Carbon-Hydrogen Bond Activation by Peralkylhafnocene and Peralkylscandocene Derivatives Citation Bulls, Al Ray (1988) Carbon-Hydrogen Bond Activation by Peralkylhafnocene and Peralkylscandocene Derivatives. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/ZSJW-KC36. https://resolver.caltech.edu/CaltechTHESIS:01172013-130132630 Abstract Chapter 1 Thermal decomposition of Cp* 2 Hf(CH 2 C 6 H 5 ) 2 (Cp* = (η 5 -C 5 Me 5 )) in benzene-d 6 cleanly affords toluene and hafnabenzocyclobutene [chemical symbol; see abstract in scanned thesis for details]. Deuterium labeling of the benzyl ligands indicates that decomposition of Cp* 2 Hf(CY 2 C 6 H 5 ) 2 (Y = H, D) proceeds primarily by α-H abstraction to form a permethylhafnocene benzylidene intermediate [Cp* 2 Hf=CHC 6 H 5 ], which rapidly rearranges to the metallated-cyclopentadienyl, or "tucked-in" benzyl complex Cp*(η 5 ,η 1 -C 5 Me 4 CH 2 )HfCH 2 C 6 H 5 . The observed product arises from rearrangement of Cp*(η 5 ,η 1 -C 5 Me 4 CH 2 )HfCH 2 C 6 H 5 to its tautomer [chemical symbol; see abstract in scanned thesis for details]. A series of meta substituted benzyl derivatives of the proposed metallated cyclopentadienyl intermediates, Cp*(η 5 ,η 1 -C 5 Me 4 CH 2 )HfCH 2 C 6 H 4 X (X = H, CH 3 , CF 3 , NMe 2 ), has therefore been prepared. The kinetics of their conversion to [chemical symbol; see abstract in scanned thesis for details] have been examined in order to probe the nature of the transition state for aryl C-H bond activation which occurs in the final steps of the rearrangement. The rates are found to be insensitive to the nature of X, suggesting that the benzyl π system is not attacked by the electrophilic hafnium center along the reaction coordinate for C-H bond activation. The structure of Cp*(η 5 ,η 1 -C 5 Me 4 CH 2 )HfCH 2 C 6 H 5 , as determined by single crystal X-ray diffraction techniques, indicates that the complex is best described as a Hf(IV) derivative containing an η 5 ,η 1 -C 5 Me 4 CH 2 Iigand, rather than a Hf(ll) η 6 -fulvene adduct. Cp*(η 5 ,η 1 -C 5 Me 4 CH 2 )HfCH 2 C 6 H 5 crystallizes in the triclinic space group P1̅ (a= 9.084(2), b = 10.492(2), c = 12.328(1) Å; α = 95.81(1), β = 96.60(1), γ = 91.15(2); Z = 2). Least-squares refinement led to a value for R of 0.048 (I > 3σI) and a goodness-of-fit of 4.37 for 4029 independent reflections. Chapter 2 Thermolysis of the singly metallated complex Cp*(η 5 ,η 1 -C 5 Me 4 CH 2 )Hf(CH 2 CMe 3 ) in toluene affords neopentane and the doubly metallated complex Cp*(η 5 ,η 1 ,η 1 -C 5 Me 3 (CH 2 ) 2 )Hf. The structure of Cp*(η 5 ,η 1 ,η 1 -C 5 Me 3 (CH 2 ) 2 )Hf, as determined by single-crystal X-ray diffraction techniques, confirms that metallation has occurred at adjacent methyl groups of the same pentamethylcyclopentadienyl ring. Cp*(η 5 ,η 1 ,η 1 -C 5 Me 3 (CH 2 ) 2 )Hf crystallizes in the space group P2/c (a= 13.775(4) Å, b = 9.516(1) Å, c = 14.183(6) Å; β = 103.965(31)°, z = 4). Least squares refinement led to a value for R of 0.036 (I > 3σ i ) and a goodness-of-fit of 2.66 for 1984 independent reflections. Cp*(η 5 ,η 1 ,η 1 -C 5 Me 3 (CH 2 ) 2 )Hf and Cp*(η 5 ,η 1 -C 5 Me 4 CH 2 )Hfl cleanly insert one equivalent of ethylene into the hafnium methylene carbon bond to form the propyl bridged species Cp*(η 5 ,η 1 ,η 1 -C 5 Me 3 (CH 2 )(CH 2 CH 2 CH 2 ))Hf and Cp*(η 5 ,η 1 -C 5 Me 3 CH 2 CH 2 CH 2 )Hfl, respectively. Cp*(η 5 ,η 1 ,η 1 -C 5 Me 3 (CH 2 )(CH 2 CH 2 CH 2 ))Hf reacts with excess ethylene to cleanly afford [chemical symbol; see abstract in scanned thesis for details]. Deuterium labeling experiments suggest that this reaction occurs via an α-H abstraction to generate the alkylidene intermediate Cp*(η 5 ,η 1 -C 5 Me 4 CH 2 CH 2 CH=)Hf, which rapidly traps ethylene to form the observed product. The metallated phenyl complex Cp*(η 5 ,η 1 -C 5 Me 3 CH 2 )HfC 6 H 5 has been shown to be in equilibrium with the benzyne complex Cp* 2 Hf(η 2 -C 6 H 4 ). Heating benzene-d 6 solutions of Cp*(η 5 ,η 1 -C 5 Me 3 CH 2 )HfC 6 H 5 in the presence of ethylene traps the benzyne intermediate and generates the hafnacyclopentene [chemical symbol; see abstract in scanned thesis for details]. Chapter 4 Relative bond dissociation energies (BDEs) have been obtained by equilibrating early transition metal alkyls and hydrides with H 2 or the C-H bonds of hydrocarbons. Thus, in benzene solution Cp* 2 HfH 2 (Cp* = (η 5 -C 5 Me 5 )) equilibrates with Cp* 2 Hf(C 6 H 5 )H and dihydrogen. From the enthalpy of the reaction, ΔH° = +6.0(3), the Hf-H (BDE) is calculated to be 0.8(3) kcal·mol -1 stronger than the Hf-C 6 H 5 BDE. Relative Sc-C 6 H 5 and Sc-alkyl BDEs have been estimated from the equilibration of the metallated complex Cp*(η 5 ,η 1 -C 5 Me 4 CH 2 CH 2 CH 2 )Sc, C 6 H 6 and Cp*(η 5 -C 5 Me 4 CH 2 CH 2 CH 3 )Sc(C 6 H 5 ), the Sc-C 6 H 5 BDE being 16.6(3) kcal·mol -1 stronger than the Sc-CH 2 CH 2 CH 2 C 5 Me 4 BDE. From a similar reversible intramolecular metallation of Cp*(η 5 -C 5 Me 4 CH 2 CH 2 CH 3 )HfH 2 to give Cp*(η 5 ,η 1 -C 5 Me 4 CH 2 CH 2 CH 2 )HfH and dihydrogen, the Hf-H BDE is estimated to be 23.0(3) kcal·mol -1 stronger than the Hf-CH 2 CH 2 CH 2 C 5 Me 4 BDE. The equilibration of Cp*(η 5 -C 5 Me 4 CH 2 C 6 H 5 )Sc-C≡CCMe 3 with the metallated scandocene derivative Cp*(η 5 ,η 1 -C 5 Me 4 CH 2 -o-C 6 H 4 )Sc and tert-butylacetylene lies very far towards Cp*(η 5 -C 5 Me 4 CH 2 C 6 H 5 )Sc-C≡CCMe 3 , so that only a lower limit for the relative Sc-alkynyl and Sc-aryl BDEs may be determined: BDE(Sc-alkynyl) - BDE(Sc-aryl) ≥ 29(5) kcal·mol -1 . These early transition metal-hydrocarbyl (M-R) BDEs correlate with the corresponding H-R BDEs (i.e. M-alkynyl > M-aryl > M-alkyl); however, the M-R BDEs increase more rapidly with s character than does the H-R BDEs. The origin of this effect is not known, but it is undoubtedly also responsible for the characteristically high M-H BDEs for transition metal hydride compounds. In an effort to probe the polarity of Sc-aryl bonds a series of scandocene derivatives capable of reversibly metallating at either of two differently substituted benzyl groups was prepared. The equilibrium constants for these metallated derivatives: (η 5 ,η 1 -C 5 Me 4 CH 2 -o-C 6 H 3 -p-X)(η 5 -C 5 Me 4 CH 2 C 6 H 4 -m-CH 3 )Sc ⇌ (η 5 -C 5 Me 4 CH 2 C 6 H 4 -m-X)(η 5 ,η 1 - C 5 Me 4 CH 2 -o-C 6 H 3 -p-CH 3 )Sc (X= H, CF 3 , NMe 2 ) were determined. The small dependence of Keq on the nature of X suggests that the Sc-aryl bond is essentially covalent with only a slight ionic character. 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: Gray, Harry B. (chair) Bercaw, John E. Schaefer, William P. Beauchamp, Jesse L. Defense Date: 17 August 1987 Record Number: CaltechTHESIS:01172013-130132630 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:01172013-130132630 DOI: 10.7907/ZSJW-KC36 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 7403 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 18 Jan 2013 17:20 Last Modified: 18 Dec 2020 02:11 Thesis Files Preview PDF - Final Version See Usage Policy. 29MB Repository Staff Only: item control page

Carbon-Hydrogen Bond Activation by Peralkylhafnocene and Peralkylscandocene Derivatives Thesis by AI Ray Bulls In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California 1988 (Submitted August 17, 1987) ii Acknowledgements One of the most difficult parts of this whole thing is trying to thank everyone who has made the last five years such an experience. Most important are all of the "people" who have ventured through the Bercaw group. Tippy and DMR helped me get started on that Cadillac of vacuum lines in 213 Noyes. Barb was a faithful suck buddy for three years. Rocco, Vanman (now known as Allan), Larry, and Ged F.L.C Parkin were and are good friends. Leon "Jeff" Gelles taught me the value of a good attitude. George Spies and Ed Schlesinger deserve a lot of thanks for getting me out of MJ and letting live in an almost regular house. I had a lot of help with the work in this thesis. Specifically, Michael Serfas, Frank Kragh, and Drs. Bernie Santarsiero, Bill Schaefer, and Dick Marsh solved the four crystal structures that are reported. Juan Manriquez is responsible for the synthesis of the substituted benzyltetramethylcyclopentadienylligands reported in chapter 4. Wayne Luekins prepared a few of the Bs ligated zirconium compounds reported in chapter 3. Oh yeah, Van man (I mean Allan) made profound contributions to chapters 1 and 5. Of course, the guy who is most responsible is John Bercaw, who has been a good friend and advisor for the past five years. Errrrr, I'm pretty sure I'm forgetting someone ... iii Table of Contents Chapter 1. Intramolecular C-H Bond Activation of Benzyl Ligands by Metallated-Cyclopentadienyl Derivatives of Permethylhafnocene. Molecular Structure of (77 5-CsMes)(f7 5 ,77 5-CsMe4CH2)HfCH2C6Hs and the Mechanism of Rearrangement to Its Hafnabenzocyclobutene Tautomer (77 5-CsMes)2HfCH2-o-C5H4. Chapter 2. Synthesis and Reactivity of Singly and Doubly Metallated Cp* 35 Containing Derivatives of Permethylhafnocene: The Molecular Structure of Cp*(775 ,77 1 ,77 1-CsMe3(CH2)2)Hf. Chapter 3. The Synthesis of Metallation-Resistant Bis(cyclopentadienyl) 64 Ligand Systems. The Molecular St ructure of [ {J.t-Me2Si)2(77 5 -CsH2-4-CMe3)2]ZrCI2. Chapter 4 . Relative Bond Dissociation Energies for Early Transition Metal Alkyl, 80 Aryl, Alkynyl, and Hydride Compounds. Equilibration of Metallated Cyclopentadienyl Derivatives of Peralkylhafnocene and Peralkylscandocene with Hydrocarbons and Dihydrogen. Chapter 5. The Mechanism of Dihydrogen Elimination from Hydrido Alkyl Derivatives of Permethylhafnocene. The Molecular Structure of Cp*2HfCH2CH(CH3)CH2. 128 1 Chapter 1 Intramolecular C-H Bond Activation of Benzyl Ligands by Metallated-Cyclopentadienyl Derivatives of Permethylhafnocene. Molecular Structure of (TJ 5-CsMes)(TJ 5,17 1 -CsMe4CH2)HfCH2C6 H 5 and the Mechanism of Rearrangement to Its Hafnabenzocyclobutene Tautomer ( 17 5-CsMes)2HfCH2-o-CsH4. 2 Abstract: Thermal decomposition of Cp*2Hf(CH2CeHs)2 (Cp* = ('7 5-CsMes)) in benzene-de cleanly affords toluene and hafnabenzocyclobutene Cp*2HfCH2-o-CeH4. Deuterium labeling of the benzyl ligands indicates that decomposition of Cp*2Hf(CY2CeHs)2 (Y = H, D) proceeds primarily by a-H abstraction to form a permethylhafnocene benzylidene intermediate [Cp*2Hf=CHCeHs], which rapidly rearranges to the metallated-cyclopentadienyl, or "tucked-in" benzyl complex Cp*(f7 5 ,'7 1-CsMe4CH2)HfCH2CeHs. The observed product arises from rearrangement of Cp*(f7 5 ,'7 1-CsMe4CH2)HfCH2CeHs to its tautomer Cp*2HfCH2-o-CeH4. A series of meta substituted benzyl derivatives of the proposed metallated cyclopentadienyl intermediates, Cp*(f7 5 ,'7 1-CsMe4CH2)HfCH2CeH<V< (X= H, CH3, CF3, NMe2), has therefore been prepared. The kinetics of their conversion to Cp*2HfCH2-o-CeH3X have been examined in order to probe the nature of the transition state for aryl C-H bond activation which occurs in the final steps of the rearrangement. The rates are found to be insensitive to the nature of X, suggesting that the benzyl1r system is not attacked by the electrophilic hafnium center along the reaction coordinate for C-H bond activation. The structure of Cp*(17 5 ,'7 1-CsMe4CH2)HfCH2CeHs, as determined by single crystal X-ray diffraction techniques, indicates that the complex is best described as a Hf(IV) derivative containing an '7 5 ,17 1 -C5 Me4CH 2 Iigand, rather than a Hf(ll) 176 -fulvene adduct. Cp*(f7 5 ,'71 -CsMe4CH2)HfCH2CeHs crystallizes in the triclinic space group P1 (a= 9.084(2), b = 10.492(2), c = 12.328(1) A; a= 95.81 (1), (3 = 96.60(1), -y = 91.15(2); Z = 2). Least-squares refinement led to a value for R of 0.048 (I > 3al) and a goodness-of-fit of 4.37 for 4029 independent reflections. 3 Introduction High valent, early transition metal, lanthanide and actinide complexes have been shown to be reactive towards both aromatic and aliphatic carbon-hydrogen bonds.£1 1 Associative mechanisms proceeding through highly ordered, polar, 4-centered transition states typically have been invoked for C-H bond cleavage by these d 0 (or d 0 fn) metal systems.£2) An electrophilic aromatic substitution in which the metal interacts initially with the arene 1r system represents an alternative mechanism for aromatic C-H bond activation.£3) This type of mechanism seems particularly feasible in view of the electron deficient, highly Lewis-acidic nature of coordinatively unsaturated d 0 metal centers. Trihapto coordination of the benzyl ligands of Zr(OR)(CH2CsHs)3 (R =2,6-di-tert- butylphenyl), as evidenced by the X-ray crystal structure, demonstrates such a 1r arene-Lewis acid interaction.£4 1 Similarly, the crystal structures of Cp*Th(CH2CsH5 )3[S] and M(CH2CsHs)4 (M =Ti, Zr, Hf)l6 1show the metal to be interacting with the 1r systems of the benzyl ligands. There are several examples of transition metal complexes serving as electrophiles in aromatic substitution reactions. Ogoshi et at. reported that arene C-H bond activation by a cationic Rh(lll) complex occurred via electrophilic aromatic metallation.£7) Shul'pin and co-workers have invoked a similar mechanism for arene activation at an anionic Pt(IV) center.£8] The intramolecular cyclopalladation of ligands such as azobenzene has also been suggested to occur by electrophilic aromatic substitution, with rearrangement of a 1r- to a-arene adduct, followed by loss of a proton leading to the observed aryl product.l9] Deeming and Rothwell, however, have shown that cyclopalladation can also occur with benzo[h]quinoline ligands constrained by chelation from forming a 1r complex.£ 1 0 1 Graham has prepared 17 2-arene complexes of the type [('7 5 -CsHs)Re(NO)(C0)('7 2-arene)t, which closely resemble the proposed intermediates of electrophilic aromatic substitutions.£ 11 1 4 Jones and Feher have developed a convincing case for the intermediacy of a 1r-arene adduct along the reaction coordinate leading to C-H bond addition to coordinatively unsaturated [Cp*Rh(PMe3)][ 12l, although this electron-rich rhodium center is not likely to be a strong Lewis acid. In order to probe the mechanism by which electron deficient, d 0 metal complexes activate aromatic carbon-hydrogen bonds we have examined the thermal decomposition of a series of bis(benzyl) derivatives of permethylhafnocene, Cp*2Hf(CH2C5H4-m-X)2 (Cp* = '7 5-CsMes; X= H, CH3, NMe2), which cleanly afford toluene (or the substituted toluene) and the hafnobenzocyclobutene complexes Cp*2HfCH2-o-C5H3X. Although the most straightforward decomposition pathway to the observed products is direct ortho C-H abstraction from a benzyl ligand by the other Hf-CH2C5Hs group, a close examination revealed that the mechanism is much more complex; both benzylidene and metallated cyclopentadienyl intermediates are involved. We report herein the synthesis, reactivity and X-ray structure determination of one intermediate, Cp*('7 5 ,'7 1-CsMe4CH2)HfCH2CsHs, and the results of a study of the mechanism of its rearrangement to Cp*2HfCH2-o-CsH4. 5 Results and Discussion Permethylhafnocene dichloride, Cp*2HfCI2 (1), reacts rapidly in benzene solution with two equivalents of the appropriate benzyl potassium reagent£13] to give the dibenzyl complexes 2-4 (equation 1). Cp*2HfCI2 + 2 K(CH2C5H4-m-X) - - - - i > Cp*2Hf(CH2C5H4-m-X)2 + 2 KCI 1 2, X= H 3, X= CH3 4, X= NMe2 (1) Isolation from petroleum ether affords 2-4 in good yields as pale yellow crystalline compounds, which may be handled briefly in air and are indefinitely stable at room temperature under an N2 atmosphere. Thermolysis of benzene or toluene solutions of 2-4 yields the hafnabenzocyclobutenes 5-7 and meta-substituted toluene (equation 2).£14) (2) 2 -4 5, X= H 6, X= CH 3 7, X= NMe 2 The conversion of complexes 2-4 to metallacycles 5-7 and toluene follows first order kinetics for 2-3 half-lives. Abstraction of an ortho-benzyl hydrogen by the neighboring Hf-benzyl group is the most direct mechanism for the thermal decomposition of 2 to 5 . Hence, we undertook an 6 investigation of the relative rates of decomposition of the meta-substituted dibenzyl complexes 5-7 in order to distinguish between an electrophilic aromatic substitution-like mechanism, proceeding through an '7 1-phenonium ion intermediate or transition state£1 51, and a mechanism in which the arene C-H a bond and Hf-benzyl bonds undergo sigma bond metathesis without interaction of the hafnium center with the arene 1r system. To probe whether this straightforward mechanism is, in fact, being followed, the isotopically labeled dibenzyl complexes (Cp*-d15)2Hf(CH2C5H5)2 (2b), Cp*2Hf(CD2C5D5)2 equation 3. -2KCI { '7 5-C5(CXs)5 }2Hf(CY2C5Z5)2 2;X=Y=Z=H 2b; X= D, Y = Z = H 2c; X = H, Y = Z = D 2d; X = Z = H, Y = D 2e; X = Y = H, Z = D (3) Thermal decomposition of 2e at 140°C occurs at roughly the same rate as 2 (kH/ko = 1.07) and produces ca. 95% C5D5CHa[ 1 6] (Table 1). A moderate deuterium kinetic isotope effect is observed (kH/ko = 3.1) when the benzyl methylene groups are deuterated (2d), and the toluene produced upon decomposition is approximately 85% C5H5CDa. Perdeuteration of the pentamethylcyclopentadienyl rings (2b) results in a small deuterium kinetic isotope effect (kHfkD = 1.1) and ca. 95% C5H5CH3. The magnitudes of the kinetic deuterium isotope effects and the labeling of the toluene obtained for the thermal decomposition of compounds 2b-e indicate a principal pathway involving a rate- limiting a-H abstraction process (Scheme 1), rather than direct transfer of an orlho-benzylic hydrogen to the adjacent benzyl group. 7 Toluene 2 5 8 9 Scheme 1. Major pathway for the thermal decomposition of Cp*2Hf(CH2CsHs)2 (2). 8 Table 1 . Rates of Decomposition of 2, and 2b-e (equation 2) and the Isotopically Labeled Toluenes Produced (140°C, C6D6)· Compound k(1 o-6s-1) Toluenea 2 5.81 (19) 2b 5.29(27) 1.1 (1) 95% C5H5CH3 2c 2.29(6) 2.5(2) 2d 1.88(6) 3.1 (2) 2e 5.40(18) 1.1 (1) 80% C5D5CD3 20% C5D5CD2H 85% C5H5CD3 15% C5H5CD2H 95% C5D5CH3 a Approximate ratio of labeled toluenes were determined by 1 H NMR analysis of the product toluene methyl resonance. Particularly indicative of this pathway are the moderate kinetic deuterium isotope effect observed for the decomposition of Cp*2Hf(CY2C5H5)2 (Y = H, 2; Y = D, 2d), and the formation of cx-d3 toluene for 2d. Thus, the hafnium benzylidene complex 8 is implicated as an intermediate. Moreover, the observation of deuterium (2H NMR) in the Cp* ligands of metallacycle product 5 upon thermolysis of Cp*2Hf(CD2C5D5)2 (2c) suggests the intermediacy of 9, which could arise from 8 by pentamethylcyclopentadienyl ring-hydrog en abstraction. Transfer of an orlho-benzylic hydrogen of 9 to the ( f7 5,'7 1-C5Me4CH2) ligand finally generates the observed product The activation parameters for the thermal decomposition of 2, over the temperature range 125-148°C (Table 2), t.H* =34(1) kcal mol-1; t.s* =+1 (3) e.u., are similar to those measured for an analo gous (rate-limiting) ex-hydrogen abstraction process in the therm al decomposition of Cp* 2Ti(CH3)2 to Cp*(f7 5,'7 1 -C5Me4CH2)TiCH3 and methane (t.H* kcal·mol-1 ; t.s* =27 =-2.9 e.u.).£17] The conversion of 8 to 9 is similar to that reported by 9 Table 2. Rate Data and Activation Parameters for the Thermal Decomposition of Cp*2Hf(CH2C5H5)2 (2) (benzene-dB) Temperature (oC) 125 1.11 (5) 132.5 2.86(1 0) 140 5.81 (19) 148 14.5(9) ~Hi = 34(±1) kcal/mol ~s* =1 (±3) e.u. Rothwell eta/. for the intramolecular addition of an aliphatic C-H bond across a tantalum-carbon double bond.£18] The appearance of ca. 20% C5D5CD2H from the thermolysis of Cp*2Hf(CD2C5D5)2 (2c) in benzene-d6 indicates that direct abstraction of a pentamethylcyclopentadienyl ring-hydrogen by a departing hafnium benzyl ligand, generating intermediate 9 without the intermediacy of 8, is a competitive pathway for the decomposition of Cp*2Hf(CH2C6H5)2. This minor pathway is likely to be most prominent when the a-hydrogen abstraction is slowed by deuteration of the benzyl methylene groups as in 2c and 2d. Also consistent with direct pentamethylcyclopentadienyl ring-hydrogen abstraction being a relatively minor pathway is the small deuterium kinetic isotope effect observed upon the thermolysis of The proposed intermediate 9 and its meta-substituted analogs 13-15 can be prepared independently in high yields by the route shown in equations 4-6. 10 (4) (5) Cp*('7 5,'7 1-CsMe4CH2)Hfl + MCH2C5H4-m-X - M+l- Cp*('7 5.77 1-CsMe4CH2)Hf(CH2CsH4-m-X) 9 ; X= H 13; X= CH3 14; X= CF3 15; X= NMe2 (6) M = K or MgCI A moderate deuterium kinetic isotope effect (kH/ko at 140°C = 3.75) is observed upon thermolysis of (Cp*-d15)2Hf(CH2CMe3)l, suggesting that formation of Cp* ('7 5,'7 1-CsMe4CH2)Hfl proceeds by the rate-limiting coupling of the neopentyl ligand and a pentamethylcyclopentadienyl ring-hydrogen. In benzene solution compounds 9 and 13-15 readily rearrange to hafnabenzocyclobutenes 5-7 and 16, but unlike the thermally robust d ibenzyl complexes 2-4, this conversion occurs slowly over several days at room temperature (equation 7). (7) 9, 13-15 5, 6, X= H X = CH 3 7, X= NMe2 16, X = CF 3 11 The lack of reactivity of the pentamethylcyclopentadienyl-metallated iodide derivative Cp*('7 5 ,'7 1 -CsMe4CH2)Hfl towards benzene (and benzene-d5) solvent, even on prolonged heating at 140°C, is striking evidence of the tremendous kinetic advantage that intramolecular C-H bond activation has over intermolecular C-H activation for these sterically encumbered permethylhafnocene derivatives. [ 191 Cp*('7 5 ,'7 1 -CsMe4CH2)HfCD2C6Ds (9-d?), obtained by treating Cp*('7 5 ,'7 1 -CsMe4CH2)Hfl with benzylpotassium-d7, decomposes much more slowly than Cp*('7 5 ,'7 1-CsMe4CH2)HfCH2C6Hs (kH/ko = 9.6 at 59°C), consistent with benzylic C-H or C-D bond cleavage in the rate-determining step. Deuterium is observed in the pentamethylcyclopentadienylligand eH NMR) of the hafnabenzocyclobutene 5-d? derived from 9-d?. Steric factors should favor activation of the benzylic hydrogen which is para rather than ortho to the substituent X in 13-15, and high field 1 H NMR spectra (400 MHz) of 6, 7 and 16 indicate X is, indeed, para to hafnium (Table 3). Moreover, the rate of conversion of 9 (X= H) to metallacycle 5 is roughly twice that measured for the conversion of 13-15 to metallacycles 6, 7 and 16, since the concentration of reactive (sterically unencumbered) C-H bonds for 9 is twice that of the compounds where X ~ H. Kinetic data and activation parameters for the rearrangement of the cyclopentadienyl-metallated benzyl derivatives 9 and 13-15 to the corresponding metallacycles 5-7 and 16 are given in Table 4. As can be seen, the rates of decomposition are nearly independent of the nature of X, and a plot of log k (at 32°C) versus the Hammett constants apara[20] gives a reasonably linear fit with p =-0.2J21 J Whereas the sign of p is in agreement with an electrophilic aromatic substitution mechanism, its small value argues against this pathway for the decomposition of 9 and 13-15, in spite of the electrophilic nature of the hafnium center. Values of p tor electrophilic aromatic substitution by cationic organic and inorganic electrophiles typically range from -1.5 to - 12J 17•22l Unfavorable steric 12 interactions between the benzyl ligand and the two pentamethylcyclopentadienylligands likely prevent the arene from coordinating to the hafnium center w ith its 1r system aligned with the vacant orbital on the permethylhafnocene moiety, as would be required for an electrophilic aromatic substitution mechanism. Similar, unfavorable steric interactions were invoked to reconcile the non-involvement of the 1r system in "sigma bond metathesis" reactions between (77 5-CsMes)2Sc-CH3 and substituted arenes CsHsX (X= CF3, H, CH3, OCH 3, NMe2) to yield (77 5-CsMes)2Sc-CsH4X and methane. [ 1 b] Moreover, models suggest considerable ring strain accompanies the transition state for the lower pathway of Scheme 2. Table 4. Rate Constants and Activation Parameters for the Thermal Conversion of Cp*('7 5 ,'7 1 -CsMe4CH2)Hf(CH2CsH4-m-X) to 5-7 and 16 (benzene-ds). X T(OC) k(1o-6s-1 ) 32 48 59 3.55(10) 19.6(6) 60.4(13) 32 48 59 1.69(12) 11.2(4) 34.9(9) 32 48 59 2.07(8) 12.2(6) 35.8(13) 32 48 59 2.16(9) 15.5(6) 46.3(13) H CH3 CF3 N(CH3)2 t.H* (kcal/mol) t.s* (e.u.) 20.2(3) -17.1 (7) 21 .6(3) -1 4.1(6) 20.3(6) -17.9(17) 22.0(4) -12.4 (11) 13 Cp*2HfCH2-o-C5H4 appears to involve attack of hafnium at an ortho benzylic C-H bond with the phenyl ring roughly parallel to the metallated cyclopentadienylligands (upper pathway of Scheme 2). Hfco . I ", ' . I : 'H - C -- -/ \. / 9 5 -J~ / \c H \ '·., SCHEME 2. Two potential pathways for the conversion of Cp*('7 5 ,'7 1 -CsMe4CH2)HfCH2C5Hs 14 The very small substituent effects for aryl C-H bond activation observed for both permethylscandocene and permethylhafnocene alkyls would appear to indicate similar mechanisms. Sigma bond metathesis involving d 0 ('7 5 -CsMes)2Sc-R derivatives almost certainly proceeds via a four-centered, non-redox mechanism, and such reactions of permethylhafnocene alkyls may well proceed in an analogous manner. On the other hand, an oxidative addition/reductive elimination sequence (lower pathway, Scheme 3) must be considered for the cyclopentadienyl-metallated hafnium derivatives 9 and 13- 15, since two A ~ ? ~~ Hf~ ~ B Scheme 3 15 Thus, a distinction between mechanisms (upper pathway of Scheme 2 or lower pathway of Scheme 3) rests on which of the two limiting resonance forms (A or B) best describes the ground states of 9 and 13-15. NMR data suggest the structure of 9 is intermediate between resonance forms A and B. The diastereotopic hydrogens of the bridging methylene are coupled to each other with 2JHH = 8.5 Hz, intermediate between typical values for geminal sp2 and sp3 hydrogens of 0-3 and 12-15Hz, respectively. The gated(1H NOE-enhanced) carbon spectrum for 9 exhibits a signal for the methylene carbon at o 71.2 with 1JcH =142 Hz, again intermediate between 1JcH values for sp2 and sp3 of 125 and 160Hz, respectively. In view of the ambiguities in these NMR data, an X-ray crystal structure determination was undertaken in hopes of distinguishing between A and B. The molecular structure of 9 is shown in Figure 1, and a skeletal view is shown in Figure 2. Unlike the unmetallated (17 5-CsMes) ring in 9, which has ring C-C distances from 1.414(1 0) to 1.424(11) A, the metallated-(17 5-CsMes) C-C distances vary from 1.396(11) to 1.460(11) A. The variations are not indicative of an 176-fulvene ligand, however. For example, the three carbon-carbon single bonds in the 17 6-fulvene formalism average 1.421 (11) A, whereas the two internal carbon-carbon double bonds are insignificantly longer (averaging 1.423(11) A). The C25M-Hf distance (2.320(9) A) is only 0.06 A longer than the C25-Hf distance of 2.258(8) A, and C25M is bent (an impressive) 37° out of the ring-carbon plane towards Hf. The C25-C25M distance of 1.453(12) A is somewhat shorter (0.046 A) than the average ring-carbon to methyl-carbon distance of 1.499(11) A. Only recently have compounds containing a coordinated fulvene ligand been structurally characterized for comparison to Cp*(17 5,17 1-CsMe4CH2)HfCH2C6H5. Marks et al. have described the structure of the closely related zirconium derivative Cp*(17 5,17 1-CsMe4CH2)ZrC6H5 and concluded that the 176 ligand has little fulvene Figure 1. The molecular structure of Cp*(I'J 5 ,1']1_CsMe4CH2)HfCH2C6 Hs (9). The themal ellipsoids are shown at the 50% probability level. ...... 0> 17 18 character.£23] Green and co-workers reported a series of fulvene complexes of the type M('7 5,'7 1-CsH4CR2)('7 6-C6HsR') (M =Moor W; R = CH3 or C5H5; R' = H or CH3).£24] Based on ring-carbon to methylene-carbon distances of 1.42(2) to 1.45(3) Aand values of 0 [2 5] ranging from 36 to 39°, these authors concluded the metal-fulvene moiety is better described as M 11 ('7 5,'7 1-CsH4CR2) than as M0 (17 6-CsH4CR2)· The similar structural features found for the tetramethylfulvene moiety for Cp*('7 5,'7 1-CsMe4CH2)HfCH2C5Hs would, by extension of Green's reasoning, argue for Cp*('7 5,'7 1-CsMe4CH2)Hf1V(CH2C5H4-m-X) (resonance form A) structures for 9 and 13-15.£26] 19 Conclusions Labeling experiments demonstrate that the thermal decomposition of Cp*2Hf(CH2C 6Hs)2 (2) occurs primarily by an a-H abstraction process to generate a transient hafnium benzylidene species, which rearranges to the metallated-cyclopentadienyl benzyl complex Cp*('7 5 ,f7 1 -CsMe4CH2)HfCH2C6Hs (9), and finally to the observed hafnabenzocyclobutene Cp*2HfCH2-o-C6H4 (5). A convenient starting material for the synthesis of a series of metallated-cyclopentadienyl derivatives of permethylhafnocene is Cp*(f7 5 ,f7 1 -C 5 Me4CH 2)Hfl, obtained in high yield by the pyrolysis of Cp*2Hf(CH2CMe3)l. The small rate differences among the series of complexes Cp*(f7 5 ,'7 1 -CsMe4CH2)Hf(CH2C6H4-m-X) (X= CF3, H, CH3, NMe2) and the relatively large negative values for the activation entropies indicate that the thermal rearrangement of Cp*('7 5 ,f7 1 -CsMe4CH2)HfCH2C6Hs to Cp*2HfCH2-o-C6H4 occurs by a highly ordered, four-centered transition state, similar to that observed for intermolecular, sigma bond metathesis reactions involving arenes and permethylscandocene alkyls.£1 b] Specifically, the cleavage of the aryl C-H bond occurs by direct attack of the electrophilic hafnium center at the carbon-hydrogen a bond, without interaction of the 1r orbitals of the arene (i.e. according to the upper pathway of Scheme 2). 20 Table 3. 1H and 13C NMR dataa Compound Assignment o(ppm)b 5 TJ -Cs(CH3)5 CH2C5H5 CH2C5H5 1.77 s 1.41 br 6.83-7.33 m TJ 5-Cs(CH3)5 CH2C5H4CH3 CH2C5H4CH3 CH2C5H4CH3 1.78 s 1.44 br 2.27 s 6.73-7.27 m coupling 0 Cp*2Hf(CH2C5Hs)2 (2) Cp*2Hf(CH2C5H4-m-CH3)2 (3) Cp*2Hf(CH2C5H4-m-NMe2)2 (4) TJ 5-Cs(CH3)5 CH2C5H4NMe2 CH2C5H4NMe2 CH2C5H4N(CH3)2 1.83 s 1.50 br 6.37-7.30 m 2.71 s TJ 5-Cs(CH3)s CH2C5H4 CH2C5H4 1.68 s 2.30 br 7.03-7.80 m 5 TJ -Cs(CH3)s CH2C5H3CH3 CH2C5H3CH3 1.62 s 2.13 br 7.52 s 7.44d 7.10 d 2.16 s Cp*2HfCH2-o-C5H4 (5) Cp*2HfCH2-o-C5H3-m-Me (6) CH2C5H3CH3 3JHH=7.0 3JHH=7.0 Cp*2HfCH2-o-C5H3-m-NMe2 (7) 175-Cs(CH3)5 CH2C5H3NMe2 CH2C5H3NMe2 1.75 s 2.37 br 7.48 d 7.18 d 6.81 dd CH2C5H3N(CH3)2 2.72 s 3JHH=7.6 4JHH=2.2 3JHH=7.6 4JHH=2.2 21 Cp*(77 5,'7CcsMe4CH2)Hf(CH2C6Hs) (9) 77 5-Cs(CH3)5 Cs(CH3)4CH2 CH2C6Hs CH2C6Hs rJ 5-Cs(CH3)5 rJ 5-Cs(CH3)s Cs(CH3)4CH2 1.88 s 1.54 d 1.96 d 1.43 s 1.52 s 1.67 s 1.95 s -0.33 d 2.13 d 6.85-7.15 m 57.25 t 12.19 q 117.5 s 71.24 t rJ 5-Cs(CH3)s CH2C(CH3)3 CH2C(CH3)3 2.00 s 0.02 5 1.27 5 rJ 5-Cs(CH3)s Cs(CH3)4CH2 1.96 5 1.42 d 1.99 d 1.53 5 1.69 5 1.98 5 2.405 13.22 q 118.91 5 74.9 t Cs(CH3)4CH2 CH2C6Hs 2JHH=7.6 2JHH=7.6 2JHH=14.4 2JHH=14.4 1JcH=113 1JcH=125 1JcH=142 Cp*2Hf(CH2CMe3)l (11) Cp*(rJ 5,rJ 1-C5Me4CH2)Hfl (12) Cs(CH3)4CH2 rJ 5-Cs(CH3)5 rJ 5-Cs(CH3)s Cs(CH3)4CH2 2JHH=7.8 2JHH=7.8 1JcH=126 1JcH=144 Cp*(77 5,rJ 1- CsMe4CH2)Hf(CH2C6H4-m-CH3) (13) rJ 5-Cs(CH3)s Cs(CH3)4CH2 Cs(CH3)4CH2 CH2C6H4CH3 CH2C6H4CH3 CH2C6H4CH3 CH2C6H4CH3 rJ 5- Cs(CH3)s 1.90 5 1.55 d 1.88 d 1.44 5 1.55 5 1.69 5 1.96 5 -0.30 d 1.91 d 2.22 5 6.43-7.07 m 56.93 t 12.22 q 2JHH=7.5 2JHH=7.5 2JHH=13.8 2JHH=13.8 1Jc H=113 1JcH=125 22 '7 5-C5(CH3)5 C5(CH3)4CH2 117.49 s 71.09 t Cp*('7 5,'7 1-C5Me4CH2)Hf(CH2CsH4-m-CF3) (14) '7 5-C5(CH3)5 C5(CH3)4CH2 C5(CH3)4CH2 CH2CsH4CF3 '7 5-C5(CH3)5 '7 5-C5(CH3)5 CH2CsH4CF3 C5(CH3)4CH2 1.84s 1.50 d 1.89 d 1.28 s 1.47 s 1.61 s 1.83 s -0.45 d 1.87 d 6.80-7.12 m 12.10 q 117.78 s 55.93 t 71.51 t 2JHH=7.6 2JHH=7.6 2JHH=14.0 2JHH=14.0 1JcH=125 1Jcw113 1JcH=142 Cp*('7 5,'7 1-C5Me4CH2)Hf(CH2CsH4-m-NMe2) (15) '7 5-C5(CH3)5 C5(CH3)4CH2 C5(CH3)4CH2 CH2CsH4N (CH3)2 CH2CsH4N (CH3)2 CH2CsH4N(CH3)2 '7 5-C5(CH3)5 '7 5-C5(CH3)5 CH2CsH4N (CH3)2 C5(CH3)4CH2 1.91 s 1.58 d 2.03 d 1.53 s 1.55 s 1.68 s 2.04s 2.67 s -0.30 d 2.13 d 6.15-7.10 m 12.24 q 117.45 s 57.82 t 71.12 t 2JHH=7.5 2JHH=7.5 2JHH=14.0 2JHH=14.0 1JcH=125 1JcH=113 1JcH=142 Cp*2HfCH2-o-CsH3-m- CF3 (16) 17 5- C5(CH3)5 CH2CsH3CF3 CH2C5H3CF3 1.54s 2.03 br 7.87 s 7.46 d 7.34 d 3JHH=7.3 3JHH=7.3 a NMR spectra were recorded in benzene-d5 at ambient temperature. b Chemical shifts are reported in parts per million (o) from tetramethylsilane. c Coupling constants are reported in hertz. 23 formula: f.wt. 539.08 Space group P1 T= 21°C z=2 V=1161(1)A3 a = 9.084(2) A a=95.81(1) 0 b = 1 0.492(2) A {3 = 96.60(1) 0 c=12.328(1)A ,..=91.15(2) Pc = 1.543(1) g/cc 0 .A MoKa = 0.71073 A p.(MoKa) = 47.5 cm-1 F(OOO) = 540 e Wmax = 1.70 Crystal size 0.55 x 0.31 x 0.34 mm 24 Table 6. Final Parametersa Hf C11 C12 C13 C14 C15 C11M C12M C13M C14M C15M C21 C22 C23 C24 C25 C21M C22M C23M C24M C25M C31 C32 C33 C34 C35 C31M X y z Ueqb 2076.9(3) 3978(8) 31 82(8) 351 0(8) 4532(8) 4775(7) 4250(10) 2344(11) 2971 (11) 5393(10) 5889(9) 771 (9) 1629(9) 11 06(8) -15(8) -166(8) 761 (11) 2660(11) 1531(11) -1005(9) -256(9) 2395(9) 1491 (12) 1852(20) 3154(24) 43703(12) 1999(9) 1 815.0(3) 1619(8) 421 (B) -152(7) 667(7) 1776(7) 2401 (10) -213(10) -1439(9) 315(1 0) 2855(9) 3245(8) 4155(7) 4050(7) 3080(8) 2520(8) 3157(11) 5199(9) 4949(8) 2755(9) 1133(9) 2036(8) 21 00(1 0) 2879(15) 3572(14) 2770(9) 1197(8) 2536.0(2) 4177(6) 3949(6) 2918(6) 2521 (6) 3279(6) 5265(7) 4729(9) 2369(8) 1577(7) 3222(8) 3721 (7) 3240(6) 2114(7) 1879(7) 2899(7) 4942(7) 3805(8) 1320(8) 831 (8) 2876(8) -149(6) -1124(7) -1 896(9) -1746(13 9(8) 685(6) 30.6(1) 43(2) 46(2) 43(2) 41 (2) 39(2) 67(3) 73(3) 67(3) 61 (2) 62(2) 48(2) 45(2) 45(2) 44(2) 47(2) 68(3) 66(3) 61 (2) 61 (2) 58(2) a x, y, z X104 ; Ueq X103 12 b U eq -- 1/3 L"L·U··(a·*a·*)a·* I J IJ I J I " a·* J ·' aUeq -- 6- 1 <aU··fU··> II II U eq 48(2) 67(3) 107(5) 119(5) 67(2) 48(2) 25 Table 7. Selected Distances and Angles Atom Atom Dist. (A) Atom Atom Atom Angle Hf Hf Hf Hf Hf Hf Hf C21 C22 C23 C24 C25 C25M C31M C22 C25 C21M C23 C22M C24 C23M C25 C24M C25M Cp*1 Cp*2 2.412(8) 2.576(8) 2.603(8) 2.448(8) 2.258(8) 2.320(9) 2.301 (8) 1.436(11) 1.396(11) 1.519(13) 1.407(11) 1.489(13) 1.410(11) 1.501 (12) 1.460(11) 1.492(12) 1.453(12) 2.224 2.146 C22 C22 C25 C21 C21 C23 C22 C22 C24 C23 C23 C25 C21 C21 C24 C31M Cp*1 Cp*1 Cp*1 Cp* 2 Cp*2 Cp*2 C25 C25M C21 C21 C21 C22 C22 C22 C23 C23 C23 C24 C24 C24 C25 C25 C25 C25 C21M C21M C23 C22M C22M C24 C23M C23M C25 C24M C24M C24 C25M C25M C25M Cp*2 C25M C31M C25M C31M C25M 109.7(7) 125.7(7) 124.4(7) 106.6(7) 128.4(7) 124.0(7) 109.5(7) 125.0(7) 125.1 (7) 107.3(7) 127.7(7) 124.6(7) 106.5(7) 120.1 (7) 118.8(7) 101.5(3) 141.6 116.3 107.0 69.1 108.9 142.7 79.0 58.7 C21 C21 C21 C22 C22 C23 C23 C24 C24 C25 Hf Hf Hf Hf Hf Hf Hf Hf C25 Cp*2 Cp* 2 Hf Hf Cp* 1 and Cp*2 are the centroid coordinates of C11 through C15 and C21 through C25 Cp* rings, respectively. 26 Experimental Section General Considerations. All manipulations were performed using glovebox or high vacuum techniques. Solvents were purified by distillation from Na metal under an N2 atmosphere followed by vacuum transfer from "titanocene". NMR solvents were purified by vacuum transfer from activated molecular sieves (4 A, Linde) and then from "titanocene". 1H, 2H, and 13C NMR spectra were obtained by using Varian EM-390 (90 MHz, 1H) and JEOL GX4000 (400 MHz, 1H; 100 MHz, 13 C; 60.4 MHz, 2H) spectrometers. Elemental analyses were performed by Larry Henling of the CIT Analytical Laboratory. Procedures. Cp*2Hf(CH2CsH5)2 (2). Permethylhafnocene dichloride 1 (1 .40 g, 2. 70 mmol) and KCH2CsH5 (0.95 g, 7.31 mmol) were stirred overnight in 20 mL of benzene. Following filtration the benzene was replaced with 10 mL of petroleum ether. Cooling to -7S°C and filtering yielded 0.90 g (52%) of yellow 2. Anal. Calcd. for HfC34H44: C, 64.73; H, 6.97. Found: C, 64.93; H, 6.S5. The isotopically labeled bis(benzyl) compounds 2b-e were prepared in an analogous manner from the appropriately labeled starting materials. Preparation of (Cp*-d15)2Hfl2. To a large glass bomb was added Cp*2HfH2 (0.46 g, 1.0 mmol) and ca. 10 mL of benzene-d5. Deuterium gas admitted into the apparatus at -196°C. This solution was heated to 115°C with stirring for 5 days. The bomb was recharged daily with fresh deuterium gas. At the end of 5 days excess CH3l was condensed into the bomb resulting in the rapid formation of (Cp*-d15)2Hfl2, which was isolated from petroleum ether (0.39 g, 55%). The total isotopic purity was ~95% (1H NMR). Cp*2Hf(CH2CsH4-m-CH3)2 (3). The diiodide derivative Cp*2Hfl2 (1.07 g, 1.52 mmol) and KCH2CsH 4-m-CH3 (0.69 g, 4.79 mmol) were stirred overnight in ca. 20 mL benzene. After filtering, the benzene was replaced with ca. 10 mL of petroleum ether. Cooling to -7S°C and 27 filtering yielded 0.4S g (4S%) of pale yellow 3. Anal. Calcd. for HfC35H4a: C, 65.59; H, 7.34. Found: C, 66.31; H, 7.54. Cp*2Hf{CH2C5H4-m-NMe2)2 (4). Compound 1 (1 .67 g, 3.21 mmol) and 1.4 g (S.09 mmol) of KCH2C5H4-m-NMe2 were stirred overnight in ca. 20 ml benzene. After filtering, the benzene was replaced with ca. 10 ml of petroleum ether. Cooling to -7S°C and filtering yielded 1.75 g {71%) of pale yellow 4. Anal. Calcd. for HfC3aH54N2: C, 63.57; H, 7.66; N, 3.90. Found: C, 63.S4; H, 7.52; N, 3.92. Cp*2HfCHro-C5H4 {5). To a small glass bomb was added 1.17 g (1.S5 mmol) of 2 and ca. 1 0 ml of methylcyclohexane. The solution was heated to 140°C for 6 days. Isolation from petroleum ether yielded 0.61 g of pale yellow 5 (61%). Anal. Calcd. for HfC27H35: C, 60.62; Hf, 6.77. Found: C, 60.S6; H, 6.S4. Cp*2HfCH2-o-C5H3-m-NMe2 (7). To a small glass bomb was added compound 4 (0.75 g, 1.0 mmol) and 5 ml of toluene. This solution was heated for 2 days at 145°C. Isolation from -7S°C petroleum ether afforded 0.1S g of pale yellow 7 (30%). Anal. Calcd. for HfC29H41 N: C, 59.S3; H, 7.10; N, 2.41 . Found: C, 59.31; H, 7.37; N, 2.23. Cp*(?J 5,?J 1-CsMe4CH2)HfCH2C5H5 (9). Cp*(?J 5,?J 1-CsMe4CH2)Hfl (O.S2 g, 1.4 mmol) and KCH2C5H5 (0.19 g, 1.5 mmol) were stirred for 1 hour in 10 ml of benzene. After filtering, the benzene was replaced with 10 ml of petroleum ether. Cooling to - 7S°C and filtering yielded 0.42 g (52%) of yellow orange 9. Anal. Calcd. for HfC27H35: C, 60.16; H, 6. 73. Found: C, 60.47; H, 6.74. Cp*2Hf(CH2CMe3)l (11 ). Permethylhafnocene diiodide (0.94 g, 1.5 mmol) and LiCH2CMe3 (0.13 g, 1.6 mmol) were stirred overnight in benzene. Following filtration the benzene was replaced with 10 ml of petroleum ether. Cooling to -7S°C and filtering afforded 0.74 g 28 (S5%) of yellow Cp*2Hf(CH2CMe3)l. Anal. Calcd. for HfC25H41I: C, 46.44; H, 6.39. Found: C, 46.15; H, 6.16. Cp*('7 5,'7 1-C5Me 4CH2)Hfl (12). To a small glass bomb was added Cp*2Hf(CH2CMe3)l (4.24 g, 6.54 mmol) and 30 ml of toluene. This solution was heated at 140°C for 2 days. Isolation from petroleum ether at -7S°C yielded 2.1 g (56%) of orange Cp*('7 5,'7 1-C 5Me4CH2)Hfl. Anal. Calcd. for HfC2oH2gl: C, 41 .79; H, 5.0S. Found: C, 42.27; H, 5.07. Cp*('7 5,'7 1-C5Me4CH2)Hf(CH2CsH4-m-CH3) (13). Cp*('7 5,'7 1-C5Me4CH2)Hfl (0.90 g, 1.6 mmol) and KCH2CsH4-m-CH3 (0.33 g, 2.29 mmol) were stirred for 1 hour in 20 ml of benzene. Following filtration the benzene was replaced with 5 ml of petroleum ether. Cooling to -7S°C and filtering yielded 0.50 g (56%) of yellow orange 10. Anal. Calcd for HfC2aH3a: C, 60.S4; H, 6.S7. Found: C, 5S.SO; H, 6.65. Cp*('7 5,'7 1-C5Me4CH2)Hf(CH2CsH4-m-CF3) (14). Cp*('7 5.'7 1-C5Me4CH 2)Hfl (0.75 g, 1.30 mmol) and 1.1 equivalents of CIMgCH2CsH4-m-CF3 were stirred in 10 ml of diethyl ether for 1 hour. The solvent was replaced with petroleum ether and the solution was filtered. Cooling to -7S°C and filtering yielded 0.57 g (72%) of yellow orange 11. Anal. Calcd. for HfC28H 35 F3: C, 55.31; H, 5.97. Found: C, 54.49; H, 5.77. Cp*('7 5,'7 1-C5Me4CH2)Hf(CH2CsH4-m-NMe2) (15). Cp*('7 5,'7 1-C5Me4CH2)Hfl (O.S1 g, 1.4 mmol) and KCH2CsH4-m-NMe2 (0.25 g, 1.46 mmol) were stirred for 1 hour in 10 ml of benzene. After filtering the solvent was replaced with petroleum ether. Cooling to -7S°C and then filtering afforded 0.37 g (45%) of yellow orange 12. Anal. Calcd. for HfC29H41 N: C, 59.S3; H, 7.10; N, 2.41. Found: C, 60.23; H, 7.17; N, 2.19. Kinetic Measurements of Thermal Decomposition. With one exception, all reactions were followed by monitoring the decrease in height of the pentamethylcyclopentadienyl 29 resonance of starting complex relative to internal (77 5-CsHs)2Fe. The rate of decomposition of 2b was followed by monitoring the decrease in integrated intensity of the (Cp*-d15)2Hf(CH2CsHs)2 signal relative to ('7 5-CsHs)2Fe. NMR spectra were recorded at timed intervals using a Varian EM-390 (90 MHz) spectrometer. Reaction temperatures were maintained using constant temperature oil baths controlled by thermoregulators and were observed to be constant to within ±1 °C. A typical experiment involved 25 mgs of hafnium complex and 6-1 0 mg of ferrocene dissolved in 0.4 ml of benzene-ds. The NMR tubes were sealed containing >1 atm of dinitrogen. Total submersion in the oil bath prevented the benzene-ds from refluxing at the elevated temperatures used in this study. Peak heights were demonstrated to be reproducible to within ±7% by repeated measurement. Each spectrum was recorded 3 times and the average peak height was used to calculate the values of k given in Tables 1,2 and 4. Errors in k represent approximately 1 standard deviation. Plots of ln(k!T) vs. 1{T were constructed from the decomposition rates and yielded the activation parameters ~Hi and ~st . Errors in ~Hi and ~s* represent 1 standard deviation calculated by error propagation of 1 standard deviation changes in the value of -~Hi/Rand ln(A!T)-1, when the measured k values are varied within their error limits. X-ray Structure Determination. Crystals of ('7 5-CsMes)('7 5,'7 1-CsMe4CH2)HfCH2CsHs (9) were obtained by slow evaporation of a saturated diethyl ether solution and sealed in glass capillaries under dinitrogen to prevent decomposition. Oscillation and zero-, first-, and second-layer Weissenberg photographs showed no evidence of twinning, and the crystal was mounted on a CAD4 diffractometer equipped with graphite-monochromated MoKa radiation. A least-squares fit to the setting angles of 25 reflections (20 of which were independent of one another) with 16°<20<26° indicated a triclinic cell and yielded the cell parameters. As neither nontrivial symmetry nor systematic absences were observed in the 30 photographs, and since an approximate calculation indicated two molecules per cell , P1 symmetry was assumed. All data (±h, ±k, ±I) with 2°<20<52° were collected, and the hkO reflections were collected twice. These data were merged to give 4529 independent observations. Three index reflections monitored every 10,000 seconds of X-ray exposure time showed a linear decay of 1.5% in F over the 136 hours required to collect the data. The data were corrected for this decay, as well as for Lorentz and polarization factors, but no correction was made for absorption due to the rounded shape of the crystal, which made an accurate representation of its shape impractical. Solution and Refinement. A Wilson plot was used to place the data on an approximately absolute scale. Variances, a 2 (1) , were assigned to the intensities on the basis of counting statistics plus an additional term, (0.014 1) 2 , to account for additional errors proportional to the intensity. A Patterson map yielded hafnium coordinates and successive structure factor-Fourier calculations located the remaining atoms. Refinement w ith all the heavy atoms anisotropic converged w ith R = ~ II Fo 1-1Fcii;L;I Fo I = 0.048 for all4455 reflections w ith Fo2 > 0 and 0.046 forthe 4187 reflections with Fo2 > 3a (Fo2 ) . The goodness of fit, [~(F02-F02 )/(n-p)] 1 12 , where n =number of data= 4529 and p =number of parameters= 254, is 4.37[27]. Hydrogen atoms were introduced at idealized positions during the refinement and repositioned two or three times, depending on the accuracy of the initial placement. Non-methyl hydrogen positions were calculated in accordance with expected bond geometry; methyl hydrogens were placed on the basis of Fourier maps in the plane where they were expected. At first the maps did not clearly s uggest all hydrogen positions, but by the last cycle the idealized positions where the hydrogens were placed agreed fairly closely with the difference maps. All hydrogens were assigned isotropic temperature factors 1 .0 A2 greater than the equivalent isotropic factors of the parent carbons. In the final least squares cycle no parameter shifted by more than 0.03 of its standard deviation. Crystal data are given in Table 5 and final parameters in Table 6. Calculations were done using the 31 programs of the CRYM crystallographic computing system£2 8] on a VAX 11/750 computer; the drawing was done using ORTEP£291 . References (1) (a) For a recent review, see: Rothwell, l.P. Polyhedron 1985,4, No.2, 177-200. (b) Thompson, M.E.; Baxter, M.B.; Bulls, A.R.; Burger, B.J.; Nolan, C.N.; Santarsiero, B.D.; Schaefer, W.P.; and Bercaw, J.E., submitted for publication in J. Am. Chern. Soc. (c) Bruno, J.W.; Smith, G.M.; Marks, T.J.; Fair, C.K.; Schultz, A.J.; Williams, J.M. J.Am. Chern. Soc., 1986, 108,40. (2) See, for example: (a) Bruno, J.W.; Marks, T.J.; Day, V.W. J. Am. Chern. Soc. 1982, 104, 7357-7360. (b) Fagan, P.J.; Manriquez, J.M.; Maatta, E.A.; Seyam, A.M.; and Marks, T.J. J.Am. Chern. Soc. 1981, 103,6650-6667. (c) Steigerwald, M.L. and Goddard, W.A. J. Am. Chern. Soc., 1984, 106, 308. (3) See for example, (a) Stock, L. "Aromatic Substitution Reactions," Prentice-Hall, Englewood Cliffs, N.J., 1968. (b) Olah, G.A. "Friedei-Crafts and Related Reactions," Vol. I-IV, Wiley-lnterscience, New York, N.Y., 1963-1964. (c) Thompson, M.E.;Bercaw, J.E. Pure and Appl. Chern., 1984, 56, 1. (4) Latesky, S.L.; McMullen, A.K.; Niccolai, G.P.; Rothwell, l.P.; and Huffman, J.C. Organometallics, 1985, 4, 902. (5) Mintz, E.A.; Meloy, K.G., Marks, T.J.; and Day, V.W. J. Am. Chern. Soc. 1982, 104, 4692-4695. (6) (a) Davies, G.R.; Jarvis, J.A.J.; Kilbourn, B.; Pioli, A.J.P. J. Chern. Soc., Chern. Commun. 1971, 677. (b) Bassi, l.W.; Allegra, G.; Scordamaglia, R.; Chioccola, G.J. J. Am. Chern. Soc., 1971, 93, 3787-3788. (c) Davies, G.R.; Jarvis, J.A.J.; Kilbourn, B. J. Chern. Soc., Chern. Commun. 1971,1511-1512. 32 (7) Aoyama, Y.; Yoshida, T.; Sakurai, K.; and Ogoshi, A. J. Chern. Soc., Chern. Commun. 1983,478. (8) Shul'pin, G.B.; Nizova, G.V.; and Nikitaev, A.T. J. Orgmet. Chern. 1984, 276, 115-153. (9) (a) Parshall, G.W. Ace. Chern. Res. 1970, 3, 139. (b) Bruce, M.l.; Goodall, B.L.; and Stone, F.G.A. J. Chern. Soc., Chern. Commun. 1973, 558. (10) Deeming, A.J. and Rothwell, I.P. Pure andAppl. Chern. 1980,52,649-655. (11) Sweet, J.R. and Graham, W.A.G. J. Am. Chern. Soc. 1983, 105, 305-306. (12) Jones, W.O. and Feher, F.J. J. Am. Chern. Soc. 1984, 106, 1650-1663. (13) Prepared by the dropwise addition of n-C4H9Li to a suspension of KOCMe3 in the appropriately substituted toluene. (14) For leading references to related metallacyclobutanes see: (a) Grubbs, R.H. Comprehensive Organometallic Chemistry, Wilkinson, G; Stone, F.G.A., Abel, E.W., Eds.; Pergamon Press, Vol. 8, 499. (b) Bickelhaupt, F. Pure and Appl. Chern. 1986, 58(4), 537-542. (c) Seetz, J.W.F.L.; Schat, G.; Akkerman, O.S.; and Bickelhaupt, F. Angew. Chern. Int. Engl. Ed. 1983, 22, 248. (15) Admittedly, the phenonium ion intermediate or transition state for this mechanism would be very strained. (16) The approximate ratios of isotopically labeled toluenes were determined by integration (1H NMR) of the residual methyl signal. (17) McDade, C.; Green, J.C.; and Bercaw, J.E. Organometallics 1982, 1, 1629. (18) Chamberlain, L.; Rothwell, I.P.; and Huffman, J.C. J. Am. Chern. Soc. 1982, 104, 7338-7340. (19) ('7 5-C5Me5)2 Hf(CaH5)1 does not liberate CaHa (to generate 12) when thermolysed (145° C) for 4 days in c 6 o6 . Furthermore, deuterium is not incorporated (2 H NMR) into 12 when it is heated for extended periods in CaDa. 33 (20) Johnson, C.D. "The Hammett Equation"; Cambridge University Press: Cambridge, 1973. (21) For the purpose of constructing a Hammett plot, the rate constant for 9 (X= H) rearranging to 5 was corrected for having two kinetically available ortho-benzyl hydrogens, i.e. the value of k (X= H) was halved. (22) Lowry, T.H. and Richardson, K.S. "Mechanism and Theory in Organic Chemistry"; Wasserman, M., Castellano, E., Eds.; Harper & Row, Publishers: New York, 1981; pp 559-580. (23) Schock, L.E.; Brock, C.P.; Marks, T.J. Organometaflics 1987,6,232. (24) Bandy, J.A.; Mtetwa, V.S.B.; Prout, K.; Green, J.C.; Davies, C.E.; Green, M.L.H.; Hazel, N.J.; Izquierdo, A.; and Martin-Polo, J.J. J. Chern. Soc. Dalton Trans. 1985, 2037, and references therein. (25) 8 is defined as the angle by which the exocyclic bond of the fulvene ligand deviates from planarity with the five ring carbons of the cyclopentadienylligand. This deviation is towards the metal center. (26) There are no indications of interaction between the hafnium and the aryl ring of the benzyl ligand (i.e. an 77 3-CH2C5H5 interaction) for 9, in contrast to the structures of other, analogous benzyl complexes (vida supra) . (27) The absorption correction was only approximate because of the irregular shape of the crystal. Because of this, there are large excursions near the hafnium atom in the final difference map (+5.05, -3.43 e A3). The high value of the goodness of fit probably results from this also. We believe the weights were chosen properly; a larger "fudge" factor for them (such as 0.05, rather than the 0.02 we used) would improve the goodness of fit, but such a high value is not indicated by the data. (28) Duchamp, D.J. CRYM Crystallographic Computing System. Am. Crystallogr. Assoc. Meeting, Bozeman, Montana, 1964. Paper B14, p. 29. 34 (29) Johnson, C.K. ORTEP, Report ORNC-3794. Oak Ridge National Laboratory, Tennessee, 1965 . 35 Chapter 2 Synthesis and Reactivity of Singly and Doubly Metallated Cp* Containing Derivatives of Permethylhafnocene: The Molecular Structure of Cp*(tJ 5 ,tJ 1 ,tJ 1 -CsMe3(CH2)2)Hf. 36 Abstract: Thermolysis of the singly metallated complex Cp*('7 5 ,?J 1-CsMe4CH2)Hf(CH2CMe3) in toluene affords neopentane and the doubly metallated complex Cp*(?J 5,?J1,?] 1-CsMe3(CH2)2)Hf. The structure of Cp*(?J 5,?J 1,?]1-CsMe3(CH2)2)Hf, as determined by single-crystal X-ray diffraction techniques, confirms that metallation has occurred at adjacent methyl groups of the same pentamethylcyclopentadienyl ring. Cp*(?J 5,'7 1,'7 1-CsMe3(CH2)2)Hf crystallizes in the space group P2/c (a= 13.775(4) A, b = 9.516(1) A, c = 14.183(6) A; f3 = 103.965(31) o, z = 4). Least squares refinement led to a value for R of 0.036 (/ > 3ai) and a goodness-of-fit of 2.66 for 1984 independent reflections. Cp*(?J 5,'71,'7 1-CsMe3(CH2)2)Hf and Cp*(?J 5,?J 1-CsMe4CH 2)Hfl cleanly insert one equivalent of ethylene into the hafnium methylene carbon bond to form the propyl bridged species Cp*(17 5 ,'7 1,?J 1-CsMe3(CH2)(CH2CH2CH2))Hf and Cp*('7 5,?J 1-CsMe3CH2CH2CH2)Hfl, respectively. Cp*(?J 5 ,?J 1,?J1-C5 Me3 (CH 2)(CH2CH2CH2))Hf reacts with excess ethylene to cleanly afford Cp*('7 5 ,'7 1,'7 1-CsMe4CH2CH2CHCH2CH2)Hf. Deuterium labeling experiments suggest that this reaction occurs via an a-H abstraction to generate the alkylidene intermediate Cp*(?J5,'71_CsMe4CH2CH2CH=)Hf, which rapidly traps ethylene to form the observed product. The metallated phenyl complex Cp*(?J 5 ,?J 1-CsMe3CH2)HfC5Hs has been shown to be in equilibrium with the benzyne complex Cp*2Hf(?J 2-C5H4). Heating benzene-d5 solutions of Cp*(?J 5,'7 1-CsMesCH2)HfC5Hs in the presence of ethylene traps the benzyne intermediate and generates the hafnacyclopentene Cp*2HfCH2CH2C5H4. 37 Introduction The 77 5-pentamethylcyclopentadienyl (Cp*) ligand is widely employed in organometallic chemistry and in general provides a stable ligand environment for reaction chemistry at metal centers. The steric bulk and the absence of activated sp2 C-H bonds in this ligand has allowed for the formation of stable monomeric metallocene hydride and alkyl derivatives of the early transition metals.£ 11 It has become recognized, however, that the Cp* ligand is not restricted to being a simple ancillary ligand, but is able to participate in reaction chemistry by metalation of its methyl C-H bonds. Several complexes containing metallated Cp*-containing derivatives have now been isolated. Bercaw and coworkers described the formation of Cp*(77 5,77 1-C5Me4CH2)TiCH3 which is formed upon thermolysis of Cp*2TiMe2.£2 1 The dime ric titanium complex, [(Cp*Ti)2-J.L-('7 5,17 1-C5Me4CH2HJ.L-0)2], reported in 1985 by Bottomley, et al., is the first example of a crystallographically characterized compound containing a 17 5,17 1 -C5Me4CH2 moeity derived from a Cp* ligand.£31 Additionally, complexes containing metallated pentamethylcyclopentadienyl ligands have been implicated as intermediates in H/D exchange reactions of Cp* ligands£4 1, in the elimination of isobutane from Cp*2Zr(CH2CHMe2)H[S1, in the conversion of Cp*2Hf(CH2CBH5)2 to Cp*2HfCH2-o-C6H4 and toluene£6 1, and in the reaction of Cp*2Sc-CH3 with benzene to give methane and Cp*2Sc-C6H5.£ 71 In light of the widespread use of pentamethylcyclopentadienylligands in transition metal chemistry, a more thorough investigation into its reactivity seems warranted. Herein are described some mechanistic aspects of the synthesis and reactivity of hafnium complexes containing singly or doubly metallated pentamethylcyclopentadienyl ligands. 38 Results and Discussion The metallated iodide complex Cp*('7 5.'7 1-CsMe4CH2)Hfl (1) described in Chapter 1 serves as a convenient starting material for a wide range of mono-metallated permethylhafnocene derivatives of the type Cp*('7 5 ,'7 1-CsMe 4CH2)Hf-R (R =alkyl or aryl). Treatment of Cp*('7 5 ,'7 1-CsMe4CH2)Hfl with neopentyllithium cleanly affords Cp*(77 5 ,'7 1-CsMe4CH2)Hfl 1 Cp*(77 5 ,'7 1-CsMe4CH2)Hf(CH2CMe3) 2 (1) undergoes a facile intramolecular rearrangement to Cp*2HfCH2-o-CsH4 [6], 2 exhibits considerable thermal stability and decomposes only at elevated temperatures. Toluene solutions of 2 cleanly eliminate neopentane over the course of 1 day at 140 o to produce the doubly metallated permethylhafnocene derivative Cp*(17 5,'7 1 ,17 1-CsMe3(CH2)2)Hf (3) (equation 2) . toluene 2 + 3 Complex 3 can be isolated in moderate yield from cold petroleum ether as air sensitive orange crystals. Teuben et al. have reported that the Ti and Zr analogs of 3 cEm be prepared in refluxing mesitylene solutions of (17 5- CsMes)2M(CH3)2[B] (conditions in which ('7 5 -c5 Me5 ) 2Hf(CH3)2 shows no evidence of decomposition after 3-5 days). The tungsten 39 analogue of 3 has been prepared by irradiation of Cp*2WH2.l9l That metalation occurs on adjacent methylene groups of the same Cp* ligand in 3 is supported by 1 H and 13c NMR data (Table 1). In order to confirm the unusual structure of Cp*(f7 5,f7 1 ,T/ 1-C5Me3(CH2)2)Hf and because of the scarcity of structural information for metallated pentamethylcyclopentadienyl containing derivatives in general, the X-ray crystal structure determination of 3 was carried out. The molecular configuration of 3 is shown in Figure 1 and a skeletal view is shown in Figure 2. Cp*(f7 5,J7 1 ,J7 1 -C5Me3(CH2)2)Hf possesses a mirror plane bisecting the two metallated methylene groups. The two methylene fragments are bent 41 o out of the plane formed by the 5 inner ring carbons of the cyclopentadienylligand (Table 3), a value consistent with other structurally characterized transition metal complexes containing fulvene-like ligands. [ 1 0] The Hf-C distances for the two metallated methylene groups average 2.346(11) A, a slight elongation compared to other Hf- C sigma bond distances. For example, distances of 2.280(13) and 2.295(14) A have been determined for the hafnium-carbon sigma bonds of Cp*2Hf(OO-tBu)(CH2CH3)[ 11 l and [Cp2HfMe] 2o[12], respectively. The averaged hafnium-methylene carbon-ring carbon angle (Hf-C(21 M)-C(21) and Hf-C(22M)-C(22)) is 65 o. Metalation of the two methyl groups also results in the hafnium "sliding forward" within the cyclopentadienyl ring system. The hafnium-ring carbon distances of the metallated Cp* vary from an average of 2.176 Afor the two carbons bearing the methylene groups (Hf-C(21) and Hf-C(22)) to 2.615(8) Afor the most distant ring carbon (Hf-C(24)). An additional consequence of double metalation is the near parallelism of the cyclopentadienyl rings, the dihedral angle between the two planes being only 12(4) o. 40 Figure 1. The molecular configuration of Cp*(ry 5,ry 1 ,ry 1 -C5Me3(CH2)2)Hf. The thermal ellipsoids are shown at the 50% probability level. 41 Figure 2. A skeletal view of Cp*('7 5 .'7 1 ,f7 1 -C5 Me 3 (CH2 ) 2 )Hf. The nonmetallated Cp* ligand has been omitted for clarity. 42 The hafnium-methylene carbon bonds of 1 and 3 exhibit enhanced reactivity as compared to typical permethylhafnocene-alkyl bonds. This heightened reactivity is presumably a consequence of the unusual bond angles and lengths of the metallated methylene groups and the resulting steric accessibility of the hafnium center. Thus, benzene solutions of 1 and 3 react with dihydrogen within seconds to generate the nonmetallated derivatives ('7 5-CsMes)2Hf(H)I and (17 5-CsMes)2HfH2, respectively. This contrasts with the hydrogenation of ('7 5-CsMes)2Hf(CH3)2, which requires temperatures of 140° C for several days before generating (17 5-CsMes)2HfH2 and methane. The hafnium-methylene bonds of 1 and 3 readily insert one equivalent of ethylene or acetylene. Heating benzene solutions of 1 in the presence of ethylene quantitatively (1H NMR) forms the propyl bridged iodide complex 4 (equation 3). (3) 1 4 Compound 4 has been characterized by 1H NMR, 13C NMR and microanalytic al data. As expecte d, the 1H and 13C NMR spe ctra of 4 exhibit four inequivalent methyl resonances for the (ry 5 ,ry 1-CsMe4CH2CH2CH2) ligand. Each of the six propyl bridge hydrogens are diastereotopic and give rise to six distinct multiplets in the 1H NMR spe ctrum. These resonances were not assigned unequivocally to specific hydrogens of the propyl bridge. The proton coupled 13C NMR sp ectrum of 4 contains three trip lets at o 26.71 CJc H = 127 Hz), o 39.29 (1JcH = 123Hz), and o 70.33 (1JcH =127 Hz) which are attributed to the propyl 43 bridge carbons of the ('7 5 ,'7 1 -CsMe4CH2CH2CH2) ligand. Complex 4 has been independently prepared by treating Cp*('7 5 ,'J 1-CsMe4CH2CH2CH2)HfH[ 13l with excess methyl iodide (equation 4). CH3I Cp*('7 5 ,'7 1-CsMe4CH2CH2CH2)HfH - - - - - - i > Cp*('7 5 .'7 1-CsMe4CH2CH2CH2)Hfl 7 - CH4 4 (4) Green and coworkers have utilized alternative methods to prepare tungstenocene and molybdocene derivatives which contain ethyl or butyl linkages between a cyclopentadienyl ligand and the metal center.£14] The reaction of the metallated iodide complex 1 with acetylene proceeds in melting benzene to cleanly generate the propylene bridged iodide complex 5 (equation 5). No + (5) HC ==CH 1 5 metathesis of the acetylene C-H bond with the Hf-C bond of 1 is observed. This lack of reactivity contrasts with that of compounds of the type ('7 5-CsMes)2ScR (R =alkyl or H) which undergo sigma bond metathesis with terminal acetylenes to give R-H and permethylscandocene acetyl ide derivatives. [ 1 5] 44 The doubly metallated complex 3 slowly (2 weeks at 35°C) incorporates one equivalent of ethylene to form the mixed methylene bridged/propyl bridged species 6 (equation 6). (6) 6 3 The titanium analog Cp*(f7 5 ,f7 1,f7 1-CsMe3(CH2)2)Ti undergoes a similar insertion of methyl phenyl ketone to yield a titanium alkoxide species.£8] Incorporation of a second equivalent of ethylene is achieved at slightly higher temperatures. An isolated sample of complex 6 (7) 6 7 (or complex 3) cleanly generates the novel a-substituted metallacycle 7 (equation 7) when heated in the presence of excess ethylene. The structural assignment of 7 is based on NMR evidence which includes 5 new carbon resonances (4 triplets and 1 doublet) in the 1H coupled 13C NMR attributed to the 3 metallacycle ring carbons and the 2 carbon bridge linking the cyclopentadienyl ring to an alpha carbon of the metallacyclobutane ring. Two potential mechanisms for the formation of 7 are outlined in Scheme 1. Mechanism A requires a rate-determining bimolecular insertion of ethylene into the 45 hafnium carbon bond of the propyl bridge of 6 to form a transient pentyl bridged species A, followed by a rapid gamma-H abstraction by the adjacent methylene bridge to yield 7. Thus, mechanism A predicts a linear dependence of the reaction rate on the concentration of ethylene. Mechanism B involves the initial abstraction of an a-H atom of the propyl bridge by the neighboring methylene group to generate a intermediate alkylidene species B. Assuming intermediate B can be efficiently trapped by ethylene, a rate independent of the ethylene concentration w ill be observed. It is found experimentally that the rate of conversion of 6 to metallacycle 7 obeys first order kinetics in the presence of a large excess of ethylene and is independent of the ethylene concentration over the range 0.1 M to 0.4 M (k =3.6±0.1 x 1o-6 sec-1 at 82 C). Additionally, a substantial kinetic deuterium isotope o effect is observed when Cp*(, 5,,1,, 1-C5Me3(CH2)(CH2CD2CD2))Hf, 6-d4 (prepared from 3 and ethylene-d4), is converted to 7 in the presence of ethylene (kH/ko =6±1 at 82 ° C). The observation of the kinetic deuterium isotope effect along with the pseudo-first order kinetic behavior for the conversion of 6-d4 to 1-d4 in the presence of C2H4 also speaks against a pre-equilibrium involving reversible ethylene elimination from the propyl bridge of 6 to reform the methylene bridge of 3. Furthermore, the reaction of 13C labeled ethylene with 6 generates metallacycle 7 in which the labeled carbon is incorporated exclusively into the primary a carbon and the f3 carbon of the metallacyclobutane ring. The data are most consistent with mechanism B, i.e., the rate of conversion of 6 to metallacycle 7 is controlled by the abstraction of an a-hydrogen off the propyl bridge by the adjacent methylene group. 46 Mechanism A A 6 1-H~t ~a-~ abstr ~low Mechanism B 8 Scheme 1. Two pote ntial m echanisms for the formation of 7 from 6 and ethylene. 47 It can be demonstrated, however, that the abstraction of an adjacent a-hydrogen by the methylene group of a metallated Cp* ligand is not a general reaction. Thermolysis of the mixed alkyl complex Cp*2Hf(CH2CMe3)CH3 (8) (prepared from Cp*2Hf(CH2CMe3)l and methyllithium), cleanly generates neopentane and the mono-metallated methyl complex 140°C Cp*2Hf(CH2CMe3)CH3 - - - - t > Cp*(77 5 ,771-CsMe4CH2)HfCH3 + C(CH3)4 8 CsDs 9 (8) Repeating the thermolysis with Cp*2Hf(CH2CMe3)CD3 (8-d3) (prepared from Cp*2Hf(CH2CMe3)l and LiCD3) yields 9 and neopentane-d1 (identified by comparison of its infrared spectra with an authentic sample of neopentane-d1). The 1H NMR spectrum of the resulting labeled 9 contains a quartet (2JoH =2Hz) at -1.1 ppm, consistent with the formation of Cp*(77 5.77 1-CsMe4CH2)Hf-CD2H (9-d2)· These results suggest that the thermal decomposition of 8 proceeds via an a-hydrogen abstraction of a methyl hydrogen by the adjacent neopentyl ligand to form a transient methylidene complex of permethylhafnocene, followed by the rapid transfer of a ring methyl C-H bond across the hafnium methylidene double bond (Scheme 2). Continued thermolysis (1 day at 140°) of Cp*(77 5 ,77 1-CsMe4CH2)HfCD2H fails to produce detectable concentrations of a-hydrogen abstraction of a methyl group hydrogen (or deuterium) by the adjacent methylene group is not in this case facile (i.e., the metallated methyl derivative is not in equilibrium with the permethylhafnocene methylidene complex). 48 8 9 ~- neopentane a-H abstr ~ Scheme 2 . Proposed mechanism for the formation of Cp*('7 5,17 1-C5Me4CH2)HfCH3 (9) The lack of reactivity shown by the metallated Cp* ligands of complexes derived from permethylhafnocene towards arene solvents, e.g. benzene and toluene, is quite surprising. Indeed, the doubly metallated complex 3 is prepared by the thermolysis of 2 in toluene. Furthermore, the lack of deuterium incorporation into the metallated Cp* ligand of 3 when thermolyzed for 1 week at 140°C in benzene-de speaks against reversible solvent activation. In contrast, the metallated permethylscandocene derivative synthesize d as shown below (equations 9-11 ). 49 Cp*2Hf(CsHs)l + Lil 10 Cp*2Hf(CsHs)l + LiCH3 10 Cp*2Hf(CsHs)CH3 11 (9) Cp*2Hf(CsHs)CH3 + Lil 11 (1 0) Cp*(77 5 ,77 1-CsMe4CH2)HfCsHs + CH4 12 (11) The zirconium analogue of 12, prepared by the thermal elimination of benzene from Cp*2Zr(CsHs)2, has been recently reported.£ 161 Complex 12 does not lose benzene or undergo solvent exchange upon prolonged thermolysis (days at 140 C) in benzene-ds. The methane produced upon thermolysis of the deuterium-labeled complex Cp*(77 5 -CsMes)Hf(CsDs)CH3 is exclusively CH3D (1H NMR), consistent with the abstraction of an ortho-phenyl hydrogen by the adjacent hafnium-methyl fragment to produce methane and a transient benzyne complex of permethylhafnocene. Rapid addition of a ring methyl hydrogen to this benzyne intermediate would result in the observed metallated phenyl complex. Transient benzyne derivatives of group IV metallocenes have ample precedent in the literature [17] and the trimethylphosphine-stabilized benzyne derivative Cp 2Zr(77 2-C6 H4)PMe3 has been crystallographically characterized.£ 181 The remaining deuteriums in 12-d4 (derived from the thermolysis of Cp*(775 -CsMes)Hf(CsDs)CH3 in cyclohexane-d12) do not reside exclusively in the phenyl ligand, but are distributed into all positions of the complex (2H NMR). This implies the existence of an equilibrium between 12 and the proposed benzyne intermediate (77 5 -CsMes)2Hf(77 2- CsH4) (equation 12). Such an equilibrium, combined with rotation about 50 (12) 12 the Hf-Cphenyl bond of 12, provides a plausible mechanism for the complete exchange of hydrogen positions within the molecule.£ 1 9] Furthermore, heating 12 in the presence of ethylene does not result in the formation of a propyl-bridged phenyl complex analogous to Cp*('7 5 ,TJ 1 -CsMe4CH2CH2CH2)Hfl4, but instead quantitatively generates (ry 5-CsMes)2Hf(CH2CH2C6H4) 13, the product expected from the trapping of (ry 5-CsMes)2Hf('7 2-C6H4) with ethylene (Scheme 3). 51 12 \ Scheme 3. Proposed mechanism for the formation of metallacycle 13 from the monometallated phenyl derivative 12 and ethylene. Thus, strong evidence is provided for a reversible hydrogen transfer between the pentamethylcyclopentadienyl ligand of ('7 5-CsMes)2Hf('7 2-CsH4) and the ortho-phenyl position of Cp*('7 5 .17 1-CsMe4CH2)HfCsHs. 52 Conclusions In summary, the syntheses and reactivities of singly and doubly metallated pentamethylcyclopentadienyl-containing derivatives of permethylhafnocene have been investigated. Metallation of Cp* ligands has been observed to occur (i) directly, via the coupling of a Cp* hydrogen with a hafnium alkyl, e.g. the formation of Cp* (77 5,77 1-C5Me4CH2) Hfl (1) and neopentane from Cp*2Hf(CH2CMes) I, or (ii) indirectly, by the intramolecular rearrangement of unstable species such as [Cp*2Hf=CH2] or [Cp*2Hf(77 2-CsH4)]. The Hf-Cmethylene bonds of Cp*(77 5,77 1 -C5Me4CH2)Hfl (1) and Cp*(l'/ 5,1'} 1 ,1'} 1 -C5Mes(CH2)2)Hf (3) exhibit enhanced reactivity towards suitable substrate molecules as compared to typical permethylhafnocene-alkyl bonds. This heightened reactivity is presumably a consequence of the unusual bond angles and lengths of the metallated methylene groups and the resulting steric accessibility of the hafnium center. For example, ethylene readily inserts into the Hf-Cmethylene bonds of 1 and 3 to form the propyl-bridged species Cp*(77 5,77 1-C5Me4CH2CH2CH2)Hfl (4) and Cp*(17 5,77 1,77 1-CsMes(CH2)(CH2CH2CH2)Hf (6), respectively. Compound 6 reacts with a second equivalent of ethylene to form Cp*(77 5 ,171 ,771_CsMe4CH2CH2CHCHi;H2)Hf (7). Kinetic data suggest that this reaction proceeds by a rate determining transfer of an a-hydrogen from the propyl bridge of 6 to the adjacent metallated methylene unit and that the resulting alkylidene species is trapped by ethylene. Intramolecular C-H bond activation is also observed to occur in the mono-metallated phenyl compound Cp*(17 5,77 1-C5Me4CH2)HfC6Hs (12), which is in equilibrium with the benzyne adduct [Cp*2Hf(77 2-CsH4)]. This benzyne adduct is trapped by ethylene to form Cp* 2HfCH2CH 2CsH4. In the absence of suitable substrates, however, metallated derivatives of permethylhafnocene exhibit remarkable thermal stability. Thus, benzene solutions of 1 or 3 are unchanged after two weeks at 140 o C. 53 Table 1. 1H and 13c NMR Data.a Compound Assignment 8(ppm) Coupl ingb Cp*('7 5,'7 1-CsMe4CH2)HfCH2CMe3 (2) '7 5·Cs(CH3)5 '7 5,'7 1-C5Me4CH2 Hf-CH2C(CH3)3 Hf-CH2C(CH3)3 1.93 s 1.53 d 1.83 d 1.58 s 1.75 s 1.98 s 2.17 s -1 .58 d 0.37 d 1.13 s 2JHH = 7.5 2JHH = 7.5 2JHH = 13 2JHH = 13 Cp*('7 5,1J 1,1] 1-CsMe3{CH2)2)Hf (3) '7 5-Cs(CH3)5 '7 5.'7 1,f7 1-CsMe3(CH2)2 '7 5.'7 1,f7 1-Cs(CH3)3(CH2)2 5 '1 -C5(CH3)5 '7 5·Cs(CH3)5 '7 5.'7 1,f7 1·C5Me3(CH2)2 1.90 s 2JHH = 7 0.93 d 2JHH = 7 0.73 d 1.46 s (2 methyls) 1.41 s (1 methyl) 118.4 s 12.00 q 1JcH=126 1JcH = 145 62.25 t Cp*(f7 5,'7 1-CsMe4CH2CH2CH2)Hfl (4) '7 5-Cs(CH3)5 f7 5,'7 1-C 5Me4CH2CH2CH2 '1 5-C5(CH3)5 '7 5-Cs(CH3)5 '7 5,'7 1-CsMe4CH2CH2CH2 f7 5,'7 1-C 5Me4CH2CH2CH2 f7 5,'7 1-C 5Me4CH2CH2CH2 f7 5,'7 1-Cs(CH3)4CH2CH2CH2 1.91 s -0.72 m 0.67 m 1.97 m 2.28 m 2.65 m 2.97 m 1.56 s 1.61 s 1.73 s 2.84 s 119.08 s 13.36 q 70.33 t 39.29 t 26.71 t 11 .38 q 12.24 q 15.14 q 17.84 q 1Jc H = 123 1JcH = 127 1JcH=123 1JcH = 127 1Jc H = 125 1JcH=125 1JcH = 125 1JcH=1 27 54 Cp*('7 5,'7 1-C5Me4CH2CH=CH}Hfl (5} '7 5-Cs(CH3)5 '7 5,'7 1-C5(CH3) 4CH2CH=CH '75,'7 1-C 5Me4CH2CH=CH '7 5,'7 1-C5Me4CH2CH=CH '7 5,'7 1-C5Me4CH2CH=CH 1.92 s 1.70 s 1.78 s 1.82 s 2.68 s 7.09 dt 7.88 dt 2.78 m 3JHH = 12, 2 3JHH = 12,3 Cp*('7 5,'7 1,f7 1-C5Me3(CH2}(CH2CH2CH2}}Hf (6) '7 5-C5(CH3)5 , 5,, 1,, 1-C5(CH3)3(CH2) (CH2CH2CH2) , 5,,1,,1-C5Me3(CH2)(CH2CH2CH2) , 5,'7 1,,1-C 5Me3(CH2)(CH2CH2CH2) '7 5-C5(CH3)5 5 '1 -C5(CH3)5 , 5,, 1,,1-C5(CH 3)3(CH2)(CH2CH2CH2) , 5,, 1,, 1-C5(CH 3)3(CH2)(CH2CH2CH2) 1.81 s 1.38 s 1.62 s 1.91 s 1.51 d 1.86 d -0.96 m -0.04 m 1.86 m 2.40 m 2.47 m 2.66 m 11.73 q 116.3 s 68.40 t 29.60 br, t 35.83 br, t 66.73 m 2JHH = 8 2JHH = 8 1JcH = 126 1J cH = 142 1JcH = 123 1JcH = 123 Cp*Hf('7 5,,1,,1-C5Me4(CH2CH2CHCH2CH2}} (7) '7 5-C5(CH3)5 , 5,,1,, 1-C 5(CH3) 4(CH2CH2CHCH2CH2) , 5,,1,,1-C5Me 4(CH2CH2CHCH2CH2) , 5,,1,,1-C5Me4(CH2CH2CHCH2CH2) , 5,,1,,1-C5Me4(CH2CH2CHCH2CH2) '7 5,'7 1,,1-C5Me4(CH2CH2CHCH2CH2) '7 5,'7 1,f7 1-C 5Me 4(CH2CH2CHCH2CH2) '7 5-C5(CH3)5 '7 5-C5(CH3)5 , 5,'7 1,'7 1- C 5(CH 3) 4(CH2CH2CHCH2CH2) 1.80 s 1.53 s 1.63 s 1.74s 2.40s -0.81 m 1.12 m 2.33m 2.60 m 3.03 m -0.40 m 0.79 m 2.93m 2.97 m 11.57 q 116.3 s 10.11 q 11.79 q 12.63 q 1JcH = 125 1J c H=1 25 1J cH= 126 1JcH = 126 55 '7 5,'7 1,f7 1-C5Me4CH2CH2CHCH2CH2 '7 5,'7 1,f7 1-C 5Me4CH2CH2CHCH2CH2 '7 5, 17 1,f7 1-C 5Me4CH2CH2CHCH2CH2 '7 5,'7 1,f7 1-C5Me 4CH2CH2CHCH2CH2 '7 5,'7 1,f7 1-C5Me 4CH2CH2CHCH2CH2 14.16 q 49.16 t 36.48 t 38.55 d 33.42 t 63.11 t 1JcH = 125 1JcH=120 1JcH = 128 1JcH = 129 1JcH=125 1JcH = 122 Cp*2Hf(CH2CMe3)CH3 (8) '1 5-C5(CH3)5 Hf-CH3 Hf-CH2C(CH3)3 Hf-CH2C(CH3)3 1.83 5 -0.58 br, s -0.17 br, 5 1.09 s Cp*(f7 5,f7 1-C5Me4CH2)HfCH3 (9) '7 5-C5(CH3)5 f7 5,'7 1-C5(CH3)4CH2 f7 5,f7 1-C5Me4CH2 Hf-CH3 1.845 1.545 1.67 s 1.73 s 1.81 s 1.41 d 1.86 d -1.07 br, s 2JHH =7.7 2JHH =7.7 Cp*2Hf(C5H5)I (1 0) '7 5-C5(CH3)5 Hf-C5H5 1.83 s 7.40 m Cp*2Hf(C5H5)CH3 (11) 11 5-C5(CH3)5 Hf-CH3 Hf-C5H5 1.73 s -0.20 s 6.7-7.2 m Cp*('7 5,f1 1-C5Me4CH2)HfC5H5 (12) 5 '7 -C5(CH3)5 5 1 '7 ,'7 -C5Me4CH2 f7 5,'7 1-C5(CH3)4CH2 Hf-C5H5 f7 5,f7 1-C5Me4CH2 1.74s 1.97 d 1.76 d 1.74 s 1.74 s 1.57 s 1.20 s 6.3-7.2 m 70.51 t 2JHH =9 2JHH =9 1JcH = 141 56 '7 5-Cs(CH3)s Hf-C5Hs 1.62 s 6 .9-7.8 m Cp*2Hf{CH2CH2C5H4) (13) '7 5-Cs(CH3)5 1.71 s Hf-CH2CH2C5H4 Hf-CH2CH2C5H4 Hf-CH2CH2C5H4 0.73 t 3.23t 6.6-7.1 m a All NMR spectra were recorded in benzene-ds and are referenced to internal tetramethylsilane. b Coupling constants are reported in hertz. Note: The following abbreviations are used to denote the multiplicity of the NMR signals: s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; br, broad. 57 Experimental Details. General Considerations. All manipulations were performed using glovebox, high-vacuum[2 0] or Schlenk techniques. Solvents were purified by distillation from an appropriate drying agent under a N2 atmosphere and either used immediately or vacuum transferred from "titanocene" or sodium-benzophenone. Benzene-dB was purified by vacuum transfer from activated molecular sieves (4 A, Linde) and then from "titanocene". 1H and 13C NMR spectra were recorded on Varian EM-390 (90 MHz, 1H), JEOL GX400Q (400 MHz, 1H; 100 MHz, 13C) and Brucker WM500 (500 MHz, 1H) spectrometers. Elemental analyses were performed by Mr. Larry Henling of the CIT Analytical Laboratory. X-ray Structure Determination of Cp*(17 5 ,f7 1,f71-CsMe3(CH2)2)Hf. Crystal data: monoclinic, space group P2/c, a 103.965(31) 0 , V =13. 775(4)A, b =9.516(1 )A, c =14.183(6)A, f3 = =1804.0(9)A3, Z =4. A single crystal was used for the photographic workup and data collection; a hemisphere (+h, ±k, ±f) of data was collected (0 scan width 1.2° plus =45° on an Enraf-Nonius CAD4 diffractometer with graphite monochromator and MoKa (>. =0. 7107 A) radiation (5054 dispersion, 0-20 scans) at ca. 4° (0)/min to 20 reflections, ca. 50 hours). The three check reflections showed no unusual decrease in intensity. The data were reduced to F 0 2 and averaged over 2/m symmetry yielding 2360 reflections. The position of Hf(1) was derived from the Patterson map, and the remaining structure was revealed by electron density maps. Hydrogen atoms were located on difference Fourier maps, and were not refined. Full- matrix least-squares refinement of scale factor, atom coordinates and anisotropic Gaussian parameters resulted in a goodness-of-fitS parameters), RF ={2:w[F0 2-(Fcfk) 2) 2/(n-v)} 112 =5.53 (n =2360 reflections, v =190 =L:IIF0 1-IF0 11L:IF0 1=0.085 (2276 reflections, I > 0), and R'F =0.079 (1984 reflections, I> 3al); each shift/error< 0.01 in the final cycle. At this stage, a large peak, ca. 10 e/A3, was found on the difference Fourier map at coordinates (0.27, 0.29, 0.54), or roughly (1/2-x, 1/2-y, 1-z) where (x,y,z) corresponds to the 58 coordinates of Hf(1); the remaining peaks in the map, remote from the Hf atoms, were less than 1 e/A3 . Hf(1) was assigned a partial population of p, and a second Hf atom, Hf(2), was added to the refinement with a partial population 1-p. Two cycles of full-matrix least-squares refinement resulted in reasonable Uij's for both Hf atoms and a refinement gave S =2.66 (v =200 parameters), RF =0.042, and R'F =0.036. The final difference Fourier map indicated peaks and valleys of height ca. 2 e/A3 around the Hf atoms (as before, presumably residuals from absorption errors), and the remainder of the map was relatively flat with peaks no greater than ca. 1 e/A3 . Apparently the structure is disordered: about 7% of the time the molecule occupies a different position (the Hf(1 )-Hf(2) distance is only 1.26A) and, presumably, a different orientation as well, which we cannot deduce because of the small scattering power of the fractional carbon atoms. We can construct no sensible twinning model to explain the phenomenon. In any event, the dramatic decreases in S and RF are ample reason for introducing Hf(2), and we are confident that the omission of the remaining 7.3% carbon and hydrogen atoms of the second molecule has not led to significant errors in the description of the molecular structure. Kinetics of the reaction of Cp*(77 5 ,'7 1 ,'7 1 -CsMea(CH2}(CH2CH2CH2}Hf ,6, with ethylene. Compound 6 (28 mg, 0.059 mmol) and FeCp2 (ca. 5 mg) were dissolved in 2.83 g of benzene-dB. Five NMR tubes were prepared from this stock solution. Varying amounts of ethylene were subsequently added to each tube, with the concentrations ranging from 0.1 M to 0.4 Mat 25°C, and the tubes were sealed. The reactions were followed by monitoring the decrease in intensity of the methyl resonance of 6 at 8 1.38 relative to internal Cp2Fe. NMR spectra were recorded at timed intervals using a Varian EM-390 (90 MHz) spectrometer. Peak heights were demonstrated to be reproducible to within± 7% by repeated measurement. Each spectrum was recorded 3 times and the averaged peak height was used to calculate the rate constant. The reaction temperature of 82 o C was 59 maintained using a constant temperature oil bath controlled by thermoregulators and was observed to be constant to within ± 1 °C. Synthetic Procedures: Cp*(TJ 5 ,1J 1 -CsMe4CH2}Hf(CH2CMe3} (2}. Approximately 20 ml of benzene was condensed into a flask charged with 1 (0.60 g, 1.03 mmol} and UCH2C(CH3)3 (0.09 g , 1.15 mmol). The solution was stirred for 2 hours, then filtered. The solvent was replaced with petroleum ether (ca. 5 ml) and the suspension was cooled to -78° C. Filtering gave 0.23 g, (43%) of 2 as a pale orange solid. Elemental analysis: Found(calculated) %C 56.03 (57.85), %H 7.58 (7.73). Cp*(TJ 5 ,'7 1 ,'7 1-CsMe3(CH2}2}Hf (3). To a thick walled glass bomb was added 2 (0.68 g, 1.3 mmol) and 30 ml of toluene. This solution was stirred for 2 days at 140 o C then transferred to a frit assembly in the glove box. The toluene was replaced with petroleum ether and the resulting suspension was cooled to -78° C. Filtering yielded 0.37 g (64%) of 3 as an orange crystalline solid. Elemental analysis: Found(calculated) %C 53.95 (53.78), %H 6.32 (6.27). Cp*(TJ 5 ,TJ 1 -CsMe4CH2CH2CH2)Hfl (4). Method (a) Approximately 20 ml of C5H 6 was condensed into a small glass bomb containing 1 (1.06 g, 1.8 mmol) and ca. 5 equivalents of ethylene. After heating this solution overnight at 110° C. the volatiles were removed in vacuo. The resulting solid was transferred to a frit assembly, suspended in ca. 10 ml of petroleum ether, cooled to -78° C and filtered. This procedure affords 0.77 g (71%) of 4 as a bright yellow solid. Elemental analysis: Found(calculated) %C 42.55 (43.83), %H 5.36 (5.52). Method (b) Iodide 4 can also be prepared by treating benzene solutions of Cp*(TJ 5 ,1J 1-CsMe4 CH2)HfH (see chapter 3) with CH3l to give product and methane. Cp*('7 5 ,1J 1-CsMe4(CH2CH=CH)Hfl (5). Approximately 5 equivalents of acetylene was condensed into a -78° C petroleum ether solution of 1 (0.78 g, 1.36 mmol). This solution 60 was allowed to warm with stirring. The volume of petroleum ether was reduced to 5 ml and the solution was cooled to -78°C. Filtering afforded 0.49 g (60%) of 5 as a pale orange solid. Elemental analysis: Found(calculated) %C 43.96 (43.98), %H 5.19 (5.20). Cp*(17 5 ,'7 1 ,f71_CsMe3(CH2)(CH2CH2CH2)Hf (6). A small glass bomb was charged with 3 (0.25 g, 0.56 mmol), ca. 10 ml of benzene and 5 equivalents of ethylene. After heating this solution for 12 days at 35 o C the volatiles were removed in vacuo. The crude product was taken into the glove box and washed with petroleum ether into a frit apparatus. Cooling to -78°C caused the precipitation of 6 as a yellow solid which collected by filtration in ca. 40% yield. Elemental analysis: Found(calculated) %C 55.35 (55.63), %H 6.95 (6.79). Cp*Hf('7 5,'7 1 ,f7 1 -CsMe4CH2CH2CHCH2CH2) (7). A small glass bomb was charged with 3 (0.60 g, 1.34 mol), ca. 10 ml of benzene and 10 equivalents of ethylene. This solution was heated overnight at 11 0°C after which the volatiles were removed in vacuo. The crude solid was transferred to a frit assembly and ca. 10 ml petroleum ether was added. After filtration the solution was cooled to -78° C. The off-white precipitate was collected by filtration in ca. 40% yield. Elemental analysis: Found(calculated) %C 56.94 (57.30), %H 6.97 (7.21). Cp*2Hf(CH2CMe3)CH3 (8). To a solution of Cp*2Hf(CH2CMe3)ll22l (0.51 g, 0.79 mmol) and N,N,N',N'-tetramethylethylenediamine (TMEDA) (1 .5 mmol) in 20 ml benzene was syringed Meli/diethyl ether (0.87 mmol). After stirring overnight the benzene was replaced with 20 ml of petroleum ether and the solution was filtered. The solvent volume was reduced to ca. 5 ml and cooled to -78° C. Filtration yielded 0.23 g of white 8 (53%). Cp*(17 5 ,'7 1 -CsMe4CH2)HfCH3 (9). Compound 8 (0.46 g, 0.86 mmol) in 5 ml of toluene was heated in a glass bomb overnight at 135° C. The solution was transferred to a sublimator and the solvent was removed in vacuo. Subliming at 110° C for 2 days yielded 0.14 g (35%) of yellow 9. Elemental analysis: Found(calculated) %C 54.88 (54.48), %H 6.63 (6.97). 61 Cp* 2Hf(C6Hs)l (1 0). To Cp*2Hfl2 (1.97 g, 2. 79 mmol) in ca. 20 ml benzene was syringed a 2-fold excess of LiC5Hs in cyclohexane/diethyl ether. After stirring overnight the solvent was replaced with ca. 50 ml of petroleum ether and the solution was filtered. The solvent volume was reduced, then cooled to -78° C. Filtration yielded 1.28 g (70%) of cream colored 10. Cp*2Hf(C5Hs)CH3 {11 ). To 10 (0.49 g, 0. 75 mmol) and TMEDA (ca. 1 mmol) in diethyl ether was syringed Meli/diethyl ether (1.35 mmol). After stirring overnight the diethyl ether was replaced with ca. 20 ml of petroleum ether and the solution was filtered. The solvent volume was reduced and cooled to -78° C. Filtration yielded 0.21 g (52%) of white 11. Elemental analysis: Found(calculated) %C 57.86 (59.93), %H 6.94 (7.08). Cp*(17 5 ,1] 1-CsMe4CH2)HfC5Hs (1 2). Compound 1 1 (0.60 g, 1.1 mmol) in 5 ml of toluene was heated in a small glass bomb for three days at 135° C. The toluene was replaced with petroleum ether and the solution was transferred to a frit assembly. The solvent volume was reduced and cooled to -78°C. Filtering yielded 0.34 g (58%) of yellow 12. (17 5 -CsMes)2Hf(C6Hs)2. Cp*2Hfl2 (1.49 g, 2.1 mmol), excess C5Hsli, and 1 ml of TMEDA were stirred in ca. 20 mls of C5H5 overnight. The benzene was replaced with petroleum ether and the solution was filtered. Cooling to -?soc and filtering afforded 0.70 g (55%) of colorless (17 5-CsMes)2Hf(C5Hs)2· (17 5 -CsMes)2Hf(CH2CH2C6H4) {13). Method (a). A sample of Cp*(17 5 ,'7 1-CsMe4CH2)HfC5Hs (12) and ca. 0.5 mls of benzene-dB were sealed in an NMR tube with excess ethylene. Heating the NMR tube overnight at 130°C cleanly generated metallacycle 13. Method (b). A benzene solution of Cp*2Hf(C5Hs)2 (0.60 g, 1.0 mmol) and excess ethylene were heated for 2 days at 120 o C. The solution was filtered and the benzene was replaced with petroleum ether. Cooling to -78 o C and filtering afforded 0.2 g 62 (36%) of off-white 13. Elemental analysis: Found(calculated) %C 58.14 (60.80), %H 7.08 (6.92). References (1) See for example: (a) Manriquez, J. M.; McAlister, D. R.; Sanner, R. D.; Bercaw, J . E. J. Am. Chem. Soc., 1976, 98, 6733. (b) Manriquez, J. M.; McAlister, D. R., Sanner, R. D., Bercaw, J . E. J.Am. Chem. Soc., 1978, 100,2716. (2) McDade, C.; Green, J. C.; Bercaw, J. E. Organometal!ics, 1982, 1, 1629. (3) Bottomly, F.; Egharevba, G. 0.; Lin, I. J. B.; White, P. S. Organometallics, 1985, 4, 550-553. (4) Bercaw, J. E. "Advances in Chemistry Series", 1978, 167, 136. (5) McAlister, D. R.; Erwin, D. K.; Bercaw, J. E. J . Am. Chem. Soc., 1978, 100, 5966. (6) (a) See chapter 1 of this thesis. (b) Bulls, A. R.; Schaefer, W. P .; Serfas, M.;Bercaw, J. E. Organometallics, 1987,6, 1219. (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. J.Am. Chem. Soc., 1987, 109, 203. (8) Pattiasina, J. W.; Hissink, C. E.; de Boer, J . L.; Meetsma, A.; Teuban, J. H.; Spek, A. L. J. Am. Chem. Soc., 1985, 107, 7758-7759. (9) Cloke, F. G. N.; Green, J. C.; Green, M. L. H.; Morley, C. P. J. Chem. Soc., Chem. Commun. , 1985, 945. (1 0) (a) Bandy, J .A.; Mtetwa, V.S.B.; Prout, K.; Green, J .C.; Davies, C.E.; Green, M.L.H.; Hazel, N.J.; Izquierdo, A.; Martin-Polo, J.J. J. Chem. Soc. Dalton Trans. 1985, 2037, and references therein. (b) Schock, L. E.; Brock, C. P.; Marks, T. J. Organometallics , 1987, 6, 232. (11) v an Asselt, A. ; Santarsiero, B.D.; Bercaw, J .E. J. Am. Chem. Soc., 1986, 108, 8290-8291. 63 (12) Fronczek, F.R.; Baker, E. C.; Sharp, P.R.; Raymond, K.N.; Alt, H.G.; Rausch, M.D. lnorg. Chern. 1976, 15, 2284. (13) Prepared as described in chapter 4 of this thesis. (14) Green, M. L. H.; O'Hare, D. J. Chern. Soc. Dalton Trans., 1985,1585-1590. (15) (a) See reference 7. (b) M. St. Clair and J . E. Bercaw, unpublished results. (16) (a) See references 2 and 6b. (b) Chamberlain, L.; Rothwell, I. P.; Huffman, J . C. J. Am. Chern. Soc. 1982, 104, 7338-7340. (17) Schock, L.E.; Brock, C.P.; Marks, T.J. Organometallics 1987, 6, 232-241 . (18) See for example (a) Erker, G.; Kropp, K. J. Am. Chern. Soc., 1979, 101, 3659. (b) Dvorak, J.; O'Brien, R. J.; Santo, W. J. Chern. Soc., Chern. Commun., 1970, 411. (19) Buchwald, S.L.; Watson, B.T.; Huffman, J.C. J. Am. Chern. Soc. 1986, 108, 7411-7413. (20) It is interesting to note that the previously mentioned zirconium analogue, Cp*('7 5 .'7 1-C5 Me4CH 2)ZrCsHs exchanges only the ortho phenyl hydrogens with the Cp* ring hydrogens, presumably due to restricted rotation about the zirconium phenyl bond (reference 17). An estimate for the barrier to phenyl rotation in Cp*('7 5 ,'7 1-CsMe4CH2)HfCsHs is provided from Cp*2Hf(CH3)(CsH4-m-CH3), which exists as a 1:1 mixture of the exo and endo isomers. Data obtained from the coalesence of the endo and exo Hf-CH3 resonances suggests li.Gl for phenyl rotation is ca. 20 kcal· mol-1 at 100° C. (21) Burger, B. J.; Bercaw, J. E. in "New Developments in the Synthesis, Manipulation and Characterization of Organometallic Compounds", ACS Symposium Series, A. Wayda and M. Darensbourg, Eds., in press. (22) See chapter 1 of this thesis. 64 Chapter 3 The Synthesis of Metalation-Resistant Bis(cyclopentadienyl) Ligand Systems. The Molecular Structure of [(J.t-Me2Si)2(77 5 -CsH2-4-CMe3)2]ZrCI2. 65 Introduction The tendency for pentamethylcyclopentadienylligands of early transition metal hydride and alkyl derivatives to undergo intramolecular metalation is well documented.£1 l Relatively facile decomposition pathways of this type may preclude the observation and study of other desirable transformations. For this reason, the synthesis of metalation-resistent bis(cyclopentadienyl) ligand systems has been undertaken. At least two methods of increasing the resistance of alkylated cyclopentadienylligands towards metalation can be imagined. One method would be to synthesize cyclopentadienyl ligands which contain alkyl groups with extremely strong C-H bonds. In view of the facility with which early transition metal complexes activate the relatively strong sp2 and sp hybridized C-H bonds of arenes and terminal alkynes[2 l, we considered this approach unlikely to succeed. The approach we chose to pursue would restrain the alkyl substituents from approaching the empty, low-lying orbitals in the equatorial plane between the cyclopentadienyl rings.£3] It is typically assumed that the activation of substrate molecules requires an initial interaction with these metal-based orbitals. Restraining the pendant cyclopentadienyl alkyl groups from interacting with these orbitals, and hence, from undergoing metalation, would be accomplished by tethering together the cyclopentadienyl rings with two bridges at adjacent carbons and placing a bulky alkyl substituent in the position "meta" to both bridges. The presence of two adjacent links between the cyclopentadienyl rings should severely restrict the rotation of the cyclopentadienyl ligands and make the conformation necessary for intramolecular metalation less favorable. The preliminary results of these efforts are reported herein. 66 Results and Discussion Initial efforts to synthesize metalation-resistant bis(cyclopentadienyl) ligands focused on the bis(ethylene) bridged compound (CsH2-1-CMe3)2(J.L-CH2CH2)22- (Figure 1). Figure 1 The singly bridged compound [(CsH3-1-CMe3)2(J.L-CH2CH2)](MgCI)2 was prepared in 50% yield by the metathesis reactions depicted in equation 1.£4] 2 6 MgCI MgCI 1) TsOCH 2CH 20Ts -2 CIMgOTs (1) 2) 2 CIMgCH 2CHMe 2 - 2 isobutane MgCI Despite considerable effort, we were unable to cleanly add a second ethylene bridge to [(CsH3-1-CMe3)2(J.L-CH2CH2)](MgCI)2[5], thus requiring us to pursue an alternative doubly bridged bis(cyclopentadienyl) ligand system . .. .. _..:..: ---- 67 The bis(dimethylsilyl) bridged bis(cyclopentadienyl) ligand [(!'-Me2Si)2(CsH2-4-CMe3)2] 2-u+2" THF2 (1) was prepared from the known mono bridged dianion (CsH3-3-1Bu)2(1 ,1'-Me2Si) 2-u+2 (Scheme 1). 1 can be isolated in ca. 50% yield from petroleum ether as a benzene-soluble white powder. "-J/ CH 3 •. Si CHj/ ~ "-J/ u+ + u+ CI 2 SiMe 2 THF 5 min. CH 3 •• Si CHj.J" AH - 2 LiCI P.E. I THF .,CH 3 Si " CH 3 + 2 n-Buli - 2 n-butane 68 Dian ion 1 reacts rapidly (::5 10 minutes) and quantitatively (1 H NMR) with ZrCI 4 at room temperature to generate [(J,L-Me2Si)2H77 5-CsH2-4-CMe3)2]ZrCI2 (2) (equation 2). + (2) 5 - 1 min. 2LICI 2 89% yield from petroleum ether as an off-white powder. In order to confirm the assignment of 2 as a monomeric zirconocene derivative the structure has been determined by X-ray diffration techniques. A molecular view of BsZrCI2 is presented in Figure 2; another view including the atomic labeling is shown in Figure 3. The most unusual feature of BsZrCI2 is the dihedral angle of 73(4) o formed by the cyclopentadienyl ring planes. Typical values for dihedral angles of the group IV metallocenes are on the order of 45-50 o . [7] As a consequence of this unusual dihedral angle, the cyclopentadienyl ring carbon to zirconium distances vary widely (Table 2). The shortest distances are from zirconium to the ring carbons bonded to the dimethylsilyl bridges, C11A, C11 B, C21A, and C21 B (Zr-C(avg) = 2.415(5) A). The largest distances are from zirconium to the carbons bearing the tert-butyl substituent, C13 and C23 (Zr-C(avg) = 2.676(5) A). The carbon-carbon distances within the two 69 Figure 2. A molecular view of 2 . The thermal ellipsoids are shown at the 50% probabi lity level. 70 Figure 3. A molecular view of 2 including atomic labels. 71 Table 2. Selected Distances and Angles for BsZrCI2 (2). Atom Atom Cl1 Zr Zr Cl2 Zr C11A Zr C11B Zr C12A C12B Zr Zr C13 C21A Zr C21B Zr Zr C22A C22B Zr C23 Zr C11A Si1 C21A Si1 C1A Si1 C2A Si1 C11B Si2 C21B Si2 C1B Si2 C2B Si2 C11A C11B C11A C12A C11B C12B C12A C13 C12B C13 C14 C13 C15 C14 C16 C14 C17 C14 C21A C21B C21A C22A C21B C22B C22A C23 C22B C23 C24 C23 C25 C24 C26 C24 C24 C27 Dist. (A) Atom Atom Atom Angle 2.427(5) 2.419(5) 2.418(5) 2.416(5) 2.570(5) 2.567(5) 2.688(5) 2.408(5) 2.417(5) 2.560(5) 2.554(5) 2.663(5) 1.870(5) 1.889(5) 1.872(7) 1.870(6) 1.871 (5) 1.878(5) 1.853(7) 1.855(7) 1.466(7) 1.427(7) 1.417(7) 1.394(7) 1.390(7) 1.512(7) 1.503(10) 1.484(10) 1.516(9) 1.448(7) 1.397(7) 1.408(7) 1.407(7) 1.408(7) 1.527(7) 1.504(8) 1.511 (9) 1.499(9) Cl1 C11A C11B C1A C2A C12A C12B C13 C13 C12B C22A C22B C23 C23 C22B Zr Si1 Si2 Si1 Si2 C11A C11B C12A C12B C13 C21A C21B C22A C22B C23 Cl2 C21A C21B C2A C2B C11B C11A C11A C11B C12A C21B C21A C21A C21B C22A 101.9(1) 92.4(2) 92.2(2) 109.2(3) 109.8(3) 105.9(4) 105.6(4) 110.4(4) 111.3(4) 106.9(4) 107.1 (4) 1 05.9(4) 11 0.5(5) 11 0.7(5) 105.8(5) 72 cyclopentadienyl rings also vary dramatically. The longest carbon-carbon bonds are the C11A-C118 and the C21A-C21B distances of 1.466(7) and 1.448(7) A, respectively. The remaining carbon-carbon bond distances are substantially shorter, ranging from the C128 to C13 distance of 1.390(7) A, to the C11A to C12A distance of 1.427(7) A. The complexes BsZr(CH2CMe3)l (3) and Cp*2Zr(CH2CMe3)l (4) were prepared by established methods[Bl, thus allowing for the relative stability of a pair of Cp* and Bs ligated zirconium derivatives to be compared. Heating benzene-ds solutions of 4 at 143°C results in the liberation of neopentane and the formation of Cp*('7 5 .'7 1-CsMe4 CH 2)Zrl (equation 3). (3) t1,2 = 12 min. • C{CH 3 ) 4 4 The disappearance of 4, as monitored by Cp* peak height versus internal Cp2Fe, follows first order kinetics for 3 half lives with a rate constant of 8.6(5) x 1o-4 s-1 (t112 = 12.5 minutes at 143 o C). In contrast, heating benzene-d6 solutions of 3 for three days at 143 o C results in no reaction 3 days (4) 73 less than 10% loss of intensity of the tert-butyl peak height versus an internal standard (equation 4). Thus, the bis(pentamethylcyclopentadienyl) ligated complex 4 decomposes~ 3 x 103 times more rapidily at 143°C than does the [(1.£-Me2Si) 2('7 5-CsH2-4-CMe3) 2] ligated complex3. In summary, a synthetically useful route to the doubly bridged bis(cyclopentadienyl) ligand [ (1.£-Me2Si)2) (CsH2-4-CMe3)2] 2-u+2 • THF2 has been developed. This ligand has been used to prepare a series of zirconocene complexes, including the dichloride [(1.£-Me2Si)2('1 5-CsH2-4-CMe3)2]ZrCI2 (2) which has been crystallographically characterized. The presence of two adjacent dimethylsilyl bridges results in an unusually large dihedral angle formed by the cyclopentadienyl rings of 2. A comparison of the thermal stabilities (at 143°C} of the complexes [(~.£-Me2Si)2('1 5 -CsH2-4-CMe3)2]Zr(CH2CMe3)l (3) and ('7 5-CsMes)2Zr(CH2CMe3)l (4) reveals that the former is~ 3 x 103 times more stable than the latter. While these results should be considered preliminary, they do suggest that early transition metal complexes of Bs can be prepared and show increased thermal stabilty as compared to their pentamethylcyclopentadienyl analogs. Efforts are presently underway to extend the range of Bs-ligated complexes to more electropositive metals such as scandium. 74 c NMR data.8 Table 1. 1H and 13 Compound Assignment o(ppm) C(CH3)3 Si(CH3)2 CsH2 THF(.B) THF(a) 1.52 s 0 .68 s 6.63 s 0 .95 m 2.92 m Couplingb [(J.L-Me2Si)2( 115-CsH2-4-CMe3)2]ZrCI2 (2) C(CH3)3 Si(CH3)2 CsH2 C(CH3)3 C(CH3)3 Si(CH3)2 Cp-C1,C2 Cp-C3,Cs Cp-C4 1.445 0.32 s 0.57 s 6.83 s 32.14 q 32.98 s -3.94 q 4 .23 q 113.8 s 136.7 d 143.4 s [(J.L-Me2Si)2('1 5-CsH2-4-CMe3)2]Zrl2 C(CH3)3 Si(CH3)2 1.60 s 0.30 s 0 .47 s 7.12 s C(CH3)3 Si(CH3)2 1.47 s 0 .23 s 0.45s 0.485 0 .50 s 7.13 s 7 .16 s 1.12 s 0.77 s CH2C(CH3)3 CH2CMe3 1JcH =125 1JcH 122 1JcH = 123 = 1JcH = 169 75 Cs(CH3)5 CH2C(CH3)3 CH2C(CH3)3 1.85 s 1.22 s 0.90 s Cs(CH3)5 Cs(CH3)4CH2 1.88 s 1.33 s 1.57 s 1.95 s 2.18 s 1.45 d not observed a All NMR spectra were recorded in benzene-d5 and are referenced to internal tetramethylsilane. b Coupling constants are reported in hertz. Note: The following abbreviations are used to denote the multiplicity of the NMR signals: s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; br, broad. 76 Experimental Section General Considerations. All manipulations were petformed using glovebox, high-vacuum£9] or Schlenk techniques. Solvents were purified by distillation from an appropriate drying agent under a N2 atmosphere and either used immediately or vacuum-transferred from "titanocene" or sodium-benzophenone. Benzene-dB was purified by vacuum transfer from activated molecular sieves (4 A, Linde) and then from "titanocene". 1H and 13 C NMR spectra were measured on Varian EM-390 (90 MHz) and JEOL GX4000 (400 MHz, 1H; 100 MHz, 13C)) spectrometers. Elemental analyses were petformed by Mr. L. Henling of the CIT Analytical Laboratory. Synthetic Procedures [(.u-Me2Si)2(CsH2-4-CMe3)2] 2-u+2"THF2 (1). Under an argon atmosphere, Me2SiCI2 (4.3 mls, 34 mmol) was slowly syringed into a well stirred tetrahydrafuran solution of [(CsH3-3-CMe3)2(P.-Me2Si)] 2-u+2[ 101 (10.7 g, 34 mmol). The resulting suspension was stirred for one hour, after which the majority of the tetrahydrafuran was removed in vacuo. Approximately 90 ml of petroleum ether was condensed into the flask and the suspension was filtered to remove the LiCI. 2.2 equivalents of n-butyllithium (47 ml of a 1.6 M hexane solution) was slowly syringed into the resulting solution. After ca. 20 minutes a white precipitate formed which was isolated by filtration. Yield: 10.0 g (57%). [(p.-Me2Si)2(CsH2-4-CMe3)2]ZrCI2 (2). Approximately 50 ml of benzene was condensed onto a mixture of 1 (1 0.0 g, 19.5 mmol) and ZrCI4 (5.0 g, 21.4 mmol). After stirring this mixture for 2 hours the suspension was filtered to remove LiCI and the benzene was removed in vacuo. Approximately 80 ml of petroleum ether was condensed onto the product. Filtering this suspension yielded 8.8 g (2 crops) of product as an off-white powder (89% yield). The dichloride is often contaminated with traces of THF. 77 [(J.£-Me2Si)2(C5H2-4-CMe3)2]Zrl2. Approximately 50 ml of benzene was condensed onto a mixture of 2 (3. 7 g, 7.3 mmol) and excess Bl3 (2.8 g, 7.2 mmol). After stirring for 1 hour the benzene was replaced with diethyl ether. Filtering the resulting suspension afforded 3.24 g of the yellow product (64%). [(J.£-Me2Si)2(C5H2-4-CMe3)2]Zr(CH2CMe3)l (4). Approximately 25 mls of benzene was condensed onto a mixture of [(J.£-Me2Si)2(C5H2-4-CMe3)2]Zrl2 (1.6 g, 2.29 mmol) and LiCH2CMe3 (215 mg, 2. 75 mmol). After stirring for 1 hour the suspension was filtered and the benzene was replaced with petroleum ether. Cooling to -78°C and filtering afforded 0.91 g (62%) of bright yellow product. Elemental analysis: Found(calculated) o/oC 49.78 (50.36), o/oH 6.69 (7.04). Kinetics of the decomposition of Cp*2Zr(CH2CMe3)l (4) and BsZr(CH2CMe3)l (3). Approximately 30 mg of 4 was sealed in a NMR tube with ca. 5 mg of FeCp2 and ca. 0.5 ml of benzene-d5. The reaction were followed by monitoring the decrease in intensity of the tert-butyl resonance of 4 at o 1.22 relative to internal Cp2Fe. NMR spectra were recorded at timed intervals using a Varian EM-390 (90 MHz) spectrometer. Peak heights were demonstrated to be reproducible to within± 7% by repeated measurement. Each spectrum was recorded 3 times and the averaged peak height was used to calculate the rate constant. The reaction temperature of 143 o C was maintained using a constant temperature oil bath controlled by thermoregulators and was observed to be constant to within ± 1°C. The thermolysis of BsZr(CH2CMe3)l (3) was followed by monitoring the decrease in the tert-resonance at 5 1.47. In order to assure an unreactive internal standard, hexamethylbenzene in benzene-d5 was sealed in a melting point capillary which was subsequently added to the NMR tube used for kinetic measurements. 78 References: (1) (a) Chapters I and II of this thesis. (b) McDade, C., Green, J. C., Bercaw, J. E. Organometallics, 1982, 1, 1629. (c) Pattiasina, J. W., Hissink, C. E., de Boer, J . L., Meetsma, A., Teuban, J . H., and Spek, A. L., J. Am. Chern. Soc., 1985, 107, 77587759. (d) Parkin, G. F. B. C. and Bercaw, J. E. unpublished results. (2) 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. Chern. Soc., 1987, 109, 203. (3) Lauher, J. W., Hoffmann, R. J. Am. Chern. Soc., 1976., 98, 1729-1742. (4) While the yield of the mono bridged dian ion [(C5Hs-1-CMes)2(JL-CH2CH2)(MgCI)2 is ca. 50%, its purity is only ca. 85% (1H NMR), which undoubtedly contributes to the lack of success in adding a second ethylene bridge. (5) The reaction of [(C5H3-1-CMe3)2(JL-CH2CH2)(MgCI)2 with one equivalent of TsOCH2CH20Ts, followed by deprotonation with n-butylithium yields an intractable mixture. Adding ScCis · THFs to suspensions of this crude mixture, followed by sublimation at 130°C, does afford (17 5-C5H2-1-CMes)2(JL-CH2CH2)2ScCI · THF in ca. 1%yield. (6) The abbreviation Bs is derived from Q.is(dimethyl§.ilyl) bridged bis(cyclopentadienyl). (7) See for example: Wailes, P. C.; Coutts, R. S. P .; Weigold, H. Organometallic Chemistry of Titanium, Zirconium, and Hafnium; Academic Press: New York, N.Y. 1974, and references therein. (8) Cp*2Zr(CH2CMes)l was prepared as described for its hafnium analogue in Chapter 1. 79 (9) Burger, B. J .; Bercaw, J. E. in "New Developments in the Synthesis, Manipulation and Characterization of Organometallic Compounds", ACS Symposium Series, A. Wayda and M. Darensbourg, Eds., in press. (1 0) Bunel, E.; Bercaw, J. E. Organometallics, Manuscript in preparation 80 Chapter 4 Relative Bond Dissociation Energies for Early Transition Metal Alkyl, Aryl, Alkynyl and Hydride Compounds. Equilibration of Metallated Cyclopentadienyl Derivatives of Peralkylated Hafnocene and Scandocene Derivatives with Hydrocarbons and Dihydrogen. 81 Abstract: Relative bond dissociation energies (BDEs) have been obtained by equilibrating early transition metal alkyls and hydrides with H2 or the C-H bonds of hydrocarbons. Thus, in benzene solution Cp*2HfH2 (Cp* = (77 5-CsMes)) equilibrates with Cp*2Hf(C5Hs)H and dihydrogen. From the enthalpy of the reaction, l:IH o = +6.0(3), the Hf-H (BDE) is calculated to be 0.8(3) kcal·mol-1 stronger than the Hf-C5H5 BDE. Relative Sc-C5H5 and Sc-alkyl BDEs have been estimated from the equilibration of the metallated complex Cp*(77 5.77 1-CsMe4CH2CH2CH2)Sc, C5H5 and Cp*(77 5-CsMe4CH2CH2CH3)Sc(C6Hs), the Sc-C5Hs BDE being 16.6(3) kcal·mol-1 stronger than the Sc-CH2CH2CH2CsMe4 BDE. From a similar reversible intramolecular metallation of Cp*(77 5-CsMe4CH2CH2CH3)HfH 2 to give Cp*(77 5,77 1 -CsMe4CH2CH2CH2)HfH and dihydrogen, the Hf-H BDE is estimated to be 23.0(3) kcal·mol-1 stronger than the Hf-CH2CH2CH2CsMe4 BDE. The equilibration of Cp*(77 5-CsMe4CH2C5Hs)Sc-C=CCMe3 with the metallated scandocene derivative Cp*(77 5.77 1-CsMe4CHro-C6H4)Sc and tert-butylacetylene lies very far towards Cp*(77 5-CsMe4CH2C5Hs)Sc-C=CCMe3, so that only a lower limit for the relative Sc-alkynyl and Sc-aryl BDEs may be determined: BDE(Sc-alkynyl) - BDE(Sc-aryl) ~ 29(5) kcal·mol-1. These early transition metal-hydrocarbyl (M-R) BDEs correlate with the corresponding H-R BDEs (i.e. M-alkynyl > M-aryl > M-alkyl); however, the M-R BDEs increase more rapidly with s character than does the H-R BDEs. The origin of this effect is not known, but it is undoubtedly also responsible for the characteristically high M-H BDEs for transition metal hydride compounds. In an effort to probe the polarity of Sc-aryl bonds a series of scandocene derivatives capable of reversibly metallating at either of two differently substituted benzyl groups was prepared. The equilibrium constants for these metallated derivatives: (7J 5,7J 1-CsMe4CH2-o-C6H3-p-X)(77 5-CsMe4CH2C6H4-m-CH3)Sc <l t> (77 5-CsMe4CH2C6H4-m-X)(77 5,7J 1- CsMe4CH2-o-C6H3-P-CH3)Sc (X= H, CF3, NMe2) were determined. The small dependence of Keq on the nature of X suggests that the Sc-aryl bond is essentially covalent with only a slight ionic character. 82 Introduction Early transition metals form many stable compounds with electronegative ligands such as halogens due to the high degree of ionic character of the M-X bond. Strong bonds between early transition metals and oxygen- or nitrogen-containing ligands (e.g. OR, NR2) have this same feature. Moreover, multiple M=X or M=X bonding may also obtain by donation of one or two electron pairs from X to empty orbitals on the metal. The later, less electropositive transition metals also form strong bonds to these heteroatoms[11; however, early metal-heteroatom bonds are generally more robust. The strength of early transition metal-carbon and early transition metal-hydrogen bonds relative to the strengths of those of the middle and late transition metals is somewhat less straightforward to predict. A priori, we m ight expect that they would also be systematically greater since the bond homolysis (LmMn-R--!> LmMn-1 + ·R) formally involves reducing the metal by one unit. Electropositive metals have higher reduction potentials (for a given oxidation state) as compared with the more electronegative transition metals to the right. Furthermore, a comparison of the electronegativities of carbon, hydrogen, and the various transition metals reveals that some ionic character (M 0 +-R 0-; R = H, C) may be expected when M is an early transition metal. Indeed, the reactions of early metal hydrides and alkyls with protic reagents such as mineral acids and water, which rapidly evolve H2 and RH, suggest just such polarization.£2 1 Little direct experimental evidence is available which is indicative of the amount of ionic character for early transition metal-carbon or -hydrogen bonds, however. Moreover, Mulliken populations obtained from quantum mechanical calculations are generally unreliable indicators of the ionic character of M-R bonds.£3] Likewise, little experiment evidence is available which reflects the dependence of the M-R BDE on the hybridization at the a 83 carbon, i.e. the quantitative differences in the series: metal-alkyl vs. metal-alkenyl vs. metal-aryl vs. metal-alkynyl. In this chapter are described some experiments designed to address these issues. Although we have thus far been unable to obtain reliable absolute bond dissociation energies (BDEs), we have been able to derive accurate relative BDEs by equilibrating early transition metal alkyls and hydrides with the C-H bonds of hydrocarbons via the rapid, reversible "sigma bond metathesis" reactions[4 l characteristic of early transition metals, lanthanides and actinides (equation 1). LnM-R + R'-H ~ LnM-R' + R-H (R, R' (1) =H, alkyl, alkenyl, aryl, alkynyl) Reported for peralkylated-scandocene and -hafnocene derivatives are (t) relative M-H and M-C BDEs (M = Sc, Hf), (it) relative bond strengths of scandium- and hafniumcarbon bonds as the hybridization at carbon is varied, and (iit) relative Sc-aryl BDEs as the aryl substitutent is systematically varied. Results Determination of Relative M-H and M-Ph Bond Dissociation Energies via Intermolecular Equilibration. As previously reported[7 l, the clean, reversible nature of the reaction of bis(pentamethylcyclopentadienyl)scandium hydride, Cp*2Sc-H (Cp* = ('7 5-CsMes)) with benzene (equation 2) permits a determination of the relative Sc-H and Sc-CaHs BDEs: the Sc-H BDE is 1.5 kcal mol-1 greater than the Sc-CaHs BDE. (2) 84 In a similar manner Cp*2HfH2 in benzene solution at~ 100 o C cleanly establishes an equilibrium with Cp*2Hf(H)(CsHs) and dihydrogen (equation 3). (3) Cp*2Hf(CsHs)2 is not observed under these conditions. The enthalpy and entropy of equilibrium 3, ~H o = +6.0(3) kcal· mol-1 and ~So = +3.5(5) e.u., are obtained from the temperature dependence of the equilibrium (Table 1, Figure 1). Table 1. Equilibrium constants for the equilibrium between Cp*2Hf(CsHs)H and Cp*2HtH 2 in CsHs (equation 3). Temperature ( 0 C) 100.5 5.27 109 5.91 129 8.57 144 12.1 171 18.6 =+ 6.0(3) kcal · mol-1 ~s o =+ 3.5(5) e.u. ~H o 85 -8.5 -9.0 -9.5 -10.0 +--~-~-..---r-.........- .........----....----0 .0022 0.0026 0.0027 0.0023 0.0024 0.0025 1 I T(K) Figure 1. van't Hoff plot for the equilibrium between Cp*2Hf(CsHs)H and Cp*2HfH2 in C5H5. 86 Attempts to extend this approach to include equilibration of alkanes with permethylhafnocene or permethylscanodcene hydrides or alkyls were unsuccessful, possibly due to interfering reactions of the C-H bonds of the pentamethylcyclopentadienyl ligands.£ 71 For example, thermolyzing Cp*2Sc-C5H5 with mesitylene, tetramethylsilane, or n-butane, either neat or with varying concentrations of benzene, results in the slow decomposition to a mixture of uncharacterized scandium-containing products. The expected reaction products, Cp*2Sc-R (R = CH2C5H3-3,5-(CH3}2, CH2SiMe3, CH2CH2CH2CH3), have been independently synthesized, and each reacts with benzene to generate Cp*2Sc-C5H5 and H-R (equation 4). Cp*2Sc-R + C5H5 R <1-- I> Cp*2Sc-C5H5 + H-R (4) =CH2C5H3-3,5-(CH3)2 =CH2SiMe3 = CH2CH2CH2CH3 Thus, the reactions of Cp*2Sc-C5H5 with these reagents appear to be so thermodynamically unfavorable as to prevent buildup of detectable concentrations of permethylscandocene derivatives other than Cp*2Sc-C5H5. Determination of Relative M-H and M-C Bond Dissociation Energies via Intramolecular Equilibration. We have previously shown that bis(tetramethylneopentylcyclopentadienyl)zirconium dihydride is in facile equilibrium with the metallated mono-hydride complex and dihydrogen (equation 5)_[5] 87 (5) With the hope that these types of equilibria are quite general, we have undertaken the syntheses of compounds which would allow intramolecular equilibration via sigma bond metathesis reactions of Sc-R with the C-H bonds of appended alkyl, aryl, etc. groups. Thus, the mixed-Cp scandocene chloride 1 has been prepared as shown in equation 6. Heating (Cp*ScCI2ln with one equivalent of Li+(CsMe4CH2CH2CH3)- yields complex 1 as an amber oil (> 90% pure by 1 H NMR). (Cp*ScCI2ln + Li+(CsMe4CH2CH2CH3)- - - - - ! > Cp*(77 5 -CsMe4CH2CH2CH3)ScCI (6) - LiCI 1 Treatment of 1 with CH3Li yields Cp*(77 5-CsMe4CH2CH2CH3)ScCH3, which rapidly eliminates methane (< 5 min. at 25°C) to afford Cp*('7 5 ,'7 1 -CsMe4C~2CH2CH2)Sc (2) (equation 7). + CH3Li, - LiCI Cp*('7 5 ,'7 1 -CsMe4CH2CH2CH2)Sc 2 (7) 88 In contrast to other permethylscandocene alkyls, cyclohexane-d12 solutions of 2 are quite stable at 140°C. If however, solutions of 2 in cyclohexane-d12 are heated with 5-10 equivalents of benzene, an equilibrium consisting of 2, CsHs, and a new species identified as Cp*('1 5 -CsMe4CH2CH2CHs)ScCsHs (3) is established (equation 8; LlH 0 = +3.5(3) kcal mol-1 and tlS 0 = 1 0.5(5) e.u.). Sc ~ 3 2 Table 2. Equilibrium constants for the equilibrium between Cp*(f1 5 ,'7 1-CsMe4CH2CH2CH2)Sc and Cp*('1 5-CsMe4CH2CH2CHs)ScCsHs in CsHs (equation 8). Temperature ( 0 C) 50 0.65 60 0.78 75 0.96 89.5 1.23 100.5 1.38 LlH o = + 3.5(3) kcal · mol-1 tlS o = + 10.5(5) e.u. (8) 89 0.4 0.2 - 0.0 0 "' Q) - ~ c -0.2 -0.4 -0.6 - + - - - r - - - r - - - - r - - - r - - - r - - - r - - - - , r - - - - - . r - - - . - - - . 0.0026 0 . 0027 0.0028 0.0029 0.0030 0.0031 1 I T(K) Figure 2. van't Hoff plot for the equilibrium between Cp*(17 5,'7 1-CsMe4CH2CH2CH2)Sc and 90 A purely intramolecular equilibration between a pendant alkyl group and a pendant aryl group can be established starting from the mixed-Cp scandocene chloride 4 (equation 9). Treating 4 with one equivalent of trimethylsilylmethyllithium quickly Cl / refluxing toluene Cl n • LICI 4 generates tetramethyisiiane and one major product identified as + LiCH 2 SiMe 3 {1 0) ca. 10 min. - SiMe 4 4 5 Heating benzene-de solutions of 5 for 12 hours does not effect the 1 H NMR spectrum. Adding small amounts of dihydrogen to the NMR tube containing benzene-de solutions of 5 91 results in H/D exchange of the propyl methyl hydrogens of 5 upon thermolysis, but does not otherwise affect the 1H NMR spectrum. An equilibrium is therefore inferred between 5 and ('7 5 -CsMe4CH2CsH3-m-CH3)('7 5 ,'71_CsMe4CH2CH2CH2)Sc which is catalyzed by (17 5 -CsMe4CH2CsH3-m-CH3)('7 5-CsMe4CH2CH2CH3)Sc-H. That dihydrogen does indeed interconvert 5 and (17 5 -CsMe4CH2CsH3-m-CH3) (17 5 .17 1-CsMe4CH2CH2CH2)Sc is supported by the following observations: (1) treating Cp*('7 5.17 1-CsMe4CH2-o-CsH4)Sc with dihydrogen (4 atm) rapidly yields Cp*(17 5 -CsMe4CH2CsHs)Sc-H, (i1) heating benzene-ds solutions of Cp*(17 5-CsMe4CH2CsHs)Sc-H under 4 atmospheres of H2 results in H/D exchange, specifically into the ortho benzylic sites. Based on previous results for H/D exchange between benzene-ds and hydrocarbons catalyzed by Cp*2Sc-H[4 1, the likely mechanism for benzylic H/D exchange involves Cp*(17 5-CsMe4CH2CsHs)Sc-H, Cp*('7 5 ,'7 1-CsMe4CH2-o-CsH4)Sc, and Cp*(17 5-CsMe4CH2CsHs)Sc-CsDs (along with H2, Relative Hf-H and Hf-alkyl BDEs can also be obtained. The mixed-Cp hafnocene refluxing xylenes ------!> - LiCI Cp*(175 -CsMe4CH2CH2CH3)HfCI2 6 (11) 92 Treating 6 with n-butyllithium under a dihydrogen atmosphere affords the dihydrido species 5 (equation 12). Cp*(YJ 5-CsMe4CH2CH2CH3)HfCI2 6 Cp*(YJ 5-CsMe4CH2CH2CH3)HfH2 (12) 2.2 n-butyiLi -2 butane - LiCI 7 Dihydride 7 has not been isolated cleanly; rather a mixture of 7 and the metallated propyl hydride 8 is obtained from petroleum ether. The equilibrium shown in equation 13 can be driven to the right by the periodic removal of H2 from the reaction vessel, thus allowing the isolation of clean samples of 8. ~ Hf ~H 7 (13) 8 Introducing dihydrogen into an NMR tube charged with 8 and benzene-da re-establishes the equilibrium shown in equation 13 with detectable concentrations of 7 and 8. The temperature dependence of the equilibrium (Table 3) indicates Li.H 0 = +15.5(3) kcal mol-1 and Li.S 0 =33.5(5) e.u. for equation 13. 93 Table 3. Equilibrium constants for the equilibrium between Cp*('7 5 ,'7 1 -CsMe4CH2CH2CH2)HfH, dihydrogen, and Cp*(77 5-CsMe4CH2CH2CH3)HfH2 in benzene-de (equation 13). 0 Temperature ( C) 70 3 .28 80 3.78 90 9.58 100.5 15.0 114 39.6 t.H o = + 15.5(3) kcal · mol-1 t.S o =+ 33(1) e.u. Relative Hf-aryl and Hf-H BDEs were also determined for comparison to values obtained via the intermolecular equilibrations described above. Thus, the metallated benzyltetramethylcyclopentadienyl hafnium hydride complex Cp*('7 5,'71_CsMe4CH2-o-CeH4)HfH has been prepared as shown in equations 14-16. refluxing xylenes Cp*(77 5-CsMe4 CH2CeHs)HfCI2 + LiCI 9 (14) 1) 2 CeHsCH2K -2 KCI Cp*(77 5-CsMe4CH2CeHs)HfCI2 9 2) H2, 80oC -2 CeHsCH3 Cp*(77 5 -CsMe4CH2C6Hs)HfH2 10 (15) 94 + H2 (16) 11 10 The equilibrium shown in equation 16 is driven to right by the periodic removal of dihyrogen from the reaction vessel. From the temperature dependence of this equilibrium a .t.H o = +8.5(3) kcal · mol-1 and a .t.S o = 19(5) e.u. were obtained. The similarity of the enthalpy for equation 16 to that for equation 3 {l:.H 0 =+6.0{3) kcal· mol-1) suggests only a small ring strain energy for metallation of the benzyl group of the ('7 5.'7 1-CsMe4CH2-o-C5H 4) ligand.£ 7 1 Table 4. Equilibrium constants for the equilibrium between Cp*('7 5 ,'7 1 -CsMe4CH2C5H 4)HfH, dihydrogen, and Cp*{17 5-CsMe4CH2C6Hs)HfH2 in benzene-d5 (equation 16). Temperature ( C} 0 59 3.56 73 5.90 86.5 9.60 98.5 17.2 113 19.0 .t.H o =+ 8.5(3) .68° =+ 19(1) 95 Attempts to equilibrate Sc-aryl and Sc-alkynyl complexes were unsuccessful, presumably due to the exceptionally strong Sc-carbon bonds of the latter. For example, Cp*('7 5 ,'7 1-CsMe4CH2-o-C6 H 4)Sc (12), prepared as shown in equations 17 and 18, reacts completely with tert-butylacetylene; no detectable amount of j Cp*('7 5 .'7 1-CsMe4CH2-o-C5H4)Sc 12 (18) When only 0.5 equivalent of Me3CCeC-H is used, no free acetylene remains. Thus the equilibrium constant for equation 19 must be greater than 102 M-1 . (19 ) 12 13 Determination of Aryl Substituent Effects on the Relative Sc- Aryl Bond Dissociation Energies. In order to d etermine the polarity of the Sc-aryl bond, a series of 96 compounds which can metallate at either of two differently substituted benzyl groups has been prepared (equation 20). X X y (20) y Any ionic contribution to the Sc-aryl BDE, and hence, the equilibrium constant, should respond to the electron releasing or electron withdrawing power of the substitutents X and Y. Each of the substituted benzyltetramethylcyclopentadienyl anions has been prepared from the appropriate benzyl Grignard reagent and tetramethylcyclopentenone, followed by dehydration of the resulting alcohol. Treatment of the isomeric mixture of tetramethylbenzylcyclopentadienes with n-butyllithium affords the lithium salt of the anion in moderate yield (Scheme 1). 97 X Scheme I 0 1. BrMgCH 2 C6 H4 -m-X 2.H 2 0 diethyl ether X X 2. n-Buli, -butane The mixed-Cp scandocene chloride compounds (17 5-CsMe4CH2C6H4-m-CH3) (17 5-CsMe4CH2C6Hs)ScCI (14), (17 5 -CsMe4CH2C6H4-m-CH3)(17 5-CsMe4CH2C6H4-m-CF3)ScCI (15), and (17 5-CsMe4CH2C6H4-m-CH3)(17 5-CsMe4CH2C6H4-m-NMe2)ScCI (16) are prepared by refluxing {(17 5-CsMe4CH2C6H4-m-CH3)ScCI2}n with the appropriate tetramethylbenzylcyclopentaienyl anion in toluene (equation 21) . 98 X 3 hr + n (21) refluxing toluene - LiCI 14, X= H 15, X= CF 3 16, X= NMe 2 Since 1H NMR analysis of the equilibrium mixture proved rather complex (even at 500 MHz), we have chosen (17 5-CsMe4CH2C5H4-m-CH3) as the reference substituted benzyl ligand. When metallated, e.g. for Cp*(1J 5,1J 1-CsMe4CH2-o-C5H3-p-CH 3)Sc, the (17 5-CsMe4CH2-o-C6H3-p-CH3) resonance appears characteristically at o 2.37. For the compounds described below, this signal consistently shifts from o 2.05-2.11 to o 2.37-2.38 on ortho-metallation (Table 5), thus serving as a reliable indicator. Treating the scandium chloride generates over the course of one hour one equivalent of methane and the isomeric mixture (1J 5 ,1J 1 -CsMe 4CH2-o-C5H3-p-CH3)(17 5 -CsMe4CH2C6Hs)Sc (17) and (17 5-C 5Me4CH 2C5H 4-m-CH3)(1J 5 ,1J 1-CsMe4CH2-o-C5H4)Sc (18) in a 1:2 ratio (Scheme 2). X 99 Scheme II 15 Heating this 1:2 mixture overnight at 110 o C in benzene-d5 does not alter this ratio. In order to establish that this 1:2 ratio of 17 and 18 is the equilibrium mixture, a small amou nt of dihydrogen was introduced to catalyze interconversion of 17 and 18, via o bond metathesis reactions with ('7 5-CsMe4CH2C5H4-m-CH3)('7 5-CsMe4CH2C6Hs)Sc-H. Again , no chang e from the stati stical 1:2 ratio is noted. 100 Treating ('7 5 -CsMe4CH2CsH4-m-CH3)('7 5-CsMe4CH2CsH4-m-CF3)ScCI with one equivalent of trimethylsilylmethyllithium rapidly generates tetramethylsilane and a ca. 1:1 mixture of ('7 5 -CsMe4CH2CsH4-m-CH3)('7 5 ,'7 1-CsMe4CH2-o-CsH3-P-CF3)Sc (19) and ('7 5,'7 1-CsMe4CH2-o-CsH3-p-CH3)('7 5-CsMe4CH2CsH4-m-CF3)Sc (20), presumably via ('7 5-CsMe4CH2CsH4-m-CH3)('75-CsMe4CH2CsH4-m-CF3)Sc-CH2SiMe3. When warmed to sooc for a few minutes, the ratio of 19 to 20 changes to 16:1. Continued heating at 50 ° C results in decomposition over a period of 1-2 hr; however, the ratio of 18 to 20 remains ca. 16:1. A much more thermally robust mixture of 19 and 20 is obtained using an alternative preparation. Treating ('7 5-CsMe4CH2CsH4-m-CH3)('7 5-CsMe4CH2CsH4-m-CF3)ScCI with one equivalent of phenyllithium generates the phenyl derivative ('7 5-CsMe4CH2CsH4-m-CH3)('7 5-CsMe4CH2CsH4-m-CF3)Sc-CsHs, wh ich is stable to loss of benzene at ambient temperature. Heating an isolated sample of (17 5-CsMe4CH2CsH4-m-CH3) (17 5-CsMe4CH2CsH4-m-CF3)Sc-CsHs in benzene-ds to 80 o C results in benzene loss and the 16:1 isomeric mixture£8] of ('7 5 -CsMe4CH2CsH4-m-CH3)('7 5 ,'7 1 -CsMe4CH2-o-CsH3-p- CF3)Sc (19) and ('7 5 , 17 1-CsMe4CH2-o-CsH3-p-CH3) (17 5-CsMe4CH2CsH4-m-CF3)Sc (20). Thermal decomposition of this mixture is much slower, requiring several days at 80 o C. The variable thermal stabilities of these compounds from batch to batch suggests decomposition is catalyzed by trace impurities, but the value of Keq is not in doubt (19 <l !> 20, Keq = 0.063(6)) Treating ('7 5-CsMe4CH2CsH4-m-CH3)('7 5-CsMe4CH2CsH4-m-NMe2)ScCI with one equivalent of LiCsHs cleanly generates ('7 5 -C 5 Me4CH2CsH4-m-CH3)('7 5 -CsMe4CH2CsH4-m-NMe2)ScCsHs. This complex has not been isolated, rather it is generated in situ and characterized by 1 H NMR. Heating benzene- ds solutions of ('7 5-C 5 Me 4CH2CsH4-m-CH3)('7 5 -CsMe4CH2CsH4-m-NMe2)ScCsHs overnight at 50 o C 101 generates CsH6 and a 1:1 mixture of ('7 5·'7 1 -C5Me4CH2-o-C5H3-p-CH3) (17 5-C5Me4CH2CsH4- m-NMe2)Sc (21) and ('7 5-C5Me4CH2CsH4-m-CH3)(f7 5,'7 1 -C5Me4CH2- o-C6H3-p-NMe2)Sc (22). As with X=H and CH3, heating solutions of the initially formed mixture at 80 o C for several hours does not alter the isomeric ratio (nor does it effect decomposition). In order to assure that a true thermodynamic equilibrium had been reached, once again a small amount of dihydrogen was introduced into a benzene-dB solution of the phenyl derivative (17 5-C5Me4CH2CsH4-m-CH3)('1 5-C5Me4CH2CsH4-m-NMe2)Sc-C6H5 to catalyze rapid interconversion of 21 and 22 via (17 5-C5Me4CH2C6H4-m-CH3)('7 5-C5Me4CH2C6H4-m-NMe2)Sc-H. After heating for 20 hr at 50°C no deviation from the original1:1 ratio of ('7 5.f7 1-C5Me4CH2-o-CsH3-p-CH3)('7 5-C5Me4CH2CsH4-m-NMe2)Sc (21) and ('7 5-C5Me4CH2C6H4-m-CH3)(f7 5,f7 1-C5Me4CH2-o-CsH4-p-NMe2)Sc (22) was noted. Thus, as for 17 and 18 no detectable substituent effect of the equilibrium constant for the two isomeric metallated benzyl complexes 21 and 22 obtains. Discussion Sigma bond metathesis has proven to be a convenient method to obtain information on the relative bond dissociation energies for these types of transition metal hydrides and hydrocarbyls. Due to interfering reactions, mostly likely involving the C-H bonds of th e (17 5-C5Me5) ligands, intermolecular equilibrations appear to be limited to the very stable hydride and phenyl derivatives. Although our data indicate that the alkynyl complexes are even more stable, side reactions such as catalytic d imerization of terminal acetylenes to gem-enynes£7) preclude equilibration. 102 By incorporating reactive groups into the cyclopentadienyl ligands and introducing systematic variations we have capitalized on the tendency for the ligand C-H bonds to participate in the a bond metathesis reactions. Moreover, the favorable entropy associated with these intramolecular metallation reactions (b.S0 driving force (Tb.S 0 =11 - 34 e.u.) provides an additional =3- 10 kcal· mol-1 at ambient temperature) to shift otherwise prohibitively endothermic equilibria to allow the concentrations of all participants to be accurately determined. These equilibrium constants (e.g. equations 8, 13, and 16) are no longer unitless, so that the issue of standard states [9] and transferability of Keq's or b.G o 's is potentially problematic. However, by comparing b.H 0 values obtained from the temperature dependence of the equilibrium constants, we can draw some meaningful comparisons. Another point of possible concern is the introduction of ring strain on metallation. Models suggest, and experiments appear to bear out (e.g. compare b.H0 values for equations 3 and 16) that there is little strain for a three-carbon link connecting the cyclopentadienyl ring carbon atom and the metal center. In the final series of experiments, those involving competitive metallation by scandium at either of two benzyl-substituted cyclopentadienyl ligands, the symmetry of the ligand system and the fact that the equilibrium constants are unitless eliminates these concerns completely. The enthalpies of the equilibrium, established by these sigma bond metathesis reactions (equation 1), allow the relative M-R and M-R' bond strengths to be determined if one assumes the Benson approximation [1 0] and cancellation of the solvation energies[ 11 •12]. In equation 3, for example, since the H-CsH5 BDE =11 0.9 kcal· mol- 1 [131, H-H BDE = 104.2 kcal· mol-1 [ 141 and the heat of solvation of H2 in benzene = -1.5 kcal · mol-1 [15], then the +6.0(3) kcal · mol-1 enthalpy of reaction indicates the BDE(Hf-H) is greater than BDE(Hf-CsH5) by (1 04.2 - 110.9 + 1.5 +6.0) = 0.8(3) kcal · mol-1. In equation 8, if the C-H BDE for Cp*(rJ 5-C5Me4CH2CH2CH2-H)Sc-CsH5 is assumed to be 98 kcal · mol-1 (the BDE for the 1 o C-H bonds of propane), then the 103 BDE(Sc-CeHs) is greater than BDE(Sc-CH2CH2CH2CsMe4) by 16.6(3) kcal· mol-1 , since BDE(H-CeHs) =110.9 kcal· mol-1 and t.H =+3.7(3) kcal · mol-1 . The following ordering for 0 Hf-R and Sc-R BDEs are calculated in an analogous manner (Scheme 3): Scheme Ill -~- BDE(Sc-C= CCMe3 ) t:. ;?: 27(5)kcal/mol J -~- BDE(H-C= CCMe3 ) BDE(Sc-H) t:. = 1.5(3) _j_ t t:. l BDE(Hf-C6 H 5 ) 1 l ! 21 (5) kcal/ mol - . BDE(H-C6 H 5 ) t:. = 7 +BDE(H-H) t:. = 22.2(3) BDE(Sc-CH 2 CH2 CH2 C 5 Me 4 ) t:. _j_ 0 .8(3)kcal/mol BDE(Sc-C6 H 5 ) t:. = 16.5(3) - - . BDE(Hf-H) t:. = 6 _L BDE(H-CH 2 CH2 CH 3 ) BDE(Hf-CH 2 CH2 CH2 C 5 Me4 ) Whereas the differences in metal-hydride and metal-phenyl bond strengths are essentially the same for Sc and Hf, the same cannot be said for the metal-phenyl vs. metai-CH2CH2CH2CsMe4 bond strengths. Some of the ca. 6 kcal· mol-1 discrepancy may be due to different ring strain energies for the pseudo trigonal scandium vs. pseudotetrahedral hafnium. On the other hand, the total strain energy for the hafnium derivative amounts to only ca. 2.5(3) kcal· mol-1, since according to equilibri~m 16, the BDE(Hf- H) is greater than BDE(Hf-CeH4CH2CsMe4) by 3.3{3) kcal · mol-1 . Whatever the 104 sources of these small effects, some general conclusions can be derived from the differences in bond strengths as the hybridization at the a carbon is varied. The order of decreasing BDEs is M-C(sp) > M-C(aryl) > M-C(sp3 ), the same order as for H-C BDEs for hydrocarbons. However, the relative differences for hydrogen-carbon bond strengths is smaller {BDE[H-C(sp)] = 132(5); BDE[H-C(aryl)] = 111; BDE[H-C(1°,sp3 )] = 98 kcal · mol-1} than those for the hafnium and scandium hydrocarbyls. Indeed, if the relative differences were the same, these equilibria would be essentially thermoneutral (i.e. t.H0 ,. 0). Additional, qualitative evidence which supports this point is provided by the intramolecular equilibration of (rJ 5 ,rJ 1-CsMe4CH2-o-CsHa-p-CHa)(rJ 5-CsMe4CH2CH2CHa)Sc (5) with ('7 5-CsMe4CH2CsHa-m-CHa}(rJ 5 ,rJ1-CsMe4CH2CH2CH2)Sc; the Sc-aryl isomer 5 being strongly favored. The origin of the effect which is responsible for the M-R BDEs increasing more rapidly with s character than do the H-R BDEs is not clear; however, it undoubtedly also results in stronger M-H BDEs than would be anticipated on the basis of the correlation suggested by Bryndza, Bercaw and co-workersJ 11 •161 For these early transition metal hydride compounds the BDE(M-H) is ca. 8 kcal · mol-1 stronger than this correlation predicts. The sensitivity of the transition metal-carbon bond energy to s character could be a reflection of considerable ionic (M 0+-ccS-) bonding, since s character at carbon increases its electronegativity. The experiments designed to probe this aspect of the Sc- aryl bonding, equilibria 20, reveal very little substituent effect on the Sc-aryl BDE[ 17l. Only for para-CFa is a measureable difference obtained. Para-NMe2 exerts its effect primarily through resonant 1r donation. Interestingly, the phenyl ring is held perpendicular to the equatorial plane of the scandocene moiety by the link to the cyclopentadienyl ligand, i.e. in resonance with the empty 1r symmetry (b2) scandium orbital. Thus, resonant donation is possible, but apparently is cancelled by the small inductive withdrawing character of the electronegative N atom. The generally small effect observed, even with the strongly withdrawing CFa 105 substituent, is quite significant, and dictates that there is little polarity for the Se-C bond in these compounds. These experimental results are in generally in accord with the picture presented by Goddard and Steigerwald[ 1 8] for the scandium-hydride of CI2Sc-H, for which the electronegative Cl ligands (and by extension the Cp* ligands) bond primarily via the Sc 4s orbitals, with the Sc-H bonding being 3d-1s in character and essentially covalent. It appears this same 3d character also dominates the Sc orbital used in the Sc-aryl bonding of compounds 17 - 22. 106 Table 5. 1 H and 13c NMR Data 8 • Compound Assignment o(ppm) Couplingb Li+(CsMe4CH2C5H4-m-CH3)" c 17 5-Cs(CH3)4CH2C5H4-m-CH3 11 5-Cs(CH3)4CH2C6H4-m-CH3 11 5-Cs(CH3)4CH2C6H4-m-CH3 17 5-Cs(CH3)4CH2C6H4-m-CH3 1.87 s 3.63 br 2.20 s 6.6-7.0 m Li+(CsMe4CH2C5H4-m-CF3)" c 17 5-Cs(CH3)4CH2C6H4-m-CF3 17 5-Cs(CH3)4CH2C6H4-m-CF3 17 5-Cs(CH3)4CH2C5H4-m-CF3 1.83 s 1.87 s 3.80 br 7.1-7.4 m Li+(CsMe4CH2C6H4·m·N(CH3)2)" c 17 5-Cs(CH3)4CH2C6H4-m-N(CH3)2 17 5-Cs(CH3)4CH2C6H4-m-N(CH3)2 17 5-Cs(CH3)4CH2C6H4-m-N(CH3)2 17 5-Cs(CH3)4CH2C6H4-m-N(CH3)2 1.93 s 1.90 s 3.60 br 2.80 s 6.3-7.0 m Cp*(t7 5-CsMe4CH2CH2CH3)ScCI (1) 17 5-Cs(CH3)5 17 5-Cs(CH3)4CH2CH2CH3) 17 5-CsMe4CH2CH2CH3 17 5-CsMe4CH2CH2CH3 17 5-C 5Me4CH2CH2CH3 1.90 s 1.85 s 2.01 s 0.79 t 1.24 m 2.35t 3JHH = 7 3JHH = 7 3JHH = 7 Cp*(f7 5,f7 1·CsMe4CH2CH2CH2)Sc (2) 11 5-C5(CH3)5 f7 5,f71.Cs(CH3)4CH2CH2CH2 5 f7 5,f7LC5Me4CH2CH2CH2 5 1 17 ,17 -C 5Me 4CH2CH2CH2 11 5-C5(CH3)5 11 5-C5(CH3)5 5 1 17 ,f7 ·CsMe4CH2CH2CH2 5 1 17 ,f7 -C 5Me4CH2CH2CH2 5 1 17 ,17 -C5Me4CH2CH2CH2 5 1 17 ,17 -C5(CH3)4CH2CH2CH2 17 ,f71.CsMe4CH2CH2CH2 1.83 s 1.57 s 1.97 s 0.77 t 2.73 tt 2.91 t 11.40 q 119.4 s not observed 38.76 t 28.9 t 117.0 s 119.6 s 138.2 s 3JHH = 6.6 3JHH = 6.6 3JHH = 6.6 1JcH = 126 1JcH = 120 1Jc H = 123 107 Cp*('7 5-CsMe4CH2CH2CH3)Sc(C5Hs) (3)1 77 5-Cs(CH3)5 77 5-Cs(CH3)4CH2CH2CH3) 77 5-CsMe4CH2CH2CH3 77 5-CsMe4CH2CH2CH3 77 5-CsMe4CH2CH2CH3 Sc(C5H5) 1.68 s 1.70 s 1.72 s 0 .82t 1.22 m 2.18 t 6.81 d 6.95 t 7.09 t 3JHH = = = = 3 JHH 3JHH =8 =8 7 3JHH = 7 3JHH 7 3JHH 7 3JHH 7 3JHH = 7 ('7 5- CsMe4CH2C6H4-m-CH3)('7 5-CsMe4CH2CH2CH3)ScCI (4) '7 5-Cs(CH3)4CH2C5H4-m-CH3 77 5-Cs(CH3)4CH2CH2CH3 '7 5-Cs(CH3)4CH2C5H4-m-CH3 '7 5-Cs(CH3)4CH2C5H4-m-CH3 77 5-Cs(CH3)4CH2CH2CH3 77 5-Cs(CH3)4CH2CH2CH3 77 5-Cs(CH3)4CH2CH2CH3 '7 5-Cs(CH3)4CH2C5H4-m-CH3 5 77 ,'7 LC 5(CH3)4CH2-o-C6H3-p-CH3 77 5-Cs(CH3)4CH2CH2CH3 77 5, '7 1-C5 (CH3)4CH2-o-C5H3-p-CH3 '7 5,'7 1-C 5 (CH3)4CH2-o-C6H3-p-CH3 77 5-Cs(CH3)4CH2CH2CH3 77 5-Cs(CH3)4CH2CH2CH3 77 5-Cs(CH3)4CH2CH2CH3 '7 5,'7 1-C 5 (CH3)4CH2-o-C6H3-p-CH3 1.85 s 1.88 s 1.98 s 2.05 s 2.11 s 3.79 b 2.33t 1.24 tq 0.81 t 6.81-7.08 m 1.53 s 2.02 s 2.08 s 2.17 s 2 .20 s 3.86 b 2.50 t 1.29 tq 0.87 t 6.945 6.70 d 8.05 d 3JHH = 8 3JHH 3JHH 3JHH =7 =7 =7 3JHH 3JHH =8 =8 Cp*('7 5-CsMe4CH2CH2CH3)HfCI2 (6) 77 5-Cs(CH3)5 77 5-Cs(CH3)4CH2CH2CH3 77 5-CsMe4CH2CH2CH3 77 5-C5 Me4CH2CH2CH3 77 5 -C5 Me4CH2CH2CH3 1.93 s 1.91 s 1.99 s 0.82 t 1.28 m 2.50 t = 3JHH 7.3 3JHH ~ 8 3JHH 8 .0 = 108 Cp*(17 5-C5Me4CH2CH 2CH3)HfH2 (7) 115-C5(CH3)5 17 5-C5(CH3)4CH2CH2CH3 115-C5Me4CH2CH2CH3 175-C5Me4CH2CH2CH3 175-C5Me4CH2CH2CH3 HfH2 2.07 s 2.10s 1.92s 0.88t 1.41 m 2.50 t 15.6 3JHH 3JHH 3JHH =8 =8 =8 1JcH 1JcH =125 =120 2JHH 2JHH =13 =13 Cp*(17 5.17 1-C5Me4CH2CH2CH2)HfH (8) 11 5 -C5(CH3)5 11 5.17 1-C5(CH3)4CH2CH2CH2 HfH 1.91 s 1.68 s 1.70 s 1.72 s 2.64s -0.60 m 0.31 m 1.90 m 2.02 m 2.58 m 2.78 m 13.2 5 1 17 ,17 -C5(CH3)4(CH2CH2CH2) 175,17 1-C5(CH3)4(CH2CH2CH2) 175.17 1-C5(CH3)4(CH2CH2CH2) 74.09t 37.58 t 27.28 t 17 5.17 1-C5Me4CH2CH2CH2 Cp*(17 5-C5Me4CH2C6H5)HfCI2 (9)d 175-C5(CH3)5 17 5-C5(CH3)4CH2C5H5 175-C5(CH3)4CH2C5H5 175-C5(CH3)4CH2C5H5 2.03s 1.97 s 2.03 s 3.80 br 6.9-7.2 m Cp*(17 5-C5Me4CH2C6H4-m-CH3)Hf(CH2C6H5)2 e 115-C5(CH3)5 17 5-Cs(CH3)4CH2C5H5 17 5-C5(CH3)4CH2C6H5 HfCH2C5H5 17 5-C5 (CH3)4CH2C5H5 HfCH2C5H5 1.75 s 1.74s 1.76 s 3.61 br 1.44 d 1.49 d 6.9-7.2 m 6.9-7.2 m Cp*(17 5-CsMe4CH2C6Hs)HfH2 (1 0) 115-Cs(CH3)5 175-Cs(CH3)4CH2C5H5 2.07 s 1.97 s 109 , 5-C5(CH3)4CH2C5H5 Hf-H 7.0-7.2 m 15.70 Cp*(ry 5,ry 1-C5Me4CH2·o-C5H4)HfH (11) ry 5-C5(CH3)5 , 5,ryLC5(CH3)4CH2-o-C5H4 11 5, '7LC5Me4CH2-o-C5H4 11 5, '7 1-C5Me4CH2-o-C5H5 Hf-H 1.95 s 1.74 s 1.87 s 1.97 s 2.02 s 3.76 d 3.92 d 7.2-7.5 m 14.15 br 2JHH = 17 2JHH 17 = Cp*(ry 5,ry 1-C5Me4CH2-o-C6H3-p-CH3)Sc 9 17 5-C5(CH3)5 ry 5,'7LC5(CH3)4CH2-o-C6H3-p-CH3 , 5,'7 1-C5(CH3)4CH2-o-C5H3-p-CH3 , 5,ryLC5(CH3)4CH2-o-C6H3-p-CH3 f7 5,'7LC5(CH3)4CH2-o-C6H3-p-CH3 1.96 s 2.03 s 1.48 s 2.37 s 4 .06 br 8.45 d 7.13 d 7.26 br Cp*('7 5-C5Me4CH2C5H5)ScCI 9 '7 5-C5(CH3)5 '7 5-C5(CH3)4CH2C5H4 , 5-C5 (CH3)4CH2C5H4 '7 5-C5(CH3)4CH2C5H4 1.88 s (7 methyls) 1.97 s (2 methyls) 3.75 br 7.0-7.2 m Cp*(f7 5,ry1-C 5Me4CH2-o-C6H4)Sc (12)9 17 5-C5(CH3)5 5 17 ,,LC5(CH3)4CH2-o-C5H4 1.95 s 2.01 s , 5,'7LC5(CH3)4CH2-o-C5H4 '7 5,'7LC5(CH3)4CH2-o-C5H4 4 .08 br 7.3-8.55 m 1.45s Cp*('7 5-C5Me4CH2C5H5)Sc-C=CCMe3 (13) ry 5-C5(CH3)5 , 5-C5(CH3)4CH2C5H4 C(CH3)3 '7 5-C5(CH3)4CH2C5H4 ry 5-C5 (CH3)4 CH2C5H4 1.88 s 1.97 s 2.07 s 1.30 s 3.73 br 7.0-7.3 m = 3JHH 7.4 3JHH = 7.4 110 (ry 5-C5Me4CH2CeH5)Sc(acac)2 d ry 5-C5(CH3)4CH2CeH5 ry 2-0C(CH3)CH(CH3)CO ry 5-C5(CH3)4CH2CeH5 2 17 -0C (CH3) CH (CH3) CO 5-C (CH3)4CH2CeH5 17 5 1.87 s 1.90 s 2.03 s 3.73 br 5.57 s 7.0-7.3 m (ry 5-C5Me4CH2CeH5)ScCI2 c ry 5-C5(CH3)4CH2CeH5 ry 5-C5(CH3)4CH2CsH5 ry 5-C5(CH3)4CH2CsH5 2.03 s 2.09 s 4.0 br 6.9-7.2 m (ry 5-C5Me4CH2CsH4-m-CH3)('7 5-C5Me4CH2CeH5)ScCI (14) ry 5-C5(CH3)4CH2CeH4-m-CH3 1.85 s (2 methyls) and ry 5-C5(CH3)4CH2CeH5 rJ 5-C5(CH3)4CH2CsH4-m-CH3 '1 5-C5(CH3)4CH2CsH4-m-CH3 '1 5-C5(CH3)4CH2CeH5 rJ 5-C5(CH3)4CH2CsH4-m-CH3 '1 5-C5(CH3)4CH2CeH5 1.95 s (1 methyl) 2.03 s (1 methyl) 2.11 s 3.70 br 3.70 br 6.8-7.2 m (ry 5-C5Me4CH2CeH4-m-CH3)(ry 5-C5Me4CH2CsH5)ScH '7 5-C5(CH3)4CH2CeH4-m-CH3 1.89 s and '7 5-C5(CH3)4CH2CsH5 rJ 5-C5(CH3)4CH2CsH4-m-CH3 '1 5-C5(CH3)4CH2CsH4-m-CH3 5 '7 -C5(CH3)4CH2CeH5 5 '7 -C5(CH3)4CH2CsH4-m-CH3 and '1 5-C5 (CH3)4CH2CeH5 1.90 s 2.02 s 2.06 s 2.09 s 3.79 br 3.79 br 6.8-7.2 m (ry 5-C5Me4CH2CeH4-m-CH3)Sc(acac)2 d rJ 5-C5(CH3)4CH2CeH4-m·CH3 '1 5-C5(CH3)4CH2CeH4-m-CH3 '7 5-C5(CH3)4CH2CsH4-m·CH3 '7 2-0C(CH3)CH(CH3)CO '72-0C(CH3)CH(CH3)CO ry 5-C5(CH3)4CH2CeH4·m·CH3 1.87 s 1.90 s 2.27 s 3.70 br 2.05 s 5.55 s 6.8-7.3 m 111 ('7 5-C5Me4CH2CsH4-m-CH3)ScCI2 e 175-C5(CH3)4CH2CsH4-m-CH3 175-C5(CH3)4CH2CsH4-m-CH3 175-C5(CH3)4CH2CsH4-m-CH3 175-C5(CH3)4CH2CsH4-m-CH3 2.20 5 2.285 2.07 5 4.33 br 6.8-7.1 m 175-C5(CH3)4CH2CsH4-m-CF3 1.80 5 and 17 5-C5(CH3)4CH2CsH4-m-CH3 175-C5(CH3)4CH2CsH4-m-CH3 175-C5(CH3)4CH2CsH4-m-CH3 17 5-C5(CH3)4CH2CsH4-m-CF3 17 5-C5(CH3)4CH2CsH4-m-CH3 and 175-C5(CH3)4CH2CsH4-m-CF3 1.82 5 1.85 5 2.035 2.08 5 3 .72 br 3.72 br 6.8-7.4 m ('7 5-C5Me4CH2CsH4-m-CH3)('7 5-C5Me4CH2CsH4-m-CF3)ScCsH5 17 5-C5(CH3)4CH2CsH4-m-CH3 1.63 5 and 17 5-C5(CH3)4CH2CsH4-m-CF3 175-C5(CH3)4CH2CsH4-m-CH3 17 5-C5(CH3)4CH2CsH4-m-CH3 17 5-C5(CH3)4CH2CsH4-m-CF3 17 5-C5(CH3)4CH2CsH4-m-CH3 17 5-C5(CH3)4CH2CsH4-m-CF3 Sc-CsH5 1.7505 1.758 5 1.761 5 2.07 5 3.45 br 3.58 br 6.7-7.5 m ('7 5-C5Me4CH2CsH4-m-CH3)('75-C5Me4CH 2CsH4-m-NMe2)ScCI (16)9 17 5-C5(CH3)4CH2CsH4-m-NMe2 1.86 5 and 17 5-C5(CH3)4CH2CsH4-m-CH3 17 5-C5(CH3)4CH2CsH4-m-CH3 17 5-C5(CH3) 4CH2CsH4-m-N (CH3)2 17 5-C5(CH3)4CH2CsH4-m-N ( CH3)2 17 5-C5(CH3)4CH2CsH4-m-CH3 17 5-C5(CH3)4CH2CsH4-m-CH3 and '7 5-C5(CH3)4CH2CsH4-m-N (CH3)2 1.89 5 1.97 5 2.21 5 2.09 5 2.59 5 3 .63 5 3 .70 5 6.4-7.2 m 112 '7 5-C5(CH3)4CH2C6H4-m·CH3 1.787 s and '1 5-C5(CH3)4CH2C6H4-m·NMe2 '1 5-C5(CH3)4CH2C6H4-m-CH3 1.792 s 1.796 s 1.86 s 2.06 s 2.51 s 3.62 br, s 3.70 '1 5-C5(CH3)4CH2C6H4-m-NMe2 6.5-7.5 m '7 5-C5(CH3) 4CH2C5H4-m-N (CH3)2 '7 5-C5(CH3)4CH2C6H4-m·CH3 '7 5-C5(CH3)4CH2C5H4-m-NMe2 '7 5-C5(CH3)4CH2C6H4-m·CH3 Sc-CsH5 ('7 5-C5Me4CH2C6H5}(f7 5,f7 1-C5Me4CH2-o-C6H3-P-CH3)Sc (17)9 f7 5,f7 1-C5(CH3)4CH2-o-C6H3-p-CH3 1.53 s and '7 5-C5(CH3)4CH2C5H5 '7 5,f7 1-C5(CH3) 4CH2-o-C6H3-p-CH3 '7 5-C5(CH3)4CH2C5H5 f7 5,f7LC5(CH3)4CH2-o-C6H3-P-CH3 f7 5,f7 1-C 5(CH3) 4CH2-o-C6H3-p·CH3 5 and 1.96 s 1.993 s 1.998 s 2 .38 s 3 .90 br, s 4.09 br, s 6.7-8.5 m '1 -C5(CH3)4CH2C6H5 ('7 5-CsMe4CH2C6H4-m-CH3}('7 5,f7 1-C5Me4CH2-o-C6H4)Sc (18)9 5 '1 -C5(CH3)4CH2C6H4-m-CH3 and '7 5,'7 1-C5 (CH3)4CH2-o-C6H4 '7 5-Cs(CH3)4CH2C6H4-m-CH3 '1 5-C5(CH3)4CH2C6H4-m-CH3 f7 5,'7 1-C5(CH3)4CH2-o-C6H4 '7 5-C5(CH3)4CH2C5H4-m-CH3 and f7 5,'7 1-Cs(CH3)4CH2-o-C5H4 5 '1 -C5(CH3)4CH2C5H4-m-CH3 1.50 s 1.93 s 1.99 s 2.04 s 2.06 s 4 .09 br, s 3 .87 br, s 6.7-8.5 m 1.39 s and f7 5,'7 1 -C 5(CH3)4CH2-o-C6H3-p-CF3 '7 5-C5(CH3)4CH2C5H4-m-CH3 '7 5,'7 1-Cs(CH3)4CH2·o-C5H3-p-CF3 '1 5-C5(CH3)4CH2C6H4-m-CH3 1.85 s 1.945 1.97 s 2.06 s 3.78 br, s 3.86 br, s 113 5 '1 -C5(CH3)4CH2C5H4-m-CH3 and f7 5,'7 1 -C5(CH3)4CH2-o-C6H3-p-CF3 6.64-8.41 m ('7 5-C5Me4CH2C6H4-m-CF3){f7 5,'7 1 -C5Me4CH2-o-C6H3-p-CH 3)Sc (20)g '7 5, '7 1 -Cs(CH3)4CH2-o-C5H3-p-CH3 and 5 '1 -C5(CH3)4CH2C5H4-m-CF3 '7 5,'7 1 -C5(CH3)4CH2-o-C5H3-p-CH3 f7 5,f7LC5(CH3)4CH2-o-C6H3-p-CH3 5 '7 -C5(CH3)4CH2C5H4-m-CF3 '7 5,'7 1 -C5(CH3)4CH2-o-C6H3-p-CH3 and '1 5-C5(CH3)4CH2C5H4-m-CF3 1.51 s 1.87 s 1.93 s 1.99 s 2.38 s 3.83 br 4.09 br 6.6-8.4 m 5 ('7 -C5Me4CH2C6H4-m-NMe2)(f7 5,'7 1 -C5Me4CH2-o-C6H3-P-CH3)Sc (21) and 5 ('7 -C5Me4CH2C6H4-m-CH3)(f7 5,f7 1·C5Me4CH2-o-C6H3-P·NMe2)Sc (22)9 f7 5,f7 1 -C5(CH3)4CH2-o-C6H3-p·CH3 f7 5,f7 1 -C5(CH3)4CH2-o-C6H3-p-NMe2 '7 5-C5(CH3)4CH2C5H4-m-CH3 '1 5-C5(CH3)4CH2C5H4-m-NMe2 f7 5,f7 1-C5(CH 3) 4CH2-o-C6H3-p-N(CH3)2 '1 5-C5(CH3)4CH2C5H4-m-N(CH3)2 '1 5-C5(CH3)4CH2C5H4-m-CH3 '7 5,'7 1 -Cs(CH3)4CH2-o-C6H3-P-CH3 '1 5-Cs(CH3)4CH2C5H4-m-CH3 '7 5 ,'7 1-C 5 (CH3)4CH2-o-C6H3-P-N(CH3)2 '7 5 ,'7 1·Cs(CH3)4CH2-o-C6H3-p-CH3 '1 5-Cs(CH3)4CH2C5H4-m-NMe2 r.J 5,'7 1-Cs(CH3)4CH2-o-C6H3-P-N(CH3)2 f7 5,f7LC5 (CH3)4CH2-o-C6H3-p-CH3 77 5-Cs(CH3)4CH2C5H4-m-NMe2 77 5 -Cs(CH3)4CH2C5H4-m-CH3 1.58 s 1.61 s 1.94 s 1.95 s 2.01 s 2.05 s 2.09 s 2.16 s 2.75 s 2.47 s 2.38 s 2.06 s 3.84s 3.95 s 4.07 s 4.16 s 6.3-8.5 m a All 1 H NMR spectra were recorded in benzene-d5 at 400 MHz unless otherwise noted. All signals are reference to internal tetramethylsilane unless otherwise noted. b Coupling constants are reported in hertz. c The 1 H NMR spectrum was recorded in THF-da at 90 MHz. d The 1H NMR spectrum was recorded in CDCI3 at 90 MHz. e The 1 H NMR spectrum was recorded in benzene-d5 at 90 MHz. f The 1 H NMR spectrum was recorded in cyclohexane-d12 and referenced to the resi dual protio impurity. 114 g The 1 H NMR spectrum was recorded in benzene-ds at 500 MHz. Note: The following abbreviations are used to denote the multiplicity of the NMR signals: s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; br, broad. 115 Experimental Section General Considerations. All manipulations were performed using glovebox, high-vacuum[ 191or Schrenk techniques. Solvents were purified by distillation from an appropriate drying agent under a N2 atmosphere and either used immediately or vacuum-transferred from "titanocene" or sodium-benzophenone. Benzene-d6 and cyclohexane-d12 were purified by vacuum transfer from activated molecular sieves (4 A, Linde) and then from "titanocene". 1H and 13C NMR spectra were recorded on Varian EM-390 (90 MHz, 1H), JEOL GX400Q (400 MHz, 1H; 100 MHz, 13C) and Brucker WM500 (500 MHz, 1H) spectrometers. Elemental analyses were performed by Mr. Larry Hen ling of the CIT Analytical Laboratory. Synthetic Procedures Preparation of Li+(Cs(CH3)4CH2C6H4-m-X)- (X= H[20l, CH3, CF3 and N(CH3)2)J21] Each ligand was prepared by the reaction of tetramethylcyclopentenone with the corresponding substituted benzyl Grignard reagent, followed by dehydration with para-toluene sulfonic acid when needed. In each case an excess of the substituted Grignard reagent (ca. 1.5 mol/mol of ketone) was allowed to react with tetramethylcyclopentenone. Due to the high boiling points of the ligands, purification and the final characterization was made by preparing the corresponding lithium salts. (a) X= H; To 54 mmol of BrMgCH2C5H5 in ca. 100 cm 3 of diethyl ether was added 36 mmol of tetramethylcyclopentenone. The reaction m ixture was refluxed overnight, then quenched with ca. 30 cm3 of H 20. The organic products were extracted into diethyl ether (2 X 50 cm 3 ). The ethereal extracts were concentrated to ca. 60 cm 3 and 0.5 g para-toluene sulfonic acid was added. The reaction mixture was stirred for 10 minutes at which time ca. 50 cm 3 of H20 and 1 g of NaHC03 were added. The organic product was extracted with diethyl ether (2 X 116 50 cm 3), dried over Na2S04 and concentrated under high vacuum to yield 7.1 g of yellow oil. To this 7.1 g of oil in ca. 100 cm 3 of diethyf ether (at -80 • C) was syringed 21 cm 3 of a 1.6 M n-butyllithium/hexane solution. After stirring for three hours at ambient temperatures a yellow precipitate had formed. This solid was coffected by filtration, washed twice with diethyl ether and dried in vacuo to yield 4.2 g of Li+(CsMe4CH2C5Hs)- (53% yield based on tetramethylcycfopentenone). (b) X= CH3; Foffowing the procedure described above (for X= H), tetramethylcycfopentenone (5.0 g) and excess BrMgCH2C5H4-m-CH3 ultimately affords 4.4 g of u+(CsMe4CH2C5H4-m-CH3)- (52% based on tetramethyfcycfopentenone). It should be noted that treatment of the initaiffy isolated organic product with para-toluene sulfonic acid is not necessary due to the alcohols spontaneous dehydration during work-up. (c) X= CF3; Again foffowing the procedure described above (for X= H), tetramethyfcycfopentenone (5.0 g) and excess CIMgCH2C5H4-m-CF3 ultimately affords 3 .8 g of Li+(CsMe4CH2C5H4-m-CF3)- (37% yield based on tetramethyfcycfopentenone). The reaction time for the deprotonation of HCsMe4CH2C5H4-m-CF3 by n-butyffithium should be kept to a mimimum (ca. 1 hour). Longer reaction times result in a darkening of the reaction mixture and a reduced yield of the lithium salt. (d) X= NMe2; Meta-N,N-dimethylaminobenzyf potassium was prepared in N,N-dimethyf-meta-tofuidine foffowing previously described proceduresl22l. Meta-N,N-dimethyfaminobenzylmagnesium bromide was prepared by stirring the potassium salt with anhydrous Mg8r2 etherate in diethyl ether, foffowed by filtration to remove the KBr. To the filtrate was carefully added 5 g of tetramethylcyclopentenone and the resulting mixture was refluxed for 2 hours. The reaction was quenched with ca. 30 cm3 of an aqueous solution containing 1 g of NaOH. The organic layer was separated and the aqueous layer was washed twice with 50 cm 3 portions of diethyf ether. The ether washings were combined and the volatiles were removed in vacuo. Residual 117 N,N-dimethyl-meta-toluidine was removed by gently heating (30-40°C) for 2-3 hours under hard vacuum. To the resulting alcohol was added 40 cm 3 of diethyl ether and 2 .0 g of para-toluene sulfonic acid. After stirring approximately 15 minutes a water layer became evident. To this mixture an aqueous solution containing 2.0 g of NaOH was added, the organic phase was separated, and the aqueous solution was washed twice with 50 cm 3 portions of diethyl ether. The solvent was removed in vacuo to afford 5.1 g of viscous yellow oil. To the 5.1 of crude ligand in ca. 50 cm 3 of petroleum ether was added 12.5 cm 3 of n-butyllithium in hexane (1.6 M). The reaction was stirred for two hours, then filtered to collect the precipitate. The product was washed with petroleum ether and dried to afford 2.0 g (21 % based on tetramethylcyclopentenone) of pale yellow product. Cp*(77 5-C5Me4CH2CH2CHs)ScCI (1). ('7 5-C5Me5)ScCI2£ 14l (4.0 g, 15.9 mmol) and Li(77 5-C 5Me4CH2CH2CH3 ) [ 2 3] (3.0 g, 17.6 mmol) were heated in benzene at 110° C for three days. The solvent was removed in vacuo and the non-volatiles were washed into a flask with petroleum ether (40 cm 3) which was then attached to a frit assembly. The resultant suspension was filtered to remove the LiCI. Removal of the solvent in vacuo affords 1 as an amber oil in near quantitative yield (90-95% pure by 1 H NMR). Cp*('7 5,'7 1-C 5Me 4CH2CH2CH2)Sc (2). Cp*(77 5-C5Me4CH2CH2CH3)ScCI (2.5 g, 6.6 mmol) and LiCH3 (0.145 g, 6.6 mmol) were stirred for one hour in ca. 20 cm 3 of benzene. The suspension was filtered and the benzene was replaced with petroleum ether. Cooling to -78° C resulted in the precipitation of product which was collected as a pale yellow solid by filtration (0.9 g, 40%). Analysis: Found(calculated) o/oC 75.22 (74.13), o/oH 9.33 (9.33). ( 77 5-C 5Me4CH2CsH4-m-CHs) ( 77 5-C5Me4CH2CH2CH3)ScCI (4). (77 5-C5Me4CH2CsH4-m-CHs)ScCI2 (0.64 g, 1.88 mmol) and excess Li(77 5-C5Me4CH2CH2CH3) (0.42 g, 2.47 mmol) were refluxed overnight in ca. 25 mL of 118 toluene. The solvent was replaced with petroleum ether and filtered to afford the product as an amber oil (essentially pure by 1 H NMR). Cp*(17 5-CsMe4CH2CH2CH3)HfCI2 (6). (17 5-CsMes)HfCI3 (1 0.2 g, 24.3 mmol) and Li(17 5-CsMe4CH2CH2CH3) (4.2 g, 24.7 mmol) were refluxed in xylenes for three days under an argon atmosphere. Removal of the xylene solvent (under reduced pressure) followed by an aqueous/methylene chloride work-upl24l gave 12.6 g (95%) of 6 as off-white crystals. Cp*(f7 5 ,f71-CsMe4CH2CH2CH2)HfH (8). A flask was charged with 6 (4.2S g, 7.S mmol) and ca. 50 cm 3 of toluene. 1 0. 7 cm 3 of a 1.6 M solution of n-butyllithium in hexane (2.2 equivalents of n-butyllithium) was added to the toluene suspension (at -7S° C) under an argon flush. The argon was then replaced with one atmosphere of dihydrogen and the solution was allowed to warm with stirring. When the uptake of dihydrogen had ceased, the toluene was replaced with petroleum ether and the suspension was filtered . Removal of the petroleum ether in vacuo afforded an amber oil which consisted of an approximately equimolar ratio of Cp*(17 5-CsMe4CH2CH2CH3)HfH2 (7) and Cp*(f7 5 ,f71 -CsMe4CH2CH2CH2)HfH (S). A benzene solution of this mixture was transfered to a small glass bomb which was heated to S0° for several hours, then evacuated to remove the generated dihydrogen. After three cycles the contents of the bomb were transfered to a frit assembly and the benzene was replaced with petroleum ether. Cooling to -?soc and filtering afforded 1.S g (50%) of 8 as a colorless powder. Analysis: Found(calculated) %C 55.34 (55.43), %H 7.07 (7.13). Cp*(17 5 -C5 Me 4CH2CsHs)HfCI2 (9). (17 5-CsMes)HfCI3 (5.1 g, 12.1 mmol) and Li+(17 5 -CsMe4CH2CsHs)- (2.7 g, 12.4 mmol) were refluxed in xylenes for three days under an argon atmosphere. Removal of the xylene solvent (under reduced pressure) followed by an aqueous/ methylene chloride work-up gave 6.5 g (90%) of 9 as an off-white solid. Analysis: Found(calculated) %C 52.24(52.40), % H 5.53(5.75) . 119 Cp*('7 5-CsMe4CH2C5Hs)Hf(CH2C5Hs)2· Approximately 20 cm3 of benzene was condensed onto a m ixture of 9 (1 .94 g, 3.26 mmol) and excess benzylpotassium (1.39 g, 11 .1 mmol). After stirring for 1 hour the suspension was filtered and the benzene was replaced with petroleum ether. Cooling to -78 o C and filtering afforded 1.69 g (73%) of pale yellow Cp* ('7 5-CsMe4CH2C5H5) Hf(CH2C5H5)2. Cp*('7 5,'7 1-CsMe4CH2-o-C5H4)HfH (11 ). A large glass bomb was charged with 1.49 g (2.1 0 mmol) of Cp*('7 5-CsMe4CH2C6Hs)Hf(CH2C5H5)2 and 25 cm 3 of toluene. The bomb was cooled to -196°C, evacuated, then refilled with one atmosphere of dihydrogen (ca. 4 atmospheres at 25°C}. The solution was then stirred at 140°C for one day, recharged with dihydrogen, and stirred for a second day at 140 o C. The 1 H NMR spectra of the crude reaction mixture showed the presence of Cp*('7 5-CsMe4CH2C5H5)HfH2 (1 0) and Cp*('7 5.'71_CsMe4CH2-o-C5H4)HfH (11). Three cycles of heating the mixture to 80 ° C followed by removal of the generated dihydrogen in vacuo produces colorless 11, which was isolated in 79% yield (0.88 g) from -78 ° C petroleum ether. Analysis: Found(calculated) %C 59.39(59.48), %H 6.51 (6.53). Cp*('7 5-CsMe4CH2C5Hs)ScCI. Cp*ScCI2 (1 .4 g, 5.5 mmol) and Li('7 5-CsMe4CH2C5Hs) (1 .2 g, 5.5 mmol) were refluxed for 1 day in ca. 25 cm 3 toluene. Filtering the reaction mixture and removing the solvent in vacuo afforded Cp*('7 5-CsMe4CH2C5Hs)ScCI as an amber oil which was essentially pure by 1H NMR. Cp*('7 5,ry 1-CsMe4CH 2-o-C5H4)Sc (12). Cp*(ry 5-CsMe4CH2C5Hs)ScCI (1.97 g, 4.62 mmol) and methyllithium (1 00 mg, 4.55 mmol) were stirred for 3 hours in ca. 20 cm3 of benzene. The benzene solvent was replaced with petroleum ether and the LiCI was removed by filtration. Cooling the solution to -78°C and filtering afforded 0.54 g (30% yield) of colorless 12. Analysis: Found(calculated) %C 79.36 (79.97), %H 8.31 (8.52). 120 Generation of Cp*(,., 5,,., 1 -CsMe4CH2-o-C6H3-p-CH3)Sc. Cp*ScCI2 (90 mg, 0.36 mmol) and u(,., 5-CsMe4CH2C5H4-m-CH3) (83 mg, 0.36 mmol) were placed into a small glass bomb with 5 cm3 of toluene and heated to 110 oC for 1 day. To this reaction mixture was added LiCH2SiMe3 (34 mg, 0.36 mmol). After stirring overnight the toluene was removed in vacuo and the scandium product was extracted with benzene-d5. This product, which was not isolated, exhibited a 1 H NMR spectrum consistent with Cp*('7 5 ,,., 1 -CsMe4CH2-o-C6H3-p-CH3)Sc. ('7 5-CsMe4CH2C6H4-meta-X)Sc(acac)2· (a) X= CH3; Sc(acac)3 (15.0 g, 43.9 mmol) and Li('7 5 -CsMe4CH2C6H4-meta-CH3) (11.2 g, 48.2 mmol) were refluxed in ca. 75 cm3 of toluene for 12 hours. The toluene was removed in vacuo and the air-stable product was washed three times (3 X 50 cm 3) with anhydrous methyl alcohol to remove the u+(acac)-. The product was dried in vacuo to yield 11.6 g (56%) of pale yellow ('7 5-CsMe4CH2C6H4-meta-CH3)Sc(acac)2. (b) X= H; This air-stable compound was prepared in an analogous manner from 4.4 g of Sc(acac)3 and 2.8 g of Li('7 5 -CsMe4CH2C6Hs) to yield 4.1 g (71%) of product. Analysis: Found(calculated) %C 67.75(68.71), %H 7.18(7.32). ('7 5-CsMe4CH2C6H4-meta-X)ScCI2. (a) X= CH3; ('75-CsMe4CH 2C6H4-meta-CH3)Sc(acac)2 (11.6 g, 24.8 mmol) and AICI3 (6.61 g, 49.6 mmol) were stirred overnight in ca. 100 cm 3 of toluene. The resulting toluene suspension was heated to 100 oCto dissolve the sparingly soluble products. Allowing the hot toluene solution to slowly cool to room temperature resulted in the precipitation of ('7 5-CsMe4CH 2C6H4-meta-CH3)ScCI2, which was collected by filtration (5.2 g, 61 %). (b) X= H; This compound was prepared in an analogous manner from 4.14 g of ('7 5 -CsMe 4CH2C6Hs)Sc(acac)2 and 2.44 g of AICI3 to yield 2.2 g (74%) of ('7 5-CsMe 4CH2C6Hs)ScCI2. Analysis: Found(calculated) % C 59.18(58.74), % H 6.00(5.85). 121 ( 17 5-CsMe4CH2CsH4-m-CH3){17 5-CsMe4CH2CsHs)ScCI (14). ( 175-CsMe4CH2CsHs)ScCI2 (0.95 g, 2.9 mmol) and Li(17 5-CsMe4CH2CsH4-m-CH3) (0.67 g, 2.9 mmol) were refluxed overnight in ca. 15 cm 3 of toluene. Filtration, followed by removal of the solvent in vacuo yielded 14 as an amber oil which was essentially pure by 1H NMR. ('7 5-CsMe4CH2CsH4-m-CH3){'7 5-CsMe4CH2CsH4-m-X)ScCI (15, X= CF3; 16, X= NMe2). (15) X= CF3; (17 5-CsMe4CH2CsH4-m-CH3)ScCI2 (0.95 g, 2.79 mmol) and Li(17 5-CsMe4CH2CsH4-m-CF3) (0.80 g, 2.80 mmol) were refluxed for 3 hours in ca. 20 cm 3 of toluene. The toluene was replaced w ith petroleum ether and the solution was filtered. Cooling the petroleum ether solution to -78 o C and filtering afforded 0. 77 g (47%) of 15 as an off-white powder. (16) X= NMe2; This compound was prepared in an analogous manner from 0.96 g (2.8 mmol) of (17 5 -CsMe4CH2CsH4-m-CH3)ScCI2 and 0.76 g (2.9 mmol) of Li(17 5 -CsMe4CH2CsH4-m-NMe2) to yield 0.78 g (50%) of 16 as an off-white powder. Analysis: Found(calculated) o/oC 73.60(75.05), o/oH 7.59(8.10), o/oN 2.34(2.50). Synthesis of ('7 5-CsMe4CH2CsH4-m-CH3)(f75 ,'71 -CsMe4CH2-o-CsH4)Sc (18) and {f7 5 ,f7 1-CsMe4CH2-o-CsH3-p-CH3)('7 5 -CsMe4CH2CsHs)Sc (17) in a 2: 1 ratio. Methyllithium powder (60 mg, 2.7 mmol) was added to a benzene solution of 14 (1.3 g, 2.5 mmol) and the reaction mixture was stirred for 2 hours. The benzene was replaced with petroleum ether and the solution was filtered. Cooling to -78 o C and filtering afforded 18 and 17 in a 2:1 mixture (0.65 g, 54%). Analysis: Found(calculated) o/oC 81.83 (82.47), o/oH 8.13 (8.18). Synthesis of ('7 5-c 5 Me 4CH2CsH4-m-CH3}(f7 5 ,f7 1-CsMe4CH2-o-CsH3-p-CF3)Sc (19) and ( 175 , 17 1-C 5Me 4CH2-o-C 6H 3-p-CH3)('7 5-CsMe4CH2CsH4-m-CF3)Sc (20} in a 1:1 (kinetic) ratio. Ca. 10 cm3 of benzene was condensed onto a mixture of 15 (0. 71 g, 1.21 mmol) and LiCH 2SiMe3 (115 mg, 1.22 mmol) . After stirring for 1 hour the benzene was replaced w ith 122 petroleum ether and the solution was filtered. Cooling the petroleum ether solution to -78 o C and filtering afforded 390 mg (59%) of the two isomers in approximately a 1:1 ratio (1H NMR). ('7 5-CsMe4CH2C6H4-m-CH3)('7 5-CsMe4CH2C6H4-m-CF3)ScC6Hs. Approximately 20 cm3 of benzene was condensed onto a mixture of 15 (0.92 g, 1.57 mmol) and phenyllithium (0.15 g, 1.79 mmol). This mixture was stirred for 1 hour at room temperature and filtered. The benzene was replaced with petroleum ether and cooled to -78°C. Filtering afforded 0.53 g (54%) of the product as an off-white powder. Analysis: Found(calculated) %C 73.86 (76.66), %H 6.80(7.08). Generation of 17 and 18 in a 16:1 (thermodynamic) ratio. Method a) Heating a sample of the 1:1 mixture of ('7 5-CsMe4CH2C6H4-m-CH3)(7J 5,'71-CsMe4CH2-o-C6H3-p-CF3)Sc and ('7 5,'7 1-CsMe4CH2-o-C6H3-p-CH3)('7 5 -CsMe4CH2C6H4-m-CF3)Sc (isolated as desribed above) to 50°C in benzene-d5 shifts the ratio to 16:1 within 1/2 hour. This ratio was measured by integrating the benzyl methyl resonance of each isomer. Once established this ratio does not change, although decomposition is observed to slowly occur over the course of several hours. Method b) Heating benzene-d5 solutions of ('75-CsMe4CH2C5H4-m-CH3) ('7 5-CsMe4CH2C5H4-m-CF3)ScC5Hs overnight at 50 oc cleanly generates benzene and a mixture of 17 and 18 in a 16:1 ratio. Samples generated in this manner do not undergo detectable decomposition until heated for several hours at 80°C. Generation of ('7 5,'7 1-CsMe4CH2-o-C6H3-p-CH3)('7 5-CsMe4CH2C6H4-m-NMe2)Sc (19) and ('7 5-CsMe4CH2C6H4-m-CH3)('7 5,'7 1-CsMe4CH2-o-C6H3-p-NMe2)Sc (20) in a 1:1 ratio. Method a) Compound 14, 22 mg, and C5Hsli, 4 mg, were sealed in an NMR tube with benzene-d5. After ca. 15 minutes the 1 H NMR spectrum showed a single species identified as ( '7 5-CsMe4CH2C5H4-m-CH3) ( 77 5 - CsMe4CH2C5H4-m-NMe2)ScC6Hs. Heating 123 (77 5-CsMe4CH2C6H4-m-CH3) (77 5-CsMe4CH2C6H4-m-NMe2)ScC6Hs overnight at 50 oc resulted in the generation of C5H5 and a 1:1 mixture of 19 and 20. Heating this sample to 80 o C did not alter this ratio. Method b). (77 5-CsMe4CH2C6H4-m-CH3) (77 5-CsMe4CH2C6H4-m-NMe2)ScC6Hs was generated as described above in a NMR tube attached to a Kontes needle valve. A catalytic amount of dihydrogen (ca. 20 torr) was admitted to the tube which was then sealed with a torch. Heating the tube overnight at 50°C resulted in the generation of C5H5 and a 1:1 mixture of 19 and 20. Measurement of Equilibrium Constants. The equilibration of Cp*('7 5 ,'7 1 -CsMe4CH2CH2CH2)HfH (B) and dihydrogen with Cp*(77 5 -CsMe4CH2CH2CH3)HfH2 (7) (equation 13) was studied in the following manner. Compound 8 (16.8 mg, 0.034 mmol) was placed in a sealable NMR tube along with 0.481 g of benzene-d5 and the tube was attached to a Kontes teflon-stoppered valve. The solvent was cooled to -78°C and the tube was evacuated. The benzene-d5 was kept frozen at -78 o C, but the remainder of the tube was allowed to warm to room temperature. Dihydrogen (282 torr) was admitted to the tube which was then sealed with a torch. The volume of space above the frozen benzene-d5 was measured, allowing for the initial amount of dihydrogen in the NMR tube to be calculated. Subsequent equilibrations were carried out in a constant temperature ethylene glycol bath which was judged to be constant to± 0.2°C. The equation describing the equilibrium constant is given below. Keq = [7] The concentrations of 8 and 7 were determined from the relative heights of the resonances at o 1.92 and 2.05, respectively. The total amount of dihydrogen in the NMR tube is given by equation 22. 124 (22) A knowledge of the total amount of dihydrogen in the NMR tube allowed for the concentration of dihydrogen in solution to be calculated as desribed elsewhere[71. In order to obtain a unitless equilibrium constant the concentration of each species was expressed as a mole fraction. The equilibrium constants for equation 3 and 16 were determined in an analogous manner. The equilibration of Cp*(77 5-CsMe4CH2CH2CH3)Sc-CsHs (3) with benzene and 2 (equation 8) was studied by sealing 29 mg (0.085 mmol) of Cp*('7 5 ,'7 1-CsMe4CH2CH2CH2)Sc (2) in a NMR tube along with 56 mg of CsHs (0.72 mmol) and 0.487 g of cylohexane-d12· The equation describing the equilibrium constant is given by: Keq = [3] The concentrations of 2 and 3 were determined from the relative heights of their respective Cp* resonances. The total amount of CsHs in the NMR tube is expressed by equation 23. (23) The amount of dissolved CsHs was calculated at each temperature using Henry's law. Concentrations were expressed as mole fractions in order to obtain unitless equilibrium constants. A lower limit for the (Sc-alkynyl) BDE relative to the (Sc-aryl) BDE was established in the following manner. Tert-butylacetylene (0.016 mmol) was condensed into an NMR tube which contained 0.033 mmol of Cp*('7 5 ,'7 1-CsMe4CH2-o-CsH4)Sc, 12, and 0.65 g of benzene-d6 and the tube was sealed. The 1H NMR spectrum revealed a 1 to 1 mixture of 12 125 and a species identified as Cp*('7 5-CsMe4CH2C5Hs)Sc-C""CCMe3 , 13; there was no evidence for unreacted tert-butylacetylene. Assuming as much as 2% tert-butylacetylene could be presence yet undetected by 1H NMR, lower values for Keq and ~Go in equation 19 are 1.4 x 102 M-1 and -3 kcal · mol-1 , respectively. If a value for ~So of -15 e.u. is assumed, ~Ho ~ 7.5 kcal· mol-1 , and since H-C""CCMe3 BDE =132(5) kcal· mol-1 [ 1 3] and H-C5H5 BDE = 111 kcal· mol-1[ 1 3], the (Sc-alkynyl) BDE must be~ 29(5) kcal· mol-1 stronger than the (Sc-aryl) BDE. References: (1) Bryndza, H. E.; Calabrese, J. C., Marsi, M.; Roe, D. C.; Tam, W.; Bercaw, J. E. J. Am. Chern. Soc. 1986, 108, 4805. (2) Caution should be exercised in concluding too much from this type of reactivity, however, since these reaction could, in large part, be due to the very strong thermodynamic driving force. Moreover, coordination of the Bronsted acid to the metal center prior to deprotonation is likely in many cases. (3) Mulliken, R. S. J. Chern. Phys., 1955, 23, 1833. (4) 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. Chern. Soc., 1987, 109, 203, and references cited therein. 5) McAlister, D. R.; Erwin, D. K.; Bercaw, J. E. J. Am. Chern. Soc., 1978, 100, 5966. (6) While ('7 5,'7 1-CsMe4CH2-o-C6H3-p-CH3)('7 5-CsMe4CH2CH2CH3)Sc (5) is the predominant product, several minor products are also observed (the largest of which represents~ 10% of the total scandium containing products). Due to the extreme solubility of 5 this complex could not be isolated free of the minor products and the identy of these minor products could not be firmly established. 126 (7) The validity of this statement depends on (i) the H-C6 H5 and ('7 5-C5Me4CH2C5H4-o-H) BDEs being equal and (it) the Hf-aryl BDE not depending on the orientation of the phenyl ring relative to the hafnocene equatorial plane. The possibility also exists that these two effects may cancel, at least in part. (8) 1H NMR integration was carried out at 50°C. (9) We have chosen units of mole fraction in calculating the equilibrium constants for these cases. (1 0) Benson, S. W. Thermochemical Kinetics, 2nd edition, Riley, New York NY 1976. (11) Bryndza, H. E.; Fong, L. K.; Paciello, R. A.; Tam, W.; Bercaw, J. E. J. Am. Chern. Soc., 1987, 109, 1444. (12) (a) Janowicz, A. H.; Periana, R. A.; Buchanan, J. M.; Kovac, C. A .; Stryker, J. M.; Wax, M. J.; Bergman, R. G. Pure & Appl. Chern. 1984,56, 13 (b) Jones, W. D.; Feher, F. J. J. Am. Chern. Soc. 1985, 107,620 (c) Bergman, R. G. Science 1984,223,902 (d) Jones, W. D.; Feher, F. J. J. Am. Chern. Soc. 1984, 106, 1650 (e) Bruno, J. W.; Marks, T. J.; Morss, L. R. J. Am. Chern. Soc. 1983, 105, 6824 (f) Wax, M. J.; Stryker, J. M.; Buchanan, J. M.; Kovac, C. A.; Bergman, R. G. J. Am. Chern. Soc. 1984, 106, 1121 (g) Bryndza, H. E.; Fultz, W. C.; Tam, W. Organometallics 1985, 4, 939. (13) McMillen, D. F.; Golden, D. M.Ann. Rev. Phys. Chern., 1982,33,493. (14) Gingerich, K. A. J. Chern. Phys. 1971, 54, 3720. (15) Cook, M. W.; Hanson, D. N.; Adler, B.J. J. Phys. Chern., 1957,26,748. (16) A roughly linear correlation is found for BDE(H-R) vs. BDE(Sc-R) (R =alkynyl, aryl, alkyl), with a slope of 0.8, not the unity value of the Bryndza/Bercaw correlation . Similar, less-than-unity slopes have been noted by others for transition metal-hydrocarbyl bonds: see reference 12(d) and the contribution by Bergman and co-workers in Polyhedron, in press. 127 (17) So far as we are aware, no substituent effects on the M-aryl (M = alkali or alkaline earth metal) BDE's have been examined. Thus, these results suffer from the lack of a point of reference. (18) Steigerwald, M. L.; Goddard, W. A. J. Am. Chem. Soc., 1984, 106,308. (19) Burger, B. J.; Bercaw, J . E. in "New Developments in the Synthesis, Manipulation and Characterization of Organometallic Compounds", ACS Symposium Series, A. Wayda and M. Darensbourg, Eds., in press. (20) An alternative synthesis of CsMe4(CH2CsHs)H has been published: Blaha, J.P.; Wrighton, M.S. J. Am. Chem. Soc. 1985, 107, 2694. (21) We have found that the meta-methoxide derivative Li+('7 5-CsMe4CH2CsH4-m-OMe)- can also be prepared by methods analogous to those described here. (22) Schlosser, M.; and Hartmann, J . Angew. Chem., Int. Ed. Engl. 1973, 12, 508. (23) Threlkel, R. S. and Bercaw, J. E. J. Organomet. Chem., 1977, 136, 1. (24) Wolczanski, P. T., Bercaw, J . E. Organometallics, 1982, 1, 793. 128 Chapter V The Mechanism of Dihydrogen Elimination from Hydrido Alkyl Derivatives of Permethylhafnocene. The Molecular Structure of Cp*2HfCH2CH(CH3)CH2. 129 Introduction There are numerous examples of transition metal complexes, although often of limited stability, which contain cis-alkyl and hydride ligands. Complexes of this type are frequently observed as the products of arene or alkane C-H bond activation at low valent transition metal centers. The reverse process of oxidative addition, reductive elimination, is a common mode of decomposition for cis-hydrido alkyl or aryl complexes, particularly for the later transition metals. l11 Studies of the thermal decomposition of early transition metal complexes containing cis hydride and alkyl ligands are rare, a notable exception being the elimination of isobutane from Cp*2Zr(CH2CHMe2)H to form pentamethylcyclopentadienyl metallated zirconium species.£2 1 In contrast, the thermolysis of benzene-de solutions of Cp*2Hf(CH2CHMe2)H has been shown to result in dihydrogen elimination concurrent with the formation of Cp*2HfCH 2 CHMeCH2 .£3) Intrigued by this latter observation, we have undertaken an investigation into the mechanism of dihydrogen loss from Cp*2Hf(CH2CHMe2)H (and from complexes of the type Cp*2Hf(R)H in general), the results of which are reported herein. Results and Discussion As previously shown by Roddick and Bercaw, heating benzene-d6 solutions of the isobutyl hydride complex Cp*2Hf(CH2CHMe2)H (1) results in the formation of a 1:1 mixture of Cp*2HfH2 and the (3 substituted metallacyclobutane Cp*2HfCH 2 CHMeCH 2 (2) plus 1/2 equivalent of isobutane.£3) In the presence of excess isobutene metallacycle 2 is formed quantitatively (equation 1). 130 toluene, 11 0°C. (1) isobutene isobutane 1 2 In order to confirm the structure of the hafnium metaffacycfe 2, the crystal structure has been determined by X-ray diffraction techniques. The molecular structure of 2 is presented in Figure 1 and the atomic labeling is given in Figure 2. As can be seen from the figures, the metaffacycfic ring of 2 adopts a puckered conformation in the solid state with the f3 carbon being displaced 0.384 A out of the plane formed by Hf, C12 and C13. This distortion from planarity may serve to minimize an unfavorable steric interaction between one of the Cp* rings and the {3-methyf substituent of the hafnacycfobutane ring . In contrast, titanacyclobutanes of the type Cp2TiCH2CH(R)CH2, which are active olefin metathesis catalysts (unlike 2), have been found to have essentially planar metaffacycfic ringsJ4 l The formation of 2 has been proposed to occur by the concerted foss of dihydrogen from 1 via a simple four-centered transition state at the Hf(IV) (Figure 3)_[3] -!-- 1 cp•,Htd .. I ' '', I I ' H-·· '' ' .':a-l Figure 3 . The proposed transition state for the conversion of 1 to 2 and dihydrogen. 131 Figure 1. The molecular structure of Cp*2HfCH2CHMeCH2 (2). The thermal ellipsoids are shown at the 50% probability level. 132 Qr--.. ~C10' \ 0 G b~,N y.- ca C9M 8 Figure 2 . The atomic labeling for Cp*2HfCH2CH(CH3)CH2 (2). 133 Table 1. Complete Bond Distances and Angles for Cp*2HfCH2CH(CHs)CH2 (2). Atom Atom Dist. (A) Atom Atom Atom Angle Hf Hf Hf Hf Hf Hf Hf Hf Hf Hf Hf Hf C11 C11 C11 Hf Hf C1 C1 C1 C2 C2 C3 C3 C4 C4 C5 C6 C6 C6 C7 C7 C8 C8 C9 C9 C10 C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C12 C13 C12 C13 C14 R1 R2 C2 C5 C1M C3 C2M C4 C3M C5 C4M C5M C7 C10 C6M C8 C7M C9 C8M C10 C9M C10M 2.563(1 0) 2.509(1 0) 2.524(9) 2.502(9) 2.539(10) 2.529(7) 2.558(9) 2.582(9) 2.558(9) 2.532(9) 2.249(1 0) 2.238(11) 1.522((15) 1.581 (16) 1.530(20) 2.225 2.251 1.382(14) 1.434(14) 1.529(17) 1.404(13) 1.493(16) 1.409(12) 1.504(16) 1.411 (13) 1.497(16) 1.778(14) 1.430(11) 1.433(12) 1.469(14) 1.403(13) 1.512(17) 1.414(13) 1.516(17) 1.389(13) 1.500(15) 1.489(17) R1 C12 C5 C1M C1M C3 C2M C2M C4 C3M C3M C5 C4M C4M C4 C5M C5M C10 C6M C6M C8 C7M C7M C9 C8M C8M C10 C9M C9M C9 C10M C10M C13 C14 C14 Hf Hf C1 C1 C1 C2 C2 C2 C3 C3 C3 C4 C4 C4 C5 C5 C5 C6 C6 C6 C7 C7 C7 C8 C8 C8 C9 C9 C9 C10 C10 C10 C11 C11 C11 R2 C13 C2 C2 C5 C1 C1 C3 C2 C2 C4 C3 C3 C5 C1 C1 C4 C7 C7 C10 C6 C6 C8 C7 C7 C9 C8 C8 C10 C6 C6 C9 C12 C12 C13 138.4 70.2(4) 109.1 (9) 126.8(9) 123.8(9) 107.8(8) 124.8(9) 127.1 (9) 108.4(8) 125.5(9) 124.3(9) 108.2(8) 126.4(9) 125.0(9) 106.3(8) 125.8(8) 127.7(8) 105.7(7) 123.8(8) 128.8(8) 109.1 (8) 123.3(9) 127.5(9) 107.6(8) 125.5(9) 126.4(9) 108.7(8) 124.6(9) 126.4(9) 109.0(8) 126.8(9) 123.1 (9) 112.5(9) 112.2(11) 112.4(11) R1 and R2 are the centroids of the C1 through C5 and the C6 through C1 0 rings, respectively. 134 Such a mechanism is consistent with the observation that the rate of conversion of 1 to 2 is independent of the concentration of added isobutene and that no intermediates are observed when the reaction is followed by 1 H NMR.[3] Thus, a mechanism involving the rate-determining insertion of isobutene into the Hf-H bond of 1 to form Cp*2Hf(CH2CHMe2)2, followed by the rapid loss of isobutane through a four-centered transition state, is considered unlikely. These observations, however, do not eliminate mechanisms involving (i) metallated pentamethylcyclopentadienyl intermediates or (ii) pentamethylcyclopentadiene intermediates derived from the migration of the hydride ligand A mechanism involving the metallated pentamethylcyclopentadienyl derivative conversion of Cp*2Hf(CH2C5H5)2 to Cp*2HfCH2-o-C5H4 and toluene proceeds through the metallated benzyl derivative Cp*('7 5,'7 1-C5Me4CH2)HfCH2C6H5[5]). Thus, Cp*('7 5,1'] 1 -C5Me4CH2)HfCH2CH(CH3)2 (3) was prepared from the metallated iodide complex Cp*(17 5.'7 1-C5Me4CH2)Hfl (4) (equation 2).£5] ~ Hf ~ 4 CIMgCH 2 CHMe 2 -Mg(CI)I ~u ~ 3 (2) 135 Thermolysis of 3 in the presence of >1 0 equivalents of added isobutene results in a complex mixture of products including ca. 30% of 2. This contrasts with the essentially quantitative conversion of the isobutyl hydride to complex 2. Thus, a mechanism involving the intermediacy of Cp*(ry 5 ,ry1 -CsMe4CH2)HfCH2CH(CH3)2 is deemed unlikely. The presence of f3 hydrogens in 1 precludes deuterium labeling studies of the hydride ligand of 2. While samples of Cp*2Hf(CH2CDMe2)D can be prepared from Cp*2Hto2 and isobutene, H/D exchange between the Cp* ligands and the deuteride position occurs upon thermolysis (presumably via Cp*2HfD2, although Cp*2Hf(CH2CDMe2)D is the only species observed by 1H NMR) and consequently makes the desired labeling study impossible. We therefore turned our attention to hydrido alkyl complexes of permethylhafnocene which lack f3 hydrogens. The method used to prepare hydrido alkyl derivatives of permethylhafnocene which lack {3-hydrogens is shown below (equations 3-5). (3) (4) Cp*2Hf(H)I + M'R Cp*2Hf(H)R + M'l (5) 5; R =CH2CsHs 6; R = CsH4-o-CH3 M' =Li or K These syntheses take advantage of the facile hydrogenation of Cp*2Hf(CH2CMe3)l to yield Cp* 2Hf(H)l and neopentane. Treating Cp*2Hf(CH2CMe3)l with D2 cleanly affords 136 Cp*2Hf(D)I. Subsequent treatment of Cp*2Hf(H)I with alkyl potassium reagents affords the corresponding hydrido-alkyl complexes 5 and 6. Mild thermolysis of benzene-ds solutions of 5 affords an equimolar mixture of Cp*2HfH2 and the hafnabenzometallacyclobutene derivative Cp*2HfCH2-o-CsH4 (7) plus one half of an equivalent of toluene (equation 6). 2 Cp*2Hf(CH2CsH5)H - - - - t > Cp*2HfCH2-o-CsH4 + Cp*2HfH2 + CsH5CH3 5 (6) 7 Traces of the ortho-tolyl hydride complex 6 are also observed in the final product mixture (the final concentration of 6 amounts to ~10% of 7 or Cp*2HfH 2 as determined by 1 H NMR). The decomposition of 5 follows first order kinetics for~ 2 half lites. Construction of an Arrhenius plot from the rate data given in Table 2 yields the activation parameters ~ H* = 25.6(5) kcal·mol-1 and ~s* = -8(2) e.u .. Table 2. Rate Constants for the Thermal Decomposition of Cp*2Hf(CH2C6 H 5)H (5). Temperature ( C) 0 46 1.1 (1) 62 7.7(2) 70.5 20(1) 4.2(1) ~H* = 25.6(5)kcal mol-1 ~s* = -8(2) e.u.c a These rate data were obtained in benzene-ds solvent in the absence of 4. The organometallic products of the reaction were 7 and Cp*2HfH2. b This rate constant was obtained in benzene- ds solvent in the presence of ca. 10 equivalents of 4. The organometallic products of the thermolysis were 7 and Cp*2Hf(H) I. c For the purpose of constructing an Arrhenius plot the rate constants obtained in the absence of 4 were halved and therefore reflect the actual rate of dihydrogen loss from 5. 137 If the thermolysis of 5 is carried out in the presence of a large excess of Cp*('7 5,'7 1-CsMe4CH2)Hfl4, first order kinetics are again observed, but with a rate constant which is approximately half that observed in the absence of 4 (Table 2). Neither toluene nor Cp*2HfH2 is detected, rather equimolar amounts of metallacycle 7 and hydride iodide 1 are the final products (equation 7). (7) 5 7 These results are consistent with the initial formation of 7 and dihydrogen upon thermolysis of 5. In the absence of an effective H2 trap, such as 4[6] , the dihydrogen reacts with another equivalent of 5 to generate Cp*2HfH2 and toluene. As the reaction proceeds and the relative concentration of 7 increases, the hydrogenation of 7 to give Cp* 2Hf(C6 H4-o-CH3)H ,6, becomes competitive with the hydrogenation of sJ 7l In the presence of ca. 10 equivalents of 4, however, the formed dihydrogen preferentially reacts with 4 to give Cp* 2Hf(H)I, and 5 decomposes at a rate which is half that observed in the absence of 4 . In an order to probe the mechanism by which dihydrogen is lost from 5, the isotopically labeled benzyl hydride derivatives Cp*2Hf(CH2C5Hs)H Sa, Cp*2Hf(CD2C5Ds)H (Cp*-d15)2Hf(CH2C5Hs)H Sf were prepared (equation 8). 138 benzene (8) -KI 5a; W=X=Y=Z=H 5b; W=X=H, Y=Z=D 5c; W=Y=Z=H, X=D 5d; W=X=Z=H, Y=D 5e; W=H, X=Y=Z=D 5f; W=D, X=Y=Z=H As can be seen from Table 3, deuteration of the benzyl methylene group (5d) has no effect on the rate of decomposition of 5 (kHiko = 1.1 (1) at 62 o C), thus precluding an a:-H abstraction process as a significant pathway. This is in contrast to Cp*2Hf(CH2C6 H 5 ) 2, which has been shown to decompose predominately by an a:-H abstraction mechanism forming an unstable hafnium benzylidene species and toluene.£51 Similarly, deuteration of the pentamethylcyclopentadienylligands (5f) has little effect on the rate of decomposition (kHiko =0.95(5) at 62°C). This result eliminates a rate limiting coupling of a ring methyl hydrogen with the hydride or the benzyl ligand as a likely mechanism, that is, the rate decomposition of 5. Furthermore, deuterium is not scrambled into the hydride position during the thermolysis (62°C) of 5f eliminating a pre-equilibrium between by rate determining transfer of an ortho-benzyl hydrogen to the cyclopentadienyl methylene group as a significant pathway. A substantial kinetic isotope effect is observed upon deuteration of the benzyl ligand (5b) (kHiko = 5.9(4) at 62 o C), consistent with an ortho-benzyl C-D bond cleavage in the rate determining step. Interestingly, deuteration of the hydride ligand (Sc) results in an inverse isotope effect upon pyrolysis (kH/ko =0.78(3) at 62°C). Likewise, an inverse isotope effect is observed when the decomposition of Se (both the benzyl and hydride ligands are deuterated) is compared to that of Sb (only the benzyl ligand is deuterated) (kH/ko = 0.57(5) at 62 o C, Figure 4). 139 Table 3. Rate constants for the thermal decomposition of deuterium labeled 5 at 62 o C in the presence of excess Cp*('7 5,'7 1 -C5Me4CH2)Hfl.a Compound 5 4.21 (7) 0.71 (3) 5.39(1) 3.94(14) 1.25(4) 4.42(7)b 5b 5c 5d 5e 5f 5.9 0.78(3) 1.1 (1) 3.4(5) 0.95(5) a These thermolyses were carried out in benzene-d5. b This value was obtained in the absence of Cp*(f7 5,'7 1-C5Me4CH2)Hfl. Hence, the value listed is half the observed rate of decomposition. 5.2 5.0 4.8 (.) 4.6 c 4.4 4.2 4.0 3.8 Oe+O 2e+5 4e+5 6e+5 8e+5 1e+6 Time (sec) Figure 4. First order plots of the decomposition of Cp*2Hf(CD2C5Ds)D (5e) and 140 Two alternative mechanisms consistent with the rate determining cleavage of an ortho-benzyl C-H bond are depicted in Schemes 1 and 2. The mechanism shown in Scheme 1 involves the concerted loss of dihydrogen via a simple four-centered transition state at the Hf(IV) center. ~ Hf-H ~H,C6 H5 ~~ + H2 ~ 7 5 '\ I =I= Scheme 1. A potential mechanism for the formation of 7 and dihydrogen from 5. Although this mechanism is consistent with the rate determining cleavage of an ortho-benzyl carbon-hydrogen bond, it does not offer a straightforward explanation for the inverse kinetic isotope effect observed upon deuteration of the hydride ligand. In theory, an inverse deuterium isotope effect for a single, very endothermic, elementary step can occur if that step involves the transfer of hydrogen from a relatively low frequency 141 -5 1! slow 7 Scheme 2. An alternative mechanism for the formation of 7 and dihydrogen from 5. 142 oscillator to a much higher frequency oscillator.£8] While the hydride is indeed being transferred to a higher frequency oscillator (a H-H bond is being formed at the expense of a Hf-H bond), the available evidence suggests that the overall reaction is to a first approximation, thermoneutral. This evidence is provided from thermochemical data£9] which indicate that the Hf-aryl bond dissociation energy (BDE) of 5 and the Hf-H BDE of 2 are approximately equal. Using the literature value of 104 kcal·mol-1 for the H-H BDE[1 0] and a value of 110 kcal·mol-1 for the ortho-benzyl BDE£11 1, and assuming a t.S o of +15 e.u. for the transformation of 5 to 7 and dihydrogen£ 121 (which makes the loss of H2 ca. 5 kcal·mol-1 more favorable at 62°C}, we can estimate that the overall free energy change is small. It should be noted that this estimate neglects any ring strain in the metallacyclic product which would serve to make the reaction more endothermic. The rapid reversible migration of the hydride ligand to the pentamethylcyclopentadienyl ring proposed in Scheme 2 is followed by the rate limiting oxidative addition of an ortho-benzyl C-H bond to the Hf(ll) center£ 13], and the final rapid elimination of dihydrogen to afford 7. This mechanism is consistent with the normal primary isotope effect observed for 5b (as was the mechanism depicted in Scheme 1). The mechanism presented in Scheme 2 is also consistent with the inverse isotope effects observed when the hydride ligand is deuterated, that is, the deuterium prefers to reside on the higher frequency oscillator (C-D rather than Hf-0). Although rare, intramolecular migrations of hydride and alkyl ligands to cyclopentadienyl rings have been directly observed. [ 14•1 5] Such migrations have, on several other occasions, been inferred. [ 1 6] Brintzinger has proposed that the hydrogenation of Cp*2Zr(CH2CMe3)X proceeds by the initial attack of dihydrogen on the cyclopentadienyl-zirconium a-bond (through a transition state analogous to that proposed in Scheme 2 for the final elimination of dihydrogen), followed by the transfer of the en do cyclopentadiene hydrogen to the neopentyl group. [ 171 143 Pyrolysis of benzene-ds solutions of Cp*2Hf(CsH4-o-CH3)H 6 in presence of 4 cleanly affords a 1:1 ratio of metallacycle 7 and Cp*2Hf(H)I (equation 9). The Excess 4, C6 D6 (9) - Cp* 2Hf(H)I 6 7 disapppearence of 6 follows first order kinetics when thermolysed in the presence of Cp*(17 5 .'7 1 -CsMe4CH2)Hfl (Table 4). In the absence of 4, the themolysis of 6 results in the formation of 7 with lesser amounts of Cp*2HfH2 as well as uncharacterized decomposition products. It is interesting to note that the activation parameters for the conversion of 6 to 7 Table 4. Rate Constants for the Thermal Decomposition of 6 in the presence of excess Cp*('7 5 ,'7 1-CsMe4CH2)Hfl (in benzene-d5). Temperature ( 0 C) 51 1.82(2) 8.39(30) 42.3(14) 49.8(12)a 64.5 80 80 0.85(4) t.H* = 24.0(5) kcal mol-1 t.s* = -11 (2) e.u. a The rate of decomposition of Cp*2Hf(CsH4-o-CH3)D in the presence of Cp* ('7 5 ,'7 1-CsMe4CH2)Hfl. and dihydrogen (t.H* =24.0(5) kcal·mol-1 and t.S* =-11 (2) e .u.) are very similar to those obtained f or the conversion of 5 to 7 and dihydrogen (see Tables 2 and 4). As w ith 5, an 144 6,-----------------------------------------~ 5 (.) c 4 E:J • complex 6-d1 complex 6 3 ~----~--~-----r----------~--~----~--~ Oe+O 1e+4 2e+4 3e+4 4e+4 Time (sec) Figure 4. First order plots of the rate of decomposition of Cp*2Hf(C5H4-o-CH3)H (6) and inverse isotope effect is observed when the hydride ligand of 6 is replaced by deuterium (kH/ko =0.85(4), Figure 4). Based on this inverse isotope effect, a mechanism analogous to Scheme 2 is tentatively proposed, that is, a rapid pre-equilibrium is established with the hydride migrating between hafnium and the Cp* ligand, followed by the rate limiting addition of a tolyl methyl C-H bond to the 14e- hafnium center. The rapid elimination of dihydrogen affords 7. It must be emphasized that the evidence for this mechanism is limited to a single inverse isotope effect; however, the suggested mechanism does allow one to envision a common decomposition pathway for various Cp*2Hf(R)H derivatives. 145 Conclusions In summary, it has been demonstrated that thermolysis of benzene-ds solutions of the permethylhafnocene alkyl hydride complexes 1, 5, and 6 results in the elimination of dihydrogen and the formation of new hafnium-containing species. In no case does decomposition proceed by the coupling of the hydride ligand with the adjacent alkyl substituent, nor does the coupling of ring methyl hydrogens with the alkyl substituent seem to be operative. For 1 the initially formed complex is Cp*2HfCH2CHMeCH2 (2), which has been crystallographically characterized. For both 5 and 6 the initially formed complex is Cp*2HfCH2-o-CsH4. Kinetic isotope effects of the decomposition of Cp*2Hf(CH2C6 H 5)H (5), are consistent with the rate limiting step cleavage of the ortho-benzyl C-H bond. Interestingly, an inverse isotope effect is found when the hydride ligands of 5 and 6 are deuterated, consistent with a rapid pre-equilibrium involving the transfer of the hydride ligand to the Cp* ring, followed by the rate limiting cleavage of an ortho-benzyl C-H bond (Scheme 2). Unfortunately, we are unable to test for an analogous inverse isotope effect for reaction 1 due to the facility with which the hydride position of Cp*2Hf(CH 2CHMe2)H (1) exchanges with the Cp* hydrogens. 146 Table 5. 1 H NMR dataa. Compound Assignment o(ppm) Cs(CH3)5 Cs(CH3)4CH2 1.90 1.57 1 .73 1.89 2 .07 CH2CH(CH3)2 0.77 d 1.13 d not observed 1.45 d 1.92 d CH2CH(CH3)2 CH2CH(CH3)2 Cp*2Hf(H)I Cs(CH3)5 Hf-H 2.03 15.3 Cs(CH3)5 CH2CsHs CH2CsHs Hf-H 1.90 1.17 6 .8-7.2 m 13.6 Cs(CH3)5 CsH4-o-CH3 CsH4-o-CH3 1.85 2.15 6.02 m ?.Om 14.0 Hf-H Couplingb 3 JHH 3 JHH =9 =9 2JHH = 7 2JHH 7 = a Spectra were recorded in benzene-ds and are referenced to internal tetramethylsilane. b Coupling constants are given in hertz. 147 Experimental General Considerations. All manipulations were performed using glovebox, high-vacuum[ 18] or Schlenk techniques. Solvents were purified by distillation from an appropriate drying agent under a N2 atmosphere and either used immediately or vacuum-transferred from "titanocene" or sodium-benzophenone. Benzene-de was purified by vacuum transfer from activated molecular sieves (4 A, Linde) and then from "titanocene". 1 H NMR spectra were measured on Varian EM-390 (90 MHz, 1H NMR) and JEOL GX400Q (400 MHz, 1 H NMR) spectrometers. Elemental analyses were performed by Mr. L. Henling of the CIT analytical department. Synthetic Procedures: Cp*2Hf(D)I. Cp*2Hf(CH2CMe3)l (1.13 g, 1.7 mmol) was placed in a large glass bomb (with a volume of approximately 400 ml) with ca. 20 ml of benzene. Deuterium gas (one atmosphere) was admitted into the bomb at -196°C. This solution was stirred at room temperature for 2 days (Note: heating results in incorporation of deuterium into the Cp* rings). The contents of the bomb were transferred to a frit assembly and the product (0.46 g, 47%) was isolated from -78°C petroleum ether. Cp*2Hf(H)I. This complex is prepared as described for Cp*2Hf(D)I and can be isolated in ca. 50% yield. The reaction time can be reduced to 1 day if the hydrogenation is carried out at 80° c. Cp*(ry 5 ,ry 1 -CsMe4CH2)Hf(CH2CH(CH3)2) (3). Cp*(ry 5 ,ry 1-CsMe4CH2)Hfl (0.66 g, 1.15 mmol) and 1.2 mmol CIMg(CH2CH(CH3)2) in d iethyl ether were stirred in benzene for 1 hour. The solvent was replaced with petroleum ether. Filtering and cooling to -78 o C 148 affords 288 mg of pale orange 3. Elemental analysis: Found(calculated) %C 55.44 (57.08), %H 7.38 (7.58). Cp*2Hf(CH2CsHs)H (5). Cp*2Hf(H)I (0.50 g, 0.87 mmol) was stirred 1 hour with benzylpotassium ( 0.14 g, 1.04 mmol) in ca. 20 ml of benzene. The solution was filtered and the solvent was replaced with petroleum ether. Cooling the solution to -78 o C and filtering affords 2 as an off-white powder (0.20 g, 43%). The deuterated analogs 5b-5f were prepared similarly from Cp*2Hf(H)I and the appropriate benzylpotassium reagent. Cp*2Hf(CsH4-o-CHs)H (6). Cp* 2Hf(H)I (785 mg, 1.36 mmol) and an excess of LiCsH4-o-CHs (250 mg, 2.55 mmol) were stirred overnight in benzene, then filtered to remove the LiCI. The product (0.30 g, 41%) was isolat ed from -78°C petroleum ether. Elemental analysis: Found(calculated) %C 59.36 (59.93), %H 6.86 (7.08). Kinetic Measurements of Thermal Decomposition. With the exception of the decomposition of 5f, all reactions were followed by monitoring the decrease in height of the pentamethylcyclopentadienyl resonance of starting complex relative to internal (17 5-CsHs)2Fe. The rate of decomposition of 5f was followed by monitoring the decrease in integrated intensity of the (Cp*-d1s)2Hf(CH2CsHs)H signal relative to (17 5 -CsHs)2Fe. NMR spectra were recorded at timed intervals using a Varian EM-390 (90 MHz) spectrometer. Reaction temperatures were maintained using constant temperature oil baths controlled by thermoregulators and were observed to be constant to within ± 1°C. A typical experiment involved 25 mgs of hafnium complex and 6-10 mg of ferrocene dissolved in 0.4 ml of benzene-ds. The NMR tubes were sealed containing >1 atm of dinitrogen. Peak heights were demonstrated to be reproducible to within ±7% by repeated measurement. Each spectrum was recorded 3 times and the average peak height was used to calculate the values of k given in Tables 2-4. Errors in k represent approximately 1 standard deviation. 149 Plots of ln(k{r) vs. 1{f were constructed from the decomposition rates and yielded the activation parameters LI.H i and LI.S i. References (1) See for example (a) Abis, L.; Sen, A.; and Halpern, J. J. Am. Chern. Soc., 1978, 100, 2915-2916. (b) Norton, J.R., Ace. Chern. Res. 1979, 12, 139. (b) Carter, W.J.; Okrasinski, S.J.; Norton, J.R. Organometallics, 1985, 4, 1376. (c) Buchanan, J.M.; Stryker, J.M.; and Bergman, R.G. J. Am. Chern. Soc. 1986, 108, 1537-1550. (2) McAlister, D. R., Erwin, D. K., Bercaw, J. E. J. Am. Chern. Soc., 1978, 100, 5966. (3) Roddick, D. M., Ph.D. Thesis, California Institute of Technology, 1984. (4) Lee, J.B.; Gajda, G.J.; Schaefer, W.P.; Howard, T.R.; lkariya, T.; Straus, D.A.; Grubbs, R.H. J. Am. Chern. Soc., 1981, 103, 7358-7361. (5) See chapter I of this thesis or: Bulls, A. R.; Schaefer, W. P.; Serfas, M.; Bercaw, J. E. Organometallics, 1987,6, 1219. (6) The addition of dihydrogen to a benzene-d5 solution of Cp*(ry 5,ry 1 -C5Me4 CH 2)Hfl immediately generates Cp*2Hf(H)I. See Chapter II of this thesis. (7) In a separate experiment, benzene-d6 solutions of 7 were shown to react with dihydrogen (ca. 4 atmospheres) to initially afford 6, which is subsequently hydrogenated to toluene and Cp*2HfH2. No Cp*2Hf(CH2C5H5)H (5) is detected by 1 H NMR. (8) (a) Bigeleisen, J. PureAppl. Chern. 1983,8 217-223. (b) Reference 1(c) and reference therein. (9) See chapter 4 of this thesis. (1 0) Gingerich, K. A. J. Chern. Phys. 1971, 54, 3720. (11) McMillen, D. F.; Golden, D. M. Ann. Rev. Phys. Chern. 1982, 33, 493. 150 12) Values for t.S o of +19 and +33 e.u. were found for the elimination of dihydrogen from Cp*('7 5-CsMe4CH2C5Hs)HfH2 and Cp*('7 5-CsMe4CH2CH2CHs)HfH2, respectively. See chapter 4. 13) The proposed Hf(ll) cyclopentadiene intermediate could also be considered a Hf(IV) metallacyclopropane derivative. (14) Davies, S.G.; Hibberd, J.; and Simpson, S.J., J. Organomet. Chem., 1983, 246, C16-C18. (15) Benfield, F.W.S.; and Green, M.L.H., J.C.S. Dalton 1974, 1324-1331. (16) (a) Jones, W.D.; and Maguire, J.A., Organometallics 1987, 6, 1301-1311. (b) Werner, H.; and Hofmann, W . Angew. Chem.lnt. Ed. Engl. 1977, 16, 794-795. (c) See reference 2. (17) Wochner, F.; Brintzinger, H. H. J. Organomet. Chem. 1986, 309, 65-75. (18) Burger, B. J.; Bercaw, J. E. in "New Developments in the Synthesis, Manipulation and Characterization of Organometallic Compounds", ACS Symposium Series, A. Wayda and M. Darensbourg, Eds., in press.