Metal-alkyl group effects on the thermodynamic stability and stereochemical mobility of B(C6F5)3-derived Zr and Hf metallocenium ion-pairs

Article abstract of DOI:10.1021/ja000810a

This paper reports significant dependence of ion-pair formation energetics and stereomutation rates upon the metal-bound alkyl substituent (R) and the solvent dielectric constant in the metallocenium series (1,2-Me₂Cp)₂MR⁺

Full text of DOI:10.1021/ja000810a

10358
J. Am. Chem. Soc. 2000, 122, 10358-10370
Metal-Alkyl Group Effects on the Thermodynamic Stability and
Stereochemical Mobility of B(C6F5)3-Derived Zr and Hf
Metallocenium Ion-Pairs
Colin L. Beswick and Tobin J. Marks*
Contribution from the Department of Chemistry, Northwestern UniVersity, 2145 Sheridan Rd., EVanston,
Illinois 60208-3113
ReceiVed March 6, 2000
Abstract: This paper reports significant dependence of ion-pair formation energetics and stereomutation rates
upon the metal-bound alkyl substituent (R) and the solvent dielectric constant in the metallocenium series
-
t
(1,2-Me2Cp)2MR+ MeB(C6F5)3 where R ) Me, CH2 Bu, CH2TMS, CH(TMS)2, and M ) Zr and Hf, as
determined by reaction titration calorimetry and dynamic NMR spectroscopy. For the ion-pair forming reaction,
(1,2-Me2Cp)2M(R)Me + B(C6F5)3 f (1,2-Me2Cp)2MR+ MeB(C6F5)3-, enthalpies in toluene solution at 25
°C for M ) Zr and R ) Me, CH2TMS, and CH(TMS)2 are -24.6(0.8), -22.6(1.0), and -59.2(1.4) kcal/mol,
respectively. Corresponding values for M ) Hf and R ) Me and CH2TMS are -20.8(0.5) and -31.1(1.6)
kcal/mol, respectively. ∆H‡reorg values for the reorganization process that interchanges the ion-pair enantiotopic
t
sites for M ) Zr and R ) Me, CH2 Bu, CH2TMS, and CH(TMS)2 are 22(1), 18(1), 17(1), and 9(2) kcal/mol,
respectively. Corresponding ∆H‡reorg values for M ) Hf and R ) Me, CH2 Bu, and CH2TMS are >24, 12(3),
t
and 15(2) kcal/mol, respectively. ∆H‡reorg values are highest in low dielectric constant solvents such as C6D12.
Activation parameters for â-Me elimination in the complexes (1,2-Me2Cp)2MCH2 Bu+ MeB(C6F5)3- for M )
t
Zr and Hf are found to be ∆H‡
(3.4) cal/mol‚K, respectively.
) 22.5(0.9) and 17.3(0.9) kcal/mol, and ∆S‡
) 4.3(3.3) and -11.9-
â-Me
â-Me
Introduction
coordinatively unsaturated metal center (M), an effective
cocatalyst/weakly coordinating counteranion (X-), and appropri-
ate conditions of temperature, pressure, and solvent.3,4
Metal-mediated olefin polymerization catalysis has experi-
enced vast growth since pioneering discoveries were made in
the 1950s which showed that systems such as TiCl4/AlClEt2
promote the polymerization of ethylene to high-density poly-
ethylene1 and propylene to stereoregular polypropylene.2 More
recently, homogeneous “single-site” systems that provide high
activity, narrow product polydispersities, and control over
macromolecular architecture, and lend themselves to experi-
mental characterization have fueled intense basic and applied
research efforts.3 Homogeneous group 4 catalyst systems
represented by I embody many of the elements required for an
active olefin polymerization catalyst, including an appropriate
ancillary ligand framework (L2), an electron-deficient and
The versatility of metallocene ligation and a counteranion
offers access to structures providing a wide range of symmetries,
activities, and enchainment stereocontrol mechanisms.3-5 Recent
advances in nonmetallocene ligation for late and early transition
metals are also promising.3c,6 The role of the central metal has
also been investigated, and although the Ziegler-Natta termi-
(1) Ziegler, K.; Holzkamp, E.; Breil, H.; Martin, H. Angew. Chem. 1955,
67, 541-547.
(4) (a) Beswick, C. L.; Marks, T. J. Organometallics 1999, 18, 2410-
2412 (preliminary communication of some aspects of this work). (b) Deck,
P. A.; Beswick, C. L.; Marks, T. J. J. Am. Chem. Soc. 1998, 120, 1772-
1784. (c) Chen, Y.-X.; Metz, M. V.; Li, L.; Stern, C. L.; Marks, T. J. J.
Am. Chem. Soc. 1998, 120, 6287-6305. (d) Yang, X.; Stern, C. L.; Marks,
T. J. J. Am. Chem. Soc. 1994, 116, 10015-10031.
(5) For recent examples, see: (a) Veghini, D.; Henling, L. M.; Burkhardt,
T. J.; Bercaw, J. E. J. Am. Chem. Soc. 1999, 121, 564-573. (b) Resconi,
L.; Piemontesi, F.; Camurati, I.; Sudmeijer, O.; Nifantev, I. E.; Inchenko,
P.; Kuzmina, L. G. J. Am. Chem. Soc. 1998, 120, 2308-2321. (c) Ewen,
J. A. J. Mol. Catal. A: Chem. 1998, 128, 103-109, and references therein.
(d) Kravchenko, R.; Masood, A.; Waymouth, R. M.; Myers, C. L. J. Am.
Chem. Soc. 1998, 120, 2039-2046. (e) Bravakis, A. M.; Bailey, L. E.;
Pigeon, M.; Collins, S. Macromolecules 1998, 31, 1000-1009. (f) Obora,
Y.; Stern, C. L.; Marks, T. J.; Nickias, P. N. Organometallics 1997, 16,
2503-2505. (g) Mitchell, J. P.; Hajela, S.; Brookhart, S. K.; Hardcastle,
K. I.; Henling, L. M.; Bercaw, J. E. J. Am. Chem. Soc. 1996, 118, 1045-
1053. (h) Giardello, M. A.; Eisen, M. S.; Stern, C. L.; Marks, T. J. J. Am.
Chem. Soc. 1995, 117, 12114-12129.
(2) (a) Natta, G. Angew. Chem. 1956, 68, 393-402. (b) Natta, G.; Pino,
P.; Mazzanti, G.; Giannini, U. J. Am. Chem. Soc. 1957, 79, 2975-2976.
(3) For recent reviews, see: (a) Gladysz, J. A., Ed. Chem. ReV. 2000,
100 (special issue on “Frontiers in Metal-Catalyzed Polymerization”). (b)
Marks, T. J.; Stevens, J. C., Eds. Topics in Catalysis, 1999, 7, 1 (special
volume on “Advances in Polymerization Catalysis. Catalysts and Pro-
cesses”). (c) Kaminsky, W., Metalorganic Catalysts for Synthesis and
Polymerization: Recent Results by Ziegler-Natta and Metallocene InVes-
tigations, Springer-Verlag: Berlin, 1999. (d) Britovsek, G. J. P.; Gibson,
V. C.; Wass, D. F. Angew. Chem., Int. Ed. Engl. 1999, 38, 428-447
(nonmetallocene olefin polymerization catalysts). (e) Jordan, R. F. J. Mol.
Catal. 1998, 128, 1 (special issue on metallocene and single-site olefin
catalysts). (f) McKnight, A. L.; Waymouth, R. M. Chem. ReV. 1998, 98,
2587-2602 (constrained geometry polymerization catalysts). (g) Kaminsky,
W.; Arndt, M. AdV. Polym. Sci. 1997, 127, 144-187. (h) Bochmann, M. J.
Chem. Soc., Dalton Trans. 1996, 255-270. (i) Brintzinger, H. H.; Fischer,
D.; Mu¨lhaupt, R.; Rieger, B.; Waymouth, R. M. Angew. Chem., Int. Ed.
Engl. 1995, 34, 1143-1170.
10.1021/ja000810a CCC: $19.00 © 2000 American Chemical Society
Published on Web 10/07/2000

Zr and Hf Metallocenium Ion Pairs
J. Am. Chem. Soc., Vol. 122, No. 42, 2000 10359
nology has traditionally referred to Ti, Zr, and Hf catalysts, some
characteristics of these systems are also observable in later
transition metal,3c,6 lanthanide,7 and actinide catalysts.4d,8 Finally,
in the past decade, cocatalyst research, building upon the initial
discovery of methylalumoxane (MAO),9 has blossomed with
the introduction of fluoroaryl boron,4,10 aluminum,4c and other
cation-anion reagents11 capable of stoichiometric precatalyst
activation.
Previous work in our laboratory focused on acquiring and
quantitatively analyzing energetic information describing the
stability and structural dynamics of I as a function of ancillary
ligation, borane Lewis acidity, metal identity, and solvent
dielectric constant in such a manner as to vary only one
component while other variables were held constant.4
much as 14 kcal/mol.4b Armed with quantitative thermodynamic
and kinetic information, the rational design of more reactive,
selective, and thermodynamically stable catalyst systems be-
comes more realistic. To this point in our investigation, the
metal-hydrocarbyl substituent (R) of the metallocenium ion pair
has remained fixed with R ) Me in all systems studied.
However, the R influence on ion-pair formation/reorganization
processes is potentially sizable in view of the close spacial
proximity of R to the catalytic center. Considering that the R
metal substituent is a growing polymer chain at most times
during the olefin polymerization process, the influence of this
substituent is expected to be of great relevance to catalyst
performance.5,12,13
In this paper, we present a full discussion of the influence of
the metallocenium metal alkyl substituent on ion-pair formation
(eq 1) and ion-pair reorganization (eq 2) processes in solvents
of varying dielectric constant for a series of complexes of the
t
formula (1,2-Me2Cp)2MR+ MeB(C6F5)3- where R ) Me, CH2 -
Bu, CH2TMS, and CH(TMS)2, and M ) Zr and Hf. It will be
seen that these processes are highly sensitive to the exact
structure of the alkyl substituent, suggesting that the thermo-
dynamic and structural dynamic behavior of “real world” single-
site catalysts will be equally sensitive to the nature of a
propagating polymer chain. We also report the first kinetic study
of metallocenium â-methyl elimination for the system where
Ion-pair formation exothermicity as in eq 1, ion pair
reorganization rates as were quantified in eq 2, and where
applicable, olefin polymerization activities were found to be
promoted by sterically encumbered, electron-donating ancillary
ligation such as (Me5Cp)2, strongly Lewis acidic boranes such
as B(C6F5)3, and Zr complexes over Hf analogues. Ion-pair
reorganization barriers were found to be lower in higher
dielectric constant solvents such as CH2Cl2, in comparison with
less polar media such as toluene. Furthermore, catalyst forma-
tion energetics were found to be highly sensitive to small
changes in the structural components enumerated above. For
instance, modification of a single aryl ring in B(C6F5)2Ar can
alter (1,2-Me2Cp)2ZrMe2 methide abstraction enthalpies by as
t
R ) CH2 Bu.
Experimental Section
General Considerations. All organometallic complexes were pre-
pared and handled under air- and moisture-free conditions using a
circulating nitrogen-filled glovebox operating at <0.1 ppm oxygen
(Vacuum Atmospheres) or a vacuum line operating at <10-5 Torr with
argon purified by passing through supported MnO and molecular sieve
columns, or by standard Schlenk line techniques. Routine NMR
characterization experiments were carried out on Varian Gemini 300,
VXRS 300, Unity 400, or Unity 500 Plus instruments. Notations for
describing NMR features are s ) singlet, d ) doublet, t ) triplet, q )
quartet, br ) broad; the notation “” refers to the appearance to the eye
(6) (a) Younkin, T. R.; Connor, E. F.; Henderson, J. L.; Friedrich, S.
K.; Grubbs, R. H.; Bansleben, D. A. Science (Washington, D.C.) 2000,
287, 460-462. (b) Svejda, S. A.; Brookhart, M. Organometallics 1999,
18, 65-74. (c) Small, B. L.; Brookhart, M.; Bennett, A. M. A. J. Am. Chem.
Soc. 1998, 120, 4049-4050. (d) Bei, X.; Swenson, D. C.; Jordan, R. F.
Organometallics 1997, 16, 3282-3302. (e) Tsukahara, T.; Swenson, D.
C.; Jordan, R. F. Organometallics 3303-3313. (f) Kim, I.; Nichihara, Y.;
Jordan, R. F. Organometallics 3314-3323. (g) Ihara, E.; Young, V. G.,
Jr.; Jordan, R. F. J. Am. Chem. Soc. 1998, 120, 8277-8278.
(7) (a) Giardello, M. A.; Yamamoto, Y.; Brard, L.; Marks, T. J. J. Am.
Chem. Soc. 1995, 117, 3276-3277. (b) Stern, D.; Sabat, M.; Marks, T. J.
J. Am. Chem. Soc. 1990, 112, 9558-9575. (c) Jeske, G.; Schock, L. E.;
Swepston, P. N.; Schumann, H.; Marks, T. J. J. Am. Chem. Soc. 1985,
107, 8103-8110. (d) Watson, P. L.; Parshall, G. W. Acc. Chem. Res. 1985,
18, 51-56.
1
of what may be a more complex multiplet. Variable-temperature H
NMR experiments were carried out on the VXR 300 instrument using
a 5-mm detection probe calibrated with methanol or ethylene glycol
temperature standards. NMR line shape analyses followed standard
methods.14 NMR tubes fitted with Teflon valves were loaded with
precisely known quantities of solid samples in the glovebox and then
interfaced to a high vacuum line where dry, deoxygenated solvent (∼0.6
mL) was vacuum-transferred in, resulting in concentrations of ap-
proximately 5-10 mM. All solvents were dried over appropriate drying
agents. Elemental analyses were performed by Midwest Microlab
(Indianapolis, IN).
Starting Materials and Reagents. Tris(pentafluorophenyl)borane
was obtained as a gift from the Dow Chemical Company (Freeport,
TX) or was synthesized according to literature procedures.15 The borane
was purified by recrystallization from pentane at -30 °C followed by
high vacuum sublimation (80 °C/10-5 Torr). LiNp was prepared by a
procedure similar to that of Schrock16,17 except Li powder (Aldrich,
<0.5% Na) was used to shorten the reaction time to 1 day. The
(8) (a) Jia, L.; Yang, X.; Stern, C. L.; Marks, T. J. Organometallics 1994,
13, 3755-3757. (b) Yang, X.; King, W. A.; Sabat, M.; Marks, T. J.
Organometallics 1993, 12, 4254-4258. (c) Lin, Z.; Marechal, J. F. L.; Sabat,
M.; Marks, T. J. J. Am. Chem. Soc. 1987, 109, 4127-4129.
(9) (a) Siedle, A. R.; Lamanna, W. M.; Newmark, R. A.; Schroepfer, J.
N. J. Mol. Catal. A: Chem. 1998, 128, 257-271. (b) Reddy, S. S.; Sivaram,
S. Prog. Polym. Sci. 1995, 20, 309-367. (c) Clarke, B. S.; Barron, A. R.
Organometallics 1994, 13, 2957-2969. (d) Sishta, C.; Hathorn, R. M.;
Marks, T. J. J. Am. Chem. Soc. 1992, 114, 1112-1114. (e) Sinn, H.;
Kaminsky, W. AdV. Organomet. Chem. 1980, 18, 99-149.
(10) (a) Chen, Y.-X.; Marks, T. J. Chem. ReV. 2000, 100, 1391-1434.
(b) Williams, V. C.; Piers, W. E.; Clegg, W.; Elsegood, M. R. J.; Collins,
S.; Marder, T. B. J. Am. Chem. Soc. 1999, 121, 3244-3245. (c) Luo, L.;
Marks, T. J. in ref 3b, pp 97-106. (d) Li, L.; Marks, T. J. Organometallics
1998, 17, 3996-4003. (e) Piers, W. E.; Chivers, T. Chem. ReV. 1997, 26,
3345-3354. (f) Sun, Y.; Spence, R. E.v. H.; Piers, W. E.; Parvez, M.; Yap,
G. P. A. J. Am. Chem. Soc. 1997, 119, 5132-5143. (g) Yang, X.; Stern, C.
L.; Marks, T. J. J. Am. Chem. Soc. 1991, 113, 3623-3625. (h) Hlatky, G.
G.; Upton, D. J.; Turner, H. W. U.S. Patent Appl. 459921, 1990; Chem.
Abstr. 1991, 115, 256897v.
(12) (a) Brookhart, M.; Green, M. L. H.; Wong, L.-L. Prog. Inorg. Chem.
1988, 36, 1-124. (b) Grubbs, R. H.; Coates, G. W. Acc. Chem. Res. 1996,
29, 9, 85-93.
(13) See: Margl, P.; Deng, L.; Ziegler, T. in the following: J. Am. Chem.
Soc. 1999, 121, 154-162; J. Am. Chem. Soc. 1998, 120, 5517-5525;
Organometallics 1998, 17, 933-946.
(14) (a) Sandstrøm, J. Dynamic NMR Spectroscopy; Academic Press:
New York, 1982; pp 77-92. (b) Kaplan, J. I.; Fraenkel, G. NMR of
Chemically Exchanging Systems; Academic Press: New York, 1980; pp
71-128. (c) Ham, N. S.; Mole, T. Prog. Nucl. Magn. Reson. Spectrosc.
1969, 4, 91-192.
(11) Sun, Y.; Metz, M. V.; Stern, C. L.; Marks, T. J. Organometallics
2000, 19, 1625-1627.
(15) Massey, A. G.; Park, A. J. J. Organomet. Chem. 1964, 2, 245-
250.

10360 J. Am. Chem. Soc., Vol. 122, No. 42, 2000
Beswick and Marks
preparation of [(1,2-Me2Cp)2ZrMe]+[MeB(C6F5)3]- (5a)4b and the
analogous Hf complex (6a) have been previously described.4b
Reaction Calorimetry. Enthalpies of reaction between metallocenes
and B(C6F5)3 were determined at 25.0 °C in toluene solution. B(C6F5)3
solutions were prepared by vacuum transferring sufficient toluene from
Na/K onto the pure solute to prepare 100 mL (nominal) of solution,
while solutions of organometallic complexes having precisely known
concentrations were prepared with Na/K-dried toluene in the glovebox.
The organometallic titrant was metered into the stirred excess Lewis
acid solution using a calibrated buret controlled by a clutched
synchronous motor, and temperature changes were recorded using a
precision thermistor amplified by a Wheatstone bridge interfaced to
an analogue recorder. The calorimeter heat capacity was determined
by monitoring temperature changes when heat was introduced with a
calibrated resistive heater. Each run consisted of several sequential
titrations carried out for each reaction; the mean average deviation is
reported for each value. Overall instrument calibrations using established
methods18 ruled out significant sources of systematic error. Other details
of the techniques and apparatus for carrying out the calorimetric
measurements have been described elsewhere.19 Data analysis followed
standard published methods.20
103.2 (CH), 73.0 (ZrCH2), 37.5 (ZrCH3), 35.9 (CH3), 35.6 (CCH3),
13.3 (CCH3), 12.9 (CCH3). Anal. Calcd for C20H32Zr: C, 66.04; H,
8.89. Found: C, 64.47; H, 8.71; and C, 64.68; H, 8.75.
Bis(1,2-dimethylcyclopentadienyl)methyl(neopentyl)hafnium-
(IV) (4b). Chlorobis-(1,2-dimethylcyclopentadienyl)(methyl)hafnium-
(IV) (2) (0.81 g, 1.96 mmol) and lithium neopentyl (0.17 g, 2.17 mmol)
were added to a Schlenk flask along with pentane (50 mL) at 25 °C.
The solution was stirred for 2 days after which time the solution was
separated from a small amount of precipitated solid by filtration. The
clear solution was cooled to obtain a white crystalline solid which was
later sublimed (90 °C/10-5 Torr) to afford the analytically pure product
(0.63 g, 1.41 mmol). Yield, 72%. 1H NMR (C6D6): δ 6.23 (“t”, 2 H),
5.45 (“t”, 2 H), 4.93 (“t”, 2 H), 2.00 (s, 6 H), 1.65 (s, 6 H), 1.10 (s, 9
H), 0.10 (s, 2 H), -0.54 (s, 3 H). 13C NMR (C6D6): δ 121.2 (s, CCH3),
120.4 (s, CCH3), 113.5 (d, 170 Hz, CH), 106.5 (d, 168 Hz, CH), 102.7
(d, 171 Hz, CH), 74.9 (t, 106 Hz, HfCH2), 38.3 (q, 115 Hz, ZrCH3),
36.4 (q, 122 Hz, CH3), 36.2 (s, CCH3), 13.2 (q, 126 Hz, CCH3), 12.8
(q, 126 Hz, CCH3). Anal. Calcd for C20H32Hf: C, 53.26; H, 7.17.
Found: C, 53.05; H, 7.16; and C, 52.78; H, 6.99.
Bis(1,2-dimethylcyclopentadienyl)(neopentyl)zirconium(IV) Me-
thyltrispentafluoro-phenylborate (5b). This complex was prepared
in situ at low temperature due to the propensity for â-methyl
elimination, resulting in complex 5a and isobutene. Bis(1,2-dimethyl-
cyclopentadienyl)methyl(neopentyl)zirconium(IV) (3b) (3.6 mg, 0.010
mmol) and tris(pentafluorophenyl)borane (6.6 mg, 0.012 mmol) were
loaded into a Teflon valved NMR tube in the glovebox. Toluene-d8
(dried over Na/K alloy) was next transferred in at -78 °C. The clear
yellow solution was then used for variable-temperature NMR experi-
Chlorobis(1,2-dimethylcyclopentadienyl)methylzirconium(IV) (1).
Bis(1,2-dimethylcyclopentadienyl)dimethylzirconium(IV)4b,17 (480 mg,
1.6 mmol) and toluene (100 mL) were added to a vacuum reaction
flask. While stirring at 25 °C, HCl gas (1.6 mmol) was slowly
introduced over the course of 2 days. Volatile components were then
removed under vacuum. The residual crude solid was recrystallized
from toluene, resulting in colorless, analytically pure product (351 mg).
1
1
ments. Above 0 °C, the solution turns to a faint yellow and the H
Yield, 73%. H NMR (C6D6): δ 5.62 (“t”, 2 H), 5.41 (“t”, 2 H), 5.18
NMR is indicative of â-elimination. 1H NMR (C7D8, -40 °C): δ 5.96
(“t”, 2.59 Hz, 2 H), 5.31 (“t”, 3.09 Hz, 2 H), 5.11 (“t”, 2.15 Hz, 2 H),
1.50 (s, 6 H), 1.27 (s, 6 H), 1.22 (s, 2 H), 0.89 (s, 9 H), 0.06 (br, 3 H).
The â-methyl elimination products appear as follows: 5a: 1H NMR
(C6D6, RT): δ 5.51 (“t”, 2.86 Hz, 2 H), 5.27 (“t”, 3.06 Hz, 2 H), 4.83
(“t”, 2.77 Hz, 2 H), 1.53 (s, 6 H), 1.23 (s, 6 H), 0.10 (s, 3 H), 0.03 (br,
(“t”, 2 H), 1.99 (s, 6 H), 1.93 (s, 9 H), 0.26 (s, 3 H). 13C NMR (C6D6):
δ 125.1 (CCH3), 124.3 (CCH3), 114.6 (CH), 108.8 (CH), 104.7 (CH),
31.5 (ZrCH3), 13.6 (CCH3), 13.2 (CCH3). Anal. Calcd for C15H21-
ClZr: C, 54.92; H, 6.47. Found: C, 54.36; H, 6.39; and C, 54.30; H,
6.44.
Chlorobis(1,2-dimethylcyclopentadienyl)methylhafnium(IV) (2).
The following preparation is analogous to that of 1. Bis(1,2-dimeth-
ylcyclopentadienyl)dimethylhafnium(IV) (2.37 g, 6.0 mmol) and toluene
(150 mL) were added to a vacuum reaction flask. While stirring at 25
°C, HCl gas (6.0 mmol) was slowly introduced over the course of six
days. Volatile components were then removed under vacuum. The
residual crude solid was recrystallized from toluene, resulting in
colorless, analytically pure product (2.08 g). Yield, 83%. 1H NMR
(C6D6): δ 5.54 (“t”, 2 H), 5.38 (“t”, 2 H), 5.14 (“t”, 2 H), 2.00 (s, 6
H), 1.95 (s, 9 H), 0.08 (s, 3 H). 13C NMR (C6D6): δ 123.6 (CCH3),
122.8 (CCH3), 113.8 (CH), 108.5 (CH), 103.7 (CH), 33.2 (HfCH3),
13.5 (CCH3), 13.1 (CCH3). Anal. Calcd for C15H21ClHf: C, 43.38; H,
5.11. Found: C, 43.28; H, 5.21; and C, 43.20; H, 5.12.
3
3 H). 19F NMR (C6D6, RT): δ -126.0 (d, JFF ) 21.26 Hz), -150.1
3
(t, JFF ) 20.51 Hz), -155.0 (“t”, 19.26 Hz). Isobutene: 1H NMR
(C6D6, RT) δ 4.74 (m, 2 H), 1.59 (“t”, 6 H).
Bis(1,2-dimethylcyclopentadienyl)(neopentyl)hafnium(IV) Me-
thyltrispentafluoro-phenylborate (6b). The in situ preparation is
analogous to that for the related Zr complex 5b with a similar excess
of B(C6F5)3. The hafnium complex also undergoes â-methyl elimination
to form 6a and one equivalent of isobutene near 0 °C. 1H NMR (C7D8,
-36 °C): δ 5.65 (“t”, 2.54 Hz, 2 H), 5.33 (“t”, 2.89 Hz, 2 H), 5.19
(“t”, 2.31 Hz, 2 H), 1.48 (s, 6 H), 1.32 (s, 6 H), 1.01 (s, 2 H), 0.90 (s,
9 H), 0.30 (br, 3 H).
Chlorobis(1,2-dimethylcyclopentadienyl)(trimethylsilylmethyl)-
zirconium(IV). A solution of trimethylsilylmethylmagnesium chloride
(Aldrich, 1.0 M in Et2O, 8.0 mL, 8.0 mmol) diluted with Et2O (20
mL) was added dropwise over a period of 1 h to a stirred solution of
dichlorobis(1,2-dimethylcyclopentadienyl)zirconium(IV)17 (2.81 g, 8.1
mmol), in a mixture of CH2Cl2 (30 mL) and Et2O (30 mL) at 0 °C.
After stirring for 12 h, the solvent was removed under reduced pressure
and the remaining solid was triturated with a pentane-toluene solution
(1:2). The crude solid recovered after removing the solvent was
recrystallized (2×) from toluene. The product was recovered as fine,
Bis(1,2-dimethylcyclopentadienyl)methyl(neopentyl)zirconium-
(IV) (3b). Pentane (50 mL) was slowly added with stirring to chlorobis-
(1,2-dimethylcyclopentadienyl)(methyl)-zirconium(IV) (0.80 g, 2.45
mmol) and lithium neopentyl (0.20 g, 2.45 mmol) in a Schlenk flask
cooled to -78 °C. The solution was then allowed to warm to 0 °C and
stirred for 5 h. The solvent was then removed under vacuum. The
resulting crude solid was recrystallized from pentane, resulting in the
1
pale yellow, crystalline product (0.49 g, 1.34 mmol). Yield, 55%. H
NMR (C6D6): δ 6.40 (“t”, 2 H), 5.50 (“t”, 2 H), 4.90 (“t”, 2 H), 1.99
1
(s, 6 H), 1.64 (s, 6 H), 1.11 (s, 9 H), 0.35 (s, 2 H), -0.39 (s, 3 H). 13
C
colorless needlelike crystals (0.77 g). Yield, 24%. H NMR (C6D6):
δ 6.33 (“t”, 2 H), 5.22 (“t”, 2 H), 5.14 (“t”, 2 H), 2.08 (s, 6 H),
1.80 (s, 6 H), 0.74 (s, 2 H), 0.09 (s, 9 H). 13C NMR (C6D6): δ 125.8
(CCH3), 125.1 (CCH3), 116.7 (CH), 106.8 (CH), 104.4 (CH), 46.2
(CH2), 13.7 (CCH3), 13.5 (CCH3), 3.6 (SiCH3). Anal. Calcd for C18H29-
ClSiZr: C, 54.01; H, 7.32. Found: C, 53.33; H, 7.20; and C, 53.37;
H, 7.21.
Chlorobis(1,2-dimethylcyclopentadienyl)(trimethylsilylmethyl)-
hafnium(IV). A solution of trimethylsilylmethyllithium (Aldrich, 6.0
mL, 1.0 M in pentane, 6.0 mmol) was added dropwise to a stirred
solution of dichlorobis(1,2-dimethylcyclopentadienyl)hafnium(IV)4b
(2.52 g, 5.79 mmol) in Et2O (80 mL) at -78 °C. The mixture was
then allowed to warm to 25 °C. After stirring overnight, the solvent
was removed under vacuum. Toluene (3 × 15 mL) was added to the
remaining solid residue, the resulting solution was filtered to remove
NMR (C6D6): δ 122.2 (CCH3), 121.0 (CCH3), 113.6 (CH), 106.2 (CH),
(16) Schrock, R. R.; Fellmann, J. D. J. Am. Chem. Soc. 1978, 100, 3359-
3370.
(17) Smith, G. M., Ph.D. Thesis, Northwestern University, Evanston,
IL, 1987.
(18) (a) Eatough, D. J.; Christensen, J. J.; Izatt, R. M. Experiments in
Thermometric Titrimetry and Titration Calorimetry; Brigham Young
University Press: Provo, Utah, 1974; pp 61-63. (b) Hill, J. O.; Ojelund,
G.; Wadso¨, I. J. Chem. Thermodyn. 1969, 1, 111-116.
(19) (a) King, W. A.; Bella, S. D.; Lanza, G.; Khan, K.; Duncalf, D. J.;
Cloke, F. G. N.; Fragala`, I. L.; Marks, T. J. J. Am. Chem. Soc. 1996, 118,
627-635. (b) Schock, L. E. Ph.D. Thesis, Northwestern University,
Evanston IL, 1988.
(20) Barthel, J. Thermometric Titrations; John Wiley & Sons: New York,
1975; Vol. 45, pp 56-76.

Zr and Hf Metallocenium Ion Pairs
J. Am. Chem. Soc., Vol. 122, No. 42, 2000 10361
solid byproducts, and then the solvent was removed under vacuum.
The remaining crude solid was recrystallized from toluene. The
analytically pure product was recovered as a colorless crystalline solid
(1.97 g, 4.05 mmol). Yield, 70%. 1H NMR (C6D6): δ 6.25 (“t”, 2 H),
5.18 (“t”, 2 H), 5.17 (“t”, 2 H), 2.13 (s, 6 H), 1.84 (s, 2 H), 0.34 (s,
2H), 0.15 (s, 9 H). 13C NMR (C6D6): δ 123.2 (CCH3), 124.3 (CCH3),
116.7 (d, 170.8 Hz, CH), 106.7 (d, 169.5 Hz, CH), 105.8 (d, 172.0
Hz, CH), 43.39 (t, 105.4 Hz, CH2), 13.6 (q, 127.1 Hz, CCH3), 13.3 (q,
127.3 Hz, CCH3), 3.9 (q, 117.2 Hz, SiCH3). Anal. Calcd for C18H29-
ClSiHf: C, 44.35; H, 6.00. Found: C, 44.53; H, 5.98; and C, 44.66;
H, 6.06.
2.42 Hz, 2 H), 1.53 (s, 6 H), 1.30 (s, 2 H), 1.21 (s, 6 H), -0.03 (br, 3
H), -0.06 (s, 9 H). 13C NMR (C7D8, -40 °C): δ 148.5 (CF, d, 1JCF
)
1
1
233 Hz), 139.5 (CF, d, JCF ) 249 Hz), 137.35 (CF, d, JCF ) 249
Hz), 127.4 (CCH3), 126.7 (CCH3), 116.9 (CH), 111.0 (CH), 110.2
(CH), 64.6 (ZrCH2), 17.1 (BCH3, br), 12.6 (CCH3), 12.5 (CCH3), -3.11
(SiCH3). 19F NMR (C7D8, RT): δ -125.4 (d, 3JFF ) 22.19 Hz), -150.1
3
(t, JFF ) 19.71 Hz), -154.8 (“t”, 20.52 Hz). Anal. Calcd for C37H32-
BF15SiZr: C, 49.77; H, 3.62. Found: C, 49.23; H, 3.92; and C, 49.37;
H, 3.82.
Bis(1,2-dimethylcyclopentadienyl)(trimethylsilylmethyl)hafnium-
(IV) Methyl-trispentafluorophenylborate (6c). Bis(1,2-dimethylcy-
clopentadienyl)methyl(trimethylsilyl-methyl)hafnium(IV) (0.21 g, 0.44
mmol) and tris(pentafluorophenyl)borane (0.24 g, 0.47 mmol) were
loaded into a flip-frit vacuum apparatus. Toluene (25 mL) was vacuum
transferred in at -78 °C, and the mixture was allowed to warm to 25
°C while stirring. After 45 min, the solvent was removed under reduced
pressure after which pentane (30 mL) was vacuum transferred in. The
solution was filtered and then cooled to afford a colorless solid. Removal
of pentane by filtration and pumping under vacuum left the beige,
Bis(1,2-dimethylcyclopentadienyl)methyl(trimethylsilylmethyl)-
zirconium(IV) (3c). Method A. A solution of methyllithium (Aldrich,
1.4 M in Et2O, 2.5 mL, 3.5 mmol) diluted with Et2O (10 mL) was
added dropwise to a stirred solution of chlorobis(1,2-dimethylcyclo-
pentadienyl)(trimethylsilylmethyl)zirconium(IV) (0.80 g, 2.0 mmol) in
toluene (30 mL) at 0 °C. After stirring for 5 h, the solvent was removed
under reduced pressure and the remaining solid was triturated with
pentane (4 × 10 mL). The analytically pure colorless powder (0.64 g)
was recovered after removing the pentane under reduced pressure. Yield,
85%. 1H NMR (C6D6): δ 6.17 (“t”, 2 H), 5.41 (“t”, 2 H), 4.94 (“t”, 2
H), 1.99 (s, 6 H), 1.67 (s, 6 H), 0.16 (s, 9 H), -0.03 (s, 2 H), -0.43
(s, 3 H). 13C NMR (C6D6): δ 122.0 (CCH3), 121.4 (CCH3), 113.4 (CH),
106.2 (CH), 103.2 (CH), 44.5 (CH2), 32.3 (ZrCH3), 13.3 (CCH3), 13.0
(CCH3), 3.9 (SiCH3). Anal. Calcd for C19H32SiZr: C, 60.08; H, 8.51.
Found: C, 59.92; H, 8.16; and C, 59.87; H, 8.00.
1
analytically pure product (0.21 g). Yield, 48%. H NMR (C7D8, -42
°C): δ 5.58 (“t”, 2 H), 5.21 (“t”, 2 H), 5.01 (“t”, 2 H), 1.57 (s, 6 H),
1.22 (s, 6 H), 0.75 (s, 2 H), -0.24 (br, 3 H), -0.04 (s, 9 H). 19F NMR
(C7D8, RT): δ -133.5 (6 F), -159.4 (3 F), -164.8 (6 F). Anal. Calcd
for C37H32BF15SiHf: C, 45.34; H, 3.30. Found: C, 45.14; H, 3.41;
and C, 45.19; H, 3.55.
Bis(trimethylsilyl)methyllithium. The following procedure is modi-
fied from that of Lappert21 and is carried out under an argon atmosphere.
A solution of bis(trimethylsilyl)methyl chloride (Aldrich, 10.0 g, 51.3
mmol) in Et2O (20 mL) was added dropwise to a refluxing suspension
of Li powder (Aldrich, ∼3 g, ∼430 mmol) in Et2O (50 mL). The
mixture was stirred for 12 h, cooled, and then filtered through Celite.
The solvent was removed from the filtrate under reduced pressure and
the resulting solid sublimed (140 °C/10-5 Torr). The product was
obtained as a colorless solid (6.62 g). Yield, 87%. 1H NMR (C6D6): δ
0.14 (s, 18 H), -2.55 (s, 1 H).
Chlorobis(1,2-dimethylcyclopentadienyl)[bis(trimethylsilyl)methyl]-
zirconium(IV). Dichlorobis(1,2-dimethylcyclopentadienyl)zirconium-
(IV)4b,17 (3.18 g, 9.16 mmol) and bis(trimethylsilyl)methyllithium (1.95
g, 11.7 mmol) were added to a Schlenk flask, cooled to -78 °C, and
then Et2O (100 mL) was added. The mixture was slowly warmed to
25 °C and was then stirred for 7 days. The solution was next filtered,
and the solvent was removed under reduced pressure. The crude solid
was recrystallized from pentane, resulting in analytically pure, tan
crystals of the product (3.51 g). Yield, 81%. 1H NMR (C6D6): δ 6.59
(“t”, 1 H), 6.18 (“t”, 1 H), 6.08 (“t”, 1 H), 5.33 (“t”, 1 H), 5.19 (“t”,
1 H), 5.03 (“t”, 1 H), 2.09 (s, 3 H), 1.99 (s, 1 H), 1.96 (s, 3 H), 1.88
(s, 3 H), 1.51 (s, 3 H), 0.34 (s, 9 H), 0.21 (s, 9 H). 13C NMR (CDCl3):
δ 130.7 (CCH3), 127.5 (CCH3), 124.2 (CCH3), 122.3 (CH), 120.8
(CCH3), 117.0 (CH), 108.6 (CH), 108.2 (CH), 107.1 (CH), 104.2 (CH),
46.5 (ZrCH), 14.0 (CCH3), 13.9 (CCH3), 13.7 (CCH3), 13.2 (CCH3),
5.7 (SiCH3), 4.7 (SiCH3). Anal. Calcd for C21H37ClSi2Zr: C, 53.39;
H, 7.91. Found: C, 53.06; H, 7.89; and C, 53.10; H, 7.97.
Method B. Chlorobis(1,2-dimethylcyclopentadienyl)methylzirconium-
(IV) (1) (0.97 g, 3.0 mmol), toluene (30 mL), and pentane (25 mL)
were charged into a reaction flask and cooled to 0 °C. Under constant
stirring, a solution of trimethylsilylmethyllithium (Aldrich, 4.0 mL, 1.0
M in pentane) diluted with pentane (20 mL) was added dropwise over
0.5 h. The resulting solution was stirred for an additional 3 h.
Precipitated solid byproducts were removed by filtration, and volatiles
were removed from the filtrate under reduced pressure. After recrys-
tallization from pentane, a colorless solid (0.56 g, 1.5 mmol) was
obtained with spectral characteristics identical to those reported in
method A. Yield, 50%.
Methylbis(1,2-dimethylcyclopentadienyl)(trimethylsilylmethyl)-
hafnium(IV) (4c). A solution of methyllithium in Et2O (1.8 mL, 1.4
M in Et2O, 2.5 mmol) was added dropwise to a stirred Et2O/toluene
solution (15 mL/20 mL) of chlorobis(1,2-dimethylcyclopentadienyl)-
(trimethylsilylmethyl)hafnium(IV) (0.78 g, 1.60 mmol) at 0 °C. The
mixture was then allowed to warm to 25 °C. After stirring overnight,
the solvent was removed under vacuum. Pentane (3 × 15 mL) was
then added to the remaining solid residue, the mixture filtered to remove
solid byproducts, and solvent subsequently removed from the filtrate
under vacuum, leaving the crude colorless solid. Purification was
achieved by preferentially crystallizing the impurities from a pentane
solution, obtaining the supernatant solution by filtration, and removing
the pentane under vacuum. The analytically pure product was recovered
as a white solid (1.97 g, 4.05 mmol). Yield, 69%. 1H NMR (C6D6): δ
6.05 (“t”, 2 H), 5.36 (“t”, 2 H), 4.93 (“t”, 2 H), 2.00 (s, 6 H), 1.69 (s,
6 H), 0.16 (s, 9 H), -0.40 (s, 2H), -0.58 (s, 3H). 13C NMR (C6D6):
δ 120.7 (CCH3), 117.3 (CCH3), 113.2 (d, 169.3 Hz, CH), 106.0 (d,
169.2 Hz, CH), 102.4 (d, 107.7 Hz, CH), 46.3 (t, 104.4 Hz, CH2),
37.4 (q, 114.9 Hz, HfCH3), 13.2 (q, 126.2 Hz, CCH3), 12.9 (q, 126.7
Hz, CCH3), 4.2 (q, 117.1 Hz, SiCH3). Anal. Calcd for C19H32SiHf: C,
48.86; H, 6.91. Found: C, 48.99; H, 6.89; and C, 49.06; H, 6.92.
Methylbis(1,2-dimethylcyclopentadienyl)[bis(trimethylsilyl)methyl]-
zirconium(IV) (3d). Chlorobis(1,2-dimethylcyclopentadienyl)[bis-
(trimethylsilyl)methyl]zirconium(IV) (2.69 g, 5.71 mmol) and Et2O (35
mL) were added to a Schlenk flask and cooled to -78 °C. MeLi
(Aldrich, 1.4 M in Et2O, 4.9 mL, 6.9 mmol) was then injected with
stirring. After warming to 25 °C and stirring for 24 h, the solution was
filtered and the solvent was removed from the filtrate under reduced
pressure. The resulting solid was recrystallized twice from pentane,
resulting in colorless, analytically pure crystals of the product (1.82
Bis(1,2-dimethylcyclopentadienyl)(trimethylsilylmethyl)zirconium-
(IV) Methyl-trispentafluorophenylborate (5c). Bis(1,2-dimethylcy-
clopentadienyl)methyl(trimethylsilyl-methyl)zirconium(IV) (0.15 g, 0.40
mmol) and tris(pentafluorophenyl)borane (0.21 g, 0.42 mmol) were
loaded into a fritted reaction apparatus. Pentane (20 mL) was vacuum
transferred in at -78 °C, and the mixture was allowed to warm to 25
°C while stirring. After 30 min, the solution was cooled to -78 °C,
filtered, and the solvent removed under reduced pressure. The solid
was then redissolved in pentane at 0 °C and filtered. Removal of the
solvent and pumping under vacuum (6 h) left the pale yellow,
1
g). Yield, 71%. H NMR (C6D6): δ 6.49 (“t”, 1 H), 6.46 (“t”, 1 H),
5.96 (“t”, 1 H), 5.27 (“t”, 1 H), 5.19 (“t”, 1 H), 4.58 (“t”, 1 H), 2.08
(s, 3 H), 1.70 (s, 3 H), 1.67 (s, 3 H), 1.40 (s, 3 H), 1.32 (s, 1 H), 0.21
(s, 9 H), 0.20 (s, 9 H), -0.29 (s, 3 H). 13C NMR (C6D6): δ 125.0
(CCH3), 122.3 (CCH3), 122.2 (CCH3), 119.0 (CH), 118.2 (CCH3),
115.5 (CH), 107.3 (CH), 107.2 (CH), 107.0 (CH), 102.1 (CH), 42.9
(ZrCH), 40.7 (ZrCH3), 13.2 (CCH3), 13.1 (CCH3), 13.0 (CCH3), 12.8
1
analytically pure product (0.12 g). Yield, 30%. H NMR (C7D8, -40
(21) Davidson, P. J.; Harris, D. H.; Lappert, M. F. J. Chem. Soc., Dalton
Trans. 1976, 2268-2274.
°C): δ 5.74 (“t”, 2.54 Hz, 2 H), 5.22 (“t”, 3.12 Hz, 2 H), 5.01 (“t”,

10362 J. Am. Chem. Soc., Vol. 122, No. 42, 2000
Beswick and Marks
(CCH3), 6.0 (SiCH3), 5.3 (SiCH3). Anal. Calcd for C22H40Si2Zr: C,
58.45; H, 8.94. Found: C, 58.28; H, 8.83; and C, 58.22; H, 8.77.
Bis(1,2-dimethylcyclopentadienyl)[bis(trimethylsilyl)methyl]-zir-
conium(IV) Methyl-trispentafluorophenylborate (5d). Methylbis(1,2-
dimethylcyclopentadienyl)-[bis(trimethylsilyl)methyl]zirconium(IV) (0.21
g, 0.46 mmol) and tris(pentafluorophenyl)borane (0.24 g, 0.48 mmol)
were added to a vacuum reaction/filtration flask. Toluene (25 mL) was
vacuum transferred in at -78 °C, and the resulting solution was warmed
to 25 °C and stirred for 30 min. The solvent was then removed under
reduced pressure leaving a viscous orange oil. Pentane (20 mL) was
then vacuum transferred in, and after stirring, the oil solidified. The
solid was collected by filtration and the solvent removed. The solid
product was washed a second time with pentane after which bright
yellow, analytically pure product (0.39 g, 0.41 mmol) was recovered.
Scheme 1. Synthetic Pathways to Substituted Metallocenium
Ion-Pair Complexes
1
Yield, 92%. H NMR (C7D8): δ 5.80 (t, 3.18 Hz, 1 H), 5.55 (d, 3.22
Hz, 2 H), 5.38 (d, 3.13 Hz, 2 H), 5.30 (t, 3.12 Hz, 1 H), 3.09 (s, 1 H),
1.53 (s, 6 H), 1.43 (s, 6 H), 1.24 (br, 3 H), -0.30 (s, 18 H). 1H NMR
(CD2Cl2): δ 6.43 (t, 3.12 Hz, 1 H), 6.26 (d, 3.12 Hz, 2 H), 6.16 (d,
3.12 Hz, 2 H), 5.96 (t, 3.12 Hz, 1 H), 3.72 (s, 1 H), 2.15 (s, 6 H), 2.05
(s, 6 H), 0.40 (br, 3 H), 0.07 (s, 18 H). 13C NMR (CD2Cl2): δ 147.2
1
1
1
(d, JCF ) 241 Hz), 137.1 (d, JCF ) 244 Hz), 136.0 (d, JCF ) 249
Hz), 129.8 (CCH3), 127.1 (CCH3), 116.2 (CH), 114.0 (CH), 113.1
(CH), 110.3 (CH), 80.7 (ZrCH), 13.7 (CCH3), 13.4 (CCH3), 3.8
(SiCH3). 19F NMR (C7D8): -124.17 (d, JFF ) 23.1 Hz), -154.78 (t,
3
3JFF ) 20.6 Hz), -157.12 (“t”, 3JFF ) 21.5 Hz). Anal. Calcd for C40H40-
BF15Si2Zr: C, 49.83; H, 4.19. Found: C, 49.39; H, 4.19; and C, 49.31;
H, 4.05.
Table 1. Formation Enthalpies in Toluene Solution at 25 °C for
-
(1,2-Me2Cp)2MR+ MeB(C6F5)3 Complexes
∆Hforma (kcal/mol)
Results
complex
M
R
5a
5b
5c
5d
6a
6b
6c
Zr
Zr
Zr
Zr
Hf
Hf
Hf
Me
-24.6(0.8)b
c
-22.6(1.0)
-59.2(1.4)
-20.8(0.5)b
c
The following sections first describe the synthesis and
characterization of the (1,2-Me2Cp)2M(R)Me precursor com-
plexes as well as the corresponding metallocenium ion pairs.
Calorimetrically determined enthalpies of methide abstraction
from the (1,2-Me2Cp)2M(R)Me series by B(C6F5)3 are then
discussed along with a kinetic study of â-methyl elimination
t
CH2 Bu
CH2TMS
CHTMS2
Me
CH2 Bu
CH2TMS
t
-31.1(1.6)
t
for the metallocenium complexes containing the R ) CH2 Bu
a Values are determined by titration calorimetry. See eq 3 for reaction
details. b From ref 4b. c â-Me elimination precludes determination at
25 °C.
substituent. The rates of ion-pair structural reorganization
processes are then measured and related to the type of substituent
and the nature of the solvent, with close attention to substituent
rotational barriers about the M-R bonds which in this case can
be an index of intramolecular nonbonded interactions.
complex 3d proceeds smoothly. Of significant interest in the
present study is the properties of neopentyl ion-pair complexes
5b and 6b. While methide abstraction proceeds cleanly at low
Synthetic Methodology. To examine metal-alkyl substituent
effects on metallocenium ion-pair formation and structural
dynamics, a series of new Zr and Hf complexes was prepared
as shown in Scheme 1. One metal-bound methyl substituent on
each of the (1,2-Me2Cp)2MMe2 complexes 3a and 4a was
replaced with chloride using 1.0 equivalent of HCl(g). Alkylation
of the resulting mixed methyl-chloro metallocenes 1 and 2 was
accomplished with either lithium or Grignard reagents to obtain
the methyl-alkyl Zr complexes 3b-d and Hf complexes 4b,c.
An alternate route to complexes 3c and 4c is to add 1.0
equivalent of trimethylsilylmethyllithium to (1,2-Me2Cp)2MCl2,
and to then use MeLi for alkylation of the second chloride.
Methide abstraction to obtain the ion-pair complexes 5 and 6
was accomplished with the strongly Lewis acidic borane,
B(C6F5)3.10c For most cases, abstraction is completely selective
for the methide functionality within 1H NMR detection limits.22
The steric demands of the alkyl substituent R ) CH(TMS)2
apparently preclude the formation of what would have been Hf
complex 4d, even starting from (1,2-Me2Cp)2HfCl2. Lappert23
reported similar experimental difficulties with less sterically
encumbered Cp2HfCl2. In contrast, formation of the related Zr
1
temperatures, as judged by H NMR, â-methyl elimination is
observed at room temperature leading to quantitative formation
of 3a and 4a in the case of Zr and Hf, respectively, along with
an accompanying equivalent of isobutene. Complexes 5 and 6
are all colorless or light-beige solids with the exception of 5d
which is bright yellow both in solution and as a solid. The new
complexes were characterized by standard spectroscopic and
analytical methodologies (see Experimental Section for data).
Ion-Pair Formation Enthalpies (∆Hform). Metallocenium
ion-pair enthalpies of formation (∆Hform) as shown in eq 3 were
measured in toluene solution by isoperibol titration calorimetry,
1
and results are presented in Table 1. In all cases, H NMR
experiments under the same conditions as the calorimetric
experiments show that methide abstraction from the neutral
metallocene by B(C6F5)3 is clean, fast, and quantitative at 25
°C. Multiple alkyl substituent abstraction from a single metal-
locene is not observed, even with a stoichiometric excess of
borane, and at no time are binuclear complexes,4c,24 such as II,
detectable by features such as a characteristically high-field
M-(µ-Me)-M+ 1H NMR signal. As noted above, 1H NMR
experiments show that metallocene-Me substituent abstraction
is exclusive over abstraction of other R groups, to the accuracy
(22) In the case of R ) CH2Ph, abstraction chemistry is not clean as a
result of competitive benzyl abstraction.
(23) (a) Lappert, M. F.; Riley, P. I.; Yarrow, P. I. W.; Atwood, J. L.;
Hunter, W. E.; Zaworotko, M. J. J. Chem. Soc., Dalton Trans. 1981, 814-
821.
(24) (a) Jia, L.; Yang, X.; Stern, C. L.; Marks, T. J. Orgnometallics 1997,
16, 842-857. (b) Bochmann, M.; Lancaster, S. J.; Hursthouse, M. B.; Malik,
K. M. A. Organometallics 1994, 13, 2235-2243.

Zr and Hf Metallocenium Ion Pairs
J. Am. Chem. Soc., Vol. 122, No. 42, 2000 10363
1
of H NMR spectroscopic detection limits.
Table 1 indicates that abstraction enthalpies (∆Hform) for
zirconocene complexes 5a and 5c are similar for R ) Me and
R ) CH2TMS. In contrast, Zr complex 5d having the R ) CH-
(TMS)2 alkyl substituent exhibits a significantly greater abstrac-
tion exothermicity, in excess of twice that of the other Zr entries
in Table 1. The physical properties of 5d, such as a bright yellow
color and limited solubility in relatively nonpolar hydro-
carbon solvents such as toluene, suggest looser ion-pairing
characteristics.24a,25 This hypothesis is corraborated by NMR
spectroscopic observables such as the upfield shift of the boron-
bound CH3 (δ ) 0.40 ppm) in toluene solution, which is
characteristic of free MeB(C6F5)3,- 4a-c,25a as is the modest ∆δ-
(m-, p-F)25b value of 2.34 ppm. Hafnium complex 6a (R ) Me)
exhibits an abstraction enthalpy 2-4 kcal/mol less exothermic
than the analogous Zr complex; however, ∆Hform for Hf complex
6c where R ) CH2TMS is significantly more exothermic than
for the Zr analogue.
Figure 1. First-order kinetic plots (top) and corresponding Eyring plots
(bottom) for the â-Me elimination processes depicted in eq 4 for (1,2-
â-Methyl Elimination Kinetics in Metallocenium Neopen-
tyl Complexes. In the case of neopentyl complexes 5b and 6b,
methide abstraction enthalpies could not be obtained by calo-
rimetry because the complexes begin to undergo â-Me elimina-
tion near -15 °C. This result is not completely surprising as
ample evidence exists for metallocenium â-Me elimination in
the literature.4d,26 However, the present transformation is unique
in being clean and quantitative, affording an opportunity to
directly observe and quantitatiVely measure and compare rates
and barriers for â-Me elimination in Zr and Hf metallocenium
complexes. Rates of eq 4 were measured at 4-5 different
temperatures over a 30 °C range in toluene-d8 solution. The
progress of the reactions was monitored by integration of at
least one 1H NMR signal from the reactant (5b for M ) Zr and
6b for M ) Hf) and product (5a for M ) Zr and 6a for M )
Me2Cp)2MCH2 Bu+ MeB(C6F5)3 in toluene-d8 where M ) Zr (5b)
and M ) Hf (6b).
t
-
Table 2. Activation Parameters for â-Me Elimination in the
Complexes (1,2-Me2Cp)2MCH2 Bu+ CH3B(C6F5)3 Where M ) Zr
t
-
and Hf
∆G‡â-Me (0 °C)
(kcal/mol)
∆H‡
∆S‡
â-Me
(cal/mol‚K)
â-Me
(kcal/mol)
complex
M
5b
6b
Zr
Hf
21.2(0.2)
20.7(0.2)
22.5(0.9)
17.3(0.9)
4.3(3.3)
-11.9(3.4)
is observed as reported in other zirconocenium systems where
â-Me elimination has been analyzed in a more qualitative
fashion.26c The present data can be fit to first-order kinetic plots
as shown in Figure 1.
1
Eyring plots of the â-methyl elimination kinetic data are also
shown in Figure 1, and derived â-Me elimination barriers
compiled in Table 2. Activation parameters for Zr complex 5b
are ∆H‡â-Me ) 22.5(0.9) kcal/mol and ∆S‡â-Me ) 4.3(3.3) cal/
mol‚K. Corresponding values for the Hf complex 6b are
∆H‡â-Me ) 17.3(0.9) kcal/mol and ∆S‡â-Me ) -11.9(3.4) cal/
mol‚K. The enthalpic barrier for the zirconium complex is ∼5
kcal/mol greater than that in the Hf complex, which is partially
offset by the entropic contribution which is ∼16 cal/mol‚K
greater for the Zr complex.
Hf). The H NMR
signals for the coproduct isobutene are identical to those for
the free olefin, arguing against significant olefin coordination
to the metal center26d in the ground state. Furthermore, the
reaction proceeds to completion, and no detectable back reaction
Ion-Pair Reorganization Processes. Increasing amounts of
data argue that metallocene olefin polymerization activity and
enchainment stereochemistry are correlated with, among other
factors, the ability to form structurally mobile ion pairs
containing an electrophilic, coordinatively unsaturated metal
center.3,4,10,24a Strongly Lewis acidic B(C6F5)3 induces a high
degree of cationic polarization at the metal center via methide
abstraction to form a metallocenium cation-methyl borate anion
ion pair.4,10,24a,27 Separation and subsequent recombination of
the ion pair exchanges diastereotopic ring Me groups via the
reorganization process shown in eq 5 and Figure 2A. This
(25) (a) Lee, R. A.; Lachicotte, R. J.; Bazan, G. C. J. Am. Chem. Soc.
1998, 120, 6037-6047. (b) Horton, A. D.; deWith, J.; van der Linden, A.
J.; van de Weg, H. Organometallics 1996, 15, 2672-2674.
(26) (a) Resconi, L.; Camurati, I.; Sudmeijer, O. Top. Catal. 1999, 7,
145-163. (b) Shaffer, T. D.; Cannich, J. M.; Squire, K. R. Macromolecules
1998, 31, 5145-5147. (c) Horton, A. D. Organometallics 1996, 15, 2675-
2677. (d) Guo, Z.; Swenson, D. C.; Jordan, R. F. Organometallics 1994,
13, 1424-1432. (e) Resconi, L.; Piemontesi, F.; Franciscono, G.; Abis, L.;
Fiorani, T. J. Am. Chem. Soc. 1992, 114, 1025-1032. (f) Eshuis, J. J. W.;
Tan, Y. Y.; Meetsma, A.; Teuben, J. H. Organometallics 1992, 11, 362-
369. (g) Kesti, M. R.; Waymouth, R. M. J. Am. Chem. Soc. 1992, 114,
3565-3567. (h) Mise, T.; Kageyama, A.; Miya, S.; Yamazaki, H. Chem.
Lett. 1991, 1525-1528. (i) Eshuis, J. J. W.; Tan, Y. Y.; Teuben, J. H. J.
Mol. Catal. 1990, 62, 277-287.
-
effectively permutes the coordination sites of the MeB(C6F5)3

10364 J. Am. Chem. Soc., Vol. 122, No. 42, 2000
Beswick and Marks
Table 3. Ion-Pair Reorganization Activation Parameters for (1,2-Me2Cp)2MR+ MeB(C6F5)3 Complexes as Determined by H NMR Line
-
1
Shape Analysis
f
i
i
∆G‡
∆H‡
∆S‡
reorg
reorg
reorg
complex
M
Zr
R
solvent
ꢀa
(kcal/mol)
(kcal/mol)
(cal/K•mol)
coalescence (°C)
5a
Me
toluene-d8
C6D5Cl
1,2-C6D4Cl2
toluene-d8
C6D12
C6D11CD3
toluene-d8
CCl2DF
CD2Cl2
toluene-d8
CCl2DF
CD2Cl2
toluene-d8
C6D5Cl
2.37
5.71
9.93
2.37
2.02
2.07
2.37
17.6(2)b
15.5c
22(1)b
11(2)b
12(2)b
18(1)g
d
19(1)
17(1)
14(3)
17(1)
d
13(2)b
-15(8)b
-10(4)b
15(2)g
d
15(2)
10(3)
17(4)
23(2)
d
110b
15.0c
5b
5c
Zr
Zr
CH2 Bu
13.8(2)g
d
14.8(2)
13.8(2)
12.2(2)
10.0(2)
d
> -6
35
35
20
-8
-25
< -60h
-100h
< -78h
e
t
CH2TMS
9.08
2.37
5d
6a
Zr
CH(TMS)2
Me
8.0(4)
d
9(2)
d
e
7(4)
d
e
9.08
2.37
5.71
9.08
9.93
2.37
2.02
2.07
2.37
9.08
Hf
e
15.7c
13(4)b
11(1)b
12(3)b
12(3)
19(2)
15(1)
15(2)
11(2)
-9(1)b
-16(2)b
-5(8)b
-6(13)
12(5)
0(3)
CD2Cl2
15.8c
1,2-C6D4Cl2
toluene-d8
C6D12
C6D11CD3
toluene-d8
CD2Cl2
13.5c
6b
6c
Hf
Hf
CH2 Bu
13.8(2)g
15.5(2)
15.2(2)
14.0(2)
13.1(2)
10
55
50
20
0
t
CH2TMS
3(5)
-5(4)
a Dielectric constant from ref 28. b Also see ref 4b. c Calculated at 25 °C from ∆H‡ips and ∆S‡ips values. d Solvent froze before the slow exchange
limit was reached. e Barrier too high to be determined by this method. f At the indicted coalescence temperature unless otherwise noted. g At 0 °C.
h Coalescence of diastereotopic TMS signals. i Estimated uncertainties are obtained from standard regression analysis. Also see ref 14a.
and R substituents. The rate of such processes provides one
measure of the “tightness” of the ion pairing.4b,10c,24a
Equilibration rates (kreorg) for eq 5 were measured as a
function of R and solvent via spectral line broadening14 in
1
variable-temperature H NMR experiments and were used to
calculate the activation parameters shown in Table 3. Line
shapes in typical experiments were found to be independent of
metallocenium complex concentration over a 7-fold range and
independent of borane concentration over a 3-fold range, arguing
against the significance of intermolecular processes such as “SE2
like” mechanisms under these conditions. Rather, it is proposed
that the principal dynamic process observed involves the borate
anion passing between symmetrically disposed sites in the
metallocenium coordination sphere (Figure 2). Representative
NMR spectra are shown for 5c in Figure 3. For Zr and Hf
complexes 5a-c and 6a-c, analysis of the exchange-broadened,
diastereotopic (see Figure 2A) ring methyl signals was carried
out. As shown in Figure 2B, the inferred orientation of the R
) CH(TMS)2 substituent in Zr complex 5d removes the
horizontal symmetry plane bisecting the metallocene wedge
which is present in the other complexes. In this case, four ring
Figure 2. Newman projections illustrating dynamic reorganization
-
processes in various (1,2-Me2Cp)2MR+ MeB(C6F5)3 metallocenium
complexes. The methylborate anion detaches from the cationic metal
center and reattaches to the opposite side. In A, two ring-methyl signals
are observed at slow exchange, and simultaneous rotation about the
M-CH2R′′ bond is postulated. In B, four ring-methyl signals are
observed at slow exchange.
reorganization activation barriers determined from one pair of
broadened MeCp signals agree well with those determined from
TMS signal broadening, as expected if the process depicted in
Figure 2B is correct. Respective activation parameters deter-
mined from ring methyl exchange and related TMS exchange
in 5d are ∆H‡reorg ) 9 and 9(2) kcal/mol; ∆S‡reorg ) 5 and 7(4)
1
Me and two TMS H NMR signals are observed in the slow-
exchange limit. Broadening of the TMS proton signals was used
to quantitate rates of ion-pair reorganization because of the
complicated pattern of four MeCp signals at low temperature,
as shown in Figure 4. However, from a limited data set,
cal/mol‚K; and ∆G‡
(-100 °C) ) 8.0 and 8.0(0.4) kcal/
reorg
mol. Rotation about the ZrsCH(TMS)2 bond is found to be
slow up to and exceeding 40 °C, as judged by variable-
temperature 1H NMR experiments on neutral precursor complex
3d (vide infra).
(27) (a) Lanza, G.; Fragala, I. L.; Marks, T. J. J. Am. Chem. Soc. 1998,
120, 8257-8258. (b) Lanza, G.; Fragala, I. L. in ref 3b, pp 45-60. (c)
Lanza, G.; Fragala, I. L.; Marks, T. J., submitted for publication.

Zr and Hf Metallocenium Ion Pairs
J. Am. Chem. Soc., Vol. 122, No. 42, 2000 10365
Figure 3. Variable-temperature 1H NMR spectra for (1,2-Me2Cp)2ZrCH2TMS+ MeB(C6F5)3- in toluene-d8 solution from -40 to +40 °C. Dynamic
processes are shown in Figure 2A (R ) CH2TMS).
the Hf complexes than for the corresponding Zr complexes with
the same R substituent and measurement solvent, even though
the corresponding ∆H‡reorg parameters for the Hf complexes are
2-6 kcal/mol smaller in each case except for R ) Me. The
reorganization enthalpies are offset by somewhat less favorable
(less positive) ∆S‡
values in the case of the Hf complexes,
reorg
suggestive of more highly organized transition states.
The data in Table 3 also indicate that higher polarity solvents
significantly reduce the barriers to ion-pair reorganization.
Solvent polarity, as indexed by dielectric constant, influences
∆H‡reorg by as much as 7-8 kcal/mol in the case of 6c in C6D12
(ꢀ ) 2.02), versus toluene-d8 (ꢀ ) 2.37), versus CD2Cl2 (ꢀ )
9.08) where the ∆H‡reorg values are 19(2), 15(2), and 11(2) kcal/
mol, respectively. In the case of R ) Me, it was previously
observed4b that activation entropies (∆S‡reorg) are somewhat less
favorable (more negative) in solvents with larger dielectric
Figure 4. 1H NMR ring-Me signal region at -126 °C for the complex
(1,2-Me2Cp)2ZrCHTMS2+ MeB(C6F5)- (5d) in CDCl2F solution.
3
The kinetic data collected in Table 3 reveal a number of
interesting trends concerning how the ion-pair separation/
reorganization barrier is related to the identity of R, the metal,
and the solvent dielectric constant.28 In regard to R effects, the
barriers for M ) Zr ion-pair reorganization decrease in the order
t
constants. However, for the larger substituents R ) CH2 Bu and
R ) CH2TMS, the activation entropies are considerably more
positive (suggestive of less organized transition states) although
values for Hf complexes are not as large as for the Zr complexes.
The enhanced solubility of R ) CH2TMS complexes 5c and
6c in aliphatic hydrocarbons versus the corresponding R ) Me
complexes 5a and 6a allows ion-pair reorganization barriers to
be measured for the first time in a minimally coordinating
saturated hydrocarbon solvent (the medium in which most large-
scale polymerizations are carried out). In cyclohexane-d12,
t
Me > CH2 Bu ∼ CH2TMS > CH(TMS)2 with reorganization
rates spanning over 8 orders of magnitude. As noted above,
the low solubility of 5d in toluene is suggestive of significant
ionic character,24a,25 and the CH3B(C6F5)3- 1H chemical shift
(δ ) 0.40 in CD2Cl2, 1.24 in toluene-d8) at 25 °C is also
indicative of a free anion rather than one tightly ion-paired with
the cation.4a-c,25,29 A similar ordering of ∆H‡reorg parameters is
observed in the M ) Hf series. For R ) Me in toluene solution,
the reorganization barrier is too large to measure by line-
∆H‡
for 6c (M ) Hf) is 19(2) kcal/mols slightly higher
reorg
t
broadening (>24 kcal/mol), while for R ) CH2 Bu and R )
than the 15(2) kcal/mol barrier in toluene-d8 solution. The high
freezing point of cyclohexane-d12 precludes a similar determi-
nation for the analogous Zr complex 5c. However, in methyl-
CH2TMS, ∆H‡reorg is significantly lower (12(3) and 15(2) kcal/
mol, respectively). The free energies of activation (∆G‡reorg) at
the indicated temperatures (Table 3) are somewhat larger for
cylohexane-d14, ∆H‡
values for 5c and 6c are found to be
reorg
19(1) and 15(1) kcal/mol, respectively, both of which are slightly
larger than the corresponding values in toluene solution. It is
interesting to note that in these low dielectric constant solvents,
reorganization barriers still do not approach the magnitudes of
barriers for complete ion-pair separation calculated for the gas
phase.27c
(28) Solvent dielectric constants (ꢀ) are taken from the following: (a)
CRC Handbook of Chemistry and Physics, 77th ed.; CRC Press: New York,
1996; p G-161 and (b) Timmermans, J. Physico-Chemical Constants of
Pure Organic Compounds; Elsevier Publishing: New York, 1950 (Vol. I)
and 1965 (Vol. II).
(29) (a) Beck, S.; Prosenc, M. H.; Brintzinger, H. H. J. Mol. Catal. A:
Chem. 1998, 41-52. (b) Horton, A. D.; de With, J. Organometallics 1997,
16, 5424-5436. (c) Wang, Q.; Gillis, D. J.; Quyoum, R.; Jeremic, D.;
Tudoret, M.-J.; Baird, M. C. J. Organomet. Chem. 1997, 527, 7-14. (d)
Bochmann, M.; Lancaster, S. J.; Hursthouse, M. B.; Malik, K. M. A.
Organometallics 1994, 13, 2235-2243.
Alkyl Substituent Rotational Barriers. Steric congestion
engendered by the R substituent bulk was also evaluated by
1
variable-temperature H NMR spectroscopy. The experiments

10366 J. Am. Chem. Soc., Vol. 122, No. 42, 2000
Beswick and Marks
reveal that the rotational barrier about the ZrsCH(TMS)2 bond
in neutral precursor complex 3d is rather large (∆G‡ ) 17.4(2)
kcal/mol at 85 °C). At 25 °C, four Cp-Me signals and two TMS
1H NMR signals are observed, as expected, for the proposed
static structure shown in eq 6. Upon heating to 110 °C, the TMS
signals broaden, coalesce, and form one signal, as
expected for fast exchange between rotamers. Solvent and
instrumental constraints preclude observation of the expected
pair of Cp-Me signals at high rotation rates, and only a broad,
coalesced signal is observed at 110 °C. NOE measurements on
5d, the metallocenium derivative of 3d, at 25 °C indicate that
the CH(TMS)2 methine proton is in close proximity to one of
the 1,2-Me2Cp rings (cf. Figure 2B). Furthermore, low-temper-
ature 1H NMR spectra of 5d at -126 °C in CDCl2F reveal four
magnetically nonequivalent ring-methyl group signals (Figure
4) and four TMS signals (integration: 3:1:1:1), indicative of
two magnetically distinct TMS groups. The pattern where three
distinct Me resonances are observed for one TMS group is
consistent with the instantaneous structure in Figure 2B, having
the added feature of restricted rotation about one CHsTMS
bond and suggestive of an agostic interaction12 as shown in
structure III or severe steric interactions or some combination
of both. Interactions of this type have been observed in
isoelectronic, neutrally charged lanthanide complexes.7a,b,30
Further support for the 5d rotomer configuration shown in
Figure 2B is the large downfield shift of the CH(TMS)2 methine
1H NMR signal (δ 3.09 ppm), presumably reflecting proximity
to the electrophilic Zr center and Cp ring currents. A similar
downfield methine shift is observed in neutral precursor 3d.
Figure 5. Born-Haber cycle for formation of L2MR+ MeB(C6F5)3
-
ion pairs from neutral metallocene and borane precursors.
somewhat higher rates of ethylene polymerization than iso-
structural Hf complexes. In the B(C6F5)2Ar series,4b higher
borane Lewis acidity leads to more exothermic ∆Hform values,
presumably via the borane methide affinity (∆HB-Me) in the
cycle (Figure 5).4b,10c Finally, increasing solvent dielectric
constant was found to decrease ∆H‡reorg, presumably because
polar solvents stabilize the charge separation in the transition
state.4b
The present discussion focuses on how alkyl group (R)
electronic and steric properties influence metallocenium ion-
pair thermodynamic stability and structural dynamics and, by
inference, the efficiency of catalyst activation, polymerization
activity, and possibly some aspects of stereoregulation. Ther-
modynamic and kinetic parameters of the ion-pair chemistry
are shown to be highly sensitive to the nature of the alkyl
substituent. Important relationships to polymerization chemistry
are also addressed including metal effects on â-methyl elimina-
tion.
Thermodynamics of Ion-Pair Formation/Catalyst Activa-
tion. R-dependent ion-pair formation thermodynamic param-
eters, as described by eqs 1 and 3 and reported in Table 1, can
be interpreted in terms of the aforementioned Born-Haber
cycle, making the pragmatic and physically reasonable assump-
tion that solvation effects in hydrocarbon solvents are small and
constant for a homologous series.4b,10c The component param-
eters of ∆Hform include the metal-methyl homolytic bond
dissociation enthalpy, D[L2M(R)-Me]; the ionization potential
of the trivalent metallocene fragment, IP (L2MR•); the electron
affinity of the methyl radical, EA(Me•);31 the methide affinity
of the borane, ∆HB-Me; and the ion-pair separation enthalpy,
∆Hips. For the present series with B(C6F5)3 as the sole methide
Discussion
A long-term goal of our single-site polymerization catalyst
research has been to investigate and better understand those
factors governing catalyst thermodynamic stability and ion-pair
structural dynamics in metallocenium systems. Key energetic
components of ∆Hform can be described by an approximate
Born-Haber cycle (Figure 5). The capacity of ancillary ligand
framework modifications such as L2 ) Cp2 f (1,2-Me2Cp)2
f (Me5Cp)2 to lower metallocene ionization potentials (IPs)
was shown previously to correlate with thermodynamic stabi-
lization of the corresponding L2MMe+MeB(C6F5)2Ar- ion pairs
(M ) Zr, Hf).4b Electron-donating and sterically sizeable (Me5-
Cp)2 ancillary ligation was found to be especially effective in
promoting ion-pair formation. Metal identity influences IP and
D[L2M(Me)-Me] in the cycle in that both values are slightly
lower for M ) Zr than for Hf. It was found4b that Zr-Me+ ion
pairs are more thermodynamically stable with respect to neutral
precursors (∆Hform is more exothermic), have somewhat lower
ion-pair reorganization barriers (eq 5; ∆H‡reorg), and exhibit
(30) (a) Klooster, W. T.; Brammer, L.; Schaverien, C. J.; Budzelaar, P.
H. M. J. Am. Chem. Soc. 1999, 121, 1381-1382. (b) Koga, N.; Morokuma,
K. J. Am. Chem. Soc. 1988, 110, 108-112. (c) Di Bella, S.; Lanza, G.;
Fragala`, I. L.; Marks, T. J. Organometallics 1996, 15, 205-208. (d) Haar,
C. M.; Stern, C. L.; Marks, T. J. Organometallics 1996, 15, 1765-1784.
(e) Giardello, M. A.; Conticello, V. P.; Brard, L.; Sabat, M.; Rheingold, A.
L.; Stern, C. L. J. Am. Chem. Soc. 1994, 116, 10212-10240. (f) Di Bella,
S.; Gulino, A.; Lanza, G.; Fragala`, I. L.; Stern, D.; Marks, T. J.
Organometallics, 1994, 13, 3810-3815.
(31) Ellison, G. B.; Engelking, P. C.; Lineberger, W. C. J. Am. Chem.
Soc. 1978, 100, 2556-2558.

Zr and Hf Metallocenium Ion Pairs
J. Am. Chem. Soc., Vol. 122, No. 42, 2000 10367
37
t
38
Table 4. Literature Ionization Potential Values for Metal
Complexes
) -0.29 eV), from SnMe4 to Sn(CH2 Bu)4 and Sn(CH2-
TMS)439 (∆IP ) -0.22 and -0.18 eV, respectively), and from
40
40
complex
GeCl4
ionization potential (eV)a,b
ref
CdMe2 to Cd(CH2TMS)2 (∆IP ) 0.0 eV), indicates the
t
progression R ) Me f CH2 Bu and CH2TMS transfers
11.88
9.33
9.01
7.75
11.5
8.89
8.67
8.71
10.4
7.42
8.33
8.58
8.33
8.64
8.60
7.55
7.18
8.89
8.51
8.58
5.72
9.3
36
36
36
39
42
37
38
39
41
41
36
36
36
36
37
38
38
37
36
36
33
40
41
40
40
increasing electron density to the metal center (more effectively
stabilizes positive charge). Furthermore, the diminution of IP
Ge(Me)4
t
Ge(CH2 Bu)4
40
41
Ge(CH[TMS]2)2
SnCl4
values from HgMe2 to Hg[CH(TMS)2]2 (∆IP ) -1.2 eV)
argues that CH(TMS)2 substitution can depress IPs by ∼28 kcal/
mol vs R ) Me. Lowering of IP values for R ) CH(TMS)2 is
also observed from SnMe442 to Sn[CH(TMS)2]241 (∆IP ) -1.5
eV), which exceeds the depression expected for a reduction in
Sn(Me)4
t
Sn(CH2 Bu)4
Sn(CH2TMS)4
SnCl2
42
41
Sn oxidation state as estimated from SnCl4 vs SnCl2 (∆IP
) -1.1 eV). Taken in total, the above data argue that the
Sn(CH[TMS]2)2
t
Ti(CH2 Bu)4
t
Ti(CH2TMS)4
progression R ) Me f CH2 Bu, CH2TMS f CHTMS2, will
t
Zr(CH2 Bu)4
have a similar effect in depressing the group 4 L2MR IP values
in Figure 5.
Zr(CH2TMS)4
Cp2ZrCl2
The thermochemical data indicate that ∆Hform values for the
present metallocenium systems are strongly influenced both by
metal alkyl and cyclopentadienyl substituents. Versus (1,2-Me2-
Cp2)ZrMe2, -CH(TMS)2 substitution at the zirconocene center
(Me5Cp)2ZrCl2
(Me5Cp)2ZrMe2
Cp2HfCl2
t
Hf(CH2 Bu)4
Hf(CH2TMS)4
(1,3-tBu2Cp)2ZrI
Hg(Me)2
-
increases the exothermicity of CH3 abstraction by 34.6 kcal/
mol, while permethylation of the Cp ligands increases ∆Hform
by somewhat less, 12.1 kcal/mol. The aforementioned IP data
and the Born-Haber analysis are consistent with the observed
large alkyl group electronic effects on ∆Hform. However, note
Hg(CH[TMS]2)2
Cd(Me)2
8.12
8.8
Cd(CH2TMS)2
8.8
-
that stabilization via release of steric congestion upon CH3
a 1.00 eV ) 23.1 kcal/mol. b Uncertainties are typically reported as
(0.1 eV or less.
abstraction is also likely operative in (1,2-Me2Cp)2Zr(Me)CH-
(TMS)2 and (Me5Cp)2ZrMe2. Within the Born-Haber cycle
description, such strain would most likely be manifested in
diminished D(L2MR-CH3) values, although IP may be influ-
enced as well. For (1,2-Me2Cp)2ZrCH(TMS)2+, it is likely that
agostic interaction III (vide supra), identified in the low-
abstractor, both the methyl radical electron affinity and the
borane methide affinity are invariant. Therefore, parameters
describing methide abstraction enthalpy differences between two
different metallocene dialkyls differing in R group can be
described by eq 7 where L2 ) (1,2-Me2Cp)2
temperature H NMR, imparts additional stabilization.43
1
∆Hips, the Coulombic and covalent stabilization accrued by
joining the metallocenium cation and the methylborate anion
(Figure 5),27 is the other major parameter defining ∆Hform. While
we know of no way to measure this parameter directly, ring or
R steric bulk which enforced greater cation-anion separation
would doubtless be electrostatically destabilizing. This qualita-
tive contention is supported by recent ab initio computational
studies on the R2Si(R′4C5)(R′′N)TiR′′′‚H3CB(C6F5)3 system to
be discussed elsewhere.27c To the extent that the magnitudes of
∆∆Hform ) {D[L2M(R)-Me] - D[L2M(R)-Me]} +
{IP - IP} - {∆Hips - ∆Hips} (7)
Bond dissociation enthalpy values for D[(Me5Cp)2Zr(Me)-
Me], D[Cp2Zr(Me)-Me], and D[(Me5Cp)2Hf(Me)-Me] of 67.2-
(1.0), 67.0(1.0), and 72.0(3.7) kcal/mol, respectively,32 argue
that M-Me bond enthalpies are relatively insensitive to
cyclopentadienyl ancillary ligand substitution (excepting cases
with extreme steric congestion33) and metal identity. The limited
thermochemical data available for Cp2M(X)R/Cp*M(X)R com-
plexes32 suggest that X = alkyl effects on D[L2M(R)-Me] should
be relatively small. Ionization potentials for trivalent Zr and
Hf metallocenes as a function of σ-alkyl substituent such as
the present measured kinetic ∆H‡
parameters of Table 3
reorg
index the “tightness” of ion pairing and hence would qualita-
tively track ∆Hips, then the decline in ∆H‡
values from R
reorg
t
) Me to R ) CH2 Bu ∼ CH2TMS by ∼4-5 kcal/mol and from
R ) Me to R ) CH(TMS)2 by ∼13 kcal/mol argues that the
t
those addressed here (R ) Me, CH2 Bu, CH2TMS, and CH-
(35) Evans, S.; Green, J. C.; Joachim, P. J.; Orchard, A. F. J. Chem.
Soc., Faraday Trans. 2 1972, 68, 905-911.
(36) Kochi, J. K. Pure Appl. Chem. 1980, 52, 571-605.
(37) (a) Green, J. C. Struct. Bonding (Berlin) 1981, 43, 37-112. (b)
Condorelli, G.; Fragala´, I.; Centineo, A.; Tondello, E. J. Organomet. Chem.
1975, 87, 311-315.
(38) Ciliberto, E.; Condorelli, G.; Fagan, P. J.; Manriquez, J. M.; Fragala,
I.; Marks, T. J. J. Am. Chem. Soc. 1981, 103, 4755-4759.
(39) Evans, S.; Green, J. C.; Jackson, S. E. J. Chem. Soc., Faraday Trans.
2 1973, 69, 191-195.
(40) Creber, D. K.; Bancroft, G. M. Inorg. Chem. 1980, 19, 643-648.
(41) Harris, D. H.; Lappert, M. F.; Pedley, J. B.; Sharp, G. J. J. Chem.
Soc., Dalton Trans. 1976, 945-950.
(42) Bancroft, G. M.; Pellach, E.; Tse, J. S. Inorg. Chem. 1982, 21,
2950-2955.
(43) There is a literature precedent for electronic stabilization effects by
this substituent: (a) Bassindale, A. R.; Taylor, P. G. in The Chemistry of
Organic Silicon Compounds; Patai, S.; Rappoport, Z., Eds.; Wiley:
Chichester, U.K., 1989; Chapter 14, pp 893-956. (b) Fleming, I. in
ComprehensiVe Organic Chemistry; Jones, D. N., Ed.; Pergamon Press:
Oxford, U.K., 1979; Chapter 13, pp 541-671.
[TMS]2) are not readily available (only a single IP value is
available for a trivalent zirconocene, [1,3-tBu2Cp]2ZrI, 5.72
eV33). Table 4 collects available IP values for homoleptic and
Zr and Hf metallocene complexes as well as for main group
elements with similar alkyl substituents. It can be seen that Zr
vs Hf IP data differ only modestly for identical ligation,34-37
while Cp methylation reduces the IP substantially.38 The trend
36
t
36
of decreasing IP values from GeMe4 to Ge(CH2 Bu)4 (∆IP
(32) (a) Schock, L. E.; Marks, T. J. J. Am. Chem. Soc. 1988, 110, 7701-
7715. (b) Simoes, J. A. M.; Beauchamp, J. L. Chem. ReV. 1990, 90, 629-
688.
(33) For an example where extreme metallocene intramolecular non-
bonded repulsion can affect bonding energetics, see: King, W. A.; DiBella,
S.; Gulino, A.; Lanza, G.; Fragala`, I. L.; Stern, C. L.; Marks, T. J. J. Am.
Chem. Soc. 1999, 121, 355-366.
(34) (a) Lappert, M. F.; Pedley, J. B.; Sharp, G. J. Organomet. Chem.
1974, 66, 271-278. (b) Bassett, P. J.; Lloyd, D. R. J. Chem. Soc. A 1971,
641-645.

10368 J. Am. Chem. Soc., Vol. 122, No. 42, 2000
Beswick and Marks
Figure 6. Reaction coordinate for formation and structural reorganization of (1,2-Me2Cp)2M(R)+ MeB(C6F5)3 ion-pairs where R ) Me, CH2-
TMS, and CH(TMS)2 in toluene solution unless otherwise noted. Features representative of ∆H‡
processes are shown as a “double hump”, in
-
form
accord with ref 4b. No information is available on ∆Hreorg
.
ion-pairing interactions are significantly greater and hence
enhance ∆Hform for smaller, less-electron-donating alkyl
substituents, which allow closer cation-anion approach/interac-
tion.
4. Polar solvents lower the barrier to ion-pair reorganization;
however, the effect is greatest for sterically unencumbered R
) Me. This may reflect inhibition of cation solvation or specific
solvent complexes (e.g., IV) in the case of larger R. The present
results include the first measurements of ∆H‡reorg in a saturated
hydrocarbon solution. The barrier is increased slightly (2-4
kcal/mol) versus that in toluene (Table 3, complexes 5c and
6c).
Ion-Pair Formation and Reorganization Reaction Coor-
dinates. The schematic potential surfaces in Figure 6 illustrate
the relative energetics of ion-pair formation from the neutral
precursors and reorganization symmetrization as a function of
R and solvent. Combined with earlier data for (1,2-Me2-
Cp)2MMe+MeB(C6F5)- and other L2MMe+MeB(C6F5)3
-
3
systems,4b these results and those in Tables 1 and 3 lead to the
following generalizations:
1. Sterically encumbered alkyl groups can greatly stabilize
metallocenium ion-pairs with respect to the neutral precursors.
2. For alkyl groups more closely approximating growing poly-
(R-olefin) chains, the stabilization versus R ) Me is less (∼0
kcal/mol for M ) Zr, R ) CH2TMS; ∼10 kcal/mol for M )
Hf, R ) CH2TMS).
3. All alkyl groups examined depress the barrier to ion-pair
reorganization (eq 5; ∆H‡reorg) versus R ) Me. The effect is
largest for sterically encumbered R groups.
5. In the (1,2-Me2Cp)2MMe+MeB(C6F5)3- series,4b ∆H‡
reorg
is only less than |∆Hform| and |-∆Hform + ∆H‡form| (estimated
formation barriers from the neutrals are small4b) for M ) Zr.

Zr and Hf Metallocenium Ion Pairs
J. Am. Chem. Soc., Vol. 122, No. 42, 2000 10369
Table 5. Propylene Polymerization Rates for Selected
However, for the larger R substituents of the present series,
∆H‡reorg is invariably less than |∆Hform|. Hence, reorganization
of the ion pair is substantially more rapid under the present
conditions than borane dissociation-reassociation.
Metallocenium Catalystsa
turnover
frequency
catalyst
T(°C)
(s-1
)
â-Methyl Elimination Kinetics. â-Alkyl elimination is an
important chain-transfer mechanism in single-site R-olefin
polymerization/oligomerization and can be the predominant
mechanism under conditions such as low monomer concentra-
tions or for specific catalysts.26e,f,h â-Me elimination can be
kinetically favored over â-H elimination in cases of specific
steric interactions, even though â-H elimination is thermody-
namically favored by ∼10 kcal/mol.26f,32 In the present com-
plexes, â-Me elimination is only realistically possible for
Me2C(Cp)(Flu)ZrMe2, + B(C6F5)3
Me2C(Cp)(Flu)ZrMe2, + B(C12F9)3
rac-Me2Si(Ind)2ZrMe2, + B(C6F5)3
rac-Me2Si(Ind)2ZrMe2, + Ph3C+B(C6F5)4
rac-Me2Si(Ind)2ZrMe2, + B(C12F9)3
rac-Me2Si(Ind)2ZrMe2, + B(C6F5)3
rac-Me2Si(Ind)2ZrMe2, + Ph3C+B(C6F5)4
rac-Me2Si(Ind)2ZrMe2, + B(C12F9)3
24
24
24
24
24
60
60
60
2
2
12
38
62
15
126
106
-
-
a Polymerizations carried out in toluene solution at 1 atm of
propylene under conditions designed to minimize mass-transfer effects.
See ref 4c.
t
-
(1,2-Me2Cp)2ZrCH2 Bu+ MeB(C6F5)3 (5c) and (1,2-Me2-
t
-
Cp)2HfCH2 Bu+ MeB(C6F5)3 (6c). In this case, the process
obeys first-order kinetics, and the enthalpy of activation for Zr
R ) CH2TMS (rate ) 160 s-1) is likely a more realistic
approximation of a growing polypropylene chain.47 Rates of
propylene enchainment in these and related complexes vary
depending upon conditions. As an example, the rate of monomer
enchainment is ∼130 s-1 for Ewen’s catalyst at 25 °C in neat
polypropylene.44b,48 Propylene polymerization turnover frequen-
cies for a number of borane-activated metallocenes are compiled
is higher than that for Hf, ∆H‡
) 22.5(0.9) kcal/mol vs
â-Me
17.3(0.9) kcal/mol, respectively. However, this difference is
partly offset by the corresponding entropies of activation,
∆S‡
) 4.3(3.3) cal/mol‚K vs -11.9(3.4) cal/mol‚K, re-
â-Me
spectively (Table 2). Thus, the â-Me elimination rates for Zr
and Hf are comparable near 0 °C as indicated by ∆G‡â-Me values
(Table 2), while at higher temperatures â-Me elimination in the
Hf complex becomes less favored versus the Zr complex. There
is also a strong dependence on ancillary ligation. For example,
in Table 5 and range from 4 to 126 s-1 4c
. If measured rates of
ion-pair reorganization (if synonymous with “chain swinging”)
were vastly lower than measured enchainment rates, it is difficult
to envision how this process could be coupled to the stereose-
lective enchainment mechanisms discussed for such systems.49
On the other hand, if ion-pair reorganization rates were far in
excess of enchainment rates, then coupling of chain swinging
to enchainment at the observed rates is not implausible; however,
site isomerization prior to enchainment could lead to stereoer-
rors. In either case, the present data indicate that the rates of
the two processes are not vastly different. This issue is being
explored further in ongoing experimental and theoretical studies.
t
the analogous complex (Me5Cp)2ZrCH2 Bu+ MeB(C6F5)3- can-
not be observed by in situ NMR at temperatures as low as -75
°C because of the facile â-Me elimination (the complex can be
stabilized as THF or RCN adducts).26c In contrast, significantly
less hindered Cp2ZrCH2 Bu+ MeB(C6F5)3 is reported to be
stable at 0 °C in toluene solution.26c
t
-
Kinetic Comparison of Ion-Pair Reorganization and
Olefin Enchainment Rates. Potentially, structural reorganiza-
tion of the metallocenium ion-pair is intimately connected with
olefin enchainment. A relevant example is the methylalumoxane
(MAO) activated form of Ewen’s iPr[Cp-1-Flu]ZrCl2 propylene
polymerization catalyst (V, eq 8; X- ) “MAO-”).5,44,45
Conclusions
The dependence of methide abstraction thermodynamics and
catalyst ion-pair structural reorganization dynamics on the size
and electronic structure of the metal-bound R substituent has
been shown to be substantial for the series of complexes having
the formula (1,2-Me2Cp)2MR+ MeB(C6F5)3-. The sterically
encumbered R ) CH(TMS)2 group strongly promotes ion-pair
formation by stabilization of the metal cationic charge and by
release of steric strain from the neutral metallocene precursor.
Ion-pair reorganization barriers generally decrease as the alkyl
substituent increases in size, and barriers are further lowered
by solvents of higher dielectric constant, able to stabilize charge-
separated intermediates. â-Me elimination activation parameters
The high syndiospecificity in propylene polymerization is
thought to result from (1) alternating olefin insertions at the
enantiomeric “left” and “right” catalyst coordination sites and
(2) enantiofacial orientation of each inserting propylene unit as
a result of steric constraints imposed by the ancillary ligand
and propagating chain structure. This mechanism requires that
the growing polymer chain “swing” from side to side concurrent
with each insertion, analogous to the ion pair reorganization
depicted in eq 5. If this is the case, then it may be expected
that the MAO-based counteranion9,46 undergoes concurrent
repositioning with each insertion. Rates in toluene solution for
methylborate anion/R repositioning in the present (1,2-Me2-
t
for R ) CH2 Bu complexes are higher for Zr than for Hf;
however, higher temperatures enhance elimination in Zr as a
result of entropic contributions. Finally, rates of ion-pair
t
(47) R ) CH2 Bu is equally representative of a polymer chain as R )
CH2TMS. Both exhibit similar ∆H‡reorg and ∆G‡reorg parameters (Table 3).
(48) Olefin polymerization rates are sensitive to the reaction conditions.
For comparison, the same catalyst activated with MAO at 50 °C, 100 psi
of propylene, and in toluene solution exhibits a turnover frequency of ∼1600
s-1 (Ewen, J. A.; Elder, M. J.; Jones, R. L.; Haspeslagh, L.; Atwood, J. L.;
Bott, S. G.; Robinson, K. Makromol. Chem., Macromol. Symp. 1999, 48/
49, 253-295). For comparison, a reasonably active metallocene ethylene
polymerization catalyst activated with B(C6F5)3 at 25 °C, 1 atm of ethylene
in toluene solution exhibits a turnover frequency of 50 s-1 (See ref 4b).
(49) We cannot exclude the possibility that the presence of olefin might
“loosen” the ion pairing in some noncoordinative (i.e., without embarking
on the enchainment reaction coordinate) or coordinative but nonpreinsertive
manner. However, it is difficult to envision how this type of mobilization
could be more effective than neat aromatic solvents.
-
Cp)2ZrR+ MeB(C6F5)3 complexes range at 25 °C from ∼0.2
s-1 for R ) Me to 2.6 × 107 s-1 for R ) CH(TMS)2; however,
(44) (a) Ewen, J. A. J. Mol. Catal. A: Chem. 1998, 128, 103-109. (b)
Ewen, J. A.; Jones, R. L.; Razavi, A. J. Am. Chem. Soc. 1988, 110, 6255-
6256.
(45) Krauledat, H.; Brintzinger, H. H. Angew. Chem., Int. Ed. Engl. 1990,
29, 1412-1413.
(46) MAO solutions are complex oligomeric mixtures; however, recent
evidence suggests that a major role is anionic stabilization of the cationic
metal center. See refs 9a,c,d.

10370 J. Am. Chem. Soc., Vol. 122, No. 42, 2000
Beswick and Marks
stereomutation and propylene polymerization are not vastly
different, suggesting that chain-swinging/anion motion may be
closely coupled in some stereoselective chain propagation
mechanisms.
Dr. C. G. Fry of the University of Wisconsin, Madison (NSF
CHE-9629688) for assistance with NMR experiments below
-100 °C, and Dr. L. Li and the Dow Chemical Company for
gifts of B(C6F5)3.
Acknowledgment. We thank the DOE (Grant DE-FG02-
86ER1351) for financial support of this research. We also thank
JA000810A