Some Mechanistic and Synthetic Aspects of the Interaction of Lewis Acids with Bis-Cyclopentadienyltitanium(IV) Alkyls and Bis-Cyclopentadienyltitanacyclobutanes Citation Ott, Kevin Curtis (1983) Some Mechanistic and Synthetic Aspects of the Interaction of Lewis Acids with Bis-Cyclopentadienyltitanium(IV) Alkyls and Bis-Cyclopentadienyltitanacyclobutanes. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/fxvt-ac87. https://resolver.caltech.edu/CaltechTHESIS:11012019-155844929 Abstract Titanocene dichloride has been shown to react cleanly with two equivalents of AlMe 3 to produce the Lewis acid stabilized titanium methylene [chemical formula; see abstract in scanned thesis for details]. Due to the interest in the utility of this complex in Wittig-type chemistry, in the metathesis of olefins, and in the synthesis of titanacyclobutanes, a study of the mechanism of formation of the complex was carried out. The proposed mechanism for formation of 1 involves an intramolecular proton abstraction by an Al-Me bond from a Ti-CH 3 group in the intermediate Cp 2 TiMeCl•AlMe 3 . The effect of Cp-ring substitution and halide substitution on the reaction rate along with a deuterium isotope effect of 3 and a large negative entropy of activation were consonant with the proposed mechanism. Specifically labelled titanacyclobutanes were prepared, and the cleavage reaction with AlMe 2 Cl to yield 1 as studied. It was found that cleavage occurred with a secondary deuterium isotope of 1.2 to 1.6 (depending upon the titanacyclobutane) and exhibited bimolecular kinetics. Surprisingly, the stereochemistry of the titanacyclobutanes was completely scrambled before cleavage to 1 and olefin. A mechanism for this isomerization was proposed to involve a rapid and reversible trans-metallation of a Ti-C bond with AlMe 2 Cl, producing a 3-aluminapropyltitanocene chloride. Following rapid inversion at the carbon adjacent to aluminum, the racemized 3-aluminapropyltitanocene chloride could cleave to 1 or close to yield isomerized titanacyclobutane. The β,β-disubstituted titanacyclobutanes proved to be good sources of the Cp 2 TiCH 2 unit, as indicated by their ready formation of bis-µ-CH 2 -bis-Cp 2 Ti ( 2 ). Some chemistry and photochemistry of 2 are reported. The Cp 2 TiCH 2 moiety is readily trapped with transiton metal or main group metal Lewis acids such as Cp^TiCl 3 or Me 3 SnCl to produce compounds such as [Cp 2 TiCl]-µ-CH 2 -[Cp*TiCl 2 ] ( 3 ) and Cp 2 Ti(CH 2 SnMe 3 )Cl. Cp 2 TiCH 2 may also be trapped as the methylene phosphine adduct Cp 2 Ti(=CH 2 )PEt 3 ( 4 ). The compounds 2 and 3 react with CO to yield insertion products which contain bridging ketene ligands which w ere characterized spectroscopically. Compound 4 also reacts with CO to produce the mononuclear ketene Cp 2 Ti(η 2 -CH 2 CO) in low yield. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Chemistry) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Bercaw, John E. Thesis Committee: Grubbs, Robert H. (chair) Dougherty, Dennis A. Evans, David A. Bercaw, John E. Defense Date: 3 June 1982 Record Number: CaltechTHESIS:11012019-155844929 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:11012019-155844929 DOI: 10.7907/fxvt-ac87 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 11882 Collection: CaltechTHESIS Deposited By: Mel Ray Deposited On: 01 Nov 2019 23:45 Last Modified: 07 Aug 2025 17:21 Thesis Files PDF - Final Version See Usage Policy. 52MB Repository Staff Only: item control page

Some Mechanistic and Synthetic Aspects of the Interaction of Lewis Acids with Bis-Cyclopentadienyltitanium(IV) Alkyls and Bis-Cyclopentadienyltitanacyclobutanes Thesis by Kevin Curtis Ott In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California 1982 (Submitted June 3, 1982) ii Titanocene dichloride has been shown to react cleanly with two equivalents of A1Me3 to produce the Lewis acid stabilized titanium f methylene Cp2 iCH2 AlMe 2 Cl (!). Due to the interest in the utility of this complex in Wittig-type chemistry, in the metathesis of olefins, and in the synthesis of titanacyclobutanes, a study of the mechanism of formation of the complex was carried out. The proposed mechanism - for formation of 1 involves an intramolecular proton abstraction by an Al-Me bond from a Ti-CH 3 group in the intermediate CP2TiMeCl•AlMe3 • The effect of Cp-ring substitution and halide substitution on the reaction rate along with a deuterium isotope effect of 3 and a large negative entropy of activation were consonant with the proposed mechanism. Specifically labelled titanacyclobutanes were prepared, and the cleavage reaction with A1Me 2 Cl to yield 1 was studied. It was found that cleavage occurred with a secondary deuterium isotope of 1. 2 to 1. 6 (depending upon the titanacyclobutane) and exhibited bimolecular kinetics. Surprisingly, the stereochemistry of the titanacyclobutanes was com- - pletely scrambled before cleavage to 1 and olefin. A mechanism for this isomerization was proposed to involve a rapid and reversible transmetallation of a Ti-C bond with A1Me 2 Cl, producing a 3-aluminapropyltitanocene chloride. Following rapid inversion at the carbon adjacent to aluminum, the racemized 3-aluminapropyltitanocene chloride could - cleave to 1 or close to yield isomerized titanacyclobutane. The {:J, {3-disubstituted titanacyclobutanes proved to be good sources of the Cp2 TiCH 2 unit, as indicated by their ready formation of bis-µCH 2 -bis-Cp2 Ti (~. Some chemistry and photochemistry of ~ are reported. iii The Cp2 TiCH2 moiety is readily trapped with transiton metal or main group metal Lewis acids such as Cp*TiCls or Me 3 SnCl to produce compounds such as [Cp2 TiCl]-µ-CH 2 -[Cp*TiC~ l (~ and Cp2 Ti(CH 2 SnMe 3 )Cl. Cp2 TiCH2 may also be trapped as the methylene phosphine adduct Cp2 Ti( =CH2 ) PEt3 (f). - - The compounds 2 and 3 react with CO to yield insertion products which contain bridging ketene ligands 'Which were characterized - spectroscopically. Compound 4 also reacts with CO to produce the mononuclear ketene Cp2 Ti(77 2 -CH 2 CO) in low yield. iv Throughout this Thesis, Cp and Cp* designate the ligands 77 5 -C 5H5 and 77 5 -C 5Me 5 , respectively. V ACKNOWLEDGMENTS I would like to thank Bob Grubbs for providing the opportunity to come to Caltech and do a little organometallic chemistry. I also thank all of the friendly people at Caltech, past and present, who made the transition to Caltech so simple. I owe special thanks to Steve Buchwald who spent many long hours helping to put this thesis into recognizable English, and to Henriette Wymar for her excellent typing. I would also like to thank my parents for their faith and support. And lastly, to Rosalie, who provided the light at the end of the tunnel. vi TABLE OF CONTENTS Page Chapter 1 1 An Investigation of the Reaction of Bis-Cyclopentadienyltitanium Dichloride with Trimethylaluminum. Mechanism of an a-Hydrogen Abstraction Reaction. Chapter 2 42 The Stereochemical Consequence of the Interaction of Dimethylaluminum Chloride with Bis-Cyclopentadienyltitanacyclobutanes. Mechanism of Titanacyclobutane Isomerization. Chapter 3 The Preparation and Characterization of a Few M-µ-CH 2 -M (M = Ti, M' = Ti, Zr) Compounds and Their Reactions with Carbon Monoxide. 79 CHAPTER 1 An Investigation of the Reaction of Bis-Cyclopentadienyltitanium Dichloride with Trimethylaluminum. Mechanism of an a-Hydrogen Abstraction Reaction. 2 INTRODUCTION ~ Since Ziegler's discovery 1 , 2 that ethylene could be polymerized at low temperature and pressure to a high molecular weight polymer using a catalyst formed from the reaction of an early transition metal halide such as TiC14 and aluminum alkyls, there has been a keen interest in the nature of the interaction of the aluminum alkyls with the transition metal center and its ligands. The insolubility of catalysts derived from TiC14 has made mechanistic studies difficult, and has led to a large effort to synthesize soluble polymerization catalysts. 3 One such soluble catalyst for the polymerization of ethylene is the system Cp2 TiCl:/AIR3-n Xn (n = O, 1, 2; R = alkyl, X =halide). 4 This catalyst system has become the best characterized (in terms of the kinetics of polymerization and subsequent side reactions) 3a of those studied. In the mid-1960's several groups studied the reaction of Cp2 TiC~ with A1Me3 and noticed that methane was evolved and that the organometallic product contained Ti-CH!!'"Al units. 5 Schrock's isolation of tantalum alkylidene compounds 6 and the possible role of such early transition metal alkylidenes in the catalysis of the olefin metathesis reaction stimulated Tebbe to reinvestigate the Cp2 TiC~/A1Me3 system. This led to Tebbe 's isolation and characterization of the Lewis acid-stabilized titanium methylene, Cp TiCH AlMe Cl 1, and other derivatives. 7 2 2 2 Tebbe's methylene compound demonstrated olefin metathesis activity, and reacted with internal acetylenes to produce the titanacyclo- 3 1 I butenes Cp2 TiCRCR'CH2 • 8 Tebbe also demonstrated that 1 was very reactive with organic carbonyl compounds, transferring a methylene unit to ketones 7 and esters to produce olefins and vinyl ethers, - respectively. Tebbe also showed that 7 reacted with propylene to form isobutylene. As part of a reinvestigation of this methanation reaction with unsymmetrical (mixed-ring) analogues of 7, 12 it was discovered - - that 7 reacted with mono-substituted olefins under certain conditions to cleanly produce titanacyclobutanes. lOa These titanacyclobutanes - derived from 7 have been the center of intense investigation in Grubbs' laboratory with respect to their properties as well defined olefin metathesis catalystslOb and their utilization in organic synthesis. 9 , 11 The - synthesis of 7 in large quantities for this research provided impetus for the work described herein. This chapter details an investigation into the mechanism of - formation of 7. This study was initiated to provide information regarding the interacti01;1 of aluminum alkyls with a transition metal center and its ligands and lend insight into the apparent a-H abstraction involved in formation o~ 1 from Cp2 TiC~ and A1Me3 • Such mechanistic information would prove valuable in extending the synthesis to the formation of higher alkylidene units. 4 RESULTS ~ Preparation of the "Mixed-Ring" Titanocenedichlorides The "mixed-ring" titanocenedichlorides (t e., those compounds having one substituted and one non-substituted cyclopentadienyl ligand) were prepared by reaction of CpTiC~ with the lithium salt of a substituted cyclopentadienide. The mixed-ring compounds prepared were Cp'CpTiC½ (Cp' = C 5H4 Me) !, (Cptms)CpTiC¼ (Cptms = C5H4 SiMe 3 ) ~' IndCpTiC½ (Ind = indenyl) ~' (Cp)CpTiC¼ ( (Cp) = 1, 2, 4-trimethylcyclopentadienyl) 1, [Cp] CpTiC¼ ([Cp] = 1, 3-diphenylcyclopentadienyl) §,_, and Cp*CpTiC¼ (Cp* = C 5Me 5) §_. The latter compound was prepared by reaction of Cp*TiC½ with CpLi. An attempt was made to prepare (1, 2, 4-triphenylcyclopentadienyl)CpTiC¼. This compound proved unstable to hydrolysis during work-up and could not be isolated in pure form. In many of these preparations, not only was the anticipated mixed-ring compound observed, but also CP2 TiC¼ and often Cp2 TiC~ (Cp = substituted Cp). For example, the preparation of Cp'CpTiC½ often resulted in a stati~tical mixture of Cp2 TiC½, Cp'CpTiC½, and Cp;TiC¼. This problem was partially circumvented by slow addition of an ether slurry of Cp'Li to a cold ether solution of CpTiC½. At best, the crude reaction mixtures contained at least 10% Cp2 TiC½ [ commercial (MeC H ) contains approximately 4% (C H ) _] 39 which 5 5 2 5 6 2 could be fractionally crystallized from Cp' CpTiC~. Typically, the synthesis of other mixed-ring compounds resulted in formation of less than 10% CP2TiC½. In one preparation of Cp*CpTiC½ from Cp*TiC½ and CpLi, the ratio of Cp*CpTiC¼ to cp:TiC¼ was approximately 2 : 1. 5 Fractional crystallization provided pure Cp*CpTiC¼• 1 H NMR spectral data for these mixed-ring compounds are listed in Table 1. Preparation of the Mixed-Ring Methylenes. Reaction of toluene solutions of the mixed-ring titanocene dichlorides with more than two equivalents of A 1Me3 resulted in the formation of analogues of 1 (eq. 1). The reactions proceeded in poor to moderate (C"p = a substituted cyclopentadienyl) yields for the mixed-ring compounds with the exception of IndCpTiC½. 1 H NMR spectroscopy of the reaction of IndCpTiC¼ with A1Me3 indicated extensive decomposition had occurred, although there were resonances in the ~ 7-8 region, indicative of formation of Ti-CH 2 -A I species. With the a i ' I exception of Cp'CpTiCH 2 AlMe2 Cl !!_ and [Cp] CpTiCH~lMe2 Cl ll_, the titanium methylene compounds were oils which crystallized only slowly, if at all. If allowed to stand undisturbed in a drybox, both (Cp)CpTiCH~IMe 2 Cl !Q and Cp*CpTiCH~IMei~112 g crystallized over a period of several weeks. 1 H NMR spectral data are listed in Table 2. Reaction of Cp2 TiClg, Cp2 TiMeC1, and CP2TiMe2 with A1Me3 and A1Me2 Cl. Reactions were followed by 1H NMR spectroscopy. Due to their lability, the aluminum alkyl adducts of the titanium species could not be isolated. 1 H NMR data for the reaction products described below appear in Table 30 One equivalent of A1Me3 reacts rapidly with Cp2 TiC¼ to form Cp2 TiMeCl • A 1Me3 _n C1n as was indicated by the appearance of a new 6 TABLE 1. 1 H NMR Spectral Data for Mixed-Ring Titanocenedichlorides Compound Solvent Cp'CpTiC½ (1) CDCla Chemical Shift (l5) * 6.54 s (CsHs) 6.33 m (C 5!!4Me) 2033 s (C 5H4 Me) 6.54 s (CsHs) J = 2Hz (C 5!!4 TMS) 0.32 s (C 5H4 TM~ 6.21 s (CsHs) 6.78 6.57 IndCpTiC½ (ID (C 6 ring, Ind) 7.36-7.68 m (Cp)CpTiC½ (1) [CpJCpTiC½ (~) 6.67 d J =3. 5 6.88 t 6.43 s (CsHs) 6.26 s (C 5)bMe2Me) 2.12 s (C 5H2Me 2Me) 2.07 s (C 5H2 Me2Me) 6.33 s (CsHs) " 7.40-7.71 m Cp*CpTiC½ (~) (C 5 ring, Ind) (C 5H3 Ph 2) 6.96 d J = lHz (C5!!2H Ph 2 ) 6.25 s (CsHs) 2.04 s (C 5Me 5) *Chemical shifts relative to residual protio-solvent resonances. CHC!s at l57. 24, CHDC½ at l55. 32. 7 TABLE 2. NMR Spectral Data for Titanocene-Methylene Compounds in C6 D6 • Chemical Shift (r;) * Compound lH: 8.28 s (CH 2) 5.67 s (CsHs) -0.28 br.s (A1Me2) 1aC: 188 (CH 2) 114 (CsHs) lH: 8.12 ABq J = 6.9Hz (CH 2) 5.78,5.39 m (C 5!!4 Me) 5.66 s (CsHs) 1.66 s (C 5H4 Me) -0 .. 32, -0. 36 2s,br (A1Me2 ) 8.47,8.24 2d, J=6.5Hz (CH2) 6.17 m (C 5!!4 TM&) 5.72 s (CsHs) o. 01 s (C 5H4 TMS) -0.28,-0.32 2s,br 1 H: 7. 99, 7. 54 5.47,5.13 5.72 (A1Me2) 2d, J =6. 8Hz (CH2) 2d, J =O. 6Hz (1, 2, 4-C 5!!2 Me3) s 1.76,1.63,1.46 3s (CsHs) (1, 2, 4-C 5H2Me3) 1sc: 180.6 (CH2) 112.9 (CsHs) 16.3,14.2 (C 5Me3 H2 ) 8 TABLE 2 {continued) Compound Chemical Shift ( ~) * 7.2-6.8 m (1, 3-C 5H3 Ph 2) 6.4-6.1 m (1, 3-C 5H3 Ph2 ) 5.62 s (CsHs) 8.23 ABq J=O. 5Hz (CH2 ) . 27 s,br lH: 7.57,6.94 (A1Me 2) 2d, J = 7. 3Hz (CH 2 ) 5.66 s (CsHs) 1.59 s (C 5Me 5 ) 2s,br. (A1Me 2) (J=125Hz) (CH 2 ) -0.19,-0. 36 1sc: 173.9 113.6 (CsHs) 121.8 (QsMes) 12.6 (CsMes) *Shifts relative to C6 DsH at o7 .15. 13 C shifts relative to f 6 D6 at o128. 0 9 TABLE 3 1 H NMR Spectral Parameters for Methyltitanium and Aluminum Mixtures Cp Resonance ( o) Ti•Me Resonance (o)* Cp2 TiMeCl 5.81 0.82 Cp2 TiMe2 5.67 0o06 Cp2 TiC¼ + 1A1Me3 5.79 0.94 Cp2 TiMeCl + A1Me 2 Cl 5.79 0.94 CP2TiMeCI + 2AIMe 3 5.73 0.90 Cp2 TiC¼ + 2A1Me 3 5.79 0.92 10 Cp resonance at o5. 79 (lOH) and a methyl resonance at oO. 94 (3H). An identical spectrum was obtained when one equivalent of A1Me 2 Cl was allowed to react with C1>2TiMeCl. Heating these solutions to 56 ec for 19h resulted in the formation of a 30% yield of a 6: 1 mixture of CP2TiCH2 AlMeClCl and 1· Further heating resulted in formation of a homogeneous blue-green solution which exhibited a broad 1 H NMR spectrum. Reaction of 1. 21 equivalents A1Me 3 with Cp2 TiMeCl resulted in formation of a mixture of 78% CpTiMeCl • AlM:e3 _nC1n (•Al designates AlMe3 _nCln) and 22% CpTiMe 2 • Similarly, when 1.19 equivalents of A1Me 2 Cl was allowed to react with CP2 TiMe 2 , a mixture of 87% CP2 TiMeCl • Al and 13% Cp2 TiMe 2 was observed, the ratio [Cp2 TiMe 2 ][A1Me2 Cl] /[Cp2 TiMeCl][AlMe 3 ] was O. 06 for both solutions. 37 - When these solutions were heated to 65 ° C for 6h, an 80% yield of 7 was formed by way of eq. 2 and 3. (3) Cp2 TiMeCl . A1Me 3 ~ 7 + CH 4 Reaction of Cp2 TiC½ with greater than two equivalents of A1Me3 resulted in formation of Cp2 TiMeCl • Al. Heating these solutions to 65 e - resulted in the formation of 7 according to eq. 1 at the same rate at which the reaction of CP2TiMe2 with A1Me 2 Cl proceeds. 11 Scheme 1 PhCO2 Me _ _ ___,_.) Ph(OMe)CCH2 + Ph(OMe)CCD2 PY 75% 25% ,-..J ,-..J 12 The reaction of C°P2 TiC ¼ with one equivalent A1Me3 followed by one equivalent of A1Me3 -dg gave rise to a 1 H NMR spectrum which indicated that Cp2 Ti(CH 3 )Cl ·Aland Cp2 Ti(C.D3 )Cl • Al were present in a 1 : 1 ratio (Scheme 1). This solution was heated to 56" C and the 1 extent of reaction was monitored by H NMR spectroscopy. As the reaction progressed '1 was preferentially formed relative to 'l.. ~ 1n· a ratio of 2. 9: 1. The sample was quenched with methylbenzoate and --- ---~yridine 9 producing 1-methoxy-1-phenylethylene-<lo and-~ in a ratio of 2. 9 ± 0.1 : 1, indicating the reaction proceeds with a moderate primary deuterium isotope effect. Close examination of the vinylic AB quartet indicated that no 1-OMe-1-Ph=CHD is formed and demonstrated that only Ti-CH 2 (CD2 )-Al groups and no Ti-CHD-Al groups were present upon reaction with methylbenzoate in the presence of pyridine. Examination of the Kinetics of Titanium-Methylene Formation. Kinetics data for the reaction of Cp2 TiC¼ with greater than 2 equi- • • valents of A1Me3 in C6 D6 at 53"C were obtained using varying concentrations of CP2TiC¼ and A1Me 3 • The reaction exhibited second-order kinetics and fit the rate law (eq. 4): Second-order plots were linear over more than two half-lives. 13 Selected kinetics data are listed in Table 4. An increase of 40-50% in kobs was observed when C~C½ was employed as solvent. Activation parameters were derived from kinetics data obtained at 38, 57, and 72" C and are listed in Table 5. 13 TABLE 4 Kinetics Data for the Reaction Cp2 TiC~ + 4A1Me 3 53 •c • 'l + AlMe 2 Cl+CH 4 [TiMeCl• Al]t=O [AlMe3]t=O 0.2 M 1.4 M 0.1 0.7 1. 8 " 002 0.7 2.3 ±0.3 " 0.2 0.6 14 TABLE 5 Activation Parameters for the Reaction T "C kobs 38 7. 7 X 10- 5 57 3o 3 X 10- 4 72 1.1 X 10- 3 AH* = 16 ± o. 2 kcal• moi-1, as* = -26 ± 1 e. u. aG;30 = 25 ± 1 kcal. mol- 1 15 Kinetics data were also obtained for certain mixed-ring compounds 14 and are listed in Table 6. The data for Cp TiC¼, 2 Cp~TiC¼, and CJl2TiBr 2 are included for comparison. An attempt was made to examine the reaction of Cp2 TiF2 with greater than two equivalents of AIMea, however, this reaction proceeds rapidly to produce Cp2 TiMe 2 which is then converted to at least 40% Cp2 TiCH 2 AlMe 2 F in lh at 65 ° C. Further heating converts this material into Cp2 TiCHzAlFMeF and other unidentified products. Monitoring of the 1 reaction by H NMR spectroscopy proved unsuitable for kinetics analysis, as the C5H5 resonance corresponding to the product(s) were fairly broad and overlapped the C5 H5 resonance of Cp2 TiMe2 • 16 TABLE 6 Observed Rates of the Reaction 53°C I I Cp2 TiC~ + 4A1Me 3 C D > C~TiCH2 AlMe 2 Cl + A1Me 2 Cl + CH 4 6 Compound 6 1 1 5 kobs x 10 M- s- Cp2 TiC12 23 Cptms CpTiC~ 8 Cp~TiC~ 7 [Cp] CpTiC~ 12 (Cp) CpTiC~ 3 Cp*CpTiC~ 1 C~TiBr2 17 (measured at 57. 5" C) 17 DISCUSSION ~ Synthesis of the Titanocenedichlorideso As mentioned above, synthesis of mixed-ring titanocene dichlorides often resulted in production of Cp2 TiC½ and Cp2 TiC¼ (Cp = substituted Cp) as significant by-productso Dormond et aL, 15 reported that ring exchange occurs when Cp2 TiC½ is reacted with one or two equivalents of LiCp; hydrolysis of the reaction mixtures with aqueous HCl gave good yields of CpCpTiC¼ and Cp 2 TiC~, respectively. It was suggested that the sterically smaller and less electron-donating Cp ligand is selectively hydrolyzed from the metal center (Scheme 2). They do not mention the observation of any Cp2 TiC½ in the reaction mixture and Scheme 2 is unable to account for the formation of Cp2 TiC~ which was observed in the work reported here. An alternative explanation for the ring exchange via a bridging Cp intermediate is outlined in Scheme 3. Such a ring exchange may occur in the reaction of Cp2 TiC~ and Tic1/ 6 (equation 5)° but must have a higher activation energy than (5) xylene . fl > 2CpT1Cls re ux for the reaction of CpTiCls with CpLi which undergoes apparent ring exchange at or below room temperature. 40 The possibility also exists that the ring exchange proceeds through a Cp bridging a titanium and a Li Cp as depicted in Scheme 4. The ability of lithium alkyls to form aggregates is well known. 17a Tuli et al., 18 report that the reaction of CpTiCla and Cp'Tl proceeds to yield Cp' CpTiC~ in greater than 85% yield. Thallium 18 Scheme 2 - - Cp2 TiC¼ + CpLi ~ Cp2 CpTiC1 + LiCl aq.HCl _ _ --~> CpCpTiC¼ (80%) + Cp2 TiC¼ (20%) 80% yield 85% yieldo Dormand, et al. Scheme 3 CpLi CpTiC1s > CpCpTiC¼ CpLi - ➔ CpCp2 TiC1 CpTiC1s _ /Cp\. _ CpCp2 TiCl - - - . . . Cp2 C1Ti TiCpC¼ --► Cp2 TiCl + Cp2 TiC¼ 'Cl/ Scheme 4 CpTiC1s CpLi - - - ~ CP2TiC¼ 19 alkyls may be less prone to aggregation, 17b and therefore are less likely to undergo ring exchange similar to that proposed in Scheme 4. How ever, CpTl is also not as reactive as the lithium analogs, and may - - react more slowly with _CpCpTiC~ to yield the species CpCp2 TiCl. Reaction of CPaTiCia, Cp2 Ti MeCl, and Cp2TiMe 2 with Aluminum Alkyls. Reaction of Cp2 TiC¼ with one equivalent of A1Me 3 proceeds rapidly to yield a solution having the stoichiometry Cp2 TiMeCl. A1Me 2 Cl. The aluminum is assumed to complex to the chlorine on titanium. 19 This complexation is weak, as the aluminum may be removed to yield Cp2 TiMeCl by pumping on the solutions or by complexation with an ether. Addition of one equivalent of A1Me 3 to "Cp2 TiMeCl • A1Me 2 Cl" 1 produces only slight changes in the H NMR spectrum, and Cp2 TiCH2 AlMe 2 Cl 1 is slowly formed along with methane and A1Me 2 Cl. The results of these reactions are summarized in Scheme 5, which indicates that one A1Me 3 methylates Cp2 TiC¼. If no additional A 1Me3 is added to this mixture, only small amounts of titanium methylene compounds are slowly for med, and after long reaction times there is apparent formation of a Ti(III) species. 22 When one additional equivalent of A1Me 3 is added, it is proposed that A1Me3 -A1Me 2 Cl exchange occurs producing an equilibrium concentration of Cp2 TiMeC1 • A1Me 3 , which then in a slow step extrudes CH 4 forming 1, requiring the observed second-order kinetics. - 7 is also produced from the reaction of either Cp2 TiMe 2 with A1Me 2 Cl, or Cp2 TiMeCl with 20 Scheme 5 slow CP2TiC¼ + A1Me3 ;= Cp2 TiMeCl • A1Me 2 Cl _ __.-, reduction to Ti(III) AIMe2c11 lAIMes , , slow Cp2 TiCH 2 AlMe 2 Cl ( Cp2 TiMeCl • A1Me 3 7 21 A1Me3 • These two reactions proceed at the same rate as Cp2 TiC¼ with 2A1Me3 • This suggests that the necessary stoichiometry for formation of 1 is Cp2TiMeCl. A1Me 3 , and that all of the above species are in equilibrium. Throughout this chapter, the Ti-Al species have been described as 1 : 1 adducts, primarily for simplicity of presentation. It should be noted, however, that due to the tendency of aluminum methyl and chloroaluminummethyl compounds to form dimers, the formation of 1 : 2 Ti-Al adducts cannot be ruled out. The observation of a deuterium isotope effect kH/kD = 2. 9 and a large, negative activation entropy of approximately -26 e. u. are con8-istent with the cyclic transition state complexes depicted in Figure 1. Polarization of the Ti-Cl bond by aluminum as in the • - resonance structure B would be expected to increase the basicity of the Al-CH 3 bonds, and concomitantly increase the acidity of the TiCH 3 a-hydrogens. 44 Extrusion of CH would then lead to the observed 4 - product 7. The 50% increase in kobs when the solvent is changed from C6 D6 to C ~C¼, a solvent of higher dielectric constant, 23 is also consistent with an increase in dipole moment in the transition state. Based on "hard-soft" considerations, 24 an aluminum center should polarize a Ti-Br bond less than a Ti-Cl bond. If resonance - structure Bin Figure 1 is important, then reaction of Cp2 TiBr2 with two equivalents of A1Me3 should proceed relatively slowly to I 1 Cp2 TiCH 2 AIMe 2 Br, and in fact, under identical conditions, Cp2 TiBr2 1 I is converted to Cp2 TiCH 2 AlMe 2 Br at 35-40% of the rate at which 1 is formed. Due to experimental difficulties, the rate at which Cp2 TiF2 22 V V :r: u C\J + Q) Q) ~ ~ <1 <! I u ~ C\J I u ::;.. u \ I ~-H ~(\J Q. t-C\J 0.. l l u u -++- -++- f() u, . CJ') LL r<) :r: I \ + I ~\:r u u C\J u, ' \ '<l:' ' ' 'I 1 u I I C\J I u \ ' \ '' <! I <l: C ::,. I I I u \_ I \ ,'l t-C\J 0.. t-C\J 0.. u u I C\J I u 'I 23 reacts with A1Me3 to produce the corresponding titanium methylene compound could not be determined accurately. That the aluminum halide bond polarization in these reactions is important is indicated by the very rapid formation of Cp2 TiMe2 and A1Me 2 F from the reaction of Cp2 TiF2 with four equivalents of A1Me3 • Apparently, the position of equilibrium of equation 8 lies in the center for X = Cl, Br but far to the right for X = F. The subsequent reaction of Cp2 TiMe 2 with A1Me 2 F proceeds to a titanium methylene complex at at least the same rate as observed for the Cp2 TiC~ reaction. In an attempt to investigate the effect of larger and more electrondonating or accepting ligands on the rate of the titanium methylene formation reaction, the mixed-ring compounds were synthesized and the kinetics of their reaction with A1Me 3 measured. These data are tabulated in Table 6 and appear in approximate order of increasing size of the substituted cyclopentadienyl ligand. With the exception of [Cp], kobs varies inversely with the steric bulk and also with the relative donor ability 25 of the substituted cyclopentadienyl ligand. [Cp] can be considered to be a relatively electron-withdrawing ligand in comparison to the other ligands. Thus, there appear to be two trends demonstrated in the rate data for the mixed ring compounds. One is that a dramatic decrease in kobs is seen with ·increasing steric bulk of the ligand, and the second being a decrease in kobs- with increasing donor strength of the ligand. A better donor ligand would - tend to decrease the Lewis acidity of the titanium center in structure B 24 (Fig. 1) and concomitantly decrease the acidity of the Ti-Me hydrogens. One might speculate that IndCpTiC¼ would react with excess A1Me3 to generate the analogous titanium methylene at a rate approximately equal to that of Cp2 TiC¼. Unfortunately,. IndCpTiC~ does not react cleanly with A1Me 3 • This is perhaps related to the relative ease with which the 775 -indenyl ligand can "slip" to 773 -indenyl. In summary, it is proposed that the reaction of Cp2 TiC¼ with two equivalents of A 1Me3 proceeds to an intermediate Cp2 TiMeCl. Al - which goes on to produce 7 via an intramolecular attack of a strong base (Al-Me) on the activated 0t -hydrogen of the titanium methyl group. This mechanistic course is similar to that observed in the intermolecular proton abstraction from Cp2 TaMe; with a strong bas~ to yield Cp TaCH (Me). 26 The latter proceeds with a deuterium isotope 2 2 effect of 3, which is similar in magnitude to that reported here. Other reactions which proceed via similar 6-center mechanisms are the "ene" reaction, ~2 and the protonolysis of trialkylborons with carboxylic acids. 43 Both of the~e reactions proceed with deuterium isotope effects of approximately 3. The "ene" reaction exhibits activation entropies of -30 to -40 e. u., similar to that reported here (-26 e. u. ). 25 EXPERIMENTAL ~ General Considerations. All manipulations of air and/or moisture sensitive compounds were performed using glove box and standard high vacuum or Schlenk line techniques. Argon used in Schlenk work was purified by passage through columns of BASF RS-11 (Chemalog) and Linde 4A. molecular sieves. Diethyl ether, C6 D6 (Merck, Sharp and Dohme) C6H6 , toluene, tetrahydrofuran, and hexane were dried and deoxygenated by stirring over CaH 2 , degassed, and were vacuum transferred onto purple sodium-benzophenone ketyl. Solvents for Schlenk work were vacuum transferred from sodium-benzophenone ketyl solutions into dry vessels equipped with teflon needle valve closures and stored under Ar. CD2 C½ was dried over CaH 2 and deoxygenated by several freeze-pump-thaw cycles. Solvents used for recrystallization of the titanocene dichlorides were used as received. Trimethylaluminum (Alfa), dimethylaluminum chloride (Texas Alkyls), and AlMe 3 (2M 'in toluene, Aldrich) were used as received. Cp2 TiC½ (strem or Boulder Scientific) was purified by Soxhlet extraction with dichloromethane. This material was then recrystallized from toluene for the kinetics studies. TiC14 (Alfa) was used as received. Methylcyclopentadiene and cyclopentadiene dimers, and indene were purchased from Aldrich. AlMe -dg, 27 Cp TiMeCl, 28 Cp TiMe , 28 3 2 2 2 C1>2 TiBr , 29 HCptms, 3o HCp*, 31 H(Cp), 32 H [Cp J, 33 and 1, 2, 4-tri2 phenylcyclopentadiene 34 were prepared by literature methods. Lithium-t-butoxide was prepared from the reaction of t-BuOH with n- BuLi followed by purification by sublimation. 26 Lithiumcyclopentadienides (except Li[ Cp] and Li(l, 2, 4-triphenylcyclopentadienide)) were prepared by dissolving the cyclopentadiene in degassed n-hexane followed by slow addition of 1.1 equivalent of 1. 6 M n-BuLi/hexane (Aldrich), Li[ Cp] was prepared by reaction of a benzene solution of H[Cp] with 1.1 equivalents of n-BuLi/hexane. Li(l, 2, 4-triphenylcyclopentadienide) was prepared similarly using toluene as solvent. The lithium cyclopentadienide suspensions thus prepared were transferred under Ar onto a medium frit via cannul~, • washed repeatedly with n-hexane, and dried under high vacuum. CpTiC¼ 16 and Cp*TiC¼ 35 were prepared by literature methods and purified by sublimation. 1 H NMR spectra were recorded using a Varian EM-390 or a JEOL FX-90Q. Probe temperatures were calculated by measuring 36 Chemical shifts are with respect to residual protioavMeOH. solvent resonances. Preparation of Cp'_C pTiC~ (1). Several syntheses of this compound appear in the literature. 15 , 16 b Due to the irreproducibility of these preparations, the following preparation is included here. A suspension of Cp'Li (0. 86 g, 10 mmole) in ether (20 mL) was slowly added dropwise via cannula to a stirred solution of CpTiC¼ (2.193 g, 10 mmole) in 30 mL ether at 0° C under Ar. The bright yellow solution slowly deposited a red microcrystalline powder. After the addition was complete, the mixture was allowed to warm to room temperature. stirring was continued for 30 min before approximately half of the ether was removed in vacuo, and 10 mL of 9M aqueous HCl was added. The red solids were filtered and washed with 9M aqueous 27 HCl until the washings were colorless. The solids were washed several times with cold ether and dried to yield 2.1-2. 4 g (80-90%) of a mixture of Cp'CpTiC¼ which typically contained 10-20% of both Cp2 TiC¼ and Cp~TiC¼. Fractional crystallization of this mixture from hot toluene yielded 60-70% Cp'CpTiC¼ as red needles which contained approximately 5-10% Cp2 TiC¼. Performing the reaction at room temperature or with a 50% excess of Cp' Li results in near statistical mixtures of Cp2 TiC¼, Cp'CpTiC¼, and Cp'2 TiC¼. 1 H NMR, !, CDCls: o6. 54, s (Cp), 6. 73, m (C 5!!4 Me), o2. 33, s (C 5H4Me). Preparation of cp;TiC¼. A slurry of LiCp' (8. 7 g, 101 mmole) in ether was slowly added to an ether solution of TiC14 (5. 5 mL, 50 mmole) under Ar at 0"C. The solution was stirred for 30 min and - worked up as for 2. The resulting solids were recrystallized from hot toluene to yield Cp~TiC¼ (8. 3 g, 60%). This material contained ca. 7% Cp2 TiC¼ (and probably ~ larger amount of !) . A small amount of this material was recrystallized from toluene, and then chloroform to yie Id a material which contained no Cp2 TiC¼; m. p. = 215-219% (decomp). Lit. 39 217-218" C (decomp). This material probably contained a small - amount of 1. 1 H NMR, Cp~TiC¼, CDCls: c5 6. 30, m (C 5H4 Me), 2. 27, s (C 5H4 Me). Preparation of CptmsCpTiC12 (~). A solution of CptmsLi (1. 24 g, 8. 6 mmole) in 10 mL THF was slowly added via cannula to a solution of CpTiCls (1. 89 g, 8. 6 mmole) in 10 mL THF under an Ar atmosphere at room temperature. After the addition was complete, the red solution was stirred for 30 m. The solvent was removed in vacuo to leave a 28 pinkish-orange powder. The solids were scraped onto a frit, and washed with 6 M aqueous HCl until the washings were colorless. The solids were then washed with 2x 5 mL of ethanol and 2x 5 mL of ether. The red powder was recrystallized from hot CHC:Is/hexane to yield 2. 1 g (76%) of deep red needles. No Cp2 TiC¼ was observed by 1 H NMR. 1 H NMR, ~' CDC:Is: o6. 54, s (Cp), 6. 78, 6. 57, 2t, J = 2Hz (C ~TMS), 0.32, s (C H TMS). 5 5 4 Preparation of IndCpTiC¼ (~. This compound was prepared as - for 2. Yields were typically in the range of 40-50% of a red-brown powder which contained ca. 15% Cp2 TiC¼. This impurity could be removed almost completely (ca. 3% remained) by washing with ether. 1 H NMR, ~' CDCla: 06.21, s (Cp), 7.68-7.36, m (C 6 indenyl ring), 6. 88, t, J = 3. 5 Hz, 6. 67, d, J = 3. 5 Hz (C 5 indenyl ring). Preparation of ( Cp) CpTiC¼ (4). This preparation is a modification of de Boer's method. 12 (Cp) Li (2. 28 g, 20 mmole) and CpTiCla (4. 38 g, 20 mmole) wer~ allowed to react under the same conditions - described for 2 (( Cp) Li is insoluble in THF, however), and worked up as for~- (Cp) CpTiC¼ can be recrystallized from hot CHCla or toluene to give scarlet plates. Yield: 2. 72-3. 25 g (48-56%). Typically, 5% or less of Cp2 TiC¼ is observed in the crude reaction mixture under these conditions. 1 H NMR, 1_, CDCla: o 6. 43, s (Cp), 6. 26, s (C 5~Me3 ), 2.12, s tc H MeMe 5 2 ) , 2 2. 07, s (C H Me~e 5 2 ) . 2 Preparation of [Cp] CpTiC¼ (5). A solution of [Cp] Li (0. 61 g, 2. 7 mmole) in 10 mL THF ( [Cp] Li solutions appear to transmit red and fluoresce blue) was slowly added via cannula to a solution of 29 CpTiClg (592 mg, 2. 7 mmole) in 10 mL THF under Ar at room temperature. The purple-red solution was stirred for 30 min and the solvent was removed in vacuo. The reaction mixture was extracted with a 1 : 1 benzene/dichloromethane mixture. Evaporation of the solvent yielded a dark purple-red microcyrstalline powder. Extensive washing of this - powder with EtOH and ether yielded 215 mg (20%) of 5. 1 H NMR, C~C~: ~6.33, s(Cp), 7.71-7.40, m(C 5H3 Ph 2), 6.96, d, J=l Hz (C 5H~Ph 2). Preparation of Cp*CpTiC~ (6). This preparation is a modification of de Boer's method. 12 , 41 A suspension of CpLi (0. 72 g, 10 mmole) in 10 mL THF was added slowly via cannula to a stirred solution of Cp*TiCla (2. 9 g, 10 mmole) in 10 mL THF under an Ar atmosphere at ambient temperature. The reaction mixture was stirred for 30 m and - worked up as for 2. Examination of the crude reaction product indicated less than 5% CptTiC~ was present (1H NMR). Recrystallization of the - red powder from hot toluene gave 1. 47 g (46%) of pure 6 as deep red columnar crystals. A small amount was recrystallized from CHClg for kinetics experiments, m.p. 184-185• C decomp). Lit. 41 185-186° C. 1 H NMR, CDClg: ~ 6. 25, s (Cp), 2. 04, s (C 5¥e 5 ). Attempted preparation of (1, 2, 4-Triphenylcyclopentadienyl) CpTiC¾. This reaction was performed similarly to the preparation of - 2, except ether was used in place of THF. In this manner Li(l, 2, 4- triphenylcyclopentadienide) (ca. 1. 47 mmole) was allowed to react with CpTiClg (307 mg, 1. 4 mmole). A purple-red solid precipitated,_and after 30 m of stirring, 'the solvent was removed in vacuo, and the reaction mixture was hydrolyzed with dilute aqueous HCl. An immediate 30 reaction occurred to give a green solution and yellow solids. The yellow s_olids were identified as 1, 2, 4-triphenylcyclopentadiene by TLC. No further attempts were made to prepare this compound. Preparation of Cp2 TiF2 • This is a procedure modified from the preparation of Cp ZrF • 29 Cp TiC~ (2. 5 g, 10 mmole) and Li(O-tBu) 2 2 2 (1. 61 g, 20 mmole) were dissolved in 20 mL THF and allowed to stir for 24 hr under Ar at room temperature to give a yellow-orange solution. The solvent was removed in vacuo . and the resulting oil was extracted with petroleum ether. The pet ether extracts were filtered under Ar through a fine frit. The petroleum ether was removed in vacuo, leaving a yellow-orange liquid. 1 H NM:R spectroscopy of this liquid dissolved in C6 D6 showed it to be >90% Cp2 Ti(O-tBu) 2 (~ 5. 98, s, l0H; 6 1.16, s, 21H) o The liquid was dissolved in 20 mL ether, and BF3 • Et 20 (2. 7 ml, 22 mmole) was slowly added by syringe. A pale orange powder precipitated rapidly. The powder was washed exhaustively with ether and dried in vacuo. The product proved to be too insoluble for NMR analysis in CDC lg and C6 D6 • Addition of one drop of D20 to these suspensions in NM:R tubes resulted in rapid dissolution of the orange suspension to give a yellow solution of Cp2 TiF2 as verified by 1 H NM:R spectroscopy. The rest of the orange powder (probably Cp2 TiF(BF4 ) or Cp2 Ti(BF4 ) 2 ) was treated with 15 mL MeOH to yield a yellow powder which was washed once with 2 mL MeOH followed by washing with ether. The yellow solid was dried in vacuo, and sublimed at 120° Cat < 1 µm Hg onto a water cooled probe. Yield: 1. 2 g (56%) of fluffy, bright yellow powder. A small portion was recrystallized from hot toluene to yield fluffy yellow needles. 31 1 H NMR: 6 6.41, t, JFH = 1.71 Hz. A small doublet (ca. 5% of total signal) appears at 6 6. 49, J = 1. 71 Hz. Preparation of Cp'CpTiCH 2 AlMe 2 Cl (8). This compound was prepared using the method of Tebbe 7 for 1· A1Me3 (9 mL 2M A1Me3 in toluene, 18 mmole) was added via syringe to an Ar flushed flask containing!. (2. 34 g, 8. 9 mmole). The resulting deep red solution was stirred for 40 h at room temperature. The solvent and the byproduct A.1Me2 Cl were removed in vacuo to yield a deep red oil. The oil was dissolved in 10 mL toluene, and more A.1Me3 was added (1 mL of 2M toluene solution, 2 mmole). This solution was stirred for ca. 12 h. Removal of the solvent in vacuo resulted in an oil which slowly solidified to a waxy mass. This mass was extracted with 40-50 mL of n-hexane. The combined extracts were concentrated and cooled to -50" C to obtain - 8 as a powder. The powder was washed once with 2-3 mL of hexane and pumped dry under high vacuum. Yield 530 mg (20%). This product - contained ca. 5-10% 7 due to the presence of a small amount of • Cp2 TiC~ in!.• Preparation of (Cp) CpTiCH 2AlMe2Cl (!Q). The preparation and isolation of this compound as a crystalline solid was first accomplished by de Boer 12 and it is this method which was reproduced and is reported here. A 1Me3 ( 4. 6 mL of 2M toluene solution, 9. 2 mmole) was added - via syringe into an Ar flushed flask containing 4 (1. 34 g, 4. 6 mmole). - The deep red solution was stirred for 28 h at room temperature. The volatiles were removed to yield a sticky solid which consisted of a 60: 40 mixture of .,.-.,... 10 and (Cp) CpTiMeCl as determined by 1 H NMR spectroscopy. A1Me 3 (1. 84 mmole) and 4 mL toluene were added to the 32 reaction flask, and the mixture was stirred for 24 h at 45 e C. Several such cycles were repeated until no (Cp) CpTiMeCl was observed. After this extended reaction time, the resulting oil contained a significant amount of toluene-insoluble red solids. 13 The insoluble solids were filtered, and the volatiles removed from the supernatant giving a deep red oil. The oil was triturated with hexane yielding a deep red mass of solids. The solids were extracted with hot hexane. The solvent was removed from the extracts yielding an oil which slowly crystallized to yield 890 mg of !Q which was ca. 85% pure by 1 H NMR spectroscopy. 500 mg of this material was dissolved in 50 mL hot hexane and filtered. The supernatant was concentrated to 15 mL and cooled to -50° C to -- yield 110 mg of pure 10 as a pale red powder. Reactions of Cp2 TiC~, Cp2 TiMeCl, Cp2 TiMe 2 , and Cp2 TiF2 with Alkylaluminums. Reactions were carried out in 5 mm NMR tubes capped with plastic caps or with latex septa (Wilmad Glass). All reactions were performed utilizing a total Ti concentration of 0. 2 M in C6 D6 solvent. Aluminum reagents were introduced via microsyringe. 1 The compositions of reaction mixtures were measured by H NMR spectroscopy and are considered accurate to within 5%. Reaction of Cp2 TiMeCl with 1 equivalent A1Me 2 Cl. A NMR tube was loaded with Cp2 TiMeCl (23 mg, 0.1 mmole), 10 µL A1Me 2 Cl (0.11 mmole), and 500 µL C6 D6 • The 1 H NMR spectrum was recorded (o 5. 79 (10 H), o 0. 94 (3 H), o -0.14 (6. 6 H)) and the sample was then placed in a bath maintained at 56 ° C. After 1 h, a trace of I i Cp2 TiCH 2 AlMeC1Cl was detected (o 7. 62, TiC~-Al). After 19 hat 56e C, a 30% yield of 1. and CpTiCH 2 AIMeC1Cl in a ratio of 1: 5 was 33 observed. The spectral lines were broad. After 30 h, the sample solution had turned blue-green, and no sharp features remained in the 1 H NMR spectrum. Identical results were obtained when C~ TiC½ (25 mg, 0.1 mmole} was allowed to react with A1Me 3 (10 µL, 0.1 mmole). Reaction of Cp2TiMe 2 with A1Me 2 Cl. A NMR tube was loaded with Cp2 TiMe 2 (20 mg, 1 mmole}, A1Me 2 Cl (ca. 1 mmole}, and 400 µL C6 D6 • The sample rapidly turned deep red upon contact of A1Me2 Cl with the yellow CP2 TiMe 2 solution. The 1 H NMR spectrum indicated the presence of Cp2 TiMeCl • Al (o 5. 73, Cp; o 0. 90, CH 3 } and Cp2 TiMe2 (o 5.69, Cp; o 0.05, CH 3 } in the ratio of 87 :13. A broad peak was also observed at a -. 23 (AlMe3 _nC1n). Integration of the methyl resonances showed the relative amounts of methyl containing species to be 2. 60 (Cp2 TiMeCl. Al}, 0. 77 (Cp2 TiMe 2), and 9. 72 (AlMe 3 _nC1n). From these integrals, the concentrations of the titanium species were calculated to be [Cp2 TiMeCl] = 0.17 Mand [Cp2 TiMe2 ] = 0. 03 M. For simplicity, the reaction under consideration was assumed to be (equation 9) and the concentrations of A1Me3 and A1Me 2 Cl were calculated to be 0. 87 M and 0. 32 M, respectively. This yields an "equilibrium ratio" of o. 06. Similarly, when Cp2 TiMeCl (22 mg, 0.1 mmole} was allowed to react with A1Me3 (ca. 1 mmole}, the ratio Cp2 TiMeCl • Al:Cp2 TiMe 2 : AlMe3 _nC1n was 2. 34:1. 34:10. 2, corresponding to an "equilibrium ratio" of 0. 06. Both samples were heated to 65 c C. After 6 h, there 37 34 - had been 80% conversion to 7. Reaction of Cp2TiC~ with 2 Equivalents of A1Me 3 • A suspension of Cp2 TiC~ (25 mg, 0.1 mmole) in 500 µ L C6 D6 was allowed to react with A1Me3 (20 µL, 0. 2 mmole) in a NMR tube yielding a deep red 1 solution. The H NMR spectrum indicated that formation of Cp2 TiMeCl • AlMe3 _nCln had occurred (1 H NMR: o 5. 79, 10 H, Cp's; o 0. 92, 3 H, TiMe). This solution was heated to 65e C. After 6 h, 1 had been formed in 80% conversion. Measurement of the Deuterium Isotope Effect. Cp2 TiC~ (25 mg, 0.1 mmole) was suspended in 500 µL C6 D6 in a NMR tube. 10 µL A1Me 3 (0.1 mmole) was added via syringe along with 142 µL of a 0. 7 M toluene solution of A1Me 3 -dg (0.1 mmole). 1 2 H and H NMR spectra of the solution were recorded. The integrals of the species Cp2 TiMeCl • AlMe 3 _nC1n were: Cp (10. 0), CH 3 (1. 5), and AlMe (7. 5). Integration of the 2H spectrum gave a ratio of TiCD3 : AlCD3 of approximately 3:13 (the signals .were somewhat overlapped). The sample was placed in a bath maintained at 56" C. The progress of the reaction was 1 monitored frequently by H NMR spectroscopy. After 2 h there had been a 37% conversion to 1 (and 1-dn). The ratio of the integrals of the cyclopentadienyl resonances to the titanium methylene resonance was 10:1.49. The resonance at o8. 28 was sharp, and did not exhibit any apparent coupling to deuterium. After 44 h, the conversion to 1 and 1-dn was complete, and methylbenzoate (19 µL, 0.15 mmole) and pyridine (12 µL, 0.15 mmole) were added to the sample. 1 H NMR spectroscopy confirmed that the reaction had yielded l-methoxy-l1 phenylethylene: H NMR: OMe, o 3.33; =C!!A!!B, o 4.63, d, o 4. 06, 35 d, JAB= 2. 8 Hz. The relative integrals of the methoxy group and the vinylic hydrogens was 1. 49 ± 0.1: 1. No 1-Ph-1-OMe=CHD could be detected by examining the center of each vinylic doublet, even with resolution enhancement of the spectrum. The limit of detection is estimated to be ± 10%. Measurement of the Kinetics of the Reaction .Cp2 TiCLz + Excess AlMe3 • A 5 mm NMR sample tube was loaded with Cp2 TiC~ (25.0 mg, 0.10 mmole) and taken into a dry box. The walls of the NMR tube were washed down with 400 µL C6 D6 , and A1Me 3 (40 µL, 0.42 mmole) was added via syringe. The sample was capped, removed from the dry bo~, and agitated briefly to allow dissolution and formation of Cp2 TiMeCl. Al. The sample was placed in the probe of the JEOL FX-90Q maintained at 53" C and spectra were obtained at regular intervals. An effective concentration of A1Me 3 at t = 0 was calculated from the relative integrals of the C5H5 and Ti-Me signals of Cp2 TiMeCl •Aland the total integral of the AlMe signal. Least squares analysis of second-order plots of lnkt vs. t yie_lded the rate constants. This procedure was ~epeated for several concentrations of the titanium and aluminum reagents and at several different temperatures. Activation parameters were obtained from least squares analysis of 1n k/T vs. 1 /T plots. 36 References and Notes 1. Ziegler, K., Holzkamp, E., Breil, H., Martin, H., Angew. -- Chem. 67, 541 (1955). 2. For an extensive review of Ziegler-Natta polymerization, see Boor Jr., J., "Ziegler-Natta Catalysts and Polymerizations", Academic Press, New York, 1979. See also Pino, P. and 3. Mulhaupt, R. A ngew. Chem. Int. Ed. Engl. !§!_, 85 7 (1980). (a) See Boor 2 pp. 115-120; (b) Ballard, D. G. H., Adv. Cata!. ~' 263 (1973); (c) Sinn, H., Kaminsky, W. Adv. Organomet. -- Chem. 18, 99 (1980); (d) Zucchini, U., Albizzati, E., and Giannini, U., J. Organomet. Chem.~' 357 (1971). 4. Natta, G., Dino, P., Mazzanti, G., Giannini, U. J. Inorg. Nucl. Chem. !!, 612 (1958). Breslow, D. S., and Newburg, N. R., -- J. Am. Chem. Soc. 79, 5073 (1957), Breslow, D. S., and Newburg, N. R., ibid. 81, 81 (1959). See also ref. 27, pp. 397415 for chemistry ·related to the CP2 Ti-Al system. 5. See reference 3(c), pp. 136-137; Heins, E., Hinck, H., Kaminsky, W. , Oppermann, G., Paulinat, P., and Sinn, H. Makromol. Chem. .,.._,.__,... 134, 1 (1970); Feay, D. C., and Fujioka, G. S., Tenth International Conference on Coordination Chemistry, 1967, Tokyo, Abstracts, p. 1. -- 6. Schrock, R. R., Accts. Chem. Res. 12, 98 (1979). 7. Tebbe, F. N., Parshall, G. W., Reddy, G. S. J. Am. Chem. Soc. !QQ, 3611 (1978); Klabunde, U., Tebbe, F. N., Parshall, G. W., Harlow, R. L., J. Mol. Catal. !!, 37 (1980). 37 References and Notes (continued) 8. Tebbe, F. N., Parshall, G. W., Ovenall, D. W., J. Am. Chem. Soc., !.Q!, 5074 (1979); Tebbe, F. N., Harlow, R. L., ibid., ~ 6151 (1980); McKinney, R. J., Tulip, T. H., Thorn, D. L., Coolbaugh, T. S., Tebbe, F. N. ibid., W, 5584 (1981). 9. Pine, S. H., Zahler, R., Evans, D. A., Grubbs, R. H. J. Am. Chem. Soc. !jg, 3270 (1980); Pine, S. H., et al., 183rd National ACS Meeting, Las Vegas, NV, Abstract 14. 10. (a) Howard, T. R., Lee, J. B., Grubbs, R. H. J. Am. Chem. Soc. !jg, 6876 (1980); Lee, J. B. , et al., ibid., W, 7358 (1981); (b) Lee, J. B., Ott, K. C., Grubbs, R. H. Submitted for publication. 11. Grubbs, R. H., straus, D. A., stille, J., Ho, S. C., Hentges, S. G., Symposium on Applications of Main Group Metal/Transition Metal Reagents in Synthesis, 183rd ACS National Meeting, Las Vegas, NV, Abstract 13, Chem. Eng. News, April 19, 1982, pp., 34-35. 12. De Boer, E. J.M., Grubbs, R.H., unpublished results. 13. After two half-lives, some of the solutions, particularly those with high concentrations of A1Me3 , began to show slight curvature. This is probably due to a competing reaction described by Sinn - and Kaminsky, refo 3( c). 7 can apparently react with the excess A1Me 3 or with another molecule of 1 producing poly-TiCH 2 Al - species., These species can be formed by allowing 7 to stand with excess A1Me3 , or in neat AlMes. After long reaction times, red, 38 References and Notes (continued) toluene insoluble solids can be isolated which have high reactivity towards acetone forming isobutylene, or with DMAP in the presence of olefins yielding titanocyclobutanes. Ott, K. C. , Grubbs, R. H. , unpublished results. 14. Kinetics were not attempted on IndCpTiC½ due to decomposition; Cp'CpTiC½ was excluded due to the inability to obtain this material free from Cp2 TiCI2 • 15. Dormond, A., Ou-Khan, and Tirouflet, J., C. R. A cad. Sci. Paris t. 278, Serie C, 1207 (1974). 16. (a) Gorsich, R. D., J. Am. Chem. Soc. .,...,.._ 80, 4744 (1958); (b) Gorsich, R. D., ibid., ~' 4211 (1960). See also ref. 30. 17. (a) Brown, T. L., Adv. Orgmeto Chem. ~' 365 (1965). (b) Lee, A. G. , "The Chemistry of Thallium", Elsevier, London, 1971. 18. Tuli, R. K., Chandra, K., Sharma, R. K., Kumar, N., Gang, - B. S., Transition Met. Chem. 5, 49 (1980). 19. Kinetics for complexation and alkylation have been extensively studied for the Cp TiC~/A1Et Cl system. Sosnovskaja 20 has 2 2 shown that for the reaction CP2 TiC½ + A1Et 2 Cl ~ Cp2 TiClCl• A1Et 2 Cl K = 140 M- • Fink, 21 has shown that the forward rate of reaction 1 is diffusion controlled, and that the equilibrium constant 3 K = 5 X 10 - 4 5 X 10 at 240 K. 39 References and Notes (continued) 20. Sosnovskaja, L. N., Fushman, E. A,,, Borisova, Lo F", and - Shupik, Ao N., J. Mol. Catal. 9, 411 (1980)0 21. Fink, G., Rotter, R., Schnell, D., Zoller, W", J. AppL Polym. -- -Set 20, 2779 (1976); Fink, G., Rattler, R., Kreiter, C. G., Ang. Makromol. Chem. ~' 1 (1981). 22. Reduction of CI½ TiC½ with A1Et 9 is rapid, producing O. 5C 2H4 and ' ' O. 5C 2H6 and the blue-green CP2TiCIAIEt 2 Cl: See ref. 3(c), - PPo 131, 136; Natta, G., Mazzanti, F., Tetrahedron, 8, 86 (1960). Heating Cp2 TiMeCl-AIMeC¼ yields a blue-green paramagnetic complex. 28 23. £: CH 2 C¼, 9. 08; C6H6 , 2. 28. "Handbook of Chemistry and Physics", 59th Edition, 1979, CRC Press. 24. Pearson, R. G., "Hard and Soft Acids and Bases", Wiley, New York, 1973. Ho, T. -L., "Hard and Soft Acids and Bases Principle in Organic Chemistry", Academic Press, New York, 1977. 25. See Calabro, D. C., Hubbard, J. L., Blevins, C. H., Campbell, A. C., and Lichtenberger, D. L., J. Am. Chem. Soc. W, 6839 (19 81) and references therein. 26. Schrock, R. R., and Sharp, P. R., J. Am. Chem. Soc. 100, """"" 2389 (1978). 27. A1Me3 -dg was prepared by reaction of HgMe 2 -d6 with Al powder in refluxing toluene. Hg-Me 2 -d6 was prepared from CD3 MgI and HgC½. Mole, T., and Jefferey, E. A., "Organoaluminum Compounds", Elsevier, Amsterdam, 1972. 40 References and Notes (continued) 28. Clauss, K. and Bestian, H., Justus Lie bigs Anno Chem. ~ ' 8 (1962) 29. 0 Druce, P. M. , Kingston, B. M. , Lappe rt, M. F. , Spalding, To Ro, and Srivastava, R. C., J. Chem. Soc. {A), 2106 (1969). 30. Cardoso, A., Clark, R. J. H., and Moorhouse, S., J. Chem. Soc. Dalton, 1156 (1980). 3L Threlkel, R. S. and Bercaw, J. E., J. Organomet. Chem. ill, 1 (1977). 32. Conia, J. -M., and Leriverend, M. -L., Bull. Soco Chem. Fr., 2981 (1970), and Ronganayakulu, K., and Sorenson, T. S., Can. J. Chem. .,...,... 50, 3534 (1972) . 33. Drake, No L., and Adams, J. R., J. Am. Chem. Soc. §1, 1326 (1939). 34. Freeman, B. H. , Gagan, J. M. F. and LLoyd, D. , Tetrahedron, 29, .,...,... 35. 4307 (1973). • Bercaw, Jo E., Marvich, R.H., Bell, L. G., and Brintzinger, Ho H., J. Am. Chem. Soc. .,...,... 94, 1219 (1972). Only one equivalent of LiCp* per TiCla· 3THF was used. Anhydrous HCl was used to convert the Cp*TiC½ to Cp*TiCis. 36. Gordon, A. J., Ford, R. A., "The Chemist's Companion", Wiley-Interscience, New York, 1072, p. 303. 37. This "equilibrium ratio" is not intended to represent an equilibrium constant. This ratio was calculated to indicate that the species for med from both reactions were present in approximately equal 41 References and Notes (continued) concentrations. An equilibrium constant would be difficult to estimate due to the complex equilibria involving all of the aluminum compounds. 38. Samuel, E., J. Organomet. Chemo _!1, 87 (1969). 39. Reynolds, L. T., Wilkinson, G., J. Inorgo Nucl. Chem. ~' 86 (1959). 40. Vitz, E o , and Brubaker, C. H. , J. Organomet. Chem. ~' C 16 (1974), have shown that ring exchange occurs photochemically. 4L Nesmeyanov, A. No, J. Organomet. Chem. §1., 55 (1973). The crystal structure of Cp*CpTiC~ is reportedo 420 Hoffman, H. M. R., Angew. Chem. Int. Ed. ~' 556 (1969). 43. Toporcer, L. H., Dessy, R. E., Green, S. I. E., J. Am. Chem. Soc. §1, 1236 (1965). 44. Goddard, R. J., Hoffmann, R., Jemmis, E. D., J. Am. Chem. -Soc. 102, 7667 (1980). .,...,.-.,,-... 42 CHAPTER 2 The stereochemical Consequence of the Interaction of Dimethylaluminum Chloride with Bis-Cyclopentadienyltitanacyclobutanes. Mechanism of Titanacyclobutane Isomerization. 43 INTRODUCTION ~ The role of the cocatalyst in the olefin metathesis 1 and ZieglerNatta polymerization7a reactions is poorly understood. It is generally accepted that the cocatalyst (usually a main group metal alkyl) serves to alkylate the transition metal complex (typically a metal halide). Little else is known about the subsequent role of the cocatalysto It is observed that many widely diverse olefin metathesis catalyst systems (typically a 2nd or 3rd row early transition metal complex with a Lewis acid cocatalyst) yield the same stereochemical results for a given substrate. This indicates that the ligands (or cocatalyst) associated with the transition metal center are not a factor in determining the stereochemistry of the product olefins. 2 There are a few systems reported, however, which give results which indicate that the ligands (or cocatalyst) are a determining factor in the stereochemical outcome. For example, the ring opening metathesis of cyclopentene is catalyzed by a variety of transition metal complexes to produce polypentenamer (equation 1). The system WF /A~Et Cla 3 with an Al/W 6 3 ratio of greater than 4 produces a polypentenamer with greater than 90% trans-double bonds. The same system with an Al/W ratio less than 1 polymerizes cyclopentene to greater than 80% cis-polypentenamer. How the Lewis acid cocatalyst interacts with the catalyst in this system is unknown. A cocatalyst-free system, W(CO) 5 (CPh2 ), polymerizes cyclopentene to greater than 90% cis-polypentenamer. 4 The preceding 44 examples indicate that the type of ligand and the oxidation state of the meta15 , 6 in addition to the presence of a cocatalyst and its concentration are factors in determining the stereochemical outcome of the metathesis reaction in certain catalyst systems. Katz has suggested that the stereochemical dependence of the olefin metathesis reaction in the presence or absence of a cocatalyst may be due in part to a reversible Lewis acid induced cleavage of a transient metallacyclobutane to a 3-metallapropyl cation. Rotation and reclosure of this intermediate would lead to loss of stereochemistry. 4 Katz also suggests that since some systems show no dependence of product stereochemistry on cocatalyst concentration there may be steric and/or electronic factors which determine the susceptibility of a metallacyclobutane to cleavage to a metallapropyl cation. Many Ziegler-Natta catalysts are similar in composition to olefin metathesis catalysts (Ziegler-Natta catalysts are typically a first-row early transition metal halide with an alkylaluminum cocatalyst). Certain syndiotactic specific catalysts are sensitive to the cocatalyst concentration 7b with respect to the degree of syndiotacticity of the resultant polymer, indicating that the Lewis acid cocatalyst is a factor in determining the product stereochemistry. Due to the nature of the olefin metathesis and Ziegler-Natta catalyst systems, the importance and scope of the interaction of the cocatalyst with the active catalytic site is difficult to determine. Reported here is an investigation of the interaction of a Lewis acid cocatalyst with a well-defined olefin metathesis catalyst. Lee and Grubbs' Sa, c investigation of the metathesis reaction catalyzed by 45 Cp2 TiCHDCH(t-Bu)CH 2 (E-~-d1 ) led to the surprising observation that I I when E-~-d1 was cleaved with A1Me2 Cl to yield Cp2 TiCH 2 AlMe 2 Cl (!) and !,-d1 , the deuterated 3, 3-dimethyl-1-butene produced was a 1: 1 mixture of the E and Z isomers (Scheme 1). Consequently, Lee found that E-~-d1 in the presence of A1Me 2 Cl was rapidly and completely isomerized before cleavage to!. and !,-d1 • The mechanism of the Lewis acid interaction with the titanacyclobutanes Cp2 TiCH 2 CRR' CH2 is described. A preliminary account of this work has appeared. Sb E-2- d I H Cp 2T ~ ~ H Scheme 1 D'VV\, :d/-- + J ( I . 2) AIMe2CI Cp:TiCH -----,.~ 2 2 AIMe 2 CI 1 + 1-dl(I) ~ 0) 47 .RESULTS ~ Preparation of~' 1, 1,-d1 , ~' and ~-di- The preparation of E-CpTi CHDCD(tBu)CH2 (~) was accomplished by an olefin metathesis reaction of E-3, 3-dimethyl-1-butene-1, 2-C½ with the thermally labile ' 2CMe2 CH ' 2 (.§.) producing ~ and isobutylene in metallacycle Cp2 TiCH excellent yields and with complete retention of configuration. The metallacycles Cp2 TiCH 2 CMe(iPr)CH2 (1), E-1-d1 , Cp2 TiCH 2 CMePhCH 2 (~) and Z-~-d were prepared by the "conventional" route 9 involving 1 I I reaction of the corresponding olefin with Cp2 TiCH~lMe2 Cl (!) in the presence of a Lewis base such as dimethylaminopyridine. - In contrast to 2, where only one equivalent of olefin is required to obtain good yields of pure metallacycle, large excesses (3-4 fold) of - - olefin are required to prepare 4 and 5 in good yields. One problem encountered in using a large excess of a-methyl-styrene in the prepa- - - ration of 5 is the inability to pump the excess olefin away from 5 at low temperatures . . This leads to decomposition of the metallacycle. Even - when 5 has been isolated and purified, the addition of a-methylstyrene results in decomposition of the metallacycle. Particular care is required during the preparation of these ,3-{3-disubstituted metallacycles due to their thermal !ability. 1 O This characteristic leads to significant stereochemical scrambling during synthesis of the deuterium labelled metallacyclobutanes 1,-d 1 and ~-diEven when the reaction mixtures were maintained at or below -40° C and were worked up rapidly, 1_-d1 -was isolated as a 4. 9 : 1 mixture of E- and Z-isomers, and ~-d1 was isolated as a 4. 3: 1 mixture of the Zand E- isomers. 48 Isomerization Reactions. - When a 0.18M solution of 3 was allowed to react with 0.1 equivalents of A1Me2 Cl, ~ was rapidly equilibrated (requiring less than 1Os) - -- yielding a 1: 1 mixture of E- and Z- 3. No 1, 1-d1 or olefin was - 1 .detected. The H NMR spectra of 3 before and after isomerization are presented in Figures 1 and 2, respectively. - Interestingly when 3, which contained a very small amount of diethyl ether (less than 2%; incompletely removed reaction solvent), was allowed to react with 0. 6 equivalents of A1Me 2 Cl, the isomerization - - of E-3 to Z-3 was very slow, and only 15% of the Z-isomer was observed after 2 hr. at room temperature. A mixture of identical - composition was obtained when 3 was allowed to react with 1. 3 equivalents of A1Me 2 Cl • Et 20 at room temperature for 2 hr. A solution of ~-d1 was prepared at low temperature, and the Z: E ratio was monitored. After 30 min at approximately -20°C, the ratio had changed from 74:26 to 64:36, and was accompanied by the appearance of 17% Z- and E-a-methylstyrene-d1 in the ratio of 64: 36. The metathesis of 14. 5 equivalents Z-a-methylstyrene-d1 - catalyzed by 5 was monitored by 1 H NMR spectroscopy at 0° C by observing the relative ratio of the vinylic hydrogens corresponding to E- and Z-{3-deutero-a-methylstyrene. The initial rate of metathesis (productive plus non-productive; i.e. , the rate of metathesis is twice the rate of isomerization) was calculated to be 1. 2 turnovers TC 1 min- 1 • The metathesis proceeded less rapidly after 30 min at 0° C, and the solution had changed in color from pale red to brown. 49 Fig. 2 2.6 2.4 2.0 2.2 1.8 1.6 PPM 1 Figure 1 (Bottom). 500 MHz H NMR spectrum of the methylene region of E-~ in C 6 D 6 , ambient temperature. Figure 2 (Top). After addition of 0.1 equiv. A1Me 2 Cl. 50 When solutions of 1-d1 or R_-d1 were allowed to react with excess A1Me 2 Cl at low temperature, the olefin which was produced was a 1 : 1 mixture of E- and Z- isomers. In some experiments where less than stoichiometric amounts of A1Me 2 Cl were used, the metallacycle that remained was also a 1 : 1 mixture of Z- and E- isomers. In all experiments involving 1-d 1 or R_-d1 , the ratio of the resulting !_-d1 to d~uterated olefin was 1. 0 : 1. 6. A competition experiment was performed to insure that all of the isomerization of 1-d1 and R_-d1 was A1Me 2 Cl induced, and not simply due to thermal nonstereoselective olefin metathesis. This experiment was carried out by reacting slight excesses of A1Me 2 Cl with ~ in the presence of 1 to 2. 6 equivalents of Z-a-methylstyrene-d1 and monitoring the amount of !_-d1 formed. No !_-d1 was detected under any circumstances; however, some isomerization (2-13%) of the deuterated a -methylstyrene was observed. A crossover experiment was performed to determine the extent of - intermolecularity of the isomerization. When 3 was reacted with excess A1Me 2 Cl at 50° C for 1 hr, GC/MS of the volatiles indicated that only 3, 3-dimethyl-l-butene-d1 and-~ were produced. None of the do- nor da-olefins were detected. The kinetics of the reaction of ~ with a slight excess of A1Me 2 Cl - to produce 1 and 3, 3-dimethyl-1-butene was also studied. Second-order plots of the reaction were linear to greater than 80% completion. The observed second-order rate constant at 294 K was 5. 8± 1. 7x 10-3 M- 1 s- 1 • An Eyring plot of In (k/T) vs. 1 /T yielded the activation parameters AS*= -38± 1 e. u. and aGt 4 = 20± 1 kcal-mole- 1 • 51 DISCUSSION ~ The A1Me 2 Cl catalyzed isomerization of i to a 1: 1 mixture of E- and Z-isomers occurs before the formation of detectable amounts of olefin or !_. 12 The rate of A1Me Cl induced isomerization of E-~-d is 2 1 several orders of magnitude faster than the rate of thermal isomerization, which has at.!~ 2 hr at 50" C. 13 2 A study of the kinetics of the reaction of ~ with A1Me 2 Cl was initiated to probe the interaction of A1Me 2 Cl with titanacyclobutanes. - The kinetics data indicated that the reaction was first-order in both 2 and A1Me 2 Cl with a fairly large negative activation entropy (-38 e. u. ). - The kinetics, along with information concerning the activity of 1 as an olefin metathesis catalyst, 14 are consistent with the mechanism - depicted in Scheme 2. The existence of the olefin-1 complex is inferred from Tebbe's olefin metathesis results. The magnitude of as* is consistent with an ordered transition state and may reflect significant Al-C and Ti-Cl bond formation. It was originally considered that preceding the cleavage of E-~-d1 to!. and !_-d1 , a reversible A1Me 2 Cl-mediated '3-hydrogen abstraction occurred to yield a titanocene all.yl cation and an aluminum hydride anion. This mechanism is depicted for E-i-d1 in Scheme 3. A rotation of a C-C bond via an r,3- to 11 1 -allyl rearrangement followed by closure to titanacyclobutane 15 would result in isomerization of the titanacyclobutane. To investigate whether the /3-H was involved in the isomerization, the labelled titanacyclobutanes f-d1 and §_-d1 were prepared. The thermal lability of these J3-J3-disubstituted titanacyclobutanes results in 52 \/ \/ <! )_(> ' ...__ ► u u + ~1~ 1 0¥ - ~~u C\J Cl. u ~ (> J C\J Q) E Q) ~ u (/) t-C\J Cl. u ( '-fJ ... t-C\J Cl. C\J Cl. -- ~ Z- 2 - d1 D Cp 2T i ~ + AIMe 2CI H E- 2 -d 1 H ,_ H Cp 2T i ~ + AIMe 2CI . Scheme 3 ... ... [CP2~-t H Cp~~ I l + [HAICIMe2r + + [HAICIMeX r ~ 01 54 moderately rapid isomerization of the deuterium labelled compounds even at -20" C. A plausible mechanism 16 for the thermal isomerization of ~-d1 is represented in Scheme 4. It is proposed that the rate determining step involves opening of the titanacyclobutane to the methylene- - - olefin complex A which is in equilibrium with the methylene complex B and olefin. The olefin can return in either of two orientations with respect to the TiCHD unit resulting either in return to Z-~-d1 or in isomerization to E-~-d1 • This is consonant with the observation that E-i-d1 isomerizes thermally at half the rate at which diphenylacetylene i I reacts to form Cp2 TiCH 2 CPh=CPh. This suggests that the barrier to rotation of Cp2 Ti=CH 2 is greater than the barrier to opening of titanacyclobutane to Cp2 Ti=CH 2 and olefin. This latter process has a ~Gt00 = 24 kcal-moi- 1 • Sc The barrier to rotation in C~ Ti=CH 2 has been calculated17 to be 19 kcal-moC 1 , which is in disagreement with our results, and may be a rather low estimate. Reaction of A1Me2 Cl with ,1-d1 or ~-d1 results in immediate cleavage of the titanacyclobutanes to !. and !_-d1 • The deuterated olefins (2, 3-dimethyl-l-butene-l-d1 or a-methylstyrene-d1 ) produced were 1: 1 mixtures of E- and Z-isomers. Due to the rapidity of the cleavage reaction, the stereochemistry of the titanacyclobutanes could not be examined except when less than a stoichiometric amount of A1Me 2 Cl was present in solution. In those cases, the titanacyclobutanes were observed to be 1 : 1 mixtures of E- and Z-isomers. However, due to the moderately fast olefin interchange discussed above, the relative contribution of thermal and A1Me 2 Cl induced isomerizations could not be determined. Therefore, competition • ~ Me Ph Z-5-d 1 A Me-.:... ./ CP2 ' Cpl,~o T ~ Ph H D H Scheme 4 1l D H~ -Me E - 5 -d 1 Cp 2Ti~Ph 0 Me ;;:::: [cpli=CHD] + =<Ph CJ1 CJ1 56 - experiments were performed. When 5 was allowed to react with A1Me2 Cl in the presence of excess Z-a-methylstyrene-d1 , no !_-d1 was observed, indicating that the reaction of ~ with A1Me 2 Cl is much faster than olefin interchange (equation 1). (1) The above experiments indicate that A1Me 2 Cl catalyzes the isomerization of 1,-d1 and ~-d1 at a rate faster than the thermal isomerization. This also suggests that it is unlikely that the /3-hydrogen is involved in the A1Me2 Cl catalyzed isomerization of the monosubstituted titanacyclobutanes such as E-~-d1 or ~- - Crossover experiments using 3 were performed to test the intermolecularity of the · reaction of 1 with A1Me 2 Cl. Any intermolecular process would result in formation of da- or ds-3, 3-dimethyl1-butene; since neither was detected, it is unlikely that the A1Me 2 Cl catalyzed isomerization involves intermediates such as free aluminum hydrides. It is also important to note that there is no a to {3 hydrogen scrambling in 1 during reaction with A1Me 2 Cl. The evidence presented above leads to the proposal that the isomerization involves a mechanism which scrambles only the a -positions of the titanacyclobutane, and which can ultimately lead to - C-C bond cleavage to yield 1. The first step of the isomerization is proposed to be a rapid, reversible transmetallation at the Ti-C bond 57 (Scheme 5). Such transmetallation reactions have precedent in both Ti18 and zr 19 chemistry. The transmetallation apparently requires that aluminum have chloride as a ligand, since A1Me 3 induced catalysis of the titanacyclobutanes is ineffective, whereas AlCls effectively catalyzes the isomerization. The 3-aluminapropyl titanocene chloride - can undergo slow C-C bond cleavage to give 1, or rapid, reversible inversion at the carbon adjacent to aluminum followed by reverse transmetallation back to titanacyclobutane. The similarity of this mechanism to Katz ,l 4 suggestion· is apparent. Inversion at the a-carbon atom in Li, Mg, and Zn a1kyls 21 via an SE reaction has been demonstrated to be fast on the NMR time 2 scale. Aluminum alkyls also undergo a -carbon inversion on the NMR time scale (Figure 1), but the mechanism is not understood in similar detail. 22 Inversion at the a -carbon in triisohexylaluminum is known 22 a to proceed with first-order kinetics, and manifests very unusual behavior in the presence of Lewis bases. The rate of inversion of triisohexylaluminum is slowed to below detectable levels when either diethyl ether is used as solvent 21 b or, curiously, when trace amounts of ether 22a are added to toluene solutions. 28 Analogously, when 3 is - allowed to react with 1. 3 equivalents of A1Me 2 Cl • Et 2O, isomerization is very slow. 23 The same extent of isomerization is observed when 3, - which contains traces of diethyl ether, is allowed to react with A1Me 2 Cl. The similarity between the lack of alkylaluminum inversion and A1Me 2 Cl induced isomerization of titanacyclobutanes in the presence of ether (or obviously, in the presence of DMAP) suggest that both processes may proceed through similar if not identical pathways. + AIMe 2CI II D '1/H 61 ,, H Me 2 Al 111D Cp2Ti~ H 61 Me 2 AI Cp 2T i ~ E- 2 -d 1 H Cp 2T i ~ Scheme 5 ~ .... .....- ► ... D Cp 2T i ~ +-=" + o~===" 1 Cp2r(;A1Me2 + 1-di Cl Me2A1-!1 H D Cp2TiS<'t CJ1 0:, 59 The Lewis acid cocatalyst A1Me 2 Cl is proposed to interact directly with a Ti-C bond in the titanacyclobutane metathesis system, which has a dramatic influence on the stereochemistry of the product olefins. Osborn et al. , 32 have reported that Lewis acids in the OW(CH 2R) 3X system interact with the "spectator" oxide, 33 creating a greater electron deficiency at the metal center facilitating a-hydrogen 34 abstraction and subsequent electrophilic attack by an olefin. Schrock has reported that AlC½ accelerates the rate of metathesis catalyzed by OW(CHCMe3 )(PEt 3 ) 2 Cl and suggests that AlC½ interacts with or removes completely a Cl- ligand, creating a vacant coordination site. Thus it appears that there are many types of cocatalyst-catalyst interactions. ·' A side from the isomerization reaction discussed above, this work revealed some features concerning the enhanced reactivity of the (3, (:3disubstituted titanacyclobutanes (1, .§_, and ~) versus the tJ-monosubstituted titanacyclobutanes such as 2. 27 The second-order rate con- - stant of the reaction of~ with A1Me 2 Cl was k r.J 6 x 10-3 M- 1 s- 1 at 21" C. The titanacyclobutanes 1, .§_, and~ react with A1Me 2 Cl at o•c at a rate - too fast to measure by NMR spectroscopy. 2 reacts with diphenylacetylene to yield Cp2 TiCPh=CPhCH2 with a half-life of approximately 1 hr Ba, c at 40° C, while the analogous (3, (3-disubstituted titanacyclobutanes react with Ph2 C2 with hall-lives of 15-20 min at room temperature. Similar effects are manifested in the secondary deuterium isotope effects observed in these reactions. The reaction of E-~-d1 or ~ with A1Me 2 Cl (Scheme 1) proceeds with a secondary deuterium isotope 60 effects of ca. 2. Sc' 35 These facts may reflect a few features of the transition state of these reactions. For the reaction with A1Me 2 Cl, the larger isotope effect exhibited by the disubstituted titanacy.clobutanes may signify a more product or titanium-methylene-olefin-like transition state. The diphenylacetylene reaction is known to exhibit first-order kinetics, and is thought to proceed via the titanium methylene intermediate Cp2 Ti=CH 2 • For this dissociative pathway, the isotope effects would be expected to be similar for mono- and disubstituted titanacyclobutenes, which is in agreement with our observations. 61 EXPERIMENT AL ~ Titanocene dichloride was purchased from strem Chemicals or Boulder Scientific and purified by Soxhlet extraction with dichloromethane. A1Me 3 (neat) was purchased from Alfa or as a 2M solution in toluene from Aldrich. A1Me 2 Cl was purchased from Texas Alkyls. AlCls was purified by sublimation. 3, 3-dimethyl-1-butene, 2, 3-dimethyl-1-butene, and a-methylstyrene were purchased from Aldrich, and stored over Linde 4.A molecular sieves. Isobutylene purchased from Matheson was used as received. Zirconocene dichloride was purchased from Boulder Scientific. LiAID4 was purchased from Merck, Sharp, and Dohme. Li(0-t-Bu) 3 AlH was purchased from Aldrich. Dimethylaminopyridine (DMAP) and diphenylacetylene were purchased from Aldrich and recrystallized from hot toluene. Cp2 ZrHCl and Cp ZrDCl were prepared by literature methods. 19 a Phenylacetylene 2 was purchased from Aldrich. t-Butylacetylene and i-propylacetylene were purchased from Farchan and used as received. Phenylacetylene-d1 was prepared by deprotonation of phenylacetylene with n-BuLi followed by deuterolysis with D2 O. Dichloromethane was stirred over P 2 O5 or CaH 2 and degassed on a high vacuum line. Pentane, hexane, and petroleum ether were stirred over concentrated H2 004 , then over CaH 2 , and vacuum transferred onto purple sodium-benzophenone ketyl dissolved in a small amount of tetraglyme. Toluene, diethyl ether, and benzene were stirred over CaH2 and vacuum transferred onto purple benzophenone ketyl. Solvents thus dried and deoxygenated were vacuum transferred 62 into dry vessels equipped with teflon needle valve closures and stored under Ar. Benzene-~ (Merck, Sharp and Dohme) and toluene-~ (Aldrich} were dried and deoxygenated by several freeze-pump-thaw cycles of sodium-benzophenone solutions. Dichloromethane-~ (Noren, Inc.) was stored over CaH2 and degassed by several freeze-pump-thaw cycles. General Procedures All manipulations of air and/or . moisture sensitive compounds were carried out using standard high vacuum, Schlenk line and dry box techniques. Argon used in Schlenk work was purified by passage through columns of BASF RS-11 (Chemalog) and Linde 4A molecular sieves. NMR spectra were recorded with a JEOL FX-90Q (89. 60 MHz 1 H, 22. 53 MHz 13 C, 13. 76 MHz 2 H} or a Bruker WM-500 (500.13 MHz 1 H, 76. 76 MHz H). Kinetics by NMR spectroscopy were run in 2 automated mode on the JEOL FX-90Q. Temperatures were calculated by measuring avMeOH 2 ~ and were constant to within± 0.1 ° C. Difference NOE's 11 were measured on the Bruker WM-500. Preparative gas chromatography was performed on a Varian Aerograph Model 920 using 20' Durapak or 6' of 10% saturated AgNO 3 /tetraethyleneglycol on Chromasorb P. GC/MS was performed on a Kratos MS25. Preparation of E-3, 3-dimethyl-1-butene-1, 2-~. This is essentially the method of Schwartz. t 9a LiAID (144 mg, 3. 4 mmole} 4 was added to a solution of Cp2 ZrC~ (4g, 13. 7 mmole} in THF under Ar. A solution of t-butylacetylene (1.12g, 13. 3 mmole) in toluene was slowly added to the resulting THF suspension of Cp2 Zr0Cl. The 63 suspension was stirred overnight and the solvent was removed from the yellow precipitate in vacuo. The precipitate was suspended in 5 ml toluene, and D2 O (250 µL, 13. 8 mmole) was added. The resulting white suspension was stirred for several hours and then cooled to 0° C. Approximately half of the reaction liquids were vacuum transferred into a collection bulb containing a few Linde 4A molecular sieves. The product E-3, 3-dimethyl-1-butene-1, 2-~ was separated by preparative gas chromatography (AgNOiChromasorb P, 6', 70° C) and was shown to be > 98% deuterated in the 1 and 2 positions (1H NMR). Yield, O. 95 g, 80%. Variable amounts of additional product may be recovered from the remaining toluene suspension by vacuum transferring the remaining volatiles from the Zr residues. Preparation of E-2, 3-dimethyl-1-butene-1-d, and Z-a-Methyl styrene. These preparations are by the method of Negishi. 26 A1Me 3 (3. 8 mL, 40 mmole) was added to Cp2 ZrC½ (5. 84 g, 20 mmole) suspended in 20 mL toluene under Ar. i-Propylacetylene (2 mL, 20 mmole) was added to this mixture via syringe. The reaction vessel was closed off from the Ar system, and heated to 45 ° C for 3 h. The volatiles were removed in vacuo, leaving sticky yellow solids which were resuspended in 5 mL toluene. The suspension was cooled to -10° C, and a large excess of D2O was added slowly. The initial reaction is very vigorous. Addition of D2O was continued until the yellow color had been discharged. The white suspension was warmed to 0° C, and approximately half of the reaction liquids were vacuum transferred into a receiver containing a few 4.A molecular sieves. The product was isolated by preparative GC (6' AgNO3 /Chromasorb P, 64 70° C). Yield 1. 25 g (74%). The E-2, 3-dimethyl-l-butene-l-d1 was deuterated to greater than 98% in the 1 position (500 MHz 1 H NMR spectroscopy). Z-a-Methylstyrene was prepared in a similar manner from phenylacetylene-d1 using CH 2 C~ as solvent. After hydrolysis of the reaction with 3M HCl, the product was extracted into ether. The combined ether extracts were dried over MgSO4 , and ether was stripped off leaving 4.1 g (8~) of a.-methylstyrene-d1 which was > 95% of the Z-isomer (1H NMR). Preparation of 1: The preparation of this compound has been previously described by Tebbe et al. l 4a A slightly modified procedure is given here. Neat A1Me3 (21 mL, 220 mmol) was added via cannula to a suspension of Cp2 TiC~ (50 g, 200 mmol) in toluene (200 ml) to give a homogeneous red solution. Evolution of methane began immediately. After stirring this solution 48 hr, all volatiles were removed by vacuum distillation into a cold trap (Caution: the A1Me 2 Cl evolved reacts violently with protic media and due precautions should be exercised). A small sample of the resulting red crystalline material may be I \ assayed for the presence of Cp2 TiClA1MeC1CH 2 (the -CH 2 - resonance appears at ~ 7. 68 as an unresolved AB quartet). This material was converted to ! by addition of an equivalent amount of A1Me 3 to the reaction mixture, which was redissolved in toluene. The resulting solution was then filtered through a pad of Celite supported on a coarse frit, and concentrated to the point of saturation ("' 150-170 mL total volume). This saturated solution was carefully layered with an equal volume of hexane or petroleum ether and allowed to stand undisturbed 65 at -20° C for 2-3 days. The supernatant was removed via cannula and the red crystalline mass was washed with several portions of petroleum ether at -20° C. The solids were dried under high vacuum. Typical yields were 30-35 g (53-61 %) . An additional crop of less pure material (10-15 g) was obtained by condensing the mother liquor as above. Preparation of Cp2 TiCH 2 CH-t-BuCH 2 (ID. This is an improvement of the original preparation. 9 To 1 (1 g, 3. 5 mmol) dissolved in - toluene (6 mL) was added 3, 3-dimethyl-1-butene (500 µL, 3. 9 mmol) and N,N-dimethylaminopyridine (DMAP) (472 mg, 4 mmol). The resulting red solution was transferred into 60 mL of vigorously stirred -20°C pentane or petroleum ether. The DMAP-A1Me 2 Cl adduct precipitated as a yellow-orange mass which was rapidly filtered to give a clear red solution. This solution was evaporated to dryness in vacuo - to yield 770-800 mg of 2 (80-83%). This crude material was recrystal- - lized from diethyl ether to give 480 mg (50%) of 2 as red needles. NMR data of 2 have previously been published. 9 - Preparation of E-Cp2 TiCHOCD(t-Bu)CH 2 (3). To a solution of§.. (238 mg, 0. 96 mmole) in Et20 under Ar at o•c was added E-3, 3dimethyl-1-butene-1, 2-~ (150 µL, 1. 2 mmole). The solution was stirred for 15 min, and warmed to room temperature. stirring was - continued for 30 min. The solvent was removed in vacuo to yield 3 (240 mg, 90%) as red needles. 1 H NMR spectroscopy indicated that 1 this material consisted solely of the E-isomer. The H NMR assignments of the stereochemistry were in agreement with the assignments - based on NOE spectroscopy of 2. 66 4 mmol) and 2, 3-dimethyl-1-butene (1. 8 mL, 14 mmol) in 3 mL CH C½ Preparation of 4 and 5: To a suspension of DMAP (427 mg, 2 - at -3o•c was slowly added a solution of 1 (1.0g, 3.5 mmol) in 3 mL of CH 2 C¼. This homogeneous red solution was stirred for 15 min while it was allowed to warm to -10° C. The solution was slowly added dropwise via cannula into 50 mL of vigorously stirred pentane at -30° C. The DMA P-A 1Me2 Cl adduct was rapidly filtered, and the resulting clear red solution was evaporated to dryness at or below -10° C to yield 480-770 mg (50-80%) of red powder. This material could be recrystallized from diethyl ether by slowly cooling saturated solutions from 0° - - to -50° C to yield red needles of 4. NMR data for compound 4 are listed in Table 1. - Compound 5 was prepared similarly, except only 1 equivalent of a-methylstyrene was used. Use of excess olefin leads to decomposition upon attempting to pump the excess a -methylstyrene from the metallacycle. This was circumvented by crystallizing the metallacycle from the pentane-methylene chloride solution at -78 • C, and washing the metallacycle free of a-methylstyrene with cold (-50° C) pentane. The material so obtained is of suitable purity for further reactions. Yield 320-420 mg (30-39%). NMR data are listed in Table 1. Preparation of E-Cp2 TiCHDC(Me)(iPr)CH2 (4-d1). To a suspension of DMAP (215 mg, 2 mmole) and E-2, 3-dimethyl-l-butene-l-d1 (900 µL, 7 mmole) in 2 mL CH 2 C¼ at -40° C was slowly added via cannula a solution of!. (500 mg, 1. 75 mmole) in 2 mL CH 2 C~ under Ar. The resulting homogeneous red solution was stirred for 15 m at -40 ° C, and added dropwise via cannula into 30 mL of vigorously stirred, cold 67 Table 1. NMR Data 4 "' 500.13 MHz H, -20" C, CD2 C¼, shifts vs. CHDC¼ at ~ 5. 320. 1 fj 5. 862, 5. 851, 2s (Cp-H), 2. 777 d, J = 9. 5 Hz (HX), 2. 047, d, J = 9. 5 Hz (Hi), o. 801, d, J = 5. 5 Hz, (He), o. 753, hep, J = 5. 5 Hz (Hn), 0.701, s(HE). 22. 53 MHz 13 C, gated decoupled, -30" C, CD2 C¼, shifts relative to CDC¼ at ~53.8~ 6111.3, 110.7, d of m (Cp's), 79.6, dd, J = 137 Hz (Ca), 42.7, d, J = 130 Hz (C 0 ), 25.1 q, J = 124 Hz (CE), 19.4, q, J = 124 Hz (Cc), 13. 4, s (C/3 ). 68 Table 1 (continued) Cps, 'Ti / Ha CpA Ph 5,...., 500.13 MHz 1 H, -20"C, CD2 C~, shifts vs. CHDC~ at o5.320. 06.0ll, s(c; ), 5.673, s(c; ), 2.818, d, J=8.8Hz(Hp)2.227, d, A * B J = 8.8 Hz (HB), 1.058, s (Me), 7.542, 7.429, 7.128, m (Ph). 22. 53 :MHz 13 C, -20" C, CD2 C~, shifts vs. CD2 C~ at o53. 8. o111. 4,· 111. 2 (Cp's), 73. 9 (CCY), 41. 2 (Me), 15. 8 (Cf3), 128. 3, 127. 8, 126. 8, 125.1 (Ph). *Assignments by difference NOE spectroscopy. 69 (-50° C) pentane. The DMAP-A1Me 2Cl was rapidly filtered, and the solvent was removed in vacuo from the resulting red solution while maintaining the temperature below -40° C. This procedure yielded 310 mg (64%) of red powder. It is important that a 3-4 fold excess of the olefin be used, otherwise, the yields are poor, and the crude material is less pure. 1 2 H and H NMR spectroscopy indicated that this material was a mixture of the E- and Z- d1 -metallacyclobutanes in the ratio of 4. 9: 1. The 1 H and 2H NMR spectra were consistent with the assignments based on difference NOE results. Preparation of Z- Cp2TiCHOC(Me)(Ph)CH 2 (~-d1). This compound - was prepared in an identical manner to 5, except that after ca. 75% of the solvent had been removed and most of the metallacycle was judged to have precipitated, the supernatant was drawn off, and the remaining red powder washed repeatedly with cold (-50° C) pentane to remove any excess olefin. The solids were dried at -40° C under high vacuum. This procedure yielded .§,-d1 (105 mg, 200k,) which was judged pure by 1 H NMR spectroscopy. 2H NMR spectroscopy indicated that this material was a mixture of Z- and E- d1 -metallacyclobutanes in the 2 1 ratio 4. 3 : 1. The H and H NMR spectra were consistent with the - assignments based on difference NOE results for 5. I I Preparation of Cp2 TiCH 2 CMe 2 CH 2 (6). This metallacycle may be - prepared as for 4. This metallacycle is considerably more stable than - - 4 or 5, and may be handled in non-halogenated solvents at room tern- perature for short periods of time without significant decomposition. Typical yields were 40-SOOk,. The metallacycle can be conveniently purified by crystallization as deep red needles by cooling diethyl ether 70 - solutions of 6 slowly to -50" C. Isomerization of 3 Catalyzed by A1Me 2 Cl. This is similar to the experiment conducted by J. B. Lee, Sb only using a catalytic amount of A1Me 2 Cl. A solution of~ (20 mg, 0. 072 mmole) in C6 D6 (400 µL) was prepared in a 5 mm NMR tube. The tube was capped with a latex 1 septum and removed from the dry box. The H NMR spectrum was recorded and indicated that the sample contained only the E- isomer. A1Me 2 Cl (7 µL of lM O. 007 mmole, solution in C6 D6 ) was added via syringe. The 1 H NMR spectrum was recorded, and indicated that the ratio of E- and Z-~-d1 was 1: 1. Reaction of ~ with AlMe 2Cl • Et 20. A solution of ~ (20 mg, O. 072 mmole) was prepared in 500 µL C6 H6 in a 5 mm NMR tube. The tube was capped with a latex septum, and removed from the dry box. The 76. 76 MHz 2II spectrum was recorded, and indicated that the solution contained only the E-isomer. A solution of A1Me 2 Cl (50 µL, 1. 84 M, 0. 092 mmole) was added to the solution via syringe. The 2 H spectrum indicated that no isomerization had occurred. After 2 hr, 5% of the Z-isomer was observed. Only a few percent 1-d1 could be observed. Isomerization of~ with A1Cl3• AlCls (10 mg, 0. 075 mmole) was added to a C6 D6 solution of ~ (20 mg, 0. 072 mmole). The H NMR 1 - spectrum was monitored and indicated that 3 had been isomerized to a 1 : 1 mixture of E- and Z-isomers. There were also indications of decomposition. The suspension slowly turned from its initial red color to brown-green. 71 Isomerization of 5-d1 • A 5 mm NMR tube was loaded with §_-d1 (15 mg). The tube was capped with a latex septum and removed from the dry box. 500 µL of toluene was very slowly added to the NMR tube which was nearly completely submerged in a -50° C bath (this must be done to insure that significant decomposition of the metallacycle does not occur). The sample was carefully dissolved while at -50° C. The 2 76. 76 MHz H spectrum was recorded in a NMR probe cooled to ca. -20° C. The ratio of trans- and cis-1_-d1 was 2. 8 : 1. Approximately 1(1)/o of the sample had decomposed to olefin by this time. This sample was allowed to sit in the probe for 30 min, after which the ratio of trans- and cis-5-d1 was 1. 8: 1 (at 17% olefin formation). (The ratio of z- and E-a-methylstyrene was also 1. 8: 1.) Nonstereoselective exchange of Z-a -methylstyrene-d1 with 2. §_ (5 mg, 0. 016 mmol) was dissolved in 500 µL toluene-~ at -20° C. 30 µL (0. 23 mmol, 14. 5 equivalents) of Z-a-methylstyrene-d1 was injected via syringe. Tµe solution was mixed well at -20° C, and placed in an NMR probe he Id at 0 ° C. The extent of isomerization of the olefin was monitored by measuring the ratio of the integrals of the peaks at fj4. 98 and 65. 33 (Z and E vinylic hydrogen resonances, respectively) by 1 H NMR spectroscopy. After 5 minutes, the ratio 1 Z/E was 3. 7: 1 corresponding to 1. 2 turnovers TC 1 min- • Isomerization of 1-d1 and 2-d1 Catalyzed by A1Me 2Cl. A solution of 1_-d1 (15 mg, 0. 05 mmole) was prepared at -50° C as described above. 2H NMR at 76. 76 MHz at below -20° C indicated that the solution contained a mixture of E- and Z-1_-d1 in the ratio of 4. 9: 1. The solution also contained ca. 5% of a mixture of E- and 72 Z- 2, 3-dimethyl-l-butene-1-di- The sample was removed from the probe, and cooled to -50e C. A solution of A1Me 2 Cl, 1 M in toluene, (100 µL, 0.1 mmole) was added via syringe, and the sample was 2 replaced in the NMR probe. The H spectrum indicated that !-d1 had been formed, and that the olefin produced was a mixture of E- and z2, 3-dimethyl-l-butene-l-d1 in the ratio of 1: 1. Approximately 50% of the metallacycle remained, and was also approximately a 1 : 1 mixture of E- and Z- isomers (natural abundance Ph-CH 2 D interferes slightly with the determination of the latter ratio). When all of the metallacycle had reacted, the ratio of (E+Z)-2, 3-dimethyl-1-butene-l-d1 to !_-d1 was 1. 6 : 1. Similarly, when ~-d1 (15 mg, 0. 048 mmole) was allowed to react with A1Me 2 Cl (100 µL of 1 M solution, 0.1 mmole) (initial ratio Z: E ~-d1 = 4. 3: 1, 5% (Z+E) a-methylstyrene-d1 ), the ratio of the Z- and E- a-methylstyrene was 1: 1, while the ratio of !_-d1 to (E+Z)-amethylstyrene was 1. 6 :.1. Reaction of 2, with A1Me 2 Cl in the Presence of Z-a -Methylstyrene-di- A toluene solution of~ (10 mg, 0. 032 mmole) was prepared at -50e C in an NMR tube as described above. Z-a-Methylstyrene-d1 (4 µL, 0. 031 mmole, 95% Z) was added via s-yringe, followed by A1Me 2 Cl (19 µL of a 2M toluene solution, 0. 038 mmole). The sample was agitated at -50e C, and then shaken vigorously as the tube was warmed to room temperature. 2 H NMR spectroscopy indicated that no !_-d1 was formed. This experiment was repeated several times, and no !_-d1 was observed. The z- a-methylstyrene was partially isomerized to the E-isomer to the extent of 2-13 (±3)% . 73 GC/MS Analysis of the 3, 3-dimethyl-1-butenes Produced on Reaction of ~ with A1Me 2 Cl. A flask was loaded with ~ (20 mg, O. 072 mmole) and removed from the drybox. Toluene (300-400 µL) was vacuum transferred into the flask, and A1Me 2 Cl (12 µL, 14 mmole) was added under a counterflow of Ar. The reaction vessel was closed, and heated at 50° C 1 h. The volatiles were vacuum transferred into a NMR tube. A sample was withdrawn from the NMR tube for GC/MS analysis: m/e (intensity) 87 (114), 86 (1894), 85 (1921), 84 (114), 71 (21,393), 70 (21,805), 57 (1 o, 996), 56 (1 o, 423), 55 (9,480), 54 (1 o, 817), indicating no <ls- or d0 - 3, 3-dimethyll-butene was produced. Kinetics of the Reaction of ~ with A1Me 2Cl. Stock C6 D6 solutions of i (34 mg, 0.125 mmole, 1. 00 mL C6 D6 , 0.125 M) and A1Me 2 Cl (14 mg, 0.151 mmole, 1. 00 mL, 0.151 M) were prepared. Samples for kinetics were prepared by adding a 200 µL aliquot of the stock solution - of 2 by syringe into a 5 mm NMR tube and then freezing the solution at -50° C in the dry box freezer. The sample was removed from the freezer and a 200 µL aliquot of the A1Me 2 Cl stock solution was added by syringe. The tube was rapidly capped, removed from the dry box and refrozen at -50" C. The sample was later thawed, and placed in the NMR probe held at constant temperature. Four measurements of the rate were recorded at 296 K. Activation parameters were derived from additional rate data recorded at 305 K and 279 K. Second-order plots were linear to greater than 80% completion. 74 References and Notes 1o For recent reviews of the subject, see: (a) Grubbs, R. H., Prog. Inorg. Chemo ~' 1 (1978); {b) Grubbs, R. H., "Comprehensive Organometallic Chemistry," G. Wilkinson, editor, Ch. 54; In press; {c) Calderon, N., Lawrence, J. P., Ofstead, E. A., Adv. Organometal. Chem. !1, 449 (1979); {d) Basset, J. M., and Leconte, M., Chem. Technol. 10, 762 {1980). ~ 2. - {a) Leconte, M., and Basset, J. M., Nouv. J. Chem. 3, 429 (1979); {b) Bilhou, J. L., Basset, J. M., Mutin, R., and Graydon, W. F. , J. Chem. Soc. Chem. Comm. , 970 (1976); (c) Bilhou, J. L. , Basset, J. M. , Mutin, R., and Graydon, W. F., J. Am. Chem. Soc. 99, 4083 {1977). ~ 3. Gunther, P., Haas, F., Marwede, G., Nutzel, K., Oberkirch, W., Pampus, G., Schon, N., and Witte, J., Angew. Makromol. Chem. 14, 87 (1970). ~ 4. Katz, T. J. and Hersh, W. H., Tetrahedron Lett. 585 (1977). 5. Under identical conditions, the polypentenamer derived from WC1a/A1Et3 and cyclopentene is trans, while the polymer derived fromMoC1 5 /AlEt 3 iscis. Natta, G., Dall'Asta, G., and Mazzanti, G., Angew. Chem. Int. Ed. ~' 723 (1964). 6. Pampus, G., and Lehnert, G., Makromol. Chem. -""-""-" 175, 2605 {1974). 7. (a) See .Boor, -J r., J., "Ziegler-Natta Catalysts and Polymeri- zations", Academic Press, New York, 1979; (b) ibid., pp~ 118-124. 8. (a) Lee, J. B., Ph.D. Thesis, Michigan state University, 1982; 75 References and Notes (continued) (b) Ott, K. C., Lee, J. B., and Grubbs, R. H., J. Am. Chem!. Soc., 1Q1, xxx (1982), in press; (c) Lee, J. B., Ott, K. C., and Grubbs, Ro H. , submitted for publication. 9. Howard, T. R., Lee, J.B., Grubbs, R.H., J. Am. Chem. Soc. ~ ' 6876 (1980). 10. Warming hydrocarbon solutions of the {3-{3-disubstituted titanacyclobutanes yield [Cp2 TiCH 2 ] 2 • Ott, K. C., and Grubbs; R. H., J. Am. Chem. Soc. .,...,_... 103, 5922 (1981). See also Ch. 3, this thesis . 11. Noggle, J. H. , Schirmer, R. E., "The Nuclear Overhauser Effect; Chemical Applications", Academic Press, N. Y., 1971, pp. 174-187. 12. The olefins are not isomerized by A1Me2 Cl after several hours at 13 60°C in benzene-d6 • 13. Lee, J. B., personal communication. 14. (a) Tebbe, F o N. ,' Parshall, G. W., and Reddy, G. S., J. Am. Chem. Soc. 100, 3611 (1978); (b) Tebbe, F. N., Parshall, G. W., """"' and Ovenall, D. W., ibid., !Ql, 5074 (1979); (c) Klabunde, U., Tebbe, F. N., Parshall, G. W., Harlow, R. L., J. Mol. Catal. - 8, 37 (1980). 15. The transformation of an allyl cation to a metallacyclobutane has been reported: Adam, G. J. A., I)avies, S. G., Ford, K. A., Ephritikhine, M. , and Green, M. L. H. , J. Mol. Catal. !!_, 15 (1980). 16. This mechanism has been studied in greater detail by Lee, J. B. for the monosubstituted titanacyclobutanes. Ba, c 76 References and Notes (continued) 17. (a) Rappe, A. K., and Goddard m, W. A., J. Am. Chem. Soc. 104, 297 (1982); (b) Rappe, A. K., Ph.D. Thesis, California Institute of Technology. 18. For example, the reaction of A1Me 2 Cl with Cp2 TiMe 2 produces Cp2 TiMeCl·AIMe3 on mixing. Chapter 1, this thesis. 19. (a) Carr, D. B., Schwartz, J., J. Am. Chem. Soc. .,..._,..._,.._ 101, 3521 . (1979); (b) Yoshida, T. , Negishi, E. , ibid. , W, 4985 (1981). 20. Reaction of E-i-d1 with 1 equiv. A1Me3 results in less than 5% isomerization in 20 min at room temperature. Ba, b 21. (a) Matteson, D. S., "Organometallic Reaction Mechanisms of the Nontransition Elements", Academic Press, New York, 1974, pp. 66-71; (b) Witanowski, M., Roberts, J. D., J. Am. Chem. Soc. ~' 737 (1966). 22. (a) Fraenkel, G., Dix, D. T., and Carlson, M., Tetrahedron Lett., 579 (1968); •(b) Eisch, J. J., J. Organomet. Chem. Rev. Ann. Surv. 4B, 314 (1968). """" 23. The reaction of i with A1Me 2 Cl• Et 2 O proceeds via first-order kinetics 24 with the same rate constant as the reaction of 2 with - diphenylacetylene. Ba, c This suggests that the rate determining step is the ring-opening to a metal-methylene complex which then reacts with A1Me 2 Cl• Et 2O to yield!. and Et2 O. This reaction is reversible. 24. Straus, D. A., Grubbs, R. H., unpublished results. 25. Gordon, A. J., and Ford, R. A., "The Chemist's Companion", Wiley-Interscience, New York, 1972, p. 303. 77 References and Notes (continued) 26. Van Horn, D. E., Negishi, E., J. Am. Chem. Soc. 100, 2252 """"' (1978). 27. The enhanced reactivity of the (3, (3-disubstituted titanacyclo- -- - -i butenes such as 4, 5, and 6 relative to 2 is most probably due to steric effects. Molecular structures of and Cp2 TiCH 2 CHPhCH 2 indicate that the (3-substituent (t-Bu or Ph) rocks into the "wedge" away from the adjacent Cp ring. .1, i, and .§_ do not have this option. As is typically the case in metal-olefin chemistry, although the disubstituted olefins are more electron rich, their steric bulk results in enhanced !ability of the metal-olefin interaction. 28. A possible mechanism for the triisohexylaluminum inversion was not discussed by Fraenkel. 22a The inversion displays first-order kinetics, ruling out an SE mechanism observed for Li, Mg, Zn, 2 21 etc. a A possible monomolecular mechanism involves an a-hydrogen migrationfroman isohexyl to the Al atom. This can be followed by rotation of the AC -c+ bond, migration of the aluminum hydride back to the a -carbonium center, resulting in inversion at the a-carbon atom. This does not account for the effect of less than stoichiometric amounts of Lewis base. Such a -hydrogen migrations are well known in early transition metal chemistry, having been observed in a tungsten system, 29 and crystallographically and spectroscopically in many of Schrock's tantalum alkylidene complexes. 30 An a -hydrogen migration has been proposed for aluminum. 31 78 References and Notes (continued) 29. Cooper, N. J., Green, M. L. H., J. Chem. Soc. Dalton, 1121 (1979). 30. (a) Schrock, R. R., Accts. Chem. Res. ,.,,,.._ 12, 98 (1979); (b) See also Schrock, et al., J. Am. Chem. Soc. 103, 160 (1981), ibid., ~ ill, 67 44 (1980), ibid. , W, 4599 (1981). 31. Ivin, K. J., Rooney, J. J., stevrart, C. D., Green, M. L. H., Mahtab, R., J. Chem. Soc. Chem. Comm., 604 (1978). 32. Kress, J., Wesolek, M., Le Ng, J. -P., Osborn, J. A., J. Chem. Soc. Chem. Comm., 1039 (1981). 33. Rappe, A. K., Goddard Ill, W. A., J. Am. Chem. Soc. "-'"'-"' 102, 5114 (1980). 34. Schrock, R. R., Rocklage, S., Wengrovius, J., Rupprecht, G., Fellmann, J., J. Mol. Catal. !!_, 73 (1980). 35. Chapter 3, this thesis. 36. Lee, J. B., Gajda, G. J., Schaefer, W. P., Hovrard, T. R., Ikariya, T., straus, D. A., and Grubbs, R.H., J. Am. Chem. Soc.~ 103, 7458 (1981). - 79 CHAPTER 3 The Preparation and Characterization of a Few M-µ-CH 2 -M (M = Ti, M' = Ti, Zr) Compounds and Their Reactions with Carbon Monoxide. 80 . INTRODUCTION ~ Transition metal methylenes have been implicated as intermediates in the olefin metathesis reaction 1 and the Fischer-Tropsch synthesis. 2 In both systems, metal methylene dimerizations are thought to be key steps. Attempts to synthesize metal alkylidene or carbene 3 complexes often results in the formation of olefins or metal-olefin complexes which arise from coupling of two alkylidene or carbene groups. 4 The coupling reactions are generally postulated to involve the bridging of alkylidene groups between two metals. Such a coupling reaction may be involved in a chain termination step of the olefin metathesis reaction. A stable bis-bridging methylene species has been isolated from an olefin metathesis system, and is the subject of this chapter. 5 Bridging methylene complexes also represent the simplest models of a surface-bound methylene. The chemistry of these species has been the subject of intense interest 6 due to the implication of the involvement of surface bound methylenes in Fischer-Tropsch chemistry. Oligomerization of methylene groups on metal surfaces to -(CH 2 )n- products and coupling of CO with µ-CH 2 groups to yield bridging ketene functionalities have been proposed to be major C-C bond forming processes in FischerTropsch synthesis. 2 The preparation, characterization, and CO coupling _ reactions of several mono- and bis-µ-CH - metal 7 species are reported herein. 2 81 RESULTS ~ Preparation and Reactivity of Bis-µ-CH 2 -[Cp2 Ti(IV)] 2 • The sterically encumbered ,9-,9-disubstituted titanacyclobutanes such as Cp2 TiCH2 CMe(i-Pr)CH2 (~) and Cp2 TiCH 2 CMe2 CH 2 (1_) decompose t0 liberate 2, 3-dimethyl-1-butene or isobutylene, an intensely permanganate-colored species. respectively, yielding Figure 1 shows a 1 H NMR - spectrum of a solution containing the products of decomposition of 2. 1 Downfield resonances at 08. 72 in the H spectrum and at t5235 (1JC-H = 125 Hz) in the 13 C spectrum were observed and are diagnostic for bridging methylenes 8 or metal-methylidene species. Infrared .spectroscopy revealed stretches at 2980 and 2920 cm- 1 and are assigned as -CH 2 - vibrations. Molecular weight determinations indicated that the intensely purple colored species is a dimer in solution. The reaction 1Me Cl (2 equiv~) producing several color changes before yielding chemistry of this dimer (1) follows. The dimer reacts readily with A 2 Cp2 TiCH 2 AlMe2 Cl in quantitative yield as determined by 1H NMR spectroscopy. Reaction with trifluoroacetic acid (TFAH) (0. 6 equiv.) at low temperature resulted in a near statistical distribution of Cp2 TiMe 2 , Cp2 TiMe(TFA), and Cp2 Ti(TFA) 2 • Compound! reacts only sluggishly in refluxing acetone to produce isobutylene. The above facts lead to the formulation of !. as bis- µ-CH 2 -[ Cp2 Ti(IV)] 2 • The reluctance of !. to react with acetone, and the lack of methylene exchange with deuterated olefins rules out!. as being formulated as [Cp2 Ti=CH 2 ] 2 • Other early transition metal alkylidene species react vigorously with organic carbonyl compounds, 1 O and other nucleophilic titanium methylene complexes exchange their methylene groups with terminal olefins. 11 8 7 1 - 6 C6D6 >00 5 PPM 4 3 [ Cp2TiCH2] 2 + 21-- 2 6 ~ 6 0 CH4 temperature to yield 1 and 2, 3-dimethyl-1-butene. - Figure 1. 89. 60 MHz H NMR spectrum of the decomposition of ~ after 2 hr in C D at room 9 2 Cp2Ti~ ~ CX) 83 - Compound 1 is surprisingly robust, decomposing only partially after several hours in refluxing toluene. The gaseous decomposition products are methane (92%) and ethane (8%). When ! was allowed to react with 12 , Cp2 T~ ( H NMR) was 1 formed, and ethylene (78%) was liberated along with methane (22%) (GC, NMR). - Photolysis of 1 at 30° C also resulted in decomposition to produce ethylene (84%; 80% hydrogenated by Cp Ti to ethane). 12 2 The rest of the hydrocarbon gases produced were CH 4 (15%) and a small amount of propylene. In an attempt to boost the propylene - yield, the photolysis of 1 was carried out under an ethylene atmosphere (ca. 1. 5 atm). After a few hours the photolysis was stopped. 1 - H NMR spectroscopy of the solution indicated that nearly all of 1 had decomposed. A new Cp resonance had grown in (,;5. 89) along with two broad multiplets (f31. 88, ~ 1. 28) which integrated 10: 4: 4. The ethylene resonance was broad and was shifted upfield from r5 5. 24 to (5 4. 64. These resonances were assigned to the titanacyclopentane Cp2 TiCH 2 (CH 2 ) 2 CH 2 , which exchanges slowly with free ethylene. 13 , 50 Methane and ethane were also observed. The photolysis was allowed to continue until ethylene was no longer detected, and the peaks t 1 assigned to Cp2 TiCH 2 (CH 2 ) 2 CH2 were absent. GC analysis of the gases above the black solution indicated that a mixture of mostly n-butane (27%), n-butenes (l{Y)k), ethane (53%) and CH 4 (9%) had been formed. Traces of ethylene, propane, propylene, and cyclopropane were detected. 84 Attempts were made to investigate the mechanism of formation of !_. Two possibilities for the formal dimerization are proposed. One involves direct dimerization of two Cp2 Ti=CH 2 fragments (Scheme 1, A). The other involves a nucleophilic attack of the methylene of Cp2 Ti=CH 2 at the titanium atom of a titanacyclobutane, displacing olefin and forming .!. (Scheme 1, B). The latter appears to be possible due to the fact that another species is observed in solution in low yield when concentrated solutions of i are allowed to decompose. This species has 1 H NMR resonances at o10. 46 (m, br) and O5. 73 (s, br) with relative integrals of 2 : 1 O. This species is tentatively formulated as the trimer [Cp2 TiCH 2 ] 3 • The signal assigned as the µ-CH 2 may be broadened by spin-spin coupling of protons which are •inequivalent due to the conformation of the TisC 3 ring. This species is proposed to arise from nucleophilic attack of Cp2 Ti=CH 2 on!. (Scheme 1, C). Assuming that the rate of trapping~ is much faster than the back reaction of Cp2 Ti=CH 2 , .both mechanisms predict the observed firstorder kinetics (kobs = 5 x 10-4 s- at 23 ° C) for the decomposition of i. 1 An experiment to distinguish the two possibilities (A and B) was designed. Decomposition of Cp2 TiCH 2 CMe(i-Pr)CD2 (~-~) to give a mixture of !_-d0 , 2 , 4 was carried out. The deuterium isotope effect on ring-opening of~-(½ is known to be 2. O. 14 With this information, one can predict the relative amounts of ethylenes produced for path A (4<io: 2~: ld4 ) and path B (2d0 : 3~: ld4 ; this assumes that there is no isotope effect on the reaction of Cp2 Ti=CH 2 with the titanacyclobutane). Unfortunately, the GC /MS data were irreproducible from injection to injection, and a spectrum containing only parent ions was not obtained, 85 Scheme A. 2 Cp 2Ti = CH 2 Cp2 TiCH2 1 • + olefin B. 2 C. 22 ► ◄ olefin + Cp2Ti = CH2 \iCp 2 86 resulting in unreliable ratios. - When 2 was allowed to decompose in the presence of E-Cp TiCHDCD(t-Bu)CH 2 -i ( ~ - ~ ) , 2 no deuterium was incorporated into 1, and therefore Cp Ti=CH preferentially inserts into the Ti-C bond of 2 2 rather than i-~. Trapping of 'Cp TiCH 2 With Other Reagents. ' 2 The {3-(3-disubstituted titanacyclobutanes are convenient sources of the 'Cp TiCH 2 ' moiety ( 'Cp TiCH 2 2 .either Cp Ti=CH 2 , 2 ' 2 designates a source of titanium methylene, or titanacyclobutane). With this in mind, 'Cp TiCH 2 ' 2 was trapped with reagents similar to A1Me Cl to generate new Lewis 2 acid stabilized metal methylenes. CpTiCls was chosen to be a good candidate. ~ or 1 readily reacted with CpTiC1 in C D to produce a 3 6 6 NMR spectrum with a new se! of singlet resonances a.-! r>~.88, 6.15, and 6. 08, integrating 2: 10: 5, respectively. Resonances due to the corresponding olefin were also observed. An orange solid began to precipitate as the NMR ~pectra broadened. A new set of broad singlet resonances grew in at o 7. 84 and 6.18, integrating 2: 15, respectively. The sample decomposed quickly to a brown-green solution. The first formed species was tentatively assigned as [C112 TiCll -µ-CH 2 -[CpTiC~] -- -- (11). The second formed species may be a dimer or oligomer of 11. Neither of these compounds were isolated due to their thermal lability. - A third, less stable species was observed when 4 was allowed to react with O. 5 equivalent of CpTiCis. This reaction rapidly yielded a deep violet species which exhibited 1 H NMR signals at o8. 68, 8. 29 (AB 87 quartet, J = 6. 5 Hz), 6. 78, s, 6. 07, s, and 5. 84, s, which integrated 2: 2: 5: 5: 5, respectively. A very broad resonance (ca. 200 Hz) was observed centered at o6. 0. No Cp2 TiC½ was detected. This species was tentatively assigned as [C1>2Ti(IV)]-bis-µ-CH 2 -[CpC1Ti(IV)] (ll) having two bridging methylenes between Cp2 Ti and CpTiCl centers. The color of this species and the position of the CH 2 resonances are very similar to that of the closely related .,.... 1. This species may arise as depicted in Scheme 5. It was anticipated that Cp*TiCls would yield a more robust adduct than CpTiCis. When .,....2 or .,.... 4 was allowed to react with Cp*TiCls, a rapid reaction occurred. A bright orange crystalline compound was isolated and exhibited 1 H NMR signals at o6. 61 (s), 6. 24 (s), and 1. 87 (s) in the ratio of 2: 10: 15. 13 C NMR spectroscopy revealed a resonance at o220 (t, JC-H = 116 Hz), in addition to Cp and Cp* resonances. The infrared spectrum of this compound exhibited stretches at 2960 and 29·20 cm -i. This compound is formulated as (Cp2 C1Ti(IV)-µ-CH 2 -[Cp*C½Ti(IV)] (.§_) containing a bridging methylene between Cp2 TiCl and Cp*TiC½ centers. Bright yelloworange C6 D6 or CD2 C¼ solutions of ~ darkened after 1 hr, and copious amounts of CP-J TiMeCl and other unidentified Cp* containing species were formed (1H NMR) . . ~ reacted slowly with acetone to produce isobutylene in contrast I i to Cp2 TiCH 2 AlMe 2 Cl, which reacts very vigorously. ~ did not react with 3, 3-dimethyl-1-butene in the presence of donor solvents such as pyridine or THF to yield Cp2TiCH 2 CH(t-Bu)CH 2 • No chemical 1 shift changes were observed in the H NMR spectrum when pyridine, 1 'c1 Cp2Ti 'c1) 1 12 1 ~ Cl TiCp2CI Cp2Ti--Tli CpCI Cp2TiCH2 Cl Cp21T'~TiCpCI I"-..,/ 11 'c1 I ~ Cp2Ti-----T i Cp Cl • Cp 2Ti -------..Ti Cp CI CpTi CH2 + C pTi Cl3 1 Scheme 5 CX> CX> 89 THF, or PEt3 was added to solutions of §_. ------ The reaction of Cp*TiCla with Cp2 TiCH 2 CH(t-Bu)CH2 is slow at 50e C in C6 D6 , and§_ decomposes under these conditions. No reaction occurs between 1 and Cp*TiCls at 60c C in C6 D6 • The compound [Cp2 C1Ti(IV)] -µ-CH 2 --[Cp*C½Zr(IV)l (§) was isolated as maroon crystals in a manner completely analogous to that - - employed for the isolation of 5. Deep maroon solutions of 6 are -- surprisingly much less thermally stable than 5. 6 is also far less soluble being nearly insoluble in C6 D6 • This property suggests that ~ 1 is oligomeric. CD2 C½ solutions of~ exhibit a H NMR signal at r, 9. 43 along with other resonances assigned to the Cp and Cp* ligands. !!. decomposes quite rapidly in CD2 C½ at room temperature to yield an orange solution containing a large amount of CP2 TiMeCl and other unidentified Cp* containing products (1H NMR). Addition of THF or PE"ta to C6 D6 suspensions of ~ does not solubilize the complex. A wide variety of other reagents were tried as traps for the 'Cp2 TiCH 2 ' moiety. Among them were (PPh3 ) 3 RhCl, CpFe(CO\I, C~ScCl, Me 3 SnCl, Me 2 SnC¼, Me3 SiCl, and CH3 I. Cp:ScCl, Me3 SiCl, - - - and Mel did not react with 2 or 4. The dimer 1 was obtained in moderate yield from these mixtures. The product from the reaction of 1: with Me SnCl was isolated in good yield and was identified as 3 Cp2 TiCH2 SnMe3 (Cl}. 15 Reaction with Me2 SnC½ yielded an analogous product. Phosphines trap 'CpgTiCH2' as the yellow compounds Cp2 Ti(=CH 2 )PRa. These compounds exhibit identical NMR spectra to complexes which ' 2 AlMe 2 Cl ' Schwartz has generated in solution by reaction of Cp2 TiCH with phosphines in the presence of hexamethylphosphoramide. 16 90 - ~ reacted with 1 equiv. PE4 to produce an equilibrium mixture of ~ 1 and Cp2 Ti( =CH 2 )PE4 (~) in the ratio of ca. 5 : 95 ( H NMR). Addition of 1 equivalent of PEt3 to 1 resulted in an equilibrium mixture of 1 and - 9 in the ratio of 70: 30. Addition of 3, 3-dimethyl-1-butene to either of these mixtures resulted in quantitative conversion to Cp2 TiCH 2 CH(t-Bu)CH 2 • A C6 D6 solution of the phosphine methylene ~ decomposed slowly - at room temperature to yield 1 in moderate yield ( 1 H NMR). In contrast - to the decomposition of 2 in the absence of phosphines, methane and a significant amount of ethylene were also produced (GC, NMR). The phosphine-methylene complexes were isolated as deep-yellow oils which when triturated with pentane gave fluffy yellow powders. In this manner Cp2 Ti( =CH 2 )PMe 2 Ph (!Q) was isolated in good yield. The phosphine-methylene species have low-field CH 2 resonances at ca. ('j 12. 5 which are coupled to the phosphorous nuclei with JP-H = 6-7 Hz, in agreement with Schwartz' observations. React ion of 1, 5, and 9 with CO. A C6 D6 solution of !. was sealed under an atmosphere of CO in a NMR tube. A new set of resonances grew in at ~ 5. 71 (s), 3. 73 (d, J = 1. 5 Hz), and 3. 65 (d, J = 1. 5 Hz) which integrated 10: 1: 1. This product is tentatively assigned as bis-77 2 -µ-(0, Cl)-OC=CH 2 -bis[Cp2 Ti(IV)] (ID (Fig. 2). ~ decomposed slowly (hours), until only Cp2 Ti(CO) 2 was observed in 60% yield (1H NMR). No other resonances were observed. Acidolysis of fresh or decomposed solutions with trifluoroacetic acid or anhydrous HCl produced several Cp-containing species ( 1 H NMR). No fragments which might have arisen from the 91 7 8 Fig. 2 92 µ -CH 2 CO moiety were detected. Reaction of CO with §_ produced a single product in 80-90% yield ( 1 H NMR) which is similar to !!• This new species is assigned as [Cp2 ClTi(IV)]-77 2 -µ-(0, Cl)-OC==CH 2 -[Cp*C½Ti(IV)] (7) (Fig. 2) based -"! 1 on the following data. The H NMR spectrum in CD2 C½ at -20" C revealed two singlets at o4. 69 and 4. 30 which integrated to 1 proton each with respect to a 10 proton singlet at o6. 41 and a 15 proton singlet at o2. 25. T.he 13 C spectrum of the same sample exhibited a double doublet (J = 10 Hz) at o215 along with another double doublet (JCH =155 Hz} at o132. These resonances are assigned as Cl and C2 of the 1 2 µ-OC =C H2 moiety, respectively. An identical product is obtained upon reaction of Cp Ti(77 2 -CH CO) 17 with Cp*TiC1 • 2 2 3 Irradiation of the Cp* protons of 1 resulted in NOE enhancement of the downfield vinylic resonance at o4. 69. Irradiation of the Cp protons did not result in any enhancement of either vinylic resonance. These NOE results suggest that the Cp2 Ti fragment is bonded to the oxygen of the OC 1 =C 2 H2 moiety, and the Cp*TiC¼ fragment is bonded to Cl. Attempts to alkylate or hydrogenate 1 were unsuccessful. - - 1 and 5 reacted rapidly at low temperature with CH 3 NC (1 equiv.) 1 to yield products which did not exhibit interpretable H NMR spectra. Reaction of a maroon solution of ~ in CD2 C¼ with CO resulted in 1 a rapid color change to deep orange. The H NMR spectrum of this mixture revealed only broad resonances in the Cp and Cp* regions. - No species analogous to 7 were identified. 93 Reaction of the phosphine-methylene complex ~ with CO yielded a mixture of products. The 1 H NMR spectrum of this mixture indicated that Cp2 Ti(CO) 2 had been formed in ca. 30% yield. Another species with a Cp resonance at o5. 66 (10H) and 2 singlets at o4. 84 (lH) and o3. 88 (lH) was identified as Cp2 Ti(CH2 CO) by direct comparison with the spectrum of an authentic sample. 30 This complex was present in 20% yield along with three or four other unidentified Cp-containing species. 94 DISCUSSION ~ - Physical and chemical characterization of 1 suggest that the compound is a bis-µ-methylene complex which arises from formal head to tail dimerization of two Cp2 TiCH 2 fragments. Steric and electronic factors favor a 1, 3- rather than a 1, 2-dititanaeyclobutane - in this system. The structure of 1 is very likely to be similar to the bis-µ -SiH2 analogue [CP2 TiSiH 2 l 2 , i5olated from the reaction of 18 . Cp2 TiC~ with KSiH 3 and structurally characterized by X-ray analysis. The 'Cp2 TiCH 2 ' dimerization is reminiscent of the thermal decomposition of Cp2 Ta(CH2 )Me which yields Cp2 Ta(C 2 H4 )Me and Cp TaMe(L) in the presence of a trapping ligand L. 4b This reaction is 2 postulated to proceed through a 1, 3-ditantalacyclobutane (Scheme 2) which cleaves to yield the tantalum ethylene product. The titanium - dimer 1 prefers not to thermally cleave to yield ethylene which would require the extrusion of the high energy Cp2 Ti fragment. When a reasonable leaving group is provided in place of Cp2 Ti, !. does react to yield ethylene as the major decomposition gas. This is the case when !. is allowed to react with 12 yielding CP2 Til2 and C2H4 • The course of this reaction is likely to be similar to that postulated for the reaction of titanacyclobutanes with 12 to yield Cp2 Til2 and cyclopropanes (Scheme 3) . 19 The reaction of the titanacyclopropane intermediate with excess iodine would produce the observed ethylene and Cp2 Til2 • Thermally, !. does not incorporate deuterium from labelled terminal olefins. The decomposition of titanacyclobutanes to !. therefore represents a chain termination reaction in this olefin metathesis system. 95 Scheme 2 L 96 Scheme 3 + 97 - The mechanism by which 1 is formed (Path A or B, Scheme 1) _could not be determined by examination of the kinetics. The crossover - experiment designed to examine the deuterium incorporation into 1 failed due to instrumental limitations. Based on the chemistry of the titanacyclobutanes with ambiphilic reagents such as A1Me2 Cl, mechanism B appears more reasonable. The formation of [Cp2 TiCH 2 ] 3 also supports this mechanism. The lack of detectable crossover between~ and~-~ is due to a difference of 2-3 kcal-moi- 1 for metallacycle cleavage between the mono and {3, ,B-disubstituted titanacyclobutanes. 49 Mechanism A, on the other hand, involves what appears to be a highly improbable event - the dimerization of two Cp2 Ti=CH 2 units which are in very low concentr~tion. It is doubtful that the lifetime of Cp2 Ti=CH 2 is sufficient to achieve this trapping reaction. The bis-µ-methylene !. has been proposed to be an intermediate in the well studied, but poorly understood, photo- and thermal decomposition of Cp TiMe • 20 . Methane, traces of ethylene and ethane, and 2 2 Cp2 Ti species are the only products of this decomposition. The proposed mechanism involves a -hydrogen abstraction by one methyl group, producing CH 4 and Cp2 Ti=CH 2 • This fragment subsequently abstracts hydrogen from the Cp rings, or, as Rausch 20 suggested, dimerizes to 1, which then decomposes to yield methane and Cp2 Ti. Since decomposition of 1 produces significant amounts of ethane and ethylene photo- - chemically or thermally, 1 can be ruled out as a possible intermediate in the decomposition of Cp2 TiMe 2 • 98 propylene. It was thought that the propylene could arise from the Photochemical decomposition of 1 produced a small amount of ethylene produced on decomposition inserting into a Ti-C bond of unreacted!_. This would give an 77 2 -µ-(Cl, C3)-propyl species, which could then undergo /3-hydrogen elimination to yield propylene. Such a transformation has been observed in an 77 2 -µ-(Cl, C3)-propyl-dicobalt compound. 21 A closer examination of the products formed when .!_ was photolysed under an atmosphere of ethylene suggests that the propylene arose from cleavage of !. to CP2 Ti=CH 2 • Cp2 Ti=CH2 can be trapped r i with ethylene to yield Cp2 TiCH 2 CH 2 CH 2 which is known to react further with ethylene, extruding propylene. 13 A small amount of cyclopropane was observed which is also indicative of formation of a small amount of Cp TiCH CH CH • 22 The photodecomposition of !. under ethylene 2 2 2 2 produced significant amounts of 1-butene and butane. The C4 products arise from catalytic dimerization of ethylene by Cp Ti12 which is 2 formed by one of several pathways (Scheme 4). The intermediate metallacyclopentane was observed by 1 H NMR spectroscopy. /3-Hydrogen elimination in the metallacyclopentane yields !-butene and regenerates Cp2 Ti. A 11 of the olefinic products are slowly hydrogenated to saturated hydrocarbons by Cp Ti. 12 2 It was found that not only can 'Cp2 TiCH 2 ' dimerize (and perhaps trimerize), but it can also be trapped by a variety of ambiphilic reagents,52 anC:I phosphines. The reaction of 'Cp TiCH ' with an M-X 2 2 (M = metal, metalloid; X = halide) bond very likely proceeds through a four-centered transition state. This is similar to the insertion of XCH 2 ZnX into Si, Sn, Pb, Hg and Ge halide bonds to produce M-CH~. 23 ~4 '- Scheme 4 ~~ ~ Cp2Ti • ( " C2H4 . hv C ;r 1 P2 • r .. ~ C2H4 hv :J T Cp2Ti 1 Cp2Ti◊ !C2H4 - Cp2Ti = CH2 hv -~ - CH4 c.o c.o 100 The compounds resulting from insertion of 'Cp2 TiCH 2' into Cp*TiCl~H Cp*ZrC¼, and Me3 SnCl were isolated. 'Cp2 TiCH 2 ' was not found to insert into Cp: ScCl, probably due to the tremendc:us steric limitation imposed by the Cp* ligands. 'Cp2 TiCH 2 ' was also not observed to insert into C-X and Si-X bonds, most likely due to the reluctance of silicon and particularly carbon to undergo frontside displacement in the proposed four-centered transition state. 'Cp2 TiCH 2 ' was also trapped as phosphine adducts which are analogous to the zirconium analogues reported by Schwartz. 16 b The Cp2 Ti( =CH 2 )(PRa) complexes react with olefins to yield equilibrium mixtures of titanacyclobutanes and the phosphine methylene complex. The position of equilibrium depends upon the steric natures of the olefin and probably the phosphine, with bulky olefins displacing the equilibrium towards the phosphine methylene complex. The thermal decomposition of the phosphine methylene complexes produces ! in moderate-yields, and also produces a small amount of ethylene. Ethylene is not observed on decomposition of titanacyclobutanes~ 1 and therefore the phosphine must be affecting the mode of dimerization of the 'Cp2 TiCH 2 ' units. It is proposed that some of the time, the phosphine methylene complex dimerizes in a head-to-head arrangement to produce an TJ 2 - µ-C 2H4 ligand which is readily lost as _ethylene (Fig. 3). Such an arrangement is postulated to be responsible 48 for the formation of ethylene from the reaction of Cp2 TiC½ with AlEta. .The similar reaction of Cp2 ZrC½ with AlEta yields a TJ2 -µ-C 2 H4 complex which has been structurally characterized. 101 CH2 Cp 2Ti ~ '-PR 3 1-PR3 1 Fig. 3 102 Reactions of the Methylene Compounds with CO. Reaction of the µ-methylene compounds bis-µ-CH 2 -[Cp2 TiCH 2 ] 2 and [Cp2 TiCl]-µ-CH 2 -[Cp*TiC½ l with CO resulted in the formation of unstable complexes described as having bis- or mono- bridging ketene (or as a -metalated enolate) ligands. These ligands are proposed to arise from the rearrangement of intermediate acyl complexes. Examples of migratory insertions of CO into metal carbon bonds to yield metal acyl complexes are well known for most d and some f transition elements. 25 Early transition metal acyls having fewer than 18 valence electrons and which are highly oxophilic 26 tend to exhibit ,-,2-bonding of the acyl via interaction of the acyl oxygen lone pairs with the metal atom. 27 This ,,2-acyl bonding mode manifests itself in adding considerable carbene character to the metal-carbon bond. 28 These features result in rapid rearrangement of the intermediate 772 -acyl to - a bridging ketene as is illustrated for the reaction of 5 with CO (Scheme 6). A .similar rearrangement has been observed for cp;Th(COCH TMS) (Cl), 29 which when heated rearranges to 2 cp;ThCl(OC(TMS)==CH 2 ). Interestingly, only one of two possible - products is observed when 5 is allowed to react with CO. Either CO preferentially coordinates and inserts into only one of the two possible Ti-C bonds, or insertion occurs at either Ti-C bond with one insertion being reversible. The 'hard' Lewis acid character of the Cp*TiC½ center relative to the 'softer' Cp2 TiCl center would lead to preferential coordination and insertion of CO into the Cp2 Ti-C bond. This would result ultimately in the formation of the observed Cp2 Ti-O arrangement (Scheme 6). 'c1 Cl H ..,_ c,., 0 7 'c1 H-XH 0 l,2shifl Cp2T( Ti Cp*c1 2 H 0'-Ti Cp*Cl2 ► Cp 2Tix0-TiCp*Cl2 ◄ ► Cp2-T.v:"f \ 1,2 shift 'ci /?~TiCp*Cl2 ◄ ► Cp2Ti / OCATiCp*Cl2 Cp2 T, I / co\ 5 co/ 'c1 Cp2 Ti ~~-c T1 p*c1 2 -0 Scheme 6 104 The insertion of Cp*TiCls into C~ Ti(77 2 -OCCH 2 ) also yields 'l· The 77 2 -ketene is thought to be in equilibrium via its O, C and C, C bound isomers. 30 Insertion of Cp*TiCls into the Ti-C bond of either - isomer would result in 7 (Scheme 7). The reaction chemistry of the µ-ketene (or a-metalated enolate) ligand remains undefined. Attempted alcoholysis, protonolysis, alkylation, and hydrogenolysis failed. Preliminary results indicate that §_ does react with ketones and aldehydes to give organometallic aldol condensation products. 31 There is little direct evidence of CO insertion into other compounds containing bridging methylenes, even though most bridging methylenes known at present contain CO as ligands. Keim 32 has reported that µ-CH 2 -Fe 2 (CO) 8 reacts to form methyl acetate in the presence of methanol via the postulated formation of an 77 2 -µ(Cl, C2)-CH C=O moiety. Shapley and Geoffroy 33 have character2 ized a triosmium cluster compound which contains a ketenylidene ligand. This compound arises from thermolysis of Os3 (CO)i 1 (µ-CH 2 ) involving an apparent carbonylation of the µ-methylene ligand. A transformation which may involve the reverse of the reactions 4 described above is the formation of µ-CH 2 -µ-CO-[Cp*CrJ? which was isolated from a mixture of CrC~, Cp*H, and the lithium enolate of acetaldehyde which was generated accidentally in situ from THF and n-BuLi. A µ-CH 2 CO species may be involved in this reaction, which subsequently decarb0nylates to yield the observed product. Carbon monoxide was also observed to react with the terminal methylene phosphine complexes to give a low yield of Cp2 Ti(772 -CH 2 CO) 30 Cp 2Ti::f 0 ~Cp*TiCl 3 Cp2Ti1 0 Scheme 7 CpTi ----- 7 ► 'c1 )lv Ticp*c1 2 1,2 shift 7 0 CJ'1 ... 0 106 which arises from the coupling of the methylene with CO, apparently via an unobserved Cp2 Ti(=CH2 )(CO) complex. Similar CO insertions into terminal metal methylenes have been observed for the Mn=CPh2 , 35 Mo=Mo=CPh , 36 Mn(anthronyl), 37 and W(=C(OMe)Ph) 38 complexes 2 yielding coordinated or free ketenes. The /3, {3-disubstituted titanacyclobutanes have been shown to be good sources of 'Cp2 TiCH2 ' for the syntheses of µ-CH 2 complexes, which are precursors of 17 2 -µ-CH 2 CO complexes. Clearly, the potential for further study in this area exists. The reactivity of 'Cp2 TiCH 2 ' with other electrophilic or ambiphilic organometallic reagents is yet to be explored. 'Cp2 TiCH 2 ' may provide interesting organometallic chemistry via reaction with organometallic substrates such as CpFe(C0) 2 CHR+ or CpW(C0) 3 (C 2 H4 )+. 107 EXPERIMENTAL ~ 2, 3-Dimethyl-1-butene-1, 1-(½ and 3, 3-dimethyl-1-butene-1, 1-(½ were prepared by carboalwnination39 , 41 of i-propylacetylene-d and 1 hydrozirconation4 o, 41 of t-butylacetylene-di, respectively, and dried over Linde 4A molecular sieves. The titanacyclobutanes Cp2 TiCH 2 CMe(i-Pr)CH 2 (~), I ' CP2TiCH 2 CMe(iPr)CD2 (i-(½), Cp2 TiCH 2 CMe 2 CH 2 (f), f E-Cp2 iCHDCD(t-Bu)tH 2 (~-(½), and Cp2 TiCH~ 1Me2 Cl were prepared as previously described. 41 CpTiCls, 42 Cp*TiCis 43 and Cp*ZrCis 44 were prepared by literature methods. cµ;scCl was a gift from Mr. M. E. Thompson. A1Me2 Cl (Texas Alkyls) was used as received. Trifluoroacetic acid, acetone, methyl iodide and chlorotrimethylsilane were dried over 4A. molecular sieves prior to use. Ethylene, carbon monoxide, and hydrogen were used as received. Triethylphosphine and dimethylphenylphosphine (strem) were used as received. Methyl isocyanide was a gift from Prof. J. E. Bercaw. THF-~, C6 D6 , and toluene-~ were vacuum transferred from sodium-benzophenone ketyl. CD2 C~ was stored over CaH 2 and degassed by several freeze-pump-thaw cycles. Aromatic hydrocarbon solvents were pre-dried over CaH2 and vacuum transferred onto sodium benzophenone ketyl. Aliphatic hydrocarbon solvents were stirred over concentrated H2 E04 , dried over CaH 2 , and vacuum transferred onto sodium benzophenone ketyl containing a small amount of tetraglyme. Solvents thus dried and deoxy- 108 genated were vacuum transferred into dry vessels equipped with teflon needle valve closures and stored under Ar. General Procedures All manipulations of air and/or moisture sensitive compounds were carried out using standard high vacuum, Schlenk line, and dry box techniques. Argon used in Schlenk work was purified by passage through columns of BASF RS-11 (Chemalog) and Linde 4A molecular sieves. NMR spectra were recorded with a Varian EM-390, JEOL FX-90Q, or Bruker WM-500. Kinetics by 1H NMR spectroscopy were obtained in automated mode on the JEOL FX-90Q. Temperatures were calculated by measuring AvMeOH 45 and were constant to within± 0.1 • C. Difference NOE's were measured on the Bruker WM-500. Infrared spectra were obtained on a Beckman IR 4240 spectrophotometer. Analyses were performed by Schwarzkopf Analytical, Inc. The molecular weight of [Cp2 TiCH2 ] 2 was measured cryoscopically in C6 H6 by a standard procedure, and also by Signer's method. 46 Preparation of bis-µ-CH 2 -bis(Cp2 TiIV) OJ• A solution of Cp TiCH AIMe c1 47 (1. 0 g, 3. 5 mmole) in CH C~ (3 mL) was slowly 2 2 2 2 added via cannula to a suspension of N, N-dimethylaminopyridine (472 mg, 3. 9 mmole) and 2, 3-dimethyl-1-butene (2 mL, 16. 2 mmole) in CH 2 C~ (3 mL) at -20° C under Ar. The resulting homogeneous red solution was slowly added via cannula to 30 mL of vigorously stirred pentane at -20° C under Ar. The resulting DMAP-A1Me2 Cl precipitate was rapidly filtered. The solvent was removed from the resulting clear - red filtrate in vacuo. The resulting red powder (2) was dissolved in 109 toluene (8 mL) at -10• C, and the solution was allowed to warm to room tenp erature. After standing at room temperature for 18 hr, the purple solution was slowly cooled to -10° C. The mother liquor was removed from the resulting purple platelets via cannula, and the crystalline mass was washed with cold toluene (1 mL) followed by n-hexane (3 x 3 mL). The purple crystals of 1 were dried in vacuo to yield "" ~ 790 mg (59%). Calculated for TiC 11H12 : C, 68. 77; H, 6. 30. Found: C, 68. 88; H, 6. 55. 1 H NMR, C6 D6 , shifts relative to C6 D5H at ~ 7.15: 5 8. 72 (CH 2 ), 6. 22 (Cp). 13 C, CD2 C½, shifts relative to CD2 C~ at <'5 53. 8: ~ 235. 2, t,. J = 126 Hz (CH 2 ), 112. 2, d, J = 174 Hz (Cp). m, C6 D6 solution: 3100, (w, br), 2980, 2920 (m, sh), 1445 (w, br), 1020 (s, sh), 810 (w, br), 790 (s, br), 435 (s, sh). Molecular weight: Calculated for [CP2 TiCH 2 ] 2 384; found 396 ± 30 cryoscopic in C6H6 ); 427 ± 20 (Signer's method). 46 The visible spectrum exhibits a fairly intense, symmetrical absorption at 530 nm. Reaction of!. with. A1Me 2 Cl~ A1Me 2 Cl (20 µL, 0. 2 mmole) was added via syringe to a solution of 1 (20 mg, 0. 05 mmole) in C6 D6 (400 µ L) in a 5 mm NMR tube. The solution changed in color from purple to brown to green to red in about 15 seconds. 1 H NMR spectros- copy revealed that the reaction had proceeded cleanly to yield i I CP2TiCH 2 AlMe 2 Cl and a small amount of Cp2 TiCH 2 AlMeC1Cl. Reaction of 1 with 3, 3-Dimethyl-1-butene-1, 1-(½. 3, 3-dimethyl- "" l-butene-1, 1-(½ (2 µL, 0. 02 mmole) was added via syringe to a solution of !. (15 mg, 0. 04 mmole) in C6 D6 ( 400 µ L). The mixture was monitored by 1 H NMR spectroscopy. After 2 days at 55 ° C, no 3, 3-dimethyl-1- 110 butene-d 2 was detected, and the CH 2 to Cp ratio of !. was 2 : 1 O. - Protonolysis of 1. A sample of 1 (1. 7 mg, O. 044 mmole) was ---------~ placed into solution with 400 µL C6 D6 in an NMR tube. The sample tube was capped with a septum, and removed from the dry box. The sample was cooled to -78"C, and trifluoroacetic acid (4 µL, 0.052 mmole) was added via syringe. The sample was agitated while maintained below -50" C. The 1 H NMR spectrum at -30" C indicated that Cp2 TiMe 2 , Cp TiMe(TF A), and Cp Ti(TF A ) 9 were formed in the ratio of 1 : 1.1 : 1. 7. 2 2 - 2 Excess 1 was also observed. Thermolysis of!.• A toluene-~ solution of 1 (10 mg in 400 µL) was prepared in an NMR tube and sealed. The sample was placed in a 1 bath maintained at 100" C for 12 hr. At this point H NMR spectroscopy o indicated that a trace of 1 remained. Two other signals were observed at oO. 79 (C 2 H6 ) and 0.15 (CH 4 ). The tube was opened under N2 , and a sample for gas chromatography was withdrawn. GC analysis (Poropak Q, 6', 110° C, FID) indicated that the gas phase contained a mixture of CH 4 (92%), C~ 6 (8%), and a trace ( < 0. 2%) of C2 H4 • Reaction of 1 with Acetone. A (CD3 ) 2 CO solution of !. (12 mg in 400 µ L) was prepared in an NMR tube, and the tube was sealed. There was no apparent reaction after 15 min at ambient temperature as judged by 1 H NMR spectroscopy. The sample was then heated to 40° C for 35 min. No appreciable reaction had occurred at this time. After heating to 50° C for 15 min, some isobutylene was detected (5 4. 68). After 12 hr at 5o•c, reaction had occurred to yield pale green solids and a pale yellow solution. Broad signals were observed inthe 1H NMR spectrum. 111 GC indicated that the gas phase contained CH 4 (93%), CJI 6 (1%), isobutylene (5%) and C2H 4 (trace). Photolysis of 1. A solution of ! (12 mg) in C6 D6 ( 400 µ L) was prepared in a pyrex NMR tube and the tube was sealed. The sample was placed in a photolysis well (Hanovia 450 W medium pressure Hg lamp) approximately 20 mm from the source. The sample was maintained at below 30° C during photolysis. After 12 hr of photolysis, the 1 solution was purplish black, and H NMR spectroscopy indicated most of ! had decomposed. The tube was opened under an N2 atmosphere, and a gas sample was withdrawn. GC of the gases indicated a mixture of CH 4 (16%), C2H 4 (3%), C2H 6 (80%), and C3H 8 (1%). Photolysis of 1 in the Presence of C2H 4 • An NMR tube was - loaded with 1 (15 mg). An 'O'-ring adapter and reflon needle valve adapter were attached to the NMR tube, and the tube was removed to the vacuum line. C6D 6 was vacuum transferred into the tube at -78 °C. The tube was vented to '700 mm Hg of C2H 4 and sealed. The sample was placed in the photolysis well ca. 20 mm from the source. The sample was maintained at 18 ° C during photolysis. After 10 hr of photolysis, the sample was removed. 1 H NMR spectroscopy indicated - that most of 1 had decomposed. A peak at o 5. 89 had grown in along with 2 broad multiplets at o1. 88 and o1. 28 and integrated ca. 10 : 4 : 4. An intense singlet at o0. 79 (C2 H 6), a smaller singlet at ~ 0. 15 (CH 4 ), and a broad peak at o 4. 64 (C 2H 4 ) was observed. After 30 hr of photolysis, the peaks at o5. 89, o1. 88, o1. 28, and o4. 64 had disappeared, while the singlets at a0. 79 (CJI 6 ) and o0.15 (CH 4 ) remained. The 112 solution was greenish-black in color. The tube was opened, and a gas sample was withdrawn. GC analysis indicated that the gas was a mixture of CH 4 (9%), C2H 4 (<1%), C2H 6 (53%), C3H 8 , C3H 6 (<1%), C4H 8 (10%), and C 4H 10 (27%), along with a trace of cyclopropane. Reaction of 1 with Iodine. A C6D6 solution of 1 (12 mg in 400 µL) was prepared in an NMR tube. The tube was capped with a latex septum and removed from the box. A saturated solution of½ (sublimed) in 20 µL C 6 D6 was added to the sample. An immediate reaction occurred, and dark solids precipitated. 1 H NMR spectroscopy indicated that Cp2 Ti~ had been forme~ (5 6. 03) along with C2H 4 (~ 5. 24) and CH 4 (o 0.15). GC of the gas phase indicated that CH 4 (22%) and CJl 4 (78%) were present. Generation of 1-d0 2 4 Followed by Reaction with Iodine. A 10 mL round-bottom flask was loaded with a stirbar and i-(½ (20 mg, 0. 072 mmole) and attached to a teflon needle valve adapter. Approximately 1 mL of p-xylene was vacuum transferred onto the titancyclobutane at -20°. • The reaction vessel was opened to a cold trap to remove the olefin as it was formed, and the sample was warmed to room temperature. The p-xylene was replenished at regular intervals. The reaction was allowed to proceed for 3. 5 hr. The p-xylene was then removed in vacuo. The purple residue was dissolved in C6 D6 • 1 H NMR - spectroscopy indicated the presence of 1. No titanacyclobutane remained. A saturated C 6D 6 solution of ¼ (200 µL) was added to the NMR sample. 1 H NMR indicated that ethylene had been formed. GC /MS of the gas phase revealed the following mass spectrum for the ethylenes: Measured mass (% intensity base) 33 (0. 3), 32 (21. 4), 113 31 (3.1), 30 (74. 0), 29 (29. 9), 28 (100. 0), 27 (66. 2), 26 (47. 7), 25 (8. 9), 24 (3. 9). Several sample injections were made and gave widely varying relative intensities at the high and low ends of the mass envelope. Reaction of 2 with E-Cp2 TiCHDCD(t-Bu)CH 2 (3-~). An NMR tube was loaded with ~ (20 mg, O. 072 mmole) and ~-~ (20 mg, 0. 072 mmole). CJI 6 was vacuum transferred into the NMR tube at -78 ° C, and the tube was sealed. The sample was shaken vigorously as the solvent thawed. The sample was allowed to react at room temperature for 2 hr. 2 H NMR spectroscopy revealed that no !_-d1 was formed, and that approximately 5% at z-~-~ had formed from the remaining E-~-~- - - Kinetics of the Reaction of 2 to Form 1. An NMR tube was loaded with~ (20 mg, O. 072 mmole) and capped with a latex septum. Toluene-~ was slowly added to the chilled (-50° C) NMR tube, and · dissolved at that temperature. Anisole (1 µL) was then added as an internal standard. The sample was placed in an NMR probe at 23" C. The rate of reaction was monitored by comparison of the Cp resonances - of 2 versus the anisole methyl resonance. First-order plots were linear to three half-lives, and the slope yielded the first-order rate constant k = 5 x 10- 4 s- 1 • Preparation of µ-CH 2 -[Cp2 TiCl][Cp*TiC12 ] (5). A small Schlenk tube was loaded with Cp2TiCH2 CMe2 CH 2 (1_) (200 mg, 0. 81 mmole) and Cp*TiCls (233 mg, 0. 81 mmole). The reaction vessel was cooled to -20° C, and toluene (4 mL) was added slowly. The reaction mixture was warmed to O" C while stirring. Approximately 30 min were required for complete dissolution of the Cp*TiCls. After stirring the mixture for 114 1 hr at 0° C, the dark yellow-orange solution was warmed to room temperature. Pentane (10 mL) was added slowly to the stirred toluene - solution to precipitate red-orange crystalline 5. This material was washed with a small amount of cold toluene followed by pentane (3 x 1 mL). The red-orange crystals were dried in vacuo .to yield 210-310 mg (54-80%). This material may be further purified by recrystallization from CH 2 C½/ petroleum ether mixtures as orange cubic crystals. 1 H NMR, C6 D6 , shifts relative to C6 D5H at r> 7 .15: o6. 61, s (CH 2 ), 6. 24, s (Cp), o1. 87, s (Cp*). 13 C, C 6 D6 , shifts relative to ~D6 at ~ 128. O, gated decoupled: o219. 6, t, J = 116Hz (CH 2 ), 129. 6, s (QsMe 5), 117. 0, d, J = 164 Hz (Cp), 13.4, q, J = 127 Hz (C ¥e m, CH C½ solution (cm- 2960, 5 1 5 ). 2 ): 2920 (m, sh), 1485, 1445 (m, sh), 1380 (s, sh), 1020 (s, sh), 825 (vs, sh), 770 (m, br), 465 (w, sh), 415 (m, br). Attempted elemental analysis gave unsatisfactory results. Calculated for C 21H27T¼Cls: C, 52. 38; H, 5. 65; Cl, 22. 08; Ti, 19. 89. Found: C, 43. 34; H, 4. 38; Cl, 19. 71; Ti, 19. 55. Preparation of µ-CH 2 -[Cp2 TiCll[Cp*ZrC½l (~) . .1 (200 mg, 0. 81 mmole) and Cp*ZrC13 (300 mg, 0. 9 mmole) were allowed to react at oe C in toluene (6 mL) for 1 hr. The resulting maroon microcrystalline powder was filtered and washed repeatedly with cold toluene. The - solids were dried in vacuo to yield 260 mg (61%) of 6. This material 1 was too insoluble in C6 D6 to obtain a H NMR spectrum. It is somewhat soluble in CD2 C½ and THF-~, but decomposed fairly rapidly to yield Cp2 TiMeCl and unidentified Cp* containing products (1 H NMR). 1 H NMR, CD2 C½, shifts relative to C!!DC½ at 6 5. 32: o9. 43, s (CH 2 ), 6. 33, 115 s (Cp), 2. 07, s (Cp*). The material from above was dissolved in CH 2 C¼ ( 8 mL) at oe C, filtered, concentrated to ca. 2 mL, and cooled to -50e C. Maroon crystals deposited from the orange solution. The crystals were washed with cold toluene and dried in vacuo to yield 25 mg (5%). Calculated for C21H27 C½TiZr: C, 48.05; H, 5.18. Found:C, 48.07; H, 5.07. IR, fluorolubemull, (cm- 1): 3110(w,sh), 2980, 2960 (m, sh), 2800 (m, sh), 1480 (m, sh), 1430 (s, br), 1380 (s, sh). - - Reactions of 1, 5 and 6. The reactivity of 5 and 6 was investi- ------------- gated using NMR tube scale reactions. A typical reaction with CO is described here. A 5 mm NMR tube was loaded with §_ (10 mg, 0. 021 mmole). 0 -ring and needle valve adapters were attached to the tube. A deuterated solvent was vacuum transferred into the tube at -196 e C. The tube was vented to 300 torr CO, and the tube sealed. The compound [Cp2 ClTi(IV)]-77 2 -µ-(0, Cl)-OC=CH 2 -[Cp*C¼Ti(IV)] (1) was formed when the tube was shaken briefly at room temperature. 1 H NMR 1, CD2 C¼, shifts relative to CHDC¼ at <S5.32: (56.41, s (Cp's), 4.69(A), 4.30(B), 2s (C=CHAHB), 2.25, s (Cp*); in C7 D8 , -20"C, shifts relative to C7 D7 H at l> 2. 09: l> 6.17, s (Cp's), 4. 89, 4. 82, 2s (C=CHAHB), 1.99, s(Cp*). 13 Cgateddecoupled, -20eC, CD2 C¼, shifts relativeto~D2 C¼ at <S53.8: 215.1, dd, J = l0Hz (~=CHAHB), 132.1, dd, J = 155 Hz (C =~HAHB), 117.1, d of m, J = 177 Hz (Cp's), 107. 3, br s (~ 5Me 5 ), 13. 0, q, J = 127 Hz (C 5~e 5 ). Difference H NOE 1 spectroscopy, CD2 C¼, -20e C: Irradiation of the Cp* resonance enhanced the downfield vinylic resonance at (5 4. 69. Irradiation of the Cp resonance resulted in no enhancement of either resonance. 116 Reaction of 1 with CO. Following the procedure described above, - 1 was reacted with an excess of CO. A set of resonances assigned to bis-77 2 -µ-(0, Cl)-OC=CH 2 -bis [Cp2 Ti(IV)] (ID were observed. 1 H NMR, C 6 D6 , shifts relative to C6 D5H at r5 7 .15: ~ 5. 71, s (Cp's), 3. 73, d, J = 1. 5 Hz, 3. 65, d, J = 1. 5 Hz (C=CHAHB) integrating 10: 1: 1, respectively. Copious amounts of Cp2 Ti(CO) 2 were also observed (6 4. 57, Cp's). Reaction of~ and 1 with PEt3 • A NMR tube was loaded with ~ (12 mg, 0. 043 mmole). The tube was capped with a latex septum. The tube was cooled to O" C, and C6 D6 was added slowly. PEt3 (7 µL, O. 047 mmole) was added by syringe. The sample was shaken vigorously 1 at room temperature. The reaction was monitored by H NMR spectroscopy. After 10 min, the reaction had reached equilibrium, and Cp2 Ti(CH 2 )PE1:s (~) and ~ were observed in the ratio of ca. 95 : 5. The solution had changed from pinkish-red to yellow. 1 H NMR, ~' C6 D6 , shifts relative to C6 D5H at t5 7.15: r512.19, d, J = 5. 9 Hz (CH 2 ), 5. 37, d, J = 2 Hz (Cp's), 1. 4-1. 05, m (P~t3 ). Reaction of .1 with PEt3 (ca. 1 : 1) resulted in an equilibrium - - mixture of 4 and 9 in a ratio of 70 : 30. Preparation of Cp2 Ti(CH 2 )PMe 2 Ph (.!Q). A small Schlenk tube was loaded with .1 (200 mg, 0. 81 mmole). The vessel was cooled to 0° C and toluene (6 mL) was added via syringe. PMe 2 Ph (120 mg, 0. 87 mmole) was added via syringe. The vessel was then partially evacuated and opened to a cold trap to remove the isobutylene produced. The reaction mixture was allowed to warm to room temperature. After 117 stirring for 15 min, the volatiles were removed to yield a deep yellow oil. The oil was triturated with pentane (3 mL) to yield a yellow powder which was washed repeatedly with pentane. The yellow solids were dried in vacuo to yield 220 mg of !.Q (83%). 1 H NMR, C6 D6 , shifts relative to C 6 D5H at o7 .15: o12. 34, d, JP-H = 6. 8 Hz (CH 2 ), 7. 08, s (PMe 2 Ph), 5. 38, d, J = 1. 5 Hz, (Cp's), 1.11, d, J = 5. 9 Hz (PMe 2 Ph}. This material was estimated to be greater than 90% pure by 1 H NMR spectroscopy. 118 References and Notes 1. a) Grubbs, R.H., Prog. Inorg. Chem. 24, 1 (1978); b) Grubbs, R. H., "Comprehensive Organometallic Chemistry", G. Wilkinson, editor, Ch. 54, in press; c) Calderon, N., Lawrence, J.P., Ofstead, E. A., Adv. Organometal. Chem. !1, 449 (1979). 2. a) Herrmann, W. A., Angew. Chem. Int. Ed. Engl. 21, 117 (1982); b) Masters, C. , Adv. Organometal. Chem. !1, 61 (1979). 3. It appears that common usage of metal- 'alkylidene' and - 'carbene' has evolved to refer to nucleophilic and electrophilic M=CR2 units, respectively. 4. a) Fischer, E. P., Plabst, D., Chem. Ber. "'-"'-"" 107, 3326 (1974); b) Schrock, R. R.; Sharp, P. R., J. Am. Chem. Soc. 100, ~ 2389 (1978); c) Casey, C. P.; Anderson, R. L., J. Chem. Soc. Chem. Comm. 895 (1975). 5. A communication of this work has appeared: Ott, K. C.; Grubbs, R. H., J. Am. Chem. Soc. "'-"'-"" 103, 5922 (1981). 6. a) For a review of the field of bridging methylene compounds, see: Herrmann, W. A., Angew. Chem. Int. Ed. Engl. !1, 800 (1978); b) See also: Dimas, P. A.; Duesler, E. N.; Lawson, R. J.; and Shapley, J. R., J. Am. Chem. Soc. !.Q1, 7787 (1980), ref. 1 for an array of known µ-CH 2 complexes. 7. Only a few bis-µ-CH 2 are known to date: a) Isobe, K., Andrews, D. G., Mann, B. E., and Maitlis, P., J. Chem. Soc. Chem. Comm., 809 (1981); b) Schmidt, G. F., Muetterties, E. L., Beno, M. A., and Williams, J. M., Proc. Natl. Acad. Sci. USA, 78, ,-.,...,; 318 (1981); c) Hursthouse, M. B., Jones, R. A., 119 REFERENCES AND NOTES Malik, K. M. A., and Wilkinson, G., J. Am. Chem. Soc., .,.._,..,_,.__ 101, 4128 {1979). 8. Herrmann, W. A. , Plank, J. , Riedel, D., Ziegler, M. L., Weidenhammer, K., Guggolz, E., and Balbach, B., J. Am. Chem. Soc. --.,... 103, 63 (1981). 9. Cp2 TiMe{TFA) and Cp2 Ti{TFA) 2 were identified by titrating Cp2 TiMe 2 with TFAH in CD2 C¼ at low temperature, and comparison of the 1 H spectra. 10. a) Pine, S. H., Zahler, R., Evans, D. A., and Grubbs, R. H., J. Am. Chem. Soc. .,.._,..,_,.__ 102, 3270 (1980); b) Schrock, R. R., J. Am . _____ Chem. Soc. ..,.___ 98, 5399 (1976). 11. a) Tebbe, F. N., Parshall, G. W., and Reddy, G. S., J. Am. Chem. Soc. !.QQ, 3611 (1978); b) Tebbe, F. N., Parshall, G. W., and Ovenall, D. W., ibid., !.Q!_, 5074 (1974); c) Klabunde, U., Tebbe, F. N., Parshall, G. W., and Harlow, R. L., J. Mol. - Catal. 8, 37 (1980); d) Krusic, P. J., and Tebbe, F. N., Inorg. Chem., (1982) in press. They report that Cp2 Ti-µ-CH 2 µ -Cl-Cp2 Ti exchanges the µ-CH 2 with terminal olefins. 12. Pez, G. P., and Armor, J. N., Adv. Organomet. Chem. ]&, 1 {1981). There are many isomeric forms of CP2Ti, some of which will dimerize olefins and acetylenes as well as hydrogenate unsaturated organic compounds. 13. This titanacyclopentane has previously been observed in solution. Lee, J. B., and Grubbs, R. H., unpublished results. Under an 120 REFERENCES AND NOTES ethylene atmosphere, its red solutions are apparently fairly stable. Whitesides has reported an attempted preparation of this compound: McDermott, J. X. and Whitesides, G. M., J. Am. -- Chem. Soc. 96, 947 (1974). 14. kH /kn was determined by reaction of 2-C½ with excess diphenyl2 -,----, 2 acetylene to produce a mixture of Cp2 TiCPh=CPhCH 2 and -C½ and 2, 3-dimethyl-1-butene-do and -'½· The ratio of 2, 3-dimethyl-11 1 2 butene-1, 1-(½ to Cp2 TiCPh=CPhCD2 was 2. 0: 1 ( H NM:R). 15. Orr, D. A., Ott, K. C., and Grubbs, R. H., unpublished results. 1 H NMR, CDCls, shifts relative to CHCls at o7. 24: o6. 24 (Cp's), 2.92, (CH 2 , JSn-H = 69.4 Hz), 0.07 (SnMe3 , JSn-H = 50.8 Hz). The same complex was isolated from the reaction of Cp2 TiC½ with Me SnCH MgI. 23 3 16. 2 a) Hartner, Jr., F. W., and Schwartz, J., submitted for publication; b) Schwartz, J., and Gell, K. I., J. Organomet. --- Chem. 184, Cl (1980). 17. Cp2 Ti(77 2 -CH2 CO) was obtained from D. A. Straus. 18. Hencken, G., and Weiss, E., Chem. Ber. 106, 1747 (1973). 19. Ho, S. C., and Grubbs, R. H.; unpublished results. 20. Rausch, M. D., Boon, W. H., and Alt, H. G., J. Organomet. ~ Chem. 141, 299 (1977). ~ 21. Theopold, K. H., and Bergman, R. G., J. Am. Chem. Soc. 102, 5694 (1980). ~ 121 REFERENCES AND NOTES 22. Photolysis of a C6 D6 solution of Cp2 TiCH 2 CH(t-Bu)CH 2 at 20° C yielded t-butylcyclopropane which was identified by comparison with the 1H NMR spectra of t-butylcyclopropane isolated from the reaction of CP2 TiCH2 CH(t-Bu~CH 2 with ½; ref. 19. 23. Seyferth, D., Andrews, S. B., and Lambert, R. L., J. Organomet. Chem. ,...,..._ 37, 69 (1972). 24. Ott, K. C., Lee, J. B., and Grubbs, R. H., J. Am. Chem. Soc. 104, .,...._,.__,,.. 25. xxx (1982) in press . a) Kuhlman, E. J. , and Alexander, J. J., Coord. Chem. Rev. ~' 195 (1980); b) Calderazzo, F. Angew. Chem. Int. Ed. Engl. ~' 299 {1977); c) Wojcicki, A. Adv. Organomet. Chem. !!., 87 {1973). 26. Keppert, D. L., "The Early Transition Metals", Academic Press, New York, 1972, Chapter 1. 27. a) Fachmetti, G. ~ Fochi, G., and Floriani, C., J. Chem. Soc., Dalton Trans., 1946 (1977); b) Fachinetti, G., Floriani, C., and Stoeckli-Evans, H., ibid., 2297 (1977). 28. Wolczanksi, P. T., and Bercaw, J.E., Accts. Chem. Res. 13, ""' 121 (1980). 29. Manriguez, J.M., Fagan, P. J., Marks, T. J., Day, C. S., and Day, V. W., J. Am. Chem. Soc. ,.,.,,,....,,.. 100, 7112 (1978). 30. Straus, D. A., and Grubbs, R. H., unpublished results. 31. Reaction of 5 with CO in the presence of a slight excess of . - benzaldehyde produced a compound with 2 inequivalent Cp's 122 REFERENCES AND NOTES (o 5. 49 and 5. 29), an ABX pattern of 2 protons centered at 3. 52, and a Cp* resonance at 61. 90. The X-proton was not located. This species was present in better than 80% yield ( 1 H NMR, C 6 D6}, and is tentatively assigned as Cp2 TiCl-r,2-µ(0l, C3)-OCHPhCH2 C= O-TiCp*C~. A similar reaction of §_, acetone, and CO yields a compound in low yield ( 1 H NMR) with a heptet (J = 1. 2 Hz} at o4. 72 and a triplet (-J = 1. 2 Hz} at o1. 58 integrating 1 : 3, respectively. 32. Kerm, W., Roper, M., and strutz, H., J. Organomet. Chem. 219, C5 (1981). -"'-""-" 33. Sievert, A. C., Strickland, D.S., Shapley, J. R., Steinmentz, G. R., and Geoffroy, G. L., Organometallics, 1, 214 (1982). 34. Halbert, T. R., Leonowicz, M. E., and Maydonovitch, D. J., J. Am. Chem. Soc. -"'-""-" 102, 5101 (1980). 35. Herrmann, W. A., and Plank, J., Angew. Chem. Int. Ed. Engl. 17, .,..._,.._ 36. 525 (1978) . Messerle, L., and Curtis, M. D., J. Am. Chem. Soc., 102, ~ 7789 (1980). 37. Herrmann, W. A., Plank, J., Ziegler, M. L., and Weidenhammer, K., J. Am. Chem. Soc. -"'-""-" 101, 3133 (1979). 38. Dorrer, B., and Fischer, E. 0., Chem. Ber. 107, 2683 (1974). 39. Carr, D. B., and Schwartz, J., J. Am. Chem. Soc. !.Q!, 3521 (1971). 40. Van Horn, D. E., and Negishi, E. i J. Am. Chem. Soc. !.QQ, ~ 225 2 (1978). 123 REFERENCES AND NOTES 41. Chapter 2, this thesis. 42. Gorsich, R. D., J. Am. Chem. Soc. ~' 4744 (1958); Gorsich, R. D., ibid., 82, 4211 (1960). - .,...,.,,.__ 43. Bercaw, .J .E., Marvich, R.H., Bell, L. G., and Brintzinger, H. H., J. Am. Chem. Soc. .,...,.,,.__ 94, 1219 (1972). Only one equivalent of Cp*Li per TiC½ • 3THF was used. Cp*TiC12 was converted to Cp*TiC13 with anhydrous HCl. 44. Wolczanksi, P. T. , Ph.D. Thesis, California Insitute of Technology, 1981, p. 117. 45. Gordon, A. J., and Ford, R. A., "The Chemist's Companion", Wiley-Interscience, New York, 1972, p. 303. 46. Clark, E. P., Ind. Eng. Chem., Anal. Ed. g, 820 (1941). 47. A procedure adapted from Tebbella is described in Chapter 2, this thesis. 48. Heins, E., Hinck, H., Kaminsky, W., Opperman, G., Raulinat, P., and Sinn, H., Makromol. Chem. 134, 1 (1970). ~ 49. t AGl = 22 kcal-moi00 1 ; ~-~: 1 AGl00 = 24 kcal-moC for the reaction of 1 + Ph 2 C2 ~ Cp2 TiCH2 CPh=CPh. Ch. 2, ref. 8a, c. 50. A similar exchange has been noted in the cp; system where the ethylene complex cp;Ti(C 2H4} is stable. Cohen, S. A., Ph.D. Thesis, California Institute of Technology, 1982, Capter 1. 51. Only traces of ethane are observed when CP2TiCH 2 CH(t-Bu}CH 2 is thermolyzed. Lee, J. B., Howard, T. R., Grubbs, R. H., J. Am. Chem. Soc. 102, 6876 (1980). 124 REFERENCES AND NOTES (continued) 52. Recently, the complex [CP2Zr(µ-Cl)(µ-CH 2 )RuCIP 2 ] 2 has been isolated from a source of 'CP2ZrCH 2 ', and is a member of this class of compounds. Sabo, S., Chaudret, B., Gervais, D., - Poilblanc, R., Nouv. J. Chim. 5, 597 (1981).