A Study of Fundamental Reaction Pathways for Transition Metal Alkyl Complexes. I. The Reaction of a Nickel Methyl Complex with Alkynes. II. The Mechanism of Aldehyde Formation in the Reaction of a Molybdenum Hydride with Molybdenum Alkyls Citation Huggins, John Mitchell (1981) A Study of Fundamental Reaction Pathways for Transition Metal Alkyl Complexes. I. The Reaction of a Nickel Methyl Complex with Alkynes. II. The Mechanism of Aldehyde Formation in the Reaction of a Molybdenum Hydride with Molybdenum Alkyls. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/5pj3-sn83. https://resolver.caltech.edu/CaltechTHESIS:09302025-210640682 Abstract I. This study reports the rapid reaction under mild conditions of internal or terminal alkynes with methyl (acetyl- acetonato) (triphenylphosphine) nickel (1) in either aromatic or ether solvents. In all cases vinylnickel products 2 are formed by insertion of the alkyne into the nickel-methyl bond. These complexes may be converted into a variety of organic products (e.g. alkenes, esters, vinyl halides) by treatment with appropriate reagents. Unsymmetrical alkynes give selectively the one regioisomer with the sterically largest substituent next to the nickel atom. In order to investigate the stereochemistry of the initial insertion, ax-ray diffraction study of the reaction of 1 with diphenylacetylene was carried out. This showed that the vinylnickel complex formed by overall trans insertion was the product of the reaction. Furthermore, subsequent slow isomerization of this complex, to a mixture of it and the corresponding cis isomer, demonstrated that this trans addition product is the kinetic product of the reaction. In studies with other alkynes, the product of trans addition was not always exclusively (or even predominantly) formed, but the ratio of the stereoisomers formed kinetically was substantially different from the thermodynamic ratio. Isotope labeling, added phosphine, and other experiments have allowed us to conclude that the mechanism of this reaction does involve initial cis addition. However, a coordinatively unsaturated vinylnickel complex is initially formed, which can undergo rapid, phosphine-catalyzed cis-trans isomerization in competition with its conversion to the isolable phosphine-substituted kinetic reaction products. II. The reaction of CpMo(CO) 3 H (la) with CpMo(CO) 3 R (2, R = CH 3 , C 2 H 5 ) at 5O°C in THF gives the aldehyde RCHO and the dimers [CpMo(CO) 3 ] 2 (3a) and [CpMo(CO) 2 ] 2 (4a). Labeling one of the reactants with a methylcyclopentadienyl ligand it was possible to show that the mixed dimers MeCpMo (CO) 3 -CO) 3 MoCp (3b) and MeCpMo (CO) 2 ≡(CO) 2 MoCp (4b) are the predominant kinetic products of the reaction. Additionally labeling the carbonyl ligands of la with 13 CO led to the conclusion that all three of the carbonyl ligands in la end up in the tetracarbonyl dimers 4a if the reaction is carried out under a continuous purge of argon. Trapping studies failed to find any evidence for the intermediacy of either [CpMo(CO) 3 ] - or [CpMo(CO) 3 ] + in this reaction. A mechanism is proposed that involves the initial migration of the alkyl ligand in 2 to CO forming an unsaturated acyl complex which reacts with la to give a binuclear complex containing a three center-two electron Mo-H-Mo bond. This complex then selectively looses a carbonyl from the acyl molybdenum, migrates the hydride to that same metal, and forms a metal-metal bond. This binuclear complex with the hydride and acyl ligands on one metal reductively eliminates aldehyde, and migrates a carbonyl ligand, to give 4a directly. The other product 3a is formed by addition of two molecules of free CO to 4a. 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): Bergman, Robert G. Thesis Committee: Unknown, Unknown Defense Date: 12 June 1980 Record Number: CaltechTHESIS:09302025-210640682 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:09302025-210640682 DOI: 10.7907/5pj3-sn83 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 17709 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 01 Oct 2025 13:19 Last Modified: 01 Oct 2025 13:24 Thesis Files PDF - Final Version See Usage Policy. 52MB Repository Staff Only: item control page

A STUDY OF FUNDAMENTAL REACTION PATHWAYS FOR TRANSITION METAL ALKYL COMPLEXES I. THE REACTION OF A NICKEL METHYL COMPLEX WITH ALKYNES. II. THE MECHANISM OF ALDEHYDE FORMATION IN THE REACTION OF A MOLYBDENUM HYDRIDE WITH MOLYBDENUM ALKYLS. Thesis by John Mitchell Huggins In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California 1981 (Submitted June 12, 1980) ii Acknowledgements I am indebted to Prof. Robert G. Bergman and the members of his research group for invaluable help and guidance over the past four years. To the late Prof. G. Dann Sargent I owe my understanding of the principles of mechanistic investigations in chemistry. I thank Prof. Robert A. Huggins and Dr. Maurice L. Huggins for the scientific curiosity I have inherited. And I acknowledge the financial assistance of the Ferris Jewett Morre Fellowship (administered by Amherst College) 1976-77. JMH iii ABSTRACT I. This study conditions of reports internal or the rapid reaction under mild terminal alkynes with methyl (acetyl- acetona to) (triphenylphosphine) nickel (1) in either aromatic or ether solvents. In all cases vinylnickel products 2 are formed by insertion of the alkyne into the nickel-methyl bond. These complexes may be converted into a variety of organic products (e.g. alkenes, esters, vinyl halides) by treatment with appropriate reagents. Unsymmetrical alkynes give selectively the one regioisomer with the sterically largest substituent next to the nickel atom. In order to investigate the stereochemistry of the initial insertion, ax-ray diffraction study of the reaction of 1 with diphenylacetylene was carried out. This showed that the vinylnickel complex formed by overall trans insertion was the product of the isomerization of corresponding ~ reaction. Furthermore, this complex, isomer, subsequent to a mixture of it and slow the demonstrated that this trans addition product is the kinetic product of the reaction. In studies with other alkynes, the product of trans addition was not always exclusively (or even predominantly) formed, but the ratio of the stereoisomers formed kinetically was substantially different from the thermodynamic ratio. Isotope labeling, added phosphine, and other experiments have allowed us to conclude that the mechanism of this reaction does involve i n i t i a l ~ addition. However, a coordinatively unsaturated vinylnickel complex is initially formed, which can undergo rapid, phosphine-catalyzed cis-trans isomerization in competition with its conversion to the isolable iv phosphine-substituted kinetic reaction products. II. The reaction of CpMo(CO) 3 H (la) with CpMo(CO) 3 R (2, R = CH3, C2Hs) at SO°C in THF gives the aldehyde RCHO and the dimers [CpMo(CO) 3 J 2 (3a) and [CpMo(CO) 2 J 2 (4a). Labeling one of the reactants with a methylcyclopentadienyl ligand it was possible to show that the mixed dimers MeCpMo (CO) 2= (CO) 2 MoCp (4b) MeCpMo (CO) 3 -CCO) 3 MoCp (3b) and are the predominant kinetic products of the reaction. Additionally labeling the carbonyl ligands of la with 13 co led to the conclusion that all three of the carbonyl ligands in la end up in the tetracarbonyl dimers 4a if the reaction is carried out under a continuous purge of argon. Trapping studies failed to find any evidence for the intermediacy of either [CpMo(CO) 3 J- or [CpMo(CO) 3 J+ in this reaction. A mechanism is proposed that involves the initial migration of the alkyl ligand in 2 to CO forming an unsaturated acyl complex which reacts with la to give a binuclear complex containing a three center-two electron Mo-H-Mo bond. This complex then selectively looses a carbonyl from the acyl molybdenum, migrates the hydride to that same metal, and forms a metal-metal bond. This binuclear complex with the hydride and acyl ligands on one metal reductively eliminates aldehyde, and migrates a carbonyl ligand, to give 4a directly. The other product 3a is formed by addition of two molecules of free CO to 4a. V TABLE OF CONTENTS I. The Reaction of a Nickel Methyl Complex With Alkynes. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results and Discussion 2 .. .. . . .... .. .. . . ... . ... . .. . . . . 6 Summary and Conclusion .............................. 37 Experimental ....•................................... 39 References .............................•............ 57 II. The Mechanism of Aldehyde Formation in the Reaction Of a Molybdenum Hydride With Molybdenum Alkyls. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..63 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71 Introduction Discussion .......................................... 83 Experimental ........................................ 91 References ......................................... 109 III. Appendix A.......................................... 112 IV. Appendix B.......................................... 125 v. The Propositions. Proposition Abstracts ............................... 129 A. A Stereoselective Synthesis Mediated by A Chiral Transition Metal Complex ............................ 132 vi B. Nucleophilicity and Transition Metal Carbonyl Anions: A Study of the Relationship Between Ion-Pairing and The Competition Between Sn2 Displacement and Electron Transfer Mechanisms ................................. 148 c. The Synthesis and Reactivity of a Novel o--Alkylµ.-Alkylidene Iron Complex. A Model for the CarbonCarbon Bond Forming Process in Fischer-Tropsch Catalysis? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 D. The Absolute Rate Constant for .c.i..§.-Trans Isomerization of Vinyl Radicals ..................... 177 E. The Study of Cationic Bisphosphine Methylnickel(II) And Platinum(!!) Complexes: The Search for Evidence of An a-Hydride Migration Process ...................... 190 1 PART I THE REACTION OF A NICKEL METHYL COMPLEX WITH ALKYNES 2 INTRODUCTION The propagation step in the low pressure polymerization of alkenes and alkynes by Ziegler-Natta catalysts is generally believed to be the concerted cis insertion of the unsaturated monomer into the metal carbon bond of the growing polymer chain. 1 The concerted cis nature of this reaction, which can perhaps better be thought of as an alkyl migration, or 1,2-addition to the alkene or alkyne, has gone unchallenged until quite recently, despite the exceedingly rare occurrance of stoichiometric models involving the 1,2-addition of transition metal alkyls to alkenes or alkynes, which should also proceed by a cis concerted pathway. The corresponding reaction involving metal hydrides is quite common, proceeding selectively cis with only a few exceptions. 2 The few stoichiometric insertion reactions of transition metal alkyls that have been reported are usually specific to a few highly "activated" alkenes or alkynes, such as tetrafluoro- ethylene, hexafluoro-2-butyne, or diphenylacetylene, 3 - 6 with a few exceptions. 7 18 More important, a number of the reactions of metal alkyl and hydride complexes with alkynes have recently been observed to give either stereorandom or exclusively trans products, sometimes under kinetic conditions. 21516 Some complexes have even been observed to give exclusively trans addition for one alkyne, and exclusively cis addition for others. Moreover, two recent studies have reported the observation of reactions involving trans addition mechanisms, by backside attack upon coordinated alkyne by a second molecule of alkylating agent. 6 3 Finding an explanation for these highly variable results has presented considerable difficulty, and this has led in turn to suggestions of al terna ti ve mechanisms for this process, some of which are shown in Scheme I. One such suggestion would account for trans addition products with a concerted trans addition pathway. 5 A more radical proposition by M.L.H. Green and coworkers 9 involves an initial a-hydride migration step, to form a metal alkylidene hydride intermediate. This intermediate can then add alkene(or alkyne) to the alkylidene ligand to form a metallocyclobutane (butene) complex. The reverse of the original a-hydride migration step specifically to the original carbon atom then leads to overall cis addition by a stepwise mechanism. The significance of these proposals is not that present evidence is sufficient to establish their viability, but rather that the existing studies cannot rule out these possibilities. In the present study we have observed the facile reaction at room temperature of a variety of alkynes with methyl(acetylacetonato)(triphenylphosphine)nickel Cl) to give vinylnickel complexes (the product of 1,2-addition) in nearly quantitative yield. The mechanism of this reaction has significant bearing on both the question of cis vs trans addition and Green's hydride migration pathway. a- In particular this reaction was observed to give both cis and trans addition products under kinetically controlled conditions; for some alkynes the trans addition product was found to be predominant. Nevertheless we have obtained evidence that this reaction involves initial cis 4 addition, giving an intermediate which forms E-and Z-vinylnickel products by kinetically controlled pathways. 10 5 SCHEME I els Concerted: I I C M-111 R C I trans Concerted= I C M-R + Ill C I a-Hydride Migration: . J-c=cH /CR2, / ·C II M,..........c......._ ◄ - ~ ~-1R2 /C=c, 6 RESULTS AND DISCUSSION A. The Synthesis and Properties of Ni(acac)CPPh 3 >ca 3 . In 1973 Yamamoto and coworkers reported the synthesis and characterization of Ni(acac)(PPh 3 )cH 2 cH 3 and Ni(acac)(PPh 3 ) 2 cH 3 (acac = 2 ,4-pentanedionato) .1 1 The square-planar geometry of the ethyl complex was firmly established in a subsequent x-ray crystal structure analysis by Cotton and coworkers. 12 13 c, The 1 H, and 31 P NMR spectra of these complexes were used to establish that the phosphine ligand is very labile at room temperature (vida infra), and that both acac exchange and {3hydride elimination processes are quite slow. These complexes have been observed to insert carbon monoxide at low temperature to give acyl derivatives which decompose by disproportionation upon warming. 13 In addition the synthesis and characterization of Ni(acac) (PPh 3 >c 6 H5 and its reaction with olefins and alkyl halides was reported recently. 8 Ni(acac)(PPh 3 )cH 3 (1) was prepared by the method of Cotton 12 for the corresponding ethyl complex from Ni(acac) 2 , PPh 3 , and Al(CH 3 ) 2 ocH 3 . The crude reaction product contains considerable excess PPh 3 which may be removed by repeated recrystallization from toluene-hexane or tol uene-acetoni tr ile mixtures. Analytic- ally pure 1 crystallizes as yellow-brown needles; 1 H NMR (C6D 6 ): 80.07 Cs, 3H, Ni-CH 3 ), 1.40, 1.90 (s, 3H each, acac-CH 3 's), 5.28 (s, lH, acac-H), 6.9-7.1, 7.6-8.0 ppm (complex, 15H, -Ph). Complex 1 is stable as a solid, but solutions decompose rapidly upon exposure to air, and it is soluble in aromatic and ether 7 solvents and insoluble in hydrocarbons. By elemental and monophosphine complex spectral analysis 1 in the solid state and is clearly a in solution. Observation of two distinct singlets in the 1 H NMR for the methyl absorptions of the acac ligand indicates a rigid square-planar geometry. The absence of splitting of the nickel methyl resonance at room temperature can be explained by the rapid exchange of phosphine on the NMR time scale. Using the temperature dependence of the 31 P NMR spectrum, Yamamoto was able to determine that the rate constant for phosphine exchange is 1.6 x 10 2 sec-l for Ni (acac} (PPh3} CH2CH3 and 2.8 x 10 3 sec- 1 for Ni (acac} (PPh3} 2CH3 in toluene. In a similar manner acac exchange was determined to be much slower; a rate constant of 1.1 x 10 1 sec- 1 was measured for Ni(acac}(PPh 3 >cH 2cH 3 at 40°C in toluene. 11 In contrast to the reported isolation by Yamamoto and coworkers of the bisphosphine complex Ni(acac}(PPh 3 >2 cH 3 , 11 we were unable to detect this species in solutions of 1 and added phosphine. Likewise we were unable to isolate a bisphosphine complex, even when 1 was recrystallized from solutions saturated in PPh 3 . The room temperature 1 H NMR spectra of 1 with and without added PPh 3 are completely superimposable, except for some broadening of the acac-methyl resonances. 14 Cooling a toluene-d 8 solution of 1 results in a downfield shift of the nickel-methyl resonance; below -75°C this resonance splits into a broad doublet, 5 Hz, at 8o.3o ppm. In the presence of added PPh 3 J= the nickel methyl resonance shifts in a similar manner. However 8 no coupling to phophorous is observed at -75°C. The 31 P NMR spectra of 1 are much more informative. The spectrum of 1 in toluene-d 8 contains one singlet at 42.5 ppm downfield from external free PPh 3 . Upon cooling to -9o c this 0 signal remains sharp shifting slightly to 43.8 ppm downfield (Figure 1) and no coalescence or broadening was observed at intermediate temperatures. Addition of PPh 3 to these solutions, however, results in drastic changes in these spectra. The 31 p NMR spectrum of solutions of 1 and one equivalent added PPh 3 exists as a very broad singlet centered at 24 ppm downfield at room temperature (Figure 2). Upon cooling the spectrum virtually disappears until temperatures below -6o 0 c are reached. At -90°C two sharp singlets at 43 and -3.8 ppm are observed in roughly equal intensity (Figure 3). In the presence of only half an equivalent of added PPh 3 analogous behavior is observed. At room temperature a broad singlet at 28 ppm downfield is observed; upon cooling to -90°C two absorptions at appear 43 ppm and -3.8 ppm in a ratio of 2:1. These spectra can be explained by the existence of a very rapid exchange between free and bound PPh 3 , which becomes slower than the NMR time scale only below about -75°C. Furthermore the integrated intensities of the absorptions for bound and free PPh 3 at -90°C, and the position of the averaged signal at room temperature, indicate that added PPh 3 is largely dissociated both at room temperature and below. These observations lead to the conclusion that solutions of 1 and added PPh 3 do not contain any observable concentration of 9 Figure 1. 31 P NMR Spectrum of 1 in toluene-d 8 at -90 C. 0 E a. a. c,o 0 0 r-f 0 N 0 rn 0 ~ 10 Figure 2. 31 P NMR Spectrum of Toluene-d 1 and One Equivalent PPh 3 in 0 8 at 30 C. a N E I 0.. 0.. ro a.,... I a a N a V1 11 Figure 3. 31 P NMR Spectrum of land One Equivalent Added PPh 3 in Toluene-d 8 at -9o•c. E 0. 0. ro 0 0 rl 0 N 0 rn 0 :j-1 0 lfJ 12 the bisphosphine complex Ni(acac)(PPh 3 >2 cH 3 . This behavior is entirely consistent with the observations by Yamamoto and Cotton for Ni(acac)(PPh 3 )cH 2 cH 3 , 11 , 12 which does not form a bisphosphine complex. It is, however, inconsistent with the reported preparation of Ni(acac)(PPh 3 > 2 cH 3 and its properties. 11 we are at a loss for a good explanation for this discrepancy. 15 The analogues of 1 Ni (acac) (PPh 3 ) co 3 (I-d 3 ), Ni (acac)- (PPh 3 > c 6 H5 (5) 8 , and Ni(acac) (PCy 3 )cH 3 (6) 16 were prepared from the appropriate aluminum reagents. Both 5 and 6 have been reported previously. These complexes all have similar properties, except that in the 1 H NMR spectrum of 6 in benzene-d 6 the nickel methyl resonance <8-0.18 ppm) is a doublet due the coupling to phosphorous (JPH = 5 Hz), indicating that the PCy 3 ligand in 6 is not labile on the NMR time scale. B. The Reaction of 1 with Alkynes Stoichiometry and Kinetics. Complex 1 reacts rapidly at room temperature with equimolar amounts of a number of both terminal and internal alkynes to give nearly quantitative yields of vinylnickel complexes formed by the 1,2-addition of the metal carbon bond to the alkyne (Scheme II). These alkynes include; diphenylacetylene, phenyl-1-propyne, phenylacetylene, dimethylacetylenedicarboxylate, tert-butylacetylene, and 4,4-dimethyl-2- pentyne. Internal alkyl alkynes such as 2-butyne and 3-hexyne also react with l; equivalent of in these cases, alkyne is consumed however, and more than one mixtures of products Scheme TI CH'3 • )Ni H-(( CH 3 LiAIH4 H+ or ► 2 0 \ or AIMe3 [~ \N. I I 3 w I-' (I)(R)C=C(R')CH PPh 3 O /C(R)=C(R')CH (CH ) (R )C=C(R')CH 3 3 (MeO C)( R)C=C(R' )CH 3 2 (H)(R)C=C(R' )CH 1 · + RC=CR' 3 3 O \PPh o ; CH-1~ ~ 3 14 result. 17 Only 1-pentyne and acetylene itself failed to give tractable products. 18 These vinylnickel complexes can be treated with either LiAlH 4 or acid to give mixtures of E-and z-alkenes in good yield (Scheme II). In this manner it was possible to confirm the structure of the organic ligand in these complexes. Similarly Ni(acac)(PCy 3 )cH 3 (6) reacts with alkynes, but much more sluggishly; for example reaction of 6 with an equimolar amount of PhC=CCH 3 , both O.lM in benzene, has a half life of about 10 hours at 4o 0 c. The reaction of 1 with alkynes follows bimolecular kinetics, first order in both 1 and alkyne. Comparable rates are observed in THF and benzene solvent. Qualitatively the relative rates of reaction follow the order: PhC=CPh ~ PhC:CH ~ Meo 2 cc:::cco 2 Me ~ PhC=CCH 3 >> Me 3 CC=CH ~ Me 3CC=CCH 3 >> CH 3C:CCH 3 ,;t CH 3CH 2c This order follows a trend = CCH 2CH 3 of increasing reactivity with increasing ability of the alkyne substituents to support 7T-back bonding in a metal-alkyne7T-bond. In addition terminal alkynes appear to react somewhat faster than their methyl substituted analogues, suggesting some steric control of alkyne reactivity as well. Complex 1 does not react with alkenes at room temperature. With prolonged heating (56°C) minimal conversions have been observed for a few highly activated alkenes (e.g. dimethyl- maleate). Other alkenes either do not react or polymerize without consumption of the nickel complex. This is in contrast to the 15 facile reaction of Ni(acac)(PPh 3 >c 6 H5 (5) with alkenes reported by Yamamoto and coworkers. 8 A significant secondary deuterium isotope effect is observed in a comparison of the reactivity of 1 and l-d 3 . Diphenylacetylene was allowed to react with an excess of a mixture of 1 and 1-d 3 at room temperature. Protonation of the resulting product gave a mixture of E-and Z-l,2-diphenylpropenesd 0 and -d 3. After purification by preparative gas chromatography, analysis of these alkenes by 180 MHz 1 H NMR and mass spectroscopy determined that k (1) /k (I-d 3 ) = 1.24 <± .05). This corresponds to a deuterium isotope effect of 3 J1.24 = 1.07 per deuterium. The reaction of 1 with alkynes is strongly inhibited by added phosphine. A plot of the second order rate constant for the reaction of 1 with diphenylacetylene at 40°C vs l/[PPh 3 J gives a straight line dependence, as shown in Figure 4. Two mechanisms which could give this observed inhibition are illustrated in Scheme I I I. The crucial di st inct ion is that in mechanism phosphine inhibition can occur only if A significant concentrations of Ni (acac) (PPh 3) 2cH 3 build up under the reaction conditions, whereas in mechanism Ba preequilibrium involving loss of phosphine accounts for the observed inhibition. solutions of 1 and added PPh 3 contain no Because detectable concentrations of this bisphosphine complex (vide supra), we conclude that the initial step in this reaction is the reversible substitution at nickel of phosphine by alkyne (mechanism B). 30 0 5 1/[P P h3]added 10 15 ( L· M- 1) + /V Figure 4. Rate of reaction of I with PhC=CPh vs PPh 3 cone. ~ 10 () CD Cf) ~ 20 ..0~ X ......... Cf) g_ 40 -;;;::- 50 60 20 I-' O'I 17 Scheme .m ~ Mechanism A llRc=cR ► Products Mechanism B ! Products 18 In a competition experiment, a mixture of PhCECPh and Ph C:C CH 3 was a 11 owe a to r ea ct wi th a def i c i ency of 1 <L = PP h 3 ) , yielding a mixture of vinylnickel complexes corresponding to a ratio k (PhC=CPh) /k (PhC:CCH 3 ) = 1.4. In a parallel experiment reaction of 6 CL= PCy 3 ) with a mixture of PhC=CPh and PhCECCH 3 gave a ratio k(PhCECPh)/k(PhCECCH 3 ) = 1.2. This result, that the reactivity ratio depends upon the nature of the phosphine, suggests that the substitution of phosphine by alkyne proceeds by an associative mechanism. A dissociative mechanism would have predicted a constant ratio independent of the phosphine or the concentration of the dissociated nickel present species. Consistent with this conclusion is the observed broadening of the acac-cH 3 resonances in the 1 H NMR spectrum of 1 upon addition of excess phosphine, which can be explained by an associative phophine catalyzed acac exchange process. 14 Taken together, these observations suggest that the squareplanar complex Ni(acac)(RC:CR)CH 3 is most likely the intermediate which precedes the insertion step. The absence of phosphine in this intermediate will be significant in our later discussion of the mechanism of this reaction. This conclusion, that the alkyne must enter a square-planar coordination site prior to insertion, has precedent in the reaction of the square-planar complexes Pt(L) 2 (X)(vinyl) with CF 3C:CCF 3 ; insertion occurs only if Xis an easily displaced ligand such as acetone, but not at all for nonlabile ligands such as chloride. 19 19 Regioselectivity. The reaction of 1 with unsymmetrical alkynes was observed to be highly regioselective, always giving only one regioisomeric vinylnickel complex. The direction of this specificity was confirmed by treatment of the vinylnickel complexes with either LiAlH 4 or acid to give mixtures of E-and zalkenes in high yield. These results, presented in Table 1, show that even alkynes with only alkyl substituents (for example 3,3dimethyl-l-butyne) give only one regioisomer . Interestingly alkynes with sterically dissimilar substituents a l w a y s ~ ~ vinylnickel complex ~ith .tM larger group nearest .tJig nickel atom . That this selectivity arises from steric and not electronic influences was confirmed by a study of the reaction of a variety of p-substi tuted diphenylacetylenes p-X-PhC = CPh (X= CH 3 , OCH 3 , Cl) with 1. The 1 H NMR spectra of the resulting vinylnickel products clearly show two sets of trans vinyl-CH 3 and two sets of downfield shifted o-phenyl proton resonances in nearly equal concentrations; in no case was a significant excess of one regioisomer observed. somewhat unexpectedly these observations are consistent with a transition state for insertion that is sensitive only to steric crowding, furthermore, the metal end of the Ni-CH 3 bond is sterically less active than the methyl group in the insertion transition state. Although sterically controlled selectivity of this kind exists in the hydrozirconation reaction, the reversed preference is observed. In the hydrozirconation reaction the 20 Table 1. Products Formed on Reaction of Vinylnickel Complexes With Acid and LiAlH . 4 + Alkyne THF Reagent Alkenes r.t. Alkyne Alkenes Reagent Yield PhC=CPh E,Z-Ph(H)C=C(CH )Ph 3 TsOHa 100% PhC= CPh E,Z-Ph(H)C=C(CH )Ph 3 LiAlH 79% Ph(H)C=C(CH ) 3 2 LiAlH Phc=ccH 3 PhC~ CH E,Z-Ph(H)C=C(H)CH t-Buc= CH E,Z-(t-Bu) (H)C=C(H)CH cH c=ccH 3 3 CH (H)C=C(CH ) 3 3 2 TsOH EtC=CEt E, Z- (Et) (H) C=C (CH ) Et 3 LiAlH t-Buc=ccH 3 t-Bu(H)C=C(CH ) 3 2 4 4 TsOH 3 3 LiAlH 78% 68% 4 4 47% TsOH aTsOH = p-Toluenesulfonic acid. bA two fold excess of 2-butyne was employed. GC of the organic products revieled 3 longer retention time products in circa 10% yield each; these products were not identified. 21 zirconium hydride preferentially migrates to the most hindered carbon in unsymmetrical alkynes. 20 Reactions of the Vinylnickel Complexes. In addition to acid or LiAlH4, a variety of reagents convert the vinylnickel complexes into Nucleophilic organic products (Scheme II, Table 2). reagents such as CH 3 Li, LiAlH 4 , and Al(CH ) 3 , as 3 well as electrophilic reagents such as H+ and I 2 give reasonable yields of their respective organic products. products formed in these reactions The organometallic have not been well- z- characterized, but importantly all gave mixtures of E-and alkene products in,variable ratios. In no case was a single stereoisomer yield. observed in good The lack of stereoselectivity in the reactions of the vinylnickel complexes required that an alternate method for determining the stereochemisty of these complexes be found. C. Stereochemistry of Addition. Reaction of 1 with Diphenylacetylene. We have examined the reaction of 1 with diphenylacetylene especially closely. A solution of 205 mg (1.15 mmol) of PhC=CPh in 1 ml toluene was added to 500 mg (1.15 mmol) of I in 20 ml toluene at room temperature to give Ni(acac) (PPh 3 ) [C(Ph)=C(Ph)CH 3 1 quantitative yield. (3) in Precipitation with hexane allowed isolation of 3 as an orange solid in 89% overall yield. The conversion of I into 3 can be conveniently monitored by NMR; no intermediates could be reagents, observed. is In benzene complete in this less reaction, than 30 0.lM minutes in both at room 22 Table 2. Reactions of the Vinylnickel Complexes, Ni(acac) (PPh ) (C(Rl}=C(R )cH ). 3 2 3 Rl R2 Reagent Products Yield Ph Ph TsOHa E,Z-Ph(H)C=C(CH )Ph 3 100% Ph Ph LiAlH E,Z-Ph(H)C=C(CH )Ph 3 79% Ph Ph A1Me 3 E,Z-Ph(CH )C=C(CH )Ph 3 3 88% Ph Ph MeLi E,Z-Ph(CH }C=C(CH )Ph 3 3 52% Ph Ph I2 E,Z-Ph(I)C=C(CH )Ph 3 66% Ph Ph CO/MeOH E,Z-Ph(MeO C)C=C(CH )Ph 2 3 39% Ph H TsOH E,Z-Ph(H)C=C(H)CH 68% Ph H I2 E,Z-Ph(I}C=C(H)CH Ph H CO/MeOH 4 3 3 E,Z-Ph(MeO C)C=C(H)CH 2 3 ap-Toluenesulfonic acid. 30% 29% 23 temperature. The vinylnickel complex 3 is an orange-red solid with solubility properties similar to 1. Treatment of a solution of 3 in THF with excess p-toluenesulfonic acid gave E-and z-1,2diphenylpropenes in quantitative yield. The 1 H NMR spectrum of 3 has acac absorptions at 81.18 and 1.78 Cs, 3H each) and 5.09 ppm Cs, lH), and a vinyl methyl resonance at 82.03 ppm Cd, JpH= 1.5 Hz, 3H). The vinyl methyl signal phosphorous CJ pH = 1.5 Hz), is split by coupling to indicating that in this complex phosphine exchange is slow on the NMR time scale. This slower phosphine exchange is general for all the vinylnickel complexes. In addition there is an unusual aromatic proton resonance, a doublet of doublets at 88.70 ppm integrating as two protons CA 2 MM' quartet, JAM= 1, JAM' =7 Hz). This low field absorption, assigned to a pair of o-phenyl protons on the /3-phenyl group of the vinyl ligand, will be discussed further below. Complex 3 is the initial product of the reaction. However after standing in solution at room temperature for several days, or heating to 56°c, 3 is converted partially into a new vinyl complex 4 . Complex 4 has a new set of acac absorptions at 85.20, 1.87, and 1.41 ppm and a new vinyl methyl doublet shifted strongly downfield at 83 . 37 ppm (JPH = 1 Hz). Additionally the absorption at 88.70 ppm in 3 is absent in 4. The similarity of 3 and 4 has prevented separation of these two compounds; however, based upon the NMR, and the fact that protonation of mixtures of 3 and 4 also give only 1,2-diphenylpropenes, we conclude that 3 24 and 4 are ~i§/~~~n§ isomeric about the double bond. This conclusion is supported by the observation that on continued heating the above solution of 3 and 4 reaches an equ~librium ratio for 3/4 of about 3.0 in about one hour at 56°c. Clearly 3 is the kinetic product of the reaction of 1 and PhC=CPh, and heating converts it to an equilibrium mixture of two isomers. Crystal and Molecular Structure of 3. Assignment of the stereochemistry of the vinyl ligand in 3 was essential to this investigation. This prompted us determination of by 3 to undertake x-ray diffraction. conclusively shown to be the Z-isomer, addition, the structure Complex 3 was the product of trans by x-ray crystallography on a single crystal obtained from a toluene-hexane solution. An ORTEP drawing of 3, excluding the phosphine phenyls, is shown in Figure 5 and ~elevant interatomic distances and angles are presented in Appendix A. Complex 3 is a square-planar vinylnickel complex with the carboncarbon double bond almost perpendicular to the plane of the complex. Both the carbon-carbon double bond distance of 1.327 A and the Nil-Cl-C7 angle of 123.7° support the characterization of the organic ligand as a vinyl group. Assuming the structure of 3 in solution is similar to that in the solid state, the x-ray study allows us to rationalize the unusual NMR chemical shift of the aromatic resonance at 88.70 ppm. This resonance is assigned to the two o-phenyl protons on the {3-phenyl group cis to the nickel for several reasons. With the vinyl ligand at right angles to the plane of the nickel 25 Figure 5. ORTEP drawing of Z-Ni(acac) (PPh )[C(Ph)=C(Ph)CH ] 3 3 ( 3) • ~ CD (..) -- CII ..( ..) - co ( ..) C0 (..) N ' -' - ( ..) 26 complex these ortho protons are placed directly over the nickel atom, al though the distance in the crystal is non-bonding 0 (greater than 2.0 A). A similar effect is observed in the strong downfield shift of the .Qll vinyl methyl in 4 relative to the trans vinyl methyl · in 3. Relative shifts of this kind are observed in the products of the reaction of 6 with PhC:CPh and PhCECCH3. In Z-Ni (acac) (PCy 3 ) [ C (Ph) =C (Ph) CH 3 ] (13) absorptions at 9.28 CA2MM' quartet, JAM= 1 Hz, JAM'= 7 Hz) and 8.0 ppm <A2MM' quartet, JAM= 1 Hz, JAM'= 7 Hz) Ni (acac) (PCy 3 ) [C (Ph) =C (CH 3 ) 2 1 (14) are observed, whereas in the only low field absorption is at 87.95 ppm CA 2 MM' quartet, JAM= 1 Hz, JAM'= 7 Hz). Since there is no triphenylphosphine in these complexes to confuse the assignment, it is clear that the ortho-phenyl protons of both the a-and /3-vinyl phenyl's are strongly shifted downfield, cis with the /3-phenyl protons shifted furthest downfield of the two. Overlap of the triphenylphosphine absorptions probably obscures the protons for the a-phenyl group in complexes 3 and 4. The observed coupling pattern in these absorptions is also consistent with their assignment as o-phenyl protons. Reaction of Ni(acac) (PPh 3 )Ph (5) with PhC;CCH 3 . The unusual result that 3, the kinetic product of the reaction of 1 and PhC:CPh, was the product of overall trans addition led us to investigate the reaction of Ni(acac)(PPh 3 )Ph (5) with PhC:CCH 3 . If this reaction also proceeded with trans stereoselectivity then 4, the isomer opposite to 3, should be the kinetic product. Contrary to this prediction, 3 was the only kinetic product of 27 the reaction of 5 with PhC:CCH 3 . This reaction proceeded at a rate comparable to that for Phc=cPh with 1, and again heating the product solution led to an equilibrium mixture of 3 and 4. Clearly 3 is the sole kinetic product of both reactions, and heating leads to an equilibrium mixture of the cis and trans vinyl products (Scheme IV). Reaction of 1 With PhC:CCB 3 . Complex 1 reacts with PhC:::CCH 3 to give Ni{acac) (PPh 3 ) [C{Ph)=C{tH 3 ) 2 1 {7) as the only product. The 1 H NMR spectrum of 7 shows two vinyl methyl signals. One at 2.95 ppm {d, JpH= 1 Hz) can be assigned a s ~ , and the other at 1.75 ppm Cd, JpH= 1 Hz) as trans to the nickel atom by analogy to 3 and 4. Treatment of 7 with LiAlH 4 gave l-phenyl-2-methylpropene in 78% yield. In this reaction, using a deuterium labeled methyl group, it is possible to distinguish~ and trans reaction pathways directly. Reaction of l-d 3 cis-co 3 /trans-co 3 -7 = 1.6 with PhC:CCH 3 gave a ratio of <± .1) as kinetic products. Likewise the reaction of 1 with PhC:CCD 3 gave a ratio of cis-CH 3 /trans-CH3 = 1.8 <± .1) {Scheme V). Both product mixtures approach an equilibrium ratio of about 1.0 after several hours. The inversion of the product ratio in these two reactions excludes a deuterium isotope effect as the cause of the observed predominance of cis addition. 21 Reaction of 1 With Other Alkynes. Using the strong downfield shift of the cis-vinyl methyl in the 1 H NMR spectrum to assign stereochemistry, it was possible to determine both the kinetic Scheme Tiz: Fast 5 ;Ph Ni + PhC=CCH 3 •• X ~ \pph3 1 ;CH 3 'Ni + PhC=CPh \pph3 ~ ► ► Slow 3 Ph\ ;Ph 0 C=C \CH3 0 \pph 3 3 4 C)tSi ~ C0I \ PPh Ph\ ;CH3 0 ;C=C 'Ni \ph 00 N Scheme -sl.. 1-d 3 Ni-CD 3 + PhC=CCH 3 1 Ni-CH 3 + PhC=CCD 3 ~ ► ► 1.0 1.8 •• •• 1.6 1.0 Ph\ ;CD 3 Ph\ /H 3 C=~ C=C Ni/ CH 3 _ · Ni/ \co 3 N \0 30 and thermodynamic ratios of cis and trans addition products for a number of alkynes. These results, summarized in Table 3, contribute to s~veral important conclusions: a} in all cases, different kinetic and thermodynamic ratios of E-and z-vinylnickel products are observed; b} the subsequent and slower isomerization of the initially obtained products to an equilibrium mixture of isomers establishes that they are the result of kinetically control led pathways; c} depending upon the alkyne involved, either the E-or Z-isomer can be the favored kinetic product; and d} the predominant thermodynamic product has the subs ti tuent on the /3-carbon larger cis to nickel. D. The Mechanism of Addition. One way to account for the stereochemical . and kinetic observations presented here would be to postulate a mechanism involving parallel concerted~ and trans addition pathways, where slight changes in the structure of the alkyne involved can strongly affect the relative rates of cis and trans addition. This mechanism would require that PhCECPh and PhC:CCH 3 , as well as CcH 3 > 3 cc=cH and Phc=cH, have opposite stereochemical preferences. Moreover this mechanism would require some pathway for cis-trans isomerization of the products that is independent of the mechanism of addition. We find these requirements rather arbitrary and prefer the alternative hypothesis that only one addition pathway exists. This requires that there exist an intermediate capable of isomerization about the carbon-carbon double bond. 31 Table 3. Stereochemistry of Addition of Nickel Complexes to Alkynes. R2>=<R1 /RI (acac)Ni\L + R2c=cR 3 1 , R 1 = CH (acac)Ni\ L Irons C/S 3 R3 5, R1 = Ph Rxn Rl 1 CH 2 Ph 3 CH 4 CD Sb CH 6b CH 3 3 3 3 3 R2 R3 Ph Ph Ph CH Ph CD Ph CH t-Bu Ph r a Ki~etic Prod ·1rTh~rmo. Prod.l - CJ.S trans trans cis 0 100 25 75 100 0 75 25 65 35 50 50 61 39 50 50 H 30 70 100 0 H 65 35 100 0 3 3 3 arelative percent. b for these alkynes determining the kinetic ratio of products is approximate, because isomerization of the products is competitive with the initial insertion reaction. 32 A mechanism of this type that reasonably accounts for our results is presented in Scheme VI. The initial step is the reversible substitution at nickel of phosphine by alkyne to form the intermediate Ni(acac}(RC:CR}CH 3 . Then, in what is probably the slow step in the reaction, the alkyne inserts into the nickel methyl bond in a £.i.§. manner to give a coordinatively unsaturated intermediate cis-.B. ~ - B is the crucial intermediate that can either add phosphine to give the coordinatively saturated cis addition product, or isomerize to give trans-B. (See section E for a discussion of the effect of [PPh 3 J on the rate of this isomerization.> Trans-B can then add phosphine to give the trans addition product. The results summarized in Table 3 require that k 1 be competitive with (if not faster than} k 2 . The predominant kinetic product under these circumstances will depend upon the relative rates of k 1 , k_ 1 , k 2 , and k 3 and not on either the ~/trans product equilibrium or the stereochemistry of the insertion step. Only in the case where both R1 and R3 are CH 3 or co 3 (Reactions 3 and 4, Table 3) can the stereochemistry of the insertion step be observed in the products. In this case k 1 and k_ 1 , and k 2 and k 3 are expected to be nearly equal; and the predominance of one isomer among the products can only arise from a competition between k 1 and k 2 . Therefore the observation of more £i..§. than trans product in reactions 3 and 4 requires that cis-B is formed first, that is the insertion step proceeds cis. 33 Scheme 3ZI ~ LI • (acac)Ni-R' + RC=CCH 3 L I 1 (acac)Ni-CH 3 + RC=CR 5, R' =Ph 1 lt lt RC=CCH 3 (acac)Ni-Ph I R R>=<CH 3 (acac)Ni • R' R' (acac)Ni >=<cH 3 cis-B trans- B }z[L] Hk3[ L] R R' (acac)Ni~CH 3 .- L 34 These stereochemical observations do not provide detailed information about the mechanism of the insertion process itself. The observed predominance of the £i§ addition product in reactions 3 and 4 argues against a direct trans addition pathway. An alternative hypothesis involving a-hydride migration (Scheme I) might also give a cis specific result, but this mechanism would predict a primary kinetic deuterium isotope effect. The observed kinetic ratio k(l)/k(l-d ,,., N 3 ) = 1.24 is too small to be primary, and clearly represents a secondary effect. On the basis of these arguments we conclude that the most likely mechanism involves a concerted cis insertion process. E. The Mechanism of Isomerization. Th e mo s t di r ect me ch an i s m f o r th e £i§ - B j;L,_g,IlS-B isomerization involves a simple unimolecular rotation about the 7T-bond (perhaps assisted by contributions from resonance forms of type 15). This mechanism, or any other involving a unimolecular 15 pathway for isomerization of B, would predict a strong phosphine concentration dependence on the ratio of cis and trans products in reactions 3 and 4, where the rate of isomerization is 35 competitive with the rate of addition of phosphine to give product. This dependence was .ntl observed: over concentrations ranging from zero to 1.0 Madded PPh 3 (.lM in 1) the £li/trans product ratio from the reaction of l-d 3 with PhC:CCH 3 changed very little. Likewise, highly charge-separated transition states can be discounted because virtually no effect upon the ~/trans product ratio was observed in changing solvent from benzene to THF, or adding 0.2M NaBPh4 in THF. In order to account for these observation, we suggest that both the isomerization and the product-forming steps involve phosphine. Phosphine can catalyze isomerization of the carboncarbon double bond if free PPh 3 has two modes of addition to the intermediate B. Addition at the nickel center gives product, whereas reversible addition to the /3-vinyl complex 16, resulting in isomerization of Addition of phosphine to the carbon leads to the double bond. vinyl ligand might be enhanced in B by delocalization of electron density toward nickel, through resonance forms of the type 15, making the/3-carbon somewhat electrophilic. Equilibration of the kinetic products would then , e-~-,c/ r. PPh 3 (acac )N 1 ✓. 4 ~, CH3 R 16 36 involve loss of phosphine to reg e nerate intermediate B (dotted arrows in Scheme VI), followed by competitive readdition of phosphine either at nickel or at carbon. This mechanism is supported by the observation that phosphine is much less labile in the vinylnickel complexes than in 1, accounting for the slow approach to equilibrium. of approach to phosphine . As expected in this mechanism the rate equilibrium is retarded by addition of excess 37 SUMMARY AND CONCLUSION The reaction of the methylnickel complex 1 with alkynes occurs rapidly under mild conditions to give vinylnickel complexes, the product of 1,2-addition to the alkyne, in high yield. A wide variety of both internal and terminal alkynes can be employed. This reaction is bimolecular, first order in both 1 and alkyne, and gives highly regioselective products. Only the regioisomer resulting from migration of the methyl ligand to the least hindered alkyne carbon is observed. Thus this system constitutes a general example of the 1,2-addition of a transition metal alkyl to alkynes. The formation of both cis and trans addition products in kinetically controlled pathways demonstrates that 1,2-addition reactions can give products which have stereochemistry that differs significantly from the stereochemistry of process itself. To account initially concerted cis methyl bond gives a for this we have the addition suggested the insertion of the alkyne into the nickel coordinatively unsaturated vinylnickel intermediate capable of isomerization of the carbon-carbon double bond at a rate competitive with product formation. Kinetic competition between the rates of isomerization and of formation of cis and trans products then leads to product stereochemistry that may be independent of the stereochemistry of addition. The observed stereochemistry is also independent of the relative thermodynamic stabilities of the cis and trans addition products, and is not affected by changes in phosphine concentration. 38 Mechanisms of this type help explain how different alkynes can give opposite stereochemistry in reactions with the same metal alkyl, even under kinetic conditions, without postulating different addition pathways in each case. 5 , 6 In this study the observation of products stereochemically distinct from the pathway for addition was dependent upon phosphine catalysis for the isomerization of an intermediate. Nevertheless, independent of following the mechanism of isomerization, the conclusion concerning stereochemical studies in organometallic addition reactions still pertains: the observation of a given ster eochem ical mode of addition, even when the observed complex is found to be the kinetic product of the reaction, does not necessarily mean that the crucial insertion step proceeds with that same stereochemistry. 39 EXPERIMENTAL General. All manipulations involving organonickel or aluminum compounds were done under either nitrogen or argon using standard Schlenk techniques or in a vacuum/Atmospheres Corporation model HE-553 inert atmosphere glove box with a model M0-40 recirculating circulating nitrogen. purification system, and continuously All organonickel compounds were stored in the glove box. All solvents were thoroughly dried and degassed prior to use. Tetrahydrofuran (THF) and diethyl ether were vacuum distilled directly from sodium benzophenone ketyl solution. Benzene, toluene, and hexanes were added to a solution of sodium benzophenone ketyl preformed in tetraglyme, distilled. then vacuum Additionally hexane was also extensively washed with sulfuric acid, potassium permanganate Clo% H2 so 4 solution), and water to remove olefins, prior to ketyl formation. Proton nuclear magnetic resonance recorded on either a Varian EM-390, (NMR) spectra were or a 180 MHz FT-NMR instrument equipped with a Bruecker superconducting magnet and a Nicolet Instrument Corp. model 1180 data system, and electrons assembled by Mr. Rudi Nunlist (U.C. Berkeley). 31 P spectra were obtained at 72.9016 MHz on the later instrument. All 1 H NMR spectra are reported in ppm downfield from tetramethylsilane, and 31 P spectra are reported in 8ppm downfield from external PPh (3 5.9 ppm vs trimethylphosphite in benzene). The following = singlet, d = doublet, dd = doublet or doublets, t = triplet, q = abbreviations are used for the observed peak multiplicities: s 40 quartet, m = complex multiplet. Most chemical shifts were determined using the residual proton absorption of benzene-d 6 absorption at 87.15 ppm, THF-d 8 at 83.58 ppm, or added cp 2 Fe 3.96 ppm. Infrared spectra were recorded on a Perkin-Elmer 237 Grating Spectrophotometer. Gas chromatography (GC) was performed on Hewlett-Packard 5750 and Varian 90-P chromatograph, with helium as the carrier gas. Two columns were employed: A) 10' x ,, 1/8" 20% SE-30 on 100/120 chromasorb P, B) 2' x 1/4" 5% SE-30 on 60-80 chromasorb w. All peak areas were determined by electronic integration using a Spectra-Physics Autolab System I integrator. Ni(acac) 2 was prepared by drying commercially available Ni(acac> 2 <H 2 o> phosphine Cloo 0 c, (Alfa) and 24 hrs. at full vacuum). diphenylacetylene Triphenyl- (Aldrich) were recrystallized from hexanes. Liquid alkynes were purified by distillation or preparative Preparation of gas chromatography where necessary. Ni (acac) (PPh 3 ) CH 3 (1). Compound 1 was prepared by analogy to the method of Cotton, et,al, 12 for the synthesis of Ni(acac) (PPh 3 )cH 2 cH 3 . A solution of 0.68 g (7.8 mmol) Al(CH 3 >2 ocH 3 22 in 7 mL hexanes was added dropwise over 20 minutes to a ice/salt bath cooled (-15°C), stirred suspension of 4.0 g (15.6 mmol) Ni(acac) 2 and 4.5 g (17 mmol) PPh 3 in 50 mL diethyl ether under nitrogen. The solution was allowed to warm to about 0°C for about 2 hours. Then the solution was cooled to -15° C again and the yellow precipitate was collected by f il tra tion. The crude product was washed three times with cold ether, residual solvent was removed at reduced pressure, and taken into 41 the drybox. This gave 5.75 gms crude product containing considerable excess PPh 3 as evidenced by NMR integration. This crude product was purified by repeated recrystallization from toluene-hexane (1/10) or toluene-acetonitrile (1/4) to give 3.5 g (51% yield) of pure I. Compound 1 does not sublime; i t has m.p. (sealed tube) 150-2°C dee.; 1 H NMR (benzene-d 6 ); 8 0.07 Cs, 3H), 1.40, 1.90 (s, 3H each), 5.28 Cs, lH), 6.9-7.1, 7.6-8.0 (complex, 15H); IR(CH 2 cl 2 ): 3300 (br.), 1580 Cs), 1520 (s), 1440 (w), 1400 (s), 1190 Cw), 1100 Cm), 1020 Cm), 860 (w), . 830 Cs) cm- 1 ; Anal. Calcd. for c 24 H25 o 2 NiP: C, 66.24; H, 5.79; Ni, 13.49%. Found: C, 65.92; H, 5.85; Ni, 13.65%. Preparation of Ni Cacac) CPPb 3 ) co 3 CI-d 3 ). Compound l-d 3 was prepared as · des er ibed above from 2.0 g (7 .8 mmol) g (7.8 mmol) PPh 3 , Al {CD 3 ) 3 (Et 2 o) 2.05 and Al(CD 3 ) 2 ocH 3 prepared in situ from (.65 g, 3.9 mmol) and CH 3 0H Cl 57 20 mL hexane-ether Ni (acac) 2 , (1/3). 1, 3.9 mmol) in Three recrystallizations from toluene- hexane yielded 0.75 g (22% Theory) of pure l-d 3 . Anal. Calcd. for C24D3H2202NiP: c, 65.79; H + D, 6.44%. Found: c, 66.20, H + D, 6.66%. Preparation of Ni(acac)(PCy 3 )cH 3 (6). 16 Compound 6 was prepared in an analogous manner from Ni {acac) 2 (7 .8 mmol), PCy 3 (7.8 mmol), and Al (CH 3 ) 2 ocH 3 (3.9 mmol), to yield 1.32 g product. The crude product was purified by crude repeated recrystallization from ether, yield 0.73 gms {21% theory) dark brown cubic crystals. 1 H NMR (benzene-d 6 ): 8-0.18 (d, J= 5 Hz, 3H), 1.68, 1.85 {s, (s), l.0-1.4, 1.4-2.2 (m), 5.28 ppm lH). 42 Clit. 16 1 H NMR: 8-0.32 Cd, J = 5 Hz), 0.75-1.75 (m), 1.69, 1.84 Cs), 5.28 ppm Cs)). Preparation of Ni(acac)(PPh 3 )Ph (5). Attempts to prepare 5 by the method of Maruyama, et.al, 8 gave very poor yields of product and large amounts of Ni(PPh 3 )n (n= 3, 4) as impurity. Results were highly variable, sometimes giving no product at all. The following modified procedure enabled the isolation of 5 sufficient for these experiments. A solution of [Al(Ph) 3 (Et 2 o>l (0.35 g, 1.3 mmol) in 25 mL toluene-ether (1/4) was added very slowly to a slurry of 2.0 g (7.8 mmol) Ni(acac) 2 and 2.15 g (8.2 mmol) PPh 3 in 35 mL diethyl ether with cooling to -78°C. Upon warming to -20°c a yellow color developed and 50 mL cold hexane was added. The solution was cooled to -78°C and filtered. Only a small amount of a yellowgreen solid was collected. The filtrate was allowed to warm to room temperature, turning yellow-brown (no precipitate). The solvent was removed to yield a red residue. Extensive washing of this crude product with ether at room temperature gave 0.55 g (14% Theory) of essentially pure 5 as a yellow powder. Compound 5 could be recrystallization from toluene-ether. 1 H NMR (benzene- d 6 >; 8 l.40, 1.72 Cs, 3H each), 5.30 Cs, lH), 6.8-7.1, 7.4-7.6 ppm (m). Clit. 8 1 H NMR: 1.50 (s,3H), 1.72 (s,3H), 5.30 (s,lH), 6.8- 7.1, 7.4-7.8 ppm (m)) The reaction of 1 with alkynes. General: The reaction of 1 with alkynes was carried out in either THF or benzene under air free conditions by one of two methods. Method A: In the glove box 43 a solution of 1 was prepared (typically around 0.1 Min 1) an equimolar amount of the alkyne was added. and The resulting solution was transferred to an NMR tube, capped, and the cap wrapped with parafilm. The reaction was then monitored by NMR over a period of hours or days depending upon the alkyne. Method ll: A solution of 1 prepared in the glove box was transferred to an NMR tube and capped with a rubber septum. Outside the drybox the al kyne was then added through the septum via syringe, and the reaction was monitored by NMR. All reactions involving heating of the reaction solution were carried out in sealed NMR tubes. Typically the solutions changed from a light yellow-brown color to red as the reaction progressed. The 1 H NMR spectra of the resulting vinylnickel complexes are summarized in Table 4. With the The exception of 3 these complexes were not isolated. structures of these complexes was deduced from their NMR spectra and through identification of the organic product resulting from cleavage of the nickel-carbon bond with acid or LiAlH 4 (Table 1). Reaction of 1 with PhC:CPh. Protonation. A solution of 205 mg (1.15 mmol) PhC:CPh in 1 mL toluene was added to 500 mg (1.15 mmol) 1 in 20 mL toluene in the glove box. After stirring at room temperature overnight the solvent was reduced by evaporation to 8-10 mL and diluted with 200 mL hexanes. An orange precipitate (280 mg) was collected. A second crop was realized by reducing the filtrate to yield a total of 450 mg (89% theory) of a mixture of 3 and 4. In a similar experiment addition of excess PhC:CPh had no effect upon the reaction, other than to accelerate it; no 44 evidence of multiple inserti o n could be detected by NMR. In a separate experiment a solution of 3 prepared in THF from 75 mg (0.17 mmol) 1 and 31 mg (0.17 mmol) PhCECPh was treated with 0.4 mL (0.4 mmol) lM p-toluenesulfonic acid in THF. Analysis of the resulting solution by gas chromatography identified E- and Z-1,2-diphenylpropenes in 104% calculated yield and a ratio of Z/E = 5/4, using naphthalene as an internal standard and molar response factors determined using an authentic sample. However experiment. this Z/E ratio was variable experiment to The 1,2-diphenylpropenes were identified by coinjection of an authentic mixture, 23 and by mass spectral analysis on a sample purified by preparative gas chromatography (column B; ll0-130°C, flow rate 60 ml/min). Reaction of 1 with PbCECPb and added PPh 3 . A solution of 235 mg (0.54 mmol) 1, 78 mg C0.298 mmol) PPh 3 , and 50 mg (0.269 mmol) cp 2 Fe (internal standard) in exactly 6 mL benzene-d 6 was prepared in the glove box. Then 96 mg (0.54 mmol) PhC:::CPh was dissolved in the solution and 1 ml aliquots were added to weighed amounts of PPh 3 . The resulting solutions were then transferred to NMR tubes, frozen and sealed off under a vacuum. Six NMR tubes were prepared in this manner containing the concentrations listed in Table 5. These NMR tubes were heated in a temperature controlled water bath at 40 <± 1)°C, and the concentration of 1 was determined by cooling the NMR tube in ice/water and integrating the nickel-methyl resonance vs cp 2 Fe. The second I order rate constants determined in this way are presented in 45 Table 4. 1 H NMR Data for vinylnickel complexes of general Chemical Shiftsa Complex: vinyl ligand =C(<CH) Rl R2 acac 2.03 6.9-7.6 Ph>=<Me (d, 1) J= (m) 3: Ph 3.37 6.9-7.6 Ph>=<Ph (d, J= 1) (m) 4: N Me 6.9-7.6 (m) 1. 78,1.18 6.9-7.6 (m) 1.87,1.41 (s,3H) 5.20 ( s, lH) 3 (s,3H) 5.09 (s,lH) ,v 7: ,-,J 1.75 1.88,1.30 (s, 3H) (d, J= 1) 5.22 ( s, lH) Ph>=<H 5.05 t-Bu>=<H 4.65 2.80 6.9-7 . 2 (dd, J= 1,7.6-7.9 (m) Me 7) 9: ,v 10: """- II: 2 . 95 6.9-7.1 (d, 1) J= 7.6-7.9 PhK: (m) 2.75 1.50 (dd,J= 1, (s) Me 6) t-Bux,e 3.12 rv J.s: (q, J=6) 1.80,1.30 (s,3H) 5.25 (s, lH) (d,J= 1.5 1.58 (s) 1.85 1.87,1.38 (s,3H) (d, J= 1) 5.28 (s,lH) 2.72 3.32 3.56 (s) Me EL;£ 1.83,1.38 (s,3H) ( s, lH) (q, J= 1) 5.22 /-\Me <s) (s) 1. 7 8, 1. 2 2 ( s, 3H) 5.15 (s,lH) I-' OL..J ~ I-' 0 I I 50 /'V Figure 6. Plot of 1/[lJ - 10 20 r-, 30 '-.... I-' L.J 40 /";J '-.... 50 60 70 N 1/[1] O 150 3-' vs time at 40°C, [PPh 7= 0.05 M. 100 Time (min) / ., 200 I I ~ O"I 47 Table 5. Rate constants for the reaction of ,..,, 1 with Phc=cPh and added PPh ["})b 3 at 40 ~ 1°C. a [PPh ] 3 k (obs) 0.0945 M 0.05 M 6.2 0.945 M 0.71 M 3.63 X 10- 3 0.855 M 0.107 M 2.98 X 10- 3 0.085 M 0.197 M 1.70 X 10- 3 0.080 M 0.43 M 7.33 X 10- 4 0.070 M 1.05 M 3.0 X 10 arate constant from plot of 1/[1] ,.., [PhC=CPh]. L·M -1 sec -1 10 -3 X -4 vs time. b[l] ,., = 48 Table 5. A sample second order plot is presented in Figure 6N The plot of k(obs) vs 1/[PPh 3 Jadded is shown in Figure 4. Reaction of 1 with ArC=CPh. Solutions of 1 ( 0.05 mmol) and equimolar amounts of ArC'SCPh (Ar= p-CH 3 o-c 6 H4 -, p-Cl-C 6 H4 -, p-Me-c 6 H4 -> in 0.5-.7 mL benzene-d 6 were prepared in the glove box, transferred to NMR tubes and capped. The reactions all proceed at room temperature to give the 1 H NMR spectra summarized here. These spectra are consistent with the presence of both possible regioisomers of the p-substituted analogs of compound 3. A.The product of 1 lH NMR: 3H), and p-Me0C6H4C'SCPh. 8 1.23 Cs, 3H), 1.85 Cs, 3H), 2.10 (two d, J = 1.5 Hz, 3.42, 3.50 ( s, 3H, ratio 1.45), 5.17 ( s, lH), 7.0-7.8 ( m) , 8.6 5 Cd, J = 9 Hz), 8. 7 5 ppm Cdd, J = 1, 7 Hz). B. The product of 1 and p-ClC6H4C::CPh. la NMR; 8 1.27 Cd, J < 1 Hz, 3H), 1.85 (d, J = 1.5 Hz, 3H), 2.03 (two d, J = 1.5 Hz, 3H), 5.12 Cs, lH), 6.9-7.6 Cm), 8.48 Cd, J = 9 Hz), 8.67 ppm Cd, J = 6 Hz). C. The product of 1 and p-MeC 6 H4 CECPh. 1 a NMR: 81.20 (s, 3H), 1.78 (s, 3H), 2.05 (d, J = 1 Hz, 3H), 2.12 Cs), 2.25 Cs), 5.05 Cs, lH), 6.7-7.8 Cm), 8.60 Cd, J = 6 Hz), 8.78 ppm (dd, J = 1,6 Hz). Reaction of 1 with Phc=ccn3 and l-d3 with PhC::CCB3. A solution of 40 mg {0.09 mmol) 1 and 11 mg {0.09 mmol) PhCECCH3 in exactly 1 mL benzene-d 6 was prepared in the glove box, transferred to an NMR tube, and capped. Monitoring by NMR showed the reaction was 70% complete in less than 30 minutes, and 49 complete in 1.25 hours at room temperature. A ratio of the ci~/~£.9.Il~ isomers could be calculated by comparing integrations for the acac-H at 8 5.22 ppm and the cis-vinyl-CH 3 at 82.95 ppm and correcting for the theoretical number the of protons. Integration of the trans-vinyl-CH 3 at 81.75 ppm was very unreliable due to the proximity of the acac-CH 3 absorption at 1.88 ppm. In this way the intial ratio was calculated to be cisCH3 ; ~£.9.Il~ - CH 3 = 1 . 7 O ( + . 1 ) . A1 t e r n a t i v e 1 y , i n a s e pa r a t e experiment in protiobenzene integration of the deuterium NMR (180 MHz FT NMR) resonances for the vinyl-co 3 at 82.95 and 1.75 ppm directly gave a ratio of trans-co 3 /.Q.ll-CD 3 = 1.87; the later experiment was considered more accurate. A similar reaction involving 68 mg (0.15 mmol) l-d 3 and 18 mg (0.15 mmol) PhC:CCH 3 in 1 ml benzene gave a ratio of .Q_llco 3/trans-co 3 = 1.56 by deuterium NMR. These product ratios were found to be almost totally insensitive to added PPh 3 or solvent. In the reaction of l-d 3 with PhC:CCH 3 in benzene-a 6 the ~/trans product ratio was 1.55 (+ .1) with no added PPh 3 , 1.67 <± .1) with 0.1 Madded PPh 3 , and 1. 8 5 ( + .1) w i th 1. 0 M added PP h 3 (both at 6 0° C) . In the r ea ct ion 0f 1 with p h C:C CD 3 the Ci s /trans pr O du Ct rat i O was 1. 6 ( + .1) in THF-d 8 , and 1.7 (+ 0.1) in 0.2 M NaBPh 4 in THF-d 8 . The Kinetic Deuterium Isotope Effect. A solution of 9 mg (0.05 mmol) PhCECPh was added to a solution of 102 mg (0.23 mmol) 1 and 114 mg (0.26 mmo : ) l-d 3 in 4.5 ml THF. After 2 hours a solution of 430 mg (2.2 mmol) p-toluenesulfonic acid in 2 ml THF so was added. The solvent was removed and the residue was extracted with H20/ether in air. The combined ether fractions were dried, then evaporated to dryness, to yield a white solid. About 1-2 mg of E-and Z-1,2-diphenylpropenes were isolated from this mixture by preparative gas chromatography (column B; 130°C, 80 ml/min). Mass spectral analysis gave a d 0 /d 3 ratio of 1.05 and 1.085 for the z-and E-1,2-diphenylpropenes, respectively. Similarly the 180 MHz NMR spectrum of Z-1,2-diphenylpropene gave a d 0 /d 3 ratio of 1.16. Averaging a p d correcting for the star ting concentrations gives a value for k(l)/k(l-d 3 ) = 1.24 X-Ray Crystal Structure of 3. <± .OS) A crystal of 3 measuring 0.36 x 0.24 x 0.28 mm obtained from a toluene-hexane solution was shown to be exclusively 3, not 4 , by examination of the NMR spectrum of the remaining crystals in the same batch. Compound 3 crystallizes in space group Pl with cell parameters: a= 17.892, b= 12.339, c= 16.732 A, 0= 106.27,/3= 73.17, y= 110.77°,z= 4. The calculated density is 1.26 gm/cc, and a density of 1.19 gm/cc was measured by flotation of a crystal from the same batch. Intensity data on 5258 reflections 28 ~ 38 were collected on a Syntex P2 1 diffractometer with monochromatic Mo Ka scanning. By monitoring ten radiation using 0-20 standard reflections it was determined that no significant crystal decomposition occurred. The locations of the two nickel and two phosphorus atoms were obtained by examination of a Patterson map. Fourier-transform electron density maps then identified the positions of the other 80 non-hydrogen atoms. The nickel and phosphorus atom thermal 51 parameters were refined anisotropically, and the other 80 nonhydrogen atoms only isotropically. A difference Fourier map identified no sites of electron density greater than 1.0 e-/A 2 0 except within 2.0 A of the nickel atoms. Finally H atom positions 0 were calculated assuming a carbon hydrogen bond. length of 0.95 A and a B value one greater than for the attached carbon atom. Final refinement after a total of four least-squares routines gave R= 0.082 and a 'goodness of f i t ' = 1.54 for all 5258 reflections (28 ~ 38°), and R= 0.053 for the 3198 reflections [F 0 > 3cr(F 0 ) l. An 0RTEP drawing of 3 is shown in Figures 5. More crystallographic data are given in Appendix A. Competition Experiment Between PhC:CPh and PhCsCCH 3 . solution of 20 mg (0.181 mmol) PhCE'CCH 3 , 30 mg A (0.168 mmol) PhC:CPh and 21 mg (.046 mmol) 6 in 1 mL benzene-d 6 was prepared in the glove box, transferred to an NMR tube, and sealed off under vacuum. After heating to 44°c for 2.5 ha ratio of Ni (acac) (PCy 3 ) [ C (Ph) =C (Ph) CH 3 l to Ni (acac) (PCy 3 ) [ C (Ph) =C (CH 3 ) 2 1 of 1.1 of <± .1) was determined by integration of the NMR spectrum the resulting solution, at this point 40% of 6 had been consumed. Correcting for the initial concentrations gives a ratio for k(PhCECPh)/k(PhCsCCH 3 ) = 1.2 (+ .1). I n a s i m i 1 a r man n e r a m i x tu r e o f 3 0 mg ( 0 .16 8 mm o 1) Ph CEC Ph , 21 mg (0.048 mmol) 1, and 13 mg 21 mg (0.181 mmol) PhCsCCH 3 , (0.049 mmol) PPh 3 (to slow the reaction at room temperature) 1.0 mL benzene-d 6 gave a value for the ratio in of k(PhC:CPh)/k(PhC:CCH 3 ) = 1.40 (+ .05) after heating to 45°C for 52 100 minutes. Another experiment involving no added PPh 3 and conducted at room temperature resulted in an identical ratio within experimental error. Reactions of the Vinylnickel Complexes. General. The vinylnickel complexes Ni(acac) (PPh 3 ) [C(Ph)=C(Ph)CH 3 ] Ni(acac)(PPh 3 )[C(Ph)=CCH)CH 3 l (8) were prepared in situ, (3) and usually in THF solvent, from equimolar amounts of 1 and either PhCECPh or PhC:CH. After about 1 hour at room temperture these solutions were treated with the appropriate reagent under air free conditions. The resulting organic products were analyzed by gas chromatogrphy and/or NMR, and identified by comparison with authentic samples in most cases. These results are summarized in Table 2. Below are descriptions of the reactions of 3 with some of these reagents; the reactions of 8 were carried out in a similar manner. LiA1B 4 . A solution of 12 mg (0.32 mmol) LiAlH 4 in 10 mL THF was added to a solution of 0.2 mmol 3, prepared as above, in 3 mL THF with 15 mg biphenyl as internal standard. At various time intervals 1 mL aliquots were removed, quenched with 1 mL saturated Na 2 so 4 and analyzed by gas chromatography (column A, 195°C, 60 ml/min). After 15 minutes a maximum yield of 85% of Eand Z-l,2-diphenylpropenes 23 was realized. However the E/Z ratio changed over the course of the reaction from an initial value of 1.1 to 1.8. Al(CB 3 ) 3 . A 1 mL solution of Al(CH 3 ) 3 (25% in hexane) was added dropwise to a solution of 0.1 mmol 3 in 2 mL benzene. 53 After 5 min, 4 hand 24 h, 1 mL aliquots were quenched with 2 mL saturated Na 2 so 4. After 4 ha maximum yield of 88% E- and Z-2,3diphenyl-2-butenes 24 were identified by gas chromatography. CH 3 Li. To a solution of 0.047 mmol 3 in 1 mL of ether was added dropwise 0.5 mL of 1.45 M CH 3 Li at room temperature. Extraction with H2 0/ether allowed isolation of a 52% yield of E- 2,3-diphenyl-2-butene 24 , identified by its mass spectrum and NMR. 12. A solution of 150 mg (0.59 mmol) 1 2 and 50µ.L pyridine in 12 mL CH 2 c1 2 was added to a solution of 0.57 mmol 3 in 5 mL THF. After stirring overnight addition of 230 µ.L CH 3 I (to precipitate PPh 3 ) and filtration through a plug of silica gel yielded 168 mg of a mixture of E-1,2-diphenylpropene, and E-and Z-l-iodo-1,2-diphenyl and propenes, in yields of 24 66% respectively. CO/MeOB. A solution of 520 mg (0.5 ml) 25% methanolic Bu 4 N+OH- in 33 mL benzene was added to a solution of 0.051 mmol 3 in 2 mL benzene under carbon monoxide purge. After 30 min CO flow was discontinued, and then after 1 h the solution was extracted with 10% HCl/ether. The combined ether layers were dried with Mgso 4 then evaporated to yield a mixture of E-and Z-methyl-2,3dipenyl-2-butenoate, in 39% yield, identified by NMR and mass spec. of a sample purified by preparative GC. Preparation of Al(CD 3 ) 3 .Et 2 o. Al(CD 3 ) 3 CEt 2 o> was prepared by the method of Krause and Wendt 25 for the corresponding protio compound. A solution of CD 3 MgI, prepared from 6.9 mL (0.108 mol) co 3 1 (99+% D, Aldrich) and 3.15 g (0.13 mol) Mg turnings in 40 mL 54 ether, was added to a solution of 4.0 g (0.03 moL freshly sublimed A1Cl 3 in 25 mL ether (the ether should be added to the AlC1 3 with cooling to control the strongly exothermic formation of the etherate) over 10 minutes with cooling from an ice bath, all under nitrogen. reflux for 5.5 h. The resulting solution was brought to a The reflux condenser was replaced . with a distillation head, condenser and receiver flask under nitrogen flow and the ether was distilled out of the flask at atmospheric pressure. The oily residue was heated with an oil bath to around 200 Cat reduced pressure, and Al(cn 3 >3 CEt 2 o> distills over as a clear liquid: b.p. 50-65°C at aspirator pressure (~15 mm Hg). Clit. 25 b.p. 68°C at 15 mm Hg) Yield 3.02 gms (60% theory>. Virtually no Al-CH 3 can be observed by NMR just the coordinated ether. This preparation was repeated with protio methyliodide to give Al(CH 3 ) 3 (Et 2 0). 1 H NMR:8 -0.42 Cs), 0.76 Ct, J = 6.5 Hz), 3.35 ppm (q, J = 6.5 Hz). Preparation of A1Ph 3 . A1Ph 3 was prepared by the method of Eisch and Kaska. 26 Under nitrogen a solution of 30.9 g C0.087 moles) HgPh 2 in 150 mL xylenes was brought to reflux in a three necked flask containing 11.1 gms C0.41 mmol) Al turnings, and equipped with a second three necked flask attached through a side arm. After heating at reflux for 4 days the solution was poured hot through the side arm into a the second flask containing 5 grns Al turnings, flask. leaving behind the Al/Hg solids in the original The solution was heated at reflux for an add .itional 2 days, then decanted hot into a third flask. Roughly two thirds of 55 the solvent was distilled out of the flask, and the solution was cooled to -30°C to give the white crystaline product. Yield 8.86 gms A1Ph 3 (59% theory); m.p. 235-42°C (lit. 26 229-242°C). The product could be recrystallized as the etherate from tolueneether. AlPh3(Et20): m.p. 127-9°C (lit. 26 129-129.5°c). Preparation of PhC;.cco 3 . 27 A 22 mL solution of BuLi (0.053 mol, 2.4 M) in hexane was added dropwise over 20 minutes to a s o 1 u ti o n of 5 . 5 mL Co. o 5 mo 1) f r e s h 1 y d i s t i 11 e d Ph csc H i n 4 o mL anhydrous ether with cooling to -20 ° C under nitrogen. The solution was allowed to warm to room temperature and 25 mL dry THF was added, followed by 3.2 mL (0.05 moles) co 3 r (Aldrich 99+% D) over 10 minutes. The reaction was exothermic, warming the solution considerably. The addition funnel was washed down with an additional 10 mL THF. After 17 hours the solution was poured over 100 ml ice/water, then extracted with ether. The combined ether extracts were dried with Mgso 4 and distilled at reduced pressure to yield 2.40 g (40% theory) PhC=CCD 3 ; b.p. 73-75°C (17 mm Hg). By ms the product was determined to be> 98 atom% D. Anal. Calcd. for C9 H5 o 3 : C, 90. 70; H + D, 9.30. Found: C, 90.23; H + D, 9.52. The substituted diphenylacetylenes p-X-PhC= CPh were prepared from PhC = CCu and the appropriate p-iodobenzene by the method of Castro and Stephens. 28 The iodobenzenes were all obtained from Pfaltz and Bauer and used without further purification. In all cases the resulting acetylene was recrystallized from methanol or hexanes until no 56 change in melting point was observed (at least twice). p-ClC 6 H 4 C=CPh : m.p. 80.5-81.5°C (lit. 29 81.5-82°C). IR(CC1 4 ): 2225 Cw), 1605 (sh), 1595 (s), 1495 (s}, 1445 Cs), 1400 (s}, 1090 Cs}, 1015 Cs), 825 (s), 685 Cs) cm- 1 • p-CH3C6H4CECPh: m.p. 70 IR(CC1 4 >: 2215 71°C. (lit. 3 0 m.p. 72-74°C) Cw}, 1600 (s}, 1515 (s), 1487 (s), 1445 (s), 685 (s} cm- 1 • 1 H NMR (C6D6): 81 _. 99 (m, 4H). p-MeOC 6 H 4 CsCPh: (s, 3H), 6.97 (m, SH}, 7.50 ppm m.p. 57-61°C Clit. 3 0, 3 l m.p. 58-60°C). IR(CC1 4 }: 2210 Cw), 1500 Cs), 1435 Cm}, 1242 Cs) ,1168 Cm}, 1030 (s}, 825 (m, 9H}. (s) cm- 1 • 1 H NMR (C6D6}: 8 3.17 (s, 3H), 6.5-7 .4 ppm 57 REFERENCES 1. For general discussions of see: a) P.W. G. Jolly, Ziegler-Nat ta Wilke, Polymerization "The Organic Chemistry of Nickel", Vol I I , Academic Press, New York, 1975, Chapter 1. b) M.M. Taqui Khan, A.E. Martell, "Homogeneous Catalysis by Metal Complexes", Vol I I , Academic Press, New York, 1974, Chapter 6. c) P. ~ Cossee, Arlman, .iL.. Catal., 1964, ~g~~L.., l, 1964, 80 ; E.J. l, 89; E.J. Arlman, P. Cossee, ~ Cat al., 1964, l, 99. 2• a) H. C. C1 a r k , K. E. Hine , .iL.. Organ o mtl ~.ID..u 1 9 7 6 , ~ , C 3 2. b) C. Rice, J.D. Oliver, 1978, ibid., Booth, A.D. Lloyd, ibid, 1972, .l.J2, 121. c) B.L. 12., 195. d) D.A. Harbourne, F.G.A. Stone, .iL.. .c.h~.m..... ~ , 1968, 1765. 3. a) H.G. Booth, Ault, R.G. .L.. Naturforsch., 1977, ~, 1139. b) ~ J.Al., 1970, 308. Hargreaves, .iL.. Che.[!k. B.L. c) J. L. Dav ids on, etal, .iL.. ~1:!k. ~ ~1:!k. .C..Qml:!k., 197 6, 714. d) R.F. Heck, .iL.. Al:!k. ~1:!k. ~ , 1964, 1l2_. e) P.L. Watson, R.G. Bergman, .iL.. Al:!k. ~1:!k. ~ , 1979, l.Q.l, 2055. 4. a) M.H. Chisholm, H.C. Clark, AQs;;_,_ ~1:!k. ~ , 1973, 202. b) H. C. Clark, R.J. Puddephatt, Iriorg. ~1:!k., 1971, lQ, 18. c) T.G. Appleton, M.H. Chisholm, H.G. Clark, .iL.. Al:!k. ~he.[!k. ~ , 1972, .2..!, 8912. d) H.C. Clark, K. von Werner, .iL.. Organom~ ~1:!k., 1975, 101, lll.Q.r.9..L ,C,hg.[!k., 1972, 347. e) H.C. Clark, R.H. Puddephatt, ll, 1269. f) H. C. Clark, G.R.C. Milne, c.s. Wong,~ Q~ggilQ!!lgh ~hgm~, 1977, llQ, Michman, M. Balog, ibid,, 1971, 265. g) M. ll, 395. h) M. Michman, L. 58 Marcus, ibid., 1976, ill, 77. i) M.D. Rausch, W.H. Boon, i.b. i .d.t.., 1 9 7 7 , .l.i.l , 2 9 9 . j ) H. Ma s a i , K. S o n a g a h i r a , N. Hagihara, Bull. ~!!h ~ Japan, 1968, 41, 750. k) W. H. Boon, M.D. Rausch, iL.. ~!!h ~ ~!!h ,C.Q.mm...., 1977, 397. 1) H. Hosseive, J. Nixon, J. Poland, .J.... Organom~~!!h, 1979, .l.ll, 107. m) Rudler-Chauvin, Rudler, ibid., 1977, UJ., 115. n) S.J. Tremont, R.G. Bergman, ibid •. 1977, il.Q., Cl2. 5. s. Otsuka, A. Nakamura, a9.Y.L Organom~ ~ID.-.., 1976, 245. 6. a) M. Michman, X.I._g.Il.§....., s. Weksler-Nussbaum, iL.. ~!!h ~ Perkin 1 9 7 8 , 2. , 8 7 2 . b ) J . J . E i s c h , R. J . Ma n f r e , D. H. Konar, iL.. Organom.e.t_._ ~!!h, 1978, 1.2.2., Cl3. c) B. B. Snider, M. Karras, R. s. E. Conn, iL.. A!!h ,ChgJ!h ~ , 1978, .l.0..0, 4624. 7. E. R. Evitt, R.G. Bergman, iL.. A!!h ,ChgJ!h ~ , 1979, l.Q.l, 3973. 8. K. Maruyama, T. Ito, A. Yamamoto, Ji. Organomet, Chem,. 1978, 155, 359; ibid., 1975, .2.Q, C28. 9. K.L. Ivin, J.J. Rooney, C.D . Steward, M.L.H. Green, R. Mah tab, iL.. ,ChgJ!h ~ ~!!h -CQ.m!!h, 1978, 604. 10. This work has been reported in part in a preliminary communication; J.M. Huggins, R. G. Bergman, iL.. A!!h ~.m.... ~ , 1979, l.Q.l, 4410. In order to minimize confusion, we have used E/Z nomenclature when referring to stereochemistry of compounds, and cis/trans nomenclature for referring to the stereochemistry of addition processes (i.e. cis addition can in principle give either and E or z product, depending 59 upon the substituents involved. see: IUPAC Commission on Nomenclature of Organic Chemistry, EJ.!il Appl. .cMI!h., 1976, fi, 11. 11. T. Yamamoto, T. Saruyama, Y. Nakamura, A. Yamamoto, Bull, .c.h,g_I!h. ~ Japan, 197 6, ti, 5 8 9; .iL.. AI!h. .c.h,g_I!h. ~ , 197 3, -2..2., 5073. 12. F.A. Cot ton, B.A. Frenz, D.L. Hunter, .iL.. AI!h. .cMI!h. ~ , 1974, il, 4820. 13. T. Saruyama, T. Yamamoto, A. Yamamoto, Bull, .cM.m.... ~ Japan, 1976, il, 546. 14. This broadening can be explained by an associative phosphine catalyzed exchange of the acac-CH 3 1 s. See ref. (11), (12), and (32). 15. One explanation for the observations of Yamamoto and coworkers (11) can account for their 31 P NMR spectra if the compoundthey assign as Ni(acac) (PPh 3 ) 2 cH 3 is really 1 with about 10% excess PPh 3 . This will give a singlet at 47 ppm at -8O°C, and a broad averaged signal at about 40 ppm at room temperature as observed. However if true there should be a small singlet at o ppm at -ao c, which they do not 0 report. 16. P. w. Jolly, K. Jonas, C. Kruger, Y-H. Tsay, .iL.. Organom..e..k Chem,, 1971, ll, 109. 17. In the reaction of 1 with 2 equivalents 2-butyne, after protonation, GC analysis identified 3 longer retention time products in circa 10% yield each. These products were not 60 identified. 18. In the reactions of 1 with acetylene and 1-pentyne polymerization of the alkyne appeared to take place. 19. H.C. Clark, C.R.C. Milne, C.S. Wong, .J..i. Organom~ .c.h.e.m....., 1977, 1-li, 265. 20. D.W. Hart, T.F. Blackburn, J. Schwartz, .J..i. Am..... ~m..... ~ , 1975, l l , 679. 21. We assume the small difference between the two product ratios is due to a secondary deuterium isotope effect of magnitude 1.8/1.6= 1.15. 22. T. Mole, .J..i. Aust, .cJ:Le..m.,_, 1963, l i , 794. 23. An authentic sample of E-and Z-1,2-diphenylpropenes was prepared from benzylphosphonium chloride, BuLi, and acetophenone by method of Wittig and Haag, Chem, .Iifil..., 1955, .fill, 654; m.p. 113-37°C; 1 H NMR (CDC1 3 ); 82.25 (d, J= 2Hz, 3H), 6.8 (d, J= 2 Hz, lH), 7.2-7.5 ppm Cm, l0H). 24. An authentic sample of E-and Z-2,3-diphenyl-2-butenes was obtained from E. Evitt. 1 H NMR (CDC1 3 ); 81.87 Cs, 3H), 2.10 Cs, 3H, Z-isomer), 7.0-7.5 ppm Cm). Clit. (CC1 4 ); 81.88 CEisomer), 2.16 ( Z-i somer); .J..i. .Q_(g_,_ ~.m.,_, 4 7 9 ( 197 6) .) 25. E. Krause, B. Wendt, Chem,~, 1923, ~ , 466. 26 . J. J. E i sch , W. C. Kaska , .J..i. A.m.,_ ~.m.,_ ~ , 19 6 6 , .fill, 2 9 7 6. 27. Adapted from method in "Preparative Acetylenic Chemistry", L. Bradsma, Elsevier Pub. Co., New York, 1971, pg 54-5. Addition of THF after formation of the anion was found to be essential. 61 c. E. Castro, .J.i. ~ Chem., 1963, 28, 3313. 28. R. D. Stephens, 29. M. S. Newman, D. E. Reid, 30. F. Scandiglia, J. D. Roberts, Tetrahedron, 1958, i, 197. 31. A. R. Katritzky, A. J. Boulton, D. J. Short,~ ~.m..... ~ , ibid., 1958, .21., 665. 1960, 1519. 32. T. Ito, T. Kirayama, A. Yamamoto, Bull. ~.m..... ~ Japan, 1976, il, 3250. 62 PART II THE MECHANISM OF ALDEHYDE FORMATION IN THE REACTION OF A MOLYBDENUM HYDRIDE WITH MOLYBDENUM ALKYLS. 63 INTRODUCTION In a recent communication Bergman and Jones 1 reported the clean bimolecular reaction of CpMo(C0) 3 H with CpMo(C0) 3 R in THF to give the aldehyde RCHO and molybdenum dimers under mild conditions (Scheme I). This reaction is significant both as an example of reductive elimination involving two metal centers, and as a model for the aldehyde forming step in the oxo process. Mechanistic studies on the cobalt catalyzed hydroformylation of alkenes by carbon monoxide and hydrogen, the oxo process, have firmly established that hydride cobalt tetracarbonyl is the active chain carrying intermediate in this system. 213 However despite evidence 2 B that the bimolecular reaction of HCo(C0) 4 with a cobalt acyl complex is the mechanism for aldehyde formation under stoichiometric conditions (25°C, 1 atm), researchers in this field have argued that direct reduction of a cobalt acyl by H 2 is responsible for aldehyde production urider reaction conditions. 2 This conclusion and most of the arguments in support - of it have been brought into considerable question by a recent re-examination of the equilibrium co 2 cco> 8 + 2 HCo(Co) 4 under high pressure and high temperature conditions . 3 The reaction of molybdenum hydride 1 with molybdenum alkyls 2 to give aldehydes, therefore, represents an important stoichiometric model system supporting the proposition that aldehyde formation occurs by the reaction of HCo(C0) 4 with cobalt acyls in the hydroformylation of alkenes. • la, Cp = 77 -C5H5 5 Cp Mo(C0) 3 H Scheme I + ► + + -½ [ CpMo(CO)2 ] 2 2c, R = CH 2Ph -½ [ CpMo(CO) 3 ] 2 THF 2b, R = C2H5 2a, R = CH 3 CpMo(C0) 3R 0 II RC-H 4 3 O'I ~ 65 The reaction of CpMo(C0) 3 H (1) with CpMo(C0) 3 R (2, R = CH 3 , C2H 5 , CH2Ph) in THF at between 25 and 50°C was observed to be bimolecular (first order quantitative yield of [CpMo(C0) 3 l 2 (3a) in each aldehyde reactant) and and [CpMo(C0) 2 l 2 the (4a) leading to a molybdenum dimers (Scheme I). Some of the observed second order rate constants are listed in Table 1. The rate of this reaction was found to be highly dependent upon the molybdenum alkyl involved, following the reactivity order c 2 H5 CH3 > > CH 2 Ph. In addition the reaction rate was found to exhibit a moderate solvent dependence, proceeding at least an order of magnitude faster in THF than in benzene, and no reaction was observed in hexane solution after 7 days at 7o 0 c. The reaction become first order (independent of the concentration of 1) in benzene. 4 The rate of reaction of 1 with 2b in THF was unaffected by 10 atmospheres of carbon monoxide pressure. These observations are analogous to some earlier work on the reaction of molybdenum alkyls with phosphines, only with the inverse solvent dependence. In 1967 Butler, Basolo, and Pearson 5 reported a kinetic study of the reaction of CpMo(C0) 3 cH 3 (2a) with phosphines and CpMo(C0) 2 (PR 3 ) [C(O)Rl phosphites (6). to give the acyl complexes They found that the reaction was first order (in 2a) and reasonably independent of the added ligand and its concentration in THF. However in toluene the reaction was found to be much slower, following second order kinetics . Similarly Craig and Green found that this same reaction was first order in acetonitrile and chloroform. 6 These results were 66 Table 1. Second Order Rate Constants for the Reaction of la with CpMo (CO) R in THF. 3 -1 -1 Complex R T (•c) 2a -CH 50 4.0 2b -cH cH 2 3 25 8.5 X 10- 4 2c -CH Ph 2 50 2.5 X 10- 5 3 k (M X sec 10- 3 ) 67 interpreted as being consistent with a mechanism involving a preequilibrium to give the unsaturated intermediate acyl complex CpMo(C0) 2 [C(O)RJ (5) followed by addition of phosphine (Scheme II). 516 The second order rate law observed in toluene was attributed to a slow direct bimolecular reaction pathway. The rate expression for the mechanism outlined in Scheme II has the form shown in eq (1). -d r 21 (1) dt Bergman and Jones 1 were able to show that reaction of 1 with phosphine complex CpMo(C0) 2 (PPh 3 ) [C(O)RJ (6b, R= c 2 H5 ) in THF does not proceed at temperatures near 50°C, but only at more elevated temperatures where loss of phosphine can occur. Under these conditions formation of aldehyde was accompanied by the observation of small amounts of 2b. Independent kinetic measurements have determined an approximate value for the ratio Bergman and Jones used crossover labeling studies to rule out an alternative mechanism involving homolysis of the metalacyl bond, to give acyl free radicals, followed by reaction with molybdenum hydride to give aldehyde. However, in the case of R = CH 2 Ph a minor competitive mechanism, involving homolysi s of the metal-alkyl bond leading to toluene, was observed. The reaction of 1 with 2 to give aldehydes is one of the few examples known for in which bond formation between alkyl and hydride ligands involving two metal centers occurs. The mechanism 2 CpMo(C0)3 R Scheme JI ◄ k _I kl ► II d~ co 1 oyR 7 I k_2 ► II 0 Products 6 CpMo(C0) 2( PR 3 ) CR -~1111 k2 [PR~ CpMo(C0) H, 3 5 -Mo-H-Mo '"' I co \\ ~ _O CO co CpMo-C-R I I co 0 O'I 00 69 of reductive elimination of alkanes and carbonyl compounds from mononuclear hydride alkyls and acyls is generally considered to be a facile intramolecular process. Yet isolable hydr ido alkyl, and dialkyl complexes have been observed to reductively eliminate alkane by both intra- and inter-molecular mechanisms. 7 , 8 At least four general mechanisms for the reductive elimination of R-H from hydride and alkyl ligands bound to two different metal centers have been identified: 1 a) alkyl migration to form an intermediate acyl complex, followed by reaction with metal hydride to give aldehyde (e.g., the reaction of 1 with 2); b) alkyl migration to form an intermediate metal acyl, followed by reaction with metal hydride to give alkane; 8 c) loss of carbonyl, followed by reaction with metal hydride to give alkane; 9 and d) metal-carbon bond homolysis to give an organic radical, 9 followed by abstraction of a hydrogen atom from the metal hydride. A fifth alternative, involving a concerted carbon hydrogen bond forming process from organic ligands bound to adjacent metal centers in binuclear and larger metal complexes, has been widely discussed.lo However no well documented examples of this "binuclear" reductive elimination mechanism exist. 11 Considering the paucity of well behaved stoichiometric examples of "bimolecular" reductive elimination reactions, and the continuing inter- vs intra-molecular controversy in mononuclear systems, the details of the mechanism of reductive elimination of aldehydes from the reaction of molybdenum hydride with molybdenum alkyls is of particular interest. Bergman and 70 Jones proposed that the bimolecular reaction of the unsaturated intermediate 5 with molybdenum hydride gives the hydride bridged , intermediate 7 containing a Mo-H-Mo three center,two electron bond. A number of stable complexes containing linear hydride bridges of this kind have been characterized. 12 However the mechanism through which this complex proceeds to give aldehyde was not established. It is this process, in which the intermediate 7 gives aldehyde, that the experiments presented here were designed to probe. 71 RESULTS The Reaction of CpMo (CO) 3 a and MeCpMo (CO)CB • As 3 described in the previous section the reaction of molybdenum hydride 1 and molybdenum alkyls 2 gives two metal containing products, the dimers [CpMo(C0) 3 J 2 (3), and [CpMo(C0) 2 J 2 (4), in roughly equal amounts. The fate of the Mo atoms can be followed by labeling one of the reactants with a methylcyclopentadienyl ligand (MeCp =~5CH 3c 5 H4 ). This can give six possible dimers 3a,b,c and 4a,b,c. CpMo(C0)3-(C0)3MoCp CpMo(C0)2=<co)2MoCp 3a 4a MeCpMo(C0)3-(C0)3MoCp MeCpMo(C0>2-<C0)2MoCp 3b 4b MeCpMo(C0)3-(C0)3MoMeCp MeCpMo(C0)2-<C0)2MoMeCp 4c 3c All six of t;Jiliese dimers are known and authentic mixtures can readily be prepared. The dimers 4a,b,c have a metal-metal triple bond, confirmed by an x-ray crystal structure on 4a; they exhibit a rich chemistry including the rapid addition of two molecules of carbon monoxide at room temperature to give the dimers 3a,b,c. 13 The dimers 3a,b,c can be photochemically 14 and thermally 13 converted to 4a,b,c provided free CO is removed from the system. Following the reaction of CpMo(CO) 3 H (la) with MeCpMo(C0) 3cH 3 (2b), MeCpMo(C0) 3H (lb) with CpMo(C0) 3cH 3 (2a), or lb with CpMo(C0) 3 Et (2b) in THF-d 8 at 50°C by 1 H NMR shows these reactions cleanly give a mixture of dimers containing the 72 resonances for the fragments CpMo (CO) 3 , and MeCpMo (CO) 2· However, CpMo (CO) 2 , MeCpMo (CO) , 3 distinguishing the two unsymmetrical dimers 3b and 4b from mixtures of the symmetrical dimers 3a + 3c and 4a + 4c is impossible by either NMR, IR or normal column chromatography. Whereas mass spectral analysis of mixtures of 3a,b,c and 4a,b,c has been achieved by others, 13 the inherent complexity of such analysis with the present mixture led us to seek another method of analysis. It was found that high pressure liquid chromatography (HPLC) using a reversed Ultrasphere-ODS, phase Altex Corp. claiming virtually complete coverage) with water/acetonitrile column (1:5) (1 x 25 cm, as eluant gives almost baseline separation of an authentic mixture of all dimers (see Figure 1). 15 Under these conditions the most polar dimer, 4a, elutes and the least polar dimer, 3c, elutes last. first Analysis of the products of the reaction of la with 2d, lb with 2a, or lb with 2b by HPLC revealed that the major metal containing products are the two unsymmetrical dimers 3b and 4b and only minor amounts of the other four dimers are formed (Figure 2). Thus two of the possible six dimers predominate in the reaction of molybdenum hydride and molybdenum alkyls when one of them contains a labeled cyclopentadienyl ligand, of the original location of the label. The independent analysis is complicated by the conversion of unreacted molybdenum hydride to the hexacarbonyl dimers in acetonitrile. Molybdenum hydride was shown to decompose in acetonitrile solution (the HPLC solvent) to 73 - 4b - 3b E c:: ~ '° (\J - -- w 4c 3c 3a u z 40 <X CD a: 0 Cl) CD <X 42 36 30 24 18 12 TIME (min) 6 0 Figure 1. HPLC chromatogram of a Mixture of 3a,b,c and 4a,b,c. f"v",/v,'\,., ,.__,rA,I\,., 74 - 4b - 3b E C: V l{) C\J w (.) z <l m a:: 0 (/') m <l - 2b 42 36 30 24 18 12 6 0 TIME (min) Figure 2. HPLC Chromatogram of the Products of the Reaction of lb and 2b. rv A./ 75 give the corresponding hexacarbonyl dimer (either 3a or 3c), in addition to some unidentified products. However, if all the molybdenum hydride has been consumed prior to the HPLC analysis, no spurious source of dimers hinders the analysis. The molybdenum alkyls and the dimers are stable to the analysis conditions. The two dimers concentration, 3b and 4b are formed independent of the in roughly equal original position of the methylcycopentadienyl "label" (i.e. lb or 2d), or the molybdenum alkyl involved. If, however, a solution of la and 2d in THF is heated to l00°C for 15 hours, a nearly thermodynamic mixture of all six dimers is formed. At this temperature conversion of the initially obtained dimers 3b and 4b to the thermodynamic mixture of all six dimers can occur by a mechanism involving the thermal cleavage of the Mo-Mo bond (Scheme 3). This series of reactions is well established to occur at temperatures above 80°c. 13 , 14 , 16 The Reaction of CpMoC 13co) 3 H with CpMoCco> 3ca 3 . hydride containing 93.8 atom % 13 co was Molybdenum prepared from [CpMo( 13 co) 3 J 2 via the standard method involving reduction, followed by protonation. 17 The labeled dimer was prepared by the repeated photochemical exchange 14 of carbonyl ligands under an atmosphere of 13 co. The carbonyl absorptions in the IR spectrum of the 13 co labeled molybdenum hydride shift from 2020 and 1948 to 1977 and 1890 cm- 1 , effect. consistent with the expected isotope Furthermore the 1 H NMR resonance for the hydride exists as a 1:3:1:3:1 quintet corresponding to coupling constants to the 76 trans and cis 13 co's of about 13 Hz. CpMo(C0) 3 H is believed to have a square-pyramidal structure. 17 Monitoring the reaction of 1 and 2 in THF by NMR shows that the primary product of the reaction is the triply bonded dimer 4, and only slowly does free CO add to give 0.5 equivalents of 3 . Therefore, our analysis of the reaction of CpMoc 13 co) 3 H with 2 concentrated upon the 13 co content of 4. To analyze for this dimer the reaction solvent was removed and then 4 was isolated by either column chromatography (Florisil, hexanes/benzene) or by HPLC (reverse phase, eluting with 20% H2 o/cH 3 CN as eluant); for this separation column chromatography proved more convenient. The resulting purified dimer was then subjected to mass spectral analysis (50 ev, 150°C inlet temp.). Since molybdenum contains seven isotopes distributed over a mass range of 9 units, in significant amounts, the mass spectrum of [CpMo (CO) 2 J 2 containing a natural abundance of the elements has parent ions at 18 different m/e values (Figure 3). Analysis of these spectra when the dimers contain 13 c carbonyls was achieved by a least squares fit of the relative intensities of the parent ions assuming each of the isotopomers gave the same relative ratios at (m/e + n) for n 13 co's. This method of analysis is explained in greater detail in appendix B. The net result is the percentage of 4 that contains n = 0,1,2,3, or 4 13 co's, and thus the average percentage of 13 co in the product. The reaction of CpMoc 13 co) 3 H (la) with 2a was carried out under two distinctly different conditions, and then analyzed for CpMo(C0)2=(C0) MoCp 2 ♦ • I ?:l+CO I CpMo(C0) 3-(C0)3MoCp Scheme m fl 3 • 2 CpMo(CO)~ + 2CO 2 CpMo(C0) -.J -.J 78 the 13 co content in 4 as described above. These results are presented in Table 2. In reaction 1 1c 13 co> 3 and 2a were heated in sealed NMR tube to 50°C for 26 hours. Under these conditions any CO liberated during the reaction cannot escape and has time to add to 4 and result in scrambling of the carbonyl labels . In reactions 2 and 3 the reaction solution was heated to 50°C under a moderately vigorous purge of nitrogen or argon; in this case liberated CO is swept out of the system. The observation of large amounts of 4a containing o, 1 and 4 13 co's immediately indicates the existence of a scrambling mechanism. It was conclusively shown that some scrambling of the CO's (presumably between product molecules) occurs under the mass spectral conditions. In a control experiment a sample of 4 from reaction 2 was mixed with some unlabeled 4 and then analyzed by mass spectrometry. The relative ratios of the 13 co containing products changed in the resulting mass spectrum, implicating exchange of CO's between the natural abundance dimer and the 13 co containing products. No conditions could be found that would completely eliminate this apparent scrambling; however, the total 13 co content of the products always remained about the same. The similarity of the total 13 co content in 4 from both experiments 2 and 3, where different reaction times are involved, suggests that any such scrambling must occur after the total CO content of the product has been determined. Curtis and Klingler have established that scrambling of the CpMo(C0) 2 fragments in 4 does occur, but this mechanism is much too slow to account for our 79 435 4:;:6 438 ._,, 4 ·":>7 4::::3, 43 1 43 ►3 440 429 4:::•=t 428 I I 426 I 441 4f7 I I I I I 443 I I I I 430 435 440 I J r -r I I 445 Figure 3 . Parent Ion Mass Spectrum of Natural Abundance 4a. ".I 80 observations. 13 In an attempt to eliminate this scrambling the product, 4, was converted to CpMo(CO) 2 (PMe 3 )cH 3 (2e) which is much more volatile. Analysis of 2e derived from reaction 3 in this way gives the results shown in Table 3. These results corroborate those in Table 2, and will be discussed in more detail in the next section. Trapping Experiments. The following trapping experiments were designed that might trap possible intermediates of the type [CpMo(CO) 3 l+ or [CpMo(CO)3l-. In all cases no significant amounts of trapping products could be observed. a) cn 3 I. Bergman and Jones 1 reported that addition of a 1O- fold excess of Mel to the reaction of la and 2b did not yield any observable CpMo(CO) 3 cH 3 , the expected [CpMo(CO) 3 ]- and CH 3 r. 18 b) LiCl. product of reaction of Addition of excess LiCl (> lM) to the reaction of la with 2a at 50°C in THF did not result in any detectable formation of CpMo(CO)Cl by NMR or IR either during or after the reaction. The presence of the intermediate [CpMo(CO) 3 ]+, a known compound, 19 would have been expected to lead to reaction with chloride ion to give CpMo (CO) 3 c1. Attempts to trap the c) intermediate [CpMo(Co> 3 1+ in the reaction of la with 2 by the processes in equations (1) and (2) were thwarted by the facile exchange of chloride and hydride ligands in the starting complexes, equations (3) and (4). 81 (1) MeCpMo(C0)3+ + CpMo(C0)3Cl (2) MeCpMo(C0)3+ + CpMo(C0) D 3 _.. MeCpMo(C0) 3 Cl + CpMo(C0) 3 + . MeCpMo(C0)3D + CpMo(C0)3+ ( 3) MeCpMo(C0)3H + CpMo(C0)3Cl ~ ~ MeCpMo(C0)3Cl + CpMo(C0)3H (4) MeCpMo(C0)3H + CpMo(C0)3D ~ ~ MeCpMo(C0)3D + CpMo(C0)3H 82 Table 2. 13 co Distribution in (CpMo(CO) 2 ) From the Reaction of 2 13 CpMo( CO) H with CpMo(CO) cH . 3 3 3 Rxn Rxn time lb a Percent of Molecules 13 With n co's Corrected Average n= 0 1 2 3 4 % 13co 26 7 21 32 29 12 59 2c 5 2 11 29 38 20 69 3d 21 1 10 28 39 22 72.5 2 + 4ae - 40 12 20 20 8 - co)H 0 1 16 83 - 93.8 CpMo( a 13 50°C. b sealed tube. c d under a purge of Ar. under a purge of N2 . ethe product of Rxn 2 was diluted with unlabeled 4a, then analyzed by mass spec. Table 3. 13co Distribution in CpMo(CO) 2 (PMe )cH Derived from 3 3 The Product of Reaction 3. n = 0 12 Percent of Molecules 13 co's With n 2 1 45 44 Corrected Average % 13co 70.5 83 DISCUSSION The reaction of molybdenum hydride (1) with molybdenum alkyls (2) in THF proceeds to give aldehyde and the two molybdenum dimers [CpMo{C0) 3 l 2 (3) and [CpMo{C0) 2 l 2 (4) as the sole metal containing products. Using a methylcyclopentadienyl ligand to label either the molybdenum hydride or molybdenum alkyl it was possible to determine the origin of the molybdenum centers in these dimers. At temperatures of 50°C and below it was found that the mixed dimers 3b and 4b, containing one labeled and one unlabeled cyclopentadienyl ligand, predominate. This result was found to be independent of which starting material was labeled. At temperatures greater than 80°C, however, a thermodynamic mixture of all six dimers was observed. At these temperatures secondary reactions involving metal-metal bond cleavage lead to equilibration of the product mixture {Scheme III). Thus the dimers 3b and 4b are the kinetic products of the reaction. These observations qualitatively suggest, but do not require, that the metal-metal bond is formed prior to or concomitant with reductive elimination of aldehyde from intermediate 7. This kind of "concerted" mechanism would give dimers that contain one metal center derived from each reactant, in a process involving the intramolecular reductive elimination of aldehyde from intermediate 7 (vid_g infra). Alternatives to a "concerted" mechanism of this kind involve either proton or hydride transfer. Both of these possibilities lead to charged I 84 intermediates that can still give unsymmetrical or mixed products under these conditions. Bergman and Jones 1 were able to exclude proton transfer on the basis of three observations: a) the much less acidic CpMo(C0) 2 (PMe 3 )H Cle) reacted with 2b at a rate (k 25 c= 2 x 10- 3 M- 1 sec- 1 ) comparable to that for la; b) addition of a 10 fold excess of CH 3 I to the reaction of la and 2b did not trap any of the molybdenum anion [CpMo(C0) 3 ]-; c) no reaction was observed between 2a and acetic acid in THF-da after 24 hours at so 0 c, 20 whereas trifluoroacetic acid reacts with 2a (presumably via direct protonation of the unrearranged alkyl) to give methane. A second alternative involving hydride transfer cannot be so easily ruled out. In this mechanism (Scheme IV) hydride transfer from 1 to the intermediate 5 gives [CpMo(Co> 3 1+ and the anionic hydride acyl intermediate 9. Reductive elimination of aldehyde from 9 to form an anionic [CpMo(C0) 2 ]- species, followed by recombination of this species with [CpMo(Co) 3 l+ could give the pentacarbonyl dimer 10. Dimer 10 can either lose or add carbon monoxide to give the observed products 4 and 3 respectively. The inability of chloride ion to trap the proposed intermediate [CpMo(Co) 3 l+ is not sufficient to exclude this mechanism, because it cannot be shown that CpMo(C0) 3 c1 is stable to the reaction conditions (e.g. displacement of chloride by CpMo(C0)n- the other charged fragment in solution). Finally there exists a "concerted" mechanism in which the bridging hydride in intermediate 7 is transferred to the metal in 85 Scheme Til ..---~ ~ CpMo(C0) C-R 2 5 l CpMorC0)3H H;y 0 R - /Mo Co I Co 9 0 II -RCH l 4 -co *)o 10 -Mo-Mo 1~1 I co to co 10 +CO _ _,.► 3 86 the acyl complex, facilitating formation of a metal-metal bond (Scheme V). In this mechanism loss of CO, hydride migration to the acyl-bound metal, and formation of the metal-metal bond (i.e. oxidative addition of the M-H bond to the second metal center) will give the intermediate 8. This hydrido-acyl complex undergoes reductive elimination of aldehyde and migration of CO from the other metal center to form [CpMo(CO) 2 J 2 (4a) directly. The direct formation of 4 in this manner is consistent with the observation that 4 is the principal metal-containing product observed by NMR at early reaction times. These mechanisms make specific predictions concerning the origin of the carbonyl's that end up in the tetracarbonyl dimer 4. In the hydride transfer mechanism outlined in Scheme IV random loss of a carbonyl ligand from intermediate 10 will lead to 4 that has 3/5, or 60%, carbonyl's that came from the molybdenum hydride. Alternatively if intermediate 7 is not CO labile, then transfer of hydride and formation of a metal-metal bond can give an intermediate of the type 11. Reductive elimination of aldehyde from 11 will give the pentacarbonyl intermediate 10. This mechanism and any other that involves 10 as the intermediate preceding dimer formation will give the same prediction. However the mechanism presented in Scheme V predicts that three of the carbonyl's in 4, or 75%, come from the molybdenum hydride. 87 Scheme Y. 0 II CpMo(C0) 2C-R 5 -co 8 7 \RLH kt)~~~=~~ ©\ ~ *co *co f!J 4 +2CO 3 88 0 CO CO f:f ~~ ~ ......-R 1 \ ' -Mo-Mo II do &\ Labeling the carbonyl's in the molybdenum hydride starting material with 13 co it now becomes possible to distinguish these mechanisms. The results presented in the Table 2, when corrected for the isotopic purity of the starting material, are in reasonable accord with the mechanism in Scheme v. Under conditions where CO liberated during the course of the reaction has the chance to escape the system, i.e. under a nitrogen purge, 69 to 72.5 % of the carbonyls in 4 were found to come from 1. This observation approaches the value of 75% predicted by the mechanism in Scheme V, i f ~ carbon monoxide liberated during the reaction is swept out of the system. Considering the inefficiency of the method employed to remove CO during the reaction this agreement seems good. Nevertheless the high degree of CO scrambling evident from the distribution of 13 co's in 4, Table 2, is worrisome. At least some of this scrambling appears to occur in the mass spectrometer. In a closed system, where liberated CO cannot escape the system and has time to react with the initially formed products, 59% of the carbonyls in 4 orginate from the molybdenum hydride. This is consistent with the fast reversible addition of CO to 4 89 to give intermediate 10, and the slow addition of a second molecule of free CO to form 3 (Scheme VI). The dimer 4 was chemically converted to the much more volatile CpMo(C0) 2 (PMe 3 )CH 3 (2e), in an effort to eliminate mass spectral scrambling. Compound 2e was found to have a 13 co content comparable to the original dimer. The distribution of 13 co label in 2e is a less sensitive measure than in 4; however, even here at least some scrambling appears to have taken place. The observed ratios of 12, 45, and 44% should be compared to the values of 4, 50, and 46% for n= o, 1, and 2 13 co•s predicted by Scheme V (Table 3). The mechanism in Scheme IV on the other hand would predict values of 18, 50, and 32% for this product. The results presented here cannot be considered conclusive. In particular a good explanation for the origin of the 13 co scrambling in 4 is lacking. However, if indeed the scrambling is occurring~~~~£ the CO content in the products has been determined (i.e. between product molecules), then the observed average 13 co content in 4 is most consistent with the mechanism in Scheme v. If this conclusion is correct, from the molybdenum alkyl, prior then the observed loss of CO to product formation strongly sug~ests that intermediate 7 loses CO, enabling the hydride to migrate to the acyl-bound metal and the formation of a metalmetal bond. Thus it would appear that aldehyde is formed from the intramolecular reductive elimination of a hydrido-acyl metal complex. 4 [ CpMo(C0) 2 ] Scheme :srr 2 .. -co ► t ast • - 10 I slow ~ 8 +co ~o co ~~ I , ~o~ -~o ~o I co co ► 3 \.0 0 [ CpMo(C0) ] 3 2 91 EXPERIMENTAL All General. manipulations involving organomolybdenum complexes were carried out either by standard Schlenk techniques or in a Vacuum/Atmospheres Corp. Model HE-553 inert atmosphere glove box with continuously circulating nitrogen atmosphere and a model M0-40 recirculating purification system. All molybdenum alkyls and hydrides were stored cold (-4o 0 c) and in the dark, except 2a which is more thermally stable. All reactions and subsequent manipulations of products were conducted so as to avoid photochemical side reactions: of NMR tubes and reaction flasks, this involved foil wrapping and the performance of column chromatography under dim, indirect light. Proton Nuclear magnetic resonance ( 1 H NMR) spectra were obtained using a Varian EM-390 90 MHz spectrometer, or a 180 MHz FT-NMR instrument equipped with a Bruecker superconducting magnet and a Nicholet Instrument Corp. 1180 data system and electronics assembled by Mr. Rudi Nunlist (U. C. Berkeley). Infrared (IR) spectra 237 were obtained Spectrophotometer. on a Perkin-Elmer Microanalyses and mass spectra Grating (MS) were obtained from the analytical facilities of the University of California at Berkeley. Mass spectral samples were prepared in capillaries by introducing the evaporating the solvent, and then compound as a solution, sealing the capillaries under an inert atmosphere. For analysis the capillaries were cracked open, inserted in the mass spectrometer probe, and evacuated with a minimum exposure to air. Photolyses were carried out through 92 pyrex using an Oriel Corp. focused beam apparatus and power supply equipped with an OSRAM 500 watt HBO Super Pressure Mercury Lamp. High pressure liquid chromatography (HPLC) employed an Altex Scientific Inc. model llOA high pressure pump, a Rheodyne model 905-42 variable volumn injector valve, and an Altex model 153 UVVIS detector equipped with a 2mm pathlength 254 nm wavelength kit. Separations were achieved using a 1 x 25 cm Ultrasphere-ODS (Altex) reversed phase column claiming complete coverage of the active sites. The solvent was degassed by continuously purging the solvent reservoir with argon. 15 Commercial samples of [CpMo (CO) 3 1 2 and [MeCpMo (CO) 3 1 2 (Cp= 775-c 5 H5 , MeCp= 7T-CH 3c 5 H4 > (Alfa) contained considerable amounts of non-volatile impurities. These impurities could be removed by column chromatography on Florisil, eluting with benzene or THF, and collecting the dimer which elutes as a red band. In less sensitive experiments (for example in the preparation of 2a) this procedure was not always followed, especially where the products could be sublimed away from these impurities. Mo(C0)6 was used without purification. NaCp and NaMeCp were prepared by the standard method from Na metal and the freshly cracked diene in THF. The compounds CpMo (CO) 3c 2 H5 (2b) , 18 CpMo(Co) 3 H (la), 17 MeCpMo (CO) 3cH 3 CpMo(C0) 3 ctt 3 (2d) , 18 (2a), 18 and Mo (CO) 3c1 13 ,lS were prepared by literature procedures, and purified by sublimation. Their IR and NMR spectra are summarized in Tables 4 and 5. 93 Table 4. 1 H NMR Data for CpMo(CO) 3 R Complexes. Chemical Shiftsa Complex: Cp R la: CpMo(CO) 3 H 5.56 -5.55 lb: MeCpMo(CO)3H 5 . 45 ( AA I BB I J = 'AB -5.45 17, JAA,= 2.1) 2a: CpMo(CO) 3 cH 3 5.45 0.40 2b: CpMo(CO) 3 Et 5.30 1. 70 1. 44 2c: CpMo(CO) 3cH 2Ph 5.29 2. 8 9 ( s) , 7.16 (m) 2d: MeCpMo(CO) 3 cH 3 5.26 0.32 CpMo(CO) Cl 3 5.75 (q, J = 6) (t, J = 6) ain 8ppm in THF-d , J is in Hz, singlets unless otherwise 8 noted. 94 Table 5. Infrared Data for CpMo(CO) 3 R Complexes. Complex: Solvent v(CO) cm -1 la: CpMo(CO) 3 H THF 2020, 1948 lb: MeCpMo(CO) 3H 13 le: CpMo( CO) 3H THF 2020, 1930 THF 1197 (w) , 1978 (s), 1915 ( sh) , 1890 ( s) 2a: CpMo(CO) 3 cH 3 THF 2015, 1930 2d: MeCpMo(CO) 3 cH 3 THF 2015, 1930 CH CL 2 2 2055, 1980 CpMo(CO) Cl 3 95 All solvents employed were thoroughly dried and degassed by the methods outlined in the previous chapter. Similarly THF-d 8 was distilled (reduced pressure) from sodium benzophenone ketyl prior to use. High pressure liquid chromatography (HPLC) solvents were not dried, but were millipore filtered to remove particulates prior to use. The Synthesis of MeCpMo(C0) 3 B, lb. The previously unknown MeCpMo(C0) 3 H {lb, MeCp ='T(-CH 3C5 H4) was prepared by the following adaptation of the standard method of synthesis for la. A solution of 1.05 g (3.72 mmol) [MeCpMo(C0) 3 J-Na+ (prepared from NaMeCp and Mo(C0) 6 by the literature method 17 > in 30 ml THF was treated with 0.7 g (4 mmol) p-toluenesulfonic acid as a THF solution in the glove box. A copious precipitate formed immediately. After 15 minutes an IR spectrum of this solution indicated only molybdenum hydride and no anion. The yellow solution was filtered and the solvent removed in vacuo to yield a red oil. This oil was distilled at reduced pressure to yield lb as a pale yellow, light and thermally sensitive oil; b.p. 55-60°C (0.01 mmHg). IR(THF) v(CO): 2015(s), 1930(s) cm- 1 . 1 H NMR(THF-ds>: 8 -5.42 (s, lH), 2.12 (s, 3H), 5.46 ppm (AA'BB' quartets, JAB= 17 Hz, JAA'= 2.1 Hz, 4H). The Synthesis of 3a, 3b, and 3c. Standard Mixture. A solution of 200 mgs (0.4 mmol) [ CpMo (CO) 3 J 2 (3a) and 220 mg (0.42 mmol) [MeCpMo(C0) 3 J 2 (3c) in 50 ml benzene was irradiated through pyrex for 12 hours with a 500 watt UV lamp, with cooling from a water bath. The solvent was removed in vacuo and the resulting 96 red residue was chromatographed on basic alumina I (2 x 25 cm column) eluting with THF. One red band was collected, and a bluegrey residue remained at the origin. The solvent was removed to yield a mixture of [CpMo(C0) 3 J 2 3a, MeCpMo(Co) 3 -cco) 3 MoCp 3b, and [MeCpMo(C0)3l 2 3c. The HPLC chromatogram of this mixture is presented in Figure 4. Coinjection of samples of 3a and 3c determined the order of elution to be 3a, 3b, 3c. Note that the relative ratios of these compounds is roughly 1:2:1. This is consistent with the expected equilibrium ratio of these isomers. IR(THF) v(CO): 2010 (w), 1960 (s), 1915 (s) cm- 1 . 1 H NMR(THF-da>: 2.15 (s), 5.30 (AA'BB' multiplet, JAB= 12 Hz), 5.40 ppm (s). The Synthesis of 4a, 4b, 4c. Standard Mixture. Following the procedure of Curtis and Klingler 12 a solution of 250 mg (0.51 mmol) 3a and 250 mg (0.48 mmol) 3c in 150 ml toluene was brought to reflux under a gentle Nz purge. After 18 h the IR spectrum of this solution indicated complete conversion to product. The solvent was removed in vacuo. The resulting residue was extracted thoroughly with hexanes and the resulting solution filtered through a 2 gm plug of florisil. The florisil plug was washed with benzene-hexanes (1:1) to yield a red solution of the products. The solvent was [CpMo(CO)zl2 (4a), removed to yield a mixture of MeCpMo(CO)z=(CO)zMoCp (4b), and [MeCpMo(CO)zl2 (4c) as a red solid. The HPLC chromatogram of this mixture is shown in Figure 5. Again a 1:2:1 ratio of isomers is observed in accordance with the expected equilibrium mixture. IR(hexanes) v(CO): 1965(w), 1898(s), 1870(s) cm- 1 . 1 H NMR(THF-ds) :8 2.05 (s), 97 E C - 3b - 3c V l() C\J w u z - 3a <l m a: 0 (/) m <l 36 30 24 18 12 TIME (min) 6 0 Figure 4. HPLC Chromatogram of an Authentic Mixture of 3a,b,c. ~ "- ,. _ 98 4b E C V I{) - C\J 4c w u - 40 z <{ m a: en 0 m <{ 30 24 18 12 TIME (min) 6 0 Figure 5. HPLC Chromatogram of an Authentic Mixture of 4a,b,c. rv ,.., ,.., 99 5.15 (AA'BB' d of quart., JAB= 5.6, JAA'= 2.1 Hz), 5.28 ppm (s). The Synthesis of [CpMoC 13 co> 3 J 2 . In a 250 ml round bottomed flask equiped with a vacuum stopcock a solution of 250 mg (0.51 rnrnol) [CpMo (CO) 3 1 2 in 100 ml benzene was degassed then charged w i th 5 . 4 rn rn o 1 13 CO (,.. 9 4 % 13 CO ) (1 e s s th an 1 at rn pr es s u r e ) . Th e solution was irradiated with a 500 watt UV lamp for 5 hours with cooling from a water bath. The solution was taken through two freeze-pump-thaw cycles to remove free co. This procedure was repeated a total of four times, the solvent was removed and the resulting residue was chrornatographed on basic alumina I eluting with THF. One red band was collected. The solvent was removed to yield a red solid. Comparing the intensities of the parent ion at 490 and 489 rn/e in the mass spectrum a rough isotopic purity of 93.7 atom% 13 co could be calculated. IR(THF) v(CO): 1967(w), 1915(s), 1870(s) crn- 1 . The Synthesis of CpMoC 13 co> 3 a. A 25 ml Schlenk flask was charged with the entire 250 mg (.... 51 rnrnol) [CpMoC 13 co> 3 1 2 in 20 ml THF, then capped with a rubber sept urn. Under ni tr oge n O. 5 ml (1.8 rnrnol) 0.65% Na/Hg amalgam was added via syringe. In less than 30 min the red solution had turned pale yellow indicating conversion to the anion. The solution was then transferred by cannula away from the amalgam to a Schlenk flask that also serves as a sublirnator bottom. A solution of 250 mg (1.5 rnrnol) p- toluenesulfonic acid in 10 ml THF was added via cannula. The yellow solution turned almost clear with the formation of a fine white precipitate. After 15 minutes an IR spectrum of this 100 solution indicated quantitative formation of the hydride. The solvent was removed in vacuo, and the residue was sublimed (50-60° C, 0.01 mmHg) to yield a somewhat impure hydride. A second sublimation yielded 140 mg (55% overall yield based upon 3a) CpMoC 13 co)H. Mass spectral analysis (see Appendix B) gave the following percentages of compounds: 1c 13 co> 3 = 82.8 ±3%, 1( 13 co) 2 = 15.9 ±3.1%, 1( 13 co) 1 = 1.2 +3.1%; or 93.8 atom% 13 co. IR(THF) v(CO): cm- 1 . 1997 (w), 1977 Cs), 1915 (shoulder, w), 1890 (vs) 1 H NMR(THF-d8): 8 5.55 (s, SH), -5.52 ppm (1:3:1:3:1 quintet, J=l3 Hz, lH). The Reactions of lb with 2a, lb with 2b, and la with 2d. In all cases the two starting materials were weighed into an NMR tube in the glove box. Then on a vacuum line THF-d 8 was condensed into the tube and it was sealed off under vacuum. The resulting solutions were heated in a temperature controlled water bath. The reaction progress was monitored by cooling the NMR tubes and recording the NMR spectrum at room temperature. Light was excluded from these solutions by wrapping the NMR tubes in foil. Analysis of the products was achieved by cracking the NMR tubes open in the dry box, removing the solvent, redissolving the resulting residue in CH 3 CN, and analyzing by HPLC. High pressure liquid chromatographic (HPLC) analysis was achieved using a 1 x 25 cm Altex Ultrasphere-ODS reversed phase column, eluting with 20% H2 o/cH 3 CN saturated with injections argon. 20-50 µ,1 completely eluted over a 35 minute time period at a flow rate of o.s ml/min. Typical pressures at this flow rate were 400-600 psi. 101 The separation achieved under these conditions is demonstrated in the chromatogram of an authentic mixture of 3a,b,c and 4a,b,c shown in Figure 1. A) A solution of 19 mg (O.O77 mmol) la and 20 mg (.073 mmol) 2d in O.4 ml THF-d 8 (approximately 57% was heated reaction) to 50(±.5)° C for hours 22 and then analyzed by HPLC. resulting chromatogram is presented in figure 6. The In addition to 2d and products arising from decomposition of la in acetonitrile solution 21 , the dimers 3a, 3b, and B) A solution of 4b were observed. 18 mg (O.O7 mmol) lb and 20 mg (O.O77 mmol) 2a in 0.4 mmol THF-d 8 was heated to 50(±.5)° C for (approximately 64% reaction) and then analyzed by HPLC. resulting chromatogram is shown in Figure 7. unreacted 2a and products hours 22 from The In addition to decomposition of lb in acetonitrile, 21 significant amounts of the dimers 3b, 3c, and 4b are observed. C) A solution of 22 mg (O.O85 mmol) lb and 27 mg C0.O98 mmol) 2b in O.58 ml THF-d 8 was allowed to react at room temperature for 48 hours, at which point no unreacted lb could be observed by NMR. HPLC analysis gave the chromatogram in Figure 2. In addition to unreacted 2b the only other compounds present in significant amounts are the dimers 3b and 4b. D) A solution of 17 mg (O.O65 mmol) lb and 18 mg (O.O69 mmol) 2a in o.78 ml THF-d 8 was heated to 1OO°c for 15 hours. HPLC analysis gave the chromatogram in Figure 8. All six dimers 3a,b,c and 4a,b,c are observed in significant concentrations. The 102 X 2d E C X V 3a &I') (\J w u - - 3b z 4b <X Q) a:: 0 en CD X <X X X -- ) V. ~ 3c A I 30 . I I I 24 18 12 6 ( min) TIME 0 I I I Figure 6. HPLC Chromatogram of the Reaction Products of The Reaction of r,.,I la and A../ 2d. xdecomposition products of molybdenum hydride 103 X - 4b ~ C: 3b ..... V lfl (\J 2a ....... LLJ (.) z <X m a:: 0 en m X <t X 3c ....... 3a ..... ~ ~J ~ I I 36 30 .._ r I I I 24 18 12 TIME (min) 6 0 I I Figure 7. HPLC Chromatogram of the Products of the Reaction of lb and 2a. l'V A./ xdecomposition products of molybdenum hydride. 104 E - - 3b C 4b V I() (\J w 3a 4~ - - 3c u z - 40 <l CD 0:: 0 V) CD '----.r------t <l 36 30 24 18 12 TIME (min) 6 0 Figure 8. HPLC Chromatogram of the Products of the Reaction of lb and 2a at 100°C for 15 hours. A./ A/ 105 relative ratio of these dimers closely resembles the equilibrium mixture in Figure 1. The products are labeled in these chromatograms. All known compounds were identified by coinjection of authentic samples. Reaction of CpMoc 13 co) 3 H with CpMo(Co> 3 ca 3 . Reaction 1: In the dry box 7 mg (O.O28 mmol) CpMo c13 co) 3 H and 21 mg (O.O8 mmol) 2a were weighed into an NMR tube. Then on a vacuum line 0.5 ml THF- d 8 was condensed into the tube and it was sealed off under a vacuum. The solution was heated to 5O(±.5)°C for 20 hours (88% consumption of hydride). The tube was cracked open in the dry box and the solvent removed. The resulting residue was chromatographed by HPLC on a reversed phase column under conditions identical to those in the previous experiment. Only 3a, 4a, and 2a and a small amount of molybdenum hydride decomposition products could be detected. 4a was collected under argon, the solvent was removed and the resulting red solid was analyzed by mass spectrometry. The relative percentages for molecules containing n 13 co's (n= o, 1, 2, 3, 4) was evaluated by a least-squares treatment of the relative intensities of the 20 line parent ion "envelope"; details of this analysis are given in Appendix B. The results of the least squares treatment are presented in Table 2. Reaction 2: A solution of 10 mg (O.O4 mmol) CpMoC 13 co) 3 H and 20 mg (O.O77 mmol) 2a in 2 ml THF prepared in a 50 ml Schlenk flask equipped with a reflux condenser capped with a rubber septum. The solution was heated to 50(±2)°C in the dark with an oil bath 106 while a vigorous purge of argon was maintained through needles in the rubber septum. After 5 hours heating was discontinued. The reflux condenser was replaced with a glass stopper and the solvent was removed. 4a was isolated by column chromatography in the dark on a 1 x 15 cm florisil column eluting first with 10% benzene/hexane (50 ml), next with 50% benzene/hexane (50 ml), then 75% benzene/hexane. A pale yellow fraction containing la and 2a eluted first, followed by a very small red band containing 3a, 4a eluted last as a yellow-orange band. No compounds appeared to remain on the column and no observable decomposition occurred. Molybdenum hydride is stable to florisil but not alumina or silica. The solvent was removed from the fraction containing 4a and mass spectral samples were prepared in capillaries. Analysis of the mass spectrum in the manner described in Appendix B gave the data presented in Table 2. Reaction 3: A solution of 20 mg (O.O8 mmol) CpMo( 13 co> 3 H and 22 mg (O.O85 mmol) 2a in 2 ml THF was heated to ~0(±2)°C for hours under a nitrogen purge. chromatography and mass Isolation of spectral analysis 21 4a by column gave the results presented in Table 2. A sample of 4a isolated from this experiment(< 1 mg) was dissolved in 1 ml THF and placed in a 5 ml flask with a teflon stopcock. On a vacuum line an excess of PMe 3 (~10 mmol) was condensed into the flask. 13 The solution was allowed to stir for 40 minutes at room temperature, and then the volatiles were removed. The resulting red solid was dissolved in 1 ml THF and 107 treated with an excess of 0.6% Na/Hg amalgam (~2 mmol) it turned a pale yellow color. After 5 minutes the solution was pipetted away from the amalgam and placed in a clean 25 ml Schlenk flask. Then an excess of CH 3 I (3 mmol) was added via syringe through a septum. After 10 minutes CpMo(CO) 2 (PMe 3 )cH 3 (2e), the solvent was removed to yield as an off-white solid. An authentic sample of 2e was prepared in the identical manner from natural abundance 4a. Mass spectral analysis in a manner identical to that for 4a gave the data shown in Table 3. Trapping Experiments using LiCl. A solution of 20 mg (O.O8 mrnol) la, 20 mg (O.O77 mmol) 2a and 120 mg (3 mmol) LiCl in 1 ml THF was heated to 5O°C in a Schlenk flask equipped with a reflux condenser for 40 hours while monitoring by IR at regular intervals. A very weak absorption at 2050 cm- 1 was observed, in addition to the expected absorptions at 2015, 1960, 1930, 1855 cm- 1 , but no strong absorption at 1980 cm-l (characteristic of CpMo (CO) 3 c1) was observed. In a similar experiment 17 mg (O.O7 mmol) la, 18 mg (O.O7 mmol) 2a and 20 mg (0.5 mmol) LiCl were weighed into an NMR tube. On a vacuum line O.65 ml THF-d 8 was condensed into the NMR tube and it was sealed off under a vacuum. The solution was heated to 50°C in a water bath and NMR spectra were recorded at regualar intervals at room temperature. No Cp resonance at 85.65 ppm for CpMo(Co) 3c1 could be observed. After 16 hours of heating the NMR tube was cracked open and an IR spectrum recorded in d 0 •THF. Again a very weak absorption at 2050 cm- 1 was observed, however 108 the absence of a band at 1980 cm-l leads to the conclusion that LiCl does not trap [CpMo(CO) 3 l+ to give CpMo(CO) 3 c1 in these reactions. The Reaction of lb with CpMoCco> 3 c1. In the dry box 10 mg (O.O38 mmol) lb, 20 mg (O.O77 mmol) 2a, and 20 mg (0.093 mmol) CpMo(CO) 3 Cl were weighed into an NMR tube. Then on a vacuum line 0.5 ml THF-d 8 was condensed in the tube and it was sealed off under a vacuum. The initial NMR spectrum after warming to room temperature already contained la, and a new MeCp resonance at 1.98 ppm, most likely MeCpMo(CO) 3 Cl. No lb could be observed. Clearly all of lb was converted to la, presumably with concomitant formation of MeCpMo(Co) 3 Cl, at a point where no reaction between 1 and 2a had taken place yet. The Reaction of CpMoCC0) 3 o with lb. In the dry box 20 mg (O.O81 mmol) CpMo (CO) 3 o, la Cd 1 ), 21 mg (0.08 mmol) lb, and 11 mg (O.O42 mmol) 2a were weighed into an NMR tube. On a vacuum line 0.5 ml THF-d 8 was condensed into the NMR tube and it as sealed off under a vacuum. After 30 minutes at room temperature the NMR spectrum revealed that virtually equal amounts of la and lb were present in solution. A similar reaction in benzene-d 6 of 10 mg (O.O4 mmol) la (d 1 ) and 15 mg (O.O6 mmol) lb also gave considerable la at room temperature in less than 10 minutes. These reactions appear to have reached equilibrium faster than their NMR spectra could be recorded. 109 REFERENCES 1. W. D. Jones, R. G. Bergman, .J.i. A!!h. ~.!lh. ~ , 1979, l.Qi, 5447; W. D. Jones, Ph.D. Thesis, Cal. Inst. of Tech., 1979. 2. a) M. Orchin, F. Heck, D. c) D. s. w. Rupilius, ~ ~ , 1972, till, 85. b) R. s. Breslow, .J.i. A.!lh. ~.!lh. ~ , 1961, 83, 4023. Breslow, R. F. Heck, ~.!lh. .l..Ilih (London), 1960, 467. d) W. Rupilius, M Orchin, .J.i. .Q.Is... ~I!l...., 1972, l l , 936. 3. N. H. Alemdaroglu, J. L. M. Penninger, E. Oltay, Monatsheft L ~I!l...., 19 7 6 , l.Q2 , 1 o 4 3 ; i b i a. , 19 7 6 , 1 o7 , 11 5 3 . 4. M. Wax, R. G. Bergman, unpublished results, a first order rate constant of 5.5 x 10- 6 sec-lat 70 C in benzene was observed 5. I. for the reaction of la and 2a. s. Butler, F. Basolo, R. G. Pearson, Inorg, ~!!h., 1967, Q, 2074. See also the analogous study on CH 3 Mn(Co) 5 ; R. J. Mawby, F. Basolo, R. G. Pearson, .J.i. A!!h. ~!!h. ~ , 1964, .a&., 3994. 6• P. J. Cr a i g , M. Gr e e n , .J.i. ~!!h. ~ ..<Al, 1 9 6 8 , 1 9 7 8 . 7. a) L. Abis, A. Sen, J. Halpern ,.J.i. A!!h. ~!!h. ~ , 1978, l.Q.Q, 2 915. b) A. Yamamoto, T. Yamamoto, Bull. .cM!!h. ~ Japan, 1975, 49, 1403. 8. a) J. J. Okrazinski, J. Norton, .J.i. AI!l.... ~!!h. ~ , 1977, 99, 295 . b) J. R. Norton, ~ Che!!h. Res., 1979, .1..2., 139. 9. J. Halpern, .J.i. A.!lh. ~.!lh. ~ , 197 9, .2.l, 21 71. 10. E. L. Meutterties, ~!!h. Rev,, 1979, ll, 91. 11. F. A. Cotton, reaction, and A. Bino have reported the reverse and oxidative addition to a quadruple metal-metal 110 bond, Angel'.LL .c..he...m..,_ .I..n.b.. ~ Engl,, 1979, 18, 332. 12. a) M. Y. Darensbourg, J. L. Atwood, W. E. Hunter, R. R. Burch, Jr., .J..... .AID...t.. ~..m..,_ ~ , 1980, l.Q.2., 3290. b) R. Bau, R. Teller, S. W. Kirtley, T. F. Koetzle, ~ ~..m..,_ ~ , 1979, 12, 176. c) H. D. Kaesz, R. Bau, M. R. Churchill, .J..... AID...t.. .c..he.ID...t.. ~ , 1967, 89, 2775. d) R. G. Hayter, ibid,, 1966, _a.a, 4376. e) L. B. Handy, P. M. Treichel, F. Dahl, R. G. Hayter, ibid,, 1966, lilt, 366. 13. a) M. D. Curtis, R. J. Klingler, .J..... Organom.e..t.,_ .c..he.ID...t.., 1978, .lil, 23. b) R. J. Klingler, W. Butler, M. D. Curtis, .J..... ,A..m..,_ .c..he.ID...t.. ~ , 1975, il, 3535. 14. a) D. S. Ginley, C. R. Bock, M. S. Wrighton, Inorg, Chi..m..,_ Ac ta. , 19 7 7 , 2..1 , 8 5 . b ) M. S. Wr i g ht o n , D. S. Gi n 1 e y , .J..... ,A..m..,_ .c..he.I!li. ~ , 1975, ll, 4246. c) D. S. Ginley, M. S. Wrighton, ibid,, 1975, 12, 3533. 15. J. King, P. Vollhardt, J. M. Huggins, in preparation. 16 • C. Gi an no t t i , G. Me r 1 e , ~ Or g an o .me t, .c..he...m....., i 9 7 6 , l.Q.2., 9 7 . 17. a) R. B. King, F. G. A. Stone, b) Inorg. Synth., 1963,2, 107. s. A. Keppie, M. F. Lappert, .J..... Organom.e..t.,_ .c..he.I!li., 1969, li, PS. c) E. O. Fischer, Inorg, Synth,, 1966, 2, 136. d) E. o. Fischer, w. Hafner, H. o. Stahl, b.. Anorg, Allg • .c..he...m....., 1955, 282, 47. 18. T. s. Piper, G. Wilkinson, .J..... Inorg. N.Y..Q.L. .c..he..mi., 1956, ~, 104. 19. a) W. Beck, K. Schloter, b b) P. M. Treichel, R. L. Nat.Y..r..f.Qllch., 1978, 33b, 1214. Shubkin, Inorg. Chelfu., 1967, 2, 111 1328. c) P. M. Treichel, K. W. Barnett, R. L. Shubkin, .J:..t.. Organom~ .cll.e.m....., 1967, 1, 449. 20. Acetic acid and la have comparable acidities (22). 21. CpMo (CO) 3 H decomposes upon dissolving in acetoni tr ile. The IR spectrum of the resulting solution contains carbonyl stretching absorptions at 1795(s) cm-l in CH 3 CN. v(CO): 2015(w), 1915(s), Analysis by HPLC confirms the presence of small amounts of 3a and a number of products with short retention times. 22. J. Frommer, R. G. Bergman,~ Am, Chern.~, in press. 112 APPENDIX A X-ray Crystal Structure Data For Z-Ni(acac) (PPh 3 ) [C(Ph)=C(CH 3 )Phl. Contents 1. Figure _l. Ortep Drawing of Full Asymmetric Unit. 113 2. Figure 2_. Schematic of Atom Numbering Scheme. 114 3. Table _l. Inter atomic Distances and Angles. 115 4. Table 2_. Fractional Coordinates. 120 113 Figure I. ORTEP drawing of full asymmetric unit. 114 Figure Schematic of atom nwnbe rin g scheme. a! :Mol ecule A C6 \cs - - 0 2 I " /Pl Cl0--Cll C4 Nil / \ '\_ /'-.......,,cl-C9 Cl2 C3-0l \ I / I C2 C23--- C2 4 ""' C25 I Cl4-Cl3 I C22........_ C26 ,___C21/ I Pl /C34.._C 3 J / ,~ I C36 C38 ""'C2?--C32'-... T 1 C28 ~C3 7,-,/ C30 " C29 Molecule B / C43 \ C42--- 03 / \ C4\l /C4g /P2 /Ni~ C40 04 "-\cl44.,.,,- c4B ct "'----cso I /C51 ."" I .,,,,,..-c62, C61 '-... C52 C39 ct / cs4--cs3 \ C56 I ---C57 "-C46 C60 C64 -......_C59~ \ C58 T I C45 I /C72 P2 C66 C73 "-c71_......... ""'c65/ '--.... L L ~ C75/ I T C70 C6 8 '--....._C69/ 115 Table 1. Interatomic Distances (i) and Angles ( 0 ). Molecule A Distances: Molecule A Distances ( cont) : Nil-Pl 2.1783 Cl9-C20 1.3812 Nil-Cl 1. 8970 C21-C22 1.3843 Nil-01 1.9101 C21-C26 1.3662 Nil-02 1. 9228 C22-C23 1. 3962 Pl-C21 1.8205 C23-C24 1.3668 Pl-C20 1.8287 C24-C25 1.3734 Pl-C33 1. 8197 C25-C26 1.3898 Ol-C3 1.2748 C27-C28 1.3826 02-C5 1.2822 C27-C32 1.3864 Cl-C7 1. 3271 C28-C29 1.4116 Cl-C9 1.4685 C29-C30 1. 3554 C2-C3 1. 5015 C30-C31 1. 3677 C3-C4 1. 3589 C31-C32 1.4037 C4-C5 1.3554 C33-C34 1.3698 C5-C6 1.5054 C33-C38 1. 4025 C7-C8 1.5268 C34-C35 1.3876 C7-15 1.4825 C35-C36 1.3575 C9-Cl0 1.3772 C36-C37 1.3623 C9-Cl4 1.4002 C37-C38 1.4146 ClO-Cll 1.4037 Cll-Cl2 1.3639 Cl2-Cl3 1.3576 Ni2-P2 2.1634 Cl3-C14 1.3964 Ni2-03 1.8970 Cl5-Cl6 1.3823 Ni2-04 1.8948 Cl5-C20 1.4087 Ni2-C44 1.8935 Cl6-Cl7 1.3883 P2-C58 1.8057 Cl7-Cl8 1.3845 P2-C65 1. 8102 Cl8-Cl9 1. 3386 P2-C71 1. 8196 Molecule B Distances: . 116 Table 1. Continued Molecule B Distances ( cont) : Molecu.le B Distances (cont): 03-C42 1.2985 C59-C64 1. 3737 04-C40 1.2770 C60-C61 1.3898 C39-C40 1.5310 C61-C62 1. 3841 C40-C41 1.3432 C62-C63 1.3534 C41-C42 1.3560 C63-C64 1. 3975 C42-C43 1.5317 C65-C66 1.4042 C44-C45 1.3093 C65-C70 1.3764 C44-C47 1.4695 C66-C67 1.3877 C45-C46 1.5546 C67-C68 1.3661 C45-C53 1.4812 C68-C69 1.3649 C47-C48 1.4110 C69-C70 1.3940 C47-C52 1.3556 C71-C72 1. 3762 C48-C49 1.4128 C71-C76 1. 3923 C49-C50 1. 3579 C72-C73 1.4127 C50-C51 1.3536 C73-C74 1. 3559 C51-C52 1.3815 C74-C75 1.3594 C53-C54 1.3960 C75-C76 1.4042 C53-C58 1.3980 C54-C55 1.3905 01 Nil Pl 175.79 C55-C56 1.3441 02 Nil Pl 87.01 C56-C57 1.3591 02 Nil 01 92.90 C57-C58 1.4053 02 Nil Cl 176.22 C59-C60 1.3843 Cl Nil Pl 93.16 Molecule A Angles: 117 Table 1. Continued Molecule A Angles (cont): Molecule A Ang:les (cont): Cl Nil 01 87.20 Cl3 Cl2 Cll 120.00 C21 Pl Nil 109.48 Cl4 Cl3 Cl2 120.86 C27 Pl Nil 122.64 Cl3 Cl4 C9 120.62 C27 Pl C21 103.32 Cl6 Cl5 C7 123.52 C27 Pl C33 102.05 C20 Cl5 C7 119.65 C33 Pl Nil 113.11 C20 Cl5 Cl6 116.83 C33 Pl C21 104.43 Cl7 Cl6 Cl5 121. 56 C3 01 Nil 124.73 Cl8 Cl7 Cl6 118.97 cs 02 Nil 123.94 Cl9 Cl8 Cl7 121.31 C7 Cl Nil 123.95 C20 Cl9 Cl8 119.87 C7 Cl C9 127.93 Cl9 C20 ClS 121.44 C7 Cl Nil 109.71 C22 C21 Pl 117.83 C2 C3 01 114.38 C26 C21 Pl 124.34 C2 C3 C4 120.40 C26 C21 C22 117.83 C4 C3 01 125.19 C23 C22 C21 121.69 cs C4 C3 125.29 C24 C23 C22 119.08 C4 02 125.86 C25 C24 C23 120.02 02 112.95 C26 C25 C24 120.17 C6 cs cs cs C4 121.17 C25 C26 C21 121.20 cs C7 Cl 123.57 C28 C27 Pl 121.89 ClS C7 Cl 122.57 C32 C27 Pl 119.11 ClS C7 cs 113.85 C32 C27 C28 118.99 ClO C9 Cl 122.35 C29 C28 C27 119.80 C14 C9 Cl 120.49 C30 C29 C28 119.47 Cl4 C9 ClO 116. 97 C31 C30 C29 122.46 Cll ClO C9 121.96 C30 C31 C32 118.02 Cl2 Cll ClO 119.58 C6 118 Tahle 1 . Continued Molecule A Angles (cont): Molecule B Angles (cont) : C31 C32 C27 121. 26 C43 C42 03 111. 55 C34 C33 Pl 120.29 C43 C42 C41 124.21 C38 C33 Pl 121.20 C45 C44 C47 124.15 C38 C33 C34 118.51 C45 C44 Ni2 126.18 C35 C34 C33 120.80 C47 C44 Ni2 109.36 C36 C35 C34 121.19 C46 C45 C44 123.77 C37 C36 C35 119.73 C52 C45 C44 122.09 C36 C37 C38 120.33 C53 C45 C46 114.13 C37 C38 C33 119.44 C48 C47 C44 119.38 C52 C47 C44 124.91 C52 C47 C48 115.60 C49 C48 C47 121.03 Molecule B Angles: 03 Ni2 P2 85.81 cso C49 C48 120.08 04 Ni2 P2 176.43 C51 cso C49 119.07 04 Ni2 03 92.88 C52 C51 cso 120.98 C44 Ni2 P2 93.21 CSl C52 C47 123.18 C44 Ni2 03 172.97 C54 C53 C45 120.23 C44 Ni2 04 88.50 C58 C53 C45 122.98 C59 P2 Ni2 112.83 C58 C53 C54 116. 76 C65 P2 Ni2 107.48 C55 C54 C53 120.59 C65 P2 C59 103.42 C56 C55 C54 121. 34 C71 P2 Ni2 124.43 C57 C56 css 120.50 C71 P2 C59 103.13 C58 C57 C56 119. 48 C71 P2 C65 103.38 C57 C58 C53 121.30 C42 03 Ni2 126.15 C60 C59 P2 119.09 C40 04 Ni2 124 . 75 C64 C59 P2 123.03 C39 C40 04 112.84 C41 C40 04 127.35 C41 C40 C39 119.81 C42 C41 C40 124.43 C41 C42 03 124.25 119 Table 1. Continued Molecule B Angles (cont) : C64 C59 C60 117.88 C61 C60 C59 120.68 C62 C61 C60 119.57 C61 C62 C63 120.91 C62 C63 C64 118.65 C63 C64 C59 _1 22. 20 C66 C56 C70 118. 44 C66 C65 P2 117.74 C70 C65 P2 123.76 C67 C66 C65 119.72 C66 C67 C68 120.60 C67 C68 C69 120.48 C68 C69 C70 119.67 C65 C70 C69 121.07 C72 C71 P2 122.15 C76 C71 P2 119.74 C76 C71 C72 118.04 C73 C72 C71 121.C3 C74 C73 C72 119.04 C75 C74 C73 121.84 C76 C75 C74 119.11 C75 C76 C71 120.89 . C6 C7 CB C9 Cl0 Cll Cl2 Cl3 Cl4 ClS Cl6 Cl7 Cl8 Cl9 C20 C21 C22 C23 C24 C25 C26 C27 C28 C29 C30 C31 C32 cs 01 02 Cl C2 C3 C4 Nil Pl Table X y Fractional Coordinates. z (6) (5) (5) (5) (6) (5) (5) (5) (5) (5) (6) (6) (6) (5) ( 5) (5) ( 6) (6) (6) (5) (6) (3) (5) (5) (5) (5) (5) (6) (5) ( 5) (5) (5) (6 ) ( 5) (7) ( 8) (7) ( 7) ( 7) (8) (7) ( 8) ( 7) (7) (8) (8) (9) (7) • (7) (8) (8) (8) ( 8) (7) (7) ( 8) (9) (8) (8) ( 7) (7) (8) ( 9) ( 8) (8) (7) 4840 5542 4399 4568 4318 4'184 4234 4946 4621 6412 7068 7881 8032 7398 6591 4139 3770 2914 2416 2747 3601 4360 3550 . 2822 2913 3712 4430 4695 5304 5239 4577 3961 4026 (3) (5) (6) (5) (5) ( 5) (6) (5) (5) (5) (5) (6) (6) (6) (5) (5) (6) (6) (6) (6) (5) (5) (6) (6) (6) (6) (5) (5) (5) (6) (6) (6) (5) 4.12 (0.12) 3.53 (0.17) 5.58 (0.21) 3.82 (0.18) 4.36 (0.19) 4.15 (0.19) 6.80 (0.24) 3.91 (0.18) 5.46 (0.21) 3.48 (0.17) 4.11 (0.18) 5.41 (0.21) 5.89 (0.22) 6.56 (0.24) 4.99 (0.2D) 3.61 (0.18) 5.35 (0.2D) 6.ll (0.22) 5.62 (0.22) 5.08 (0.2l) 3.95 (0.13) 3.38 (0.17) 5.38 (0. 2 l) 6.33 (0.21) 5.97 (0.22) 5.53 (0.21) 4.51 (0.13) 3.54 (0.17) 4.84 (0.1'3) 5.93 (O.n) 5.81 (0.2:3) 5 .66 co .2:n 4 . 2 8 ( 0 • 1 il) 1572 4389 716 ll04 3 36 56 9 -372 4854 3711 4411 4309 4238 4254 4352 4441 4964 5828 5611 4511 3655 3862 3338 3232 2795 2445 25 34 2976 4452 3986 2748 2030 2457 3687 B 6552 8196 8838 8091 7401 6679 5931 1297 6 75 8233 7553 7 578 8294 8976 8958 8703 8693 8677 8694 8715 8717 6041 6531 6199 5369 4873 5213 3134 3235 2982 26 4 3 2534 2788 z y 389 (7) 389(14) 410 (7) 380(14) --------------------------2166 (5) 4976 (3) 3.93 (0.ll) 526ll(l3) 39544(18) U22 ------- Ull ------- --------8166 ( 3) X 65425 ( 12) ------------------------29967 (9) 73 930 (6) 51551 ( 6) 2. 449 (7) 406(14) U33 ------118 ( 5) 118 (11) Ul2 ------- -ll7 (5) - 57 (ll) Ul3 -------- 5 (5) 65 ( 11) U23 ------ f-J N 0 X (continued) y z B 03 04 C39 C40 C41 C42 C43 C44 C45 C46 C47 C48 C49 C50 C51 C52 C53 C54 C55 C56 C57 C58 C59 C60 6942 6458 7158 7900 8551 9404 6799 3753 4 4 06 6 806 6143 6152 6798 7426 7432 3794 3783 3834 3893 3911 3870 9232 9500 --------------------------( 5}' 7876 (4) 5.07 (0.13) 810 --------8530 (3) X (3) (6} (5) (6) (5) ( 7) (5) (5) (6) (5) (7) ( 8) (6) (6) (5) ( 5) (5) (6) (6) (6) (6) (5) (6) 845 1009 973 1062 998 1141 808 85 -52 2047 2391 35 8 7 4425 4095 29 32 1328 1708 2878 3674 3353 2180 1776 2633 y (5) ( 8) (7) ( 8) (8) (10) (7) ( 7) ( 8) ( 7) (10) (12) ( 9) ( 0) (8} ( 7) ( 7) ( 8) ( 9) (10) (9) (7) ( 8) (19) 8207 7099 7434 6919 7143 6564 9817 -127 - 965 144 114 36 9 664 712 437 2[3 9 1160 1554 1113 263 -154 9733 9244 z (4) (6) (6} (6) (6} (7) (5) (5) (6) (5) (7) ( 8) (6) (6) (5) (5) (5) (6) (7) (7) (6) (5) (6) 94735(13) 4.73 6.63 4.81 4.90 5.20 9.00 4.13 4.26 6.11 4.46 8.14 9.16 6. 66 5.41 4. 92 4 . 31 4.69 5.72 7.10 7.72 6.33 3.61 5.98 379 (7) 313 (13) 695 ( 8) 529(15) U33 ------- U22 ------- (0.12) (0.24) (0.1 9) (0.21) (0.20) (0.30) (0.19) (0 .18) (0.22) (0.19) ( 0. 28) (0.33) ( 0. 24) (0.21) (0. 20) (J .19) (0.19) (0. 22) (0.25) (0.27) (0.24) (0.17) (0.22) B 4 66 (7) 425(14) ull (0.23) (0 .20) (0.18) (0. 20) ------- . 3.91 5.05 5.03 6.11 5 .14 4491 z (5) (6) (5) (6) (6) 84188 (13) 6607 7326 76 33 7237 6505 P2 y (7) ( 8) (8) (8) (8) Ni2 X 2887 2766 3525 4432 4583 ------------------------76322 (6) 7100 (10) 88218 (6) (5) (5) (5) (6) (5) 5528 4903 4445 4604 5241 ----------------------------------6191 (5) 5708 (4) 3789 ( 7) 3.15 (0.16) 2. C33 C34 C35 C36 C37 C38 Table 256 (6) 167 (12) Ul2 ------- -74 (5) -60(11) Ul3 ------- 101- (6) 74 (11) U23 ------- I-' N I-' 10149 (7) 10523 (6) 10248 (6) 9610 (5) 11043 ( 5) 11513 (5) 11126 (5) 10293 (5) 9826 (5) 10205 (5) 1940. (5) 1793 (5) 2064 • (6) 2466 (5) 26 02 (5) 2343 (5) HSl HS2 HS3 HS4 HS 5 !-1S6 HS7 HS 8 HS 9 HSl0 HSll HS12 HS13 HS14 0.116739 0.272330 0.214594 0.22 6 918 0.303181 0.348345 0.230460 0.212312 0.342580 0.499232 0.523070 0.451275 0.240500 0.798219 0.655458 0.513092 0.428094 0.485158 0.712489 0. 5 8 38 76 0.47 8 924 0.400598 0.428158 0.536168 0.349010 0.303754 y ---------0.044699 7.77 7.39 5.82 4.48 3.54 4.31 5.18 4.99 4.50 4.04 3.58 4 . 80 6.00 5.65 5.52 4.34 (0. 1~1) (0.l"i') ( 0. 1 ~') ( 0. 2(1) ( 0. 2(1) ( 0 .15) (0. lE) (0.17) ( 0 . 2 (1) (0. 23) (0.2~) (0. 21) (0 .lS) (O.n) (0. 2'i') (0. 26) B ----------·- 0.633481 0.225769 0,240832 0.376183 0.499767 0.349502 6.638456 0.762955 0. 813126 0.746227 0.621885 0.577305 0.566263 --------0.397262 z 6.37 7.52 6.94 6.6 7 5.69 4.62 4.94 6.28 6.09 7.12 6.14 5.85 7.16 B ----------5.45 THE CONVERSION OF BIJ TO UIJ FOR I.NE.J INCLUDES MULTIPLICATION 3641 (10) 9401 (7) ( 9) 10059 (7) 3794 ( 9) 2995 10571 (6) 1972 (7) 10389 (5) 1271 (5) 547 (7) 1661 (7) 1635 (5) (8) 2446 2219 (5) 2135 ( 8) 2460 (5) (7) 1050 2123 (5) 256 (7) 1527 (5) 154 ( 7) 9561 (5) (7) • 9522 (5) 1158 1595 (9) 8759 (6) 8056 ( 6) 1002 (8) -12 (8)' 8060 (6) -437 8821 (5) (7) --------0.742979 X z --------- --------- y UIJ HAS BEEN MULTIPLIED BY 10**4. BY 1/2. C73 C74 C75 C76 ·en C61 C62 C6J C64 C65 C66 C67 C6 8 C69 C70 C71 --------- X Table 2:.. (continued) I-' Iv Iv y -0 . 004613 0.029512 0.0 84546 0.09 3996 0.177600 0.381625 0.526800 0. 469304 0. 565158 0.571797 0. 680502 0 . 7 88274 HS48 HS49 HS50 HS51 1. 003105 0.875150 0 .7 53146 0.754037 0.883574 0.126341 0.22 9917 0.28 8081 0.245763 0.147128 o:695544 0. 833424 0. 860299 0.751501 0 . 614344 0.413781 0.266222 0.181625 0. 237853 0. 383187 0.149141 0.21 608 0 0.140 64 1 -0.004594 -0.077962 0.192398 0.401579 0.252366 0.423325 0.449220 0 .313754 0.136666 0.156134 0.231999 0.130110 ' -0.041984 -0.116125 -0.050 632 0 . 082959 0.268011 0.323445 0.186921 0.426837 0.417746 0.421 576 0.436978 0.449iJ02 0.661569 0. 622729 • 0 . 435280 0.288284 0.322117 0.114592 0.312739 0.4 49908 0.394423 0 .193885 0.452751 0.347 583 0.359953 0.878 536 0.904219 1. 014875 1. 10 5506 1. 073197 z -------- 0.149159 0.197050 0.265037 0. 288772 0.245073 0.987171 0.9 23 06 0 1.00 2822 1. 145084 1. 210933 0.703386 0.708861 0.831789 o. 948070 0.9 44885 0.870718 0.866718 0.866723 0.873404 0.873112 0 . 375758 0.3 8289 7 0.392984 0 . 393654 0.391351 -------0.117478 HS23 HS24 ·HS25 ES26 !1527 HS28 HS29 HS30 HS31 HS32 H533 HS34 HS35 HS36 HS37 HS38 HS39 HS40 HS41 HS42 HS43 HS44 HS45 HS46 HS47 -------- X (continued) 0.246607 0.229334 0.271174 0.92 2 081 1. 032929 1. 099657 1. 048945 0.942588 2. HS15 HS16 1-1S17 HS18 HS19 HS20 HS21 HS22 Table 9.22 10.62 7.66 6. 56 5.87 7.14 6.77 6.56 5.54 5.21 5 .57 , 6.24 6.09 5.47 5.21 6.63 7.15 7.56 5.85 6.29 7.12 6.78 6. 28 • 5.18 5.80 6.70 7.95 8.85 7.57 6.98 6.86 5.31 6.82 8.76 8.41 6.10 5.75 B w f.-J N 0.791604 0.9300 0.8919 0.87673 0.4918 0.43124 0.4360 0.66362 0.605 0.62027 0.9805 0.94076 0.947 0.082 0.01221 0.0727 0.6070 0.55241 0.570 38 HS58 !-IS59 HS60 HS61 !-IS62 HS63 HS64 HS65 HS66 HS67 HS68 HS69 HS70 -------- X y 0.4126 0.4939 0.40034 -0. 0"/95 -0.12202 -0.1 35 0.65578 0.705 0.74934 0.6795 0.61245 0.628 0.4620 0.48757 0.401 0.3875 0.47286 0.38913 0.034 0.1195 4 0.0118 0.15787 0.1 838 0.04486 0.181 0.302 0.37574 0.3643 -0.1062 -0.0 6010 -0.01153 z -------0.045076 0.1274 0.07091 -0. 00447 0.0092 -0.08886 -------0.275121 (continued) HS52 HS53 HS54 HS55 HS56 HS57 Table 2. 7.64 7.68 7. 68 7.68 9.93 • 9. 93 9.93 6.36 6 . 36 6.36 7.05 7.05 7.05 6.14 6.59 6.59 6.59 7.64 7.64 B f-' N .i::,. 125 APPENDIX B Analysis of the Mass Spectral Data 126 Molybdenum has seven isotopes in significant natural abundance, distributed over a mass range of 9 units, resulting in a mass spectrum for natural abundance [CpMo(C0) 2 J 2 (4a) containing parent ions at 18 different m/e values ranging from 426 to 443 (Figure 3, Chapter II). The relative intensities of these ions under the conditions employed in this study (50 ev, 150°C inlet temperature) are summarized in Table 1. The observed parent ion intensities for the 13 co labeled 4a for reactions 1, 2, and 3 (Table 2, Chapter II) is also presented in Table 1. In all cases these relative intensities are the averaged result of at least 4 measurements. This data were analyzed by a least squares fit to the equation; Im= am x c 0 + a(m-l) x c 1 + a(m- 2 ) x c 2 + a(m- 3 ) x c 3 + a(m-4) x C4 Where; Im= observed relative intensity at mass m. am= observed normalized intensity of natural abundance 4a at mass m. en= calculated coeficient, the relative contribution of [CpMoc 13 co>n< 12 co> 4 _nJ 2 to the observed mass spectrum. The values of C0 , c 1 , c 2 , c 3 , and c 4 calculated in this manner were then converted into percentages to give the data in Table 2, Chapter II. This analysis ignores changes in the (m + 1) contribution to these relative intensitities as a result of the 127 replacement of natural abundance CO by labeled 13 co. This calculation was accomplished using the least squares regression routine available at the u. C. Berkeley Computer Center as part of the Statistical Package for the Social Sciences program (SPSS). 1 The statistical error in these calculations was quite good(~ 1.5%). The sensitivity to throwing out weak data was within these error limits. With 20 data points (m/e = 427 - 446) and only five variables the system is highly overdetermined. A similar analysis was used in evaluation of the mass spectral data for 2e. It is expected that errors in the mass spectral data collection, and through differences in reaction conditions highly outweigh the error inherent in this analysis of the mass spectra. Reference 1. Nie, N. H., Hull, C. H., Jenkins, J. G., Steinbrenner, K., Bent D. H., "SPSS-Statistical Package for the Social Sciences", Second edition, McGraw-Hill, New York, N.Y., 1975, pg 320-367. 128 Table 1. The Observed Relative intensities of the Parent Ions in The Mass Spectra for 4a. Reaction# a rn/e nat. ab. b 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 0.1460 0.0198 0.1870 0.3518 0.4471 0.4699 0.9506 0.6304 1.0 o. 9101 0.8834 0.6581 0.7184 0.2399 0.3533 0.0527 0.0736 0.0100 - - 1 2 - 392 929 1628 2712 4664 6624 8400 10352 12032 13248 13376 12704 11248 9024 6720 4368 2664 1110 418 - 3 - 379 1086 1850 3056 5144 7840 10576 13248 16000 17280 18528 18432 16544 14096 10832 7224 4480 2392 1032 471 206 - 918 2644 4600 7232 12544 19232 25888 33152 40128 44480 46912 47296 42880 36480 28192 19392 11904 6472 2700 1156 - aSee Table 2, Chapter II. bnatural abundance, normalized to 1 434 = l.O 129 PART III THE PROPOSITIONS Abstracts A. A Stereoselective Synthesis Mediated by a Chiral Transition Metal Complex. A series of selective reactions of organic ligands bound to a chiral transition metal complex, where the metal center is chiral, will be employed to effect the overall selective synthesis of chiral organic products. This stereoselective synthesis involves the use of an optically active phosphine ligand to allow the separation of diastereomeric mixtures of organometallic compounds, and the ability of an optically active transition metal complex to induce stereoselective reactions of organic ligands. B. Nucleophilicity and Transition Metal Carbonyl Anions: A Study Of the Relationship between Ion-Pairing and the Competition between Sn2 Displacement and Electron Transfer Mechanisms. The ion-pairing equilibria of CpFe(C0) 2 -Na+ (1) will be studied by infrared and conductivity methods. Knowledge of the ion-pairing equilibrium position under a number of conditions · will be employed to analyze the relative rates of Sn2 and electron transfer processes in the reaction of 1 with alkyl halides. In particular cyclopr opylcarbinyl halides will be used to evaluate the contribution of electron transfer processes to 130 the observed rate. In the second part of this proposal the relative rates of the reaction of a number of metal anions with alkyl halides under "pure" Sn2 conditions, involving only free ions, will be re-evaluated. C. The Synthesis and Reactivity of a Novel c,--Alkyl-µ-Alkylidene Iron Complex. A Model for the Carbon-Carbon Bond Forming Process in Fischer-Tropsch catalysis? Two novel synthetic approaches to the preparation ofµalkylidene complexes are proposed. These methods will be investigated as routes to the CT-alkyl-µ-alkylidene complexes Cp(CO)Fe<µ-cH 2 > cµ.-cO)Fe(R) Cco> 3 , and their vinylidene analogues. Once prepared, the reactions of these complexes will be studied. In particular reactions that lead to carbon-carbon bond formation are sought. D. The Absolute Rate Constant for Cis-Trans Isomerization of Vinyl Radicals. It is proposed to measure the absolute rate constant for the cis-trans isomerization of vinyl radicals by using a steady state analysis on the reduction of vinyl fluorides by sodium naphthalenide. The rate of reduction of the vinyl radicals to stable vinyl carbanions is assumed to be 1.6 x 10 9 M- 1 sec- 1 , the value obtained for the reduction of alkyl radicals to carbanions by sodium naphthalenide. 131 E. The Study of Cationic Bisphosphine Methylnickel (II) and Hethylplatinum(II) Complexes: The Search for Evidence of an aHydride Migration Process. It is proposed to study the reactions of cationic rnethylnickel and rnethylplatinurn complexes of the type [ (L) 2 M(CH 3 ) (acetone) l + PF 6 - with phosphines and alkenes. These reactions will be analyzed in an effort to gain evidence for the intermediacy of an a-hydride migration process. 132 A STEREOSELECTIVE SYNTHESIS MEDIATED BY A CHIRAL TRANSITIONMETAL COMPLEX INTRODUCTION In organic synthesis the selective construction of asymmetric centers remains one of the most difficult challenges for the synthetic chemist. The use of organotransition metal reagents in stereoselective syntheses has \ been growing rapidly in recent years. 1 To date, however, most of these efforts have centered around the asymmetric induction provided by chiral ligands, in particular chiral bisphosphine ligands. Recently Brunner and his coworkers 2 have studied both the synthesis and properties of a wide variety of asymmetric transition metal complexes, where the chiral center is at the metal atom. Brunner has proposed that the use of such chiral transition metal complexes in asymmetric synthesis can result in high optical yields due to the ability of the chiral metal atom to closely approach the reactive center when the substrate is bound to the metal complex. 2 a An asymmetric synthesis of this kind involving a chiral transition metal complex is proposed here. The complexes synthesis and reactivity of the cationic alkene [CpFe(C0) 2 (alkene)l+ has been extensively studied and the use of these organoiron complexes in organic synthesis has been reviewed recently. 3 , 4 These complexes have been shown to undergo a number of highly selective reactions of the organic 133 ligand. For example it was found that protonation of the <T-allyl complex CpFe (CO) 2 [ CH 2 C (H) =CH 2 J (la) with dry DCl occurs at the 3 position to give exclusively [CpFe(C0) 2 CcH 2 =C(H)(CH 2 D)]+ (2a). Furthermore reduction of the 2a with a number of hydride reagents can give either CpFe(C0) 2 [CH(CH 3 ) 2 J (3a) or CpFe(C0) 2 (CH 2 CH 2 CH 3 ) (4a). Some reagents have been found that reduce 2a selectively to 3a. 3 -S In this later reaction a new asymmetric center is generated at the carbon atom bound to iron. The stereochemistry of this carbon center is determined by which face of the alkene is bound to iron in 2a, and is independent of whether hydride attack occurs exo or endo to the metal center. Thus the stereochemistry of the carbon center in 3a is predetermined in the enantiomeric cationic complex 2a. These reactions and the stereochemistry of 2a are shown in Scheme I. If one of the carbonyl ligands in the complexes 1, 2, 3, or 4 is substituted by a phosphine ligand then the metal center becomes chiral. Brunner has prepared an analogue of la CpFe(CO)(PR 3 *>cH 3 using the optically active phosphine ligand (S)-PPh 2 [N(CH 3 )-CH(CH 3 ) (Ph)], diastereomeric isomers, PPh 2 R*. 2 , 6 This complex has two related by inversion of stereochemistry at the metal center, which can be separated by high pressure liquid chromatography. A similar approach, substituting the complexes 1, and 2 with an optically active phosphine, will be employed in this proposal. 134 SCHEME I la /"-" ~ H- I ~~Fe,,,/. 0 c·· j i,,•• C0 CH H c/ CH R 3 2 ~ '+ "',_,Fe,,,, o c··· , '···co ~ 3a 2a ~ ,r,..../ CH R 2 H+ 135 THE PROPOSAL It is proposed to [ CpFe (CO) (PPh 2 R*> (CH 2 =CHR)] + synthesize (2) the chiral complexes in high optical purity. Then the reduction of these complexes to the the secondary alkyliron complexes CpFe(CO) (PPh 2 R* ) [CH(CH 3 ) (R)] (3), will be investigated. The subsequent convertion of 3b into organic products without loss of optically activity will result in the overall stereoselective synthesis of a chiral secondary center from achiral alkenes using stoichiometric react1ons of a transition metal complex. Two somewhat distinct approaches to the preparation of the required optically active alkene complexes 2b will be investigated. A. The first approach involves the substitution of the complexes la, which are readily prepared from CpFe(Co> 2 Na and allylic halides, with the optically active phosphine PPh 2 R* to give a mixture of two diastereomers. These diastereomers will then be separated by high pressure liquid chromatography in a manner analogous to that employed by Brunner and coworkers. 5 Protonation of the resolved diastereomers should give optically active 2b (Scheme II). The optical purity of 2b will depend upon both the optical purity of lb and the amount of asymmetric induction in the protonation step. Two suggested conformations of the 0--allyl ligand in the transition state for protonation, leading to different stereochemistry, are shown in Scheme III. These conformers are related as object and mirror image, and since the geometry about the metal center is chiral it can 136 SCHEME IT hv separate 137 SCHEME ill or 138 reasonably be expected that one of these transition states will be of lower energy than the other. Previous studies found that deprotonation of cylic alkene complexes to give analogues of la occurs only if the proton can obtain a configuration trans to iron, supporting the hypothesis that the transition states for both protonation and deprotonation are highly organized. 4 Separation of the optically active complexes lb should be aided by asymmetric induction in the phosphine substitution step, where the influence of the optically active center in the phosphine ligand results in the predominance of one isomer. An effect of this kind was observed by Brunner and coworkers. 4 Loss of the organic substrate can be minimized by epimerization of the unwanted diastereomer followed by repetition of the separation proceedure. For example it has been shown that the optically active complex CpFe(CO)(PPh 2 R*)cH 3 looses most of its optical activity upon heating to 96°C for 10 hours. 6 B. An alternative approach involves the separation of the diastereomeric alkene complexes 2b. Using an optically active phosphine four possible diastereomers can result (Scheme IV). These alkene complexes can be prepared by a number of methods; a) protonation of the displacement O"-allyl complexes lb (approach A), of isobutene [CpFe(CO) CL) (isobutene)]+, and c) from the b) complex by nucleophilic attack of [CpFe (CO) CL)]- on epoxides followed by protona tion (Scheme IV). In this last case retention of the stereochemistry in the epoxide is observed. 3 ' 4 l39 SCHEME N CpFe(CO)(IJ(=<r + C /V 140 Separation of this diasteromeric mixture of complexes by high pressure liquid chromatography could prove difficult considering the number of isomers. However it should be noted that the isomers a & c and b & d have the olefin coordinated on the same side, and will therefore give the same stereochemistry at carbon upon reduction. Thus only separation of a & c from b & d becomes necessary. Treatment of the undesired isomers with Na! or r 2 in acetone allows the recovery of the alkene substrate in good yield. 4b Green and Nagy 3 [CpFe(CO) 2 (CH 2 =CH(CH 3 ))]+ reported that the reduction of with NaBH 4 gave exclusively CpFe (CO) 2 (CH (CH 3 ) 2 (3a) and no CpFe (CO) 2 (CH 2 CH 2 CH 3 ) (4a). However Rosenblum and coworkers 4 reported that this same reduction gives (2a) a 3a/4a ratio of 1:3. Reduction with NaBH 3CN, 7 CpFe(CO) (PPh 3 )H, 5 and HFe(CO) 4 - NEt 4 + have been reported to give exclusively 3a. This reduction is strongly effected by the presence of electron withdrawing groups on the alkene, causing attack of hydride to occur at the substituted carbon. Likewise both carbon and heteroatom nucleophiles have been observed to add preferentially at the substituted alkene carbon. For the purposes of this proposal the selective reduction of the cationic alkene complexes to secondary alkyliron complexes is desired. An investigation of the selectivity of this reaction as a function of reducing agent and functionality on the alkene will be required. The optically active alkyliron complexes 3b can then be converted to a variety of useful organic functionalities 141 without loss of optical activity (Scheme V). 8 For example corbonyla tion in methanol gives the corresponding methyl ester with retention of conf iguration, 8 and bromination in cs 2 gives the corresponding bromide with inversion of stereochemistry at carbon. 819 DISCUSSION This proposal involves the stereoselective conversion of either a terminal alkene or allylic halide into optically active secondary esters or halides using stoichiometric organometallic reactions (Scheme VI). This goal is achieved by taking advantage of the ability of an asymmetric transition metal center to induce stereoselectivity in reactions that occur at either a-- or 17"-bound ligands, and through the use of an optically active phosphine ligand to produce diastereomeric organometallic complexes which can be separated. Two distinct approaches are proposed. The first approach described here involves the synthesis of the diastereomeric a-allyl complexes lb from allylic halides, CpFe(Co) 2 Na, and an optically active phophine, PPh 2 R* . Separation of these diastereomers by high pressure liquid chromatography, followed by protonation, gives the optically active alkene complex 2b. Selective reduction of this complex to a secondary alkyliron complex then generats a new chiral center at carbon. The second approach involves the synthesis of mixtures of diastereomeric alkene complexes 2b, followed by separation of the resulting four 0 I~ R .:::,__,_ II /c-c" MeO CH 3 R • I [o], Me OH ~ 3b I 3 _, CH ~ I H L-•11 Fe ./ R ~ 'C~ * oc H -s:--R c1 , CHCl 2 - - - -3 c-c ....- - = c( 'cH \ 3 SCHEME V cs 2 Br2 3 CH I 3 R'--t-I Hl, CH ~ ~ H R ~'--C-Br ► I f-' N ~ 143 SCHEME "fil • * R~ H X ====~ ► I* R-C H-C-Y 2 I CH 3 H '* Y R- C- =====~ ► I CH 011 3 0 II X = CI, Br, I ; Y = I, Br ,-C-0 R,-C-C I 144 diastereomers .. Again alkyliron complexes. reduction will give optically active This second approach should give optically pure products, provided the diastereomers of 2b can be cleanly separated, and selective reduction of 2b can be achieved. The protonation of lb, however, is expected to yield variable amounts of asymmetric induction, depending upon the nature of the substituent R, and the phosphine ligand, resulting in products of variable optical purity. In all cases the organic substrate that ends up in metal complexes with undesirable stereochemistry can be recovered and used again to produce more product. The beauty of this synthesis is that the separation of diastereomers occurs after the stereochemistry in the product has been determined, i.e. in the alkene complex, but prior to product formation. And all of the steps including the reduction are reversible, 3 , 4 allowing recovery of the organic substrate. In conclusion it is proposed to use a series of stoichiometric reactions of asymmetric transition metal complexes to effect the overall conversion of achiral organic substrates into chiral organic products in high optical yield. This stereoselective transformation employes the use of an optically active phosphine ligand to allow separation of diastereomeric mixtures of organometallic compounds, and the ability of an optically active transition metal center to induce stereoselective reactions of organic ligands. This proposal is unique in that it is one of the first examples of the use of an asymmetric transition metal complex, where the metal center is 145 chiral, to effect a stereoselective transformation of an organic substrate. 146 REFERENCES 1. a) D. Valentine, Jr., J. W. Scott, Synthesis, 1978, .2., 329. b) A. J. Birch, D. H. Williamson,~ React., 1976, ll, 74. c) J. D. Morrison, w. F. Masler, M. K. Neuberg, 1976, ~ ~, .2..2., 81. d) H. B. Kagan, Pure Appl. ~.m....., 1975, 43, 401. e) w. s. Knowles, M. J. Sabacky, B. D. Vineyard,~ Chem.....~, 1974, .l.3..2., 274. f) F. J. McQuillin, Prog. Q!:£L.. ~.m....., 314. 1973, Ji, g) R. E. Harmon, s. K. Gusta, Brown, ~.m..... Revu 1973, TI, 21. h) L. Marko, B. Heil, .Rfil[._, 1973, ~, 269. i) E. I. Klabunovski, E. D. J. ~ s. Levitina, Russ. ~.m..... Rev,, 1970, l l , 1035. j) K. K. Babievskii, V. K. Latov, Russ, ~.m..... .Rfil[._, 1969, ll, 456. 2. a) H. Brunner,~ ~.m..... ~ , 1979, 12, 250; Angeli£. ~.m..... .l.11.h Ed, Engl., 1971, l.Q, 249; Ann.,_ .N.a.X.... Acad. ~ , 1974, 2.1..2., 213. b) H. Brunner in "Organic Chemistry of Iron", Vol 1, _E. A. Koerner von Gustorf, F.-w. Grevels, I. Fischler, Ed. Acdadernic Press, 1978, p. 299. 3. M. L. H. Green, P. L. r. Nagy, ih. Organorn~ ~.m....., 1963, l, 58; ih. ~.m..... ~ , 1963, 189. 4. a) M. Rosenblum,~ ~.m..... ~ , 1974, 2, 122. b) w. P. Giering, M. Rosenblum, ih.~.m.....~.c..h..e..m.....,CQ.m.m....., 1971, 441. c) A. M. Rosan, M. Rosenblum, J. Tancrede, .S.Q~, 1973, .2..2., 3062; w. P. Giering, M. ih. A.m..... ~.m..... Rosenblum. J. Trancrede, .J....t.. Arn, Chern,~, 1972, 94, 7170. 5. T. Bodnar, s. L. La Croce, A. R. Cutler, ih. Am..... .C.hg.m..... ~ , 147 1980, 1.Q.2., 3292. 6. 7. a) H. Brunner, M. Muschiol, w. Nowak, L Naturf .. 1978, J...l,Q, 407. b) H. Brunner, H. Vogt, L Naturf., 1978, 1..1,Q, 1231. s. M. Florio, K-M. Nicholas,~ Organom~~~, 1976, ll2., Cl7. 8. a) P. L. Bock, D. J. Boschette, J. • R. Rasmussen, J. P. Demers, G. M. Whitesides, .J..... A~~~~, 1974, ll, 2814. b) G. W. Whitesides, D. J. Boschette, ibid., 1971, il, 1529. 148 NUCLEOPHILICITY AND TRANSITION METAL CARBONYL ANIONS: A STUDY OF THE RELATIONSHIP BETWEEN ION-PAIRING AND THE COMPETITION BETWEEN Sn2 DISPLACEMENT AND ELECTRON TRANSFER MECHANISMS. Introduction The reaction of transition metal anions with alkyl halides or tosylates is one of the major synthetic routes for preparation of transition metal alkyl complexes. 1 the The similarity between this reaction and ordinary organic displacement reactions has led to a number of attempts at a comparison of transition metal anions with tranditional organic and main group nucleophiles. 2 , 3 The conclusion that these reactions proceed by an Sn2 displacement mechanism is supported by a number of observations; a) the reaction invariably follows bimolecular kinetics, b) the relative rates of reaction normally follow the order 1° > 2°> 3° for the carbon center, and I > Br s OTs > Cl for the leaving group, and finally c) the reacton of [CpFe(co> 2 1-M+ ( M -- Na + , Li+) with a diastereomeric primary bronsylate or bromide, 4 and the reaction of Na 2Fe(C0) 4 with an optically active 2-octyl tosylate, 5 were observed to result in inversion at carbon. The observation of inversion of stereochemistry at carbon is usually considered conclusive proof for an Sn2 displacement mechanism. The generality of this conclusion for all substrates and metal anions has become suspect recently, as the facility of electron transfer processes in organometallic reactions has 149 become increasingly clear. 6 For example the related reaction, the oxidative addition of alkyl halides to transition metals has been shown to occur by electron transfer pathways in some cases. 7 However more convincingly the reaction of [CpFe(Co> 2 -JNa+ with cyclopropylcarbinyl iodide in THF was observed to give about 30% of the ring opened product. 8 In this case ring opening would be expected to arise from a unimolecular rearrangement of a free cyclopropylcarbinyl radical, mechanism. These two indicating an electron transfer mechanisms, Sn2 displacement and dissociative electron transfer, are presented in Scheme I. Two studies have attempted to evaluate the relative nucleophilicity of transition metal anions by examining their relative rates of reaction with Mel or MeOTs. 213 An early study by Dessy, Pohl, and King 2 found the relative reactivity of metal anions with Mel in DME to follow the order: CpFe(Co> 2- > Re(Co) 5- > CpW(C0) 3 - > Mn(C0) 5 - > CpMo(C0) 3 - > CpCr(C0) 3 - > Co(C0) 4 -. More recently Pearson and Figdore 3 have studied the rates of the reaction of a number of metal nucleophiles with Mel and MeOTs in THF. Whereas CpCr(C0) 3 - was found to react with Mel faster than CpMo(C0) 3 - and CpW(C0) 3 -, and likewise CpFe(Co> 2 - faster than CpRu(C0) 2 -, Mn(C0) 5 - was found to react an order of magnitude .§..l..Q~tl than Re(C0) 5 -. To further confuse things the entropy of activation for the reaction of MeOTs with CpCr(C0) 3 - was large and negative CpMo(C0) 3 - (-20.2 e.u.), small but still negative (-2.5 e.u.), but positive for CpW(C0) 3 - The explanation for these results presented by for .PA (11.3 e.u.). the authors l50 SCHEME I Sn2 H M- + R~ / C-X ~ R' Electron Transfer M- + R-X M-R l 151 evaluated the MeI/MeOTs ratio in terms of hard and soft nucleophiles and leaving groups. However the observation of highly variable entropies and enthalpies of activation is a serious warning sign that strong solvent influences effecting the ground and transition states differently are purturbing these systems. 9 These results are both confusing and contradictory, making conclusions about relative nucleophil ici ty of the metal anions by the authors dubious at best. To a large extent these studies of the reaction of transition metal anions with alkyl halides have ignored, or else dealt inadequately with, solvent effects and the influence of ion-pairing on both the rate of reaction and the competition between electron transfer and Sn2 displacement mechanisms, with two exceptions. lO,ll The ion-pairing of transition metal anions with alkali metal cations has been extensively studied for [Co(C0) 4 J-, 12 [Fe(C0) 4 J- 2 ,lO [RFe(C0) 4 )-,lO,l 3 [CpFe(C0) 2 J-, 14 and [Mn(C0) 4 Ll- (L = PMe 2 Ph, PPh 3 , P(OPh) 3 , CO).ll And in all cases both tight and loose, or solvent separated ion pairs have been identified. Most significantly, however, all these ions are substantially if not entirely tight ion paired in THF solution, the solvent in which almost all of the kinetic and mechanistic data available was obtained. An exhaustive study of the reaction of Na 2 Fe(Co) 4 with alkyl halides and tosylates has shown that the rate of Sn2 displacement is highly dependent upon the nature of the ion-pairing1 tight ion pairs are almost competely unreactive, and the mono anion appears 152 to react somewhat faster than the free di anion in NMP solution.S,lO In this system no evidence for the participation of electron transfer pathways was observed. In a second study the rate of the reaction of NaMn(CO) 5 with PhCH 2 c1 in THF was found to be inhibited upon addition expectations. 11 These results of HMPA, lead to contrary to serious questions concerning the relaibility of existing relative rate comparisons. Without knowledge of the position of the tight vs sol vent separated ion pair equilibrum, and the relative rates of displacement for each of these ions in a given solvent, comparisons are meaningless. Furthermore, other than in the specific cases where the stereochemistry has been determined, the existence of at least a partial contribution to the rates from electron transfer processes cannot be ruled out. In fact in the one system that has been studied for a number of leaving groups, the CpFe (CO) 2- anion, the reaction was found to proceed by Sn2 displacement for primary tosylates or bromides, 4 but for primary iodides electron transfer processes were identified. 8 The Proposal It is the intention of this proposal to carefully examine the relative rates of Sn2 displacement and electron transfer reactions for [CpFe (CO) 2 1-M+ as a function of solvent, ion- pairing equilibria, and added sequestering agent. Then with this knowledge in hand the relative rates of a series of metal anions with alkyl halides and tosylates under identical conditions, 153 where an Sn2 pathway is assured, will he re-examined. Two important issues need to be addressed in this study. First it is neccessary to determine the position of the tight vs loose or solvent separated ion-pair equilibrium as a function of solvent, counter-ion, and added sequestering agent. Secondly it will be neccessary to determine the relative rates of Sn2 displacement vs electron transfer pathways as a function of the structure of the ion pair, the counter ion, the solvent, and the leaving group. Only then will it be possible to examine the relative rates of the reaction of a number of transition metal anions with alkyl halides and tosylates, where only free ions and non-coordinating counter ions are involved, such that the relative nucleophilicities of these anions in solution can be evaluated. The cyclopentadienyl (dicarbonyl) iron anion (1) in THF exhibits a complex set of infrared absoptions at 1877, 1862, 1806, 1786, 1770 cm- 1 . Parnell and Jackson were able to show that three distinct species exist in THF. 14 The proposed structures for these species are presented in Figure 1. The solvent separated ion C where the counter ion is R4 N+ has absorptons at 1865, 1788 cm- 1 . The weakly associated ion pair Bis believed to have a strong interaction between the counter ion and one carbonyl, thus the strongly shifted absorption at 1770 cm- 1 . Finally the contact ion pair A has absorptions at 1877, 1806 cm1 . As expected the salts of the non-coordinating counter ion NR + 4 exists exclusively as the solvent separated ion pair C, and s i mi 1 a r 1 y addition of HM PA to a TH F so 1 u ti on of the sodium s a 1 t 154 SCHEME .II ~ K_ ~ ~ Kl Fe ~ Fe + .,,,,, Fe,//~,.. ~ /\ ~ /\ co· j ·co CO CO·~Na CO CO Na A B C 155 of 1 results in formation of the ion pair C. The first part of this proposal will involve the quantitative evaluation of the equilibrium between structures A, B, and C for a variety of solvents and counter ions. Solvents such as tetrahydrof uran (THF), N-methylpyr rolidinone (NMP), hexamethylphosphoric triamide (HMPA) will be used, along with the counter ions Li+, Na+, K+, and the non-coordinating bistriphenylphosphine imminium cation (PPN+). An accurate evaluation of K2 , the equilibrium between C and A+ B can be obtained from a conductimetric titration of 1 in THF or NMP using a crytand such as Krytofix 222 as titrant. Krytofix 222 has a Na+ complexation constant of 10 8 M-l im MeOH. Provided a sharp breaking point at one molar equivalent added cryptand is observed, and the presence of triple ions can be avoided, then after correcting for mobility differences; K2 = a= A= A= a 1 <1 - a > degree of dissociation = A; A0 conductance at zero added cryptand conductance at one equivalent added cryptand 0 An approximate value of the equilibrium conatant K1 , the equilibrium between A and B, can be obtained in the following manner. In THF with one equivalent added crown ether only the species Band C exists in solution. 14 Evaluation of this equilibrium constant by conductimetry in a manner Ldentical to that just described gives K1 under these conditions. Then knowing the ratio of B/C the extinction coeficients for the absorptions for the infrared absorptons at 1745, and 1785 cm-l attributable 156 to these ions can be determined. Furthermore in ether or ether/THF it may be possible to evaluate the extinction coeficients for the IR absorptions for structure A independently. These extinction coeficients will then allow the equilibrium constant K1 to be measured by IR, assuming Beer's law behavior. The equilibrium constant K1 evaluated in this manner is bound to be less accurate than K2 determined by conductimetry due to assumptions concerning extinction coeficients and Beer's law behavior. However since the solvent separated, or free ions are the species that will be of the most interest here this situation will not represent a large problem,~ infra. In the second part of this proposal an attempt will be made to evaluate the ratio of Sn2 vs electron transfer processes using the a cyclopropylcarbinyl halides or tosylates under conditions where 1 exists either as solvent separated, or free ions, structure c, or as entirely "tight" ion pairs, structure A or B. First using the cyclopropylcarbinyl tosylate the rate of Sn2 displacement can be evaluated without interference from electron transfer processes. Alkyl tosylates have not been observed to undergo dissociative electron transfer reactions involving the carbon-oxygen bond. If as expected the rate of displacement for the solvent separated ion pair is faster than for tight ion pairs, then evaluation of the rate of displacement for the tight ion pair may be complicated by minor amounts of free ions present even when K1 is very small. The work of San Filippe 8 suggests that in the reaction of 157 [CpFe(C0) 2 1-Na+ with cyclopropylcarbinyl iodides in THF at least 30% of the reaction pathway involves electron transfer. Whereas the reaction of the corresponding bromide was observed to involve less than 5% of an electron transfer pathway. leaving group dependence upon the This kind of competition between nucleophilic and electron transfer processes has been observed for the reactions of main group anions. 8 , 15 Re-evaluation of the ratio of ring opened to ring closed product under solvent and counter ion conditions such that only tight or free ion pairs (A, and B, or C) exist in solution will allow the relative rates of electron transfer and Sn2 displacement for each of the ions to be determined separetly. The best way to evaluate these ratios for each of the limiting cases is to determine the ratio of cycl ized/uncycl ized product for various concentrations of added crytpand in a solvent such as THF, where in the absence of a sequestering agent the ions exist mostly as tight ion pairs. Then ploting the observed ratio vs K2 at that concentration of cryptand and extrapolating to K2 = O will give the ratio in the total absence of free ions. This same ratio will be determined under different solvent and counter ion conditions, with and without added crytand to dissociate the ion pairs. The rate of Sn2 displacement is expected to be highly dependent upon the leaving group, the solvent, and nature of the ion pair involved. In particular the results with Na2Fe(C0)4 suggests that solvent separated ion pairs should be more reactive than more tightly bound ion pairs in Sn2 reactions. On the other 158 hand no clear prediction can be made for the rate of electron transfer processes from different ion pair structures. Furthermore determining the rate of electron transfer from each of the ion pairs A, B, and C will be considerably more difficult than for the Sn2 pathway. However if the absolute rate of "pure" Sn2 displacement for each ion structure is known, from the alkyl tosylate studies, and a correction for the change in leaving group can be made, then, using the ratio of cyclized to uncylized product from the reactions of cyclopropylcarbinyl iodide, the approximate rate constants for electron transfer for each limiting structure can be calculated. Repeating this calculation for a series of counter ions should show a stronger dependence for the rate of electron transfer from A or B than for Casa function of counter ion. The actual situation is in fact even more complicated. the analysis just described assumes that reaction of a metal anion with a cyclopropylcarbinyl halide by an electron transfer process will always lead to ring opening. However this conclusion requires that the organic radical formed upon disociation of the halide ion a) escapes the solvent cage where electron transfer has taken place, and b) rearranges faster than it proceeds to give product. In an analogous reaction of CpV(Co) 3 H- with alkyl halides a dissociative electron transfer mechanism was identified, and trapping of alkyl radicals was found to be very fast Ck 25 = 2 x 10 7 sec- 1 ), competitive with the rate of ring opening for the cyclopropylcarbinyl radicai. 16 Some of these 159 problems can be avoided by using the diastereomeric halides employed by Whitesides, 4 and studying the stereochemistry of displacement. The third part of this proposal involves the evaluation of relative rates of displacement for variety of transition metal anions, a given substrate and a under conditions of solvent and counter ion where an Sn2 mechanism and free ions is virtually assured. Using a non-coordinating counter ion, such as PPN+, and a polar solvent such as NMP or HMPA only totally dissociated free ions will be involved. Thus the influence of tight vs loose ionpair equilibria upon the rates of reaction can be eliminated. Using cyclopropylcarbinyl halides the participation of electron transfer processes can be evaluated. Then comparisons of the relative rates of displacement, under pseudo first-order conditions, by a Sn2 pathway will allow the evaluation of the relative anions. nucleophilicities of a number of transition metal The anions to be studied will be [ Co (CO) 4 1- , [Fe(CO) 4 ]- 2 , [Rh (CO) 4 1 - , [RFe(CO) 4 ]-, [CpFe(L) 2 ]-, [CpRu(CO) 2 l-, [Mn(CO) 5 ]-, [Re(CO) 5 ]-, [CpCr(CO) 3 ]-, [CpMo(CO) 3 ]-, [CpW(CO) 3 ]-, (L = PPh 3 , dipos, CO). 17 The interesting comparisons within this series will be between; isostructural anions of the first, second, and third rows, and between isoelectronic anions of the first row. Furthermore a comparison of the effect of substituting carbonyls with phophines for a given metal will be examined for the anion [CpFe(L) 2 l-. Only two studies of the reactivity of transition metal 160 anions with alkyl halides have carefully considered the effects of ionLpairing upon the rates of reaction. In both cases solvent, and substrate were chosen so that only an Sn2 process needed consideration. In the present study the effect of ion-pairing upon the competition between dissociative electron transfer and Sn2 displacement mechanisms is evaluated. complementary, These studies are and together define the "boundary" conditions within which Sn2 processes are assured. Under these purely Sn2 conditions with the influence of ion-pairing and the nature of the counter ion virtually eliminated the relative rates of a variety of transition metal anions with alkyl halides will be reexamined. Only a comparison of the reactivity of free metal anions a by common mechanism in the same solvent will be meaningful in evaluating the relative nucleophilicity of the metal centers involved. 161 References ~ ~.!!h. ~ , 197 0, .l, 41 7. 1. R. B. King, 2. R. E. Dessy, R. Pohl, R. B. King, .J..... A.!!h. ~.Ilh. ~ , 1966, 88, 5121. 3. R. G. Pearson, P. E. Figdore, ibid,, 1980, J..Q2., 1541. 4. P. L. Bock, D. J. Boschette, J. R. Rasmussen, J. P. Demers, G. M. Whitesides, ibid,, 1974, ll, 2814; P. L. Bock, G. M. Whitesides, ibid,, 1974, .2..2., 2826. 5. J. P. Collman, R. G. Finke, J. N. Cause, J. I. Brauman, ibid,, 1977, ll, 2515. 6. J., K. Kochi, Pure Appl. ~.!!h., 1980, .5...2., 571. 7. J. K. Stille, K. 8. a) J. San Filippe, J. Silbermann, P. J. Fagan, J...... A.Ilh. ~.Ilh. s. Y. Lau, ~ ~.!!h. ~ , 1977, J..Q, 434. ~ , 1978, l..Q.Q, 4834. b) P. J. Krusic, P. J. Fagan, J. San Filippe, ibid., 1977, ll, 250. 9. s. Winstein, A.H. Fainberg, ibid,, 1957, ll, 5937. 10. J. P. Collman, R. G. Finke, J. N. Cawse, J. I. Brauman, ibid,, 1978, l..Q.Q, 4766. 11. M. Y. Darensbourg, D. J. Darensbourg, D. Burns, D. A. Drew, ibid,, 1976, 98, 3127. Note: in this study electron transfer processes can not be ruled out. 12. a) W. F. Edgell, S. Chanjamsri, ibidu 102, 147. b) Edgell, M. T. Yang, N. Koizumi, W. F. ibid., 87, 2563. c) W. F. Edgell, J. Lyford, A. Barbetta, C. I. Jose, ibid., 93, 6403. 13. M. Y. Darensbourg, D. J. Inorg, Che.!!h., 1978, 17,297. Darensbourg, H. L. c. Barros, 162 14. a} K. H. Pannell, D. Jackson, J.i. Am.... ~.m..... ~ , 1976, .2,Ji, 4 4 4 3. b} M Ni ta y, M. Ro sen b 1 um , J.i. 0 r g an om~ ~I!l.u 1 9 7 7 , l.li, C23. 15. P. B. Chock, Halpern, J., J.i. Am.... ~.m..... ,S_QQ_,_, 1969, il, 582. 16. R. J. Kinney, w. D. Jones, R. G. Bergman, J.i. A.m..... ~.m..... ~ , 1978, J..QQ, 7902. 17. a} R. B. King, F. G. A. Stone, Inorg. Synth., 1963, 2, 99; ibid., 1963, 2,196. b} E. O. Fischer, ibid., 7,136. c} T. s. Piper, G. Wilkinson, J.i. Inorg. Nucl. ~.m....., 1956, .J.., 104. d} H. Brunner, w. A. Herrmann, "Anorganish-Chemisches Praktikum fur Fortgeschrittene- Metallorganische Chemie", Fachbereich Chemie und Pharmazie, Universitat Regensburg, 1979. 163 THE SYNTHESIS AND REACTIVITY OF NOVEL a--ALKYL--µ.-ALKYLIDENE IRON COMPLEXES. A MODEL FOR THE CARBON-CARBON BOND FORMING PROCESS IN FISCHER-TROPSCH CATALYSIS? INTRODUCTION The catalytic hydrogenation of carbon monoxide by FischerTropsch catalysts, and related systems, will undoubtably become increasingly important as coal replaces petroleum as the major feedstock for organic chemicals over the next several decades. Even though Fischer-Tropsch Catalysis has been known since the 1920's, the mechanism of this heterogeneous process remains poorly understood. This situation does not arise for a lack of mechanistic proposals. 1 Two major mechanistic schools of thought prevail. The first argues that insertion of carbon monoxide into a metal carbon bond, a process that is facile in homogeneous systems, followed by reduction of the resulting acyl species, results in extention of the carbon chain (Scheme I). 1 A second group argues that carbon monoxide dissociatively absorbs on the catalyst surface under reaction conditions to give surface carbon and oxygen atoms. 2 This surface carbon species is then sequentially reduced to alyklidyne, alkylidene, and then alkyl species (Scheme II). 1 - 3 The major uncertainty in this later mechanism remains the mechanism of carbon-carbon bond formation. As early as 1926 Fischer and Tropsch proposed that surface methylene intermediates simply polymerize. 3 In support of this hypothesis is the observation l64 SCHEME I H H, I /0 H C I 0I J R ~ .?77777 /7777 -H 0 2 H .I 77 /7777 CH 3 H I 77 777 I H'-- ,,,.o H C I 7777 -H 0 2 etc. 165 SCHEME TI g IJ 777 ~ /777 C ► ~ 1?\ 777 H I /7 I -H0 2 H I yH3 /777 /// H H H I .,77 "c/ I\ /777 yH3 9H2 /777 etc. ---• etc. 166 that diazomethane, a methylene precursor, polymerizes over metal catalysts to give linear hydrocarbons. 4 Alternatively extention of the growing alkyl chain could involve the insertion of a surface methylene species into a surface bound alkyl, Scheme 11.2,3 Homogeneous models for the reduction of carbon monoxide by the mechanisms in Scheme II have been quite few, 5 and no such examples for any of the carbon-carbon forming steps proposed for this mechanism exist. It is the intension of this proposal to synthesize transition metal complexes that might be expected to exhibit reactivity that model the proposed reaction of a surface alkylidene species with the growing alkyl chain. Surface alkylidenes are most likely bound to two adjacent metal centers, thus bridging µ-alkylidenes represent the most reasonable models for these surface species. In this proposal binuclear complexes that contain both µ-alkylidene and ~-alkyl ligands will be prepared and their reactivity studied. In addition the synthesis and reactivity of the related µ-vinylidene complexes will also be explored. THE PROPOSAL It is proposed to synthesize the complexes [Cp(CO)FeCµ-cR 2 )<µ-co)Fe(R') (C0)3l (1, R = H, CH3, R' = CH3, C(O)CH3), and the vinylidene analogues [Cp(CO)Fe Cf-A-C=C(R) 2 > <µ-co>Fe CR') (CO) 3 1 (2). Two novel synthetic approaches for the preparation of these transition metal complexes will be investigated. Both approaches 167 involve the reaction of a nucleophilic irom complex either RFe (CO) 4-Na+, or Na 2Fe (CO) 4, with an electrophilic carbon ligand of a second complex. Na 2 Fe(Co> 4 has been characterized as a "supernucleophile", readily reacting with one molecule of an alkyl bromide to give the complexes RFe(CO) 4-Na+. These complexes in turn react with alkyl halides to give ketones. 6 This later reaction is believed to proceed type [R(O)Cl (R)Fe(Co> 3 , through an intermediate of the which could not be observed. The corresponding Osmium complexes are quite stable and the mechanism of reductive elimination in these complexes has been the subject of considerble study. 7 The first approach involves the reaction of (R') Fe (CO) 4-Na+ with the cationic complexes [Cp(L) (CO)Fe 2 =cR 2 ]+ (3) , and [Cp(L) (CO)Fe 2 =C=CR 2 l+ (4)lO (L = CO, PPh 3 , Scheme III). In this reaction attack of the nucleophilic metal on the alkylidene carbon, and the formation of a metal-metal bond, will give a bridging µ-alkylidene complex. The resulting complex, however, may be an acyl-µ-alkylidene complex resulting from migration of the alkyl ligand to carbonyl in RFe(CO) 4- prior to attack upon the alkylidene complex. It has been shown that this migration process is involved in the reaction of RFe(co> 4- with alkyl halides. 6 Furthermore it is well established that nucleophiles react with group VI-VIII alkylidene complexes by attack at carbon, 8 and recent work has established similar behavior for mononuclear vinylidene complexes as we11. 10 In fact some known µ-vinylidene complexes were prepared in this way. 9 168 SCHEME ill 1co R + / CpFe=~ \ R L I N + co I Cp \ e=C= C R2 + R' Fe(co4- Na+ L \ ~R2 C /C, P;Fe, Fe(C0) 3 L C/~ C(O)R' O 2 rJ L=CO,PPh 3 R=H,CH ,Ph 3 169 The second synthetic approach to be employed here involves the reaction of RFe(C0)4-Na+ with CpFe(L)2-CH2-X (L= co, PPh3, X = Cl, OMe, SMe 2+>lO (Scheme IV). In this reaction displacement of the leaving group by the metal nucleophile is expected to be followed by formation of a metal-metal bona, and the bridging of a carbonyl, resulting in the µ-alkylidene structure shown. It is hoped that this process will be competitive with reductive elimination from the iron atom. The complexes Cp(CO) 2FeCH 2x (X= Cl, OMe) have been used as a precursors for the unstable cationic alkylidene complex [Cp(CO) 2Fe=CH 21+.ll Alternatively use of the Fe(CO)- 2Na 2 + dianion as the nucleophile followed by alkylation of the resulting binuclear anion may yiel a the desired o--alky 1-µalkylidene complexes (Scheme IV). An interesting extension of this method involves the use of iron alkyl complexes with the leaving group at more distant positions on the alkyl ligand. Reactions of these complexes with metal nucleophiles could then result in binuclear complexes linked by carbon chains of two, three, or more carbon atoms. A complex of this A has recently been prepared by Theopold and Bergman, Cp (CO) Co-Co (CO) Cp. 12 The second part of this proposal involves a study of the reactions of the complexes 1, 2, and 3. Of greatest interest will be reactions involving bona formation between the two organic ligands. This reaction if it gives CpFe(L) 2 CcR 2R1 ) can be thought of as similar to the reductive elimination of two 0--bouna organic ligands from R2Fe(CO)4. This should be a facile thermal process. 6 170 SCHEME IV co I Cp~e-CH X 2 co I RX A/ co 1 CpFe-(CH )-x \ n + R Fe(CO( co ~ (CH2)n / ' Cp(CO)Fe--Fe(CO) "'-c/ \ 3 O C(O) R X = Cl -OMe -OCR ' ' ' + -SMe 2' 3 +N -PPh 3 171 However in the absence of thermal or ligand induced reactivity, photolysis, or oxidation of the complex may be employed. Considerable evidence has been aquired in recent years that oxidation often induces carbon-carbon bond coupling reactions.13 Alternatively the alkylidene ligand could "insert" into the iron alkyl bond to give R'C (0) CR 2Fe (CO) 3 -(L) 2FeCp (Scheme V). DISCUSSION As with all synthetic studies of this kind the stability of the products cannot be predicted directly. However it has been noted that bridging alkylidene and vinylidene complexes are generally more thermally stable than their mononuclear anologues, thus the proliferation of these complexes that have been prepared in recent years. 10 , 14 Stoichiometric models for the proposed chain propagating steps in the Fischer-Tropsch reaction have been lacking, largely because of the absence of known complexes that model the structure of the surface bound intermediates in this reaction. The increase of characterizable µ.-alkylidene and µ.-vinylidene complexes recently suggests that reasonable models for the reactivity This these surface intermediates may be found. proposal is one such attempt at the synthesis of model complexes that exhibit Fischer-Tropsch reactivity. In conclusion some novel synthetic procedures for the preparation of µ.-alkylidene complexes are proposed. These methods will be employed to prepare binuclear iron complexes containing 172 both a µ-alkylidene and a c,--alkyl ligand. The reactivity ~f these o--alkyl-µ-alkylidene complexes will be studied. In particular reactions that induce carbon-carbon bond formation are of interest. It is believed that the observation of carbon-carbon bond forming substantiate processes in mechanistic these complexes proposals for should the help catalytic hydrogenation fo carbon monoxide that involve the active participation of surface alkylidenes in the growth of the carbon chain. SCHEME V Fe(CO) L 3 2 0 l I /V /c, Cp(L)Fe-Fe(C0) L 2 'c/ \ 11 CR R' 0 2 173 REFERENCES 1. a) G. H.-Olive, 136. b) s. Olive, Angew_._ .c.h.e..m.... .I.nL. ~ , 1976, 1.5., M. A. Vannice, Cata!. Rev.-Sci. ~ , 1976, 14(2), 153. c) G. A. Mills, F. W. Steffgen, 159. d) H. J. Storch, N. Golumbic, ~ R. Fischer-Tropsch and Related Syntheses 0 , ~ , 1973, fil.2.l, G. Anderson, nThe Wiley Interscience, New York, 1951. 2. a) J. G. Ekerdt, A. T. Bell, .J..... .c.at,._, 1979, .2..8., 170. b) G. C. Low, A. T. Bell, ibid,, 1979, 51, 397. c) J. G. McCarty, H• Wi s e , .i!2i.!L.., 1 97 9 , 5..2 , 40 61 • d ) A. Za g 1 i , E• J • L. Falconer, C. A. Keenan, ibid., 1979, .5..6., 453. e) P. Bileon, J. N. Helle, W. M. H. Schatler, ibid.,1979, 58, 95. f} R. A. Rabo, A. P. Risch, M. L. Poutsma, ibid,, 1978, 5-.l, 295. g) v. Ponec, Cata!. Rev,-Sci. .E. ns.u 1978, 1...8., 151. 3. a) F. Fischer, H. Tropsch, Brenstoff~m.,_, 1926, 2, 97. b) S. R. Craxford, E. K. Rideal, .J_._ .c.h.e..m.... ~ , 1939, 1604. 4. a) F. D. Mango, I. Dvoretzky, .J..... Am.a. ~ID.a.~, 1966, .§Ji, 1654. b) H. Meerwein, H. Rathjen, H. Werner,~ Dtsch, ~llh. ~ , 1942, 1610. 5. a) J. S. Bradley, G. B. Ansell, E. W. Hill, .J..... A.m.... .CMm..... ~ , 1979, l.Qi, 7417. b) J. S. Bradley, ibid,, 1979, .l.Ql., 7419. c) M. Tachikawa, Kiester , .J..... ~.m.... ~ E. L. Meutterties, .C.hg.m.... .C.Q.mm...., 19 7 9 , in press; o. 214. B. d > R. Calvert, J. R. Shapley, .J..... A.m.... ~llh. ~ , 1977, il, 5225. e) R. G. Calvert, J. R. Shapley, A. J. Shultz, J. M. Williams, S. L. Suib, G. D. Stucky, ibid,, 1978, l.QQ, 6240. 174 6. a) J. P. Collman, A££.&. .c..h.e..m.... ~ , 1975, Collman, R. G. Finke, J. N. Cause, J. a, 342. b) J.P. A.m.... I. Brauman, k ~.m.... ~ , 1977, il, 2515; ibid., 1978, l.QQ, 4766. 7. J. R. Norton,~ ~.m.... ~ , 1979, 1.2., 139; and references therein. 8. a) H. Berke, ..Z..... Naturf., 1980, .3....5.,Q, 86.b) A. B. Antonova, N. E. Kolobova, P. k v. Petrovsky, B. v. Lokshin, N. s. Obszyuk, Organo.m~ ~!!Lt.., 1972, .lll, 55. c) v. C. Andrianov, Yu. T. Struchkov, N. E. Kolobova, A. B. Antonova, N. s. Obczyuk, ibid., 1976, l..2..2., C33. d) T. V. Ashworth, J. A. K. Howard, .c.h_g_.m.._ ~ ~.m.... ,C,Q.m.m...., 1979, 42. F. G. A. Stone, k 8. Se e f or ex a rn p 1 e : a ) C. P. Ca s e y , S. W. Po 1 i ch now s k i , k ~.m.... .,S_Q_Q_._, Gladysz, 1977, il, 6097. b) ibid., 1979, .JJU, Dahan, Jeannin, k o. Fi s c he r , w. w. 5440. K. Wong, w. Am.... Tarn, J. A. c) Levisalles, Rudler, Organorn~ ~.m...., 1980, lJU., 233. d) E. He 1 d , .i.b.i_g_._, 1 97 6 , l.12. , c5 9 . e) E. o. Fischer, Selrnayr, Kriessl, Schubert, ~.m....~, 1977, .l.l.Q, 257 4. 10. a) M. Brookhart, J. R. Tucker, T. c. Flood, J. Jensen, k Am.... ~.m.... ~ , 1980, l..Q2, 1203. b) A. Solar, k Davidson, J. P. Organorn~ ~.m...., 1978, .l...5...5., CS. c) A. Davidson, J. P. Selegue, k Am.... .c.h_g_.m.._ ~ , 1978, l.QQ, 7763. d) R. D. Adams, A. Davidson, J. P. Selegue, ibid., 1979, l..Q.l, 7232. e) M. I. Bruce, A.G. Swincer, R. G. Wallis, k Organorn_g_t_._ ~.m...., 1979, ill., cs. 11. a) P. W. Jolly, R. Pettit, ibid., 1966, 88, 5044. b) M. L. 175 H. Green, M. Ishaq, R. N. Whiteley, .il... ~.m...... ~ J.Al, 1967, 1508. c) M. Brookhart, J. R. Tucker, T. C. Flood, J. Jensen, iI.... AL ~h~L SQ£.&., i~M, 102, 1203. d) S. Brandt, P. Helquist, ibid., 1979, l.Q.l, 6473 . e) J. A. Labinger, J:..... 0rganom~ ~.m......, 1980, ill., 287. f) D. L. Roger, Culbertson, ibid., 1977, E. , c. ill., 297. 12. K. H. Theopold, R. G. Bergman, .iI.... Am..... ~.m...... ~ , in press. 13. J. K. Kechi, Pure Appl. ~m...., 1980, .22., 571; and references therein. 14. a) w. A. Herrmann, Angeli... ~.m...... .I.nL ~ Engl., 1978, il, 800. b) w. A. Herrmann, r. Bernal, et.al. , .iI.... 0rganom~ ~.m......, 1979, 165, Cl7. c) E. o. Fischer, V. Kiener, ibid., 1970, 23, 215. d) M. B. Hursthoue, R. A. Jones, K. M. Abdul Malik, G. Wilkinson, .il... Am..... ~.m...... ~ , 1979, l.Ol., 4128. e) N. E. Kolobova, Antipu, Yu. 69. f) A. B. Antonova, D. M. Khitrova, M. Yu. T. Struckhov, .iI.... 0rganom~ ~.m......, 1977, 137, N. E. Kolobova, Yu. T. Stauchkov, A. N. Nesmeyanov, G. G. Aleksandrov, A. B. Antonaova, K. N. Anisiniov, ibid,, 1976, .ll.O, C36. g) c. R. Eady, B. F. G. Johnson, J. Lewis, .iI.... Chem.....~ Dalton, 1977, 477. h) M. Green, A. Laguna, J. L. Spencer, King, M. S. ibid,, 1972, F. G. A. Stone, Saran, .il... Am..... ibid., 1977, 1010. i) R. B. ~.m...... ~ , 1973, 95, 1811. j) li, 1784. k) O. S. Mills, A. D. Redhouse, .il... Chem~ ~.m ~.mm....., 1966, 444. 1) H. W. Sternberg, J. G. Shukys, C. D. Donne, R. Markby, R. A. Freidel, I. Wender, .il... AL rh~.m SQ.Qi, 1959, ~l., 2339. m) 0. S. Mills, A. D. 176 Redhouse, Inorg, .c..h.iI!h. Acta,, 1967, l., 61. 177 THE ABSOLUTE RATE CONSTANT FOR CIS-TRANS ISOMERIZATION OF VINYL RADICALS INTRODUCTION Vinyl radicals 1 can be produced through the addition of free radicals to alkynes, 2 ' 3 through irradiation in an ESR cavity, 4 , 5 through the decomposition of a,,8-unsaturated peresters, 6 - 8 through the irradiation of vinyl iodides, 9 and through the reduction of vinyl chlorides by sodium naphthalenide. 10 In all cases mixtures of both ~- and .z.-alkenes are obtained upon abstraction of hydrogen from hydrogen donors. ESR studies of the ethenyl radical, spectrum characteristic of CH 2 cw, observed a two species that are rapidly interconverting, consistent with a non-linear ethenyl radical interconverting between cis and trans geometrical isomers. The half life for this process was determined to be between 3 x 10-a and 3 x 10-lO sec at -160°to -104°C, which assuming a frequency factor of 1013 corresponds to a barrier to isomerization of about 2 kcal/mole. 4 The authors implicated electron tunneling in this process. By contrast the 1-methyl vinyl radical was found to be configurationally stable at -172°c, indicating a barrier to interconversion of greater than 2.7 kcal/mole. 4 ' 5 Early studies found that in the thermal decomposition oftbutyl ~- and z-a,,8-dimethylpercinnamates, and the corresponding t-butyl ~- and .z.-a-methyl and- a-phenylpercinnamates yielded Zand ~-alkenes in ratios of 1.16 and 1.5 respectively, independent 178 of the starting isomer. 6 • 7 These results can be rationalized either by a symmetrical intermediate, or a pair of rapidly interconverting isomers, where in both cases steric influences control the observed product ratio. Support for this steric control of the product ratios was found in the effect of changes in the hydrogen donor upon the observed ZI~ product ratio; for example cumene gave higher ZIE ratios than cyclohexane. 6 , 7 Sargent and Browne reduced isomerically pure~- and z-3chloro-3-hexene with sodium naphthalenide and observed a different ratio of z- and ~-alkenes depending upon the starting isomer, with retention of stereochemistry in the predominant isomer. This observation allows for the conclusion that there exist distinct cis and trans vinyl radicals, not a single linear intermediate. More importantly this experiment was able to demonstrate that the rate of trapping of the vinyl radical by sodium naphthalenide, to give the configurationally stable vinyl carbanion, is faster than the rate of ~-trans equilibration for disubstituted vinyl radicals. THE PROPOSAL It is proposed to study the reduction of a series of three pairs of~- and z-vinyl fluorides by sodium naphthalenide in 1,2dimethoxylethane (DME) solvent. By examining the EIZ ratio of the alkene products it will be possible to determine the absolute rate of isomerization of vinyl radicals. The series of vinyl fluorides to be studied are:~- and Z-3-fluoro-3-hexene (la, lb), E- and ~-2-fluoro-2-butene (2a, 2b), and E- and Z-l-fluoro-1- 179 deutero-1-propene (3a, 3b). X lb '==<F 2b ~F '=<F la 2a '==<D F 3a It has been adequately established that the reduction of alkyl halides by sodium naphthalenide, as well as other aromatic radical anions, proceeds by an initial one-electron transfer, followed by dissociation of halide. The resulting alkyl radical is reduced to a carbanion by another molecule of naphthalene radical anion. The resulting carbanion then abstracts a proton from solvent.lO,ll This mechanism is outlined in Scheme I. In this mechanism k 3 is rate limiting and halogen dependent. The order of relative rates goes X = F <<Cl= Br= I. When X = Cl, Br, or I the reaction proceeds at a rate approaching the rate of mixing of the reacants, while florides typically have rate constants of 1 x 10- 4 M- 1 sec- 1 . 11 The second order rate constant kz has been measured by examining the extent of cyclization in the reduction of 5-hexenyl fluoride. 11 In that study it was established that k 2 = 1.6 x 10 9 M-l sec-l or almost the diffusion con trolled 1 im it. It is an ass umpt5.on of this proposal that the rate constant k 2 for the reduction of the vinyl radical to the vinyl carbanion is the same as for the reduction of alkyl 180 SCHEME I R' + Na· X + N s- + RH a Na, + N~ is • so d'iurn nap h tha 1 eni'd e. b I f sg 1~ • d'imet h_oxyethane (DME), then S- is CH =CHOCH + CH o Na 3 2 3 181 raaicals to carbanions. Basea on analogy to the reauction of alkyl halides the porposed mechanism for the reauction of isomeric vinyl fluorides is presentea in Scheme II. In this scheme k 2 is assumed to be the same for both cis and trans vinyl radicals. Basea upon this mechanism the kinetic analysis goes as follows. Starting from the .Z.-isomer (I), for example, a steady state approximation is made on the concentration of the opposite cc·) Starting from~- vinyl radical isomer, the ti.e_-vinyl radical vinyl fluorides (II) this assumption is made on the trans-vinyl radical concentration. Therefore, starting from I: (T•) a cc· 1 dt = cc· l solving; = k 1 [T•l 0 - k_ 1 [C•l - k 2 [C• l [N~l kl = (1) • k -1 + k 2 [N-l [T • l [T• l = concentration of trans-vinyl radical [ C: l = concentration of cis-vinyl radical [N-l = concentration of soaium naphthalenide [Tl = concentration of trans-alkene proauct [Cl = concnetration of cis-alkene proauct where the observed proauct ratio is airectly proportional to the concentration of these radicals; . [Cl) ( = [Tl obs . = [T• l k [T•l [N-l 2 under equilibrium conditions; cc· l [T • l cc· l k2[c•1 [N-l = kl k -1 k_l (2) . >> k 2 [N-l ( 3) 182 therefore at equilibrium; [Cl) ( [Tl eq ( 4) • under non-equilibrium conditions k_ 1 is comparable to k 2 CN-l, substituting eq. Cl) and (4) into eq. (2); [Cl) = ( [Tl obs k -1 ( [ Cl / [ Tl ) eq (5) solving for k_ 1 ; ( 6) (([Cl/[Tl)eq/([Cl/[Tl)obs) - 1 using eq (4) and solving for k 1 = k - l ( [ Cl / [ Tl ) eq (7) A similar analysis may be employed for the opposite isomer. However, the ratios ([Cl/[Tl) 0 bs and ([Tl/[Cl) 0 bs from~- and~alkenes will not necessarily be the same, and must be distinguished. The ratio ([Cl/[Tl>eq' on the other hand, must of necessity be identical starting from either vinyl fluoride isomer. This ratio can be evaluated by either of two methods. The first involves reducing the concentration of sodium naphthalenide until further dilution leads to no change in the ratio ([Cl/[Tl) observed, always maintaining an excess of reducing agent. If this does not slow the rate of reduction sufficiently, then a change in either the metal cation or the aromatic radical anion is capable of changing k 2 by orders of magnitude. 11 , 12 In particular, a change to either potasium naphthalenide or sodium H T R R>=<H I R H >==< F R SH k3 • Na,N- SCHEME JI Na, N.!. l T. R H R R~ k 2 H RR H R~R k2 N~, N.!. l c· R~ ·R . -I H kl k SH k3 JI F H C H R>=<R H Na, N_;_ R>==<R w 00 f-J 184 anthracenide should work. It should be noted that the ([C]/[T])eq is independent of the rate constant k 2 and the reducing agent, in must merely be established that equilibrium has been established between the two interconverting isomers. THE EXPERIMENT All reactions must be carried out under an inert atmosphere. The sodium naphthalenide can be prepared by depositing a sodium mirror on the wall of a flask, then transfering in the solution of naphthalene in DME. The solutions can be titrated with H2o to determine their molarity. 11 Addition of the vinyl fluoride in DME solution to the filtered solutions of sodium nathalenide can be done either by syringe or through a break seal as described by Garst, 11 a syringe pump may be particularly useful in this addition. All reactions should be carried out with a 5-10 fold excess of reducing agent, so that the concentration can be known accurately. Reactions of alkyl fluorides typically have halflives of 3-5 minutes. The .E- and .z.-alkenes can be analyzed satisfactorily on a number of vapor phase chromatography (VPC) columns without workup. The column employed by Sargent and Browne was an AgN0 3/glycerine column. Yields can be measured by addition of alkanes as internal standards and comparisons of output integrations after establishing molar resonce factors with known samples. The synthesis of the vinyl fluorides required by this proposal can be achieved by one of two possible synthetic routes (Scheme III). The E- and z-vinyl fluorides can be separated by 185 SCHEME 1IT I) FCI0 Ph P=CHR 3 3 2) Nol + [Ph/-CHRF] I- j 80% Phli Ph P=CRF 3 R' Ph P=CRF + 3 R' 'c =o _ ____. / F 35 -50% H>=<R H ci s/trans R\dR NBS, Br 2 R (HF )x· pyr Br>==<F R Li/LiC0 60-70% R\_..JR r-\ H F cis/trans 65% 3 186 column chromatography, 13 or by preparative VPC techniques. Vinyl halides form peroxides easily and therefore the compounds should be chromatographed through alumina prior to use. NMR, IR, and mass spetral data should be sufficient in assigning the structures of the various isomers. However, not suffice, if these methods do then treatment of the pure isomers with t- butyllithium at -120°C followed by quenching with a proton source will allow correlation with the corresponding hydrocarbons. 16 DISCUSSION There are a number of assumptions inherent in this proposal. First, it is assumed that the mechanism for the reduction of vinyl fluorides by sodium naphthalenide is the same as for alkyl florides. Most importantly, it is assumed that the rate of reduction of the vinyl radical to the vinyl carbanion occurs at the same rate as was measured for alkyl radicals. There is no way that the validity of this assumption can be tested directly1 however, considering the almost diffusion controlled rate of this process this assumption does not seem very drastic. Secondly, it is assumed that the reduction of the ti.,s.- and trans-vinyl radicals procedes at the same rate. That is, that there is no steric barrier like that which was encountered for hydrogen abstraction. The results of Sargent and Browne are encouraging in this regard. Furthermore, a comparison of ([Tl/ [Cl) eq obtained for compound la,b and 3a,b should indicate if any significant steric barrier exists. Thirdly, it is assumed that the vinyl carbanions are 187 stable to £ll/trans isomerization under reaction conditions. There is good evidence for this stability. 17 In particular, the two electron reduction of vinyl chlorides by Na/NH 3 goes with at least 96% retention of stereochemistry. 18 Most importantly there is the steady state assumption. This, . however, can be tested. A plot of k_ 1 vs [N-l should yield a straight line with a slope of zero. This lack of a concentration dependence will work both to confirm the correctness of the steady state assumption and the kinetic analysis derived from it, but also indirectly it will confirm the mechanism in Scheme II. In conclusion it is proposed to measure the absolute rate constant for the cis-trans isomerization of vinyl radicals by using a steady state analysis on the reduction of vinyl fluorides by sodium naphthalenide. The rate of reduction of the vinyl radical isomers to the stable vinyl carbanions is assumed to be 1.6 x 10 9 M-l sec- 1 , the value obtained for the reduction of alkyl radicals to carbanions by sodium naphthalenide. 188 REFERENCES 1. a) o. s i n am u r a , Top i c s .i.n s t e r e o c he m.i, 19 6 9 , .i, 1 . b > w. G. Bentrude, AruL.. .RfilLt. Phys, Chem., 1967, JJl, 300. 2. a) P. S. Skell, R. G. Allen,~ Am.... ~.m_._ ~ , 1958, .a,Q, 5997. b) A. A. ibid., Oswald, 1964, ll, K. Griesbaum, 2877. c) W. T. B. B. Hudson, Jr., Dixon, J. Foxall, G. H. Williams, D. J. Edge, B. C. Gilbert, H. Kazarians-Maghaddam, R. o. c. Norman, .J..... Chem, ,SQQ_._ Perkin ll, 1977, 827. 3. J. A. Kampmeier, G. Chen, ibid,, 1965, fil, 2608. 4. R. w. Fessenden, R. H. Schuler,~ .cM.m.... Phys,, 1963, ll, 2147. 5. F. L. Cochran, F. L. Adrian, v. A. Bowers, ibid,, 1964, .i.Q, 213. 6. J. A. Kampmeier, R. M. Fantazier, ~ Am.... ~.m_._ .£Q.Q..., 1966, 1959, 5219. 7. L.A. Singer, N. D. Kong, T.eL Lett., 1966, 2089. 8. L.A. Singer, N. D. Kong, .J..... Am... Chem, ,SQQ_._, 1966, .fill, 5213. 9. R. C. Neuman, Jr., G. D. Holmes, ~ ~ ~.m...., 1968, ll, 4317. 10. G. D. Sargent, M. w. Browne,~ Am.... ~.m_._ .£Q.Q..., 1967, li, 2788. 11. a) J. F. Garst, F. E. Barton II, ibid,, 1974, il, 523. b) J. F. G a r s t , ~ Chem,~, 1970, .i, 400. 12. Unpublished results, J.M. Huggins, Undergraduate Thesis, Amherst College, Amherst, Mass., 1976. 13. M. Schlosser, M. Zimmermann, Synthesis, 1969, 75. 189 14. G. Olah, M. Nojima, I. Kerekes, ibid., 1973, 780. 15. L. H. Briggs, J. P. Bartley, P. S. Rutledge, .J..t_ ~.m_._ ~ Perkin i, 1973, 806. 16. H. Neumann, D. Seeback, ~ Lett,, 1976, .5..2., 4839. 17. D. J. Cram, "Carbanion Chemistry", Academic Press, Inc., New York, N. Y., 1965, pg. 133-35. 18. M. c. Hoff. K. w. Greenlee, C. E. Boord,~ Am..... ~.m..... SQfu., 1951, ll, 3329. 190 THE STUDY OF CATIONIC BISPHOSPHINE METHYLNICKEL(II) AND PLATINUM(II) COMPLEXES: THE SEARCH FOR EVIDENCE OF AN a-HYDRIDE MIGRATION PROCESS. INTRODUCTION The process of a-hydride migration is represented in equation Cl). Evidence for this reaction only exists for a few (1) (R = H, C) metal complexes. 1 - 5 The most convincing evidence exists for multiply alkylated complexes of Cr, Nb, and Ta (MRn, n 2. 3), and in a report by Green and Cooper in which a tungsten hydridocarbene complex was trapped with P(CH 3 ) 2 Ph. 1 This reaction is presented in Scheme I. The complex 2 was isolated from the reaction of the alkene complex 1 with P(CH 3 ) 2 Ph at room temperature. Upon heating to 50°C in acetone for 1 1/2 hours this complex was converted with loss of ethene to a second product 5 in about 70% yield. Complex 5 exhibits an IR absorption at 1938 cm- 1 , and 1 H NMR absorptions at 8 1.16 and -12.11 ppm (ratio 2:1) which led to the structural assignment shown in Scheme I. Heating 5 for an additional 8 days at 70°C led to a 50:50 mixture of 5 and 6. 1 These observations led Green and Cooper to conclude that they had observed the equilibrium 3 ~ 4, an a-hydride migration process. The kinetic product of the reaction is attack of phophine at carbon to give 5, the thermodynamic product being attack at tungsten leading to 6. 191 SCHEME I /~ Cp2W'CH + 3 I rv 70°C + Cp W-CH 3 2 k3 1PM~Ph + ,PMe Ph 6 N ~+ Cp2W-=-CH2 + c~c~ 4 rv 3 /'V Cp2~ CH kl 2 3 k.zl PMelh + C W/CH2 PM~Ph P2 \ H 5 ,..._, 192 THE PROPOSAL It is proposed to look for evidence of an a-hydride migration process in the reactions of complexes of the type [{L)2M{R){solvent)]+pp6- {7: M = Ni, Pt; R = -CH3, -CH2Ph; L = bulky phosphi te ligand). The complex [ {L) 2 Pt {CH 3 ) {acetone) J +pp 6- {L = P{CH 3 ) 2 Ph) has been prepared by Chisholm and Clark. 6 It is stable at room temperature, but decomposes upon heating. The acetone ligand can be displaced easily to form 77"-complexes with a number of alkynes. The chemistry of these methylplatinum{alkyne) complexes has been extensively studied. 6 , 7 The complexes 7 will be prepared at low temperature, 6 , 8 , 9 then allowed to react with a number of "trapping" reagents at room temperature or below. The preparation of these complexes and two different approaches to trapping an intermediate hydridocarbene complex 8 are summarized in Scheme II. By analogy to the results of Green and Cooper phophines such as P{CH 3 ) 2 Ph, PPh 2 {CH 3 ), or P{CH 3 ) 3 will be used in an effort to observe complex 9. This complex will be characterized by IR <v<MH)), 1 H NMR {M-H, M-CH 2 P), and by chemical trapping with dry DCl to give HD, {CDH 2 PR 3 ) +, and L2 M{H) {Cl) _lO Other attempts to confirm the existence of 8 are based upon the presently favored mechanism for the olefin metathesis reaction. 11 Treatment of 7-d 3 with ethene-d 0 can result in one of three possible reaction pathways. {l) Reversible formation of a hydr idometallocyclobutane complex {11) results in no net reaction. Complex 11 is symmetrical and 193 SCHEME JI + L M-C H C H CH 2 2 3 2 12 /"'.J i L, /Cl M L/ 'CH Al L H I+ L2tJ 3 l AgPF6 j- +/o 3 7 rv lPR3 + 1 PR 3 'M L/ 'CH 3 10 r..J L,+~CH 2 M L/ 'H CH=CH 2 2 L, ~ + CH 1 M L/ 'H 8 /V 1lp R3 + L,M/CH L/ 'H 9 IV or ~ 11 r,..I 'M L/ 'CH L 6 2 PR 3 2 194 can reverse to ethene and 8 to give either CH 2 =cH 2 and 8-d 3 , or CH 2 =cn 2 and 8-d 1 . Detection of CH 2 =cn 2 in the recovered ethene would confirm the existence of this pathway. (2) The hydridometallocycobutane complex (11) could be isolated as a stable product. Complex 11 is a 16 electron Ni(II) or Pt(II) complex, normally a stable electron count for this oxidation state. (3) Alternatively 11 could react to give other products by one of three pathways: a) reductive elimination to give cyclopropane, b) /3-hydride elimination to give propene, 11112 c) a-hydride transfer to give a cationic propyl complex (12), and possibly polymerization by a repetition of this process. This propyl complex, if observed, will contain a mixture of propyl groups, CH2CH2CD3 and -CD2CH2CH2DIt cannot reasonably be predicted which of these pathways will be observed. However, with the exception of the poylmerization of ethene each process will give products that are unique and will necessitate the postulate of 8 as an intermediate in their formation. Alternative approaches involve starting with a cationic benzyl complex [ CL) 2 M(CH 2 Ph) (acetone) l + (13) and looking for formation of styrene and 7 in the reaction with ethene. Or conversely allowing the cationic methyl complex 7 to react with styrene and looking for the formation of 13 and ethene. In addition 13 may react with phophines to form complexes of the type 9 more readily, because of stabilization of the resonance 195 structures for 8 that place the positive charge on carbon, increasing its electrop hility. DISCUSSION Two characteristics of the system studied by Green and Cooper were recognized as being significant in the design of this proposal. (1) The complex [Cp 2 W-CH 3 J+ (3) is a 16 electron unsaturated complex, a-elimination gives [Cp 2 W(H)=CH 2 J+ (4) a saturated 18 electron complex. In addition this compl~x is a cation with strong 77"-acid ligands, the cyclopentadienyl factors contribute to destabilize 3 ligands. These with respect to 4. ( 2) Green and Cooper observed a situation where k 2 > k 3 (Scheme I). Furthermore the thermodynamic equilibrium of 5 and 6 is roughly 50:50, thus this kinetic preference for attack at carbon (to give 5) is probably due to greater steric hindrance to attack at the metal center. In addition a resonance form for 4 , cp 2 M(H)-CH 2 +, places the positive charge on carbon, increasing its electrophilici ty. Consideration of these points led to the design of the present system. [L 2 M(CH 3 )(acetone)J+ can be considered an unsaturated 14 electron M(II) complex if the acetone ligand is ignored. The ligands Lare to be strong 77"-acids to maximize destabilization of complex 7 relative to 8 (Scheme II), and bulkyl so as to favor nucleophilic attack at carbon rather than the metal. In summary, it is proposed to study the reactions of 196 cationic methylnickel and methylplatinum complexes of the type [(L) 2 M(CH 3 )(acetone)J+PF 6 - with phosphines and alkenes. These reaction will be analyzed in an effort to gain evidence for the intermediacy of an a-hydride migration process. 197 REFERENCES 1. a) N. J. Cooper, M. L. H. Green, .J..... .c.h..e.Ilk. ~ ~Ilk. ~ID!Il..L, 1974, 208. b) N. J. Cooper, M. L. H. Green, ibid., 1974, 761. c) M. Canestrari, M. L. H. Green, ibid, 1979, 913. d) M. canestrari, M. L. H. Green, .J..... ~ I l k . ~ Dalton Trans., 1979, 1121. 2. A. Sanders, L. Cohen, w. P. Giering, D. Kennedy, c. V. Magatti, .J...... &h Chern,~, 1973, .2.5., 5430. 3. w. A. Herrman, G. Kruger, R. Goddard, I. Bernal, .J...... Organornet, Chem., 1977, l.iQ, 73. 4 . a) M. Mich rn an , H. H. Ze i s s , ibid., 1 9 6 8 , 13 , ·p 2 3 . b) R. P. A. Sneeden, H. H. Zeiss, AngeYLL .c.h..e.Ilk. .I..n.t.,_ Ed, Engl.~ 1968, 2, 9 51. c) M. Tsutsui, H. U. ze i ss, .J..... .A!Il..L .c.h..e.Ill.L ~ , 1959, fil.., 6090. 5. a) R. R. Schrock,~ .c.h..e.Ilk. ~ , 1979, .l..2., 98. b) R.R. Schrock, G. w. Parshall, .cM!Il..L Rev., 1976, li, 243. c) R. R. Schrock, .J...... .A.m... ~h~.m.a. ~Q.Q....., 1975, ~2, 6577. d) R. R. Schrock, .J...... Organornet. Chern., 1976, 12.2., 209. 6. M. H. Chisholm, H. C. Clark, .J..... .A!Il..L ~ I l k . ~ , 1972, ll, 1532. 7. a) H. C. Clark, K. von Werner, .J..... Organorn~ ~Ilk., 1975, l.Qi, 347. b) H. C. Clark, H. H. Puddephatt, Inorg • .c.h.e.!Il..L, 1972, 11, 1269. 8. a) K. A. Jensen, z.... Anorg • .c.h.e.Ilk., 1936, 22..2., 265. b) L. M. Venanzi, .J..... .c.h.gI!l..!.. ~ , 1958, 719. c) G. Boothe, J. Chatt, ibid., 196 5, 3 23 8. d) o. Dahl, Ac ta, ~h~Ilk. Seang_,_, 196 9, 21., 198 2342. 9. a) H. F. Klein, H. H. Karsch, b) J. Chatt, B. L. Shaw, .J..... ~m.,_ ~ , ~m.,_ ~ , 1972, l.Q2, 2628. 1959, 705; 1960, 1718. c) M. L. H. Green, M. J. Smith, .J..... ibid,, .c.h.e_m.,_ ~ , 1 9 7 1 , 6 3 9 • d) G. Wi 1 k e , G. He r r rn an n , Ang e~ ~m.,_ .I..nL.. .EJh. Engl,, 1966, 2, 581. 10. a) M. L. H. Green, T. Saito, P. J. Tanfield, .J..... .c.h.e_m.,_ ,S_Q£,_ J.Al., 1971, 152. b) J. Chatt, B. L. Shaw, .J..... ~m.,_ ,S_QQ,_, 1962, 5075. 11. a) N. Calderon, E. A. Of stead, w. A. Judy, Angew..L ~m.,_ .I..n.ti. ~ Engl,, 1976, ll, 401. b) ~m.,_ S Q ~ ~ , .c..hgm.,_, 1978, R. J. Haines, G. J. Leigh, 1975, .i, 155. c) R. H. Grubbs, Prog, Inorg_, li, 1. d) c. P. Casey, s. w. Polichnowski, .J..... AID.&. Chern,~, 1977, .2.2.., 6097. 12. E. Ephritikhine, M. L. H. Green, .J..... ~m.,_ .£Q£_,_ .c_~m.,_ ,CQmm.,_, 1976, 619; ibid,, 1976, 926.