Design, Synthesis, Characterization, and Reactivity of Group III ansa-Metallocenes Containing Bulky Alkyl and Silyl Substituents Citation Abrams, Michael Benjamin (1998) Design, Synthesis, Characterization, and Reactivity of Group III ansa-Metallocenes Containing Bulky Alkyl and Silyl Substituents. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/2v44-pj33. https://resolver.caltech.edu/CaltechTHESIS:09052025-171738932 Abstract The preparations of a series of silyl-bridged bis(cyclopentadiene) complexes containing bulky alkyl and silyl substituents are reported. Incorporation of tertiary alkyl groups are required in order to direct trimethylsilyl groups exclusively adjacent to the silyl linkers. This 1,2,4- substitution is crucial for the directed syntheses of chiral ansa-metallocenes. Transmetallation of Me 2 Si{2-SiMe 3 -4-[2-(2-CH 3 -C 10 H 14 )]-C 5 H 2 ) 2 2- affords exclusively racemo-chiral) Group III metallocenes, whereas transmetallation of Me 2 Si{3-[2-(2-CH 3 -C 10 H 14 )]- C 5 H 3 } 2 2- (lacking the α-TMS groups) proceeds with the opposite enantioselectivity, generating exclusively mesa- (achiral) metallocenes. Treatment of meso-Me 2 Si{3-[2-(2-CH 3 -C 10 H 14 )]-C 5 H 3 } 2 MCl-(LiCl)-(THF) 2 (meso- AdpMCl-(LiCl)-(THF) 2 ) with allylmagnesium bromide affords mesa-AdpM(ƞ 3 -C 3 H 5 ) (M = Sc, Y). Hydrogenolyses of these mesa-metallocenes generate highly reactive 14-electron d 0 metal hydride complexes. The methyladamantyl substituents on the Cp rings inhibit formation of bridging hydride dimers, enabling the reactive mesa-AdpMH monomers to react readily with available solvent or metallocene C-H bonds. Treatment of [rac-Me 2 Si{2-SiMe 3 -4-[2-(2-CH 3 -C 10 H 14 )]-C 5 H 2 } 2 Y(μ-Cl)] 2 ([ra c-Abp Y(μCl)] 2 ) with LiCH(SiMe 3 ) 2 affords the bulky yttrium-alkyl complex rac-AbpYCH(SiMe 3 ) 2 . This alkyl complex readily polymerizes ethylene, but does not react with cx-olefins. Hydrogenolysis ofrac-AbpYCH(SiMe 3 ) 2 generates the homochiral hydride dimer [rac-AbpY(μ-H)] 2 . The bulk of the methyladamantyl substituents inhibits formation of the heterochiral hydride dimer. Efforts are made to assess the ability of the methyladamantyl groups to inhibit the hydride monomer-dimer equilibria that deactivates previously reported group III Ziegler-Natta catalysts. The hydride dimer [rac-AbpY(μ-H)] 2 is an ex-olefin hydrogenation catalyst, but does not oligomerize or polymerize α-olefins. [rac-AbpY(μ-H)] 2 also reacts readily with dihydrogen, allene, and 2-butyne; examples of σ-bond metathesis, alkyne insertion, and β-hydride elimination have been observed for [rac-AbpY] complexes. The fluxional behavior exhibited by a series of d 0 -metallocene ƞ 3 -allyl complexes is investigated. Lineshape analysis of NMR studies performed on these complexes indicates activation barriers of ~8-14 kcal/mol for processes involving dissociation of the allyl C=C double bond as the rate determing step. Investigation of C s -symmetric linked scandocenes and yttrocenes allows elucidation of a second allyl rearrangement mechanism which operates concurrently with olefin dissociation: in-plane rotation rotation of the intact ƞ 3 -allyl ligand. Allyl rotation is generally fast relative to the rate of olefin dissociation. Ligand, metal, and solvent effects on the observed rates are discussed. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Chemistry) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Bercaw, John E. Thesis Committee: Grubbs, Robert H. (chair) Bercaw, John E. Lewis, Nathan Saul Myers, Andrew G. Defense Date: 6 January 1998 Record Number: CaltechTHESIS:09052025-171738932 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:09052025-171738932 DOI: 10.7907/2v44-pj33 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 17668 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 12 Sep 2025 10:43 Last Modified: 12 Sep 2025 11:36 Thesis Files PDF - Final Version See Usage Policy. 59MB Repository Staff Only: item control page

Design, Synthesis, Characterization, and Reactivity of Group III ansa-Metallocenes Containing Bulky Alkyl and Silyl Substituents Thesis by Michael B. Abrams In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy Division of Chemistry and Chemical Engineering California Institute of Technology Pasadena, California 1998 (Submitted January 6, 1998) 11 C 1998 Michael B. Abrams All Rights Reserved 111 For my parents lV Acknowledgments I have been very fortunate to have been able to work for and with John Bercaw during my stay at Caltech, and I am deeply appreciative of the support, guidance, and generosity that he has provided over the course of my graduate career. I consider myself very lucky to have been able to work for an advisor who granted me the freedom to explore what I found interesting, both in and out of the laboratory. I can not thank my family enough for their assistance and enthusiasm. My parents have been sources of unwavering support, and I am truly indebted to them for all that they have provided . I've also truly appreciated the kind thoughts and deeds of my grandparents. I would especially like to thank my brother for being a true confidant. He has always been there for me with a receptive ear and a much-enjoyed and appreciated perspective on just about everything. In addition, the almost frightening ability of Le Chevre to motivate me to drop everything and run to Tucson, Las Vegas, San Francisco, Berkeley, Oakland, and other locales in North America too numerous to mention have resulted in some of my most enjoyable memories and outrageous stories of the last several years. The Bercaw Mafia has provided a unique assortment of people with whom I am happy to have spent the last seven (?) years. Bryan Coughlin (SB#l) taught me all that I ever wanted to know about vacuum line techniques; I blame him personally for getting me started in organosilicon chemistry. Andy Kiely, in addition to being a terrific linemate and softball coach, is one of the few people who truly knows what does and what does not go into a whiskey sour (viva Diver Cocktails™!). It has also been a pleasure to share a vacuum line with Chris (Jag) Brandow, whose dance moves are overshadowed only by his fine singing voice. Many thanks to Alex Muci for not booting me off the vac line while he was getting started, and for reading the next to last installment of allyl misery. Donnie Cotter's Organometallics seminar was the first one I saw here at Caltech, and his unabashed mid-seminar belch certainly made me take notice V of the Bercaw group; he also took me on my first backpacking trip in the Sierras, and the beauty of Little Five Lakes Basin and the Kaweah Gap more than made up for me wanting to kill him during every step of the entire 3000+ foot slog up and over Black Rock Pass. Roger Quan taught me how to rule the Blue Box with an iron fist (glove?), and demonstrated the subtleties involved in throwing a softball from third to first base. I always enjoyed Eric Kelson's presence on backpacking trips and at softball games. I can't thank Sharad (Bahooka) Hajela enough for his help in and out of lab; he's a great person with whom to talk chemistry, and his home has provided fantastic food, usually way too much drink, and shelter from the occasional Hurricane. Bob Blake knew the importance of and introduced me to Big Wednesday, as well as Little Thursday. As well as being a fine chemist and a source of fruitful discussions, Tim (Vandalism is Cool) Herzog is a close friend and was a great roommate (may the Disco Party pictures survive forever!) and hiking partner. Thanks to Eva Birnbaum for helping motivate me to go to New Mexico, and to Shannon Stahl for relentless cheerfulness on backpacking trips and for debates on the nature of fish tacos. Many thanks to Susan (Donut Dilemma) Brookhart, my running partner, sometimes climbing partner, and faithful dream analyzer. It's been amazing to work in the same labs as Antek Wong-Foy, who is poised to lead the Bercaw Mafia into the 21st century; thanks for your enthusiasm both on and off of the 2nd floor of Noyes. My fourth year as a grad. student provided a litter of new Hogs: in addition to the previously mentioned WWF, we gained Cory Nelson and Steve Miller, with whom I have had the pleasure of sharing multiple Hogball innings. Jeff (A.J., Turbo, Sparky) Yoder, my partner in allyl crime, has been a great person with whom to discuss all things fluxional, preferably over a steaming trough of Phillips BBQ. It's been really fun sharing the NookTM with Deanna Zubris (despite her choice of calendars) and going to P-Chem seminars with Paul (Notorious) Chirik. Many thanks to Annita Zhong for not giving me grief about my progress writing props, and to Seva (Can You Dope It?) Rostovtsev for harassing me on a nearly daily basis about my progress writing props. VJ Numerous Postdocs have influenced my life here at Caltech. Gui Bazan introduced me to the wonders of the Dining Society, Gerrit Luinstra was in the San Francisco hotel room when Screaming Weasels were invented/ discovered/born, and Dave Antonelli showed me how the hell to find The Room. I really enjoyed trying to understand Jon (Limus) Mitchell talk chemistry over the roar of the vac lines and glove box and stereo back when we were in 201; also great fun was trying to understand him talk about anything in our frequent forays outside of Noyes. Eugene Mueller, Stephanie Chacon, and Antonio Pastor kept conversations in the Orange Room lively, and I'm indebted to Lin Wang for being a great Beer Coach, and to (Where's) Jim Gilchrist for discussions on all things organic, at BI social hours, and at parties in the toilet seat statue. Matt Holtcamp has killed, skinned, and eaten one of almost every animal on this planet, and it was fun having someone with whom to commiserate while interviewing for jobs. Cyrille Loeber added enantiomerically pure metallocenes to the confusion of allyl rearrangements. Thanks to Shigenobu Miyake for continuing to bring fame and fortune to compounds containing Adamantyl substituents, and to Dario Veghini for dinner parties, discussions about chemistry, runs to Lucky Baldwin's, and for aiding my fluency in Italian. Chris Levy arrived just in time to make and enliven this year's hike in the Sierras, and I hope that rest of the group drags him along on several more (or vice versa); I am forever indebted to him, as well, for braving Crenshaw Blvd. on New Years Eve for a mid-thesis Phillips run. There are also several people outside of my research group who I'd like to thank. Chris Kenyon and Ken Brameld have been fantastic climbing partners and occasionally as successful poker players, without whom grad school would have been no fun whatsoever. Mike (Aeropuerto) Woolf , Ken and Jen Butvill, Alison McCurdy, Chris Karp, Maria Khader, Eric Rubinstein, Arnel Fajardo, Tom (P) Shields, Dave (Slug) Lauterbach, and all the deadbeats who showed up for Poker Nights™ truly helped make life fun. Vll Abstract The preparations of a series of silyl-bridged bis(cyclopentadiene) complexes containing bulky alkyl and silyl substituents are reported. Incorporation of tertiary alkyl groups are required in order to direct trimethylsilyl groups exclusively adjacent to the silyl linkers. This 1,2,4- substitution is crucial for the directed syntheses of chiral an sa-metallocenes. Transmetallation of Me2 Si {2-SiMe3 -4-[2-(2-CH3 -C 1J-I 14 )]-C5 H 2 ) / affords exclusively racema- (chiral) Group III metallocenes, whereas transmetallation of Me2 Si{3-[2-(2-CH3 -C 1oH 14 ) ] C5 H 3 ) / (lacking the cx-TMS groups) proceeds with the opposite enantioselectivity, generating exclusively me sa- (achiral) metallocenes. Treatment of mesa- Me2 Si{3-[2-(2-CH3 -C 1J-I 14 )]-C5 H 3 ) 2 MCl-(LiCl)-(THF)2 (mesoAdpMCl-(LiCl)-(THF) 2 ) with allylmagnesium bromide affords mesa-AdpM(ri 3 -C3 H 5 ) (M = Sc, Y). Hydrogenolyses of these mesa-metallocenes generate highly reactive 14-electron d 0 metal hydride complexes. The methyladamantyl substituents on the Cp rings inhibit formation of bridging hydride dimers, enabling the reactive mesa-AdpMH monomers to react readily with available solvent or metallocene C-H bonds. Treatment of [r11c-Me2 Si {2-SiMe3 -4-[2-(2-CH3 -C 1oH 14 )]-C5 H 2 ) 2 Y(µ-Cl) ]2 ([ra c-Abp Y(µCl)lz) with LiCH(SiMe3 ) 2 affords the bulky yttrium-alkyl complex ra c-AbpYCH(SiMe3 ) 2 . This alkyl complex readily polymerizes ethylene, but does not react with cx-olefins. Hydrogenolysis ofrac-AbpYCH(SiMe3 ) 2 generates the homochiral h ydride dimer [racAbpY(µ-H)lz. The bulk of the methyladamantyl substituents inhibits formation of the heterochiral hydride dimer. Efforts are made to assess the ability of the methyladamantyl groups to inhibit the hydride monomer-dimer equilibria that deactivates previously reported group III Ziegler-Natta catalysts. The hydride dimer [rac -AbpY(µ-H)] 2 is an ex-olefin hydrogenation catalyst, but does not oligomerize or polymerize cx-olefins. [ra c-AbpY(µ-H)h also reacts readily with dihydrogen, allene, and 2-butyne; examples of CJ-bond metathesis, alkyne insertion, and ~-hydride elimination have been observed for [rac-Abp Y] complexes. 3 The fluxional behavior exhibited by a series of d0 -metallocene 11 -allyl complexes is investigated. Lineshape analysis of NMR studies performed on these complexes indicates activation barriers of -8-14 kcal/mol for processes involving dissociation of the allyl C=C double bond as the rate determing step. Investigation of C5 -symmetric linked scandocenes and yttrocenes allows elucidation of a second allyl rearrangement mechanism which operates 3 concurrently with olefin dissociation: in-plane rotation rotation of the intact 11 -allyl ligand. Allyl rotation is generally fast relative to the rate of olefin dissociation. Ligand, metal, and solvent effects on the observed rates are discussed. viii Table of Contents Dedication m Acknowledgments 1v Abstract v 11 Table of Contents v111 List of Figures 1x List of Tables xv List of Schemes xv11 Chapter 1 Discovery and Development of Homogeneous Ziegler-Natta Catalysis 1 Chapter 2 16 Design of Sterically Tailored bis-(Cyclopentadienyl) Ligand Arrays with Extensive ~-Alkyl Substituents for Syntheses of Group III ansaMetallocenes Appendix A: 1 H and 13C NMR Data for 19-54 Chapter 3 Reactivity of C5 - and C2 -Symmetric Group III Metallocenes with Dihydrogen and with Unsaturated Organic Substrates Appendix B: 1 H and 13C NMR Data for 16-25 Chapter 4 Investigation of Intramolecular Allyl (C3 H 5 ) Rearrangement Mechanisms and Evaluation of the Strength of Olefin Binding in d 0 -Metallocene - Alkyl - Olefin Complexes 78 92 136 142 lX Appendix C: Varaible Temperature 1 H NMR Data for 15-18 188 Appendix D: Selected gNMR Data 193 List Of Figures Chapter 1: Figure 1. Olefin insertion vs. cr-bond metathesis for Cp*2 Sc-CH3 . 6 Figure 2. The Cossee-Arlman mechanism for olefin insertion. 10 Figure 3. The original Green-Rooney mechanism. 11 Figure 4. The modified Green-Rooney mechanism. 11 Figure 1. Racemic and meso conformations of singly-substituted unlinked metallocenes (M = group IV). 18 Figure 2. Linked mixed-ring Cp complexes as chiral ansametallocenes. 19 Chapter 2: Figure 3. 24 Figure 4. Tailored ligand arrays for diastereoselective metallocene syntheses. 29 Figure 5. Initial cyclohexyl-substituted target ligand array. 30 Figure 6. Syntheses of cyclohexylcyclopentadieny1-trialkylborane adducts. 31 Figure 7. Dichlorodimethylislane fails to cleanly link 20-(BR3 ) 11 . 31 xi Figure 23. Syntheses of [2-(2-Methyl)-Adamantyl] cyclopentadienyls. 45 Figure 24. Unsuccessful attempts to generate 2-(2-methyl)-Adamantyl substituted analogues to 28 and 38. 46 Figure 25. Deprotonation of 42 affords exclusively [1,2-(SiMe3)i-4(MeAd)-C5 H 2 ]Li (44). 47 Figure 26. Syntheses of H 2 Adp (45), Li2 Adp (46), H 2 Abp (47). 48 Figure 27. Deprotonation of H 2 Abp (45) affords a single constitutional isomer of the desired 2-(2-methyl)-Adamantyl substituted ligand array. 48 Figure 28. 500 MHz H NMR spectra of a) H 2 Abp (45) (C6 D6 ); 49 Figure 29. 500 MHz 1 H NMR spectra of K2 Abp (47) (THF-d 8 , with added 18-crown-6). 50 Figure 30. Formation of exclusively mesa- AdpMCl-LiCl-(solvent)z complexes; M = Sc (48), Y (49). 52 Figure 31. Metallation using the [Abp] dianion results in exclusively racemo- metallocenes (M = Sc or Y). 54 Figure 32. Metalation of 47 affords rac-AbpMCl-(KCl)-(THF)z, which is subsequently converted to rac-AbpMCl(THF) and [rac-AbpM(µ-Cl)]z (M=Sc, Y). 55 Figure 1. Insertion of allene to give 17 3 -allyl and 17 3-methallyl complexes. 94 Figure 2. cr-bond metathesis of [Cp*2 Y(µ-H)]z with allene. 95 1 Chapter 3: Xll Figure 3. The relatively uncongested (C5 H 5 )i ligand array permits formation of bimetallic complexes inactive for alkyne oligomeriza tion. 97 Figure 4. Isospecific Ziegler-Natta polymerization catalysts. 100 Figure 5. Transition states for olefin insertion into Y-H and Y-CH2P bonds of (R,S)-BnBpY 101 Figure 6. Group III bridging hydride dimers. 102 Figure 7. Monomeric/homochiral-/heterochiral hydride dimer equilibria 103 Figure 8. The results of increased steric crowding on formation of (µ-Hh dimers. 104 Figure 9. Intra- and intermolecular ligand C-H activation. 105 Figure 10. Only small quantities of (µ-H) 2 dimers are observed upon hydrogenolysis of meso-AdpM(C3H 5 ) (M=Sc, Y) 109 Figure 11. Possible intramolecular C-H bond activation. 110 Figure 12. One of several possible dimers resulting from intermolecular C-H activation. 110 Figure 13. Possible equilibria involving mes o-Ad pM-H, trimethylphosphine, and dihydrogen (M = Sc, Y). 111 Figure 14. 500 MHz 1 H NMR spectra (C6D 6 ) of rac-Abp YCH(SiMe3 )i 113 (18). Figure 15. rac-Abp YCH(SiMe3)i (18) polymerizes ethylene but does not react with a-olefins. 114 Figure 16. 500 MHz 1 H NMR spectra (cyclohexane-d d of [rac-AbpY(µ-H)h (20). 116 Figure 17. Allene undergoes a-bond metathesis with 19. 121 xiii 1 Figure 18. 500 MHz H NMR spectrum (C6 D6 ) of rac-AbpYCH=C=CH2 (21); * = residual H 2 C(SiMe3 )i from hydrogenolysis of 18. 122 Figure 19. Formation of an equilibrating mixture of propargyl and 123 allenyl complexes are observed from the reaction of 2-butyne with CnY(CH3h- Figure 20. Alkyne insertion into 19, followed by ~-H elimination and subsequent a-bond metathesis generates a pair of diastereomeric yttrium-methylallyl complexes. 124 Figure 21. 300 MHz 1 H NMR spectra (cyclohexane-dd of racAbp YCH=C=CHMe (25a,b) (18). 125 Figure 22. Y-C bond rotation for the diastereomeric pair 25a,b. 126 Chapter 4: Figure 1. Transition metal-alkyl-alkene complexes as possible 144 intermediates in Ziegler-Natta polymerization systems (M = 14electron, d 0 : Ti\ Zr+, Ht, Sc, Y, La; R = -H, -CH3 , -CH2 CH3 , etc.; P = growing polymer chain) . Figure 2. Spectroscopically characterized d 0 metal-alkyl-alkene complexes. 145 Figure 3. Use of 173 -allyl species as model complexes for estimating the strength of M-olefin interactions in metal-alkyl-alkene complexes. 145 Figure 4. Proposed allyl rearrangement mechanisms: (i) dissociation 146 1 of the allyl C=C double bond to generate an 17 -allyl complex, rotation about the allyl C-C single bond, and recoordination of the allyl alkene fragment, (ii) a 180° in-plane rotation of the coordinated allyl ligand. Figure 5. d 0 Group IV metallocene 17 3 -allyl complexes . 148 XlV Figure 6. Related metallocene d 0f1 Ln complexes. 149 Figure 7. Group III metallocene allyl complexes investigated. 150 Figure 8. Variable temperature 1 H NMR spectra of Cp*2 Sc(C3H 5 ) (15) (500 MHz, toluene-d 8 ). 152 Figure 9. gNMR simulation of Cp* and (C3H 5 ) resonances of 15 using 17 3 -17 1 processes (ia) and (ib ). 155 Figure 10. Eyring plot for olefin dissociation for 15 in toluene-d 8 . 157 Figure 11. Origin of statistical factor included in two-site exchange approximation. 158 Figure 12. Frontier molecular orbital diagram for Cp2 M(C3H 5) (M=Sc, Y). 159 Figure 13. Variable temperature H NMR spectra of AdpSc(C3H 5 ) (16) (500 MHz, Et2O-d 10 ). Figure 14. gNMR simulation of (C3H 5 ) and vinylic Cp resonances 167 3 1 of 16 (500 MHz, Et2 O-d 10 ) using only 17 -17 processes (ia) and (ib). Figure 15. gNMR simulation of (C3H 5 ) and vinylic Cp resonances of 16 (500 MHz, Et2O-d 10 ) using only 17 3 -rotation. 168 Figure 16. gNMR simulation of (C3H 5 ) and vinylic Cp resonances 1 of 16 (500 MHz, Et2O-d 10) using 17 3-17 processes (ia) and (ib) and 17 3 -rotation. 169 Figure 17. Possible 173 -rotation transition states for meso-metallocenes. 175 Figure 18. Variable temperature H NMR spectra of rac-IpSc(C 3H 5 ) (20) (500 MHz, toluene-d 8 ). 177 Figure 19. gNMR simulation of (C3H 5 ) and vinylic Cp resonances of 3 1 20 (500 MHz, toluene-d 8 ) using only 17 -11 processes. 180 1 1 159 xv LIST OFTABLES Chapter 2: Table 1. Metallation enantioselectivities for a series of ethylene-bridged group IV metallocenes 21 Table 2. Racemic/meso ratios for tetramethylethano-bridged group IV metallocenes 22 Table 3. Racemic/meso ratios for dimethylsilyl bridged group IV meta 11 ocenes. 23 Table 4. Observed group III enantioselectivities for dimethylsilylbridged ligands. 25 Table 5. Racemic/meso ratios for additional dimethylsilyl bridged group IV metallocenes. 26 Table 6. Racemic/meso ratios for indene- and indenyl-based ansametallocenes. 27 Table 1. Proton NMR assignments for 19-54 79 Table 2. Proton decoupled 13C assignments for 53 91 500 MHz 1 H NMR chemical shifts and one-bond Y-H coupling constants for the bridging hydride resonance of [rac-AbpY(µ-H))i (20). 115 Appendix A: Chapter 3: Table 1. Appendix B: XVI Table 1. Proton NMR assignments for 16-25 133 Table 2. Proton decoupled 13C NMR assignments for 18, 20 137 Table 1. Observed coalescence temperatures and gNMR-derived 17 3 -17 1 rate constants for 15 (toluene-d 8) 156 Table 2. Observed coalescence temperatures for 15 (THF-d 8 ) and gNMR-derived 17 3 -17 1 rate constants 158 Table 3. Olefin dissociation barriers form es o- metallocenes 170 Table 4. Activation barriers to 173 -rotation activation, obtained from two-site exchange approximation (I) 171 Table 5. Activation parameters for olefin dissociation, and 298 K rate constants obtained from extrapolation of estimated liHt liSt • 172 Chapter 4: f Table 6. Activation parameters for in-plane rotation form es o172 metallocenes obtained from Eyring plots of low temperature coalescence data Table 7. In-plane rotation barriers for 16 - 19, extrapolated to 298 K 173 Table 8. Olefin dissociation activation energies for race mometallocenes 181 1 Appendix C: Variable Temperature H NMR Data for 15-18 Table 1. Variable temperature 300 and 500 MHz 1H NMR data for Cp*2 Sc(C3H 5 ) (15) 188 Table 2. Variable temperature 500 MHz 1 H NMR data for mesoAdpSc(C3H 5 ) (16) 189 xvii Table 3. Variable temperature 500 MHz 1 H NMR data for m es oAdp Y(C3H 5 ) (17) 190 Table 4. Variable temperature 500 MHz 1 H NMR data for mes oDpSc(C3H 5 ) (18) 192 Appendix D: Selected gNMR Data Table 1. gNMR Parameters for Cp*zSc(ri 3-C3H 5 ) (15), toluene-d 8, 500 MHz Table 2. gNMR Parameters formeso-AdpSc(ri -C H (16), Et O-d 500 MHz, Allyl and Vinylic Cyclopentadienyl Protons Table 3. gNMR Parameters for rac-IpSc(17 3 -C3H 5 ) (20), toluene-d 8, 500 MHz, Allyl and Vinylic Cyclopentadienyl Protons 195 Table 4. gNMR Parameters for rac-IpSc(ri -C H 500 MHz, Isopropyl Protons 196 193 3 ) 3 3 3 ) 5 5 , 2 (20), toluene-d 10 , 8 194 LIST OF SCHEMES Chapter 3: Scheme 1. Generation of cationic group IV M-alkynyl complexes. 95 Scheme 2. Catalytic olefin oligomerization by Cp*2 M-alkynyl complexes (M=Sc, Y, Zr+). 96 Scheme 3. Insertion of internal alkynes. 98 Scheme 4. Syntheses of ri 5 -pentadienyl complexes. 98 Scheme 5. Organolanthanide catalyzed dimerization of internal alkynes 99 XVlll Scheme 6. Dimer dissociations and a-bond metathesis steps required to account for H/D scrambling into 20. 117 Scheme 7. Olefin hydrogenation catalyzed by [rac-AbpYH] (19). 119 Scheme 8. H/D scrambling is not observed between 20-d 2 and perprotio 1-olefins (R=CH2 CH2C3 , CMe3 ). 118 Chapter 4: 1 154 17 3 -17 Allyl rearrangements in mes o-metallocenes. 1 164 Scheme 3. 17 3 -Rotation form es o-metallocenes. 165 Scheme 4. 180° In-plane rotation for C2 -symmetric nnsa-metallocenes. 178 Scheme 5. Implications of olefin dissociation and subsequent M179 and C-C single bond rotations for C2 -symmetric metallocenes. 3 Scheme 1. 17 -17 Processes for Cp\Sc(C3H 5 ) (15). Scheme 2. 1 Chapter 1 Discovery and Development of Homogeneous Ziegler-Natta Catalysis Abstract 2 Introduction 3 References and Notes 13 2 Abstract A brief introduction to the discovery of and recent developments in the field of Ziegler-Natta catalysis is provided in order to provide context for the experiments described in Chapters 2, 3, and 4. 3 Introduction The discovery that group IV transition metal-halides and certain aluminum alkyls may be used to catalyze the polymerization of ethylene was made by Karl Ziegler more than 40 years ago.l The extension of this type of catalyst mixture to the polymerization of cx-olefins was made shortly thereafter in the laboratories of Giulio Natta.2 Since then, intense research on heterogeneous ex-olefin polymerization systems have lead to the development of catalysts with remarkable activities and regio- and stereochemical selectivities; billions of pounds of polyolefins are produced annually using these Ziegler-Natta catalysts.3 Complicating characterization of the active site of and detailed mechanistic studies on these systems are the heterogeneous nature of the polymerization reactions and the fact that the catalysts are often poorly-defined mixture of MgC12 -supported metal halide, trialkylaluminum, and ester or amine donor species. Efforts to develop homogeneous catalysts that rivaled the activities and selectivities of the heterogeneous systems, yet which proved more amenable to mechanistic investigation and rational catalyst design, were prompted by the observation that solutions of titanocene dichloride and alkyl aluminum species could be used to catalyze the polymerization of ethylene.4,5 A major breakthrough in the development of homogeneous catalysts was the discovery by Sinn and Kamisnky6 that the addition of methylalurninoxane (MAO) to any of a variety of group IV metallocenes generated highly active olefin polymerization catalysts. The MAO cocatalyst, generated by the addition of exactly one equivalent of water to trimethylaluminum, presumably consists of a mixture of linear and cyclic oligomers containing the -Al(O)(CH3 )- repeat unit? &.7 \ r~~ n+2 I R ~ [Al(O)(CH3 )]x R R n R 4 Another significant advance in the use of metallocenes as ZieglerNatta catalysts was the finding by Brintzinger8 that chiral metallocenes could be utilized as the precursors to catalysts for the stereospecific polymerization of propylene and other a-olefins. The use of rigid C2 -symmetric group IV metallocenes, such as racemo-ethylenebis(tetrahydroindenyl)zirconium dichloride (rac-EBTHIZrC12, 1), in conjunction with sufficient MAO to afford ~1000-fold excesses of Al:Zr were found to catalyze the polymerization of propylene to isotactic polypropylene. Both the activity and stereospecificity of this system approached those of the commercially employed heterogeneous systems. (1) nd------- Isotactic poly(propy lene) Tm=170°C Hard, translucent nd (2) [Al(O)(CH3)]x Atactic poly(propylene) Low melting Amorphous, few uses One drawback to the use of this type of chiral catalyst was the fact that isolation of the racemo- isomers was a prerequisite for the production of isotactic polypropylene. The achiral mesa- conformer (2), which was typically formed in two-fold to ten-fold excesses along with the desired racemometallocenes, generate atactic polypropylene. Although methods were 5 developed for the purification and isolation of exclusively the C2-symmetric isomers, these separations were typically tedious and resulted in significantly diminished yields of available catalyst precursors. Investigations into the electronic requirements for the generation of active homogeneous metallocene catalysts indicate that part of the role of the MAO cocatalyst is to generate a cationic 14-electron d 0 -metallocene alkyI.9-11 Watson first demonstrated that 14-electron d 0 -metal complexes could act as ethylene polymerization catalysts.12 The use of Cp*2 LuCH3 , which is isolectronic with cationic [Cp2 M-Rt (M=group IV), to demonstrate olefin insertion, ~-H elimination, ~-methyl elimination, and a-bond metathesis showed that neutral metallocenes could be used as model complexes for the fundamental steps involved in olefin polymerization, supporting the notion of 14-electron d 0 active sites for the group IV systems. Further support for the involvement of 14-electron metal centers as the active polymerization catalysts came from research in which a single alkyl substituent was removed from neutral group IV metallocenes to generate complexes of the form [Cp2 M-Rt [Ar (M=Ti, Zr, Hf; A="weakly coordinating" counterion), which initiate olefin polymerization. Convenient methods for the removal of an alkyl group involve protonolysis by dimethylanilinium (C6H 5 NH(CH3 h) salts, first developed by Turner,13-15 or by abstraction using the highly Lewis acidic trityl cation ((C6H 5 hC+)_l6,17 Fluorophenylborate counterions (B(Arp) 4-) such as [B(C6F5 ) 4 r 14,l8 and [B(C6F4SiR3\]-19 have proven less prone to side reactions with or deactivation of the metal cation than have conventional counterions such as [BF4L [PF6L and [BPh4 T-20 The highly Lewis acidic B(C6F5h may be alternately be used to generate complexes such as 3 which are active for ethylene and propylene polymerization.21,22 6 Uncertainties in the roles of the various counterions in the initiation, propagation, and termination of polymerization by the [Cp2 MRr [B(ArF) 4 catalysts are obviated by the use of isoelectronic neutral group III metallocene complexes as Z-N catalysts. Previous work in the Bercaw group has shown that scandocene and yttrocene derivatives may act as single component catalysts for the polymerization of ethylene and the stereospecific r polymerization of a-olefins. At low temperatures, the permethylscandocene methyl complex 4 was shown to be an active catalyst for the living polymerization of ethylene.23,24 Although4 reacts with propylene, a-bond metathesis results in the generation of methane and the vinylic complex 5. Unfavorable steric interactions between the bulky pentamethylcyclopentadienyl ligands and the alkyl group of incoming monomer explain the inability of Cp*2 Sc-CH3 to oligomerize, or even insert, propylene (Figure 1). (4) (5) Figure 1. Olefin insertion vs. a -bond metathesis for Cp*2 Sc-CH3 (4) . The use of dimethylsilyl linked cyclopentadienyls to reduce the effective steric congestion at the metal center has been demonstrated by the ability of Me2Si(C5 Me4 )iSc(H)PMe3 (6, OpScH(PMe3 )) to catalyze the head-totail dimerization of a-olefins and the cyclization of a,CD-dienes.25 A similar reactivity towards olefinic substrates was observed for [m es o-Me2 Si(3-CMe3C5H 3)iSc(µ-H)b (7, [mes o-DpSc(µ-H)i]), whose reduction in steric crowding at 7 the metal center is manifested by the dimerization of the electron-deficient monomeric Sc-H species responsible for catalysis. Further reductions to the bulk of the ancillary ligand array was achieved by replacement of one of the cyclopentadienyl rings with an alkylamido fragment, resulting in derivatives of Me2 Si(C5 Me4 )[N(CMe3 )] (abbreviated Cp*SiNR) such as 8 ([Cp*SiNRSc(µ-H)(PMe 3 )h and 9 ([Cp*SiNRSc(µ-CH 2CH2CH2CH3)h. 26, 27 The Cp*SiNRSc systems polymerize ethylene and propylene at room temperature, and were the first examples of single component ex-olefin polymerization catalysts. The group IV derivatives utilizing the Cp*SiNR and related ligand arrays have generated considerable interest as possible commercial ethylene/1-alkene copolymerization catalysts.28-30 Me3C\ N M~ ~ _ _,,,PMe 3 / ""-H--7 ~ Me3P / N / i-Me \ CMe3 (8) The introduction of additional silyl substituents to the cyclopentadienyl rings can be used to generate single component isospecific olefin polymerization catalysts. The C2 -symmetric yttrocene 10 ({rac-[Me2 Si(2SiMe3-4-CMe3 -C5H 2hY(µ-H) h, [Bp Y(µ-H)h) catalyzes the stereospecific polymerization of propylene, 1-butene, 1-pentene, and 1-hexene to isotactic poly( cx-olfins) _31 8 n+2 10 R R ~R n Synthesis of the [Bp] ligand array proved to be a major breakthrough in the construction of chiral metallocenes, as exclusively racemo-metallocenes are formed upon metallation of [Bp ]2- (M=Sc, Y, La, Ti, Zr),32,33 thus eliminating the need for tedious separation of race mo - from undesired m es o conformers. Unfavorable steric interactions between the trimethylsilyl groups in the narrow portion of the metallocene wedge are believed to direct the syntheses of exclusively C2 -symmetric (race mo) metallocenes. 9 Me.... Me"'"" Me3 Cl CMe 3 ···\,( ~CMe3 ) .... / / "; · -CI Me )S( CMe3 racemo-AbpMCl meso-AbpMCl (not observed) The use of trimethylsilyl groups adjacent to a silyl- linking unit may be employed to direct the formation of racemo-metallocenes incorporating linkers other than the dimethylsilyl moiety in [Bp]. The use of an enantiomerically pure 1,1'-bi-2-naphthol to generate a chiral linking unit resulted in the directed synthesis of diastereomerically pure metallocenes .34 In addition to the TMS-TMS interactions in the narrow portion of the metallocene wedge preventing formation of mes o-metallocenes, steric interactions between the TMS groups and the 3,3' -methine positions of the naphthalene rings are thought to induce enantioselective metallation of the [(C 2Ji 120 2 )Si(2-SiMe3 -4-CMe3 -C5H 2hf- ([BnBp )2-) ligand array. 10 R,S-[BnBp]Y-R S,S-[BnBp ]Y-R (11) (not observed) Initial mechanisms for the C-C bond forming step in olefin polymerization invoked direct addition of the olefin across the M-alkyl bond (Figure 2), the so-called Cossee-Arlman mechanism.35,36 This mechanism, however, did not account for the empirical observation that two vacant metallocene orbitals appeared to be a prerequisite for olefin insertion; although they contain d 0 metal centers, 16-electron metal dialkyls such as Cp 2 Zr(CH3 h do not insert olefins. Figure 2. The Cossee-Arlman mechanism for olefin insertion. Rooney and Green37,38 noted the availability of multiple oxidation states for the group V metals, and proposed a mechanism in which there is aelimination from a M-alkyl species; subsequent cyclization and reductive elimination would serve to regenerate a propagating metal-alkyl (Figure 3). The fact that cationic group IV and neutral single component group III and lanthanide metallocene complexes, which can not undergo the formal oxidation resulting from a-elimination, are active polymerization catalysts efficiently refutes this proposed mechanism. 11 H Cp,,,,, 1_,,,H ··M-C c P_,.-"'CH P 2 ~.,,,.,.--CH2 P H 2c Cp,,,,,. \ Cp_,.--M"'- / CHR C ;" \ HH Figure 3. The original Green-Rooney mechanism (M=Group V). In a modification to the Green-Rooney mechanism, Green and Brookhart39 proposed that one vacant orbital was required for olefin coordination and that a second vacant orbital accommodates an cx-agostic interaction which facilitates olefin insertion into the metal-alkyl bond (Figure 4) . The requirement for cx-agostic assistance in olefin insertion has generated considerable debate. While Bercaw,40,41 Brintzinger,42,43 and Stille44 have developed systems which provide evidence for cx-agostic interactions, other experiments devised in the laboratories of Grubbs45 and Brintzinger43 show no evidence for such interactions, and implicate a transition state such as A as that for olefin insertion. Figure 4. The modified Green-Rooney mechanism. One drawback to mechanistic studies on 7-11 results from the tendency of the highly electron deficient metal hydrides and alkyls to dimerize. Although the bridging hydride dimers 10 and that of 11 promote the 12 stereospecific polymerization of a-olefins, the polymerizations are slow; typical reactions produce enchainment of several hundred monomer units after days in neat olefin. This sluggishness is attributed to the low concentration of 14-electron monomeric Y-H species in solution, a direct result of the monomer-(µ-H)i dimer equilibria which heavily favor the inactive 16-electron dimers. It has been shown that bridging d 0f 1 - metalhydrides react ~ 108 -10 10 times slower than the corresponding terminal hydrides.46 The equilibria for metallocenes incorporating tert-butyl substituents in the open portion of the wedge (such as 7, 10, and 12) is sufficiently weighted towards the dimer so that monomeric Y-H species are never observed in solution by 1H NMR spectroscopy. 2 (never observed directly) Although the advent of single component catalysts allowed kinetic studies on systems that afforded atactic polyolefins, similar investigations had not been performed on isospecific systems. The availability of a monomeric C2 -symmetric Y-H species would be invaluable for such mechanistic studies. Rates of processes such as olefin insertion into metal hydride (initiation), insertion into metal alkyls (propagation), ~-H or ~-Me elimination (termination) could then be measured directly by techniques such as NMR spectroscopy. In addition, insertion of an unsymmetric gem-disubstituted olefin to a M-H species would result in a chiral alkyl unit in a chiral metallocene framework, ideal for characterization by techniques such as XRay diffraction. The stabilities of the previously observed (µ-H)i dimers had prevented studies such as those described above. The replacement of the tert-butyl groups in the open portion of the metallocene wedge with more extensive alkyl substituents was proposed to be a means of perturbing the monomer-dimer equilibria to disfavor formation 13 of the (µ-Hh dimers, thus favoring formation of the active 14-electron monomers. Crucial to the directed syntheses of C2 -symmetric metallocenes, however, was the placement of the trimethylsilyl substituents adjacent to the silyl linker; the results of such changes in alkyl substitution on ligand syntheses, and the abilities of the resultant metallocenes to insert a-olefins, remained to be investigated. Chapter 2 describes the syntheses of such ligand arrays incorporating cyclohexyl-, 2-adamantyl-, and 2-(2-methyl)-adamantyl- substituents, and the contents of Chapter 3 concern the elucidation of the ability of 2-(2-methyl)adamantyl substituents to perturb monomer-dimer equilibria and prevent formation of (µ-H) dimers. Chapter 4 is a mechanistic study on allyl rearrangement processes that arose initially from the characterization of several complexes derived from ligand arrays incorporating methyladamantyl groups. References and Notes 1. Ziegler, K.; Holzkamp, E.; Breil, H.; Martin, H. Angew. 541. Chem. 1955, 67, 2. Natta, G. Angew. 3. Thayer, A. N. Chem. Eng. News. 1995, 73, 15. 4. Natta, G.; Pino, P.; Mazzanti, G.; Giannini, U. J. Am. Chem. Soc. 1957, 79, 2975. 5. Breslow, D. S.; Newburg, N. R. J. Am. Chem. Soc. 1957, 79, 5072. 6. Andersen, A.; Cordes, H. G.; Herwig, J.; Kaminsky, W.; Merck, A.; Mottweiler, R.; Pein, J.; Sinn, H .; Vollmer, H. J. Angew. Chem., Int. Ed. Engl. 1976, 15,630. 7. Mason, M. R.; Smith, J. M.; Bott, S. G.; Barron, A. R. J. Am. Chem. Soc. 1993, 115, 4971. 8. Kaminsky, W.; Kulper, K.; Brintzinger, H. H.; Wild, F. R. W. P. Angew. Chem. Int. Ed. Engl. 1985, 24, 507. Chem. 1956, 68,393. 14 9. Jordan, R. F. Adv. 10. Long, W. P.; Breslow, D. S. Liebigs. Ann. Chem. 1975, 463. 11. Harlan, C. J.; Bott, S. G.; Barron, A. R. J. Am. Chem. Soc. 1995, 117, 6465. 12. Watson, P. L. J. Am. Chem. Soc. 1982, 104,337. 13. Hlatky, G. G.; Turner, H. W.; Eckman, R. R. J. Am. Chem. Soc. 1989, 111, 2728. 14. Turner, H. H. Eur. Pat. Appl. 1988, 277004. 15. Hlatky, G. G.; Turner, H. W. Eur. Pat. Appl. 1988, 277003. 16. Ewen, J. A.; Elder, M. J. Eur. Pat. Appl. 1991, 426637, 426638. 17. Chien, J. C. W.; Tsai, W.-M.; Rausch, M. D. J. Am. Chem. Soc. 1991, 113, 8570. 18. Turner, H. W. Chem. 19. Lia, L.; Yang, X.; Ishihara, A.; Marks, T. J. Organometallics 1995, 14, 3135. 20. Bochmann, M. J. Chem. Soc., Dalton Trans. 1996, 255. 21. Yang, X.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1991, 113, 3623. 22. Yang, X.; Stern, C. L.; Marks, T. J. J. Am. Chem . Soc. 1994, 116, 10015. 23. Thompson, M. E.; Baxter, S. M.; Bulls, A. R.; Burger, B. J.; Nolan, M. C.; Santarsiero, B. D.; Schaefer, W. P.; Bercaw, J. E. J. A1n. Chem. Soc. 1987, 109,203. 24. Burger, B. J.; Thompson, M. E.; Cotter, W. D.; Bercaw, J. E. J. Am. Chem. Soc. 1990, 112, 1566. 25. Bunel, E. E., Ph.D. Thesis, California Institute of Technology, 1989. 26. Shapiro, P. J.; Bunel, E. E.; Schaefer, W. P.; Bercaw, J. E. Organometallics 1990, 9 867. Organomet. Chem. 1991, 32,325. Abstr. 1989, 110, 58290a. I 27. Shapiro, P. J.; Cotter, W. D.; Schaefer, W. P.; Labinger, J. A.; Bercaw, J. E. J. Am. Chem . Soc. 1994, 116, 4623. 15 28. Canich, J. M. Eur. Pat. Appl. 1990, 420436. 29 . Canich, J. M.; Hlatky, G. G.; Turner, H. W. U. S. Patent 1990, 542236. 30. Stevens, J. C.; Timmers, F. J.; Wilson, D. R.; Schmidt, G. F.; Nickias, P. N.; Rosen, R. K.; Knight, G. W.; Lai, S. Eur. Pat. Appl. 1990, 416815. 31. Coughlin, E. B.; Bercaw, J. E. J. Am. Chem. Soc. 1992, 114, 7606. 32. Coughlin, E. B., Ph.D. Thesis, California Institute of Technology, 1994. 33. Chacon, S. T.; Coughlin, E. B.; Henling, L. M.; Bercaw, J. E. J. Organomet. Chem. 1995, 497, 161. 34. Mitchell, J. P.; Hajela, S.; Brookhart, S. K.; Hardcastle, K. I.; Henling, L. M.; Bercaw, J.E. J. Am. Chem. Soc. 1996, 118, 1045. 35. Cossee, P. J. J. Catal. 1964, 3, 80. 36. Arlman, E. J. J. Cata!. 1964, 3, 99. 37. Brookhart, M.; Green, M. L. H. J. Organomet. Chem. 1983, 250, 395. 38. Laverty, D. T.; Rooney, J. J. J. Chem. Soc. Faraday Trans. 1983, 79,869. 39. Brookhart, M.; Green, M. L. H.; Wong, L. Prag. Inorg. Chem. 1988, 36, 1. 40. Piers, W. E.; Bercaw, J. E. J. Am. Chem. Soc. 1990, 112, 9406. 41. Piers, W. E.; Kohn, R. D.; Herzog, T. A.; Bercaw, J. E .. Manuscript in preparation . 42. Leclerc, M. K.; Brintzinger, H. H.J. Am. Chem. Soc. 1995, 117, 1651. 43. Brintzinger, H. H.; Krauledat, H. Angew. Chem. Int. Ed. Engl. 1990, 29, 1412. 44. Barta, N. S.; Kirk, B. A.; Stille, J. R. J. Am. Chem. Soc. 1994, 116, 8912. 45. Clawson, L.; Soto, J.; Buchwald, S. J.; Steigerwald, M. L.; Grubbs, R. H.J. Am. Chem. Soc. 1985, 107, 3377. 46. Stern, D.; Sabat, M.; Marks, T. J. J. Am. Chem. Soc. 1990, 112, 9558. 16 Chapter 2 Design of Sterically Tailored bis-(Cyclopentadienyl) Ligand Arrays with Extensive ~-Alkyl Substituents for Syntheses of Group III ansaMetallocenes Abstract 17 Introduction 18 Results and Discussion 30 Con cl us ions 55 Experimental 56 References and Notes 74 Appendices 78 18 Introduction The appeal of using chiral organometallic compounds as catalysts for enantioselective transformations has prompted the development of a wide variety of chiral metallocenes. Chiral early transition-metal cyclopentadienylbased complexes have found considerable success as stereospecific ZieglerNatta a-olefin polymerization catalysts.11 Similar Cp-based compounds have proven successful in promoting transformations related to organic synthesis, such as the asymmetric hydrogenation of imines,2,3 enamines,4 and olefins,5 and the enantioselective hydrosilylation of ketones.6 From the search for chiral catalysts have evolved several strategies for the construction of chiral Cp-based organometallic complexes. Unlinked substituted cyclopentadienyls, such as those depicted in Figure 1, may adopt both chiral and achiral conformations. The freedom of the Cp rings to rotate about the M-centroid axis permits interconversion between C2 -symmetric (and chiral) structures, in which the Cp substituents are staggered, and C5 -symmetric (and achiral) structures, in which the Cp substituents adopt eclipsing conformation. Equilibration between these limiting structures is generally rapid on the NMR timescale, thus diminishing the possible stereodirecting effects of the chiral conformer. A judicious choice of Cp substituents, however, may be used to exploit this racemo mesa Figure 1. Racemic and meso conformations of singly-substituted unlinked metallocenes (M = group IV). chiral-achiral interconversion to observe new and potentially useful reactivities; if exchange between the C2 and C5 structures is slow on the polymerization timescale, polymers containing blocks of both isotactic and atactic sequences may be produced. Erker7-9 has used chiral substituents on 19 Cp rings, and Waymouthl0,11 has employed achiral phenylindenyl ligands, to generate catalysts that appear to function in this manner. Linking the two cyclopentadienyls with a short tether effectively inhibits these rotations. Organometallic complexes with such tethered linked arrays are commonly referred to as "a nsa" -metallocenes.12 Incorporation of a single substituent onto one of the Cp rings results in an overall C1 -symmetric complex (Figure 2). Marks has successfully utilized chiral R substituents (menthyl, neomenthyl, phenylmenthyl) in order to induce diastereoselective metallation in a variety of metallocenes; the group III, group IV, and lanthanide metallocenes. Derivatives of these C1 -symrnetric complexes have found use as olefin hydrogenation, hydroamination / cyclization,13,14 and polymerization15 catalysts. Figure 2. Linked mixed-ring Cp complexes as chiral ansa- metallocenes. Although additional methods have been developed for generation of chiral metallocenes, 16-18 the most prevalent means is through the use of a short tether to connect a pair of symmetrically substituted Cp rings. Incorporation of a transannular linker does not necessarily result in the formation of chiral complex, as C5- (mesa) and C2 -symrnetric (rac) isomers of a given ansa-metallocene are possible. Major drawbacks in the use of ansametallocenes are the facts that both rac- and mesa- complexes are typically formed upon metallation, and that separation and isolation of the racenwisomer is required for asymmetric catalysis. Modification of both the inter-Cp tether and the Cp substituents themselves have been employed in order to encourage formation of the rac- isomer. 20 The catalyst originally reported by Brintzinger for the stereospecific polymerization of propylene to isotactic poly(propylene) utilizes the ethanobridged bis(tetrahydroindenyl) zirconium dichloride 1. Metallocenes incorporating such a 2-carbon linker are particularly attractive because of the ease with which the linker is incorporated into the ligand framework. Synthesis of bridged metallocenes of this sort typically involve treatment of two equivalents of the appropriate Cp anion with 1,2-dibromoethane. Although the 2-atom interannular bridge significantly hinders rotation of the Cp rings between rac and mes o isomers, several energetically discernible conformations of -(CH2 CH2)- linked metallocenes have been observed crystallographically and detected using dynamic NMR techniques;l9,20 interconversion between the various isomers is generally rapid in solution. The use of a transition metal halide and the alkali metal salt of the appropriate ligand is commonly used to generate ansa- metallocenes.21 These metallations typically generate both rac and mes o products, the exact ratio of which is often dependent not only on the particular metal and ligand substituents, but also the workup procedure. For ethano-bridged Cps containing only ~-alkyl substitutents,22 the selectivity for formation of racemo- metallocenes is generally poor, with equimolar quantities of rac and mes o complexes formed irrespective of the bulk of the ~-alkyl group (Table 1, R = -CH3 , -CH2CH3 , -CH(CH3h, -C(CH3h). 21 R Rp Rp M"'"Cl '°'Cl M··" Cl '°'Cl 1 a Rp Rp racemo mesa 2 M Ra Rp Synthetic Photostationary rac 1neso rac meso 1 1.3 * * 1 1.5 * * 2a22 Ti H 2b 22 Ti H CH3 CH2 CH3 2c22 Ti H CH(CH3)i 1 1.8 * * 2d22 Ti H C(CH3h 1 2.0 * * 2e23 Ti 1 2.5 1 Ti CH3 CH2 CH3 2 2£23 CH3 CH3 2.6 1 4 1 2g23 Ti CH3 C(CH3h 1.6 1 15 1 Table 1. Metallation enantioselectivities for a series of ethylene-bridged group IV metallocenes. * = not investigated The substituent adjacent to the linker is able to exert considerably more influence on the metallation enantioselectivity than is the ~- substituent, as observed by incorporation of methyl groups a- to the ethano linker.23 These ratios may occasionally be further enhanced by photoisomerization of the original mixtures, permitting isolation of pure rac emo- complexes by fractional crystallization or by column chromatography. Incorporation of a more sterically demanding linker results in only a modest improvement in the observed rac:m es o ratios, as shown for a related series of tetramthylethano-bridged titanocenes and zirconocenes (Table 2).24 The effects of different reaction conditions and workup procedures are illustrated by the range of multiple rac:m es o ratios denoted as multiple entries in Table 2. 22 R R~ M ..... ,c1 --.....,Cl R~ mesa racemo 3 M Ra RB rac meso 3a Ti H C(CH3h 2.5 1 3b Ti H Si(CH3h 0.6 1 1.8 1 3c Ti H CH(CH3h 1 1 3d Ti H CH2C6H 5 1 1 3e Ti H C(CH3)2C6H 5 1.2 1 2.2 1 1.2 1 2.2 1 1.5 1 2 1 1.5 1 2 1 3£ 3g 3h Ti Zr Zr H H H C(CH3)iC6H 5 C(CH3h Si(CH3h Table 2. Racemic / meso ratios for tetramethylethano-bridged group IV metallocenes24 Silyl linking units have found widespread use as a result of the simplicity with which they are incorporated into ligand ensembles, and because of their ability to "tie back" the Cp rings, often affording an effective reduction in steric bulk in the front of the metallocene wedge. The insertion of a single-atom silyl-linker is often obtained through the use of dichlorodimethylsilane rather than 1,2-dibromoethane in the ligand synthesis. 23 As is observed for the ethano-bridged complexes, rac:m es o ratios for the group IV metallocenes employing a dimethylsilyl linker are largely unaffected by the nature of the ~-alkyl substituent when the a-substituent is a hydrogen. Increasing the steric bulk of the a-substituent generally results in a considerable increase in selectivity for the racemo metallocene, with smaller effects imparted by modification of the ~- substituent (Table 3). racemo mesa 4 M Ra RB rac meso 4a25 Zr H C(CH3)3 1 1 4b25 Zr H Si(CH3h 1 1 4c25 Zr H 1 1 4d25 Zr H C(CH3hC6H 5 C(CH2) 5C6H 5 1 1 4e25 Zr C(CH3h 2 1 4£25 Zr CH3 CH3 CH(CH3)2 6 1 4g26 Ti Si(CH3h C(CH3h 1 * 4h26 Zr Si(CH3)3 C(CH3h 1 * Table 3. Racemic/meso ratios for dimethylsilyl bridged group IV metallocenes; * = not observed A major breakthrough in the synthesis of chiral metallocenes was the discovery that incorporation of a trimethylsilyl group adjacent to the silyl linker results in the exclusive formation of C2 -symmetric metallocenes. The ligand is constructed by the procedure outlined in Figure 3. The ligand array, abbreviated [Bp], exists in its doubly protonated form ([BpH2]) as a mixture of rapidly interconverting isomers, due to facile silyl- and hydrogen- migrations about the cyclopentadiene rings.27 Double deprotonation yields exclusively 24 [Me2 Si(2-SiMe3 -4-CMe3-C5 H 2hf ([Bpf), the isomer in which the silyl groups reside a- to the sily 1 linker. l 2 [ Me3 C ~ Li+ 1) Me2SiC12 2) 2 eq. n-BuLi -2:fr exclusive product + isomers [Bpf Transmetallation of [Bpf affords exclusively racemo- metallocenes. Unfavorable steric interactions between the trimethylsilyl groups in the narrow portion of the metallocene wedge precludes formation of them es oisomer. This is true not only for the group IV metallocenes, but also for the group III derivatives (Table 4). 25 R13 Me2 s· a M-Cl Ra racemo R13 mesa 5 M Ra RB rac meso Sc y H C(CH3h * 1 H C(CH3h 1 1 Si(CH3h C(CH3h 1 * sd31 Sc y Si(CH3h C(CH3h 1 * se32 La Si(CH3h C(CH3h 1 * sa28 5b29 sc30 Table 4. Observed group III enantioselectivities for dimethylsilyl-bridged ligands; * = not observed In the absence of the metallation-directing a -trimethylsilyl groups, additional steric congestion on the Cp rings serves to increase the metallation selectivity, as shown in Table 5. The most substantial effects, again, result from incorporation of increasing steric bulk in the position adjacent to the Silinker . 26 R3 R4 Rs M~~i R4 R3 mesa racemo 6 M R2 R3 R4 Rs rac 1neso 6a Zr CH3 H 5.6 1 Zr CH3 H H 6b CH3 CH3 15.7 1 6c Zr H C(CH3h H 2.7 1 6d Zr H H H 7.3 1 6e Hf H 1 Hf H 100 1 6g Hf H H CH3 H CH3 H 13.3 6£ CH3 CH3 CH3 CH3 H 1 2.8 6h Hf H CH3 H H 1.1 1 CH3 H Table 5. Racemic / meso ratios for additional dimethylsilyl bridged group IV metallocenes33 Chiral metallocenes incorporating indenyl- and indenyl-derived rings have often resulted in catalysts with especially high activ ities and / or stereospecificities. The metallocene dichloride precursors to these catalysts, however, are formed with generally poorer enantioselectivity than the formally Cp-based complexes described previously (Table 6) . 27 Cl Cl 7 10 M R 734 Zr H sa35,36 Zr H 8b35,36 Zr 8c35 rac meso 1 2-10 H 1 1 CH3 H 1 1 Zr CH2CH3 H 1 1 sa35 Zr CH3 CH(CH3)i 1 1 8e37 Zr H C6Hs 1 1 sf37 Zr CH3 C6Hs 1 1 8g37 Zr CH3 1-naphthyl 1 1 937 Zr CH3 C6Hs 1 2 1oa36,37 Zr H 1 * tob36,37 Zr CH3 1 1 11i37 Zr CH3 ti ti R2 C6Hs Table 6. Racemic/ meso ratios for indene- and indenyl-based ansametallocenes; *= not observed; t = mixture of 5- and 6- indenyl substituted metallocenes were inseparable; ti = mixture observed, but rac I mesa ratio was not reported 28 The use of chiral linking units to induce not only enantioselectivity (rac vs. mesa) but diastereoselectivity in the metallation step has been, until recently, met with little success. Ligands employed in initial studies resulted in either mixtures of rac and mes o (ligand 12)38 or exclusively mes o (ligand 13)39 metallocenes, and generally low overall yields. 12 13 The chiral binaphthyl-bridged ligand (14)40 affords exclusively chiral metallocenes, although X-Ray diffraction indicates that the steric constraints of the linker result in considerable deviations from the desired C2metallocene symmetry. The use of enantiomerically pure cyclacene ligand 15 affords only one rac- metallocene (M=Ti);41 a crystal structure indicates an unsymmetric disposition of the indenyl ligands, despite their apparent equivalence in solution (1 H NMR). The use of a 2,2'-biphenyl linker (16; R=H, CH3)42 results in the formation of diastereomerically pure metallocenes, although in low overall (c.a. 3-18%) yield. ,.)~ ~ /? R R 14 15 16 Considerably more successful are the use of chiral cyclohexane linkers and silyl linkers containing chiral binaphthol-derived substituents (Figure 4). 17 metallates in high yield to give a single metallocene as a result of steric 29 interactions between the indenyl moieties and the isopropyl groups on the cyclohexane linker;43 the high diastereoselectivity of 18 is a result of steric interactions between the a-trimethylsilyl substituents on the Cp rings (interactions between which are responsible for enantioselective metallation), and the 3,3'-naphthalene methine positions.44 1) TiC13 2) HCl, 0i exclusively Me MC13 (THFh_ 5 (M = Sc, Y) • 8::s· M-Cl exclusively Figure 4. Tailored ligand arrays for diastereoselective metallocene syntheses. As described in Chapter One, a monomeric chiral group III complex would be invaluable for mechanistic studies on a single component isospecific catalyst. Generation of such a complex hinges on the ability to prevent formation of bridging-hydride dimers commonly observed for group III and lanthanide metallocene derivatives. We reasoned that replacement of the tert butyl substituents in 5 and 18 with more extensive alkyls may perturb the hydride monomer-dimer equilibrium by virtue of intermolecular steric interactions between these new ~-substituents. The choice of substituent must induce the 1-(linker)-2-(SiMe3 )-4-(alkyl) Cp substitution required to generate exclusively racemo - metallocenes (as in compounds 4g. 4h, Sc-e, 18), thus eliminating the need for separation of m es o and rac isomers. The synthesis of several novel ligand arrays incorporating extensive ~substituents is reported herein, as are the resultant group III metallation enan tioselecti vi ties . 30 Results and Discussion Efforts to generate ansa- metallocenes with extensive alkyl substituents in the ~- positions focused initially on the synthesis of the cyclohexylsubstituted analogue to the [Bp] ligand array, as shown in Figure 5. Me,,, ... S. H Me- Figure 5. Initial cyclohexyl-substituted target ligand array. Following the general protocol for the parent system,32 the initial steps of the ligand synthesis involved construction of alkyl-substituted cyclopentadienyls. Pentamethylenefulvene (19), synthesized using the procedure of Stone and Little,45 is easily reduced using borohydride reagents such as LiB(CH2 CH3hH (Super-Hydride) and KB[CH(CH3 )CH2 CH3 hH (KSelectride), resulting in trialkylborane adducts of cyclohexylcyclopentadienyllithium (20) and -potassium (21) (75-80% yields, Figure 6). The coordinated trialkylboranes are not removed either by repeated washings of the complexes with petroleum ether or by extended heating under dynamic vacuum. Attempts to generate cyclohexyl-substituted ligand arrays from the trialkylborane adducts 20-(BR3 )n and 21-(BR3 )n ultimately proved unsuccessful. The coordinated trialkylborane appears to interfere in the linking reaction of 20-(BR3)n and 21-(BR3)n with Me2 SiC12 , as the borane free salts 20 and 21 may be successfully employed to generate the dimethylsilylbridged 22 (vide infra). Dichlorodimethylsilane reacts readily with 20-(BR3 )n, 31 LiB(CH2CH3hH[ ] ~ C y ~ Li+· [B(CH2 CH3 ) 3 Ji [20-(BR3)n1 l K B ~ [ c y ~ K·· [B(sec-BulJln [21-(BR3)n1 Figure 6. Syntheses of cyclohexylcyclopentadienyl-trialkylborane adducts. yet does not consistently generate 22. The mixture of resultant products may be a mixture of 22, the mono-cyclopentadiene product 23, and species resulting from deprotonation of 23 by additional equivalents of 20-(BR3 )n; more thorough analyses of linking reactions are presented later (vide infra) . 1 /2 eq. Me2 SiC12 20-(BR 3)n - - - - - - - mixture of products Cy~_~Cy S1·, + H3C CH3 ~ 22 Figure 7. Dichlorodimethylislane fails to cleanly link 20-(BR3 )n. The trimethylsilyl- substituted Cp 24 ([1-SiMe3-3-Cy-C5H 3 ]K; Cy= cyclohexyl = C6H 11 ) is obtained from the treatment of 20-(BE~)n with chlorotrimethylsilane and subsequent deprotonation with potassium-tertbutoxide (Figure 8), the reaction of this dipotassium salt with 1/ 2 equivalent of dichlorodimethylsilane failed to generate the diprotio- form of the desired ligand array, 25. The origin of the difficulties in obtaining 25 from the route shown in Figure 8 was not investigated in great detail, but analogous linking reactions on 2-adamantyl-substituted cyclopentadienyls indicates that 1trimethylsilyl-3-(secondary alkyl)-cyclopentadienyls are too sterically congested to permit linking with Me2SiC12 (vide infra). 32 Me3Si SiMe3 Cy~Si~Cy ✓ H3C ·,, CH3 (25) Figure 8. Alternate synthetic attempts to obtain Me 2Si[(Cy)(SiMe3 )-C5H 3 )h (25). Alternate routes into the synthesis of 20/ 21 were investigated in the hopes of obtaining borane-free lithium- or potassium cyclopentadienyl salts. Albeit in low yield (:5; 30%), Diels-Alder dimers of cyclohexylcyclopentadiene + Cy~ + + LiAlH4 ------;- Figure 9. Alternate syntheses of 20 and 21. ([CyCpHh) are generated cleanly from the reaction of the appropriate Grignard reagent with cyclopenten-2-one (Figure 9). Improved yields (>70%) of [CyCpHh are obtained from use of LiAlH4 in conjunction with 19. 33 Either the lithium salt resulting from the reaction of 19 with LiAlH4 or the potassium salt generated from treatment of CyCpH with KOCMe3 may be used to generate the linked complex Me2 Si(Cy-C5H 4) 2 (22) by the slow warming from -78 °C to 25 °C of a tetrahydrofuran solution of the appropriate anion with 0.5 equivalents of Me2 SiC12 . A single regioisomer Li2 [Me2 Si(3-CyC5H 3)i] (26) is obtained as a white petroleum ether insoluble powder (60% yield), from deprotonation of 22 (Figure 10) with two equivalents of n butyllithium. + isomers (25) (26a, M = Li; 26b, M = K) Figure 10. Synthesis of doubly protonated cyclohexyl-substituted ligand array. Trimethylsilyl- groups are introduced to the Cp rings by treatment of 26 with a slight excess of chlorotrimethylsilane. As is common for silylsubstituted cyclopentadienes, the multiply silylated complex 25 exists in solution as a mixture of regioisomers due to facile silyl and hydrogen migrations about the cyclopentadiene ring, although the silyl substituents tend to reside on sp3-hybridized ring carbons.27,32 1 H NMR spectroscopy indicates a mixture of products; complete assignments were not made because of multiple overlapping resonances (> 18 Cp, 9 Me2Si signals) . Resonances for the major product (by integration of the -SiMe3 resonances, (8 0.10 to -0.10, C6D 6 ) are consistent with those expected for the isomer of 25 drawn in Figure 10, in which all silyl substituents reside on sp3 - hybridized carbons. Mass spectroscopy on the interconverting isomers provides a parent peak at (m / z) = 496 (molecular weight of C3J1 52Si3 = 497.09 g / mol). 34 In contrast to the tert-butyl-substituted [Bp] ligand array, which deprotonates to give a single regioisomer (Figure 10),3 2 the doubly-protonated cyclohexyl-substituted ligand array 25 reacts with a variety of deprotonating agents (e.g. n-BuLi, LiCH2 SiMe3 , KOCMe3 , KN(TMSh, LiN(TMSh, etc.) to afford a mixture of products. Presumably, this mixture includes not only the desired Me2Si(2-SiMe3 -4-Cy-C5H 2)i isomer (27a), but also regioisomers 276 (Me2Si(3-SiMe3-4-Cy-C5H 2 ) 2 ) and the mixed (1,2,4)(1,3,4)-substituted complex 27c (see Figure 11). Also possible are structures analogous to 27a-c in which the cyclohexyl groups reside adjacent to the dimethylsilyl linker rather than to the trimethylsilyl group. Various permutations of deprotonating agent, solvent, and reaction temperature fail to result in the exclusive formation of the desired isomer 27a; the most successful set of reaction conditions towards the formation of a single isomer is the slow warming from -78 °C to 25 °C of a THF solution of KOCMe3 and 25. The major regioisomer is then be obtained in approximately 20-25% yield after repeated (>3) recrystallizations from cold THF / petroleum ether solutions. The failure of the cyclohexyl-substituted ligand array to provide a single geometric isomer upon deprotonation is attributed to the inability of the cyclohexyl group to exclusively direct the trimethylsilyl group adjacent to the dimethylsilyl linker prior to deprotonation. Figure 11 shows some of the possible isomers of 25, and the results of deprotonation thereof; an alkyl substituent that is not sufficiently bulky to disfavor the sampling of the isomers indicated in the two leftmost columns will result in a mixture of undesired constitutional isomers upon deprotonation. I ..... ::::r ~ (t) ~ rt) >-t i:: 0 (t) '"1 0 0 (t) ~ ..... [fl [fl [fl &S [fl .. . < (t) [fl CJl N ........ (t) rt- 0 Pl Pl (t) '-< ::; ro 5- 0......., ~- 0 0 ....... 0 ::;rt-:=..; cr' ::; (") '"1 'Up;· ft b.. ::::r ~ [fl (t) ~ rt- ro ro [fl ro >< rt- '"1 Q.. ~ - (/) C C o ::::r Pl 1----'• ........ ::; rt- ~ (t) ..... ..... f""'T' w [fl i:,i C ro CJ) 5· N '"1 [fl '1 <:< &. (")ro <; '"1 ::::r (t) 0 (t) (") ~ rt- rt::::r rtrt(t) ::::r ;: :;.: ::; rt- [fl .... • 8 ro C 'D '"1 'U 0 (t) rt- :=..:'--<: ........ ................ Mp_. CTQ ::::r 0 ::::r (t) ~ ~ '"1 ;1°(008 (") 7" (") p_. ~ (") C::S'--<:(f'J Q '"1 'U (f'J rt- "T1 ()Q. j) '::::: Cy SiMe2 Me ! SiMe2 I ~ 'Q-siMe3 Cy Me ! l r [Me2Si(3-TMS-4-Cy-C5H2hf(27b) (27a) (desired isomer) Me 3Si Cy Cy9sf~SiMe Me. - 2 1-1 + ] cy-efsiM~SiMe3 r l Me,· sfMe - 2 H+ cy--0::;?i~ . SiM~;~s1Me3 Cy [HJ migrations -o::SiMe2 I . Cy .,. __ SiM~SiMe3 . . jf Cy [HJ migrations ~ -o::SiMe2 I I ~ . Cy SiM*S1Me3 jt Cy [HJ migrations)) (27c) [Me 2Si(2-TMS-4-Cy-C5H 2)(3-TMS-4-Cy-C5H 2) ]2- C ~Si~ . Y~ .. ~~SiMe3 SiMe3 Cy . t·101:1s j) ..[HJ m1gra I Cy SiMe 2 C y - 9 ~ S i M e3 SiMe3 Cy SiMe3 - j) [HJ migrations j J 9 Cy [HJ migra tions SiMe3 . -ctS1Me2 C y .,. __ ~~---S i M e3 SiMe3 Cy [SiMe3J migrations SiMe2 ~:::::----. S i M e3 ::::= [Me 2Si(2-TMS-4-Cy-C 5H 2h] - 2 Cy~~~R:e3 l Me - 2 H+ j SiMe3 Elf ~ Cy~SiMe~ Cy I 3 SiMe 2 SiMe3 Cy - o : :S i M ~ Cy jt s·M SiMe 3 2 pr Cy [SiMe3J migrations SiMe ~~ - 1/" [HJ migrations Cy • · [HJ m1gra hons ~ j) Cy SiMe2 QSiMe Cy~ /4 3 -.Me SI 3 Cy [SiMe 3J migrations S~~e3 Cy~Si~ uJ Vl 36 It was hoped that replacement of the cyclohexyl groups with even larger alkyl substituents, namely 2-Adamantyl (Ad) groups, could provide a ligand array that could be deprotonated to afford the single Me Si(2-SiMe -4alkyl-C H Li) isomer (29). 2 5 2 3 2 Me,, .. . M~ 5 ? + isomers 28 Figure 12. Target ligand incorporating 2-Adamantyl substituents. Condensation of 2-Adamantanone with freshly cracked cyclopentadiene provides fulvene 30, a bright yellow crystalline solid, in high yield (>90%, Figure 13). Reduction of the fulvene with borohydride reagents gives results similar to those observed with pentamethylenefulvene: rn] MBRH [~ 3 M - (BR3)n Figure 13. Syntheses of 2-Adamantyl substituted cyclopentadienyl anions. (M = Li, R = -CH CH M = K, R = CH(CH )CH CH ; 2 3 ) 3 2 3 37 alkylborane - Cp adducts of the desired alkyl-substituted cyclopentadienes. Borane-free (2-Ad)-C5 H 4 Li (31) and (2-Ad)-C5 H 4 Na is obtained from treatment of 30 with LiAlH4 or NaAl(OCH2 CH2 0CH3)iH2 , respectively (2-Ad = 2Adamantyl = 2-(C 1Ji1s)). The (2-Ad)-C5H 4 Li and Na salts react readily with chlorosilane reagents 1 to give (2-Ad)(silyl)-substituted cyclopentadienes. H NMR spectroscopy indicates that both 32 and 33 (Figure 14) exist as single isomers in solution (THF-d 8 , C6D6 ). 32 shows three vinylic and one allylic cyclopentadiene protons, a single SiMe3 group, and a single set of Ad resonances; 33 shows similar cyclopentadiene resonances and a doublet for the diastereotopic Si-Me groups. 32 is converted to [1-SiMe3 -3-(2-Ad)-C5H 3 ]Li (34) by treatment with excess n-BuLi. Similarly, 34 is obtained from treatment of 33 with excess CH3 Li, both deprotonating the cyclopentadiene ring and reacting with the remaining Si-Cl group. In contrast to both the tert-butyl- and cyclohexylsystems, (2-Ad)-C5H 4Li can not be linked using dichlorodimethylsilane; treatment of 2 equivalents of 31 with Me2 SiC12 results only in the formation of 33, with the second equivalent of Li salt failing to react at room l [ ~ A d Li+ (31) Me2 SiC12 Me3 SiCl Me3Si"- ----- LiCl (32) 98% yield !-LiCl n-BuLi Cl I Me1••si (33) Me~ ~ A d ~ Ad 2 eq CH3 Li - LiCl -C~ !-C H 4 10 [M e s s ~ Ad] u+ >90% yield (34) Figure 14. Syntheses of [1-SiMe3 -3-(2-Adamantyl)-C5H 3 ]Li, 34. temperature. Heating a 31/33 mixture (80 °C, THF) results only in decomposition of the starting materials. 38 The exact nature of the side reaction / decomposition mechanism operating in this system is not known, although the presence of appreciable quantities of AdCpH (protonated staring material) in the product mixture supports the two pathways shown in Figure 15. After generation of the chlorosilyl- substituted cyclopentadiene 33, additional equivalents of cyclopentadienyllithium may react with the remaining Si-halide functional group (pathway i) to give the desired linked species 28; the remaining 31 may alternately serve to deprotonate the doubly allylic cyclopentadiene hydrogen in 33 (pathway ii), thus generating one equivalent of AdCpH (the other products are unlikely to be identified due to subsequent decomposition). Competing pathways analogous to (i) and (ii) are presumed to account for the inability to cleanly link 24 with Me2 SiC12 , as described previously. [ ~ A d ] Li+ H ~Ad Me••·s/ ,__ Me- l ~ A d H (28) (31) ! ?? Figure 15. Possible reactivity pathways for 33. The tendency for silyl- groups to reside on sp3-hybridized cyclopentadiene ring carbons may, however, be exploited in order to disfavor the side reaction arising from pathway (ii). The trimethylsilyl-substituted complex 34 reacts with Me2 SiC12 to generate the doubly-silylated complex 35; the single isomer of 35 observed in solution is that with both silyl 39 3 substituents residing on the same sp -hybridized carbon. This isomer has no allylic (i.e. acidic) hydrogens available for deprotonation by additional equivalents of 34, and should therefore be expected to resist deprotonation/ decomposition (pathway ii) in a manner analogous to that shown in Figure 15. Indeed, deprotonation of 35 is found to be slow; there is, in fact, no additional reaction between 35 and additional equivalents of 34 at room temperature. Extended heating (days at 80 °C) results in the slow decomposition of the starting materials to give a mixture of intractable, unidentifiable products. The resistance of 35 to deprotonation is a consequence of the desire of multiple silyl substituents to reside on a single sp 3-hybridized cyclopentadiene carbon. The stability of the Si-Cl bond is attributed to a combination of steric influences from the adjacent methyl groups on the potential linker and those of the alkyl and silyl substituents on 34. 34 34 ~ Figure 16. Reaction of 34 with Me2SiC12 generates a robust doubly silylated cyclopentadiene, 35. An alternate route into the synthesis of 28 involves synthesis of a silyllinked complex whose linker unit is subsequently functionalized after the linking step. The unlinked, (trimethylsilyl)(trichlorosilyl)- substituted complex 36 is generated from the reaction of equimolar quantities of 31 and SiC14 (Figure 17); the single isomer of 36 observed in solution is, again, that with both silyl substituents residing on the same sp 3-hybridized carbon. Additional equivalents of 31 react with the -SiC13 moiety of 36 within 24 hours (25 °C, THF) to generate the dichlorosilylene bridged complex 37, as an air and moisture sensitive pale yellow solid. The dimethylsilylene bridged species 28 is obtained as a mixture of rapidly interconvering isomers27 from treatment of 37 with an excess of CH3Li followed by quenching with methanol 40 and an aqueous workup, similar to the procedure reported by Marks and coworkers. 46 [Me3 S ~ A d ] L i + SiC14 - LiCl (31) isomers (28) sublimes @ 150-160 °C (<1µ) - 45 - 65% overall yield + isomers Figure 17. Syntheses of doubly protonated, 2-Adamantyl substituted ligand arrays (28, 38). No deprotonation of 36 by 31 is observed. Based in part on the stability of 35, the robustness of the Si-Cl functional group in 36 is attributed to steric protection by the adjacent silyl methyl groups, although steric effects are not easily separated from electronic effects of switching from Me2 SiCl to SiC13 substituents. It was hoped that 37 would prove to be a versatile building block for a variety of metallocene precursors due to the presence of reactive Si-Cl bonds in the Cp linking unit. In addition to reacting with small nucleophiles such as CH3 Li, 37 reacts with (±)1,l'-bi-2-naphthol, resulting presumably in the C 2oH 12O 2 Si-linked bis(cyclopentadiene) 38. The chelating nature of the binaphthol moiety should favor formation of 38 over structures in which only one of the binaphthol O-H bonds reacts, or polymeric structures in 41 which Bn groups bridge multiple Si- linkers. Unfortunately, positive 1 identification by H NMR of the product mixture is complicated by the silyl and hydrogen migrations that rapidly interconvert the various regioisomers of 39,27 and positive identification by deprotonation to give a recognizable BnSi(Ad-TMS-Cp)2 dianion is unsuccessful (vide infra) . As is observed for the mixture of isomers of 25, deprotonation of the 2Adamantyl-substituted ligand precursors 28 fails to produce single isomers of the desired dianionic ligands. Mixtures of products are obtained when 28 is treated with any one of a variety of alkyllithium reagents, amides, or alkoxides (e.g. CH3 Li, n-BuLi, LiCH2SiMe3 , LiCH(SiMe3 )i, LiN(SiMe3 )i, KN(SiMe3)i, KOCMe3 , NaOCMe3 ). Analogous mixtures of products are obtained upon deprotonation of the binaphthol-containing ligand precursor 38. X. ,X i [A ~ ~ ~ : e j [X2 Si(2-TMS-4-Ad-C5 H 2)i] 2+ X"···s· x- l Ad X. ,x Ad [A ~ ~ ~ i M e j + isomers 29: X2 Si = Me 2Si 39: X2 Si = (C 20H 12O2 )Si [X2 Si(2-TMS-4-Ad-C5 H 2 )(3-TMS-4-Ad-C 5H 2 ) f' + rA~Sf~iMej ~~S( Ad [X2 Si(3-TMS-4-Ad-C5 H 2)i] 2- Figure 18. Deprotonation of 2-Adamantyl-substituted ligand precursors 28, 38 result in mixtures of constitutional isomers. If the mixture of products observed upon deprotonation are merely the kinetic products and are different from the thermodynamic deprotonation products, then an equilibration to the thermodynamically favored isomers may be feasible by introduction of a small amount of a proton source, 42 followed by subsequent deprotonation. Multiple cycles of this sequence should eventually convert all of the initially formed species to the thermodynamically favored species. Attempts to drive the observed deprotonation mixture to such a product (or products) by addition of a small amount of 28 to either the lithio- or potassium deprotonation products (days at 80 °C, THF-d 8 ) result only in the observation of unidentifiable additional products. The failures of the secondary alkyl (cyclohexyl- and 2-Adamantyl) substituents to direct the trimethylsilyl substituents exclusively adjacent to the Si- linker, and the success of the tert-butyl groups in inducing such regioselectivity,3 2,44 suggests that tertiary alkyls are necessary in order to obtain the 1,2,4- substitution required for the directed synthesis of exclusively C2-symmetric racemo- metallocenes. Accordingly, three tertiary alkyls are considered for incorporation into ligand ensembles directed towards the generation of monomeric C2symmetric Group III metallocene hydride complexes: rnethylcyclohexyl, 1Adamantyl, and 2-(2-Methyl)-Adamantyl groups. Methylcyclohexylcyclopentadienyllithium is prepared pentamethylenefulvene and CH3 Li; it is important that the reaction mixture remain cold (petroleum ether/diethyl ether, ~o °C) in order to prevent deprotonation of the fulvene to yield methane and (cyclohex-1ene)cyclopentadienyllithium. 25 °C 0°C Figure 19. Synthesis of methylcyclohexylcyclopentadienyllithium. 43 Although the appropriate starting material could be made cleanly, efforts to synthesize a complete ligand array incorporating MeCy substituents were terminated at this point due to the likelihood of cyclohexane chair-chair interconversion perturbing the stereospecificity of any eventual catalyst derived from this tertiary alkyl. The high stereospecificities of 4g, 4h, and Sd, and catalysts derived from 18 are believed to result, in part, from the rigidity of the ligand arrays; the fact that methylcyclohexyl-derived ligand arrays are likely to possess more than one energetically accessible conformation (i.e. methyl group axial versus methyl group equatorial, as shown in Figure 20) prompted the search for a more rigid tertiary alkyl substituent. Me M _RCH3 ~ Me,,,. . - M ~ 5l 3 Figure 20. Possible methylcyclohexyl-Cp chair-chair interconversions. A ligand array incorporating 1-Adamantyl groups as the ~-alkyl substituents would provide the desired steric rigidity; synthesis of such a ligand is not currently feasible, however, as a result of difficulties encountered in forming a 1-Adamantyl - cyclopentadienyl C-C bond. Almost entirely unsuccessful are attempts to couple 1-AdMgBr and cyclopenten-2one. GC/MS analysis indicates that only approximately 2% of the highly reactive 1-AdMgBr, generated in situ ,47,48 reacted with cyclopenten-2-one to give (1-adamantyl)cyclopenten-1-ol or 1-adamantylcyclopentadiene. The vast majority of the product mixture (>97%) had NMR spectra and mass spectral (ml z) ratios consistent with adamantane and adamantane dimers. Also unsuccessful are attempts to generate (1-Ad)-CpH from 1-AdLi and cyclopenten-2-one and from NaCp and 1-bromoadamantane. Mg [lG-MgBrl7 IQ lG--Gl + 6 44 HO + + ~98% Figure 21. Synthesis of 1-adamantyl-cyclopentadiene. The rigid tertiary alkyl most amenable to incorporation into ansa metallocenes is the 2-(2-methyl)-adamantyl group, whose Sa,b and Sc-e analogues [Me2 Si{3-[2-(2-CH3 -CHJI 14 )]-C5H 3 ) 2 and (Me2Si(2-SiMe3-4-[2-(2-CH3C 1J-l 14 )]-C5H 3 )z)2", abbreviated [Adpf and [Abpf, respectively) are shown below. Although the [Adp] ligand array is not expected to result in exclusively racemo- metallocenes (due to the absence of -SiMe3 groups adjacent to the linker), metallocenes derived from this ligand array may be utilized to gauge the ability of the methyladamantyl groups to perturb group III hydride monomer-dimer equilibria. f [Adp] 2- [Abp]2- Figure 22. Target [Adp] ([Me2Si{3-[2-(2-CH3-C 10H 14 )]-C5H 3 ) 2 ]) and [Abp] (Me2 Si(2-SiMe3 -4-[2-(2-CH3 -C 1J-I 14 )]-C5H 3 }z) dianions. 45 Although considerably slower than the previous borohydride and LAH reductions, 30-40 gram quantities of [2-(2-methyl)-adamantyl]cyclopentadienyllithium (39) are obtained in high yield (>90%) from the reaction of CH3 Li with fulvene 30. Slightly more convenient for subsequent manipulations are the DME and TMEDA adducts of 39 (39-DME and 39TMEDA, respecitively), which are obtained from the addition of the appropriate ether or amine prior to isolation of the Li salt. Efforts to obtain a methyladamantyl-substituted dichlorosilyl-linked species analogous to 37 were undertaken, as such a complex would be of considerable synthetic utility due to the large number of metallocene precursors readily available from functionalization of the SiC12 linker. 40 is obtained upon treatment of 39 with Me3SiC1, filtration of LiCl, and subsequent deprotonation with n-butyllithium. The dimethoxyethane and tetramethylethylenediamine adducts 40-DME and 40-TMEDA are more easily manipulated species than the DME-free complex, as the former precipitate readily from petroleum ether, thus facilitating isolation and purification; the free salt 40 forms a viscous paste in petroleum ether, from which residual butyllithium can not be removed. Li (39) DME Li - DME (39-DME) TMEDA Li-TMEDA (39-TMEDA) Figure 23. Syntheses of [2-(2-Methyl)-Adamantyl] cyclopentadienyls. Unfortunately, all attempts to generate linked bis(Cp) complexes from 40 (or the DME or TMEDA adducts) and Me2 SiC12 or SiC14 are unsuccessful, irrespective of the presence or absence of chelating ethers or amines. 40- 46 complexes react readily with Me2 SiC12 and SiC14 , yet often result in a mixture of products analogous to those formed via the pathways indicated in Figure 15. (39-TMEDA) Me3 SiCll - LiCHMEDA Me3 Si""'- ----=::--.._ ~MeAd H n-BuLi j DME [Me3S~MeAd] u +. DME (40-DME) jMe SiCI, ~Me SiCI SiCI✓ Me2 C 1 S 1 ~ · ~ MeAd Me3 S1 40-DME 3 2 Cl3s ~ · ~ MeAd Me3Si ~ ;\<40-DME Me3S · ~ ~ MeAd Me3Si (41) Me3s · p ~ MeAd X,, .. . (X=Me, Cl) x- 51 ~ /~MeAd Me3 Si Figure 24. Unsuccessful attempts to generate 2-(2-methyl)-Adamantyl substituted analogues to 28 and 38. MeAd = 2-(2-methyl)-Adamantyl); TMEDA = N,N,N',N'-tetramethylethylenediamine; DME = 1,2dimethoxyethane. The regiochemistry upon deprotonation of the multiple silylated cyclopentadienes provides a means of ascertaining the relative bulk of the methyladamantyl substituents. Of interest is the ability of the methyladamantyl groups to direct various silyl- substituents about the ring, relative to that of both the cyclohexyl- and 2-adamantyl substituents employed previously. The doubly silylated cyclopentadiene 41 exists as a single isomer in solution (determined by 1H NMR). Upon deprotonation, the 47 resultant cyclopentadienyllithium salt shows only 1,2-(SiMe3)i-4-MeAd substitution (MeAd = 2-(2-methyl)-Adamantyl); none of the l,3-(SiMe3)i-4MeAd regioisomer, in which the two vinylic Cp and the two trimethylsilyl 1 groups are inequivalent, is observed in the H NMR of even the crude reaction mixture. This indicates that the 2-MeAd substituents should be sufficiently bulky to direct trimethylsilyl groups exclusively a- to the silyl linking unit to generate [Abp f. Me3 Si ~ M e A d (41) Me3 Si /n-BuLi' Me3S ~ M e A J u+ [ Me3 S1 j ( 42) IMe3S i ~ ~ e A d l u + S1Me3 not observed l j Figure 25. Deprotonation of 41 affords exclusively [1,2-(SiMe3) 2 -4-(MeAd)C5H 2 ]Li (42). Using the strategy employed successfully for 25, dichlorodimethylsilane is used to link two equivalents of 39-TMEDA. Slow warming from -78 °C of a THF solution of 39-TMEDA and Me2 SiC12 yields multigram quantities of the doubly protonated ligand 43, H 2 [Adp ], in c.a. 80% yield after removal of LiClTMEDA by extraction of 43 into petroleum ether and Kugelrohr distillation. The success of this reaction is somewhat surprising, given the failure of Me2 SiC12 to cleanly link the 2-adamantyl-substituted 31; the origin of this discrepancy is unclear. Deprotonation of 43 provides Li2 [Me2 Si(3-MeAd-C5H 3 h] (Li2 Adp, 44) as a white petroleum ether insoluble powder (-50% yield after recrystallization). Trimethylsilyl groups are incorporated onto each ring by addition of a slight excess of Me3 SiCl to 44 (THF). After a purification procedure which involves deprotonation with KOCMe3 and then hydrolysis, H 2 Abp (45) may be isolated as a tan air and moisture-stable solid (38% yield). 48 H 2 eq. 39 -TMEDA ~MeAd Mei,, .. / ._.... 51 -2 LiCl-TMEDA M~ ~ M e A d Me2 SiC12 H + isomers (43) 2 eq.n-BuLij- 2 eq. C4H 10 2 eq. Me3 SiCl - 2 LiCl H2 Abp (45) A:§;J_MeAd Mei,. ( 5 Me" ~ M e A d Li2 Adp (44) Figure 26. Syntheses of H 2 Adp (43), Li2 Adp (44), H 2 Abp (45). 1 As expected, the H NMR of 45, in either C6D6 or THF-d 8, indicates a mixture of isomers due to facile silyl- and hydrogen migrations about the cyclopentadiene rings. Treatment of this mixture with any of a variety of nucleophiles, however, results in the double deprotonation of H2 Abp to yield a single isomer of [Abpf. Li2 Abp (46) and K2 Abp (47) may be obtained from the treatment of 45 with LiCH2 SiMe3 or KN(SiMe3 )i, respectively. The 500 MHz 1 H NMR spectra of the mixture of isomers of 45 and the single constitutional isomer 47 are shown in Figures 27 and 28, respectively. The Me3S \ ~ Me, . .__ r--MeAd _-_2_H_+ 5 Me--~MeAd Me3 S1 + isomers (45) [Abp ]2[Me2 Si(2-TMS-4-MeAd-C 5H 2 )i]2 (46 and 47) no other regioisomers observed Figure 27. Deprotonation of H 2 Abp (45) affords a single constitutional isomer of the desired 2-(2-methyl)-Adamantyl substituted ligand array. 'T".1 ..... °'0 °'..__, n ~ CJl ..__, 'U ~ ~ 0 ......, Pl >"i () ,..... (1) en 'U ~ ~ ~ ::r: N ::r: z ~ 0 0 (Jl N 00 >"i (t) e oq ! ,, !: I , 1 1 7.0 ,.,- 5.0 4.0 3.0 I I 2.0 I ''-.../ ' I I... • ~ 1.0 1 l'1 i PPM I ''II·ii11 1l i LL I I 1 /I ~ JJJv ' j I Si(CH3)z 'T ·.-r,,· ,-r-,- 7 '--_; •1 ) n.,, !1\ CHTC10H14 CHrC1oH14 I r I I I I r r r r 7 T 7-, r T r r r r I r r r r I r r r r I r r r , I r , 6.0 r , , , -, r I r r - , r --~~:GLJL~-/~~---- -- - I + isomers Me2 Si[CsH 3(SiMe 3)(CHTC10H 14)h Me2 Si[CsHiSiMe3)(CH3 -C10H 14)lz C 6D5H Me3 Si ~:::si Me3 Si Si(CH3h r,/:>, \0 "T1 ..... ~ • l,,C) ~ I Cl:J ~ p.. p.. (!) p.. ~ §:: ' 00 ~ ::r: 'Ti ~ -...::i .__,,, ~ ,,.,......_ "'d ~ ~ 0 ......., ~ "°1 rt- () (!) "'d (/l :;:::::J ~ z ::r: N ...... ::r: ~~ IQ ::i ~ 0 Q N l't) '"I i:: (JQ I 6.0 1 1 1 1- r I I r 1 I 5.0 I I I I I ) I Me2 Si[CsfI 2(SiMe3)(CH3-C10H14)h Me,... Me_:Si I 2 KN(SiMe 3)2 ! H 2Abp (mixture of isomers) I I 4.0 I I 'I I I ,J I solv • K+2 ! I I '-......-, __ _ I I 3.0 I I I 18-Crown-6 I I I I I I I 2.0 I I CH3-C10H14 solv _/ 1.0 /:.__ solv CHTC10H14 - - /i j PPM )'~.)-~_.__/ ti~...... } .._,__ ______ _ 1 Si(CH3h Si(CH3h 0 CJl 51 1,2,4 substitution of 46 and 47 are assigned on the basis of similar chemical reactivities of Abp-derived metallocenes (vide infra) to related tert-butyl substituted complexes, whose 1,2,4-substitutions have been confirmed by single crystal X-ray crystallographic methods.31,44 The dilithio salt 37 is soluble not only in ethereal solvents such as THF and Et20, but also in hydrocarbon solvents such as toluene and petroleum ether, which precluded its isolation and purification. Analogous solubility problems are observed for the dilithio salt of the analogous tert-butyl substituted ligand; efforts to precipitate 46 through addition of coordinating bases are not undertaken because of previous difficulties encountered in the transmetallation of DME- and TMEDA-containing dianionic ligands.32 In contrast to the dilithio salt, 47 is insoluble in hydrocarbon solvents, and is almost completely insoluble in ethereal solvents such as THF. Only upon addition of stoichiometric quantities of the K+-coordinating crown ether 18-Crown-6 could 47 be rendered sufficiently soluble in THF for 1 characterization by H NMR spectroscopy. Addition of potential donor ligands such as DME fails to enhance the solubility properties of 47. Efforts to obtain Group III metallocenes of both 44 and 47 were undertaken in order to assess the ability of the methyladamantyl groups and the a- trimethylsilyl groups to induce formation of racemo- metallocenes. The reaction of Li2 Adp (44) with ScC13 (THF h_ 5 proceeds readily in THF over 12 hours at room temperature. Examination of the crude product mixture indicates the formation of exclusively m es o-AdpScCl-LiCl-(THFh (48) (Figure 30), as evidenced by the inequivalent linking silyl-methyl 1 resonances in the H NMR spectra of 48; no racemo products are observed. 52 S1 /\ Me Me MC13 (THFh_ 5 / MeAd ~ MC13 (THFh, 5 / (M = Sc, Y)'\ MeAd THF I M ..... ,.,c1,,. .... L. -......,. C1..---- \ THF MeAd MeAd MeAd M)i . ...... ~-CI..___L ,...,,,THF MeI meso-AdpMCl-LiCl-(THF)i ., / l"' Cl THF MeAd racemo-AdpMCl-LiCl-(THFh (not observed) Figure 30. Formation of exclusively mes o- AdpMCl-LiCl-(solvent) 2 complexes; M = Sc (48), Y (49) . Although NMR spectroscopy of the crude reaction mixture indicates a high (>85%) yield of 48, separation of 48 from LiCl generated during transmetallation proved problematic, as both 48 and LiCl are soluble in THF and insoluble in toluene and diethyl ether. An identical selectivity form es o- metallocene formation is observed in the reaction between 44 and YC13 (THFh_ 5 (48 hours, 25 °C, THF). mes o-Adp YCl-LiCl-(THFh (49) is obtained as a tan solid (56% yield) after removal of LiCl through a series of toluene and Et2 O washes and filtrations. A small amount (~ 18%) of the tetrahydrofuran originally coordinated in 49 is replaced by diethyl ether over the course of these washings; two equivalents of solvent appear coordinated to Li throughout the workup procedure. 53 The assignment of 48 and 49 as LiCl-bis(solvent) adducts is based on the established tendency of highly Lewis acidic group III and lanthanide metallocenes to coordinate Li-halides,49 and the observation of two equivalents of coordinating solvent in the metallocenes' 1H NMR spectra. Repeated attempts to grow single crystals of 48 and 49 suitable for X-ray diffraction to confirm the solid state structures were unsuccessful. The formation of exclusively mes o-metallocenes is consistent with previous findings that even relatively bulky alkyl substituents residing in the ~- positions of singly linked cyclopentadienes do not necessarily lead to an increased selectivity for formation of racemo- complexes. Increasingly bulky ~- substituents, in fact, seem to favor the formation of mes a- complexes; analogous transmetallations using the tert-butyl substituted ligand array [Me2 Si(3-CMe3 -C5H 3)z]Li2 (abbreviated [Dp ]Li2 ) showed exclusive formation of [mes o-DpScCl]n, and a 1:1 mes o:racemo ratio in the formation of [Dp YClJn.28 The thermodynamic preference for formation of mes a- metallocenes of singly silyl-linked bis(3-alkyl-cyclopentadienyl) ligand arrays has been demonstrated previously by Jordan and co-workers.SO The transmetallation of [Abpf (46/47), which contains trimethylsilyl substituents adjacent to the Me2 Si linker to generate group III metallocenes proceeds with diastereoselectivities opposite to those observed for [Adp]. Treatment of K2 Abp (47) with either ScC13 (THFh 5 or YC13 (THFh 5 (refluxing THF, 48 hours) results in the formation of exclusively racemo- metallocenes. Steric interactions in the narrow portion of the metallocene wedge apparently disfavor formation of the mesa- complexes, in which the SiMe3 groups adopt eclipsed conformations. 54 2- (46or47) Si /\ Me Me meso-AbpMCl (not observed) racemo-AbpMCl Figure 31. Metallation using the [Abp] dianion results in exclusively racemometallocenes (M = Sc or Y).51 Solutions of refluxing THF are required to promote transmetallation; the metallations are unsucessful when attempted in toluene or benzene. The limited solubility of K2 Abp in THF is undoubtedly a factor contributing to the extended periods of heating required to drive the metallations to completion. When followed by 1 H NMR spectroscopy, treatment of K2 Abp and ScC13 (THFh.s in THF-d 8 results in the formation of a C2 -symmetric kinetic product, assigned as the KCl-bis(solvent) adduct rac-AbpScCl-(KCl)-(THFh (50) (Figure 32),52 which is then converted to two new products, each of which contains exclusively racemo-metallocenes (51, 52). 55 racemo-AbpMCl-KCl-(THF)i (50, M=Sc) [racemo-AbpM(µ-Cl)]z homochiral (51 M=Sc- 53 M=Y) I I I Figure 32. Metalation of 47 affords rac-AbpMCl-(KCl)-(THFh, which is subsequently converted to rac-AbpMCl(THF) and [rac-AbpM(µ-Cl)h (M=Sc, Y). 51 and 52 are assigned as the THF adduct rac-AbpScCl(THF) and the chloride bridged dimer [racemo-AbpSc(µ-Cl)h.51 These assignments are supported by the previous observation of multiple transmetallation products for [Me2Si(2-SiMe3 -4-CMe3)i]K2 with ScC13 (THFh.s, which are assigned as the analogous THF adducts and (µ-Clh dimers.30 Also possible, however, are equilibria involving homochiral and heterochiral dimers of [rac-AbpYCl]. As the transmetallations are performed in THF, it is unlikely that the monmeric, base-free [rac-Abp YCl] is a stable metallation product under the observed conditions. Conversion of the kinetic product to 51 and 52 begins prior to total consumption of the starting materials, and the ratio of 51 to 52 does not deviate from the 1:3 ratio attained after 48 hours at 80 °C. Removal of solvent (THF) followed by extended heating in benzene-d 6 does not result in 56 additional conversion of 51 to 52; decomposition to unidentifiable products is, instead, observed. A similar mixture of products is observed in the transmetallation of 47 with YC13 (THFh_ 5; the 2:1 mixture of thermodynamic products observed upon treatment of 47 with YCliTHFh_ 5 (THF, 48 hours, 80 °C) are assigned as the homochiral chloride-bridged dimer [rac-AbpY(µ-Cl)h (53) and racAbp YCl(THF) (54). Alternate possible transmetallation products, however, include the homochiral and heterochiral dimers of [racemo-Abp YCl] and racAbp YCl-(KCl)-(THFh. Nonstoichiometric quantities of THF (~0.2 equivalents THF-d 0 per Y) accompany the yttrocenes (as determined by 1 H NMR spectroscopy), despite repeated lyophilizations and exposures to dynamic vacuum. Although the ratio of major:minor products varies somewhat depending on the workup conditions, the ratio is largely independent of the amount of residual THF present. A single racemo- organometallic product is observed (1H and 13C NMR spectroscopy), however, when the [Abp YCl] product mixture is dissolved in C6D6 . Based, in part, on the previously observed thermodynamic stabilities of (µ-Cl) 2 dimers,30,32,44 this product is assigned as the chloride-bridged dimer 53. When benzene is removed in vacuo and replaced by THF-d 8 , the two racemo- products 53 and 54 are observed in a c.a. 2:1 ratio. This disruption of (µ-Cl) 2 dimers in the presence of donor ligands (THF-d 8 ) may serve as preliminary evidence for the ability of the methyladamantyl substituents to destabilize dimer formation relative to monomeric (albeit donor-stabilized) complexes. Conclusions The syntheses of a series of multiply silylated cyclopentadienes incorporating either cyclohexyl- (25), 2-adamantyl- (28, 38), or 2-(2-methyl)adamantyl groups (45) are described. These cyclopentadienes exist in solution as a mixture of rapidly interconverting isomers as result of the ability of silylated cyclopentadienes to undergo spontaneous hydrogen and silyl 57 migrations. Mixtures of constitutional isomers result from the double deprotonation of the cyclohexyl- and 2-adamantyl-substituted ligand arrays. The products of these deprotonations are believed to be constitutional isomers of the desired dianionic ligands. Double deprotonation of the 2-(2methyl)-adamantyl substituted cyclopentadienes results in exclusive formation of the [Me2 Si(2-(SiMe3 )-4-[2-(2-CH3-C 10H 14 )]-C5H 2 b]M2 regioisomer (M = Li (46); K (47)) . These results show that a tertiary alkyl group is required in order to direct the trimethylsilyl substituents selectively to the position adjacent to the silyl linker. The incorporation and 2,4-disposition of trimethylsilyl and alkyl substituents is essential for the formation of chiral metallocenes. Transmetallation of 47 with MC13 (THFh_ 5 result in exclusively racemometallocenes. Steric interactions between the trimethylsilyl substituents m the narrow portion of the metallocene wedge are believed to preclude formation of themeso- (achiral) conformer. Transmetallation of Me2 Si/3-[2(2-CH3 -C 1J{ 14 )]-C5H 3 ) 2 ]Li2 (46) proceeds with enantioselectivity opposite to that of 47, resulting in exclusively mesa- metallocenes 48 and 49. Experimental Section General Considerations: All air and/ or moisture sensitive compounds were manipulated using standard high vacuum line, Schlenk, or cannula techniques or in a dry box under a nitrogen atmosphere, as described previously.53 Argon and hydrogen gases were purified and dried by passage over columns of MnO on vermiculite and activated molecular sieves. Solvents were stored under vacuum over titanocene54 or sodium benzophenone ketyl. Pentamethylenefulvene was prepared according to the procedure of Stone and Little.45 ScC13 (THFh_ 5 was prepared using the procedure of Manzer;SS YCliTHFh.s was prepared by an identical route. LiCH(SiMe3)i was prepared according to the procedure of Cowley.56 SiC14 , Me2 SiC12 , Me3SiCl, and Et3 N were dried over CaH2 and stored under vacuum. LiCH2 SiMe3, KOCMe3 and KN(TMS)i were purchased from Aldrich and sublimed prior to 58 use. LiAlH4 was purchased from Aldrich and dissolved in diethyl ether prior to use. TMEDA was stored over Na, and DME was dried first over sodium benzophenone ketyl and then stored over potassium benzophenone ketyl. 18-crown-6 was dried over activated 4A sieves and sublimed twice prior to use. Vitride was purchased from Kodak and used as received . All other reagents were purchased from Aldrich and used as received. NMR spectra are recorded on a Bruker AM500 (500.13 MHz for 1 H, 13 125.76 MHz for C) or a GE QE-300 (300 MHz for 1 H, 75.5 MHz for 13C) spectrometer. All shifts reported for proton and carbon NMR are referenced to solvent. NMR tube reactions are carried out using flame sealed NMR tubes or J. Young type Teflon-valved NMR tubes purchased from Wilmad. Mass spectra were acquired with the help of Dr. Peter G. Green at the Caltech Environmental Analysis Center using an HP 59880B Particle Beam LC/MS Interface to an HP 5989A Mass Spectrometer Engine; air /water exclusion was attempted using the HP 1090 Liquid Chromatograph's Solvent Delivery System and Variable Volume Injector. Elemental analyses were performed at the Caltech Elemental Analysis Facility by Fenton Harvey'. Preparation of [Cy-C5H 4 ]K-B(sec-Bu)3 (21-BR3 ): A 50 ml roundbottom flask is charged with 2.0 g (13.7 mmol) of pentamethylenefulvene, and is attached to a small frit with a 50 ml receiving roundbottom flask . After evacuation, 25 ml Et20 is vacuum transferred into the flask, affording a clear yellow solution upon warming to room temperature. After cooling the roundbottom and its contents to O °C with an ice-water bath, 13.6 mmol potassium tri-secbutylborohydride (K-Selectride, 13.6 ml, 1.0 M solution in THF) is injected into the flask against an argon counterflow. The ice bath is removed, and the resulting pale yellow solution is allowed to stir at 25 °C for 10 hours. Solvent is removed in vacuo, and the resulting tan solid is dried under dynamic vacuum for 1 hour. Diethyl ether (25 ml) is vacuum transferred onto the solid, and a small amount of finely divided white solid is filtered away from the ether solution, and is washed twice with diethyl ether. Solvent is removed in vacuo, leaving a tacky tan solid in the receiving roundbottom flask. A new 50 ml flask is attached to the frit assembly in the glove box, and petroleum ether (35 ml) is vacuum transferred onto the solid. The petroleum ether (PE) insoluble material is isolated by filtration, and is washed twice with PE, affording 2.037 grams of beige powder (10.9 mmol, 80% yield). 59 Preparation of (SiMe3 )(Cy)C5H 4 : A 250 ml roundbottom flask is charged with 17.2 mol of 21 and is attached to a swivel frit assembly. THF (250 ml) is transferred onto the solid. Chlorotrimethylsilane (2.5 ml, 19.7 mmol, 1.15 eq.) is added at -78 °C. The solution is warmed 55 °C and stirred for three hours. After cooling to room temperature, solvent is removed in vacuo, and petroleum ether (15 ml) is distilled into the flask. LiCl is removed by filtration, and is washed twice with petroleum ether. Solvent is removed in vacuo, leaving a clear yellow oil. A short path distillation (60 °C, < 1µ) yielded 1.76 g (8.0 mmol, 47% yield) of a yellow oil. Preparation of [1-SiMe3 -3-Cy-C5H 3 ]K (24): In a glove box, a 250 ml roundbottom flask is charged with 1.75 g (8.0 mmol) of (SiMe3 )(Cy)C5 H 4 and 120 ml THF. Potassium tert-butoxide (888 mg, 7.92 mmol) is added slowly to the stirred solution. After stirring for 21 hours at room temperature, solvent is removed from the clear orange solution in vacuo. Excess tert-butanol is removed by heating under dynamic vacuum (6 hrs, 65 °C). The flask is attached to a swivel frit assembly, and 1.59 g of a white petroleum ether solid is isolated (6.15 mmol, 77% yield). Preparation of [Cy-C5H 4 ]Li (20): In a 250 ml roundbottom flask is placed 40 mmol of pentamethylenefulvene, and the flask is attached to a medium porosity swivel frit assembly. Diethyl ether (125 ml) is vacuum transferred onto the fulvene and is warmed to 0 °C, resulting in a clear yellow solution. LAH (1.0 M in Et2 O, 50 ml, 50 mmol) is syringed into the flask against an Ar counterflow over the course of 30 minutes, and the solution is stirred at 0 °C for 25 minutes. The 0 °C bath is removed, and the solution is stirred at 25 °C for an additional 12 hours. The white diethyl ether-insoluble material is isolated by filtration and is dried in vacuo, affording 3.998 g of a flocculent white solid (26 mmol, 65% yield). Preparation of Me2 Si(Cy-C5H 4) 2 (22): A 100 ml roundbottom flask is charged with 3.934 g (25.5 mmol) of 20, and is attached to a medium porosity swivel frit . Tetrahydrofuran (80 ml) is distilled onto the solid, and the solution kept at -78 °C while 1.15 ml (9.46 mmol) Me2 SiC12 is vacuum transferred first into a graduated cold finger and then into the THF solution. The solution is stirred 60 for 11 hours as the dry ice/acetone bath slowly warmed to room temperature, leaving a yellow solution with grey precipitate. Solvent is removed in vacuo, and the resulting material is dried under dynamic vacuum for 4 hours. Petroleum ether (60 ml) is distilled into the roundbottom, and a grey solid (LiCl) is removed from the yellow solution by filtration; the solid is washed 5 times will small portions of PE. The petroleum ether is removed in vacuo, leaving 3.062 g (8.68 mmol) of a yellow oil (92%) . Preparation of [Me2 Si(3-Cy-C5H 3 ) 2 ]Li2 (26a): A 500 ml roundbottom flask is charged with 7.999 g (22.7 mmol) of 22, and is attached to a large swivel frit equipped with a 90° needle valve and a 500 ml receiving roundbottom. Petroleum ether (250 ml) is distilled onto the yellow oil, producing a clear yellow solution when warmed to O C. A 1.6 M solution of n-butyllithium (31 ml, 49.6 mmol) is syringed into the solution against an argon counterflow, and the solution is stirred at O °C for 2 hours. The resulting solution, containing a large amount of white precipitate, is stirred at 25 °C for an additional 2 hours, at which time the solution is filtered, and the solid 0 washed is washed 5 times with small aliquots of petroleum ether. Solvent is removed and the white solid is dried in vacuo, affording approximately 5.66 g (15.5 mmol) of 26a (68%). Preparation of [Me2 Si(3-Cy-C5H 3 ) 2 JK2 (26b): In the glove box, a 25 ml roundbottom flask is charged with 570 mg (1.62 mmol) 22, and 15 ml of THF is added vi a pipet. To this clear yellow solution is added 357 mg KOCMe3 (sublimed, 3.18 mmol, 1.96 eq.); the frit assembly is placed on the vacuum line, and the solution is stirred at 25 °C for 16 hours. Solvent is removed in vacuo, and the resultant brown solid is dried under vacuum (70 °C) for 9 hours. Petroleum ether (20 ml) is distilled into the flask, and 404 mg of a tan PE insoluble solid is isolated after filtration (0.93 mmol, 57%). Preparation of Me2 Si[(SiMe3 )(Cy)C5H 3 ]2 (25): A medium porosity swivel frit is affixed with a 90° needle valve and two 100 ml roundbottom flasks, one of which contains 1.214 g (3.33 mmol) of 26a. The frit assembly is evacuated, and 50 ml of THF is distilled onto the solid. Trimethylchlorosilane (1.05 ml, 8.27 mmol) is vacuum transferred into a graduated cold finger, and is then transferred into the reaction flask at -78 °C. The dry ice / actone bath is 61 removed, and the solution allowed to warm with stirring to room temperature; solvent is removed from the solution after approximately 44 hours of stirring at 25 °C. Petroleum ether (50 ml) is distilled onto the product mixture, and the resulting white precipiate (LiCl) is filtered easily from the bright yellow solution. The solid is washed three times with petroleum ether, and then PE is removed in vacuo, affording 1.293 g (2.60 mmol) of orange oil after Kugelrohr distillation (79%). Mass spectroscopy on the mixture provides a parent peak at (m/z) = 496 (expected for C3J{ 52Si3 = 497.09 g/mol). Preparation of Admantylfulvene (30): To a stirred solution of 2- adamantanone (24.85 g, 165.4 mmol) and freshly cracked cyclopentadiene (66.0 ml, 800.6 mmol) in 500 ml of methanol is added 17.0 ml (204 mmol) of pyrrolidine. Yellow solid began precipitating from the burgundy-colored solution within several hours of pyrrolidine addition. After stirring for 11 days, the solution is acidified with 33 ml of glacial acetic acid and diluted with 400 ml of water and 100 ml of diethylether. The liquids are transferred to a 2liter separatory funnel, to which is added an additional 100 ml of water and 400 ml ether. The aqueous layer is washed three times with 375 ml and once with 75 ml of ether, and the combined organic layers are washed once with 150 ml water and once with 150 ml brine. After drying over and gravity filtering from magnesium sulfate, the organic solution is rotovapped to dryness, leaving an orange-brown solid. Bright yellow crystals (28.69 g, 144.7 mmol, 88% yield) are obtained upon recrystallization from petroleum ether. Elemental analysis: Calculated (Found, with V2Os added) C: 90.83 (90.97, 90.92); H: 9.17 (8.97, 9.05). Preparation of [(2-Ad)-C5H 4 ]Li (31a): Adamantylfulvene (30) (10.829 g, 55 mmol) is placed in a 500 ml roundbottom flask on a large swivel frit assembly, and 250 ml Et2 O is distilled in from Na / benzophenone ketyl. At 0 °C, 60 ml of a 1.0 M solution of LiAlH4 (Et2 O) is syringed into the clear yellow solution against an Ar counterflow. The solution is stirred at 0 °C for 30 minutes, and then at room temperature overnight. A large amount of white solid is isolated by filtration, washed 3 times with Et2 O, and dried under vacuum, affording 12.27 g of 31a-(Et20) 0_2 (55 mmol, 100%). The solid may be then subjected to an aqueous workup by making a suspesion of 31a in diethyl 62 ether (3.85 g 31a/150 ml EtiO, 250 ml roundbottom flask), adding water (2.5 ml) via syringe pump (0.17-0.25 µ1/min) (USE EXTREME CAUTION WHEN QUENCHING RESIDUAL AlH3 ), followed by 2.5 ml of an aqueous 15% NaOH solution. More water (15 ml) is added to the flask, and the contents of the flask are transferred to a 250 ml separatory funnel. The aqueous layer is washed once with 20 ml EtiO, and the combined organics are dried over magnesium sulfate and filtered. Solvent is removed in vacuo, leaving a light green oil ((2-Ad)-C5H 5 ) in ~70% yield. Deprotonation with n-butyllithium proceeds by adding 40 ml petroleum ether to a solution of 1.4606 g (7.29 mmol) of (2-Ad)-C5 H 5, cooling the solution to O °C, and syringing 4.0 ml of a 2.0 M solution of n -BuLi in pentane into the cooled roundbottom flask against an Ar counterflow. The solution is stirred at room temperature for three hours, during which time white solid precipitates from solution. Solvent is removed in vacuo, and 930 mg of a white petroleum ether insoluble material is isolated and washed once with fresh pet ether (4.30 mmol, 59% yield of 31a). Elemental analysis: Calculated (Found, with V2O5 added) C: 83.26 (86.55, 87.81); H: 13.54 (9.70, 10.41). Calculated (Found, without added oxidant) C: 83.26 (85.48, 84.80); H: 13.54 (9.90, 9.77). Preparation of [(2-Ad)-C5H 4 ]Na (31b): A 100 ml roundbottom flask is charged with 1.027 g (5.18 mmol) 30. Petroleum ether (50 ml) is distilled onto the yellow solid and is warmed to O C. Vitride (NaAl(OCH2 CH2 OCH3) 2H 2 , 70% solution in toluene, 2.0 ml, 7.22 mmol) is syringed into the clear yellow solution against an Ar counterflow, and the solution stirred at O °C for 0 approximately 10 minutes. After warming to room temperature and stirring for an additional 18 hours, solvent is removed in vacuo, and the resulting solid is washed twice with petroleum ether and then filtered from toluene. 31b is isolated as a toluene-insoluble white solid (599 mg, 2.69 mmol) in 52% yield. Preparation of (SiMe3 )(2-Ad)C5 H 4 (32): A 500 ml roundbottom flask equipped with a 180° needle valve is charged with 6.765 g (32.8 mmol) of 31a. Onto this white solid is distilled 200 ml of THF. Trimethylchlorosilane (4.9 ml, 38.6 mmol) is vacuum transferred onto the suspension at -78 °C. The solution is stirred at -78 °C for 30 minutes, and then warmed to room temperature and stirred overnight. Solvent is removed in vacuo, and the flask is brought into 63 the glove box and attached to a large swivel frit assembly. Petroleum ether (250 ml) is vacuum transferred onto the crude reaction mixture, generating a colorless solution with a small amount of white precipitate. The solid (LiCl) is filtered away from the colorless solution, and is washed twice with small portions of petroleum ether. Removal of solvent from the clear, colorless solution gave 8.773 g (32.2 mmol) of a viscous oil (98%). Preparation of (SiMe2 Cl)(2-Ad)C5 H 4 (33): A 100 ml roundbottom flask is charged with 559 mg (2.71 mmol) of 31a and attached to a medium porosity swivel frit assembly. THF (40 ml) is distilled onto the solid, giving a clear and colorless solution. The solution is cooled to -78 °C, onto which dimethylchlorosilane (0.5 ml, 4.12 mmol, 1.53 eq.) is transferred. The solution is allowed to warm slowly to room temperature as the dry ice/ acetone bath slowly evaporates. After 12 hours of stirring at 25 °C, volatiles are removed from the solution in vacuo. A grey petroleum ether (25 ml) insoluble solid is removed by filtration, leaving a clear yellow liquid. Solvent is removed, affording 780 mg of a clear yellow oil (2.67 mmol, 98% yield). Preparation of [1-SiMe3 -3-(2-Ad)-C5H 3 ]Li (34): 32 (8.773 g, 32.2 mmol) is placed in a 500 ml roundbottom flask attached to a 180 °C needle valve. Petroleum ether (300 ml) is transferred into the flask, and the solution brought to 0 C. n-Butyllithium (1.6 M in hexanes, 24.0 ml, 38.4 mmol) is syringed dropwise into the solution against an Ar counterflow followed by 60 minutes of stirring at 0 0 C. After warming to room temperature, the solution is stirred for 22 hours, accompanied by the evolution of gas and precipitation of a large amount of white solid. The solution is degassed and brought into a glove box, where the solid is washed and suction filtered 4 times through a coarse filter frit with a total of 175 ml of petroleum ether. The petroleum etherinsoluble white solid is dried in vacuo, affording 8.77432 g (31.4 mmol) of 34 (98%) . Elemental analysis: Calculated (Found) C: 77.63 (75.56, 74.97); H: 9.79 (9.50, 9.56); N: 0.00 (0.00, 0.00). 0 Preparation of (SiMe2 Cl)(SiMe3 )(2-Ad)C5H 3 (35): In a glove box, a J-Young NMR tube is charged with 15.3 mg (0.0550 mmol) of 34. The teflon valve is sealed and the NMR tube evacuated on the vacuum line. THF-d8 (0.4 ml) is distilled onto the solid and the solution is warmed to room temperature, 64 giving a clear and colorless solution. The tube is cooled with a LN2 bath, and (using a calibrated gas bulb) 0.0275 mmol of Me2SiC12 is vacuum transferred into the tube from CaH2 . The tube is sealed and warmed slowly to room temperature. Half of the starting 34 had reacted within 15 hours at room temperature. No further reaction was observed (3 days at room temperature) . Heating the NMR tube resulted in decomposition to unidentifiable products (days, 80 °C). Preparation of (SiC13 )(SiMe3 )(2-Ad)C5 H 3 (36): A 25 ml roundbottom flask is charged with 136 mg (0.49 mmol) 34 and is placed on a small medium porosity swivel frit. Toluene (13 ml) is distilled onto the solid and the suspension warmed to room temperature, then cooled to -78 °C. Silicon tetrachloride (0.24 mmol) is measured into a calibrated 42.5 ml gas bulb, and is then transferred into the reaction flask at -78 °C. The dry ice/ acetone bath is allowed to warm to room temperature and the solution is stirred at 25 °C for 4 days. The white solid is filtered away from the yellow solution and is washed three times with small aliquots of toluene. Solvent and . volatiles are removed in vacuo from the clear yellow toluene solution, leaving 136 mg of a clear yellow oil (0.38 mmol, 78%). Elemental analysis: Calculated (Found) C: 67.34 (63.64, 62.90); H: 8.49 (8.21, 8.53); N: 0.00 (0.00, 0.00). Mass spectral analysis indicates a parent peak at (m/ z) = 406 (C 18H 27Si2 Cl3 = 405.98), with no peak at (m/z) = 643 (corresponding to Cl2 Si(AdTMSCpH)i = C 36H 54 Si3 Cl2 ). Also present in the mass spectrum are smaller peaks at m/z = 389 and 369, corresponding to the products resulting from hydrolyses of one and two, respectively, of the Si-Cl bonds in 36, during introduction into the spectrometer. Preparation of Cl2 Si[(SiMe)(2-Ad)C5H 3 ]2 (37): A 25 ml roundbottom flask is charged with 303 mg (1.09 mmol) of 34. Tetrahydrofuran (12 ml) is distilled onto the white solid, giving a clear and colorless solution at 35 °C. Silicon tetrachloride (0.535 mmol) is transferred into a 42.5 ml gas bulb and then into the reaction flask, which had been cooled to -78 °C. The dry ice / acetone bath is allowed to warm slowly to 25 °C as the solution is stirred overnight. Solvent and volatiles are removed in vacuo, and the roundbottom flask is attached to a medium porosity swivel frit assembly. Petroleum ether (13 ml) is distilled onto the light yellow solid, resulting in a yellow solution and fine 65 white precipitate. The white solid (LiCl) is removed by filtration and is washed three times with small portions of petroleum ether. Solvent is removed in vacuo from the solution, leaving 245 mg (0.382 mmol) of a faint yellow microcrystalline solid (72% yield). Elemental analysis: Calculated (Found) C: 67.34 (66.97, 66.84); H: 8.49 (8.54, 8.59); N: 0.00 (0.00, 0.00). Mass spectroscopy indicates a parent peak at (m/z) = 642 (expected for C36H 54 Si3 Cl2 = 642.07) . Preparation of Me2 Si[(SiMe3 )(2-Ad)C5 H3 ]z (28): A 250 ml roundbottom flask is charged with 2.366 g (3.67 mmol) of 37 and placed on a fine porosity swivel frit assembly. Diethyl ether (150 ml) is distilled onto the solid, and the solution is warmed to 0 C. Methyllithium (1.4 M in Et2O, 13.0 ml, 18.2 mmol) is syringed into the flask against an Ar counterflow and the ice bath is allowed to warm slowly to room temperature as the solution is stirred overnight. The faint yellow solution with white precipitate is cooled to -78 °C and 0.8 ml (19.74 mmol) CH3 OH is syringed into the flask. The solution is stirred and warmed slowly to room temperature for 4 hours. Solvent and volatiles are removed in vacuo and 20 ml petroleum ether is distilled onto the product mixture. The white solid (LiCl) is removed from the light yellow solution by filtration, and is washed twice with petroleum ether. Solvent is removed in vacuo , affording 986 mg (1.64 mmol) of a light yellow microcrystalline solid (45% yield). Elemental analysis: Calculated (Found) C: 75.91 (76.90, 77.56); H: 10.08 (10.24, 10.42); N: 0.00 (0.00, 0.00). The yellow solid is placed in a small sublimator and yellow solid began collecting on the cold finger as the heating bath temperature reached 147 °C. The sublimator is heated at 165 °C for 4 0 hours, at which time 407 mg (0.41 mmol) of light yellow solid could be scraped from the cold finger. Preparation of [(±)C20H 12O2 ]Si[(SiMe3 )(2-Ad)C5H 3 ] 2 (38): A 100 ml roundbottom flask is charged with 976.5 mg (1.521 mmol) of 37 and 433.2 mg (1.513 mmol, 1.00 eq.) of (±)1,l'-bi-2-naphthol, attached to a 180° needle valve, and degassed on a vacuum line. THF (50 ml) is distilled into the flask (-78 °C), as is 0.6 ml Et3 N (4.305 mmol, 2.85 eq.). The solution is allowed to warm to room temperature for 10 hours. Solvent is removed and the resultant yellow solid dried in vacuo. The flask is attached to a medium porosity frit assembly (in the glove box); 50 ml of diethyl ether is distilled onto the solid, resulting 66 in a yellow solution and a considerable amount of white solid. Et3 N-HC1 is removed by filtration, and the remaining Et2 O-soluble yellow solid is dried under dynamic vacuum and then subjected to an Et2 O filtration through a fine frit. Solvent is removed in vacuo, leaving 974.3 mg of a microcrystalline 1 yellow solid ( H NMR indicates the presence of 0.71 equivalents of residual diethyl ether); 1.07 mmol, 71 % yield. Preparation of [(MeCy)C5 H4 ]Li: A 100 ml roundbottom flask is charged with 1.359 g (8.58 mmol) of pentamethylenefulvene, and is attached to a medium porosity swivel frit assembly. Petroleum ether (50 ml) is distilled onto the solid, generating a clear yellow solution. The solution is cooled to -78 °C and 6.5 ml 1.4 M CH3Li (9.1 mmol) is syringed, dropwise, into the solution against an Ar counterflow. The resulting cloudy yellow solution is stirred at -78 °C for 15 minutes, and is allowed to warm slowly to room temperature with stirring. After 26 hours of stirring, the solution is filtered, leaving a tan petroleum ether-insoluble powder, which is dried in vacuo (1.266 g, 7.52 mmol, 88% yield). Preparation of {[2-(2-Me-C 10H 14 )]-C 5H 4 }Li-DME ([MeAd-C 5H 4 ]Li-DME) (39DME). A 100 ml roundbottom flask equipped with a 180° needle valve is charged with 1.427 g (7.19 mmol) of 30. Diethyl ether (50 ml) and dimethoxyethane (1. 1 ml, 10.6 mmol) are distilled onto the solid and the resulting clear yellow solution is cooled to 0 C. Methyllithium (1.4 M in Et2 O, 6.0 ml, 8.4 mmol) is syringed into the solution against an Ar counterflow and the solution is stirred at 0 °C for 20 minutes. After warming to room temperature, the yellow solution with white precipitate is stirred at 25 °C for 70 hours. Solvent is removed in vacuo and the reaction flask is affixed to a coarse swivel frit; 50 ml of petroleum ether is vacuum transferred 0 into the flask and a white solid is removed from the colorless solution by filtration. The petroleum ether insoluble material is washed 3 times with aliquots of petroleum ether, and dried in vacuo, yielding 1.947 g of a flocculent white solid. 1 H NMR spectroscopy (THF-d 8 ) indicates that the solid is 39 accompanied by a non-stoichiometric amount of DME (0.55 equivalents); isolated yield = 6.50 mmol, 90%. Elemental analysis: Calculated (Found) C: 80.99 (77.98, 78.44); H: 9.92 (10.06, 10.19); N: 0.00 (0.00, 0.00); these elemental 67 analysis numbers are more consistent with the presence of 0.88 equivalents of coordinated DME: C: 78.21; H: 10.04; N: 0.00. Preparation of [MeAd-C5H4 ]Li-TMEDA (39-TMEDA) : A 2-liter, 3-necked roundbottom flask, equipped with a rubber septum and a 180° needle valve, is charged with 83.30 g (420 mmol) of 30. With stirring, 1100 ml of diethyl ether and 64 ml (424 mmol, 1.01 eq.) of N,N,N',N'-tetramethylethylenediamine are cannulaed into the flask. The solution is cooled to 0 °C and 300 ml of 1.4 M methyllithium (Et2O, 420 mmol, 1.00 eq.) is cannulaed into the flask over 1 hour. White solid began to precipitate from the otherwise clear and bright yellow solution during the methyllithium addition. The solution is allowed to warm slowly to 25 °C and is stirred at 25 °C for approximately 6 days. Solvent is removed in vacuo, and the remaining white solid is washed four times with 150 ml aliquots of petroleum ether. The petroleum ether insoluble material is dried in vacuo for 6 hours, providing 132.5 g (393.7 mmol, 94% yield) of a fluffy white solid. Elemental analysis: Calculated (Found, with V2Os added) C: 78.51 (77.95, 78.43); H: 11.10 (10.30, 11.03); N: 8.33 (8.29, 8.46). Preparation of (SiMe3 )(MeAd)C5H 4 : A 100 ml roundbottom flask is charged with 1.537 g (4.57 mmol) of 39-TMEDA and attached to a small, fine porosity swivel frit assembly. Tetrahydrofuran (50 ml) is distilled onto the solid, yielding a clear light yellow solution at 0 C. The solution is cooled to -78 °C, and 0.8 ml (6.30 mmol) of Me3 SiCl is vacuum transferred into a graduated cold finger, and then into the reaction flask. The -78 °C bath is removed and the solution allowed to warm to room temperature and stir for 12 hours. Solvent is removed in vacuo and 50 ml petroleum ether is distilled onto the tan solid. White solid (LiCl) is removed by filtration and is washed twice with petroleum ether. Solvent and volatiles are removed from the clear yellow solution, leaving 1.282 g (4.47 mmol) of tan solid (98%). 0 Preparation of [1-SiMe3 -3-MeAd-C5H 3 ]Li-DME (40-DME): II-125 & 129 In the glove box, a 1-liter roundbottom flask is charged with a magnetic stir bar, 23.475 g (81.91 mmol) of (SiMe3 )(MeAd)C5H 4 and 450 ml of petroleum ether. A 180° needle valve is attached to the flask and the apparatus is placed on the vacuum line and cooled to 0 C. n-Butyllithium (1.6 M in hexanes, 60 ml, 96 0 68 mmol) is syringed into the flask against an argon counterflow over the course of 25 minutes. The solution is stirred at 0 °C for 10 minutes, at which time the ice bath is removed and the solution allowed to warm to room temperature and stirred overnight, leaving an extremely viscous brown paste at the bottom of the otherwise pale yellow solution. Solvent is removed in vacuo, leaving 25.8623 g of tan powder. The flask is brought into a glove box, and 2.0636 grams of the powder is removed for other reactivity studies. To the remaining 23.799 g (~80 mmol) of tan powder is added 250 ml of petroleum ether. Dimethoxyethane (9.4 ml, 90.4 mmol) is pipetted dropwise into the colorless solution containing the gooey orange solid, accompanied by the immediate precipitation of white solid. The solution is stirred vigorously overnight in order to break up clumps of starting material. The white solid formed is isolated by suction filtration, is washed three times with 50-60 ml portions of petroleum ether, and is dried in vacuo, affording 26.8962 g of flocculent white solid (93% yield). Elemental analysis: Calculated (Found) C: 72.19 (71.82, 72.29); H: 10.29 (10.23, 10.39); N: 0.00 (0.00, 0.00). Preparation of (SiMe2 Cl)(SiMe3 )(MeAd)C5H3 : In the glove box, a J-Young NMR tube is charged with 14.2 mg (0.037 mmol) of 40. Into the tube is vacuum transferred 0.4 ml THF-d 8 (-78 °C), and the contents of the tube warmed to room temperature, affording a clear yellow solution. The contents of the NMR tube are frozen with a liquid nitrogen bath, and 0.018 mmol Me2 SiC12 is vacuum transferred from CaH2 to a calibrated 3.3 ml gas bulb to the J-Young tube. The contents of the NMR tube are warmed to room temperature and agitated overnight. After 15 hours at room temperature, 1 H NMR spectroscopy indicated that all of the Me2 SiC12 had reacted, as had approximately half of the starting Li salt. This product mixture changed neither after an additional 6 hours of agitation at 25 °C, nor after 16 additional hours of heating at 80 °C. The presence of diastereotopic Me2 SiCl methyl 1 groups, and the absence of allylic cyclopentadiene resonances, in the H NMR spectrum of the new compound are characteristic of a doubly silylated, alkylsubstituted cyclopentadiene. Preparation of (SiC13 )(SiMe3 )(MeAd)C5H3 : In the glove box, a J-Young NMR tube is charged with 13.0 mg (0.044 mmol) of 40. THF-d 8 (0.4 ml) is vacuum transferred into the tube, and the contents of the tube shaken and warmed to 69 room temperature. Using a calibrated 3.3 ml gas bulb, 0.045 mmol of SiC14 is vacuum transferred into the tube (LN2); the contents of the tube are warmed to room temperature and are agitated for 90 minutes, at which time 1H NMR spectroscopy indicates that all of the starting material has been consumed. Preparation of (SiMe3 ) 2 (MeAd)C5H 3 (41): In the glove box, a J-Young NMR tube is charged with 18.9 mg (0.049 mmol) of 40-DME. THF-d 8 (0.5 ml) is transferred into the tube, and the solution warmed to room temperature. The contents of the NMR tube are then cooled with a liquid nitrogen bath, and Me3 SiCl (0.062 mmol) is vacuum transferred from CaH2 into a calibrated 42.5 ml gas bulb, and then into the cooled NMR tube. The contents of the tube are warmed quickly to room temperature, and the resulting clear, pale yellow solution is agitated overnight. After 11.5 hours of stirring, 1 H NMR spectroscopy indicates that all of the starting material had been consumed. Preparation of [1,2-(SiMe3 ) 2 -4-MeAd-C5H 2 ]Li (42): II-151 A J-Young tube containing 0.049 mmol of MeAd(TMS) 2 CpH and 0.5 ml of THF-d 8 is brought into a glove box, and 9 mg (0.083 mmol) of CH3Li-LiBr is added to the clear, very pale yellow solution. After 4 hours of stirring at room temperature, 1H NMR spectroscopy indicated that all of the starting material had been consumed. The presence of a single vinylic cyclopentadiene resonance (relative intensity 2H), one singlet for the trimethylsilyl groups (18H), and one MeAd resonance (singlet, 3H), are diagnostic for the assigned 1,2-(TMS)i4-MeAd substitution about the Cp ring. Preparation of Me2Si(MeAd-C5H4)2 (H2Adp, 43): A 2-liter, 3-necked roundbottom flask, equipped with a 180° needle valve and a rubber septum, is charged with 80.82 grams (237.7 mmol) of 39-TMEDA. THF (400 ml) is added via cannula, and the contents of the flask are cooled to -78 °C. Dichlorodimethylsilane (13.5 ml, 111.1 mmol, 0.46 eq.) is vacuum transferred from CaH2 into an evacuated 50 ml roundbottom flask / 180° needle valve assembly. After filling the small flask with argon, the M~SiCh is syringed slowly into the 2-liter flask containing the white THF suspension. The THF solution is slowly warmed to room temperature, and is stirred at room temperature for an additional 36 hours. THF is removed in vacuo, leaving a viscous yellow oil, which is brought into a dry box. The petroleum ether 70 insoluble white solid (LiCl-TMEDA) is removed by suction filtration, and the bright yellow petroleum ether solution is transferred to a 1 L roundbottom flask with a 180° needle valve. Solvent is removed in vacuo, leaving a large quantity of viscous orange oil. MeAd-C5H 5 , a side product from the linking reaction, is removed by Kugelrohr distillation (15 hours, 100 °C, <1 µ), affording 88 mmol of 43 (79% yield) as a dark orange / brown glassy solid upon cooling to room temperature. Elemental analysis: Calculated (Found, with V 2Os added) C: 84.21 (83.15, 82.97); H: 10.00 (10.30, 9.75); N: 0.00 (0.00, 0.00); (Found, without added oxidant) C: (83.22, 83.12); H: (8.73, 8.58); N: (0.00, 0.00). Preparation of [Me2Si(3-MeAd-CsH3)i]Li2 (Li2Adp, 44) : A 500 ml roundbottom flask equipped with a 180° needle valve is charged in a dry box with 18.0 mmol of 43. Petroleum ether (160 ml) is vacuum transferred onto the orange solid and the solution is cooled to 0 C. A 1.6 M solution of nbutyllithium in hexanes (24 ml, 38.4 mmol, 2.1 eq.) is syringed into the 0 roundbottom against an argon counterflow over a period of five minutes. The solution is warmed slowly to room temperature and is then stirred at 25 °C for 16 hours. Volatiles are removed from the stispension of white solid in a faint yellow solution and residual solvent is removed from the beige solid by an additional 2 hours of drying under dynamic vacuum. The flask is brought into a dry box and attached to a fine porosity swivel frit and a 500 ml roundbottom flask . Petroleum ether (175 ml) is vacuum transferred onto the solid and the solution is filtered after warming to room temperature. Solvent is removed and the beige petroleum ether insoluble material is dried in vacuo for 8 hours, yielding 8.20 grams of a pale peach-colored solid. 43 is recrystallized from Et2O, producing an initial crop of 4.76 g (9.58 mmol, 53.2% yield) of ligand; additional ligand may probably be recovered from the remaining 1.95 grams of Et2O soluble material through additional recrystallizations. Elemental analysis: Calculated (Found, with V2Os added) C: 82.20 (79.91, 78.31); H: 9.35 (10.00. 9.54); N: 0.00 (0 .00, 0.00) . Preparation of Me2Si[(SiMe3 )(MeAd)-C5H3)i] (H2Abp, 45): A 1-liter roundbottom flask is charged with 22.0336 g (42.8 mmol) of 43 and 500 ml of THF. A 180° needle valve is attached to the flask and the solution is degassed on a vacuum line and then cooled to -78 °C. Trimethylchlorosilane (13.2 ml, 104 mmol, 2.43 equivalents) is vacuum transferred from CaH2 into a 71 graduated cold finger and then onto the Li2 Adp suspension. The solution is stirred at -78 °C for 3 hours, and is then allowed to warm to room temperature and stirred at 25 °C for 12 hours. Volatiles are removed in vacuo, leaving a large amount of yellow foam; the foam is dried overnight by exposure to dynamic vacuum, and then brought into a glove box. A bright yellow solution and white solid result from the addition of 350 ml of petroleum ether to the foam. The solid (LiCl) is removed by filtration and is washed twice with small aliquots of petroleum ether. Solvent is removed from the combined organics, leaving a large amount of yellow foam. The low boiling (<110 °C) species are removed by Kugelrohr distillation, leaving 20.17 g of a dark orange glassy solid that could be crushed into an orange solid (32.1 mmol, 75% yield). Elemental analysis: Calculated (Found, with V2Os added) C: 76.34 (75.49, 75.17); H: 10.27 (8.73, 9.44); N: 0.00 (0.00, 0.00). Calculated (Found, without added oxidant) C: 76.34 (75.50, 75.51); H: 10.27 (10.27, 10.52); N: 0.00 (0.00, 0.00). Use of the following purification procedure results in improved isolated yields of yttrocenes: 18.0 mmol of the once-Kugelrohred 45 is placed in a 500 ml roundbottom flask, to which is added 275 ml of THF. KOCMe3 (4.312 g, 38.4 mmol, 2.13 eq.) is added slowly to the clear yellow solution; white solid began precipitating from the solution within several minutes. Solvent is removed in vacuo after stirring at room temperature for 45 hours, and the resultant tan solid dried in vacuo. Diethyl ether (225 ml) is transferred onto the resulting solid, and the suspension is cooled to 0 C. A 0.5 M aqueous solution of NH4 Cl (110 ml) is syringed slowly into the suspension, and allowed to warm to room temperature. The orange organic layer is washed three times with 30 ml aliquots of water, and the combined aqueous layers are washed once with 30 ml of ether. The combined organics are dried over magnesium sulfate and filtered. Solvent is removed from the clear yellow solution, leaving an orange foam,. Removal of low boiling species is attained by a lengthy Kugelrohr distillation (15 hours of heating at 100-110 °C), leaving 7.52 g (12 mmol) of 45 (38% yield based on 44). 0 Li2 Abp (46) A 50 ml roundbottom flask is charged with 789 .9 mg 45 and 247.0 mg LiCH2 SiMe3 , and the flask is attached to a coarse frit assembly. THF (25 ml) is distilled onto the solids (-78 °C), and the solution is warmed to -40 °C. 72 After stirring at -40 °C for three hours, the solution is allowed to warm slowly to room temperature, and is stirred at 25 °C overnight. Volatiles are removed and the resulting foam dried in vacuo. Petroleum ether (25 ml) is distilled onto the foam, resulting in an almost completely clear orange solution, thus precluding removal of 46. Solvent is removed from the flask, and 936 mg of 1 tan solid is isolated. H NMR spectrosopy indicates one equivalent of residual THF accompanying the doubly deprotonated 46/LiCl mixture. K2 Abp (47): A 100 ml roundbottom flask is charged with 2.5001 g (3.97 mmol) of 45 and 1.6056 g (8.05 mmol, 2.03 eq.) of KOCMe3 . The flask is attached to a 180° needle valve and degassed on the vacuum line. Toluene (50 ml) is distilled onto the solids (-78 °C) and the solution allowed to warm to room temperature. After stirring at 25 °C for 32 hours, solvent is removed in vacuo, leaving a brown solid. The flask is attached to a medium porosity swivel frit and 40 ml petroleum ether are distilled onto the reaction mixture. A light brown petroleum ether insoluble material is isolated by filtration from the dark brown solution and the solid is washed twice with small aliquots of pet ether. The solid is dried by heating (9 hours, 130 °C) under dynamic vacuum (<1µ), affording 2.2815 g (3.22 mmol) of 47 as a pale tan solid (81 % yield). Elemental analysis: Calculated (Found, with V2Os added) C: 68.09 (67.33, 67.18); H: 8.88 (9.13, 8.59); N: 0.00 (0.00, 0.00). Calculated (Found, without added oxidant) C: 68.09 (65.64, 65.56); H: 8.88 (9.59, 9.48); N: 0.00 (0.00, 0.00). Preparation of meso-AdpScCl-LiCI-(THFh (48): In a 100 ml roundbottom flask attached to a medium porosity swivel frit is placed 2.09 g (11.79 mmol) of 47 and 1.54 g of ScC!J(THF)3_5. THF (50 ml) is vacuum transferred into the flask, and the resulting solution is allowed to warm to room temperature. After 45 minutes of stirring, all the reagents had fully dissolved, producing a clear orange solution. THF is removed after stirring at room temperature for 12 hours, and the resulting tan foam is dried in vacuo. In a glove box, the foam is crushed into a finely divided solid and the frit assembly is brought back to the vacuum line, where 50 ml of petroleum ether is transferred onto the solid and allowed to warm to 25 °C. The finely divided tan and white petroleum ether-insoluble species are removed by filtration and are washed three times with small portions of PE. Removal of solvent from the 73 petroleum ether solution provided 374 mg (0.52 mmol) of AdpScCl-LiCl(THF)2. Additional AdpScCl-LiCl-(THF)2 comprised much of the 1.86 g of white petroleum ether insoluble solid (as evidenced by 1 H NMR); the exact quantity of PE insoluble AdpScCl is not determined. As subsequent reactions with allylmagnesium bromide prove tolerant of LiCl, these alkylations are often run using the crude 48 product mixture. Preparation of meso-AdpYCl-LiCl-(solvh (49): A 500 ml roundbottom flask, equipped with a magnetic stir bar and a 180° needle valve, is charged with 11.1465 grams of Li2Adp (22.44 mmol) and 9.2344 grams of YCl3(THF)3 (22.43 mmol, 1.0 eq.). The flask is evacuated and 250 ml THF is vacuum transferred onto the solids at -78°C. The solution is stirred and allowed to warm to room temperature. After 4 days, 150 ml of solvent is removed in vacuo. The remaining THF is removed in vacuo after an additional 12 hours of stirring at room temperature, leaving approximately 19.65 grams of tan solid. From that 19.7 grams of solid, 13.73 grams are placed in a 500 ml roundbottom flask, which is attached to a fine porosity frit and another 500 ml flask. Diethyl ether (150 ml) is vacuum transferred onto the solid, warmed to room temperature with stirring, and then removed in vacuo. This procedure is repeated with one 150 ml aliquot of toluene, and then with two aliquots of Et20. Another 150 ml of diethyl ether is vacuum transferred onto the solid, and the Et20 insoluble material is removed by filtration, and is washed twice with small portions of Et20. Solvent is removed from the dark yellow ether solution, and the resulting tan foam is washed with an additional 75 ml Et20, which is removed in vacuo. After drying the resulting foam for 5 hours under dynamic vacuum, 7.19 grams of a tan solid is isolated. NMR spectroscopy indicates the presence of 0.16 equivalents of toluene, 0.37 equivalents of THF, and 1.74 equivalents of Et20, providing a yield of 8.76 mmol of Adp YCl-LiCl (56%, based on Y) . Preparation of rac-AbpScCl (SO, 51, 52): A J-Young tube is charged with 15.7 mg (0.022 mmol) of 49 and 8.4 mg (0.022 mmol, 1.0 eq.) ScC13 (THF)3 and is degassed on the vacuum line. THF-d 8 (0.4 ml) is vacuum transferred onto the solids and the tube is warmed to room temperature, resulting in a suspension of white solid. The presence of 50 in solution is evident after 20 hours at room temperature (1H NMR) . The cloudy white suspension is then placed in 74 an 85 °C bath, and NMR spectra taken periodically. Resonances attributed to 51 and 52 appear as those of 50. Total conversion of 50 to 51 and 52 is observed after 24 hours at 85 °C. Preparation of rac-AbpYCI (53 and 54): A 500 ml roundbottom flask, equipped with a water cooled condensor and a 180° needle valve, is charged with 8.529 g (12.05 mmol) of 47 and 4.95 g (12.0 mmol) of YC13 (THFh and is evacuated on the vacuum line. THF (230 ml) is distilled onto the solids and the solution is heated to reflux. After 60 hours of refluxing, the solution is cooled to room temperature and solvent is removed in vacuo. Benzene (20 ml) is vacuum transferred onto the solid, the solution cooled to -78 °C, and benzene removed in vacuo. The apparatus is brought into a glove box, and the flask attached to a medium porosity swivel frit assembly. The petroleum ether soluble species are isolated by extraction into 200 ml of pet ether; 15% of the petroleum ether soluble species are set aside for other experiments. The diethyl ether insoluble portion of the remaining PE soluble species are washed with diethyl ether and isolated by a cold (-78 °C) filtration from 25 ml of Et2O, leaving a tan solid which is dried in vacuo . Benzene (100 ml) is distilled onto the petroleum ether soluble and diethyl ether insoluble material, warmed to room temperature, and the volume of the solution reduced in vacuo to ~40 ml. The clear burgundy-colored solution is cooled to -10 °C with an ice/NaCl/water bath, and benzene is removed in vacuo. After 5 hours, the flask is heated to 55 °C under exposure to dynamic vacuum, affording 3.3325 g of pale tan solid. 1 H NMR spectroscopy indicates 0.22 equivalents of residual THF (per Y), 0.07 equivalents of residual C6H 6 , and a 1.5:1 ratio of 53:54 (4.31 mmol of Y, 42% yield) . Elemental analysis does little to confirm the assignment of 53 and 54 in the solid state. More than one permutation of rac-Abp YCl-(KCl)-(THFh, racAbpYCl(THF), and [rac-AbpY(µ-Cl)h (in quantities consistent with the observed 1 H NMR [Abp] ligand intensities) provide virtually identical %C and %H values (see below). Furthermore, these permuations of rac-Abp YCl(KCl)-(THFh, rac-AbpYCl(THF), and [rac-AbpY(µ-Cl)h are inconsistent with the amount of residual THF observed by 1H NMR; inclusion of the observed quantity of THF results in expected %C and %H values that equally poorly fit the observed data. 75 %C %H Observed (with added V?O~) 61.96, 61.80 8.14, 8.11 Observed (without added oxidant) 61.65, 61.94 8.36, 8.40 mixture of homo- and heterochiral [AbpY(µ-Cl)] ? 64.71 8.54 1.5 [AbpY(µ-Cl)] ? + 1 AbpYCl(THF) 64.18 8.39 1.5 [AbpY(µ-Cl)] ? + 1 AbpYCl-(KCl)-(THF)? 61.81 8.70 0.5 [AbpY(µ-Cl)] 7 + 3 AbpYCl(THF) 64.28 8.52 0.5 [Abp Y(µ-Cl)] 7 + 3 Abp YCl-(KCl)-(THF)? 57.08 7.41 3 Abp YCl(THF) + 1 Abp YC1-(KC1)-(THF) 7 62.10 8.33 1 Abp YCl(THF) + 3 Abp YC1-(KC1)-(THF)7 58.32 7.90 stoichiometry of [AbpYCl](THF)n?? 64.13 8.34 stoichiometry of [Abp YCl] (KCl)n 11 (THF)n ?? 63.05 8.24 References and Noles 1. Brintzinger, H. H.; Fischer, D.; Mulhaupt, R.; Rieger, B.; Waymouth, R. N. Angew. Chem. Int. Ed. Eng. 1995, 34, 1143. 2. Willoughby, C. A.; Buchwald, S. L. J. Am. Chem. Soc. 1992, 114, 7562. 3. Willoughby, C. A.; Buchwald, S. L. J. Org. Chem. 1993, 58, 7627-7629. 4. Lee, N. E.; Buchwald, S. L. J. Am. Chem. Soc. 1994, 116, 5985. 5. Broene, R. D.; Buchwald, S. L. J. Am. Chem. Soc. 1993, 115, 12569-12570. 6. Carter, M. B.; Scihott, B.; Gutirrez, A.; Buchwald, S. L. J. Am. Chem. Soc. 1994, 116, 11667. 7. Erker, G.; Nolte, R.; Aul, R.; Wilker, S.; Kruger, C.; Noe, R. J. Am. Chem. Soc 1991, 113, 7594. 8. Erker, G.; Aulbach, M.; Knickmeier, M.; Wingbermuhle, D.; Kriiger, C.; Nolte, M.; Werner, S. J. Am. Chem. Soc. 1993, 115, 4590. 76 9. See also a) Erker, G; Temme, B. J. Am. Chem. Soc., 1992, 114, 4004. b) Chen, Z.; Eriks, K.; Halterman, R. L. Organometallics, 1991, 10, 3449. 10. Coates, G. W.; Waymouth, R. M. Science 1995, 267,217. 11. Hauptman, E.; Waymouth, R. M.; Ziller, J. W. J. Am. Chem. Soc. 1995, 117, 11586. 12. The latin prefix "ansa", meaning "handle", has been used by Brintzinger to describe such a tethered arrangement of cyclopentadienyl rings. Smith, J. A.; Von Seyerl, J.; Huttner, G.; Brintzinger, H. H.J. Organomet. Chem., 1979, 173, 175. 13. Giardello, M. A.; Conticello, V. P.; Brard, L.; Gagne, M. R.; Marks, T. J. J. Am. Chem. Soc. 1994, 116, 10241. 14. Giardello, M. A.; Conticello, V. P.; Brard, L.; Sabat, M.; Rheingold, A. L.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1994, 116, 10212. 15. Giardello, M. A.; Eisen, M. S.; Stern, C.; Marks, T. J. J. Am. Chem. Soc. 1993, 115, 3326-7. 16. Mengele, W.; Diebold, J.; Troll, C.; Roll, W.; Brintzinger, H. H. Organometallics 1993, 12, 1931. 17. Grossman, R. B.; Tsai, J. C.; Davis, W. M.; Gutierrez, A.; Buchwald, S. L. Organometallics 1994, 13, 3892. 18. Herrmann, W. A.; Morawietz, M. J. A.; Priermeier, T. Angew. Int. Ed. Engl. 1994, 33, 1946. 19. Burger, P.; Diebold, J.; Gutmann, S.; Hund, H . U.; Brintzinger, H. H. Organometallics 1992, 11, 1319. 20. Piemontesi, F.; Camurati, I.; Resconi, L.; Balboni, D.; Sironi, A.; Moret, M.; Ziegler, R.; Piccolrovazzi, N. Organometallics 1995, 14, 1256. 21. For alternate means of synthesizing metallocenes see a) Diamond, G. M.; Rodewald, S.; Organometallics, 1995, 14, 5. b) Jordan, R. F. Diamond, G. M.; Jordan, R. F.; Petersen, J. L. Organometallics, 1996, 15, 4030. c) Jordan, R. F. Diamond, G. M.; Jordan, R. F.; Petersen, J. L. Organom etallics, 1996, 15, 403045. 22. Collins, S.; Hong, Y.; Taylor, N. J. Organom etallics 1990, 1990, 2695. Chem. 77 23. Collins, S.; Hong, Y.; Ramachandran, R.; Taylor, N. J. Organometallics 1991, 10, 2349. 24. Gutmann, S.; Burger, P.; Hund, H. U.; Hofmann, J.; Brintzinger, H. H.J. Organomet. Chem. 1989, 369, 343. 25. Wiesenfeldt, H.; Reinmuth, A.; Barsties, E.; Evertz, K.; Brintzinger, H. H.J. Organomet. Chem. 1989, 369, 359-370. 26. Chacon, S. T.; Coughlin, E. B.; Henling, L. M.; Bercaw, J. E. J. Organomet. Chem. 1995, 497, 161. 27. Jutzi, P. Chem. 28. Bunel, E. E., Ph.D. Thesis, California Institute of Technology, 1989. 29. Bunel, E. E.; Bercaw, J. E. Unpublished 30. Mitchell, J. P.; Bercaw, J. E. , unpublished results. 31. Coughlin, E. B.; Bercaw, J. E. J. Am. Chem. Soc. 1992, 114, 7606. 32. Coughlin, E. B., Ph.D. Thesis, California Institute of Technology, 1994. 33. Miya, S.; Mise, T.; Yamazaki, H., in Catalytic Olefin Polymerization; Tominaga, K. and Soga, K., Eds.; Elsevier: Amsterdam, 1990, p. 531. 34. Kaminsky, W.; Kiilper, K.; Brintzinger, H . H.; Wild, F. R. W. P. Angew. Chem. Int. Ed. Engl. 1985, 24, 507. 35. Spaleck, W.; Antberg, M.; Rohrmann, J.; Winter, A.; Bachmann, B.; Kiprof, P.; Behm, J.; Herrmann, W. A. Angew. Chem. Int. Ed. Engl. 1992, 31, 1347. 36. Alternate syntheses of these complexes have been reported in which no rac:m es o ratios are reported. Stehling, U.; Diebold, J.; Kirsten, R.; Roll, W.; Brintzinger, H. H.; Jungling, S.; Miilhaupt, R.; Langhauser, F. Organometallics, 1994, 13,964. 37. Spaleck, W.; Kiiber, F.; Winter, A.; Rohrmann, J.; Bachmann, B.; Antberg, M.; Dolle, V.; Paulus, E. F. Organometallics 1994, 13, 954. 38. Rheingold, A. L.; Robinson, N. P.; Whelan, J.; Bosnich, B. Organometallics 1991, 11, 1869. Rev. 1986, 86,983. results . 78 39. Bandy, J. A.; Green, M. L. H.; Gardiner, I. M.; Prout, K. J. Chem. Soc. Dalton Trans. 1991, 11, 2207. 40. Burk, M. J.; Coletti, S. L.; Halterman, R. L. Organometallics 1991, 10, 2998. 41. Hollis, T. K.; Rheingold, A. L.; Robinson, N. P.; Whelan, J.; Bosnich, B. Organometallics 1992, 11, 2812. 42. Huttenloch, M. E.; Diebold, J.; Rief, U.; Brintzinger, H. H.; Gilbert, A.; Katz, T. J. Organometallics 1992, 11, 3600-3607. 43. Chen, Z.; Halterman, R. L. J. Am. Chem. Soc. 1992, 114, 2276. 44. Mitchell, J. P.; Hajela, S.; Brookhart, S. K.; Hardcastle, K. I.; Henling, L. M.; Bercaw, J. E. J. Am. Chem. Soc. 1996, 118, 1045. 45. Stone, K. J.; Little, R. D. J. Org. Chem. 1984, 49, 1849. 46. Fendrick, C. M.; Schertz, L. D.; Day, V. W.; Marks, T. J. Orgn nometallics 1988, 7, 1828. 47. Molle, G.; Bauer, P. J. Am. Chem. Soc 1982, 104, 3481. 48. Molle, G.; Bauer, P.; Dubois, J. E. J. Org. Chem 1983, 48, 2975. 49. Edelmann, F. T., in Comprehensive Orgnnometnllic Chemistry II; Abel, E. W., Stone, F. G. A., Wilkinson, G. and Lappert, M. F., Eds.; Pergamon Press: Tarrytown, NY, 1995; Vol . 4, p. 11. 50. Diamond, G. M.; Jordan, R. J.; Petersen, J. L. Organom.etallics 1996, 15, 4045. 51. Although presumably equimolar quantities of the (R) and (S) enantiomers of the chiral rac-AbpYMCl metallocenes are formed during metalation, for clarity only the (S) metallocenes are drawn explicitly. 52. The stability of KCl-(solventh adducts of group III ansa- metallocenes has been observed previously. Zubris, D. L.; Bercaw, J. E. Unpublished results. 53. Burger, B. J.; Bercaw, J.E. New Developments in the Synthesis, Manipulation, and Characterization of Organometallic Compounds; ACS Symposium Series, 1987 357, 79 54. Marvich, R. H .; Brintzinger, H . H.J. Am. Chem. Soc. 1971, 93, 2046. 55. Manzer, L. E. Inorg. 56. Cowley, A. H.; Kemp, R. A. Synth. React. Inorg. Metal-Org. Chem. 1981, 11, 591. Synth. 1982, 21, 135. Appendix A: 1 H and 13 C NMR Data For 19-54 Table 1. Proton NMR assignments for 19-54 (500 MHz, C6D 6 , 25 °C unless otherwise noted) 6 (ppm) Compound Assignments J H-H Pen tameth y leneful vene Cy (2H) (19) (4H) (300 MHz) (4H) Cp (4H) SiMe3 (9H) (SiMe3 )(Cy)C5H 4 (300 MHz) SiMe3 (~2 isomers) Cy (10H) (lH) Cp, allylic Cp, allylic (lH) Cp, vinylic (3H) Cp, vinylic [1-SiMe3 -3-Cy-C5H 3 ]K (24) SiMe3 (9H) Cy (llH) (THF-d 8 , 300 MHz) [Cy-C5H 4 ]Li (20) (THF-d 8 , 300 MHz) (lH) Cp (3H) Cy (7H) (3H) (lH) (lH) Cp (4H) 1.28 (m) 1.41 (m) 2.35 (m) 6.59 (m) -0.07 (s) 0.19 (d) 1.10-1.40 (overlapping) 1.60-1.75 (overlapping) 1.90-2.00 (overlapping) 2.38 (m) 2.88, 2.94 (m) 3.20 (s) 6.11 (s), 6.42 (s), 6.63 (m) 6.19 (m), 6.24 (m), 6.79 (m), 6.88 (m) 0.27 (s) 1.27, 1.49, 1.79, 1.93 (overlapping) 2.44 (m) 5.65, 5.72, 5.80 (m) 1.20-1.40 (overlapping) 1.60-1.80 (overlapping) 1.90 (d) 2.39 (m) 5.45 (s) 80 Table 1. (continued) Compound Assignments 8 (ppm) JHH [Cy-C5H 4 ]K (THF-d 8 , 300 MHz) Me2 Si(Cy-C5H 4)i (22) (2': 2 isomers) [Me2 Si(3-Cy-C5H 3 ) 2 ]Li2 (26a) (THF-d 8 , 300 MHz) [Me2 Si(3-Cy-C5 H 3) 2 ]K2 (26b) (THF-d 8 ) 1.21, 1.26, 1.31, 1.35, 1.39, 1.42, 1.67, 1.74 I 1.83, 1.90 (m) 2.37 (t oft) (lH) 11.8, 3.7 Cp (2H) 5.32 (m) (2H) 5.35 (m) Me2Si (3H) -0.18 (s) MeiSi (2H) 0.13 (s), 0.15 (s) Cy (lOH) 1.10 - 1.40 (m) (6H) 1.60 - 1.85 (m) (3H) 1.95 (m), 1.97 (m) (2H) 2.27 (t oft), 2.37 (t oft) Cp, allylic (0.5H) 2.92 (s), 2.97 (m), 3.44 (d) Cp, allylic (0.5H) 3.36 (s) Cp, vinylic (4H) 6.14, 6.47, 6.62 (m) Cp, vinylic (lH) 6.19, 6.21, 6.80, 6.91 (m) 0.46 (s) MeiSi (6H) Cy (20H) 1.16, 1.20, 1.22, 1.28, 1.34, 1.63, 1.68, 1.80 (m) (lH) 2.32 (t oft) 12.0, 3.2 5.50 (m) Cp (lH) 2.6 (lH) 5.58 (m) 2.6 5.65 (m) (lH) 2.6 0.85 (s) MeiSi (6H) 1.15 - 1.50 (m) Cy (12H) 1.70 - 1.80 (m) (4H) 1.87 - 1.95 (m) (4H) 2.43 (m) (2H) 5.72, 5.75, 5.78 (m) Cp (6H) Cy (lOH) 81 Table 1. (continued) Compound Assignments Me2Si[ (SiMe3 )(Cy)C5H 3h Me3Si (18H) (25) (>2 isomers) MeiSi (6H) Cy (14H) (7H) (2H) Cp, allylic Cp, vinylic (4H) {MeiSi[Cy(SiMe3)Cp]2)K Me3 Si (18H) Me2Si (6H) 2 (major isomer, THFCy (22H) d 8 I 18-crown-6) Cp (4H) Adamantylfulvene (30) Ad (2H) (2H) (8H) (2H) Cp (4H) [(2-Ad)-C5H 4 ]Li (31a) Ad (l0H) (THF-d 8 ) (2H) (lH) Cp (4H) (degenerate) Ad (2H) [(2-Ad)-C5H 4 ]Na (31b) (3H) (THF-d 8) (7H) (2H) (lH) Cp (4H) (degenerate) o (ppm) -0.04, -0.03, -0.02, 0.02, 0.03 (s) 0.10, 0.14, 0.18, 0.22, 0.27, 0.28, 0.32 (s) 1.15 - 1.42 (m) 1.60 - 1.80 (m) 1.86 - 2.20 (m) 2.95, 3.03, 3.17, 3.20 (m) 6.15, 6.23, 6.29, 6.36, 6.44, 6.50, 6.54, 6.59, 6.69, 6.75 (m) 0.04 (s) 0.33 (s) 1.20 - 1.35, 1.60 - 1.75, 2.47 (m) 5.79, 5.80 (m) 1.65 (t) 1.72 (quartet) 1.75 (t) 3.11 (s) 6.62, 6.65 (m) 1.42, 1.67, 1.75, 1.89, 1.91, 1.93, 1.94, 2.20 (s) 2.21 (d) 2.96 (s) 5.58 (s) 1.39 (d) 1.66, 1.75 (s) 1.87, 1.89, 1.91, 1.92, 2.21 (m) 2.28 (d) 3.02 (s) 5.59 (s) J H-H 2.8 2.0 2.6 11.5 11.8 11.7 11.9 82 Table 1. (continued) Compound (SiMe3 )(2-Ad)C5H 4 (32) (2 isomers) (SiMe2 Cl)(2-Ad)C5H 4 (33) (2 isomers) Ass ignments o (ppm) SiMe3 SiMe3 -0.05 (s) 1.54, 1.56, 1.71, 1.74, 1.81, 1.83, 1.88, 1.90, 2.10, 2.24, 2.29 (m) Cp, allylic 2.62 (s), 2.66 (s), 2.82 (s), 3.00 (s) Cp, allylic Cp, vinylic Cp, vinylic 2.86 (s), 3.25 (s) 6.22 (s), 6.49 (s), 6.66 (s) (3H) 6.26 (s), 6.28 (s), 6.83 (m), 6.91 (m) 0.12 (s) 0.15 (s) SiMe2Cl 0.43 (s) 0.44 (s) Ad (2H) (14H) 1.47 (m) 1.71, 1.77, 1.79, 1.85, 1.87, 1.91 (in) 2.17 (m) (3H) (lH) Cp, allylic Cp, allylic (lH) Cp, vinylic (lH) Cp, vinylic Cp, vinylic (lH) Cp, vinylic (lH) Cp, vinylic [1-SiMe3 -3-(2-Ad)C5H 3 ]Li (34) (THF-d 8 ) 0.21, 0.19 (s) Ad SiMe2Cl (3H) SiMe3 (9H) Ad (2H) (14H) Cp (lH) (lH) (lH) J H-H 2.77 (s) 2.89 (s) 3.03 (m) 3.42 (s) 6.12 (s) 6.15, 6.21 (m) 6.42 (m) 6.57 (m) 6.83 (m) 6.92 (m) 0.10 (s) 1.43 (d) 1.68, 1.76, 1.88, 1.90, 1.91, 2.19, 2.21, 2.96 (s) 5.76 (m) 5.79 (m) 5.82 (m) 11.8 2.5 2.1 2.6 83 Table 1. (continued) Compound Assignments 8 (ppm) (SiMe2 Cl)(SiMe3 )(2Ad)C5H 3 (35) (THF-d 8 ) Me3Si (9H) MeiSi (6H) Ad 0.04 (s) 0.22, 0.26 (s) 1.42, 1.56, 1.68, 1.56, 1.68, 1.72, 1.75, 1.78, 1.88, 1.90, 1.92, 1.95, 1.97 (m) 2.21 (s) 2.83 (s) 6.28, 6.57, 6.79 (m) (SiC13 )(SiMe3 )(2Ad)C5H 3 (36) Cl2 Si[ (SiMe3 )(2Ad)C5H 3 lz (37) :::::2 isomers MeiSi[ (SiMe3 )(2Ad)C5H 3lz (28) :::::2 isomers Cp, allylic (lH) Cp, vinylic (3H) SiMe3 (9H) Ad (2H) (SH) (3H) (lH) (lH) (2H) Cp (3H) Me3Si (SH) (lOH) Ad (SH) (44H) (lH) (lH) Cp (3H) Me3Si (18H) Me2Si (16H) Ad (2H) (15H) (SH) Cp, allylic (2H) Cp, vinylic (3H) 0.05 (s) 1.47, 1.49, 1.51 (s) 1.69 (s), 1.72 (m) 1.75, 1.80, 1.83 (m) 1.93 (d) 1.99 (d) 2.14, 2.72 (s) 6.28, 6.41, 6.72 (m) -0.06 (d) . 0.17 (s) 1.47, 1.49, 1.51, 1.53 (m) 1.73, 1.79, 1.82, 1.87, 1.89, 1.98, 2.01 , 2.05, 2.22 (m) 2.63 (s) 2.78 (d) 6.52, 6.70, 6.76 (m) -0.06, -0.02, 0.00, 0.02, 0.04, 0.05 (s) 0.14, 0.17, 0.36, 0.38 (s) 1.52 (d), 1.62 (d), 1.66 (d) 1.77, 1.88, 1.92 (m) 2.24 (d), 2.38 (m) 2.94 (s), 2.74 (s) 6.45, 6.53, 6.64, 6.77, 6.82 (m) JH-H 12.6 10.8 16.0 13.6 84 Table 1. (continued) Compound Assignments o (ppm) J HH [(±)C2J-l 120 2 ]Si[ (SiMe3)(2 Me3Si (18H) -Ad)C5H3h (38) MeCyCpLi (THF-d 8) [MeAd-C5H 4 ]Li-DME (39-DME) (THF-d 8 ) Ad (2H) (SH) (lH) (18H) (3H) Cp, allylic (2H) Cp, vinylic (2H) (overlaps with Bn) Bn (4H) (2H) (SH) MeCy (3H) MeCy (6H) (2H) (2H) Cp (2H) (2H) MeAd (3H) MeAd (2H) (SH) -0.42, -0.27, -0.21, -0.04, 0.02, 0.08, 0.10, 0.12, 0.14, 0.17, 0.19, 0.21, 0.23 0.83 - 1.04 (overlapping) 1.21 - 1.49 1.50 - 1.70 1.70 - 2.10 2.10 - 2.50 2.50 - 3.45 6.21 - 6.70 6.70 - 7.04 7.05 - 7.14 7.18-7.78 1.14 (s) 1.42 - 1.49 (m) 1.52 - 1.55 (m) 1.83 m 5.50 (m) 5.53 (m) 1.27 (s) 1.43 (m) 1.59, 1.66, 1.68, 1.84, 2.08 (m) 2.24 (m) (2H) 2.39 (m) (2H) CH30CH2CH20CH3 3.27 (s) CH30CH2CH20CH3 3.43 (s) 5.55, 5.56 (m) Cp (4H) 2.7 2.7 11.0 10.4 10.0 85 Table 1. (continued) Compound Assignments 8 (ppm) [MeAd-C5H 4 ]Li-TMEDA (39-TMEDA) MeAd (3H) MeAd (2H) (12H) 1.30 (s) 1.42 (d) 1.59, 1.65, 1.69, 1.83, 2.07, 2.22 (s) 2.39 (d) 2.13 (s) (SiMe3 )(MeAd)C5H 4 [1-SiMe3 -3-MeAdC5H 3 ]Li-DME (40-DME) (2H) Me2NCH2CH2N Me2 (12H) M~N CH2CH2NM~ (4H) Cp (4H) Me3Si (9H) MeAd (3H) MeAd (2H) (2H) (SH) (lH) (2H) (lH) (lH) Cp, allylic (lH) Cp, vinylic (3H) Me3Si (9H) MeAd (3H) MeAd (lH) (lH) (4H) (lH) (overlaps w /DME) (lH) (overlaps w /DME) CH3OCH2CH20CH3 CH30CH2CH2OCH3 Cp (3H) J H-H 5.9 5.7 2.26 (s) 5.56, 5.57 (m) -0.05 (s) 1.24 (s) 1.58 (d) 1.64 (d) 1.70, 1.78, 1.85 (s) 2.04 (d) 2.15 (d) 2.20 (d) 2.26 (d) 3.21 (s) 6.12, 6.48, 6.65 (m) 0.46 (s) 1.44 (s) 1.65 (d) 1.70 (d) 1.85, 1.84, 1.82 (m) 2.00 (s) 2.35, 2.41 (m) 2.57 (d) 2.72 (m) 2.39 (s) 2.76 (s) 6.16, 6.19, 6.35 (m) 12.4 12.8 11.1 12.8 12.3 12.0 12.0 10.3 11.9 11.8 86 Table 1. (continued) Com__E_ound Assignments (SiMe2 Cl)(SiMe3)(MeAd Me3Si (9H) M~Si (6H) )CsH3 (THF-d 8) MeAd (3H) MeAd (lOH) (SiC13)(SiMe3)(MeAd)C5 H3 (THF-d 8 ) (SiMe3h(MeAd)C5H 3 (41) (THF-d 8 , 300 MHz) [1,2-(SiMe3h-4-MeAdC5H 2 ]Li (42) (THF-d8) Cp (3H) Me3Si (9H) MeAd (3H) MeAd (2H) (6H) (lH) (3H) (2H) Cp (3H) Me3Si (18H) MeAd (3H) MeAd (2H) (lH) (4H) (2H) Cp (3H) Me3Si (18H) MeAd (3H) MeAd (2H) (2H) (2H) (SH) (2H) (2H) Cp (2H) 8 (ppm) 0.04 (s) 0.23, 0.27 (s) 1.20 (s) 1.56 (s), 1.86 (s), 2.06 (m), 2.08 (m), 2.23 (m) 6.23, 6.57, 6.79 (m) 0.13 (s) 1.21 (s) 1.57 (d) 1.72 (br), 1.87 (s) 1.99 (d) 2.11 (s) 2.22 (d) 6.25, 6.61, 6.97 (m) -0.03 (s) 1.19 (s) 1.55 (d) 1.68 (m) 1.86 (s) 2.10 (m) 2.22 (d) 6.17, 6.49, 6.67 (m) 0.17 (s) 1.26 (s) 1.42 (d) 1.59 (s) 1.65 (d) 1.68, 1.82, 2.09 (s) 2.22 (d) 2.27 (d) 6.11 (s) J H-H 11.3 13.0 10.4 9.8 10.5 11.5 17.3 11.0 11.9 87 Table 1. (continued) Compound Assignments 8 (ppm) JH H M e2 Si (MeAd-C5H4)2 (H2Adp) (43) M~Si (SH) MeAd (6H) MeAd (43H) Li2Adp (44) (THF-d 8) Me2Si[ (SiMe3 )(MeA d)C5H3)2] (H2Abp) (45) (~2 isomers) Cp, allylic (2H) Cp, vinylic (6H) Me2Si (6H) MeAd (6H) MeAd (28H) Cp (2H) (2H) (2H) Me3Si (18H) Me2Si (6H) MeAd (9H) MeAd (23H) (6H) (23H) Cp, allylic (lH) Cp, vinylic (4H) -0.18, -0.14, -0.13, 0.14, 0.17 (s) 1.13, 1.20, 1.22, 1.24, 1.25 (s) 1.58, 1.60, 1.63, 1.65, 1.71, 1.78, 1.85, 2.06, 2.11, 2.14, 2.16, 2.20, 2.26, 2.27 (m) 3.01, 3.04, 3.31 (s) 6.18, 6.53, 6.66 0.24 (s) 1.27 (s) 1.41 (d) 5.9 1.58 (s) 1.66 (d) 6.4 1.69 (s), 1.83 (s), 2.10 (s) 2.24 (d) 5.6 2.41 (d) . 4.9 5.70 (t) 2.6 5.90 (t) 2.1 5.91 (t) 2.5 -0.10, -0.07, 0.37 (s) 0.17, 0.30, 0.37, 0.41 (s) 1.24, 1.26, 1.28, 1.29 (s) 1.65 - 1.72 (m) 1.86 (m) 2.08 - 2.36 (m) 3.57, 3.64 (s) 6.31, 6.66, 6.79, 7.10, 7.13 (m) 88 Table 1. (continued) Compound Assignments o (ppm) J HH Li2Abp (46) (THF-d 8) K2Abp (47) (THF-d 8 /18-Crown-6) mes o-AdpScCl-LiCl(THF)2 (48) (THF-d 8 ) Me3Si (18H) M~Si (6H) MeAd (6H) MeAd (9H) (2H) (4H) (SH) (2H) (2H) Cp (2H) (2H) Me3Si (18H) Me2Si (6H) MeAd (6H) MeAd (4H) (2H) (SH) (2H) (4H) (4H) (4H) Cp (4H) Me2Si (3H) (3H) MeAd (25H) MeAd (6H) Cp (2H) (2H) (2H) -0.17 0.41 1.27 1.60, 1.69 (m) 1.83 (m) 2.12 (m) 2.23 (m) 2.38 (m) 2.54 (m) 6.04 (m) 6.24 (m) 0.18 (s) 0.52 (s) 1.14 (s) 1.30 (m) 1.52 (m) 1.60 (d), 1.65 (s) 1.78 (m) 2.10 (s) 2.18 (d) 2.48 (d) 5.84, 6.14 (m) 0.41 (s) 0.55 (s) 1.42, 1.47, 1.49 1.53 t 1.63. 1.69, 1.72, 2.22, 2.25, 2.29, 2.32, 2.50 (m) 2.16 (s) 5.79 (t) 5.89 (t) 6.23 (t) 11.8 12.1 11.6 11.0 I 2.7 2.1 2.5 89 Table 1. (continued) Compound Assignments 8 (ppm) J HH mes o-AdpYCl-LiCl(solv)2 (49) rac-AbpScCl-(KCl)(THF)i (50) (THF-d 8 ) MeiSi (3H) (3H) MeAd (6H) MeAd (28H) Cp (2H) (2H) (2H) Me3Si (18H) Me2Si (6H) MeAd (6H) MeAd [rac-AbpSc(µ-Cl)h (51) (THF-d 8) rac-AbpScCl(THF) (52) (THF-d 8) Cp (4H) Me3 Si (18H) Me2 Si (6H) MeAd Cp (2H) (2H) Me3 Si (18H) Me2 Si (6H) MeAd (6H) MeAd [rac-Abp Y(µ-Cl)h (53) Cp (2H) (2H) Me3 Si (18H) Me2Si (6H) MeAd MeAd Cp (3H) 0.44 (s) 0.57 (s) 1.69 (s) 1.28, 1.43, 1.55, 1.59, 1.82, 2.15, 2.22, 2.24, 2.29, 2.49 (m) 5.84 (m) 5.92 (m) 6.20 (m) 0.26 (s) 0.56 (s) 1.18 (s) 1.19, 1.22, 1.28, 1.40, 1.56, 1.66, 2.11, 2.21 (m) 5.97 (s), 6.23 (s) 0.28 (s) 0.77 (s) 1.20 - 1.70 2.05 - 2.35 6.28 (d) 6.59 (d) 0.26 (s) 0.84 (s) 1.56 (s) overlap with those of 51 6.23 (d) 6.56 (d) 0.56 (s) 0.99 (s) 1.70 (s) 1.29, 1.46, 1.50 - 2.00, 2.20 - 2.25 (br, overlap) 6.65 (s), 6.87 (br) 2.7 2.0 2.5 1.6 1.7 1.0 1.7 90 Table 1. (continued) Compound Assignments 8 (ppm) [rac-AbpY(µ-Cl)h (53) (THF-d 8 ) Me3 Si (18H) Me2 Si (6H) 0.27 (s) 0.77 (s) 1.74 (s) 1.28 (d) 1.36 (d) 1.43 (d) 1.50, 1.52, 1.57, 1.61, 1.69 (overlapping) 1.73 (s) 2.11 (s) 2.19 (d) 2.25 (d) 2.48 (d) 6.28 (d) 6.54 (d) 0.29 (s) 0.84 (s) 1.74 (s) overlap with major product resonances 6.31 (d) 6.58 (d) MeAd MeAd Cp (4H) rac-Abp YCl(THF) (54) (THF-d 8 ) Me3 Si Me 2 Si MeAd MeAd Cp J H-H 9.8 10.2 12.0 12.1 12.5 12.5 2.0 2.0 2.0 1.9 91 13 Table 2. Proton decoupled C assignments for 53 (125 MHz, 25 °C) Cmpd solvent Assignments [rac-AbpY(µ-Cl)h (53) C6D6 8 (ppm) Me2 Si Me3 Si 3.34 MeAd 28.58, 28.77, 29.94, 34.44, 34.57, 1.13 34.65, 35.37, 37.15, 40.29, 40.91, 42.00 Cp 118.67, 120.44, 124.29, 130.70, 147.70 [rac-AbpY(µ-Cl)h (53) THF-d 8 Me2 Si Me3 Si MeAd 1.05 3.06 29.20, 29.32, 29.40, 34.67, 34.89, 34.97, 35.47, 35.77, 37.03, 40.96, 41.57 Cp 117.37, 119.53, 122.40, 131.17, 146.47 92 Chapter 3 Reactivity of C5 - and C2 -Symmetric Group III Metallocenes with Dihydrogen and with Unsaturated Organic Substrates Abstract 93 Introduction 94 Results and Discussion 106 Conclusions 127 Experimental 129 References and Notes 132 Appendices 136 93 Abstract The reaction of mesa- Me 2Si{3-[2-(2-CH3-C 10H 14 )]-C 5 H 3 }2MCl-(LiCl)(THFh (meso-AdpMCl-(LiCl)-(THFh) with allylmagnesium bromide afford meso-AdpM(ri3-C 3H 5) (M = Sc, Y) in good yield. The allyl ligands appear 11 3 coordinated in solution, but variable temperature NMR spectroscopy indicates fluzinal behavior associated with the allyl moieties. The results of hydrogenolysis of meso-AdpM(ri 3-C 3H 5 ) are consistent with a-bond metathesis to generate propene and highly reactive meso-AdpMH monomers. The use of methyladamantyl groups as metallocene ~substituents effectively prevents formation of bridging hydride ((µ-Hh) dimers; [meso-AdpM(µ-Hh] account for only ~3% of the overall organometallic hydrogenolysis products. The highly reactive [meso-AdpMH] monomers react readily with available C-H bonds in solution; intermolecular cyclopentadienyl C-H activation is believed to be the predominant [mesoAdpMH] decomposition mechanism. Treatment of [rac-Me 2Si/2-SiMe 3-4-[2(2-CH3-C 10H 14 )]-C 5H 2 bY(µ-Cl)h ([rac-AbpY(µ-Cl)h) with LiCH(SiMe 3 h affords the bulky yttrium-alkyl complex rac-Abp YCH(SiMe 3 h (18). Although ethylene is readily polymerized by 18, no reaction is observed between aolefins and 18. Hydrogenolysis of 18 results in the formation of 20, the homochiral dimer of [rac-AbpY(µ-H)h Neither the monomeric rac-AbpYH (19) nor the heterochiral [rac-Abp Y(µ-H)h dimer are observed in solution. Although variable temperature NMR studies indicate that dissociation of 20 into 19 is not rapid on the NMR timescale at 25 °C, evidence for the dissociation of 20 at room temperature is obtained from H / D scrambing experiments. Dimer 20 is an active a-olefin polymerization catalyst, but does not oligomerize or polymerize a-olefins. The allenyl complex 21 is formed from the C-H activation of allene by 19. Treatment of 20 with 2-butyne results in alkyne insertion to generate the yttrium-butenyl complex 22. Complex 22 2 readily undergoes ~-H elimination from an sp -hybridized carbon to generate a methylallene- complex, the C-H bonds of which rapidly undergo a-bond metathesis with 19 to generate methylallenyl complexes 25a,b . 94 Introduction The reactions of early transition metal metallocene-hydride and -alkyl complexes with unsaturated organic substrates have been studied extensively over the last 40 years.I Cationic group IV and the isoelectronic neutral group III 14-electron d 0 metallocenes have found extensive use as homogeneous aolefin polymerization catalysts. Similar catalysts may often be employed to promote the oligomerization or polymerization of alkynes.2-4 The ability of an olefin or alkyne to undergo insertion into or a-bond metathesis with a given bis(cyclopentadienyl)metal complex depends considerably on the degree of substitution and saturation of the substrate and on the steric influences of the metallocene. The often competing mechanisms of insertion chemistry and a-bond metathesis dominate the reactivities of allene and alkynes with metallocenebased Ziegler-Natta catalysts. Cationic group IV metaliocenes such as [Cp* ZrH(THF)][BPh and [Cp* ZrCH ][B(4-C H F) favor insertion of allene to generate r/-allyl and 17 -methallyl complexes (Figure 1).5,6 Additional ] 2 ] 4 2 3 6 4 5 3 insertions of allene into and decomposition of 2 are observed in the presence of excess allene; subsequent insertions into 1 are not observed, presumably due to the nonlability of the THF ligand . Although insertion of allene is observed for the monomeric Cp* 2Sc-H (3),7 allene does not insert into (1) (2) (2) (3) Figure 1. Insertion of allene to give 17 -allyl and 11 -methallyl complexes. 3 3 95 [Cp* 2 Y(µ-H)h; a-bond metathesis with one of the allene C-H bonds instead gives dihydrogen and the allenyl complex Cp* 2 YCH=C=CH 2 (Figure 2).8 .H Cp* 2Y-C ~ c "'c--H \ H Figure 2. a-bond metathesis of [Cp*2Y(µ-H)h with allene. Subtle differences exist between the reactivities of cationic group IV complexes and those of neutral group III and Ln complexes with terminal alkynes. Terminal acetylenes undergo 1,2-insertion into cationic group IVmethyl complexes to generate M-vinyl species.9 Although these vinyl complexes may be trapped in the presence of coordinating ligands,10 they are usually quite reactive and undergo rapid a-bond metathesis with C-H bonds of additional alkyne monomers to generate M-alkynyl complexes (Scheme 1).11,12 + == + H R Cp*z>Cp*2M-CH3 - - - - - - - - ~ R (1,2-insertion) H3C H--R l (cr-bond metathesis) j '-.....- H3 ~ R + Cp* 2M - - - - R Scheme 1. Generation of cationic group IV M-alkynyl complexes. Once generated, these metal-alkynyl species act as efficient catalysts for the dimerization of terminal olefins. The catalytic cycle, involving 1,2insertion to give a metal-alkenyl-yne, followed by a-bond metathesis to regenerate the alkynyl complex, is shown in Scheme 2 (A). 96 R /R I Cp•,M -- R (1,2-insertion) H A (cr-bond metathesis) R RI Cp*2 R R R H---R H === (cr-bond meta thesis) B R R (2,1-insertion) H---R R Cp*2 \>-----R R Scheme 2. Catalytic olefin oligomerization by Cp*2 M-alkynyl complexes (M=Sc, Y, Zr+). Rather than initially undergoing 1,2-insertion, terminal alkynes undergo a-bond metathesis with Cp*2 Sc-CH3 and Cp*2 YCH(SiMe3 )i to generate M-alkynyl complexes and one equivalent of alkene?,13 Like their group IV counterparts, though, these alkynyl complexes rapidly catalyze the dimerization of alkynes (A, Scheme 1). H ~ R Cp*2Sc-CH3 - - - - - - C p *2Scr-----R + CH4 (3) R + H2C(SiMe3)z (4) (5) H---R Cp*2Y-CH(SiMe3)z~------ Cp*2Y ~ Variations to the R substituent and to the ligand array have marked effects on the ability of the metal complexes to oligomerize acetylenes. Use of alkynes with R=-CH3 and -CH2 CH2 CH3 provide both alkyne dimers (A, Scheme 2) and trimers (B, Scheme 2), while acetylenes with R=-CMe3 , -Ph, and -4-C6H 4 Me are only dimerized. Although trimethylsilylacetylene undergoes multiple insertions, it fails to undergo a-bond metathesis to regenerate a catalytically active M-C=R complex.11 The less sterically 97 congested (C5H 5h framework permits formation of unusual bridged complexes upon additon of just one equivalent of the activator [PhNMe2 H][B(4-C6H 4 F\] to Cp2 Zr(CH3)i (Figure 3)) 2 neither the starting materials nor the organometallic products react with additional equivalents of alkyne. Cp 2Zr(CH3 )i + [PhN(CH3)iH][B(4-F-C6 H 4) 4 ] + H~~-R - C~, -PhN(CH3)i (several steps) R )l0 Cp2ZKrCr.,[BPh'4 ] H 3C R (R=-CMe3) Figure 3. The relatively uncongested (C5H 5 )i ligand array permits formation of bimetallic complexes inactive for alkyne oligomerization. The observed reactivity of internal alkynes towards Cp2 M-R (M=Sc, Y, Ln, Zr+) is highly subject to subtle variations to the metal, ligand, and substrate.14 The vast majority of cationic group IV-alkyl complexes insert 2butyne and various other internal alkynes, affording M-alkenyl species (Scheme 3). In the absence of additional coordinating ligands, the relatively unhindered metallocenes (R2 =R3 =CH3 or CH2 CH3 , Cp2 M ,;z= Cp\Zr,) often permit insertion of a second equivalent of alkyne to generate 115-pentadienyl complexes (Scheme 4).15 By contrast, the sterically congested alkenyl complex Cp\Zr+ -C(CH3 )=C(CH3h does not insert additional 2-butyne, instead activating a C-H bond of butyne by a-bond metathesis to give Cp*2 Zt-CH2C=C-CH3 and 2-methyl-2-butene.1 5 Cp*2 Zr+ -CH3 also fails to react cleanly with H 3 C-C=C-SiMe3 , presumably due to the steric bulk of the substrate.16 98 Scheme 3. Insertion of internal alkynes; Cp 2 = (C5H 5 )i, (CH3 Cp)i, (Me3 SiCp )2, (Me3 CCp)i, ethylenebis(indenyl), Cp*2 ; (R2,R3 ) = (CH3 , CH3 ), (CH2CH3 , CH2 CH3 ), (C6H 5 , C6H 5 ), (CH3 , Si(CH3h) .5,15,16 Scheme 4. Syntheses of 175 -pentadienyl complexes; R=CH3 or CH2 CH3 . Permethylyttrocene alkyl- and hydride- complexes show no insertion chemistry with internal alkynes. A C-H bond in 2-butyne undergoes a-bond metathesis with [Cp*2 Y(µ-H)h (4) to give dihydrogen and the propargyl complex Cp*2 YCH2C=CCH3,1 7 which is unreactive towards additional equivalents of alkyne. The hindered Cp*2 YCH(SiMe3)i is completely unreactive towards 2-butyne. Switching to larger lanthanides leads to a pronounced increase in reactivity. Both Cp\LaCH(SiMe3)i (6) and Cp\CeCH(SiMe3 ) 2 (7) undergo abond metathesis with and catalyze the dimerization of internal alkynes (Scheme 5).18 Bulky alkynes such as H 3 C-C=C-C(CH3) 3 and H 3 C-C=C-Si(CH3h are not dimerized by 6 or 7, although do react by a-methyl C-H bond activation. 6 and 7 do not react with 3-hexyne, even after several months at 80 °C. 99 ----=-=-=--R Scheme 5. Organolanthanide catalyzed dimerization of internal alkynes; Ln = La, Ce; R= Me, Et, n-Pr. Considerably more research has focused on the use of metallocene catalysts in olefin polymerization than has touched on the area of alkyne insertion and oligomerization. The use of chiral metallocenes, such as ethylenebis(tetrahydroindenyl)ZrC1 2 / MAO ((EBTHI)ZrC1 2, (8/MAO) as stereospecific Ziegler-Natta a-olefin polymerization catalysts is well documented.19,20 Although mechanistic studies have been performed on selected cationic group IV systems, the fact that neutral group III metallocene complexes, such as rac-Me 2Si(2-SiMe 3-4-CMe 3 )iY-H (rac-[Bp]YH, 9b) and rac(C 20 H 120 2)Si(2-SiMe 3-4-CMe 3)iY-H (rac-[BnBp]YH, 10b), are single component catalysts (require no activator or cocatalyst) makes them particularly well suited for mechanistic studies. Several factors contribute to the stereospecificity with which these C 2 symmetric catalysts promote olefin polymerization. The high isotacticities exhibited by the polymers produced indicate a very high degree of enantioselectivity for olefin insertion into M-alkyl bonds. Isotopic labelling studies on olefin hydrogenation, hydrodimerization, and oligomerization using rac-(EBTHI) complexes indicate a considerably lower (and opposite) enatiofacial selectivity for insertion into [Zr+-H] bonds.21-24 Similarly, although insertion of a-olefins into the Ti-ethyl bond of rac- 100 R n+2 / Me • H -CH(SiMe3)i----1.._ [(5)-[Bp]Y-H] n1g;:s· R n+2 ~ (9b) Me3S (5)-[Bp ]Y-CH(SiMe3)z R n+2 (9a) highly isotactic poly(cx-olefin) (>95% mmmm pentads) / M~s· Q. · ~ y CH(S"M ) H2 o --S~/-~ 1 e3 2 - Me3S ~ [ (R,5)-[BnBp]Y-H] (10b) (R,5)-[BnBp ]Y-CH(SiMe) 2 (10a) Figure 4. Isospecific Ziegler-Natta polymerization catalysts.25 ethylenebis(indenyl)Tt- 13CH2 CH3 proceeds with high selectivity, analogous insertions into the Tt- 13CH3 bond proceed with very low enantioselectivity.26,27 The use of an isotopically labelled and chiral 1-pentene (D3 CCD2 CHDCH=CD2 ) in conjunction with the diastereomerically pure single component catalyst 10b shows that insertion into the Y-H bond proceeds with low (34%) enantiomeric excess, and that subsequent insertion into the Ypentyl species is achieved with high (>40:1) enantioselectivity.28 These results suggest the models for olefin insertion shown in Figure 5. Steric interactions between the C2 -symmetric ligand array and incoming olefin are such that transition state A, in which the R group of the olefin is directed away from the bulky ligand tert-butyl group, is favored only slightly over that of B, resulting in low observed e.e.'s for insertion into the Yhydride. The role of the ligand array is much more pronounced in 101 subsequent insertions, when the growing polymer chain (P) is directed into the open region, away from the tert-butyl groups. With the growing chain in this orientation, assisted by an a-agostic interaction with one of the two methylene hydrogens, the key steric interaction between the metallocene complex and the incoming monomer is that between the olefin alkyl group (R) and the Ca-C~ bond of the yttrium-alkyl; the favored transition state (C) is that in which the R group adopts a trans- orientation to the growing polymer chain. ~SiMe3t ~ S i M e3 t H_. Y__\c--::::::-CH?- Me3Si ccb~ A slightly favored si insertion (70% of product from this TS) B slightly disfavored re insertion (30% of product from this TS) ~ S i M e3 t ~ i M e3 t --1___ IH, H1PcbsY-\.::;;:-CH2 p H Me3S~ C strongly favored re insertion (>95% of product from this TS) D strongly disfavored si insertion (<5% of product from this TS) Figure 5. Transition states for olefin insertion into Y-H and Y-CH2 P bonds of (R,S)-BnBp y_25 Although useful in mechanistic studies, the practical use of group III complexes as Ziegler-Natta catalysts are complicated by the highly electron 102 deficient nature of the metal alkyls- and hydrides. Common pathways leading to catalyst deactivation are formation of bridging hydride ((µ-Hh) dimers and activation of solvent, and often ligand C-H bonds by a-bond metathesis. Metal-alkyl species such as the bis(trimethylsilyl)methyl complexes 9a and 10a often initiate the polymerization of ethylene, but are too crowded to permit insertion of a-olefins. Hydrogenolysis of these M-alkyl complexes serves a convenient method for the formation of less congested M-H complexes which may serve as olefin polymerization catalysts. In the absence of extreme steric congestion imparted by the Cp* 2 ligand array, terminal group III metallocene-hydrides are extremely unstable, and dimerize to form (µ-H)z dimers (Figures 6 and 7). The monomer-dimer equilibria which interconvert the 14-electron monomer and the 16-electron dimers are so heavily weighted in favor of the inactive dimer that monomeric M-H species are never observed in solution for 9 and 10. Rapid dissociation into monomeric M-H complexes is a prerequisite for active group III and Ln- hydride catalysts; the terminal hydrides are estimated to be 8 to 10 orders of magnitude more reactive than their (µ-H)z counterparts.29 (12) Figure 6. Group III bridging hydride dimers. Careful investigation of the hydrides formed from the chiral 9a and 10a show two types of (µ-H)z dimers: homochiral (formed from the dimerization of two metallocenes of like chirality, 9c and 10c), and heterochiral (formed from dimerization of a pair of metallocenes of opposite chirality, 9d and 10d). When a mixture of (R) and (S) 9a or 10a metallocenes are used to generate YH complexes, a 1:1 mixture of the homochiral and heterochiral dimers are observed immediately upon hydrogenolysis; over the course of several hours, 103 the mixture is converted entirely to the thermodynamically stable homochiral dimer. (9c, 10c) homochiral (µ-Hh dimer (9b, 10b) never observed (9d, 10d) heterochiral (µ-Hh dimer Figure 7. Monomeric/homochiral-/heterochiral hydride dimer equilibria; (R2Si = Me 2Si, (±)C 20 H 12O 2Si). The use of sterically demanding pentamethylcyclopentadienyl ligands may be exploited to permit manipulation of monomeric M-H complexes. Hydrogenolysis of Et 2Si(C 5Me4 )(C 5 H 4 )MCH(SiMe 3 h (M=Lu, Y) permits observation of a monomeric Lu-H complex prior to formation of "flyover" (µHh dimers; observation of the corresponding Y-H monomer is reported, 1 although the doublet in the H NMR diagnostic for a monomeric 89 Y-H species is not observed.29 Although the monomeric Cp* 2Y-H is not observed in solution, monomeric Cp* 2Sc-H is the (moderately) stable hydrogenolysis product of Cp*2Sc-CH3 ? 104 H2 Cp*2Sc-CH3 - Cp*2Sc-H (3) Figure 8. The results of increased steric crowding on formation of (µ-Hh dimers . The destabilization of (µ-Hh dimers is generally accompanied by an increased reactivity towards H-H and C-H bonds. 3 reacts rapidly with H2 in solution, even at -80 °C. Arene C-H bonds are also readily activated by 3; although slow at -80 °C, H/D exchange between Cp\Sc-D and C6H 6 is >95% complete within 5 minutes at 25 °C. Activation of 1° C-H bonds also commonly leads to formation of M-R complexes such as Cp*2 M-Ph (M=Sc,7 Y,31 Lu32) and Cp\M-C6H 4 -MCp*2 (M=Lu32) . Electron deficient early transition metal complexes are also known to activate exposed C-H bonds of the metallocene ligand array, as shown in Figure 9. Although considerably slower than 1° C-H bond activation, examples of both inter- and intramolecular 2° C-H activation are well documented.7,31,33-35 105 ··"H Z r....., Figure 9. Intra- and intermolecular ligand C-H activation. The attraction of using sterically congested pentamethyl cyclopentadienyl ligands to generate moderately stable monomeric 14electron d 0 metal complexes is offset by the inability of these complexes to undergo multiple ex-olefin insertions. The monomeric hydride 3 readily inserts one equivalent of propene at -80 °C; the resultant Sc-CH2 CH3 complex, however, does not insert additional equivalents of olefin, and is subject to <Jbond metathesis with available C-H bonds to form Sc-vinyl complexes. In the presence of only one or two equivalents of olefin, or in the presence of excess olefin at elevated temperatures, Cp*2 Sc-CH3 is shown to insert one equivalent of propene to generate Cp*2 Sc-CH2 CH(CH3) 2 . Both a-bond metathesis with additional equivalents of propene and ~-H elimination from 106 the isobutyl complex, however, are much faster than insertion of additional olefins. Results and Discussion The incorporation of bulky methyladamantyl groups as ansametallocene ~-substituents is designed to perturb the monomer-dimer equilibria commonly observed for group III and Ln hydrides. Although they are unlikely to serve as stereospecific olefin polymerization catalysts, hydride complexes ultimately derived from C5 - symmetric scandocene 13 and yttrocene 14 (meso-AdpScCl-LiCl-(THF) 2 and meso-AdpYCl-LiCl-(THF)i, respectively) should provide preliminary evidence for the ability of the methyladamantyl groups to disrupt formation of (µ-H)i dimers. Of greater interest, though, are the effects of incorporation of bulky ~-substituents into chiral metallocenes such as those derived from the [Abp] ligand array. Disruption of the monomer-dimer equilibria may result in highly active and enantioselective ex-olefin polymerization catalysts. ~~~ti~. lHF Me2S~/Sc.;;.._Cl---- \ ~ MeAd THF <~?4 THF THF MeAd meso-AdpScCl-(LiCl)-(THF)2 MeAd meso-Adp YCl-(LiCl)-(THFh (14) (13) 3s· MeA 3 M s· ~ Q MeAd ~ MeAd \,--cl_ _ I SiM ~ M s l ......... ,THF Me2S~1 ---cl--]iMe iMe2 e2 ~ - - C l q Me,S~MeAd , , MeAd [rac-AbpY(µ-Cl)h (15) Me3S~ M e A d rac-Abp YCl(THF) 107 Treatment of Et2 O solutions of mes o-AdpMCl-LiCl-(THFh (M = Sc, Y) with a slight excess (-1.1 equivalents) of C3H 5 MgBr (-78 °C), followed by stirring at 25 °C, results in the formation of mes o-AdpSc(ri 3 -C3H 5 ), 16, and meso-AdpY(ri -C H 17, as yellow diethyl ether soluble solids. Residual THF and LiCl in 13 and 14 neither impede the alkylation nor coordinate to Sc or Y in the resulting ri -allyl complexes, and may be removed from 16 and 17 by a series of petroleum ether and toluene washings. The ri coordination of the allyl moiety may serve to prevent coordination of donor solvent and Li / Mghalide to the electrophilic metal centers. 3 ), 3 5 3 3 - ~MgBr 13 (5) - LiX, MgX2 ~MgBr 14 (6) - LiX, MgX2 Although the allyl ligands appear ri -coordinated m solution, variable temperature NMR spectroscopy indicates fluxional behavior associated with the allyl ligands. A detailed investigation of the fluxional behavior of these, and other, group III metallocene allyl complexes is described in the following chapter (vide infra) . These studies indicate that ri 1 -allyl complexes are kinetically accessible, although ri 1 - complexes are never observed directly (Equation 7) . Upper limits on the barriers to dissociation of the allyl C=C double bond (generation of an ri 1 -allyl species), as determined by lineshape analysis of the variable temperature NMR spectra, are found to be 12.0 kcal/mol (290 K, Et O-d for 16, and 15.6 kcal / mol (361 K, Et O-d and 16.0 kcal/mol (370 K, toluene-d for 17. 3 ) 2 ) 2 10 ) 8 10 108 Q MeAd MeAd meso-AdpM(17 3 -C3H 5) A!SP--MeAd M0, ... . ~ ~ 51 M.,- ~ ~ (7) MeAd meso-AdpM(ri 1-C 3 H5 ) The role of 17 1 -allyl complexes in the reactivity of meso-AdpM(C 3 H 5 ) is not known. Although the 14-electron 17 1-species generated by dissociation of the allyl C-C double bond are expected to be much more reactive than the ground state 16-electron 17 3-complexes, reactions such as hydrogenolysis may proceed by associative mechanisms in which olefin displacement follows coordination of substrate to the metal center. Treatment of 16 or 17 with excess H 2 (1 atm, 25 °C, C 6 D 6 ) results in the generation of propylene from a-bond metathesis of the allyl ligand with dihydrogen; the monomeric 14-electron metal hydride species meso-AdpMH, the initial organometallic product derived from hydrogenolysis, is not observed during or following hydrogenolysis. A small amount (~3%) of the overall product mixture from the hydrogenolysis of 17 is identified as the bridging hydride dimer [meso-AdpY(µ-H)h on the basis of the appearance of a 1 1:3:1 triplet (8 4.65, Jy_H = 31.1 Hz) due to coupling to the 100% natural abundance 89 Y (I= 1/2); no other resonances attributable to Y-H species were observed. The absence of a corresponding (µ-H)i resonance for the Sc hydrogenolysis product [meso-AdpSc(µ-H)] 2 is attributed to broadening from quadropo 1ar 455 c. 109 - unidentifiable decomposition products (>95%) Figure 10. Only small quantities of (µ-H)i dimers are observed upon hydrogenolysis of meso-AdpM(C 3H 5) (M=Sc, Y) The organometallic products resulting from hydrogenolysis of 16 and 17 do not react with a-olefins. The propylene generated from initial hydrogenolysis of the allyl ligand is never converted to propane via insertion into a M-H species and subsequent a-bond metathesis with H 2, despite the presence of excess dihydrogen. No reaction is observed between the organometallic products and added olefinic substrates such as 1-pentene. Identification of the major organometallic hydrogenolysis products is complicated by the number of possible decomposition pathways, including solvent C-H bond activation, intramolecular ligand C-H activation, and intermolecular ligand C-H activation. Deuterium labeling studies and hydrogenolyses in nonaromatic solvents (C 6Dii), intended to identify arene (solvent) C-H activation, are inconclusive. Molecular models indicate that each methyladamantyl group in [Adp] has two sterically accessible C-H bonds which may undergo a-bond metathesis (Figure 11; these products share structural similarities to the 2° C-H bond activation products observed previously by Bercaw (Figure 9).35 Although hydrogenolysis with D 2 fails to result in an observable decrease in intensity of the methyladamantyl resonances (1H NMR), these observations are 110 complicated by the multiple overlapping resonances of the thirty-four methyladamantyl protons. M-H Me· .. 5 M~ Figure 11. Possible intramolecular C-H bond activation. Despite the presence of the bulky methyladamantyl substituents, the C52 symmetry of the meso-Adp metallocenes results in exposed sp -hybridized cyclopentadienyl C-H bonds that may undergo c,-bond metathesis with M-H species to form dimers similar to those observed from [CpCp*ScCH3 (PMe3 )] (Figure 9).33 Multiple complexes resulting from intermolecular C-H activation are possible as a result of the exposed vinylic C-H bonds in the 4and 5- positions of the Cp rings; the possible dimer resulting from activation of the 4-CH bonds is shown in Figure 12. MeAd MeAd M\ ,, s···•' _J MeA MeAd Me u Me Figure 12. One of several possible dimers resulting from intermolecular C-H activation. 111 Efforts to trap [Adp YH] with a coordinating ligand are largely unsuccessful. Hydrogenolysis of a solution of 17 containing excess trimethylphopshine does not result in the clean formation of mesoAdp YH(PMe 3); a mixture of at least four products is obtained (1H NMR). Dihydrogen and trimethylphosphine pressure dependencies of the relative ratios of the products suggest equilibria between species such as those shown in Figure 13, in addition to the possible C-H activation products outline above. Figure 13. Possible equilibria involving meso-AdpM-H, trimethylphosphine, and dihydrogen (M = Sc, Y). These overall results are consistent with the hydrogenolyses of mesoAdpM(C 3H 5 ) to generate extremely reactive 14-electron meso-AdpMH monomers, which appear to react with any of a variety of accessible C-H bonds. The incorporation of methyladamantyl groups as ~- substituents appears to inhibit formation of (µ-H)z dimers, in contrast to the related tertbutyl complex [meso-Me 2Si(3-CMe 3-C 5H 3)zSc(µ-H)h (11). It was hoped that the rac- [Abp] ligand array resulting from the net addition of trimethylsilyl groups onto [Adp] would provide metallocene hydride species less prone to irreversible deactivation than those derived from 16 and 17; molecular models indicated that the staggered conformations of the methyladamantyl and SiMe 3 groups in rac-AbpMH provide considerable steric hindrance to the approach of two M-H monomers, thus 112 rendering intermolecular ligand C-H activation less likely than that shown in Figure 12. Although the use of (C3H 5 )MgBr provides a facile means of generating mesa- group III allyl complexes, analogous reactions with the [rac-AbpY(µCl)h / [rac-AbpYCl(THF)] mixture fails to reproducibly provide pure samples of Abp Y(C3H 5 ). Uncertainties in the Y:Mg stoichiometry may lead to formation of bis(allyl)yttrates along with the desired Abp Y(C3H 5 ) . Bis(trimethylsilyl)methyllithium reacts rapidly with [rac-AbpY(µ-Cl)h (benzene or toluene) to generate Abp YCH(SiMe3)i, 18. A beige powder, 18 is isolated in 25-35% yield after removal of solvent, removal of KCl by Et2O filtration, and recrystallization from cold (-78 °C) pentane. The loss of C2 symmetry in the 1H and 13C NMR spectra of 18, (benzene-d6 , toluene-d6, cyclohexane-d 12 , methylcyclohexane-d 14 ) evidenced by the inequivalence of the Cp rings and substituents, is indicative of hindered rotation about the Y-C bond (Figure 14). (18) 15 MeAd ? - s i M e3 1-fy Cl (8) MeAd ~ S i M e3 (TMS\,Y )TMS)a (TMS)[ j'(TMS)6 Me3S ~ MeAd Me,S~ MeAd 18 is an active catalyst for the polymerization of ethylene (1 atm. C2H 4, 25 °C, toluene). No reaction is observed, however, between 18 and either 20 equivalents of I-butene or 230 equivalents of 1-pentene (days, 25 °C, C6 D6 ) . Steric congestion at the metal center imparted by the methyladamantyl and bis(trimethylsilyl)methyl groups appears to prevent both olefin insertion and cr-bond metathesis. "Tl ..... ...... _ _,/,11 ~ 00 '---" 1 111 7.0 1111 111111 6.0 '1 l~ -- CDH 11~ 6 5 '!'•' I N [rac-Abp Y(µ-Cl)h + LiCH2 (SiMe3)i Me,, .. . 51 Me, Me3 3 I I I I I I I 5.0 l I I I I I I I I I I -, 4.0 I I I I I I 3.0 I I I I l J I I i- 2.0 I , ~ Vil~ .'J~\ i' 1 1 .. f · ,~ ; , 1.0 - ,,-----, - - - I, .;ti JJ 1 J! ~ \ h I ~'-............. PPM 7 ~-- / / :!: CH(SiMe3 ,'', ;1 ,I' ., ' j ,' , f I.I t\ Si(CH3h CH3-C10H14 Si(CH3)i CH3-c10H 14 • · ~-, 'e3 - - - - - - - - - - - - - - - - - 1~1\____J' V. . , Me2 Si[Csf{ 2(SiMe3)(CH3-C10H 14)b ~ ro ':::...., ----CJ) "'Cl 5= >-< n ::c I r, :;:, .... >-1-, 0 '---" "'tj ~ Pl '"1 n ,...,. ro "'Cl (/) ~ ~ z ::c N ::c ~ CJl 0 0 ~ >--1 re ()Q ~ '"1 f---' f---' vJ 114 R 18 I no reaction Figure 15. rac-Abp YCH(SiMe 3 )z (18) polymerizes ethylene but does not react with a-olefins. 18 also fails to react with bulky terminal alkynes such as trimethylsilylacetylene and tert-butylacetylene at room temperature. This is in marked contrast to Cp* 2YCH(SiMe 3 )z, which rapidly undergoes cr-bond metathesis with terminal alkynes to generate Y-alkynyl species and H 2C(SiMe 3)z.36 Heating 1:1 18:alkyne solutions (80 °C, C6D 6 ) results in the evolution of H 2C(SiMe 3 )z, although the resultant organometallic products can not been positively identified as either alkynyl products resulting from crbond metathesis or alkenyl-yne complexes arising from insertion of additional equivalents of alkyne to a Y-alkynyl complex. The reaction of 18 with dihydrogen proceeds cleanly and quantitatively via cr-bond metathesis, resulting in the formation of H 2C(SiMe 3 ) 2 and a single hydride dimer, [rac-AbpY(µ-H)]z (20), as the sole organometallic product 1 (Figure 16). The presence of a 1:2:1 triplet in the H NMR spectra of 20 due to coupling to equivalent 89 Y nuclei is diagnostic for the presence of a symmetric, hydride-bridged yttrocene dimer (Table 1). None of the monomeric rac-Abp YH (19), which should exhibit a characteristic Y-H doublet and a distinct set of ligand resonances, is observed under any hydrogenolysis conditions. 18 H2 I -H2C(SiMe3)z Me3S ~ ~ , MeAd Me2S1 J-H Me,~MeAd (19) (not observed) (9) 115 1 Solvent 8 (ppm) benzene-di; 4.76 31.3 cyclohexane-d, 7 4.48 31.0 4.39 31.2 me thy lcyclohexane-d 1 ,1 Jv-H (Hz) Table 1. 500 MHz 1H NMR chemical shifts and one-bond Y-H coupling constants for the bridging hydride resonance of [rac-AbpY(µ-H)h (20). Variable temperature NMR studies rule out the possibility of an energetically accessible homochiral-heterochiral dimer equilibrium in methylcyclohexane. Hydrogenolysis of 18 at 246 K (C 7D 14 ) confirms that the triplet observed in the room temperature hydrogenolysis is the kinetic hydrogenolysis product as well as the thermodynamic product. Neither additional hydride resonances nor changes in the chemical shift of the observed (µ-H)i triplet, are observed upon cooling room temperature solutions of 20 to 213 K, indicating the absence of exchange between homochiral and heterochiral dimers on the NMR timescale. This result also indicates that dissociation of 20 to 19 is slow on the NMR timescale. Assignment of 20 as the homochiral dimer is based, in part, on the previously observed stabilities of 9c and 10c (previous homochiral dimers), compared to the instabilities of the heterochiral dimers 9d and lOd.37 Further support for the assignment of 20 as the homochiral dimer comes from the upfield chemical shift of the (µ-H)i resonance (cf. 9c: 8 4.87, J=31 Hz, C6D 6 ; 10c: 8 4.98, J=31.4 Hz, C D 10d: 8 5.97, J=31.4 Hz, C D Single crystals grown by cooling a concentrated cyclohexane solution of 20 show severe twinning, and ; 6 6 ) , 6 6 are not suitable for study by X-ray diffraction techniques. Efforts to trap the highly reactive 19 monomer by addition of donor ligands are only partially successful. Hydrogenolysis of 18 in the presence of 1.1 equivalents of THF shows a doublet (8 5.52, J=72.8 Hz, C6D6 ) in the region of the 1H NMR spectrum in which the Y(µ-Hh Y and YH(THF) resonances are typically observed, suggesting the formation of rac-Abp YH(THF) (cf. Cp* 2YH(THF): 8 6.17, J=81.7 Hz, C 6D 6). Formation of rac-AbpYH(THF) is accompanied by the formation of an intractable mixture of additional organometallic products, as evidenced by the presence of multiple Cp, methyladamantyl, and Si-Me resonances. Following the hydrogenolysis by 'T1 ..... -....., N '---' -....., ::r: 1=I ,......,_ >-< ► Ci "D (j I i:::i ~ 0>-;-, -....., N ....~ (t) >< Pl :::i (t) 0 ::,-- n '-<! n ,......,_ Pl n ..... '"I (t) "D :;:;::, en ~ z ::r: .... ::r: N ~ 0 0 U7 . N~ 0 O"I ,......,_ (t) 1-1 C CJCl 1 ..-i~ , 1 r I 6.0 I "---- t L_J' I r 1 ,- I -, ,- r ·1 5.0 1 1 - , I I I ~ )_ ;. r I I Y(µTH)zY Me2 Si[Csf{ 2(SiMe 3)(CH3-C10H 14)h ____.) Ii I Me••s· Me ... 1 j 4.0 t ~'-------------->·~··, ...____ H2 rac-AbpYCH(SiMe3)z - - - - I Me~ - ,- , I J 1 I -, I I 3.0 I I I I I ' I I 2.0 '' I 1.0 -,---.--~r,,- PPM VL . - j'v'J~\)\L._ __ _}1': ,_ _____ _,1/ .· , , - CH3-C10H;4 ;i Si(CH3)} CH -c H Si(CH3)z 3 10 14 i¢lJ: H 2C(SiMe 3)z 0\ f---l f---l 117 1 H NMR in neat THF-d 8 does not provide additional information on the fate of [rac-AbpYH]. The hydride dimer 20 is not cleaved by added PMe3; similarly, hydrogenolysis of 18 in the presence of excess PMe 3 results only in 20, with no evidence for the formation of rac-AbpYH(PMe 3 ). Evidence for dissociation of 20 into monomeric rac-Abp YH, however, is obtained from H/D scrambling experiments. The deuteride-bridged dimer 20-d 2 may be generated by placing 20-d0 under one atmosphere of D2. The NMR spectrum of 20-d 2, which may also be obtained from treatment of 18 with D 2, is identical to that of 20-d0 , save for the absence of the (µ-H)i triplet. Complete conversion from 20-d 0 to 20-d 2 is observed after several hours at room temperature, indicating a series of dimer dissociations and cr-bond metathesis steps outlined in Scheme 6. 1/2 AbpY(µ-HhYAbp ~ ~ (20-do) Abp Y(µ-H)(µ-D)YAbp (20-d 1 ) 1/2 AbpY(µ-D)iYAbp 19-do ~ ~ [Abp YH] (19-d 0 ) ~~ l9-d 0~ ~ (20-d 2 ) Scheme 6. Dimer dissociations and a-bond metathesis steps required to account for H/D scrambling into 20. Dissociation of 20 into 19 and subsequent a-bond metathesis with excess dihydrogen is, however, slow on the NMR timescale at 25 °C, as evidenced by the presence of a sharp singlet for free H 2 accompanying the characteristic (µ-H)i triplet of 20 (benzene-d6, cyclohexane-diz, methylcyclohexane-d 141 25 °C). Rapid exchange would result in an averaged H 2 and M-hydride signal, as is observed for monomeric Cp* 2ScH? 20 is also relatively unreactive towards the C-H bonds of aromatic solvents . Whereas 118 the permethylscandocene hydride monomer readily undergoes cr-bond metathesis with the aryl C-H bonds of benzene, deuterium incorporation into 20 (in the absence of D2 ) is not observed after several hours in C6 D6 or C7D 8 (25 oc). Although 20 does not readily activate solvent C-H bonds, the overall stability of 20 is greatly diminished in aromatic solvents, compared to that in aliphatic hydrocarbons. Decomposition of 20 in C6D6 is observed within minutes in the absence of excess dihydrogen (25 °C). Under one atmosphere of H 2, 20 is stable in benzene-d 6 or toluene-d 8 for several (<12) hours. By contrast, cyclohexane-d 12 and methylcyclohexane-d 14 solutions of 20 are stable for weeks at room temperature. The decomposition mechanisms operating are not currently understood, and are beyond the scope of this study. It does not appear that 19 or 20 activates the C-H bonds of benzene to afford Y-phenyl complexes, as no reaction is observed between 20 and excess C6H 6 (10 equivalents per yttrium) after 24 hours at 25 °C in cyclohexane-d 12 . The stability imparted by the introduction of trimethylsilyl substituents in 20 (either by virtue of the reduction in the number of available C-H bonds or due to steric hindrance provided by the racemo- conformations of the resultant metallocenes) compared to the rapid decomposition of [meso-AdpMH] indicates that decomposition of the meso- Adp complexes may occur primarily by intermolecular cyclopentadienyl C-H bond activation. Hydrogenolysis of 18 in the presence of excess 1-butene (20 equivalents per Y) results in the formation of 20 and butane (<45 minutes, 1H NMR). The proposed catalytic cycle for olefin hydrogenation is shown in Scheme 7. 119 rac-Abp YCH(SiMe3)z H2 t -H C(SiMe h 2 3 rac-AbpYH jt [rac-AbpY(µ-H)h Scheme 7. Olefin hydrogenation catalyzed by [rac-Abp YH] (19) . The influence of the methyladamantyl groups on the reactivity of 19 is most easily observed by the introduction of a-olefins to solutions of 20. Treatment of 20 with excess a-olefin does not result in the formation of poly(a-olefin). 20 fails to generate polymers even under reaction conditions analogous to those successfully employed in a-olefin polymerization by 9c (90% v /v 1-pentene/ cyclohexane, 11 days, 25 °C). n+2 20 .. V IR ---i/\1------- Ji1µR R (10) n When followed by 1H NMR, 20 begins to react within 20 minutes of the addition of excess 1-butene or 1-hexene (20 equivalents per Y) to a solution of 20. Complete reaction/ decomposition of 20 is observed within 120 minutes at 25 °C, although there is virtually no consumption of alkene. 20 is not converted cleanly to either an identifiable yttrium-alkyl or yttrium-alkenyl complex; the organometallic products of these reactions show broad and unassignable NMR resonances. The use of 3,3-dimethylbutene38 does not result in significantly cleaner reactions with 20 than do the a-olefins mentioned above. Deuterium labeling indicates that 19 does not insert a-olefins and then rapidly undergo P-H elimination from a transient yttrium-alkyl complex. Treatment of 20-d 2 with 2 equivalents of d0 - 1-pentene (C 6 Dd shows neither H incorporation into the bridging deuteride positions of 20 nor a change in the 120 relative intensities of the vinylic resonances of the free olefin. A similar lack of HID scrambling is observed for a 1:1 mixture of 20-d 2 and neohexene. R I [rac-AbpY(µ-D)J~----[rac-AbpYD] \? ' [rac-AbpYCH 2CHDR] 20-d 0 [rac-Abp Yh(µ-D)(µ-H), -- [rac-Abp YH] 19 -d1 20-d1 19-d 0 Scheme 8. H/D scrambling is not observed between 20-d 2 and perprotio 1olefins (R=CH 2CH 2C 3, CMe3 ). In contrast, the reaction of 20 with allene proceeds smoothly over 24 hours at room temperature, affording the allenyl complex 21. The proposed mechanism for formation of 21 is shown in Figure 17; dissociation of 20 into monomeric 19 is followed by a-bond metathesis with a C-H bond of allene to generate dihydrogen and 21. There is no evidence for insertion of allene into 19 to generate rac-AbpY(ri3-C 3H 5 ). The NMR spectrum of 21 (Figure 18) shows equivalent Cp rings and substituents, indicating rapid rotation about the Y-C bond. The C 2- symmetry of the [rac-Abp] ligand array renders the two terminal allenyl methylene hydrogens magnetically inequivalent, regardless of the rate of Y-C bond rotation. 121 20 - [19] Figure 17. Allene undergoes a-bond metathesis with 19. Catalytic amounts of trimethylphosphine are reported to catalytically induce dissociation of and equilibration between rac-BpLa(µ-Hh dimers, indicating an associative pathway for dimer dissociation39 The observed rates of consumption of 20 and generation of 21 are not affected (qualitatively) by the addition of PMe 3 . Terminal alkynes such as H-C=C-CMe 3 and H-C=C-SiMe 3 do not react cleanly with 20, instead resulting in the decomposition of 20 over several hours at room temperature. Particularly intriguing, though, is the reaction of 20 with 2-butyne. Given the observed preference for 20 to undergo cr-bond metathesis with rather than undergo insertion with allene, the propargylic product resulting from C-H activation was expected to be the kinetic product "Tj ~ c.n ~ 0 ~ C (1) n °' ► 0 '. ----N -!::: N I I [rac-Abp Y(µ-H)h H2C=C=CH 2 I I I 1 ___.)~..-....._. _ J,.___,______ _ 1 7.0 I 6.0 - ,,--r,-"77,--.--r-r , I I I C6D5Hj I , , 1 , , , Me2Si[Csf!iSiMe3)(CH3-C10H14)h ::c ~- - n ::cII nII n ?' r-< ,-1 'lj c:,' I ....... r, (Jl ;;::, ..., 1--'• (Jl '-<; O 0 ............... oq CJ (1) °' !:l .____, 0 0.. ----~ '-<; s C;:::r ::,- s o n ~ ~ (Jl ....... 'lj ~ ~ wro ~ ----~ z n ::c ~ N,... e. ::c ~ 0 (Jl (1) ,-1 'I- II re ~ i= oq" 5.0 \ ~. 1 , , , , --~ 1 '-.. 1 -,- , 3.0 , , , , I "-- YCH=C=CH2 4.0 i--i---•, -,- , r - 1 - , - , , / Me 3Si Me;:si Me11,, Me~ 1 , -, - , - 2.0 * i V \ .... 1.0 __ PPM ',, I Ii Ji r, _j.!LJvJ. i. LJ : I ,1; ,_ I], ~ Si(CH3h CHTC10H14 CHTC10H14 H Si(CH3h N N ........ 123 with the Y-CH3 bond of CnY(CH3h to give an equilibrium mixture of the propargyl and allenyl complexes shown in Figure 19. K<< 1 rn (CH3h y---~f H2 : -- c F~ H 3C Figure 19. Formation of an equilibrating mixture of propargyl and allenyl complexes are observed from the reaction of 2-butyne with CnY(CH3h. Surprisingly, treatment of 20 with one equivalent of 2-butyne (C6Dii) proceeds over the course of minutes at room temperature to generate the butenyl complex 22, the product of alkyne insertion into the Y-H bond of 19 (Figure 20). The overall C2-symmetry of 22 indicates rapid rotation about the Y-C bond. Formation of the propargyl complex 23 is unlikely based on the observed downfield chemical shift (1 H NMR) of the single proton assigned as Ha (4.76 ppm), the only vinylic proton resonance observed. The observed NMR spectrum is also inconsistent with the allenyl complex 24. The C2symmetry of the rac-Abp ligand array renders the allenyl methylene hydrogens (Hb1 and Hb2) magnetically inequivalent, regardless of the rate of rotation about the Y-C bond. 124 20 MeAd (25a) (25b not shown) Figure 20. Alkyne insertion into 19, followed by ~-H elimination and subsequent a-bond metathesis generates a pair of diastereomeric yttrium-methylallyl complexes. The butenyl complex 22 undergoes a rearrangement over the course of 48 hours to give a ~ 1:1 mixture of two new complexes (25a,b) the NMR spectra of 25a,b are consistent yttrium-methylallenyl complexes (Figure 21). A mechanism accounting for formation of 25a and 25b is given in Figure 20. ~Hydride elimination from 22 generates a 1,2-butadiene-complex, which then rapidly undergoes a-bond metathesis with the available Y-H bond to generate dihydrogen and yttrium-methylallyl complexes. The ~-hydride elimination step from an sp2 -hybridized carbon to generate methylallene has been noted for 2-butenyl derivatives of iridium41 and molybdenum,42 and is implicated in the observed rearrangements of Cp*2 M(C(CH3 )=CHCH3 )H (M=Zr, Hf) to the corresponding crotyl-hydrides (equation 11).43 125 r-1 ~ L 5 t ~ I I i ~ £ :;i ~o .-i r 4 L ! Vl r - ,,- di 1l ~ \I L .::r- ~X: .Q u I :i: £ #u . a :i:J / ~ u 6[ / ui,1,,~ [3jfLl _ ;- '- 1 3' u l a ;,-- er, ::J ::E : Ji I\ ~ ::E <ii' - . ::E : + r M :i: y -i "=' v I - ;t. -l \.) 5' >- >- ., t\ ::E ~ ~ 0' ) Figure 21. 300 MHz 1 H NMR spectra (cyclohexane-dn) of racAbp YCH=C=CHMe (25a,b). - ~ I f - iiiM Ji iL 126 (11) The mixture of observed final products arises from the two possible CH bonds with which the chiral rac-Abp YH moiety can react.44 The observed C 2- symmetry of the Abp ligand arrays of each of the diastereorners of 25 is indicative of rapid rotation about the Y-C bonds (Figure 22). (25a) (25b) Figure 22. Y-C bond rotation for the diastereomeric pair 25a,b . The observed 1H NMR spectra are inconsistent with the formation of either 23 or 24. The observed relative intensities of the vinylic Cp and methylallyl methylene hydrogens are inconsistent with both the propargyl complex 23 and the allenyl complex 24. Furthermore, 23 should exhibit overall C 2- symmetry due to facile rotations about the Y-C bond; this is 127 inconsistent with the 4 Cp, 2 SiMe 3, 2 SiMe 2 (linker), and 2 MeAd resonances actually observed. Similarly, rotation about the Y-C bond in the allenyl complex 24 (as is observed for 21) would result in an observed C 2- symmetric complex by NMR, which runs contrary to the pairwise [Abp] ligand resonances observed. The possibility of hindered rotation about the Y-C bond in 24 accounting for a complex with overall C1- symmetry is refuted on the basis of the pair of alkenyl-methyl resonances in the observed NMR spectra; the substitution of the alkenyl moiety in 24 is such that the methyl group resides in the plane defined by Y-Ca-C~, which results in a single CH 3 resonance, regardless of the rate of Y-C bond rotation. An entirely coincidental 1:1 mixture of 24 and 25 accounting for the observed spectra is theoretically possible, although unlikely based on the extreme bias of the previously observed CnY equilibrium, which heavily favors the alkenyl complex at room temperature. Canel us ions Incorporation of methyladamantyl substituents into the ~-positions of group III ansa-metallocenes is a successful strategy for preventing the formation of bridging hydride dimers. Only very small quantities (~3% of the overall product mixture) of the (µ-H)i dimers are formed upon hydrogenolysis of meso-AdpM(C 3H 5 ). The observed reactivity of mesoAdpM(C 3H 5 ) with H 2 is consistent with formation of highly reactive 14electron d0- metal-hydride species that, since formation of (µ-H)i dimers is disfavored because of the bulky methyladamantyl substituents readily activate available C-H bonds. Intermolecular Cp C-H bond activation is thought to be the preferred mode of decomposition for the mesa- complexes, although other possible mechanisms include intramolecular (methyladamantyl) and solvent C-H activation. rac-AbpYCH(SiMe 3)i (18), obtained from the treatment of [rac-AbpY(µCl)h, is an active ethylene polymerization catalyst. Although the monomeric [rac-AbpYH] (19) is not observed directly from the hydrogenolysis of 18, the ability of the methyladamantyl groups to perturb the hydride monomerdimer equilibria are evident from the formation of only a single (µ-H)i dimer, 128 20. Variable temperature NMR experiments show that the heterochiral [rac- AbpY(µ-H)h dimer is not in equilibrium with 20 (the homochiral dimer) at 25 C, and the heterochiral dimer is not formed even as a kinetic hydrogenolysis product en route to the thermodynamically stable 20. The methyladamantyl groups that prevent formation of the 0 heterochiral dimer also appear to inhibit multiple a-olefin insertions and sbond metathesis with arene substrates. The observation that 20 is a 1-butene hydrogenation catalyst indicates that the [Abp] ligand array is not too sterically bulky to prevent insertion of a single olefin monomer into 19, although steric factors undoubtedly play a major role in the inability of 20 to act as olefin polymerization catalyst. Similar steric factors are believed to preclude the approach of arene C-H bonds, thus rendering 20 considerably more inert to H/D scrambling in perdeuterated aromatic solvents than are complexes such as Cp*2Sc-H. Both allene and 2-butyne react cleanly with 20. Activation of a C-H bond in allene results in the formation of the yttrium-allenyl complex 21; no insertion of allene to generate 11 3 -allyl complexes is observed. Insertion of 2butyne into the Y-H bond of 19 gives the butenyl complex 24. This reactivity is somewhat surprising, given the tendency of allene to resist insertion into 19 and the previously observed C-H activation of 2-butyne with Cn Y(CH3 h to generate an equilibrium mixture of yttrium-propargyl and yttrium-allenyl complexes. The [Abp Y]-butenyl complex undergoes ~-hydride elimination from an sp2- hybridized carbon to generate a methylallenyl complex which rapidly undergoes cr-bond metathesis with available 19 to produce 25a,b, a diastereomerically related pair of yttrium-methylallyl compounds . Assessment of whether dissociation of 20 proceeds by a dissociative pathway or an associative pathway is complicated by the formation of only the homochiral dimer (thus precluding homochiral-heterochiral equilibration studies) and the fact that the observed reactivity of 20 is a function of both the dissociation of the (µ-H)i dimer and the subsequent reaction with substrate. Variable temperature NMR studies in the presence of excess dihydrogen show no evidence for a rapid dissociation of 20 into monomeric 19 at 25 °C. Decomposition of 20 occurs within minutes in the presence of substrates such as olefins, yet 20 reacts over the course of 24 hours in the presence of analogous concentrations of allene. The addition of trimethylphosphine fails to result in a marked increase in the observed rate 129 of the reaction with allene, suggesting a dissociative mechanism for formation of the 14-electron monomer 19. Experimental Section General Considerations: All air and/ or moisture sensitive compounds were manipulated using standard high vacuum line, Schlenk, or cannula techniques or in a dry box under a nitrogen atmosphere, as described previously.45 Argon, hydrogen and deuterium gases were purified and dried by passage over columns of MnO on vermiculite and activated molecular sieves. Solvents were stored under vacuum over titanocene46 or sodium benzophenone ketyl. Allylmagnesium bromide was purchased from Aldrich and used as received. LiCH(SiMe3h was prepared using the method of Cowley47 and sublimed prior to use. Ethylene was dried in a toluene solution of (iBuhAl prior to use. 1-butene, 1-hexene, 3,3-dimethylpropene, 2-butyne, tertbutylacetylene, and trimethylsilylacetylene were stored over activated sieves, and 1-pentene was stored over Na/K. Allene was purchased from Aldrich and purified by several cycles of freeze-pump-thaw . NMR spectra were recorded on a Bruker AM500 (500.13 MHz for 1 H, 125.76 MHz for 13C) or a GE QE-300 (300 MHz for 1 H, 75.5 MHz for 13C) spectrometer. All shifts reported for proton and carbon NMR were referenced to solvent. NMR tube reactions were carried out using flame sealed NMR tubes or J. Young type Teflon-valved NMR tubes purchased from Wilmad. Elemental analyses were performed at the Caltech Elemental Analysis Facility by Fenton Harvey. Preparation of meso-AdpSc(T1 3 -C3H5) (16). In a 100 ml roundbottom flask was placed 2.5 mmol of AdpScCl-LiCl-(THF)2 and a magnetic stir bar. Approximately 50 ml of Et2O was vacuum transferred into the roundbottom, and the maroon-colored solution was warmed to room temperature with stirring. After cooling the solution back to -78 ·c, 2.5 ml of allylmagnesium bromide (1.0 M solution in Et2O, 1.0 eq.) was injected against a heavy argon counterflow. The solution was stirred at -78 •C for 15 minutes, after which time the cooling bath was removed and the solution was allowed to warm to room temperature. After stirring at 25 •C for 6 hours, Et2O was removed from 130 the dark orange solution in vacuo, and the brown foam was left open to dynamic vacuum for 10 hours. The 100 ml roundbottom flask was brought into the glove box and attached to a medium porosity pencil frit equipped with a 100 ml roundbottom flask and a 90° needle valve. Two 20 ml aliquots of petroleum ether were successively vacuum transferred onto and off of the ./ brown solid, followed by a similar procedure using 20 ml of toluene, and two additional petroleum ether washes (40 ml each); the purpose of these washings was to remove residual THF from the LiCl and Mg-halide byproducts prior to filtration. The final petroleum ether solution was filtered, and PE was removed in vacuo, leaving a dark yellow powder, which was dried in vacuo for approximately 6 hours, affording 480 mg (0.84 mmol, 34% yield) of AdpSc(ri 3-C3H5). Contamination of the AdpScCl starting material with excess ,LiCl does not affect the purity or yield of the (ri 3-C3H5) product, as the crude reaction mixture from Li2Adp and ScCl3(THF)3 can also be alkylated cleanly with C3HsMgBr. Anal. Calcd. for C37Hs1SiSc: C, 78.11; H, 9.05. Found (with V2Os added): C, 78.85, 75.78; H, 10.50, 9.98; (without oxidant) C, 74.90, 75.50; H, 9.13, 9.57. Preparation of meso-AdpY(ri 3 -C3H5) (17). A 250 ml roundbottom flask, charged with 3.24 mmol of Adp YCl-LiCl-(solv)2 is attached to a medium sized coarse frit; 115 ml of diethyl ether is vacuum transferred onto the solid and warmed slightly with stirring, giving a clear tan solution. Against a heavy counterflow of argon, 3.6 ml of allymagnesium bromide (1.0 M in Et2O, 1.1 eq.) is syringed into the solution at -78 •C. After 10 minutes of stirring at -78 •C, the cooling bath is removed and the solution allowed to warm to room temperature. After three hours of stirring at room temperature, 60 ml of Et2O is removed in vacuo, and the remaining orange solution was stirred at room temperature for another five hours. The solution is filtered, and the fine white insoluble material is washed once with a small portion of Et2O. Ether is removed in vacuo, leaving a large amount of an orange foam, which is dried in vacuo overnight. The roundbottom flask containing the Et2O soluble species is put on a fine pencil frit, and 40 ml of benzene is vacuum transferred onto the solid. The solution is warmed to room temperature and filtered. The benzene is removed in vacuo, and the benzene-soluble species is dried for 6 hours by exposure to dynamic vacuum, leaving a yellow microcrystalline solid (1.699 g, 2.77 mmol, 85% yield). 131 Preparation of rac-AbpYCH(SiMe3)z (18): A 25 ml roundbottom flask is charged with 276.6 mg (0.358 mmol) of [rac-AbpY(µ-Cl)]2 and 166.38 mg (0.358 mmol) of LiCH(SiMe3)2 and attached to a 180° needle valve. The flask is evacuated on the vacuum line, and toluene (15 ml) is distilled onto the solids (-78 °C); the clear orange solution is warmed to room temperature and stirred at 25 °C for 4 hours. Toluene is removed from the yellow solution with white precipitate, and benzene (10 ml) is distilled onto the resulting white solid, cooled to -10 °C, and removed in vacuo. The flask is brought into the box, attached to a small swivel frit assembly, and evacuated. The diethyl ether (10 ml) soluble species are again dried by freezing a benzene solution (5 ml) and slowly removing solvent under reduced pressure. The pentane (1 ml) insoluble species are isolated by a cold (-78 °C) filtration, affording 77.1 mg of white solid (0.088 mmol, 25% yield). Elemental analysis: Calculated (Found, with added V20 5 ) C: 64.46 (61.00, 61.16); H: 9.34 (8.76, 8.45); N: 0.00 (0.00, 0.00), Calculated (Found, without added oxidant) C: 64.46 (60.61, 60.98); H: 9.34 (8.53, 9.08); N: 0.00 (0.00, 0.00). Preparation of [rac-AbpYH)i (20): A J-Young NMR tube is charged with 15.2 mg of 18 and evacuated on the vacuum line. Cyclohexane-d 12 (0.4 ml) is distilled on from titanocene. The solution warmed to room temperature affording a tan solution with a small amount of finely divided white precipitate, and one atmosphere of dihydrogen is admitted to the NMR tube. Within 2 minutes of the addition of H 2 , the solution is clear and light tan. 1H NMR indicates complete conversion to 20 within several minutes at 25 °C . Preparation of rac-AbpYCH=C=CH2 (21): A J-Young NMR tube is charged with 15.0 mg (0.0171 mmol) of 18. Cyclohexane-d 12 (0.4 ml) is distilled onto the solid from titanocene (-78 °C), and 1 atm H 2 admitted to the tube. The solution is warmed to room temperature, and agitated overnight. The contents of the NMR tube are frozen (-78 °C), and excess dihydrogen removed in vacuo. Allene (0.0690 mmol, 4.0 equivalents) is transferred into the NMR tube using a liquid nitrogen bath. The solution is warmed to room temperature, and 1H NMR spectra are taken periodically. Half of the starting 18 has reacted after 3.5 hours at room temperature, and complete conversion to 21 is observed after 25 hours at 25 °C. Volatiles are removed in vacuo, and 132 cyclohexane-d 12 (0.4 ml) is vacuum transferred back onto the light yellow solid. Preparation of rac-AbpYC(CH3 )=CH(CH3 ) (22) and rac-AbpYCH=C=CH(CH3 ) (25): A J-Young tube is charged with 11.4 mg (0.0130 mmol) of 18 and is evacuated on the vacuum line. Cyclohexane-d12 (0.4 ml) is distilled onto the solid, the tube is sealed, and the solution warmed to room temperature with shaking. After cooling to -78 °C, 1 atmosphere of dihydrogen is admitted to the tube, the tube is sealed, warmed to room temperature and agitated for 30 minutes. The solution is cooled to -78 °C and volatiles are removed in vacuo. 2-Butyne (0.0135 mmol, 1.04 equivalents) is vacuum transferred into the tube from a calibrated gas bulb, the tube is sealed, and the solution is warmed to 25 C. Proton NMR spectra are taken periodically to monitor the extent of the reaction. Conversion to 22 is observable within 20 minutes at room temperature; 25% of 20 is consumed after 60 minutes at 25 °C. Conversion of 22 to 25 is observed within 4 hours of addition of 2-butyne, and the reaction has gone to completion after 43 hours at 25 °C. 0 References and Notes 1. Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. Principles and Applications of Organotransition Metal Chemistry; University Science Books: Mill Valley, CA, 1987, p. 577. 2. Edelmann, F. T., in Comprehensive Organometallic Chemistry II; Abel, E. W., Stone, F. G. A., Wilkinson, G. and Lappert, M. F., Eds.; Pergamon Press: Tarrytown, NY, 1995; Vol. 4, p. 11. 3. Bochmann, M., in Comprehensive Organometallic Chemistry II; Abel, E. W., Stone, F. G. A. and Wilkinson, G., Eds.; Elsevier Science, Inc.: Tarrytown, NY, 1995; Vol. 4, p. 273. 4. Guram, A. S.; Jordan, R. F., in Comprehensive Organometallic Chemistry 11; Abel, E.W., Stone, F. G. A. and Wilkinson, G., Eds.; Elsevier Science, Inc.: Tarrytown, NY, 1995, p. 589. 133 5. Jordan, R. F.; LaPointe, R. E.; Bradley, P. K.; Baezinger, N. Organometallics 1989, 8, 2892. 6. Horton, A. D. Organometallics 1992, 11, 3271. 7. Thompson, M. E.; Baxter, S. M.; Bulls, A. R.; Burger, B. J.; Nolan, M. C.; Santarsiero, B. D.; Schaefer, W. P.; Bercaw, J. E. f. Am. Chem. Soc. 1987, 109,203. 8. Yoder, J. C.; Bercaw, J. E. , Unpublished results. 9. For examples of 2,1 insertions of acetylenes into M+-R complexes, see: Guram, A. S.; Jordan, R. F.; Taylor, D. F. f. Am. Chem. Soc., 1991, 113, 1833. 10. Guram, S. A.; Guo, Z.; Jordan, R. F. J. Am. Chem. Soc. 1993, 115, 4902. 11. Horton, A. D. J. Chem. Soc. Chem. Commun. 1992, 185. 12. Horton, A. D.; Orpen, A. G. Angew. Chem. Int. Ed. Engl. 1992, 31, 876. 13. den Haan, K. H .; Wielstra, Y.; Teuben, J. 1987~ 14. For reactions of internal alkynes with non-Cp containing early transition metal complexes, see: a) Uhrhammer, R.; Black, D. G.; Gardner, T. G.; Olsen, J. D.; Jordan, R. F. f. Am. Chem. Soc., 1993, 115, 8493. b) Hajela, S. P.; Schaefer, W. P.; Bercaw, J. E. J. Organomet. Chem . , 1997, 532, 45. 15. Horton, A. D.; Orpen, A. G. Organometallics 1992, 11, 8. 16. Horton, A. D.; Orpen, A. G. Organometallics 1991, 10, 3910. 17. Heeres, H.J.; Heeres, A.; Teuben, J. H. Organometallic s 1990, 9, 1508. 18. Ionic radius (6 coordinate): Sc3+ 0.75A; y 3+ 0.90A; La3+ 1.03A; Ce 3+ 1.07A; Zr4 + 0.72A; Shannon, R. D. Acta. Cryst. , Sect . A, 1976, 32, 751. 19. Brintzinger, H. H.; Fischer, D.; Mulhaupt, R.; Rieger, B.; Waymouth, R. N. Angew. Chem . Int. Ed. Eng. 1995, 34, 1143. 20. Bochmann, M. f. Chem. Soc., Dalton Trans. 1996, 255. 21. Pino, P.; Galimberti, M. J. Organomet. Chem. 1989, 370, 1. 134 22. Krauledat, H.; Brintzinger, H. H. Angew. Chem. Int. Ed. Engl. 1990, 102, 1412. 23. Waymouth, R.; Pino, P. J. Am. Chem. Soc. 1990, 112, 4911. 24. Pino, P.; Cioni, P.; Wei, J. J. Am. Chem. Soc. 1987, 109, 6189. 25. The (R,S) designation refers to metallocenes derived from (R)binaphthol, which results in metallocenes of (S) stereochemistry. 26. Longo, P.; Grassi, A.; Pellecchia, C.; Zambelli, A. Macromolecules 1987, 20, 1015-1018. 27. Zambelli, A.; Pellachia, C.; Oliva, L. Makromol. Symp. 1991, 48/49, 297. 28. Gilchrist, J. H.; Bercaw, J.E. J. Am. Chem. Soc. 1996, 118, 12021. 29 . Stern, D.; Sabat, M.; Marks, T. J. J. Am. Chem. Soc. 1990, 112, 9558. 30. den Haan, K. H.; Wielstra, Y.; Teuben, J. Organometallics 1987, 6, 2053. 31. Booj, M.; Deelman, B. J.; Duchateau, R.; Postma, D. S.; Meetsma, A.; Teuben, J. H. Organometallics 1993, 12, 3531. 32. Watson, P. L. J. Chem. Soc. Chem. Commun. 1983, 276. 33. Bunel, E. E., Ph.D. Thesis, California Institute of Technology, 1989. 34. Hajela, S.; Schaefer, W. P.; Bercaw, J. E. Acta. Crystallogr. 1992, C48, 1771. 35. Bercaw, J.E. Adv. Chem. Ser. 1978, 167, 136. 36. den Haan, K. H.; Wielstra, Y.; Eshuis, J. J. W.; Teuben, J. H . J. Organomet . Chem. 1987, 323, 181. 37. Mitchell, J. P.; Hajela, S.; Brookhart, S. K.; Hardcastle, K. I.; Henling, L. M.; Bercaw, J. E. J. Am. Chem . Soc. 1996, 118, 1045. 38. Although the analogous reactions with 1-pentene are not clean, Cp*(C5HsBN(CHMeih)HfH(PMe3) cleanly inserts neohexene to afford the Hf-CH2CH2CMe3 complex. M.S. Thesis, California Institute of Technology, 1996. Chem. Macromol . 135 39. Levy, C. J.; Bercaw, J.E. , Unpublished results. 40. Hajela, S.; Schaefer, W. P.; Bercaw, J. E. J. Organomet. Chem. 1997, 532, 45. 41. Schwartz, J.; Hart, D. W.; McGiffert, B. J. Am. Chem. Soc. 1974, 96, 5613. 42. Bottrill, M.; Green, M. J. Am. Chem. Soc. 1977, 99, 5795. 43. McDade, C.; Bercaw, J.E. J. Organomet. Chem. 1985, 279,281. 44. Although only the reaction of 1,2-butadiene with metallocenes of (S) chirality are shown explicity, it is assumed that the (R) metallocenes react with both available 1,2-butadiene methylene hydrogens to give the enantiomeric pair of the diasteromeric pair of yttrium-butadienyl complexes shown. 45. Burger, B. J.; Bercaw, J. E. New Developments in the Synthesis , Manipulation, and Characterization of Organometallic Compounds; ACS Symposium Series, 1987 357, 46. Marvich, R. H.; Brintzinger, H. H.J. Am. Chem. Soc. 1971, 93, 2046. 47. Cowley, A . H.; Kemp, R. A. Synth. React. Inorg. Metal-Org. Chem. 1981, 11,591. 136 1 13 Appendix B: H and C NMR Data For 16-25 Table 1. Proton NMR assignments for 16-25 (500 MHz, C6 D 6 , 25 °C unless otherwise noted) Compound o (ppm) Assignments J H-H 0.23 (s) mes o-AdpSc(ri 3-C3Hs) M€2Si (3H) (16) mes o-Adp Y(ri 3-C3Hs) (17) rac-Abp YCH(SiMe3)i (18) (3H) MeAd (6H) MeAd (29H) Cp (2H) (2H) (2H) ri 3-H2CCHCH2 (lH) M€2Si (3H) (3H) MeAd (6H) MeAd (29H) ri 3-H2CCHCH2 (syn) (2H) 11 3-H2CCHCH2 (anti) (2H) Cp (2H) (2H) (2H) 3 ri -H2CCHCH2 (lH) CH(SiMe3)i Me3 Si (43H) Me2 Si (6H) MeAd (51H total) MeAd Cp (4H) 0.69 (s) 0.85 (s) 1.46, 1.49, 1.55, 1.57, 1.59, 1.62, 1.66, 1.67, 1.70, 1.78, 1.92, 2.06, 2.08, 2.39 (m) 5.54 (m) 5.98 (m) 6.55 (m) 7.19 (m) 0.41 (s) 0.72 (s) 0.97 (s) 1.45, 1.48, 1.51, 1.54, 1.56, 1.58, 1.60, 1.65, 1.78, 1.86, 1.94, 1.98, 2.00, 2.04 (m) 2.66 (s) 3.39 (d) 5.79 (m) 5.98 (s) 6.12 (s) 6.72 (m) 0.26 0.28, 0.32, 0.35, 0.36 (s) 0.91, 0.93 (s) 1.63, 1.68 1.43, 1.48, 1.59, 1.66, 1.71, 1.80, 1.92, 2.10, 2.14, 2.19, 2.43, 2.47 (br) 6.47, 6.46, 7.10, 7.12 (d) 2.6 2.4 2.5 12.3 7.4 2.5 7.7, 1.3 (JY-H) 2.0 (h-H) 137 Table 1. (continued) Compound Assignments 8 (ppm) rac-Abp YCH(SiMe3)i CH(SiMe3)i Me3 Si (36H) Me2Si (6H) MeAd (6H) MeAd (16H) Y(µ-H)z Y (lH) not located 0.01, 0.03, 0.23, 0.24 (s) 0.89, 0.90 (s) 1.55, 1.56 (s) 1.58, 1.61, 1.66, 1.79, 1.81 (br) 2.20, 2.24, 2.27, 2.31 (br) 6.27 (d) 6.28 (d) 6.87 (d) 6.89 (d) not located 0.01, 0.03, 0.23, 0.24 (s) 0.89, 0.90 (s) 1.54, 1.56 (s) 1.60, 1.62, 1.66, 1.69, 1.79, 1.82, 2.17, 2,24, 2.32 (br) 6.27, 6.88 (d) not located 0.24, 0.28, 0.34, 0.35 (s) 0.92, 0.94 (s) 1.60, 1.66 (s) 1.20 - 1.60 (br) 1.68 - 2.00 (br) 2.10 - 2.32 (br) 2.44 (m) 6.43, 6.44, 7.07, 7.09 (s) 0.46 (s) 0.96 (s) 1.66 (s) 1.83, 1.97, 1.99, 2.06, 2.23, 2.30 (br, overlapping) 4.76 (t) Cp (4H) 6.54, 6.92 (s) C6D12 (12H) Cp (4H) rac-Abp YCH(SiMe3) 2 C7D14 CH(SiMe3)i Me3 Si (38H) Me2Si (4H) MeAd MeAd Cp (3-4H) rac-Abp YCH(SiMe3)i (18) CH(SiMe3)i Me3 Si (36H) C7Ds Me2Si (6H) MeAd (6H) MeAd (7H) (llH) (6H) (4H) Cp (4H) Me3 Si (18H) rac-[AbpY(µ-H)] 2 (20) Me2 Si (6H) C6D 6 (300 MHz) MeAd (6H) MeAd J H-H 1.9 1.9 1.7 1.5 1 htt= 31.3 138 Table 1. (continued) Compound Assignments 8 (ppm) rac-[AbpY(µ-H)h (20) Me3Si (18H) Me2Si (6H) MeAd (6H) MeAd (4H) (14H) (2H) (4H) (4H) Y(µ-H)i Y (lH) 0.26 (s) 0.83 (s) 1.42 (s) 1.45 (d) 1.60, 1.62, 1.65, 1.71, 1.81 (overlapping) 2.10 (s) 2.13 (d) 2.23 (d) 4.48 (t) Cp (4H) Me3Si (18H) Me2Si (6H) MeAd (6H) MeAd (18H) (SH) (3H) (2H) Y(µ-H)i Y (lH) 6.24, 6.64 (s) 0.17 (s) 0.73 (s) 1.31 (s) 1.50, 1.53, 1.55 (overlap) 1.72 (br) . 2.02 (br) 2.12 (br) 4.39 (t) Cp (4H) 6.15, 6.55 (s) C6D12 rac-[AbpY(µ-H)h (20) C7D14 J H-H 12.0 12.0 10.7 1 htt= 31.0 1 hH=3 1.2 139 Compound Assignments allenyl (lH) 0.29 (s) 0.91 (s) 1.30 (s) 1.46, 1.49, 1.56 (overlapping) 1.61, 1.65, 1.71 (overlapping) 1.77 1.93, 1.96 2.04 2.11, 2.14, 2.16, 2.18 (overlapping) 2.71 (dd) allenyl (lH) 3.50 (dd) allenyl (lH) Cp (2H) Cp (2H) Me3 Si (18H) Me2 Si (6H) YC(CH 3 )=CH(CH 3 ), MeAd (overlapping) YC(CH3 )=CH (CH3) Cp (4H) 4.39 (t) 6.45 (d) 7.19 (d) 0.23 (s) 0.85 (s) 1.18 - 2.40 rac-AbpYCH=C=CH 2 (21) Me3 Si (18H) Me2 Si (6H) MeAd (6H) MeAd (3H) (9H) (2H) (6H) (2H) (6H) racAbp YC(CH3 )=CH(CH3) (24) 8 (ppm) 4.76 (m) 6.71, 6.29 (s) J H-H 3.1, 7.7 3.8, 7.7 2.9 2.0 1.9 140 Table 1. (continued) Compound Assignments 8 (ppm) racAbp YCH=C=CH(CH3) (25 a,b) Me 3Si (36H) Me 2Si (12H) MeAd YCH=C=CH(CH3) (6H) MeAd (overlapping) YCH=C=CH(CH3) (2H) YCH=C=CH(CH3) (2H) Cp (8H) 0.16, 0.23 (s) 0.80, 0.86 (s) 1.21, 1.40 (s) 1.55, 1.62 (s) JH-H 1.40 - 2.40 (br) 2.94, 2.96 (m) 5.39, 7.10 (m) 6.12, 6.30, 7.02, 7.14 (s) Table 2. Proton decoupled 13C NMR assignments for 18, 20 (125 MHz, 25 °C unless otherwise noted) Assignments 8 (ppm) Cmpd solv rac-Abp YCH(SiMe3)2 C6D6 (18) Me 2 Si Me 3Si Y-C MeAd Cp 0.76, 0.96 3.05, 3.19, 6.08, 6.32 1 30.94 ( J(89 y_ 13q = 38.8 Hz) 28.24, 28.29, 28.38, 28.41, 31.13, 33.24, 33.59, 33.73, 33.89, 34.08, 34.48, 34.72, 34.74, 35.23, 37.06, 37.19, 38.95, 39.58, 39.94, 40.49, 41.82, 42.21 121.40, 122.50, 125.79, 126.44, 127.89, 131.42, 151.57, 152.77 141 Table 2. (continued) Cmpd solv rac-Abp YCH(SiMe3)2 C6D12 (18) Assignments o (ppm) Me2Si Me3Si Y-C MeAd 0.58, 0.72 2.74, 2.90, 5.71, 5.91 1 31.41 ( J( 89 y_ 13 q = 39.7 Hz) 27.00, 28.65, 28.76, 28.81, 31.17, 33.36, 33.90, 34.06, 34.20, 34.39, 34.60, 34.86, 34.91, 35.39, 37.25, 37.44, 39.13, 39.99, 40.32, 40.51, 41.86, 42.27 121.45, 122.51, 125.81, 126.36, 127.97, 128.51, 128.56, 131.47, 151.52, 152.78 1.03 3.12 28.75, 28.99, 34.23, 34.37, 34.55, 34.86, 34.98, 37.91, 38.76, 40.24, 42.02 121.83, 122.18, 122.41, 126.00, 152.91 Cp rac-(Abp YH)2 (20) C6D12 Me2Si Me3Si MeAd Cp 142 Chapter 4 Investigation of lntramolecular Allyl (C3H 5 ) Rearrangement Mechanisms and Evaluation of the Strength of Olefin Binding in d 0 -Metallocene - Alkyl - Olefin Complexes Abstract 143 Introduction 144 Results and Discussion 150 Conclusions 181 Experimental 182 References and Notes 185 Appendix C: Varaible Temperature 1 H NMR Data for 15-18 188 Appendix D: Selected gNMR Data 193 143 Abstract Variable temperature 1H NMR spectroscopy indicates fluxional behavior for a number of group III metallocene allyl (ri 3 -C3H 5 ) complexes. Rearrangement mechanisms in which allyl C=C double bond dissociation from the metal center are the rate determining steps are supported by simulation and lineshape analysis of the VT spectra. Activation barriers to olefin dissociation have been determined for Cp*2Sc(ri 3 -C3H 5 ) (6H:): =8.0 :): :): :): kcal/mol, 6S =-14.4 e.u., toluene-d 8; 6H =5.5 kcal / mol, and 6S =-23.3 e.u., THF-d 8 ), mes o-Me2Sil3-[2-(2-CH3 )-C 1J-{ 14 ]-C5H 3 )zSc(ri 3 -C3H 5 ) (6G:J: =12.0 kcal/mo 1, toluene-d8 ), mes o-Me2Sil3-[2-(2-CH3 )-CHff 14 ]-C5H 3 l2 Y(T] 3-C3 H 5) :): :): (6G =15.6 kcal/mol, Et2 O-d 10; 6G =16.0 kcal / mol, toluene-d 8 ), and mes oMe2 Si(3-CMe3 -C5H 3)zSc(T] 3 -C3H 5 ) (6G:J: =13.4 kcal/mol, toluene-d 8 ), mes oMe2Si(2,4-CHMe2 -C5H 2)zSc(T]3 -C3H 5 ) (6G:J: =14.9 kcal / mol, toluene-d 8 ), racem oMe2Si(2,4-CHMe2-C5H 2)zSc(ri3 -C3H 5 ) (6G:J: =14.7 kcal / mol, toluene-d 8 ; 6G+ =14.6 kcal/mol, THF-d 8 ), and (S,R)-(C 2J-{ 10O2 )Si(2-SiMe3 -4-CMe3-C5H 2) 2Sc(11 3 -C3H 5) :): :): (6G =13.6 kcal/mol, toluene-d 8 ; 6G =12.4 kcal/mol, THF-d 8 ). The presence of donor solvents does not dramatically affect the rate of olefin dissociation. A second rearrangement mechanism, which involves 180° rotation of the intact 11 3 -allyl moiety, has been found to operate in those metallocenes whose ancillary ligand arrays adopt rigid mesa-geometries. Lineshape analysis indicates that the rate of 113-rotation is generally : : : 1 order of magnitude faster than olefin dissociation for a given mes o -metallocene. Experimental difficulties preclude confirmation of the 17 3 -rotation mechanism for the racemo-metallocenes. 144 Introduction 0 Formally d - early transition metal - alkyl - olefin complexes are of interest because of their possible roles in Ziegler-Natta catalysis.1,2 Although they have not been observed directly, the existence of d 0 transition metal alkyl - alkene complexes as genuine intermediates has been implicated in the chain propagation step of the Z-N polymerizations of ethylene and cx-olefins (Figure 1). :j: + Figure 1. Transition metal-alkyl-alkene complexes as possible intermediates in Ziegler-Natta polymerization systems (M = 14-electron, d Ti\ Zr+, Hf\ Sc, Y, La; R = -H, -CH -CH CH etc.; P = growing polymer chain). 0 : , 3 , 2 3 The few examples of structurally characterized d 0 -metal - alkyl - olefin model complexes reported utilize olefinic groups linked by hydrocarbon or ether tethers to the metal center; the strengths of the M-olefin interactions were evaluated by dynamic NMR techniques. Olefin dissociation, followed by recoordination of the opposite face of the olefin, is believed to be rapid at -100 °C for the d yttrocene - alkyl - alkene complex 1, as evidenced by the equivalence of the diastereotopic Cp* and geminal-methyl groups in its 1H 0 NMR spectrum (LiG+ ~ 10.4 kcal/mol, toluene-d 8 ).3,4 Similarly, olefin dissociation and subsequent recoordination is believed to be the process which renders the diastereotopic Cp* and gem-methyl groups equivalent m the 1 H NMR spectrum of the cationic zirconocene complex 2 (LiG+ = 10.7 kcal/mol, CD2 Cl2 ),5 although the extent to which solvent and counteranion assist this alkene displacement is unclear. 145 --« -4(~ ~~ ~ (1) &f-tt +~ Zr ~ 0 ( ) MeB(C 6F5 h 2 (not observed directly) ~ ~ #' Zr-O 0 ~ MeB(C F h \(B 6 5 (not observed directly) 0 Figure 2. Spectroscopically characterized d metal-alkyl-alkene complexes. 17 The allyl (C3H 5- ) ligand is ubiquitous in organometallic chemistry, with or 17 -allyl complexes known for virtually every transition metai.6-8 11 1 3 - 3 - Allyl complexes are, themselves, special examples of tethered transition metal - alkyl - olefin complexes, where the barrier to the rc-cr (17 3-17 1) rearrangement shown in Figure 3 may serve as an estimate of the strength of the M-olefin interaction. Variables such as the transition metal, ligand ensemble, and solvent utilized may, in principle, be varied in order to ascertain their relative effects on the strength of olefin binding. Figure 3. Use of 17 3 -allyl species as model complexes for estimating the strength of M-olefin interactions in metal-alkyl-alkene complexes. Complicating the assessment of M-olefin bond strengths by NMR techniques is the fact that more than one mechanism may be contributing to dynamic behavior. The two most widely accepted unimolecular mechanisms for explaining the observed spectroscopic behavior of M-allyl complexes are: 146 An overall process involving dissociation of the allyl C=C double bond, rotation about the allyl C-C single bond, and subsequent recoordination of the C=C double bond to regenerate an 17 3-fragment (mechanism (i), Figure 4);9-11 and 173-rotation,9,10,12,13 which involves a 180° rotation of the intact 173fragment about an axis perpendicular to the plane of the allyl ligand (mechanism (ii), Figure 4). HI C 3 /-- ,......._ H C,, 'C' ' M \ H2 H4 H 'l (i) - H 1 \ H, ~C:::-,_ H3 c ..''M~cy H \ H4 - H I Hl (ii) /,,C,......._ H3 _ 'c,, 'C' ' M \ H2 H4 Figure 4. Proposed allyl rearrangement mechanisms: (i) dissociation of the allyl C=C double bond to generate an 17 1 -allyl complex, rotation about the allyl C-C single bond, and recoordination of the allyl alkene fragment, (ii) a 180° in-plane rotation of the coordinated allyl ligand. Interestingly, only two complexes reported previously, the monomeric, homoleptic allyls of Mo(IV) (3) and W(IV) (4) have been unambiguously shown to undergo both 17 3-17 1 and 17 3 -rotation. Detailed magnetization transfer experiments indicate that process (i) has been found to be faster than process (ii) for Mo(C3H 5 )4,l0 and although both exchange processes are believed to occur for W(C3H 5)4,l0 no rate data for either (i) or (ii) has been reported. If the ground state structure of the related Zr(C3H 5)4 (5) is assumed to be thesame as that of the molybdenum and tungsten compounds, then both mechanisms (i) and (ii) operate in the Zr system. Although the slow exchange limit is not observed (-90 °C), 17 3-rotation is reported to be the faster of the two processes.9,l4 147 (3) (4) (5) Of the bent-sandwich d 0- metallocene alkyl complexes reported previously, what is striking is the lack of understanding of the factors that promote fluxional allyl behavior. Although exceptions to these trends exist, the majority of the isolectronic group IV complexes, both neutral and cationic, possess static 11 3 -allyl ligands. Addition of strong CT-donor and weak n-acceptor ligands, such as THF, PMe3 , and methylimidazole, tends to induce fluxional behavior of the allyl ligand; addition of carbon monoxide, which acts as both a good CT-donor and n-acceptor, fails to induce fluxional allyl behavior in either the cationic Cp*2 Zr complex 7bl5 or the neutral aminoborole 9b.16 Comparison of dynamic behavior for the isoelectroi1ic d 0f11 Ln(allyl) complexes provides less definitive conclusions regarding the factors affecting allyl fluxionality; neither the presence nor absence of donor solvents and ligands can be used to reliably predict allyl fluxionality in related Ln(11 3-allyl) and (crotyl) complexes, shown in Figure 6. 148 -1Rfv 8 BPI-¼ 8 BPhi ~ (6a) (7a) toluene-ds static at -18 °C CDiCli "nearly static" at -78 °C toluene-d8 fluxional at 25 °C (Sa) (9a) benzene-d6 static at 80 °C benzene-d6 static at 80 °C ~ N ... R ~ N ...R \ !(co --- \ c ·co ~V~'IA (9b) (7b) benzene-d6 slow allyl in plane rotation at 25 °C CD2Cl2 static at 25 °C (7c) (Sb) (9c) (9d) ClJ:2Cl2 fluxional at 25 °C benzene-d6 fluxional at 25 °C benzene-d6 fluxional at 25 °C benzene-d6 fluxional at 25 °C Figure 5. d0 Group IV metallocene ri3-allyl complexes.15-17 149 (10) (11) toluene-d8 static at 25 °C toluene-d 8 static at 90 °C on magnetization transfer timescale (12) (13) benzene-d6 and THF-d 8 static at 25 °C cyclohexane-d 12 THF-d 8 static at 25 °C fluxional at 25 °C ;>/21 ~-\\_/(Me Sm-¥ ~ toluene-d8 (14) equivalent Cp* at -80 °C THF-d 8 equivalent Cp* and syn-anti coalescence Figure 6. Related metallocene d 0fn Ln complexes.13,18,19 150 Results and Discussion Reported herein is a mechanistic investigation of the fluxional 0 behavior of d - metallocene complexes 15-21. Variable temperature 1 H NMR spectra have been obtained for 15-21, and lineshape analysis performed on these complexes in order to ascertain whether more than a single allyl (15) Cp*2 Sc(ri3-~H5 ) -v Se-v (19) meso-IpSc(ri 3-C3H 5 ) (20) (21) rac-IpSc(ri3-C3Hs) (S)-BnBpSc(T) 3-C3H 5) Figure 7. Group III metallocene allyl complexes investigated. 151 rearrangement mechanism is required to explain the observed temperature dependent behavior of Cp*2 Sc(ri 3 -C3H 5 ) (15) and metallocene allyl complexes 3 meso-AdpSc(ri 3 -C 3H 5 ) (16), meso-AdpY(ri -C3H 5 ) (17), and meso-DpSc(ri 3-C3H 5 ) (18) [Adp = Me2 Sil3-[2-(2-CH3 )-C 1J-{ 14 ]-C5H 3 l2 , Dp =Me2 Si(3-CMe3-C5H 3 h] . In addition, lineshape analysis of the variable temperature spectral data has been performed form es o-IpSc(ri 3 -C3H 5 ) [Ip =Me2 Si(2,4-CHMe2-C5 H 2) 2 ] (19), racemo3 3 IpSc(ri -C3H 5 ) (20), and (S,R)-BnBpSc(ri -C3H 5 ) [BnBp = (C 2J-l 100 2 )Si(2-SiMe3 -4CMe3-C5 H 2h] (21), whose syntheses and NMR spectra have been reported elsewhere.20,21 Activation parameters have been determined, when possible, for the barriers to olefin dissociation for complexes 15-21, and for the barriers to in-plane allyl rotation for 16-19, with the hope of correlating these rearrangement barriers with particular changes in metal, ligand array, and solvent. Investigation of the Permethylscandocene System --2p{_v ~ ~ TH \ I (15) ,,,, C "-' ~S;,C ' H H C I ~ The 500 MHz H NMR spectra (toluene-d of Cp* Sc(ri -C H are shown as a function of temperature in Figure 8. At low temperatures (<225 K), the allyl ligand is clearly static on the NMR timescale, as evidenced by the inequivalent Cp* resonances (1.73 and 1.86 ppm) and an AM2X2 splitting pattern for the allyl moiety (H3 : 3.93 ppm; H 5 : 2.05 ppm). As the probe temperature is raised, the Cp* resonances coalesce (257 K), followed by the appearance of an averaged Cp* signal at yet higher temperatures (>269 K). Accompanying the averaging of Cp* resonances are the broadening and coalescence (286 K) of the resonances for the syn and anti allyl methylenehydrogens, and the appearance of a new signal (T>309 K) at a chemical shift equal to the weighted average of the individual syn and anti resonances. 1 3 ) 8 ) 2 3 5 152 0 °' C') L!) 0 N ~ -.:I' =:'. ~ <fl ~ z N l: "l:::00 0 0 I Cl) ~ I..() Cl) i-.. .....;:::l ro i-.. Cl) 0.. Cl) ;:::l ....... Cl ,/ :r:-u \"' :r: 0 ..... ro~ .....i-..u Cl) 0.. E Cf) Cl) ~ Cl :r: I :r:-u ~~ CJ)- u ,....l: ~ °' ► I u CJ) ::_ U"l - ('<) Cl) ....... ,..Cl .....ro i-.. ro > Figure 8. Variable temperature 1 H NMR spectra of Cp*2 Sc(C3H 5 ) (15) (500 MHz, toluene-d 8 ). 153 Similar spectral trends are observed for 15 in THF-d 8; low temperature spectra (:'.S: 233 K) indicate a static 17 3-allyl ligand, while fluxional behavior is observed as the probe temperature is raised. Coalescence of the Cp* signals is observed at 229 K, and the syn and anti allyl methylene averaging occurs with a coalescence temperature (Tc) of ~273 K (300 MHz) . The simplest explanation for the observed fluxional behavior of 15 1 involves the rr-cr (allyl 17 3 -17 ) processes shown in Scheme 1. Dissociation of 1 the allyl C=C double bond generates an 17 -allyl intermediate (15-171); rotation about the M-C single bond in 15-171, followed by recoordination of the alkene fragment to form an 17 3 -species (pathway ia), exchanges the environments of the two Cp* rings. An alternate pathway from intermediate 15-17 1 is rotation about the allyl C-C single bond, followed by recoordination of the alkene fragment (ib ), serving not only to exchange the Cp* ligands, but also to exchange the syn and anti substituents on one of the allyl methylene carbons (that which remains coordinated to the metal throughout the transformation). As rotation about the M-C and C-C single bonds are expected to be facile in an 17 1 -allyl complex, both pathways (ia) and (ib) are likely upon generation of 15-17 1. Lineshape analysis22 confirms that the combination of (ia) and (ib) may 1 be used to simulate the observed H NMR spectra of 15. Using only the 173-17 1 processes depicted in Scheme 1, the experimental data could be accurately modeled over the entire temperature range investigated solely varying the rate of alkene dissociation (k1). Olefin dissociation to generate 15-17 1 was modeled as the rate determining step, with subsequent rapid M-C and C-C single bond rotations occurring prior to regeneration of an 17 3- complex. In addition, olefin dissociation from the left side of the metallocene wedge was modeled as occurring with equal probability as dissociation from the right side of the wedge. A stackplot of the simulated NMR spectra (toluene-d 8 ) as a function of temperature is shown in Figure 9. At the limit of slow exchange (k1 = 0), inequivalent Cp* signals and an AM2X2 splitting pattern for the allyl moiety are observed. As the rate of alkene dissociation is increased, averaging of the 154 Net exchange of Cp* rings 3 1 Exchange of Cp* rings syn f2anti on one methylene carbon Scheme 1. 11 -ri Processes for Cp\Sc(C3H 5 ) (15). 155 tj< 0 rl >< *0... 0 u C') 0 C: rl ~ >< II 0 C') ~ \0 rl ..... -s:!; II >< ~ 0 00 ..... ...., ·- 0 t::: rl + II >-, ~ ..... t::: (f} X 0 00 \0 II ..... ~ 0 0 ~ II ..... ~ N X 0 II ..... ~ E NO QJ t::: ·-:5 QJ s X \() 0 rl >< 0 0 Lf) 0 rl ~ >< II 0 0 ~ ~ ..... II ..... ~ Figure 9. gNMR simulation of Cp* and (C3H 5 ) resonances of 15 (500 MHz, toluene-d 8 ) using 17 3-17 1 processes (ia) and (ib ). Q. Q. 156 3 1 sec- ) is followed by coalescence of the syn and anti Cp* signals (k1 == 1.4 x 10 4 1 signals (k1 == 1.3 x 10 sec- ). Observation of a single, averaged allyl methylene resonance (an AX4 splitting pattern) is observed at k1 > 2.0 x 106 -1 sec . A linear Eyring plot for olefin dissociation (Figure 10) may be generated from the observed 300 and 500 MHz coalescence temperatures in conjunction with the corresponding rate constants obtained from gNMR simulations (Table 1). Activation parameters of LiH\ 3_ri 1 = 8.0 kcal/mol and LiS+ 11 3_ri 1 = -14.4 e.u. for olefin dissociation are obtained from the slope and intercept of this plot. These activation parameters provide a iiG\ 3_ri 1 of 11.7 kcal/mol at 259 K (the Cp*-Cp* coalescence temperature); this is in general agreement with the LiG\p•-cp• (12.4 ± 0.7 kcal/mol) obtained from application of the simple twosite exchange approximation I and equation II, where kcp•-Cp*,Tc is the rate constant for Cp* exchange at Tu and L'.iv is the peak separation (in Hz) of the two averaging signals in the limit of slow exchange. 00 kcp•-Cp*,Tc = 2 rc/ ✓2 (Livoo ) (I) LiG+ = -RT[ln(k/T) - 23.76] (II) Coalescence Spectrometer Field Tc (K) Strength (MHz) k 1 (sec- 1) k 1 (sec- 1) from gNMR from (I) C5 Me5 300 250 3.8 X 102 1.7 X 102 C5 Me5 500 257 6.8 X 102 3.0 X 102 C3H 5 syn-anti 300 292 4.8 X 103 * C3H 5 syn-anti 500 299 6.4 X 103 * Table 1. Observed coalescence temperatures and gNMR-derived 17 3 -17 1 rate constants for 15 (toluene-d 8); * = not available using this method 157 4.0 3.5 y = 16.507 - 4009.9x R1' 2 = 0.998 3.60e-3 4.00e-3 3.0 2.5 2.0 i::; c--- .5 1.5 1.0 0.5 0.0 -0.5 -1.0 3. 20e-3 3.40e-3 3.80e-3 4.20e-3 1/T Figure 10. Eyring plot for olefin dissociation for 15 in toluene-d 8 . A statistical factor of two is included in equation I in order to account for the fact that, in addition to the (productive) processes shown in Scheme 1 which serve to exchange Cp* environments, there exist degenerate 17 3-17 1 1 transformations in which the 17 -allyl simply recoordinates to M without net C-C or M-C bond rotations (Figure 11). Simulation of 15 in THF-d 8 , with olefin dissociation again modeled as the rate-determining step for the allyl rearrangement (tabulated in Table 2), provides activation parameters of .1G+ 173_111 , 229 K = 10.8 kcal/ mol, L1H\ 3_11 1 = 5.5 kcal/mol, and .1s\ 3_171 = -23.3 e.u .. For comparison, the two-site exchange approximation I provides .1G+ = 11.3 (± 1.0) kcal / mol for Cp*-Cp* exchange. 158 exchanges Cp* environments no net effect on Cp* environments Figure 11. Origin of statistical factor included in two-site exchange approximation. Coalescence Spectrometer Field Tc (K) Strength (MHz) k 1 (sec- 1) k 1 (sec-1) from gNMR from (I) C5 Me5 300 233 2.9 X 102 1.4 X 102 C3H 5 syn-anti 300 273 1.9 X 103 * Table 2. Observed coalescence temperatures for 15 (THF-d 8 ) and gNMR- derived 11 3 -17 1 rate constants; * = not available using this method The activation barriers to olefin dissociation of -11-12 kcal / mol are comparable to the barriers observed for tethered olefin complexes 1 and 2 (:,;; 10.4 kcal/mol and 10.7 kcal/mot respectively). The activation barriers for 15 may be separated into enthalpy barriers of 5-8 kcal/ mol, and large and negative entropy barriers (-14 to -23 e.u.) . Unlike the previously observed group IV metallocene allyl complexes, the presence of donor solvents appears to have little effect on the rate of olefin dissociation; extrapolation of the ~Gt values for olefin dissociation to 298 K (using the ~Ht and ~St obtained from the Eyring plots) provides ~Gt 113 _111 = 12.3 kcal / mol (toluene) and 12.4 kcal/mol (THF). 159 The use of lineshape analysis does not provide a great deal of 3 information regarding the rate of 11 -rotation for 15. As simulations using 3 1 only the 11 -11 processes (ia) and (ib) explain the observed data, process (ii) need not necessarily be invoked. Control experiments in which lineshape simulations include process (ii) (with k2 s:; (10 x k1 )) do not significantly alter the overall spectral simulations or fit to the observed spectra. Lineshape simulation using only the 11 3-rotation process fails to provide the syn-anti averaging of the allyl methylene observed at high temperatures for 15. 3 Given the available data, though, it is possible that 11 -rotation is occurring, albeit at a rate considerably slower than is olefin dissociation. Examination of the molecular orbital diagram for 15 provides one explanation for a slow rate of allyl rotation. The frontier MO diagram for a d 0bent metallocene allyl complex is shown in Figure 12a.23 Although rotation of the intact 11 3 -allyl fragment results in no net loss of overlap between the la1 frontier MOs, partial rotation does disrupt the bonding interaction of the lb 1 MO. Figure 12. Frontier molecular orbital diagram for Cp2 M(C3H 5 ) (M=Sc, Y). 160 Investigation of meso-Metallocenes The ansa-scandocene allyl complex AdpSc(C H (16) exhibits markedly different temperature-dependent behavior from that of 15, in part due to the C5 symmetry of themes o- ancillary ligand array, which places the 3- and 3' -Cp ) 3 5 substituents [2-(2-CH3 )-C 1J1 14 = 2-(2-Methyl)-Adamantyl = MeAd] on the same side of the metallocene wedge. Shown in Figure 13 are the variable temperature 1 H NMR spectra of AdpSc(C H in Et 0-d At low temperatures, an AGMPX splitting pattern is exhibited for the allyl ligand, indicating the absence of allyl rearrangement on the NMR timescale. This static structure is supported by the inequivalence of all of the vinylic cyclopentadienyl protons as well as the inequivalence of the methyladamantyl substituents in the low temperature (<194 K) NMR spectra. As the probe temperature is raised, the resonances for the allyl syn hydrogens (H5 1 and H 5 2) broaden and coalesce, as do the resonances for the ) 3 5 . 2 10 allyl anti hydrogens (Hai and Ha2) (~215 K) . Continuing to raise the temperature (246 K) results in the appearance of two new allyl resonances, one located midway between the previous signals for H 5 1 and H 5 2, and the other resonance midway between those of Ha 1 and Ha2; the integrated areas of each of these new resonances are equal to twice those of the individual signals in the AGMPX spectra. This corresponds to averaging of syn and of anti hydrogens (H 51 ~ H 52, Ha1 ~Ha2), with no syn -anti (Ha ~HJ exchange. Accompanying these changes at higher temperatures are averaging of the Cp rings and substituents of the ancillary ligand array. 161 As the temperature of the sample is raised further, the averaged H 5 and the averaged Ha resonances (2.28 ppm and 3.89 ppm, respectively) in the AM2X2 spectra broaden, coalesce, and a new signal appears, located equidistant from the previous resonances. This A~ splitting pattern, where all allyl methylene hydrogens are magnetically equivalent indicates not only rapid syn-syn (anti-anti) exchange, but also concurrent rapid exchange between the syn and anti allyl positions. This progression from AGMPX - AM2X2 - A~ splitting patterns for meso-AdpSc(ri3-C H is also observed for meso-AdpY(r(C H (17), formesoDpSc(r(C3H 5 ) (18) (acquired from treatment of [meso-DpSc(µ-H)h with excess allene), and form es o-IpSc(C3H 5 ) (19), and holds true for all solvents in which spectra have been obtained. ) 3 ) 3 5 excess CH =C=CH 2 2 eq. 2 5 (18) 2 A 500 MHz spectrum of 18 exhibiting an A~ splitting pattern could not be obtained due to temperature limitations of the spectrometer employed for this study; the expected A\ splitting pattern is observed using a 300 MHz spectrometer at temperatures above 358 K. Similarly, the highest temperature spectra obtained for 17 (361 K, Et2 0-d 10 ), and 19 (377 K, toluene-d 8 ) are above the coalescence temperatures of the H5 and Ha resonances, and exhibit very broad features centered at approximately 2.80 ppm and 3.11 ppm (the predicted weighted averages of the chemical shifts for H5 and Ha), respectively. The highest temperature (355 K) at which spectra were recorded for 17 in toluene-d 8 is, as well, above the temperature required for the characteristic intermediate-temperature AM X splitting pattern (298 K), where the allyl resonances observed are approaching, but have not quite reached, syn-anti coalescence. 2 2 162 ,..... 'Cl C'"l ~ 'tj< 'tj< C'"l ~ I.I) C'"l C'"l ~ C'"l ,..... C'"l 'Cl 'tj< N ,..... C'"l N ~ ,..... I.I) N ~ ,..... \ I N 0 I ~ I / i I s! ~ <; I i ' ::_ q !- I.I) I - ~ - j ---'° - ::_ a ~) I I -- ~ ::_- o - - c--.: - \I - Figure 13. Variable temperature 1 H NMR spectra of AdpSc(C 3H 5 ) (16) (500 MHz, Et2 O-d 10 ). 163 All of the spectral differences resulting from temperature changes for 15-19 are reversible; cooling of the NMR samples after obtaining high temperature spectra results in the regeneration of the previously observed lower symmetry species. Decomposition is observed, though, after prolonged heating of solutions of any of these compounds. The low temperature AGMPX splittings exhibited by 16-19 are indicative of nonfluxional allyl ligands on the NMR timescale; the Cssymmetry of them e s a-ligand arrays renders the two syn allyl substituents (Hsi and Hsi) as well as the two allyl anti substituents (Ha 1 and Ha 2) inequivalent. The high temperature AX4 splitting patterns may be explained by n-cr rearrangement pathways analogous to those discussed earlier for the Cp*2 Sc system. As shown in Scheme 2, rotation about the C-C single bond of an 17 1 allyl complex serves to exchange the syn and anti substituents on one of the allyl methylene carbons (Ha 1 and Hs1, as drawn); multiple iterations of the 17 31 17 processes will average all four of the methylene hydrogens. There are two viable mechanistic possibilities, however, that will generate the AM2X2 allyl splitting patterns characteristic of the intermediate temperature range NMR spectra of all of the m es o complexes. Rotation about the M-C single bond of the 17 1 -allyl species (pathway ib ), which serves only to exchange the Cp* rings for 15, has the net effect for the mes a- complexes of not only exchanging the magnetic environments of the two Cp rings, but of also exchanging the environments of the two syn allyl substituents (Hs1 ~Hsi) and exchanging the two an ti substituents (Ha 1 ~Ha 2 ). The observation of syn-anti averaging at higher temperatures is indicative of 1 the presence of an 17 intermediate (pathway ib) . As pathways ia and ib are 1 expected to be facile from an 17 -intermediate, syn-syn (anti-anti) averaging via pathway (ib) is expected to accompany syn-anti exchange. The observed synsyn and anti-anti averaging at lower temperatures than those where syn-anti averaging is seen may, therefore, simply result from the fact that L'ivHsl-Hs2 and 6 vHsl-Hs2 are both less than 6VHsl-Hal and 6VHs 2-Ha 2• An alternate mechanism, however, may explain the existence of spectra which indicate syn-syn (and anti-anti) averaging but not syn-anti 3 averaging. The in-plane rotation of an intact 17 -allyl complex (Scheme 3) 164 Net exchange of Cp rings Exchange of Cp rings synf2syn and ant~ anti exchange synf2anti on one methylene carbo1 Scheme 2. 17 3 -17 1 Allyl rearrangements in mes o-rnetallocenes. 165 serves to exchange the two syn hydrogens and to exchange the two anti hydrogens, as well as interchange the magnetic environments of the two Cp rings; this is identical to the net result of pathway (ia) shown in Scheme 2. The observed NMR spectra, therefore, could also be explained by the existence of two distinct rearrangement pathways, with both processes slow on the NMR timescale at low temperatures (resulting in AGMPX splitting patterns),11 3 -rotation (providing Cp-Cp, syn-syn, and anti-anti averaging) rapid 3 1 at intermediate temperatures, and the 11 -11 processes (ia) and (ib) (syn-an ti exchange and the observed AX4 allyl patterns) becoming rapid only at the higher temperatures investigated. 180° 173 - allyl rotation rate= k2 Scheme 3. 17 3-Rotation form es o-metallocenes. Distinguishing between these possibilities in the absence of lineshape analysis proved exceptionally difficult. The temperature range over which the Cp substituents coalesce was quite narrow; Eyring plots relying on these data points often showed considerable scatter and deviations from linearity, and, unsurprisingly, resulted in chemically nonsensical activation parameters. The utility of observing Cp or allyl coalescence temperatures on spectrometers of different field strength (300 vs. 500 MHz) was negated by the fact that the change in Tc was comparable to the uncertainty associated with the temperature measurement itself. As averaging of all four methylene 3 1 hydrogens requires multiple iterations of the 17 -17 process, the fact that equation I applies rigorously only to two-site exchange processes dictates that I cannot be used to determine k1 at the observed syn-anti Tc Interpretation of the low temperature coalescences is further complicated by the fact that averaging of the ligand resonances may result from 113 -rotation, and the syn- 166 3 1 anti averaging from 17 -17 processes, a situation which can not be substantiated using only Equation I and Erying plots derived from its application. Lineshape analysis, however, proved invaluable in distinguishing between the two mechanistic possibilities outlined in Schemes 2 and 3. Simulation of themes o-scandocenes and yttrocene allyl complexes using only the 17 3 -17 1 exchange process utilized in previous simulations of 15 Jailed to successfully model the entire temperature range investigated experimentally, 3 1 indicating that the low temperature 17 -17 pathway observed for 15 is n at the low energy allyl rearrangement pathway form es a complexes 16-19. Representative NMR simulations generated using only the 17 3 -17 1 exchange processes are shown in Figure 14 for 16. Although the observed low and high temperature spectra can be adequately simulated using only the 17 3 -17 1 mechanism, the characteristic AM2X2 splittings observed for the allyl ligands at intermediate temperatures can not be obtained solely by varying the rate of allyl double bond dissociation from the metal. Furthermore, Eyring plots obtained from the experimentally determined coalescence temperatures with the corresponding best-fitting rate constants (as determined using gNMR) were often plagued with the same problems associated with (previous) Eyring plots generated solely from the application of equation I. Similarly, simulation of the VT NMR spectra using only the 173rotation mechanism did not mimic the experimental spectra over the entire temperature range investigated. Implementation of the 173-rotation process as the sole rearrangement pathway successfully models both the low and intermediate temperature spectra, yet fails to generate the A~ splitting pattern observed experimentally at elevated temperatures, as shown for 16 in Figure 15. 3 The results of simulations using concurrent 17 -rotation and the 17 3-17 1 pathways shown in Scheme 2 successfully mimic the observed spectra over the entire temperature range investigated. The simulated spectra for 16 using • . l1er b arner • 173 -17 1 rearrangement a 1ow-energy 17 3 -rotation process an d h 1g 167 N ::r:: C0 0 ,--I N >< ::r:: \.0 C0 0 C"! N II rl ~ ,--I >< \.0 0 ,--I N II ::r:: rl ~ 0 0 N II ::r:: [',. rl ~ OC) 'tj< ,--I II rl ~ N ::r:: 0 II 0.. u rl ~ 0.. u 0 NO 0.. u Q.J .s f; Ei ::r:: N ::r:: If) 0 ,--I >< N ~ ,--I II rl ~ N ::r:: 'tj< 0 ,--I N >< ::r:: OC) [',. ,--I II rl ~ C0 0 ,--I >< OC) OC) OC) II rl ~ Lf:o I'- 0 Figure 14. gNMR simulation of (C 3H 5 ) and vinylic Cp resonances of 16 (500 MHz, Et20-d 10 ) using only ri 3-17 1 processes (ia) and (ib ). E 0. 0. 168 .::: >- N l:: <Fl l:: L!") \0 L!") N II l:: N \0 \0 ~ N II ~ (1j N ~ l:: N l:: L!") ['.. rl II N ~ N l:: 0 t---: ('i") II N ~ N l:: 0 II 0... u N ~ 0... u E 0. 0. 0... u l:: N l:: ['.. 0 rl >< rl ['.. rl II N ~ N l:: L!") 0 rl >< ('i") rl t---: N l:: ('i") 0 rl II >< N ~ N N C'l II N ~ Ll?o r----o Figure 15. gNMR simulation of (C 3H 5 ) and vinylic Cp resonances of 16 (500 MHz, Et20-d 10 ) using only ri 3 -rotation. "T1 ~ Ul C/l N~ ,..... Pl 9C - ,.... • wn Pl _:j '"d :::; < ....... p_. :::; p_. ~ 0 ,..... ....... Er 0 ., I (;J _:j O"I ~ >-n 0 C/l ('[) n :::; Pl 0 ;:::l -Pl ;:::l ,ct ('[) C/l CJ"' n '"d .___,, ....... ,-.._ ~ 5.i. ....... C/l ,-.., n- ('[) '-< C/l C/l ~ 0 >-t % Vl::r:: ...... .___,, >-n 0 ,-.._ ....... :::; CIQ :::; C/l 'c' 5· 0 ,_.~ I 0 -- l'::J z~ N ::r:: ~CIQ o· O"I ,-.._ 0 i::::: >-t rt> aq" 1 -- -- -- - - Cp Cp 207 K: k2 = 260 Hz k 1 = 6.47 Hz 215K: k2 =550Hz k1 = 15.4 Hz 201 K: k2 = 172 Hz k1 = 3.24 Hz - Hsyn 197 K: k2 = 35.0 Hz k 1 = 2.00 Hz Hanti Hsyn+anti ppm ~ - - - - - - ~ - - ~ - - - low temp limit: k2 = 0 Hz . ., r. k1 = 0 Hz 246 K: k2 = 1.96 x 104 Hz -4-------' k1 = 259 Hz 290 K: k2 = 7.13 x 105 Hz k = 5.20 x 103 Hz 1 344 K: k 2 = 1.71 x 10 7 Hz k = 7.42 x 104 Hz methine Cp r---' \!) °' 170 mechanisms are shown in Figure 16. Only when ri3-rotation is fast relative to 3 1 the 11 -11 processes, however, do the simulated spectra fit the observed spectra 1 in all temperature regimes; making the 113-11 processes the low energy rearrangement pathways fails to generate spectra containing sharp AM2 X2 allyl splitting patterns. It is worth noting that the use of spectral simulation is intended primarily as a means of judging whether the observed spectral trends could be explained qualitatively by a single rearrangement mechanism. Rate constants and activation parameters are obtained from these simulations, but caution must be exercised when interpreting these numbers, especially when extrapolating rate constants over large temperature ranges. The temperature where the allyl syn-anti coalescence occurs (i.e. where the AM2X2 and AX4 splitting patterns are averaged) is the key temperature at which to obtain information on the rate of olefin dissociation. As ri3-rotation does not provide a means of affecting syn-anti exchange, this coalescence is solely a result of pathway (ib). Results of optimization of the rates of olefin dissociation for the various mes a-complexes, as well as the associated LlG:t: s calculated using equation II are given in Table 3. Activation barriers to olefin dissociation for :j: themes o- complexes are all on the order of 14 kcal/mol. These LlG values are each valid only at the Tes indicated, which span a roughly 80 °C temperature range for the entire series of metallocene / solvent pairs . :j: LlG ri3-ril,Tc Cmpd solvent Tc (K) (kcal/mol) 16 EtiO-d 10 290 12.0 17 Et2O-d 10 361 15.6 17 toluene-d 8 370 16.0 18 toluene-d 8 327 13.4 19 toluene-d 8 362 14.9 Table 3. Olefin dissociation barriers form es o- metallocenes 171 When pathways (ia) and (ib) are slow (as was determined by lineshape 3 analysis for 16-19), 11 -rotation may be treated as a two-site exchange mechanism, and thus evaluated using equations I and II. Activation energies 3 for 11 -rotation, determined using this approximation, for the mes ometallocenes are given in Table 4. t 6 G ri3-rot,Tc Cmpd solvent Tc (K) (kcal/ mol) Eqns I, II 16 EtzO-d 10 215 9.7 (± 0.6) 17 Et2 O-d 10 208 9.9 (± 0.7) 17 toluene-d 8 207 10.6 18 toluene-d 8 246 12.2 (± 1.6) 19 toluene-d 8 259 12.2 (± 0.9) Table 4. Activation barriers to 113-rotation activation, obtained from two-site exchange approximation (I) Extrapolation of Activation Barriers for Complexes 16-19 Complicating the direct comparison of olefin dissociation barriers is the fact that the 6Ht and 6St may not both be independently obtained from the observed NMR spectra. Rate constants for olefin dissociation may then be extrapolated to various temperatures using Equation III (Table 5). Based primarily on the value obtained for 15 in toluene-d 8 , a 6St of -10 e.u. is utilized for these calculations. t t k = T exp [-(6H /RT) + 6S /R + 23.76] (ill) . . . th e approximate YJ 3 -11 1 rates, t h e rate o f ri 3-rotation may Incorporatmg then be optimized in order to best approximate the observed NMR spectra at several temperatures, generally those where distinct coalescences could be observed. These optimizations were performed with the Tj 3-ri 1 process operating at a rate determined by application of Equation III. After a series of k 2 had been evaluated, Eyring plots were generated for each of the 172 metallocene/ solvent pairs; the activation parameters resulting from these plots are given in Table 6. t L1H 113-111 (kcal/mol) L1G 113-111,298 K t Cmpd solvent t 6S 113-TJ 1 (e.u.) 16 Et2O-d10 -10 9.1 12.1 8.2 X 103 17 Et2O-d10 -10 12.0 14.9 6.8 X 10 1 17 toluene-d 8 -10 12.3 15.3 3.9 X 10 1 18 toluene-d 8 -10 10.1 13.1 1.6 X 103 19 toluene-d 8 -10 11.3 14.3 2.1 X 10 2 1 k1, 298 K (sec- ) (kcal/mol) Table 5. Activation parameters for olefin dissociation, and 298 K rate constants obtained from extrapolation of estimated 6Ht, 6St Cmpd solvent t ,15 113-rot (e.u.) t L1H 113-rot (kcal/mol) t (kcal/mol) 16 Et 2O-d 10 6.4 11 9.6 17 Et2O-d10 -24 5.1 10.1 17 toluene-d 8 -20 6.4 11.6 18 toluene-d 8 -13 9 12.2 19 toluene-d 8 9.9 14.5 11.9 L1G 113-rot,Tc Table 6. Activation parameters for in-plane rotation for mesa- metallocenes obtained from Eyring plots of low temperature coalescence data Despite the small temperature range over which ligand coalescences were observed, the considerable scatter observed in the Eyring plots, and the number of approximations required to arrive at the final activation free energies, comparison of the 6G\ 3_ rot resulting from extrapolation using the gNMR-derived 6H\ 3_rot and ,15\ 3-rot show a surprisingly good correlation with (are within experimental error of) those obtained from application of equations I and II. As was performed for olefin dissociation, the evaluation of 6H\ 3 _rot and 6St113 _rot permits extrapolation of 6Gt113 _rot and k 2 for the various 173 metallocenes to a common temperature (Table 7), although these values are derived ultimately from estimated values of ~Ht 173_171 and ~St 173_171. t ~ G 173-rot,298 K k2 298 K (sec-1) Cmpd solvent 16 Et2O-d 10 9.1 17 E~O-d 10 12.3 17 toluene-d 8 12.4 18 toluene-d 8 12.9 5.35 X 103 2.25 X 103 19 toluene-d 8 11.5 2.10 X 104 (kcal/mol) 1.33 X 106 6.42 X 103 Table 7. In-plane rotation barriers for 16 - 19, extrapolated to 298 K Regardless of the individual ~Ht and ~St values, judging only by qualitative assessment of the VT spectral trends, large differences exist between 15 and themes o metallocenes 16-19. In the former, olefin dissociation is the lowest energy rearrangement pathway, whereas 17 3-rotation 1 is fast relative to 173-11 (which becomes prominent at higher temperatures) for the latter. This may be explained by the relative congestion in the metallocene wedge imparted by the C5 Me5 rings relative to the lesssubstituted m es o- complexes, whose metal centers may additionally be more open as a result of "tying back" the Cp rings with the dimethylsilyl linking units. Regarding the relative rates of 17 3-rotation and 173-11 1 for them e s acomplexes 16-19, qualitative assessment of the gNMR simulations indicates 3 that the rate of 11 -rotation (k2) is generally (usually : :"'. 1 order of magnitude) greater than olefin dissociation (k1). Extrapolations of the calculated ~Gt to common temperatures (298 K, Table 2) are for the most part consistent with this observation. The one exception to this trend is 18 (toluene-d 8); the rate of 11 3-rotation still exceeds that of 173-111, but only by roughly a factor of two, as determined by extrapolation of the g-NMR derived activation parameters . Caution must be exercised when comparing the 298 K values of k1 and ~G\ 3_111' as these are extrapolations based on estimat ed values of 3_111 . Nonetheless, a pronounced increase in the rate of olefin dissociation (2 orders ~s\ 174 of magnitude) is observed when switching from Adp Y to the slightly more congested AdpSc system. The rate of olefin dissociation also appears largely independent of solvent at room temperature, with 17 exhibiting comparable 1 17 3 -17 activation barriers in both diethyl ether and toluene. 3 The rate of 17 -rotation appears largely independent of solvent (17, toluene vs. Et2 O), but large changes in k1 (3 orders of magnitude) are observed when changing the transition metal (16 vs. 17, Et2 O), and smaller changes in rate are associated with variation of the Cp substituents (18 vs. 19). Some of the changes in rate may be easily rationalized, such as a decreased barrier to rotation when switching to a ligand with smaller alkyl substituents in the open portion of the metallocene wedge (3 and 3' substituents) (19 vs. 18); other trends are more puzzling (such as the large decrease in the rate of 17 3rotation when switching from yttrium to the slightly smaller scandium, or the smaller decrease in rate when 3,3' substituents are changed from MeAd to CMe3 ). As mentioned earlier, allyl C=C double bond dissociation was modeled as occurring with equal probability from each side of the metallocene wedge. Although this is a reasonable assumption for the Cp*2 ligand array, its validity in the cases of themes o- complexes is less certain, as the C5 symmetry of the ligand array makes dissociation from the two sides of the wedge energetically inequivalent. The 17 3 -17 1 process was modeled assuming olefin dissociation from exclusively one side of the metallocene wedge, in order to determine whether such a preference might be observed experimentally. Although there is undoubtedly some lateral preference for dissociation, the fact that 173 -rotation, which provides syn-syn and anti-anti exchange, was determined to occur rapidly with respect to C=C dissociation leads to the observation that dissociation from one side of the metallocene wedge is, using the available NMR data, indistinguishable from dissociation from both sides of the wedge. Modeling C=C dissociation from one side of the wedge in the absence of 173 -rotation provided spectra that bore little likeness to any of the experimentally observed spectra. 3 The clockwise and counterclockwise rotations of the 17 -allyl ligand pass through energetically inequivalent transition states in the framework of a me so- ligand array (Figure 17). The 173-rotation LiG:t:s given in Tables 6 and 7, therefore, technically reflect only the pathway with the lower energy 175 activation barrier. Not enough is currently known about the precise steric interactions between the allyl ligand and the 3,3' metallocene alkyl substituents in order to determine which of the two possible transition states is of lower energy. The availability of two possible rotation pathways does not significantly affect the analyses presented above, as the two pathways are experimentally indistinguishable for mes o- complexes. Figure 17. Possible 11 3-rotation transition states for mes o-metallocenes. Lineshape Analyses of racemo -Metallocenes 20-21 Spectral simulations were undertaken in order to try to understand the rather complicated variable temperature NMR spectra exhibited by two d 0 scandocenes whose ancillary ligand arrays exhibit C2-symmetry: rac-IpSc(C3H 5) (20), and the enantiomerically pure (R,S)-BnBpSc(C3H 5 ) (21) (the (R) refers to 176 Me • -g;;;si Sc-v Me3 Si (20) (21) rac-lpSc(ri3-C3Hs) (R,S)-BnBpSc(ri3-C3 H 5) the chirality of the binaphthol moiety, and the (S) to the configuration at the metal center).24 The C2 - symmetries of the ligand arrays lead to pronounced differences in the observed NMR spectra, compared to either 15 or 16-19. The NMR spectra of 20 and 21 at low temperatures show AGMPX allyl splitting patterns and inequivalent Cp rings and substituents (Figure 18), consistent with static allyl moieties. As the probe temperature is raised, the Cp rings and substituents are rendered equivalent. As the temperature is increased further, there is a simultaneous broadening of all four of the allyl methylene hydrogen resonances. After a wide temperature range in which the allyl methylene hydrogens are broadened out into the baseline, a single resonance appears at a chemical shift equal to the weighted average of the four previously inequivalent methylene hydrogen signals (an AX4 splitting pattern) with no sign of the intermediate AM2X2 spectra previously observed for the mes o- complexes. This behavior may be explained, in part, by observing the results of the 1 3 17 -rotation and 17 3 -17 rearrangement processes in a C2 -symmetric ligand framework. A single iteration of the in-plane 17 3 -rotation mechanism (Scheme 4) provides Cp-Cp exchange simply by a change in directionality of the allyl methine carbon. Inspection of the allyl ligand before and after the transformation shows no change in the magnetic environments of the individual methylene hydrogens; the fact that 17 3 -rotations affect no change in the environments of the allyl methylene hydrogens results in an AGMPX allyl splitting pattern, regardless of the rate of 17 3-rotation. 'Tl ..... ...... N 8 CJl 0 0 ,-... ------ 0 N ,-... ------ U1 ::r:: (.;J n (') (fl '"d ....... ('") I :;:::i "' ....... 0 Pl (') ...... '"-1 (1) '"d IJl ~Al .... ::r:: '"-1 (1) 2" (1) '"-1 p.) '"d ~ • (1) ,...,. (1) cr" ........ s;· '"1 p.) < (1) I :;:::,... ::::i (1) 2" 0 ' ~ ~ ::r:: ~ ro '"1 i:: OQ II s I a I C s - -- - - _J l ' j' - J J 22s 3.0 25 PPM 1111111111111111111111 1 1111 1 11 1 1 1 11 1 111 1 1 1 111 1 1111111 1 11 11 1 3.5 . - -- ~ U J,_ b J u uL235 __ A_____ _)V~-- _flV(__jUC 193\___207 JiJ Ii _J_ , 1 -__ ___AA 4.0 ~ - - · ..JL - ----"' ---- - · -~ ---- . _ _____ ____l__l ___ _1_ ___ j __325 _/\___ __ _j \ _ -~1272_, 290 _A__J- -''-301 _ __ __L ___~ L~ 310 ------- - - -- j__ __ __ J _ ___ _ __ __L _J_ __ L __343 U 'L ' 25 404 2 359 - -- l~ _j___J___393 ------- . ---- ______ _l__L__j_ __378 I;I~_):~Aj-1~~' _JL _ jl - - fl_ - ---- ~ '"~ _J\ _J\___J - l_ __ __ AA 45 a H,,,..C~_Sc . . _, ' ✓/✓ C, H _c~ Me••·Si~ Me~~ A½V H-,-H racemo-Me2Si[2,4-( CHMe2)i-C 5H 2] 2Sc( C 3H 5) Variable Temperature 500 MHz 1H NMR Spectra, toluene-d8 'l I-' '-;J 178 180° 173 - allyl rotation rate= k2 ~R' I H s2 "- C c--/-1-~--c ✓/ I M H s1 \ R~H"' R Scheme 4. 180° In-plane rotation for C2-symmetric ansa-metallocenes. The results of olefin dissociation followed by rotation about the allyl CC single bond for 20 and 21, however, are quite different from the systems investigated previously; a non-pairwise exchange of hydrogens (i.e. Ha 1 --- H 52; H a2 --- Ha 1; H 51 --- H a2; H 52 --- H 51) is observed for the racemo- complexes (Scheme 5). Olefin dissociation followed by rotation about the M-C single bond has the same net effect as in-plane 17 3 -rotation; for the race ma-complexes 20 and 21, this is manifested by exchange of the Cp rings and no net effect on the allyl resonances. Again, spectral simulation must be used to obtain information on the rate of olefin dissociation, which exchanges allyl substituents by non-two-site exchange mechanisms. A stackplot of the gNMR-derived spectra for 20 using only the 17 3 -17 1 processes is shown in Figure 19. The values of ~G+ and k1 obtained from lineshafe analysis similar to that performed on 15-19 are given in Table 8. As the ~G values are valid only at the coalescence temperatures indicated, direct comparison of the rac-Ip and BnBp systems is not possible. Consistent with the results obtained for 15 and 17, the nearly identical coalescence temperatures and ~G+ s in THF and toluene indicates a lack of solvent dependence on the rate of olefin dissociation. 179 Net exchange of Cp rings No effect on allyl resonances Exchange of Cp rings Hal- Hs2 Hsl- Ha2 Ha2-Ha1 Hs2- Hsl Scheme 5. Implications of olefin dissociation and subsequent M-C and C-C single bond rotations for C2-syrnrnetric rnetallocenes. 180 Lf) t0 t0 0 0 ,-< ,-< ,-< >< >< tj1 0 ,-< 0 >< >< '-1'1 ,-< '-1'1 ,-< 00 N ~ \0 II ,...... II ,...... II ,...... II ,...... ~ ~ ~ ~ 0 0 00 ·.c c::ro II ,...... ~ + 0 0 ,-< II ,...... c::>, V) ~ :::c 0 00 II ,...... I ~ I I 00 N II ,...... l ~ 0 ) ~ ( ' N :::c II ,...... ~ " ./ 0... u (l) c:: ~(l) 8 :::c I ' l 1 ..,/ 0 NO -..., ' 1 ' ·~ 1 ---... ___, = ' 1 -=<-.._ _,, -- l J J J J ~ Figure 19. gNMR simulation of Cp and (C 3H 5 ) resonances of 20 (500 MHz, toluene-d8 ) using 11 3 -11 1 processes (ia) and (ib ). LI: 0 I"-- 0 E 0.. 0.. 181 Compound Solvent 20 t Tc (K) k1 Tc (sec-1) ~G n 3-nl Tc (kcal/mol) toluene-d 8 359 8400 14.7 20 THF-d 8 350 5200 14.6 21 toluene-d 8 330 6400 13.6 21 THF-d 8 335 6800 13.8 Table 8. Olefin dissociation activation energies for racemo- metallocenes 3 1 Although it is clear that the 17 -17 processes are operating for 20 and 21 at higher temperatures, it is not possible to determine whether this process is responsible for the averaging of the ligand resonances at low temperatures (as was seen for 15), or whether in-plane rotation is operating as the low energy rearrangement pathway (as was observed for themes o- metallocenes) . The errors associated with extrapolating rate constants over considerably large temperature ranges precludes the identification of the rearrangement mechanism operating at low temperatures. Conclusions The variable temperature NMR spectra of d 0 -metallocene allyl 3 complexes 15-21 indicate static 17 -allyl ligands at low temperatures. Common 1 to 15-21 is olefin dissociation to generate transient 17 -allyl species. Lineshape analysis of the observed spectra provides activation barriers to olefin dissociation of ~10-14 kcal/mol for all of these complexes, although the coalescence temperatures at which the ~Gt were obtained span a 140 degree temperature range. The symmetry constraints imposed by rigid meso-metallocenes 16-19 permits confirmation of a second type of allyl rearrangement: in-plane rotation of the intact 17 3 -allyl moiety. Although conclusive evidence of the existence of the latter process could not be obtained for racemo-complexes 2021, in-plane rotation was found, qualitatively, to be fast (rate constants 182 generally ~1 order of magnitude larger) relative to olefin dissociation for 1619. Rate constants derived from the activation parameters obtained for olefin dissociation and 11 3 -allyl rotation for 16-21 are rigorously valid only at the observed coalescence temperatures; this fact, coupled with the inability to independently determine .1H+ and .1S+, precludes a quantitative comparison of rate constants for the various processes and metallocene / solvent pairs at a common temperature, especially when the comparisons require extrapolation over large temperature ranges. Experimental Section General Considerations: All air and / or moisture sensitive compounds were manipulated using standard high vacuum line techniques or in a dry box under a nitrogen atmosphere, as described previously.25 Argon and hydrogen gases were purified and dried by passage over columns of MnO on vermiculite and activated molecular sieves. Solvents were stored under vacuum over titanocene26 or sodium benzophenone ketyl. Allene was purchased from Aldrich and used as received. Cp*2 Sc(17 3-C3H 5 ),27 mesoAdpSc(11 3-C3H 5 ), meso-AdpY(11 3 -C3H 5 ),28 and [m es o-DpScHJi,29 were prepared according to published procedures. mes o-IpSc(11 3 -C3H 5 ) and rnc-IpSc(11 3C3H 5 )21 were synthesized and characterized by Jeffrey C. Yoder. Cyrille Loeber was responsible for the synthesis of (R,S)-BnBpSc(11 3 -C3H 5 )20_ Unless otherwise specified, all NMR spectra were recorded on a Bruker 1 AMS00 (500.13 MHz for H) spectrometer. Lineshape analyses were performed on a 7300/180 Power Macintosh using gNMR (Cherwell Scientific Publishing, 1992-1996) version 3.6 for Macintosh. Preparation of meso-Me2Si(3-CMe3-C5H3hSd11 3-C3H5) [DpSd11 3-C3H5)] (15). A J-Young NMR tube was charged with approximately 15 mg (0.0218 mmol) of [DpScH]2 and 0.4 ml C6D6. The contents of the tube were cooled to -78 °C, and volatiles were removed on a high vacuum line. Allene (0.111 mmol, 5.1 eq) was transferred from a lecture bottle into the NMR tube at -78 °C via a 42.5 ml gas bulb, the NMR tube sealed, and the contents of the tube w as brought to 183 room temperature. After heating for 6 hours at 80 °C, 1 H NMR spectroscopy indicated that all of the [DpScH] had been consumed. Volatiles were removed from the NMR tube in vacuo, and the residual yellow solid was dried under dynamic vacuum for 30 minutes. 1 H NMR (25 "C, C6D6): o7.50 (HITu m, J=12.3 Hz, lH), 6.75 (C5H3, t, J=2.5 Hz, 2H), 5.96 (C5H3, br, 2H), 5.29 (C5H3, t, J=2.6 Hz, 2H), 4.33 (Ha, br, 2H), 2.25 (H5 , br, 2H), 1.08 (CMe3, s, 18H), 0.70 (SiMe:z, s, 3H), 0.22 (SiMe:z, s, 3H). Lineshape Simulations Spectral simulations using gNMR version 3.6 are based on general NMR theory.30-32 Dynamic spectra are determined using a standard Liouville representation of quantum dynamics, as described by Binsch.33,34 Lorentzian lineshapes with HWHM = 5.00 Hz were utilized, unless otherwise noted. Control experiments showed that changes to the linewidth did not visibly affect the rate of a particular exchange required to produce a given coalescence. As systems with more than 12 exchanging nuclei may not be simulated using gNMR v.3.6, complicated systems such as 16-22 are generally broken down into two separate simulations: that of the allyl ligand and vinylic cyclopentadienyl fragments, and that of the silyl- and alkyl- substituents on the Cp rings. Frameworks for 16-22 were drawn on C.S. Chemdraw Pro v.3.5.1 (CambridgeSoft Corporation, 1995) and then cut and pasted into gNMR. The experimentally determined low temperature (slow exchange) chemical shifts and coupling constants were then entered for the various metallocene complexes. Exchange permutations then determined for each complex and then entered into the appropriate "Exchange" subroutine. Simulated spectra were obtained by entering the appropriate rates into the individual permutation routines and then using the "Calculate" function. Stackplots were generated by cutting and pasting a series of individual optimized spectra from gNMR into a single MacDraw Pro l.5vl file; this transfer causes a reduction in the resolution of the simulated spectra of a factor of 4. For each combination of ligand array and solvent, 1 H NMR spectra were evaluated for 15-22 using the ri 3-ri 1 processes (rate= k1 ) as the sole exchange mechanisms. The rate constant k1 was evaluated as the rate of 184 3 olefin dissociation from an 17 -complex. Contributing to k1 were processes involving rotations about both the resultant allyl C-C and the M-C single bonds. Each of these two types of rotations has the possibility of resulting in a productive rearrangement (one in which there is a net change in the allyl or Cp magnetic environments), or a nonproductive rearrangement (where there are no net changes in the allyl or Cp magnetic environments). Consequently, k 1 is the sum of the rates of four distinct exchange processes: generation of an 1 17 -intermediate and (1) a "productive" rotation about the C-C single bond; (2) a "non-productive" rotation about the C-C single bond; (3) a "productive" rotation about the M-C single bond; (4) a "non-productive" rotation about the M-C single bond. To simplify the lineshape analyses, only "productive" rearrangements were utilized to generate the simulated spectra; k1 was adjusted to take into account "nonproductive" rearrangements, which were modeled as occurring with equal likelihood as the "productive" rearrangements. Unless otherwise noted, generation of 17 1 -intermediates were modeled as occurring with equal probability from both sides of the metallocene wedge; rotation about the M-C single bonds of these 17 1 -intermediates result in identical exchanges of allyl and cyclopentadienyl hydrogens. In order to further simplify calculations, only a single M-C rotation permutation was utilized, with appropriate statistical factors and included in the reported value of k1 . Control experiments were performed in which both "productive" and "nonproductive" rearrangements were explicitly simulated; excluding the identity operation(s) ("nonproductive" rearrangements), when appropriate, resulted in no observable change to the NMR spectrum for a given value of k 1 when the appropriate correction factor is applied to the remaining contributor to k1 . Similarly, control experiments showed no visible change in the value of k1 required to generate a given coalescence when utilizing either one or two distinct processes to simulate the two possible M-C single bond rotations. The variable temperature NMR spectra for 16-22 were simulated using the following procedure: (1) k1 was optimized to best simulate the high temperature (syn-anti) 3 coalescence, with k2 (the rate of 17 -rotation) set temporarily equal to 0. 185 (2) 6H+ 11 3 - ril was evaluated using equations (II) and (III), with 65+ 17 3 _171 = -10 e.u. (3) k1 was calculated at several temperatures using the 6H+ 11 3 _11 1 and 65+ 17 3 _ 11 1 from step (2) (4) k2 was optimized to best fit the observed spectra at the temperatures used in step (3); these optimizations of k2 were performed with concurrent olefin dissociation, using the k1 rates determined in step (3) . (5) Activation parameters for 173 -rotation were derived from Eyring plots of the k2 values from step (4) (6) An approximate rate of 17 3 -rotation was calculated for the high temperature (syn-anti) coalescence using the activation enthalpy and entropy values derived in step (5) (7) k1 values were reoptimized to simulate the observed spectra, this time including concurrent 17 3 -rotation (k2 rates from steps (4) and (6)) References and Notes 1. Bachmann, M. J. Chem . Soc. Dalton. Tran s 1996, 255. 2. Brintzinger, H. H.; Fischer, D.; Mulhaupt, R.; Rieger, B.; Waymouth, R. N . Angew. Chem. Int. Ed. Eng. 1995, 34, 1143. 3. Casey, C. P.; Hallenbeck, S. L.; Pollock, D. W.; Landis, C.R. J. Am. Chem . Soc. 1995, 117, 9770. 4. Casey, C. P.; Hallenbeck, S. L.; Wright, J. M.; Landis, C. R. J. Am. Chem. Soc. 1997, 119, 9680. 5. Wu, Z.; Jordan, R. F.; Petersen, J. L. J. Am. Chem. Soc. 1995, 117, 5867. 6. Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. Principles and Applications of Organotransition Metal Chemistry; University Science Books: Mill Valley, CA, 1987, p . 175. 7. Jolly, P. W.; Mynatt, R. Adv. Organomet. Chem. 1981, 19, 257. 186 8. Vrieze, K., in Dynamic Nuclear Magnetic Resonance Spectroscopy; Jackman, L. and Cotton, F. A., Eds.; Academic Press: New York, 1975, p. 441. 9. Kreiger, J. K.; Deutch, J. M.; Whitesides, G. M. Inorg. 1535. 10. Benn, R.; Rufinska, A.; Schroth, G. J. Organomet. Chem 1981, 217, 91. 11. Becconsall, J. K.; Job, B. E.; O'Brien, S. J. Chem. Soc., A 1967, 423. 12. Faller, J. W.; Chen, C.-C.; Mattina, M. J.; Jakubowski, A. J. Organomet. Chem. 1973, 52,361. 13. Evans, W. J.; Ulibarri, T. A.; Ziller, J. W. J. Am. Chem. Soc. 1990, 112, 2314. 14. Landis, C.R.; Cleveland, T.; Casey, C. P. Inorg. 15. Tjaden, E. B.; Casty, G. L.; Stryker, J.M. J. Am. Chem. Soc. 1993, 115, 9814. 16. Pastor, A.; Kiely, A. F.; Henling, L. M.; Day, M. W.; Bercaw, J. E. J. Organomet. Chem. 1997, 528, 65. 17. Antonelli, D. M.; Tjaden, E. B.; Stryker, J. M. Organometallics 1994, 13, 763. 18. Jeske, G.; Lauke, H.; Mauermann, H.; Swepston, P. N.; Schumann, H.; Marks, T. J. J. Am. Chem. Soc. 1985, 107, 8091. 19. Tsutsui, M.; Ely, N. J. Am. Chem. Soc. 1975, 97, 3551. 20. Abrams, M. B.; Yoder, J.C.; Loeber, C.; Day, M. W.; Bercaw, J. E., Manuscript in preparation. 21. Yoder, J.C.; Day, M. W.; Bercaw, J.E. Manuscript 22. Lineshape analysis was perfomed using gNMR Version 3.6 for Macintosh as described in the Experimental Section and in Appendix B. 23. Lauher, J. W.; Hoffmann, R. J. Am. Chem. Soc. 1976, 98, 1729. 24. Schlogl, K. Top . Stereochem . 1967, 1, 39. Chem. 1973, 12, Chem. 1995, 34, 1285. m preparation,. 187 25. Burger, B. J.; Bercaw, J. E. New Developments in the Synthesis, Manipulation, and Characterization of Organometallic Compounds, 1987; 357, . 26. Marvich, R. H.; Brintzinger, H. H.J. Am. Chem. Soc. 1971, 93, 2046. 27. Thompson, M. E.; Baxter, S. M.; Bulls, A. R.; Burger, B. J.; Nolan, M. C.; Santarsiero, B. D.; Schaefer, W. P.; Bercaw, J. E. J. Am. Chem. Soc. 1987, 109, 203. 28. Abrams, M. B., Ph.D. Thesis, California Institute of Technology, 1998. 29. Bunel, E. E., Ph.D. Thesis, California Institute of Technology, 1989. 30. Pople, J. A.; Schneider, W. G.; Bernstein, H. J. High Resolution Magnetic Resonance; McGraw-Hill: New York, 1959, p. 1. 31. Binsch, G., in Dynamic Nuclear Magnetic Resonance Spectroscopy; Jackman, L. M. and Cotton, F. A., Eds.; Academic Press: London, 1975, p. 45. 32. Steigel, A., in NMR, Basic Principles and Progress; Diehl, P., Fluck, E. and Kosfeld, R., Eds.; Springer-Verlag: Berlin, 1978; Vol. 15, p. 1. 33. Binsch, G. J. Am. Chem. Soc. 1969, 91, 1304. 34. Binsch, G. J. Magn. Res. 1978, 30, 625. , Nuclear 188 1 Appendix C: Variable Temperature H NMR Data for 15-18 1 Table 1. Variable temperature 300 and 500 MHz H NMR data for Cp*?Sc(CiHc;) (15) Solvent Temp (K) 8 (ppm) Assignment JH-H (Hz) toluene-d8 7.06 (quintet, lH) 12.9 326 H 2 CCHCH2 2.89 (s, 4H) H 2 CCHCH 2 1.85 (s, 30H) C,Mec; 7.15 (m, lH) 194 14.8 H 2 CCHCH2 3.93 (d, 2H) H 2 CCHCH 2 , anti 2.05 (m, 2H) H 2 CCHCH 2 , syn 1.86 (s, 15H) C5 Me5 1.73 (s, 15H) Cc;Mec; Et2 O-d 10 (300 MHz) 333 6.97 (quintet, lH) 12.6 H 2 CCHCH2 (avg) 2.74 (d, 4H) 12.0 H 2 CCHCH 2 1.90 (s, 30 H) Cc;Mec; 6.97 (m, lH) 210 H 2 CCHCH2 3.62 (d, 2H) 15.6 H 2 CCHCH 2 , anti 1.88 (d, 2H) 4.6 H 2 CCHCH 2 , syn 1.93 (s, 15H) C5 Me5 1.83 (s, 15H) Cc;Mes 189 1 Table 2. Variable temperature 500 MHz H NMR data for meso-AdpSc(C3H 5 ) (16) Temp (K) 8 (ppm) Solvent JH-H (Hz) Assignment Et2 0-d 10 181 246 7.30 (m, lH) 6.74 (lH) 6.70 (lH) 6.18 (lH) 5.92 (lH) 5.52 (lH) 5.49 (lH) 16 4.16 (d, lH) 15 3.64 (d, lH) 11 2.61 (d, lH) 9 2.41 (d, lH) 2.27, 2.16, 2.08, 1.88, 1.68, 1.57, 1.47, 1.32 (m, 28H) 0.95 (s, 3H) 0.84 (s, 3H) 0.74 (s, 3H) 0.25 (s. 3H) 7.22 (m, lH) 6.67 (m, 2H) 2.3 5.98 (m, 2H) 5.50 (m, 2H) 2.6 3.89 (d, 2H) 14.3 2.28 (br, 2H) 2.35, 2.17, 1.83, 1.76, 1.70, 1.67, 1.62, 1.47 (m, 28 H) 0.90 (s, 6H) 0.71 (s, 3H) 0.24 (s, 3H) H 2 CCHCH2 Cp Cp Cp Cp Cp Cp H 2 CCHCH 2, anti H 2 CCHCH 2, anti H 2 CCHCH 2, syn H 2 CCHCH 2, syn MeAd MeAd MeAd Me2 Si Me?Si H 2 CCHCH2 Cp Cp Cp H 2 CCHCH 2, anti H 2 CCHCH 2, syn MeAd MeAd Me2Si Me?Si 190 344 7.15 (m, lH) 6.62 (2H) 5.96 (2H) 5.51 (2H) 3.07 (4H) 2.30, 2.18, 2.16, 1.81, 1.79, 1.70, 1.64, 1.48 (m, 28H) 0.92 (s, 6H) 0.69 (s, 3H) 0.22 (s, 3H) H 2 CCHCH2 Cp Cp Cp H 2 CCHCH 2 MeAd MeAd Me2 Si Me7 Si Table 3. Variable temperature 500 MHz 1 H NMR data for meso-AdpY(C3H 5 ) (17) Solvent Temp (K) 8 (ppm) Et2 0-d 10 263 6.67 (m, lH) 6.19 (m, 2H) 5.97 (br, 2H) 5.68 (m, 2H) 3.19 (d, 2H) 2.48 (d, 2H) 2.15, 2.10, 1.95, 1.85, 1.81, 1.68, 1.58, 1.48 (m) 1.70 (s, 6H) 0.67 (s, 3H) 0.36 (s, 3H) JH-H (Hz) Assignment 3 15 7.5 H 2 CCHCH2 Cp Cp Cp H 2 CCHCH 2, anti H 2 CCHCH 2, syn MeAd MeAd Me2 Si Me?Si 191 194 6.72 (rn, lH) 6.24 (lH) 6.20 (lH) 6.13 (lH) 5.93 (lH) 5.67 (2H) 3.34 (d, lH) 3.12 (d, lH) 2.56 (br, lH) 2.41 (br, lH) 2.13, 2.02, 1.93, 1.87, 1.55 1.46 (rn) 1.69 (s, 6H) 0.70 (s, 3H) 0.38 (s, 3H) 6.68 (m, lH)?? 6.03 (br, 2H) 5.91 (s, lH) 5.86 (s, lH) 5.77 (s, lH) 5.75 (s, lH) 3.53 (d, lH) 3.39 (d, lH) 2.81 (br, lH) 2.64 (br, lH) 2.45, 2.00, 1.87, 1.76, 1.59, 1.38, 1.26 (rn, 28H) 0.93 (s, 3H) 0.87 (s, 3H) 0.78 (s . 3H) 0.50 (s, 3H) 13 13 I toluene-d 8 186 14.3 13.4 H 2 CCHCH2 Cp Cp Cp Cp Cp H 2 CCHCH 2, anti H 2 CCHCH 2, anti H 2 CCHCH 2, syn H 2 CCHCH 2, syn MeAd MeAd M e2 Si Me7 Si H 2 CCHCH2 Cp Cp Cp Cp Cp H 2 CCHCH 2, anti H 2 CCHCH 2, anti H 2 CCHCH 2, s yn H 2 CCHCH 2, s yn MeAd MeAd MeAd M e2 Si Me7 Si 192 309 6.66 (m, lH) 6.09 (m, 2H) 5.95 (m, 2H) 5.74 (m, 2H) 3.32 (d, 2H) 2.61 (d, 2H) 2.05, 2.02, 1.99, 1.93, 1.90, 1.85, 1.78, 1.65, 1.61, 1.59, 1.57, 1.55, 1.52, 1.44 (m) 0.98 (s, 6H) 0.70 (s, 3H) 0.38 (s, 3H) 12.3 2.4 2.6 15.1 6.6 H 2 CCHCH2 Cp Cp Cp H 2CCHCH 2, anti H 2CCHCH 2, syn MeAd MeAd Me2 Si Me7 Si 1 Table 4. Variable temperature 500 MHz H NMR data for meso-DpSc(C3H 5 ) (18) Solvent Temp (K) 8 (ppm) JH-H Assignment (Hz) toluene-d 8 210 7.41 (m, lH) 6.66 (t, lH) 6.54 (s, lH) 5.91 (m, lH) 5.70 (m, lH) 5.17 (m, lH) 5.11 (m, lH) 4.48 (d, lH) 4.16 (d, lH) 2.31 (d, lH) 2.04 (d, lH) 1.05 (s, 9H) 1.00 (s, 9H) 0.67 (s, 3H) 0.20 (s, 3H) 2.2 br br 2.5 2.5 15.3 15.3 7.3 H 2 CCHCH2 Cp Cp Cp Cp Cp Cp H 2CCHCH 2, anti H 2CCHCH 2, anti H 2 CCHCH 2, syn H 2CCHCH 2, syn CMe3 CMe3 Me2 Si Me?Si 193 297 373 7.44 (quintet, lH) 6.91 (m, 2H) 5.90 (2H) 5.23 (t, 2H) 4.28 (d, 2H) 2.52 (s, 2H) 1.07 (s, 18H) 0.68 (s, 3H) 0.21 (s, 3H) 7.44 (quintet, lH) 6.75 (t, 2H) 5.95 (t, 2H) 5.28 (t, 2H) 3.22 (br) 1.09 (s, 18H) 0.69 (s, 3H) 0.22 (s, 3H) 12.3 br 2.7 12.5 12.4 2.6 2.2 2.8 H 2 CCHCH2 Cp Cp Cp H 2 CCHCH 2, anti H 2 CCHCH 2, syn CMe3 Me2 Si Me?Si H 2 CCHCH2 Cp Cp Cp H 2 CCHCH 2 CMe3 Me2 Si Me?Si Appendix D: Selected gNMR Data 3 Table 1. gNMR Parameters for Cp*2 Sc(ri -C3H 5 ) (15), toluene-d 8 , 500 MHz Molecule# 1 (Concentration= 1.00000) (C5 Me5 ) Shift W (Hz) J (Hz) Nucleus Group Name #n (ppm) 1 7.150 11 lH 5.00 21 2.050 5.00 7.00 31 3.930 5.00 14.80 41 2.050 5.00 7.00 3.930 51 5.00 14.80 Molecule# 2 (Concentration = 5.00000) (C5 Me5) W (Hz) J (Hz) Shift Nucleus Group #n (ppm) Name 5.00 1.860 13 lH 2 3 4 0.00 0.00 0.00 0.00 0.00 0.00 194 Molecule# 3 (Concentration = 5.00000) (C3H 5) W (Hz) J(Hz) Nucleus Group Shift Name #n (ppm) lH 13 1.730 5.00 Permutations Mol # Nucl Name 1 1 1 1 1 2 3 lH lH lH lH lH lH lH Group #n 11 21 31 41 51 13 23 113rotation 1-1 1-4 1-5 1-2 1-3 3-1 2-1 11 -111, C-C 3 1 11 -11 , C-C 3 1 11 -11 , M-C 1-1 1-3 1-2 1-4 1-5 3-1 2-1 1-1 1-2 1-3 1-5 1-4 3-1 2-1 1-1 1-4 1-5 1-2 1-3 3-1 2-1 3 3 Table 2. gNMR Parameters forme s o-AdpSc(11 -C H 3 ) 5 (16), Et 0-d 2 Allyl and Vinylic Cyclopentadienyl Protons · Molecule# 1 (Concentration= 1.00000) (C3H 5 ) W (Hz) J (Hz) Nucleus Group Shift Name #n (ppm) 1 11 2.610 lH 5.00 21 4.160 5.00 0.00 31 7.300 5.00 11.00 41 2.410 5.00 0.00 3.640 51 5.00 0.00 Molecule# 3 (Concentration = 1.00000) (C5H 3) Shift W (Hz) J (Hz) Nucleus Group Name (ppm) #n 1 5.00 lH 11 6.740 5.00 21 6.180 0.00 5.00 5.520 31 0.00 , 500 MHz, 10 • 2 3 4 16.00 0.00 0.00 9.00 15.00 0.00 2 0.00 195 Molecule # 4 (Concentration = 1.00000) (C5H 3 ) Nucleus Group W (Hz) J (Hz) Shift Name #n 1 (ppm) lH 11 5.00 6.700 21 5.00 0.00 5.920 5.00 0.00 31 5.490 Permutations Mol # Nucl Name 1 1 1 1 1 3 3 3 4 4 4 lH lH lH lH lH lH lH lH lH lH lH Group #n 11 21 31 41 51 11 21 31 11 21 31 3 11 rotation 1-4 1-5 1-3 1-1 1-2 4-1 4-2 4-3 3-1 3-2 3-3 3 2 0.00 3 3 1 11 -111, C-C 11 -171, C-C 11 -11 , M-C 1-2 1-1 1-3 1-4 1-5 4-1 4-2 4-3 3-1 3-2 3-3 1-1 1-2 1-3 1-5 1-4 4-1 4-2 4-3 3-1 3-2 3-3 1-4 1-5 1-3 1-1 1-2 4-1 4-2 4-3 3-1 3-2 3-3 3 Table 3. gNMR Parameters for rac-IpSc(17 -C 3H 5 ) (20), toluene-d 8 , 500 MHz, Allyl and Vinylic Cyclopentadienyl Protons Molecule# 1 (Concentration = 1.00000) (C3H 5 ) W (Hz) J (Hz) Nucleus Group Shift Name #n (ppm) 1 7.420 lH 11 5.00 2.550 21 5.00 8.00 3.780 5.00 31 15.00 1.910 5.00 41 8.00 4.510 5.00 51 15.00 Molecule# 2 (Concentration = 1.00000) (C5H 2 ) W (Hz) J (Hz) Shift Nucleus Group 1 Name #n (ppm) 5.00 6.370 lH 11 5.00 0.00 21 5.230 2 3 4 0.00 0.00 0.00 0.00 0.00 0.00 2 196 Molecule# 3 (Concentration = 1.00000) (C5H 2 ) Nucleus Group W (Hz) J (Hz) Shift Name 1 #n (ppm) lH 11 6.550 5.00 0.00 21 5.00 5.160 2 Permutations Mol # Nucl Name 3 1 Tl -ri C-C I 11 -111, C-C T1 -ri M-C 1-1 1-4 1-5 1-3 1-2 3-1 3-2 2-1 2-2 1-1 1-5 1-4 1-2 1-3 3-1 3-2 2-1 2-2 1-1 1-2 1-3 1-4 1-5 3-1 3-2 2-1 2-2 1 1 1 1 1 2 2 3 3 lH lH lH lH lH lH lH lH lH Group #n 11 21 31 41 51 11 21 11 21 113rotation 1-1 1-2 1-3 1-4 1-5 3-1 3-2 2-1 2-2 3 3 1 I Table 4. gNMR Parameters forrac-IpSc(ri 3 -C3H 5 ) (20), toluene-d 8 , 500 MHz, Isopropyl Protons Molecule # 1 (Concentration = 1.00000) Shift W (Hz) Nucleus Group #n (ppm) Name 11 3.100 5.00 lH 1.320 23 5.00 1.320 33 5.00 Molecule # 2 (Concentration = 1.00000) Shift W (Hz) Nucleus Group #n (ppm) Name 2.960 11 5.00 lH 1.160 23 5.00 1.160 5.00 33 J (Hz) 1 2 7.00 7.00 0.00 J (Hz) 1 2 7.00 7.00 0.00 197 Molecule # 3 (Concentration = 1.00000) Nucleus Group Shift W (Hz) Name #n (ppm) lH 11 5.00 2.530 23 1.100 5.00 5.00 33 1.100 Molecule # 4 (Concentration = 1.00000) Nucleus Group Shift W (Hz) Name #n (ppm) lH 5.00 11 2.430 23 5.00 1.060 5.00 33 1.060 Permutations Mal# Nucl Name 1 1 1 2 2 2 3 3 3 4 4 4 lH lH lH lH lH lH lH lH lH lH lH lH Group #n 11 23 33 11 23 33 11 23 33 11 23 33 173rotation 2-1 2-2 2-3 1-1 1-2 1-3 4-1 4-2 4-3 3-1 3-2 3-2 J (Hz) 1 2 7.00 7.00 0.00 J (Hz) 1 2 7.00 7.00 0.00 3 17 -171, C-C 3 17 -171, C-C 3 17 -171, M-C 2-1 2-2 2-3 1-1 1-2 1-3 4-1 4-2 4-3 3-1 3-2 3-2 2-1 2-2 2-3 1-1 1-2 1-3 4-1 4-2 4-3 3-1 3-2 3-2 2-1 2-2 2-3 1-1 1-2 1-3 4-1 4-2 4-3 3-1 3-2 3-2