as bond formation. Because most bonds involve two electrons, many p-orbital redox processes are two electron transfers. 3 Analogous to bonding responsible for two-electron transfers in p-bonded compounds, metal-metal bonding could favor a single twoelectron process. A two-electron process is favored over two oneelectron steps by the metal-metal bond energy. Many examples of metal-metal bonded complexes are known. 4 Singly as well as multiply bonded bimetallic complexes are substantially stabilized by metal1e + + 2 e M-M M-M metal interactions. Some bimetallic compounds have shown a tendency toward two-electron electrochemistry. l(CO) 5Mnl 2 and [CpCr(CO) 3 l 2 are reduced by two electrons to form (CO) Mn - and CpCr(Co);. 5 5 The reductions are electrochemically irreversible due to dissociation. In several bimetallic macrocycles where dissociation does not occur, 97 two-electron processes are not observed. 6 Well separated oneelectron transfers are obtained. However, the distance between M = Cu, Ni, Co, Fe metals is too long to permit metal-metal bonding. The fulvalene bimetallic complexes offer promise of twoelectron processes. The bidentate fulvalene is capable of holding the metals close together whether or not there is a metal-metal bond. Two-electron oxidative or reductive cleavage of a metal-metal bonded fulvalene complex would not totally separate the metals. Since only one electron is transferred per metal, ligand reorganization should not be too severe to prevent electrochemical reversibility. The redox potential of the the two-electron couple, LxM:2 /LxM 2 , should be one half the metal-metal bond energy more positive than the single electron couple, LxM+/LxM. Only a single two-electron redox potential should result. 98 Several metal-metal bonded fulvalene complexes which may serve as two-electron transfer agents are: 7 ' 8 :r-----...s::::::5· I ~\ 'M~ ~ The metal-metal bonds of the first two compounds will cleave upon two-electron oxidation. Two-electron reduction should cleave the metal-metal bond of all three complexes. Several related complexes have been studied. The bis-fulvalene complexes of vanadium, chromium, iron, cobalt, and nickel all have two well separated oneelectron potentials. 9 The separation of one-electron redox potentials of greater than 230 mV indicates the metals of these bis-fulvalenes are close enough to interact. A significant two-electron bond does not form because the bis-fulvalene ligand environment keeps the metals 99 out of bonding distance. The molybdenum and niobium fulvalenes are better suited to two-electron transfer because their larger size permits metal-metal bonding at the 2. 9-3. 4 'A distance permitted by the fulvalene. lO Unless solvent interacts strongly with the non-bonded complexes, two-electron reversible electrochemistry of these secondrow fulvalenes appears imminent. 100 REFERENCES ~ 1. A. J. Bard, "Encyclopedia of Electrochemistry of the Elements," Vol.1-7, Dekker, 1979. 2. C. E. Moore, "Atomic Energy Levels," Vol. I and II, National Bureau of Standards, 1949. 3. W. M. Clark, "Oxidation-Reduction Potentials of Organic Systems," Williams and Wilkens, 1960. 4. (a) W. C. Trogler, H.B. Gray, Accts. Chem. Res., .,.._,.,_ 11, 232 (1978). (b) H. B. Abrahamson, C. C. Frazier, D. S. Ginley, H. B. Gray, J. Lilienthal, D. R. Tyler, M. S. Wrighton, Inorg. Chem., 1§., 1554 (1977). 5. R. E. Dessy, P. M. Weissman, R. L. Pohl, J. Am. Chem. Soc., 88, .,.._,.,_ 5117 (1966) . 6. C. Spiro, R. R. Gagne, J. Am. Chem. Soc., 102, 000 (1980). ~ 7. (a) J. C. Smart, C. J. Curtis, Inorg. Chem., 11., 3290 (1978). (b)ibid., 16, 1788(1977). - - .,.._,.,_ 8. T. Y. Struchkov, Y. K. Slovohotov, A. J. Yanovskii, D. A. Lemonovskii, Isv. Akad. Naulc SSSR, Ser. Kim., ~' 1421 (1979). 9. B. L. Pinsky, Ph.D. Thesis, University of California, Berkely, 1979. 10. M. R. Churchull, J. Wormald, Inorg. Chem., !!, 1970 (1969). 101 PROPOSITION ID ~ The synthesis of several novel substituted cyclopentadienyl tantalum and niobium compounds is proposed. Hydrogen abstraction from alkyl or allyl ligands can yield coordinated ligands which are normally difficult to prepare. The resulting compounds are potentially interesting with respect to reductive elimination, carbonhydrogen bond activation and small molecule reactions. 102 Intramolecular hydrogen abstraction occurs in several reactions of early transition metal complexes. Abstraction of a phenyl hydrogen to form a benzyne complex has been implicated in the decomposition of diphenyl zirconocene. 1 The benzyne complex, C Me Ta(CH \C H , 5 5 3 6 4 was isolated from the decomposition of C5Me 5 Ta(CH 3 ) 3 q>. 2 Presumably, hydrogen abstraction of a phenyl hydrogen by either the metal or a methyl preceeds reductive elimination of methane to form the benzyne complex. A similar reaction involving C5 Me 5 Ta(CH 3 ) 3 CH 2 C(CH 3 ) 2 CH 2 produced the trimethylene methane complex, C5Me 5Ta(CH 3 ) 2 C(CH2 ) 3 and methane. 3 A closer look at the mechanism indicates methane reductive elimination follows hydrogen transfer from the methyl allyl * CpTa(CH 3 )3 y to yield the trim ethylene methane product. A similar mechanism is postulated for the decomposition of bis-permethyl cyclopentedienyl titanium dimethyl. 4 Thermolysis of Cp:Ti(CD ) yielded only CD • 3 2 The resulting titanium product is the fulvene complex. 4 103 These intere sting ligand transformations are mediated by an unsaturated metal center. Hydrogen abstraction in the tantalum complexes, CpTa(CH 3)~, followed by reductive elimination regenerates unsaturation of the metal. Consecutive hydrogen abstraction and reductive elimination cycles of appropriately substituted tantalum complex should produce interesting products. Several possibilities include: -----►➔ CpTaCH 3 / ~ In general, three experiments are envisioned for each cyclic ligand: decomposition of the cyclo ally 1, decomposition of the cycloalkyl, and substitution of the cyclic ligands to block hydrogen abstraction. Several interesting compounds are expected which may lend them- 104 selves to studies of reductive elimination and carbon-hydrogen bond activation. The proposed decomposition of the cyclohexallyl-substituted complex is shown below. Two consecutive hydrogen abstractions CpTaCH 3 6 CpTaCH 3 6 yield the sixteen-electron 77 5 -cyclohexadienyl complex. The final hydrogen abstraction yields the eighteen-electron cyclopentadienyl benzene tantalum methyl hydride. In light of the unusually high stability of a zirconium alkyl hydride, 5 the possibility of reductive elimination to form CpTaBz should be interesting. Decomposition of the cyclohexyl rather than cyclohexallyl could yield the dihydride tantalum sandwich complex. The driving force for the reaction is the eventual eighteen-electron complex. Whether or not any of the intermediates will be isolable seems unfathomable at this point. However, some steps may be prevented by methyl substitution. 105 CpTa ( CH 3) 3 6 CpTa( CH 3 ) 2 -CH4 6 -CH4 bH, > -CH4 ~_..,.-H CpTa V 6 ~'H Presumably, a hydrogen must point toward the metal for hydrogen abstraction to occur. 6 Another group besides hydrogen should prevent hydrogen abstraction. 5, 5 Dimethyl cyclohexallyl cannot easily form a benzene complex. Trans-4, 5-dimethyl cyclohexallyl cannot transfer the last hydrogen to tantalum unless there is a hydride shift. Cpb 106 Similar considerations for the cyclopentallyl and cyclopentyl are shown below. Once again, consecutive hydrogen abstractions are > CpTaCH 3 0 facilitated by the unsaturated metal. The use of trans-4, 5-dimethyl cyclopentallyl could permit a closer look at the 1-5 hydride shift. 107 These syntheses of bis-cyclopentadienyl tantalum complexes are 1, 5hydride ?H,~ shift > ~ H , ) , ---C-H-,.- CQTa(CH 3 ) 2 H particularly attractive for bis-permethylcyclopentadienyl tantalum compounds. Cp:Ta.Xn cannot be prepared by conventional techniques. 7 These alkyl hydrides lend themselves to reductive elimination studies. Hopefully the alkyl hydrides will be produced under mild enough conditions to permit their isolation. Reductive elimination of methane from Cp2 Ta(CH 3 ) 2H yields Cp2 TaCH 3 which is known to be very reactive. 8 Providing the reductive elimination is facile at low temperature, Cp2 Ta(H)CH2 could be isolable via the a-elimination reaction. Decompositions of both the four- and three-membered rings should lead to similar products. All the cycloalkyl decompositions are interesting reactions yielding products conducive to considerable investigation. The coordinatively unsaturated products lend themselves to further study with small molecules such as ethylene, carbon 108 monoxide and hydrogen. The alkyl hydrides are also potentially reactive due to their expected facile reductive elimination. Finally, the low temperatures involved in the hydrogen abstraction may allow isolation of reactive intermediates. 109 REFERENCES ~ 1. G. Erker J. Organomet. Chem., 134, 189 (1977). 2. (a) S. J. McLain, R. R. Schrock, P. R. Sharp, M. R. Churchill, ~ W. J. Youngs, J. Am. Chem. Soc., .!.Q!, 263 (1979). (b) M. R. Churchill, W. J. Youngs, Inorg. Chem., _!1, 1697 (1979). 3. C. J. Curtiss, J. E. Bercaw, manuscript in preparation. 4. (a) J. E. Bercaw, J. Am. Chem. Soc., ~' 5087 (1974); (b) R. Schmidt, J. E. Bercaw, manuscript in preparation. 5. D. R. McAlister, D. K. Erwin, J. E. Bercaw, J. Am. Chem. Soc., 100, 5966 (1978). 6. T. H. Whitesides, J.P. Neilan, J. Am. Chem. Soc.,~' 63 (1976). 7. R. R. Schrock, private communication. 8. R.R. Schrock, P . R. Sharp, J. Am. Chem. Soc., .,..._,.._,.__ 100, 2389 (1978). 110 PROPOSITION IV ~ Abstract: The mechanism of propene insertion into a hydride followed by reductive elimination of propane is probed. The products of the deuterium promoted reaction should indicate the reaction pathway. 111 0 The high reactivity of several d zirconium complexes toward hydrogen has provoked controversy regarding the reaction mechanism. Deuterium is observed to exchange with cp; ZrH2 at -80 ° C forming primarily HD. 1 Hydrogen also significantly promotes the reductive elimination of alkanes from several zirconocene alkyl hydrides. 2 , 3 Oxidative addition of hydrogen to the <f' complexes seems unlikely. Two mechanisms have been proposed to explain the promoting effect of hydrogen. The first explanation involves a four-center exchange reaction whereby the two atoms of H2 and the two atoms of the cleaving metal substituted bond interact. 3 The mechanism explains why the alkane ,, H Cp2 Z~ I + R iR Cp2 Zr, + H R D2 D2 :> ► H * Cll2 ~r •. ,,,,,~ D •.. ,, , ,n' I R Cp2 ~r, .... ,~ D,,,.. .. Di H • Cp2 Zr'n + R-D I . ~ ~ Cp2 Zr, + HD D product of reductive elimination is exclusively deuterium substituted. It also rationalizes the incorporation of deuterium into the hydride position of the alkyl hydride, Cp2ZrHR, prior to alkane formation. However, an efficient pathway involving H -D exchange catalyzed by 112 Cp2 ZrD2 followed by intermolecular deuteride transfer also explains the alkyl hydride deuterium incorporation. This seems feasible considering the dimeric nature of Cp2 ZrRH. Additionally, Cp2 ZrH 2 is expected to a very efficient catalyst for H-D exchange from D2 • A second mechanism involves intitial reductive elimination of the hydrogen and the Cp*. 2 Oxidative addition of D is then possible. 2 * CP2Zr ,....R 'H ~ Cp* Only alkyl deuterium results from reductive elimination. In contrast to the normal ring system, deuterium does not exchange with alkyl hydride prior to reductive elimination. If a four-center mechanism is responsible for the H-D exchange, it is hard to believe the alkyl hydride would not exchange with D2 • cp;ZrH2 rapidly exchanges with D2 at -80° C. The second mechanism does not allow H-D exchange for the alkyl hydride. H-D exchange can occur from the dihydride because hydride Cp* reductive elimination still leaves one hydride on zirconium. 113 Both mechanisms bespeak the high energy of the simple reductive elimination product, Cp2 Zr. Whereas most transition metal alkY:l hydrides undergo facile reductive elimination,' 4 Cp ZrHR 2 goes to extremes to avoid forming Cp2 Zr. Several early transition metal biscyclopentadienyl complexes share zirconiums aversion for the unsaturated structure. 5 CP2Nb-X and Cp Ta-X are extremely 2 unstable sixteen electron complexes. 6 Several olefin hydrides of bis-cyclopentadienyl niobium and tantalum readily undergo olefin migratory insertion at low temperatures. 7 Although the simple migration to form the alkyl seems obvious, the olefin hydride Cp2M(CH 2 =CHR)H may shun the unsaturated complex Cp2 Nb-R' in favor of the intramolecular gymnastics proposed for Cp2 ZrRH. The deuterium promoted migratory insertion and reductive elimination of the niobocene propene hydride 9 is capable of uncovering .this devious reactivity. If the hydride migrates to the ring prior to oxidative addition of deuterium, olefin insertion will yield a secondary C Nb~ P2 ' H deuterio-propyl group. At this point, several possibilities exist. Reductive elimination of the propyl deuteride followed by deuterium trapping would yield 1, 2-dideuteriopropane. If the hydrogen moves back from the ring first, either 1, 2-dideuteriopropane or 114 C1>zNb~ ~ Nb/"( / Cp 'n ~ 'n t> C1>zNb~ 'H D ~ D y l) YD D /H C1>zNb« t) oYY [) D 2-deuteriopropane will result, depending on how the hydrogen comes back. If the olefin inserts prior to deuterium oxidative addition, only 1-deuteriopropane will result. Both schemes assume olefin insertion will yield only the primary alkyl complex. This seems reasonable considering other metallocene hydride additions to olefins. 8 The yield of 2-deuteriopropane differentiates the two schemes as to whether or not the olefin insertion is reversible. In scheme Il, deuterium incorporation into the secondary carbene is not allowed. The presence of two deuteriums on the secondary carbon of propane would indicate reversible olefin insertion in scheme I. The two possible reaction modes are thus clearly differentiable by analysis of the deuterium:-substituted propane products. 115 REFERENCES ~ 1. J. E. Bercaw, Adv. Chem. Ser., 167, 136 (1978). 2. D. R. McAllister, D. K. Erwin, J. E. Bercaw, J. Am. Chem. ~ Soc., 100, 5966 (1978). 3. K. I. Gell, J. Schwartz, J. Am. Chem. Soc., !.QQ, 3246 (1978). 4. (a) J. Evans, S. J. Okrasinski, A. J Prioula, J. R. Norton, J. Am. Chem. Soc., ~' 4000 (1976). (b) L. Abis, A. Sen, J. Halpern, J. Am. Chem. Soc., 100, 2915 ~ (1978). (c) E. L.Muetterties, P. L. Watson, J. Am. Chem. Soc., .,.,_,,..,_,.._ 100, 6978 (1978). 5. J. L. Thomas, J. Am. Chem. Soc., .r-.r-. 95, 1838 (1973) . 6. (a) E. K. Barefield, G. W. Parshall, F. N. Tebbe, J. Am. Chem. Soc., .r-.r-. 92, 5235 (1970). - (b) L. J. Guggenburger, P. Meakin, F. N. Tebbe, J. Am. Chem. -Soc., 96, 5420 (1974). /"'-"- 7. F. N. Tebbe, G. W. Parshall, J. Am. Chem. Soc., .r-.r-. 93, 3793 8. (a) D. W. Hart, J. Schwartz, J. Am. Chem. Soc., "-"' 96, 8115 (1974). (b) C. A. Bartolo, J. Schwartz, J. Am. Chem. Soc., "-"' 97, 228 (1975). 9. A.H. Klazinga, J. H. Teuben, J. Organomet. Chem., 157, 413 ~ (1978). 116 PROPOSITION V The generation of acetylene and substituted acetylene carbonium ions is proposed by oxidation of acetylene radicals by copper II. Formation of a stable copper acetylide and the high endothermici ty for the process HC:C• ~ HC:C~' are problems which can hopefully be overcome by the proper choice of solvent and supporting media. 11 7 Although nucleophilic substitutions at acetylene carbons are not common, such substitutions have been observed 1 to be facile. Unlike substitutions at saturated carbons, there has been no evidence to implicate an acetylene carbonium inn as an intermediate in an S Nl type reaction. The reason for this is probably the high heat of formation of HC=C® which is 397 kcal/mole. Indeed, reported attempts have failed to produce such a cation when a haloacetylene was reacted with silver oxide. Diazotation of acetylene amine followed by loss of nitrogen to form the desired cation 1s complicated by the instability of acetylene amines. Another approach is to take two steps to reach the high energy cation, each step being relatively endothermic 1n terms of the acetylene. Step one would be to generate an acetylene radical and step two would be to oxidize that radical to the acetylene carbonium 1On. An important requirement for step two is that it must be as fast as radical coupling in order to avoid simple dirnerization of acetylene radicals. Two ways to generate acetylene radicals which allow a variety of acetylene substituents are the loss of CO or 2 CO from an acyl ar acyloxy radical respectively. Although 2 these reactions have not been utilized to generate acetylene radicals in the past, these fractionations hve been used to 118 generate a variety of alkyl radicals including methyl radical. The acyl radical can be generated by radical abstraction of a hydrogen atom from an aldehyde, 3 and the acyloxy radical can be generated by the thermolysis of a peroxy-diacid. 4 The latter is preferred as one might get complications from radical addition to the acetylene rather than hydrogen abstraction from an aldehyde. exothermic than loss of CO. 5 Also, loss of CO 2 is more In order to oxidize the acetylene radical, the oxidation rate must be faster than the radical formation and comparable to radical dimerization. For this reason, conventional strong oxidants may not be fast enough to catch the acetylene radical. In an extensive series of organic radical oxidations, Kochi found that copper II was an efficient radical scavenger 6 and that the rate of oxidation approached diffusion controlled. 7 Although copperIIseems to be an unlikely choice due to its relatively low oxidation potential, 8 Kochi observed extensively rearranged products attributable to carbonium ion rearrangement from the oxidation of neopentyl and cyclobutyl radicals. No coupled product from the corresponding radicals was detected. 9 Mechanistic studies of the oxidationof carbon radicals implicated unstable organocopper III whose decomposition could be controlled by choice of solvent and the variation of counterions of copper II. 10 119 In particular, it is known that use of acetonitrile as solvent significantly increases the oxidation potential of copper II. 11 Non-coordinating counterions increased the amount of oxidation products attributable to carbonium ion intermediates. Also, LiC10 4 , associated with the "special salt effect" observed in nucleophilic 14 . . . . . su b st1tut1ons, increase d t h e apparent car b on1um-1on derived products. 10 Sufficient control of copper II + Cu I X Cl0 4 -J RNu as an oxidant appears to be present to make oxidation of acetylene radicals possible. Because copper III is not very stable in general, 12 the acetylene copper complex will probably decompose somehow. Once the copper III acetylide has formed, will it decompose to form the carbonium ion and copper I or will it merely cleave homolytically as it formed? The formation of only diacetylene would indicate homolysis predominates over oxidation. If the acetylene carbonium ion does form, trapping ought to be relatively easy. 120 Solvent trapping is most likely. Trapping by acetonitrile would yield acetylene amides on aqueous workup. 13 ffi HC::C-N:£C-CH 3 0 H O II - 2 - HC=C-NH-C-CH 3 Although it may be difficult to prove that acetonitrile traps free acetylene carbonium ions rather than the copper III acetylide complex, other acetylenes with different substituents may show carbonium ion reactivity. For example, i-propyl acetylene or t-butyl acetylene cations may rearrange to the more stable allene cations. Methyl acetylene cation may react by in a similar fashion. Phenyl substituted acetylenes should stabilize the cations. Additionally, varying the phenyl substituents could significantly alter the acetylide reactivity. By sufficiently stabilizing the positive charge in the phenyl ring, vinylidene reactivity might 121 be observed. C=-C.: 12.2 References 1. S.I. Miller, J.I. Dickstein, Accts. Chem. Res., ,.. 9, 358(1976) 2. J. K. Kochi, R. D. Gilliom, J. Amer. Chem. Soc. , _.... 86, 5251 ( 1964) 3. H.E. DeLaMare, J.K. Kochi, F.F. Rust, J. Amer. Chen!. Soc., _.... 85, 1437 (1963) 4. a) J.K. Kochi, R. V. Subramanian, J. Amer. Ch€m. Soc., 87, 4885 (1965) ·"" b) J. K. Kochi, R. V. Subramanian, J. Amer. Chem. Soc. , ,.._,.._ 87, 1508 (1965) 5. Sldney W. Benson, ''Thermochemical Kenetics", Wiley & Sons, New York, N. Y. 1968, pp. 178-204 6. J. K. Kochi, F. F. Rust, J. Amer. Chem. Soc. , 84, 3946 (1962) --- 7. C. L. Jenkins, J. K. Kochi, J. Org. Chem. , 33, _.... ~103 (1971) 8. J.E. Huheey, ''Inorganic Chemlsti\y: Ftlnctples of Sti-ucture a-nd Reactivity", Harper& Row, New York, N. Y. 1972 p. 259 9. C. L. Jenkins, J. K. Kochi, J. Amer. Chem. Soc. , .,._,.,, 94, 843 ( 1972) 10. C. L. Jenkins, J.K. Kochi, J. Amer. Chem. Soc., ,.._,.._ 94, 856 (1972) 11. F.A. Cotton, G. Wilkenson, "Advanced Inorganic Chemistry", Wiley&Sons . New York, N. Y. 1972 p. 906 12. Ibid. , p. 922 --- 13. J. Rtter, P. Miniari. J. Amer. Chem. Soc. , 70, 4045, 4048 (1948) 14. H. L. Goering, H. Hopf, J. Amer. Chem. Soc . ., ~ ' -1·224 ( 1.>971}