Synthesis, structures, and reactivity of weakly coordinating anions with delocalized borate structure: The assessment of anion effects in metallocene polymerization catalysts

Article abstract of DOI:10.1021/ja002820h

The formation of adducts of tris(pentafluorophenyl)borane with strongly coordinating anions such as CN^- and [M(CN)₄]^2- (M = Ni, Pd) is a synthetically facile route to the bulky, very weakly coordinating anions [CN{B(C

Full text of DOI:10.1021/ja002820h

J. Am. Chem. Soc. 2001, 123, 223-237
223
Synthesis, Structures, and Reactivity of Weakly Coordinating Anions
with Delocalized Borate Structure: The Assessment of Anion Effects
in Metallocene Polymerization Catalysts
Jiamin Zhou, Simon J. Lancaster,† Dennis A. Walker,† Stefan Beck,†
Mark Thornton-Pett, and Manfred Bochmann*,†
Contribution from the School of Chemistry, UniVersity of Leeds, Leeds LS2 9JT, U.K.
ReceiVed July 31, 2000
Abstract: The formation of adducts of tris(pentafluorophenyl)borane with strongly coordinating anions such
as CN- and [M(CN)4]2- (M ) Ni, Pd) is a synthetically facile route to the bulky, very weakly coordinating
anions [CN{B(C6F5)3}2]- and [M{CNB(C6F5)3}4]2- which are isolated as stable NHMe2Ph+ and CPh3+ salts.
The crystal structures of [CPh3][CN{B(C6F5)3}2] (1), [CPh3][ClB(C6F5)3] (2), [NHMe2Ph]2[Ni{CNB(C6F5)3}4]‚
2Me2CO (4b‚2Me2CO), [CPh3]2[Ni{CNB(C6F5)3}4]‚2CH2Cl2 (4c‚2CH2Cl2), and [CPh3]2[Pd{CNB(C6F5)3}4]‚
2CH2Cl2 (5c‚2CH2Cl2) are reported. The CN stretching frequencies in 4 and 5 are shifted by ∼110 cm-1 to
higher wavenumbers compared to the parent tetracyano complexes in aqueous solution, although the M-C
and C-N distances show no significant change on B(C6F5)3 coordination. Zirconocene dimethyl complexes
L2ZrMe2 [L2 ) Cp2, SBI ) rac-Me2Si(Ind)2] react with 1, 4c or 5c in benzene solution at 20 °C to give the
salts of binuclear methyl-bridged cations, [(L2ZrMe)2(µ-Me)][CN{B(C6F5)3}2] and [(L2ZrMe)2(µ-Me)]2[M{CNB-
(C6F5)3}4]. The reactivity of these species in solution was studied in comparison with the known [{(SBI)-
ZrMe}2(µ-Me)][B(C6F5)4]. While the latter reacts with excess [CPh3][B(C6F5)4] in benzene to give the
mononuclear ion pair [(SBI)ZrMe+‚‚‚B(C6F5)4-] in a pseudo-first-order reaction, k ) 3 × 10-4 s-1, [(L2-
ZrMe)2(µ-Me)][CN{B(C6F5)3}2] reacts to give a mixture of L2ZrMe(µ-Me)B(C6F5)3 and L2ZrMe(µ-NC)B-
(C6F5)3. Recrystallization of [Cp′′2Zr(µ-Me)2AlMe2][CN{B(C6F5)3}2] affords Cp′′2ZrMe(µ-NC)B(C6F5)3 6, the
X-ray structure of which is reported. The stability of [(L2ZrMe)2(µ-Me)]+X- decreases in the order X )
[B(C6F5)4] > [M{CNB(C6F5)3}4] > [CN{B(C6F5)3}2] and increases strongly with the steric bulk of L2 )
Cp2 , SBI. Activation of (SBI)ZrMe2 by 1 in the presence of AlBui3 gives extremely active ethene
polymerization catalysts. Polymerization studies at 1-7 bar monomer pressure suggest that these, and by
implication most other highly active ethene polymerization catalysts, are strongly mass-transport limited. By
contrast, monitoring propene polymerization activities with the systems (SBI)ZrMe2/1/AlBui3 and CGCTiMe2/
1/AlBui3 at 20 °C as a function of catalyst concentration demonstrates that in these cases mass-transport limitation
is absent up to [metal] ≈ 2 × 10-5 mol L-1. Propene polymerization activities decrease in the order
[CN{B(C6F5)3}2]- > [B(C6F5)4]- > [M{CNB(C6F5)3}4]2- . [MeB(C6F5)3]-, with differences in activation
barriers relative to [CN{B(C6F5)3}2]- of ∆∆Gq ) 1.1 (B(C6F5)4-), 4.1 (Ni{CNB(C6F5)3}42-) and 10.7-12.8
kJ mol-1 (MeB(C6F5)3-). The data suggest that even in the case of very bulky anions with delocalized negative
charge the displacement of the anion by the monomer must be involved in the rate-limiting step.
Introduction
leads to a dramatic increase in activity and is the basis of the
use of ionic systems [LnMR]+[R′B(ArF)3]- as highly effective
olefin polymerization catalysts.6 In particular, Marks has
demonstrated that suitable elaboration of the perfluoroaryl borate
counteranion can result in significant enhancements in stability
and activity of ionic metallocene catalysts.7,8 However, in view
of the fact that the preparation of such anions is costly and
involves not insignificant synthetic effort, we have become
The profound effect of counteranions on the reactivity and
catalytic activity of cationic early transition-metal complexes
is well-established.1-4 For example, while BPh4- has a tendency
to form π-complexes and tight ion pairs,5 the weak coordinating
ability of its perfluoro analogue [B(C6F5)4]- and its congeners
* To whom correspondence should be addressed. E-mail: m.bochmann@
uea.ac.uk. Fax: +44 1603 592044.
(5) See, for example: (a) Bochmann, M.; Karger, G.; Jaggar, A. J. J.
Chem. Soc., Chem. Commun. 1990, 1038. (b) Horton, A. D.; Frijns, J. H.
G. Angew. Chem., Int. Ed. Engl. 1991, 30, 1152. (c) Schaverien, C. J.
Organometallics 1992, 11, 3476. (d) Calderazzo, F.; Englert, U.; Pampaloni,
G.; Rocchi, L. Angew. Chem., Int. Ed. Engl. 1992, 31, 1235. (e) Solari, E.;
Musso, F.; Gallo, E.; Floriani, C.; Re, N.; Chiesi-Villa, A.; Rizzoli, C.
Organometallics 1995, 14, 2265.
(6) Turner, H. W. Eur. Patent Appl. 0277 004, 1988. (b) Ewen, J. A.;
Elder, M. J. Eur. Patent Appl. 0426 637, 1990. (c) Ewen, J. A.; Elder, M.
J. Eur. Patent Appl. 0427 697, 1991. (d) Yang, X.; Stern, C. L. Marks, T.
J. J. Am. Chem. Soc. 1991, 113, 3623. (e) Ewen, J. A.; Elder, M. J.
Makromol. Chem., Macromol. Symp. 1993, 66, 179. (f) Ewen, J. A. Stud.
Surf. Sci. Catal. 1994, 89, 405.
† Current address: School of Chemical Sciences, University of East
Anglia, Norwich NR4 7TJ, UK.
(1) (a) Lupinetti, A. J.; Strauss, S. H. Chemtracts: Inorg. Chem. 1998,
11, 565. (b) Reed, C. A. Acc. Chem. Res. 1998, 31, 133. (c) Strauss, S. H.
Chem. ReV. 1993, 93, 927.
(2) (a) Bochmann, M.; Jaggar, A. J. J. Organomet. Chem. 1992, 424,
C5. (b) Bochmann, M. Angew. Chem., Int. Ed. Engl. 1992, 31, 1181.
(3) Yang, X.; Stern, C. L.; Marks, T. J. Organometallics 1991, 10, 840.
(4) (a) Siedle, A. R.; Lamanna, W. M.; Newmark, R. A.; Stevens, J.;
Richardson, D. E.; Ryan, M. Makromol. Chem., Macromol. Symp. 1993,
66, 215. (b) Chien, J. C. W.; Song, W.; Rausch, M. D. J. Polym. Sci., Part
A: Polym. Chem. 1994, 32, 2387.
10.1021/ja002820h CCC: $20.00 © 2001 American Chemical Society
Published on Web 12/15/2000

224 J. Am. Chem. Soc., Vol. 123, No. 2, 2001
Zhou et al.
Scheme 1
interested in methods of generating new bulky anions in a more
straightforward manner. One such method is the facile adduct
formation between commercially available B(C6F5)39 and strongly
coordinating anions.10 With this aim in mind we have recently
reported the synthesis of [CPh3][CN{B(C6F5)3}2] (1) and
[CPh3]2[Ni{CNB(C6F5)3}4] (4c) and their use as activators for
zirconocene-based ethene polymerization catalysts in a prelimi-
nary communication.11 In contrast to borates of the type
[R′B(ArF)3]- and [B(ArF)4]-, which contain one negative charge
per boron atom, the charge in the oligoborates 1 and 4 is
delocalized and reduced to 0.5/B (structure A). Both for steric
Figure 1. Structure of the [CN{B(C6F5)3}2]- anion in 1, showing the
atomic numbering scheme. Ellipsoids are drawn at 40% probability
level.
and electronic reasons it seemed reasonable to expect that as
counterions for cationic early transition-metal catalysts such
(7) (a) Jia, L.; Yang, X.; Ishihara, A.; Marks, T. J. Organometallics 1995,
14, 3135. (b) Chen, Y. X.; Stern, C. L.; Yang, S.; Marks, T. J. J. Am. Chem.
Soc. 1996, 118, 12451. (c) Jia, L.; Yang, X.; Stern, C. L.; Marks, T. J.
Organometallics 1997, 16, 842. (d) Chen, Y. X.; Marks, T. J. Organome-
tallics 1997, 16, 3649. (e) Chen, Y. X.; Stern, C. L.; Marks, T. J. J. Am.
Chem. Soc. 1997, 119, 2582. (f) Chen, Y. X.; Metz, M. V.; Li, L.; Stern,
C. L.; Marks, T. J. J. Am. Chem. Soc. 1998, 120, 6287. (g) Li, L.; Marks,
T. J. Organometallics 1998, 17, 3996. (h) Metz, M. V.; Schwartz, D. J.;
Stern, C. L.; Nickias, P. N.; Marks, T. J. Angew. Chem., Int. Ed. 2000, 39,
1312. (i) Chen, E. Y. X.; Marks, T. J. Chem. ReV. 2000, 100, 1391.
(8) For related boranes, see: (a) Ko¨hler, K.; Piers, W. E. Can. J. Chem.
1998, 76, 1249. (b) Ko¨hler, K.; Piers, W. E.; Jarvis, A. P.; Xin, S.; Feng,
Y.; Brasakis, A. M.; Collins, S.; Clegg, W.; Yap, G. P. A.; Marder, T. B.
Organometallics 1998, 17, 3557. (c) Williams, V. C.; Piers, W. E.; Clegg,
W.; Elsegood, M. R. J.; Collins, S.; Marder, T. B. J. Am. Chem. Soc. 1999,
121, 3244. (d) Williams, V. C.; Dai, C.; Li, Z.; Collins, S.; Piers, W. E.;
Clegg, W.; Elsegood, M. R. J.; Marder, T. B. Angew. Chem., Int. Ed. 1999,
38, 3695. (e) Williams, V. C.; Irvine, G. J.; Piers, W. E.; Li, Z.; Collins,
S.; Clegg, W.; Elsegood, M. R. J.; Marder, T. B. Organometallics 2000,
19, 1619.
(9) (a) Massey, A. G.; Park, A. J.; Stone, F. G. A. Proc. Chem. Soc.
[London] 1963, 212. (b) Massey, A. G.; Park, A. J. J. Organomet. Chem.
1964, 2, 245. (c) Pohlmann, J. L. W.; Brinckmann, F. E. Z. Naturforsch. B
1965, 20b, 5.
(10) Anions derived from adducts between B(C6F5)3 and alcohols, silanols
and oximes and their use in polymerization catalysis are known: Siedle,
A. R.; Lamanna, W. M. U.S. Patent 5 ,416, 177, 1995 and ref 4a.
(11) (a) Lancaster, S. J.; Walker, D. A.; Thornton-Pett, M.; Bochmann,
M. Chem. Commun. 1999, 1533. (b) Compound 1 and related anions have
been independently reported in a patent: LaPointe, R. E. WO 99/42467,
1999.
anions would show reduced ion pairing and hence increased
catalytic activity. Preliminary results with (SBI)ZrMe2/1 mix-
tures (SBI ) rac-Me2Si(Ind)2) indicated that this expectation
was justified.11a We report here the evaluation of oligoborates
as activators in high-activity ethene and propene polymerization
catalysts.
Results
Anion Synthesis. Stirring a suspension of KCN with 2 equiv
of B(C6F5)3‚Et2O in diethyl ether at room temperature for 12 h
leads to the slow dissolution of the solid to give a clear solution.
Replacing the ether solvent with dichloromethane and adding
1 equiv of chlorotriphenylmethane gives an orange-yellow
solution from which [CPh3][CN{B(C6F5)3}2] (1) is isolated in
45% yield (Scheme 1). Crystalline 1 is not affected by storing
in air at ambient conditions for extended periods of time.
Crystals suitable for X-ray diffraction were grown from dichlo-
romethane/light petroleum mixtures. The structure is shown in
Figure 1, and selected bond lengths and angles are given in
Table 1. The CN-bridged geometry is further confirmed by the
observation of two 11B NMR signals at δ -11.94 and -21.68
and the νCN stretching frequency at 2305 cm-1. Structurally
related cyanoborates are known, viz. [CN(BH3)2]- and [CN-

Anion Effects in Olefin Polymerization
J. Am. Chem. Soc., Vol. 123, No. 2, 2001 225
Table 1. Selected Bond Distances [Å] and Angles [deg]
[CPh3][CN{B(C6F5)3}2] 1
that used for 1 fails to give the nitrate adduct but generates
[CPh3][ClB(C6F5)3] (2) in good yield. The same product is
obtained in the absence of KNO3. The structure of 2 was
confirmed by X-ray crystallography; the anion is identical to
that obtained by Erker et al. from the reaction of (Ar2PCMe2-
Cp)2ZrCl2 and B(C6F5)3 (Ar ) p-tolyl).15 By contrast, the
reaction of [PhNMe2H][HSO4] with B(C6F5)3‚Et2O in dichlo-
romethane affords [PhNMe2H][HSO3{OB(C6F5)3}] (3).16
The tetracyanometalates [M(CN)4]2- are characterized by very
high formation constants in aqueous solution, 1031 for M ) Ni
and 1042 - 1051.7 given for M ) Pd.17 These anions therefore
promised to produce very stable adducts with B(C6F5)3. Stirring
a mixture of K2[Ni(CN)4] with B(C6F5)3‚Et2O in dichloro-
methane did indeed give the expected product, K2[Ni{CNB-
(C6F5)3}4] (4a), as a white solid. Cation exchange with [NH-
Me2Ph]Cl afforded the corresponding dimethylanilinium salt 4b,
B(1)-N(1) 1.593(2)
B(2)-C(2) 1.583(2)
C(2)-N(1) 1.144(2)
B(1)-C(111) 1.1636(2)
B(1)-C(131) 1.1644(2)
N(1)-C(2)-B(2) 174.50(14)
B(1)-C(121) 1.1646(2)
C(2)-N(1)-B(1) 173.55(14)
N(1)-B(1)-C(111) 108.41(11) N(1)-B(1)-C(121) 103.71(11)
N(1)-B(1)-C(131) 104.75(11) C(111)-B(1)-C(121) 111.02(11)
[CPh3][ClB(C6F5)3] 2
B-Cl 1.928(2)
B-C(21) 1.649(3)
B-C(11) 1.646(3)
B-C(31) 1.648(3)
C(11)-B-Cl 105.10(14)
C(31)-B-Cl 103.31(14)
C(11)-B-C(31) 115.5(2)
C(21)-B-Cl 111.69(14)
C(11)-B-C(21) 107.7(2)
C(21)-B-C(31) 113.2(2)
[HNMe2Ph]2[Ni{CNB(C6F5)3}4]‚2Me2CO 4b
Ni-C(1) 1.851(2)
Ni-C(2) 1.856(2)
+
while the CPh3 salt 4c was made directly from Ph3CCl, K2-
C(1)-N(1) 1.143(3)
N(1)-B(1) 1.582(3)
[Ni(CN)4] and B(C6F5)3‚Et2O. Crystals of 4b‚2Me2CO were
grown from dichloromethane containing small amounts of
acetone. The palladium compounds A2[Pd{CNB(C6F5)3}4] 5a,
(A ) K) and 5c (A ) CPh3) were obtained by analogous
procedures as crystalline solids.
Coordination of B(C6F5)3 to [M(CN)4]2- significantly raises
the C-N stretching frequencies. The nickel complexes 4 show
infrared absorptions at ∼2236 cm-1, essentially independent of
the cation, compared to 2123.5 cm-1 for K2Ni(CN)4 in aqueous
solution.18 The corresponding frequency for the Pd complex 5c
is observed at 2243 cm-1, ∼110 cm-1 higher than that in K2-
[Pd(CN)4] (2135.8 cm-1).
The structures of 4b‚2Me2CO, 4c‚2CH2Cl2, and 5c‚2CH2Cl2
were determined by X-ray diffraction. In all cases the anions
are square-planar; the anion of 5c is shown in Figure 2. The
M-C and C-N distances of the B(C6F5)3 adducts are remark-
ably similar to those found in the solid-state structures of K2-
Ni(CN)4 and Rb2Ni(CN)4‚H2O;19,20 for example, the average
Ni-C distance in 4c (1.856(2) Å) is essentially identical to that
in K2Ni(CN)4 (1.87(3) Å). The C-N bond distances in K2Ni-
(CN)4 and Rb2Ni(CN)4‚H2O are almost identical to those in 4b,
4c, and 5c, (as far as is possible to judge in view of the
comparatively large standard deviations in the parent cyano
complexes). A comparison of bond distances in cyano com-
plexes is given in Table 2.
C(2)-N(2) 1.143(3)
N(2)-B(2) 1.578(3)
B(1)-C(111) 1.636(3)
C(1)-Ni-C(1)* 180.0
C(1)-Ni-C(2) 90.56(9)
N(2)-C(2)-Ni 174.8(2)
C(2)-N(2)-B(2) 178.2(2)
B(1)-C(121) 1.639(3)
C(1)-Ni-C(2)* 89.44(9)
N(1)-C(1)-Ni 178.3(2)
C(1)-N(1)-B(1) 177.5(2)
N(1)-B(1)-C(111) 105.9(2)
[CPh3]2[Ni{CNB(C6F5)3}4]‚2CH2Cl2 4c
Ni-C(1) 1.855(2)
Ni-C(2) 1.862(2)
C(1)-N(1) 1.145(2)
N(1)-B(1) 1.574(2)
C(2)-N(2) 1.146(2)
N(2)-B(2) 1.574(2)
B(1)-C(111) 1.635(3)
C(1)-Ni-C(1)* 180.0
C(1)-Ni-C(2) 87.15(7)
N(2)-C(2)-Ni 175.2(2)
C(2)-N(2)-B(2) 175.6(2)
B(1)-C(121) 1.642(3)
C(1)-Ni-C(2)* 92.85(7)
N(1)-C(1)-Ni 176.5(2)
C(1)-N(1)-B(1) 174.8(2)
N(1)-B(1)-C(111) 103.65(14)
[CPh3]2[Pd{CNB(C6F5)3}4]‚ 2CH2Cl2 5c
Pd-C(1) 1.989(2)
Pd-C(2) 1.995(2)
C(1)-N(1) 1.126(3)
N(1)-B(1) 1.577(3)
C(2)-N(2) 1.126(3)
N(2)-B(2) 1.571(3)
B(1)-C(111) 1.639(3)
C(1)-Pd-C(1)* 180.0(2)
C(1)-Pd-C(2) 92.95(8)
N(2)-C(2)-Pd 174.7(2)
C(2)-N(2)-B(2) 175.5(2)
B(1)-C(121) 1.645(3)
C(1)-Pd-C(2)* 87.05(8)
N(1)-C(1)-Pd 176.6(2)
C(1)-N(1)-B(1) 175.3(2)
N(1)-B(1)-C(111) 106.4(2)
Cp′′2Zr(Me)NCB(C6F5)3‚C7H8 6
Zr-C(14) 2.254(3)
Zr-N(1) 2.259(2)
Reactivity Studies. The solution chemistry of the new
activators 1 and 4 in reaction with various metallocenes was
studied by NMR spectroscopy.
Some time ago we showed that the reaction of [CPh3]-
[B(C6F5)4] with L2ZrMe2 (L2 ) Cp2 or SBI; SBI ) rac-Me2-
Si(Ind)2) in dichloromethane at -60 °C formed the methyl-
N(1)-C(16) 1.149(3)
B(1)-C(111) 1.639(4)
Zr-C(2) 2.511(3)
C(16)-B(1) 1.614(4)
Zr-C(1) 2.508(3)
Zr-C(3) 2.530(2)
Zr-C(5) 2.486(3)
Zr-C(4) 2.496(3)
B(1)-C(111) 1.639(4)
C(14)-Zr-N(1) 102.60(10)
N(1)-C(16)-B(1) 174.1(3)
C(16)-B(1)-C(121) 102.6(2)
C(1)-Si(1) 1.878(3)
Zr-N(1)-C(16) 175.1(2)
C(16)-B(1)-C(111) 111.2(2)
C(16)-B(1)-C(131) 105.0(2)
(15) Bosch, B. E.; Erker, G.; Fro¨hlich, R.; Meyer, O. Organometallics
1997, 16, 5449.
(16) Adducts of B(C6F5)3 with neutral (a-d) and anionic (e-h) oxo and
nitrido complexes have recently been described: (a) Galsworthy, J. R.;
Green, M. L. H.; Mu¨ller, M.; Prout, K. J. Chem. Soc., Dalton Trans. 1997,
1309. (b) Galsworthy, J. R.; Green, J. C.; Green, M. L. H.; Mu¨ller, M. J.
Chem. Soc., Dalton Trans. 1998, 15. (c) Doerrer, L. H.; Galsworthy, J. R.;
Green, M. L. H.; Leech, M. A. J. Chem. Soc., Dalton Trans. 1998, 2483.
(d) Doerrer, L. H.; Galsworthy, J. R.; Green, M. L. H.; Leech, M. A.; Mu¨ller,
M. J. Chem. Soc., Dalton Trans. 1998, 3191. (e) Doerrer, L. H.; Graham,
A. J.; Green, M. L. H. J. Chem. Soc., Dalton Trans. 1998, 3941. (f) Barrado,
G.; Doerrer, L. H.; Green, M. L. H.; Leech, M. A. J. Chem. Soc., Dalton
Trans. 1999, 1061. (g) Abram, U.; Kohl, F. J.; O¨ fele, K.; Herrmann, W.
A.; Voigt, A.; Kirmse, R. Z. Anorg. Allg. Chem. 1998, 624, 934. (h) Abram,
U. Z. Anorg. Allg. Chem. 1999, 625, 839. See also Scott, R. N.; Shriver, D.
F.; Lehman, D. D. Inorg. Chim. Acta 1970, 4, 73.
(BPh3)2]-; they are characterized by closely similar νCN
frequencies (2305 and 2255 cm-1, respectively).12,13 The latter
was found to be prone to dissociation into BPh3 and [Ph3BCN]-.
In the case of the much stronger Lewis acid B(C6F5)3 such a
dissociation process appears to be disfavored.14
Oxo anions represent a weaker type of donors and form less
stable complexes. Stirring a mixture of KNO3, Ph3CCl, and
B(C6F5)3‚Et2O in dichloromethane in a procedure analogous to
(12) Wade, R. C.; Sullivan, E. A.; Berscheid, J. R.; Purcell, K. F. Inorg.
Chem. 1970, 9, 2146.
(13) Giandomenico, C. M.; Dewan, J. C.; Lippard, S. J. J. Am. Chem.
Soc. 1991, 103, 1407.
(14) For a discussion of Lewis acid strengths of boranes see: Luo, L.;
Marks, T. J. Top. Catal. 1999, 7, 97.
(17) Beck, M. T. Pure Appl. Chem. 1987, 59, 1703.
(18) Kubas, G. J.; Jones, L. H. Inorg. Chem. 1974, 13, 2816.
(19) Vanneberg, N. G. Acta Chem. Scand. 1964, 18, 2385.
(20) Dupont, L. Acta Crystallogr. 1970, B26, 964.

226 J. Am. Chem. Soc., Vol. 123, No. 2, 2001
Zhou et al.
Table 2. Comparison of Bond Distances in Metal-Cyano Complexes [Å]
complex
M-C1
M-C2
C1-N1
C2-N2
N1-B1
N2-B2
K2[Ni(CN)4]
4b
4c
Rb2[Pd(CN)4]‚H2O
5c
1.90(3)
1.84(3)
1.15(5)
1.43(3)
1.145(2)
1.164(26)
1.126(3)
1.11(5)
1.851(2)
1.855(2)
1.967(20)
1.989(2)
1.856(2)
1.862(2)
2.038(23)
1.995(2)
1.143(2)
1.146(2)
1.093(28)
1.126(3)
1.582(3)
1.574(2)
1.582(3)
1.574(2)
1.577(3)
1.571(3)
Scheme 2
not possible to obtain a clean spectrum without decomposition
products. The formation of the zwitterionic B(C6F5)3 adduct,
Cp2ZrMe(µ-Me)B(C6F5)3, and one other species, presumably
Cp2ZrMe(µ-NtC)B(C6F5)3, could be detected.
The dimethylsilyl-bridged complex forms a more stable
binuclear complex. Decomposition of [{(SBI)ZrMe}2(µ-Me)]-
[CN{B(C6F5)3}2] occurs within hours; complete decomposition
of a sample could be detected after 24 h at 298 K in benzene
solution. Two main products are detected: the B(C6F5)3 adduct,
(SBI)ZrMe(µ-Me)B(C6F5)3, and one other species, assigned to
(SBI)ZrMe(µ-NtC)B(C6F5)3. Mononuclear cation-anion pairs
of the type [L2ZrMe+‚‚‚(C6F5)3BCNB(C6F5)3-] are not observed.
The reactions of Cp2ZrMe2 and (SBI)ZrMe2 with the nickel
complex 4c again give initially the corresponding binuclear
µ-Me cations as the only observable metallocene products.
Because of the poor solubility of 4c in benzene, measurements
were initially carried out in CD2Cl2. The primary product,
[{(SBI)ZrMe}2(µ-Me)]2[Ni{CNB(C6F5)3}4], proved surprisingly
stable in the chlorinated solvent at room temperature, although
some decomposition was apparent after ∼15 min.
Figure 2. Structure of the [Pd{CNB(C6F5)3}4]2- anion in 5c, showing
the atomic numbering scheme. Ellipsoids are drawn at 40% probability
level.
bridged binuclear cations [(L2ZrMe)2(µ-Me)]+, even in the
presence of excess trityl salt.21 Conversion to the mononuclear
species [L2ZrMe]+ occurred only on warming but was ac-
companied by rapid decomposition in the chlorinated solvent.
For comparison with the behavior of activators 1 and 4, the
reactions of L2ZrMe2 with [CPh3][B(C6F5)4] were reinvestigated
in benzene-d6 at room temperature.
Treatment of [CPh3][B(C6F5)4] with an excess of L2ZrMe2
generates the binuclear zirconocene cations [(L2ZrMe)2(µ-Me)]+-
[B(C6F5)4]- in benzene solution at 298 K. The zirconocene
concentration was kept low (0.6 < [Zr]tot < 2.5 mmol L-1) to
prevent the formation of higher aggregates and the precipitation
of ionic species as oils. Even in samples with a reactant ratio
of 1:1 only the binuclear cations are observed as the initial
products. The slow formation of mononuclear cationic zir-
conocene complexes [L2ZrMe+‚‚‚B(C6F5)4-] is seen after a few
minutes. In the case of L2 ) SBI the kinetics of this transforma-
tion were determined. The reaction of [{(SBI)ZrMe}2(µ-Me)]-
[B(C6F5)4] with excess [CPh3][B(C6F5)4] proceeds cleanly to
give the ion pair [{(SBI)ZrMe)+‚‚‚B(C6F5)4-] and follows
pseudo-first-order kinetics, k ) 3 × 10-4 s-1 (298 K). The
mononuclear complex is stable in benzene under these condi-
tions. A comparable slow reaction was observed for the Cp2-
ZrMe2 system, which forms a far less stable binuclear cation.
The reaction of L2ZrMe2 with 1 in C6D6 proceeds in a similar
manner, forming [(L2ZrMe)2(µ-Me)][CN{B(C6F5)3}2] as the
initial product at 25 °C (Scheme 2). The binuclear ion pair
formed from the Cp2ZrMe2 system, [(Cp2ZrMe)2(µ-Me)+‚‚‚
(C6F5)3BCNB(C6F5)3-], proved to be relatively unstable at room
temperature. The spectrum shows complete decomposition of
a benzene solution with [Zr]tot ) 2 mmol/L after 5 min. It was
For reactions with 4c in C6D6 the addition of up to 20 vol %
1,2-difluorobenzene was required to achieve adequate solubility.
Mixture of 4c with excess L2ZrMe2 in C6D6/C6H4F2 gave [(L2-
ZrMe)2(µ-Me)]+2[Ni{CNB(C6F5)3}4]2-, as expected (L2 ) Cp2,
SBI), together with unreacted L2ZrMe2. These binuclear com-
plexes are more stable than their [CN{B(C6F5)3}2]- analogues;
a sample of [(Cp2ZrMe)2(µ-Me)]2[Ni{CNB(C6F5)3}4] decom-
posed completely only after 60 min, while a sample of [{(SBI)-
ZrMe}2(µ-Me)]2[Ni{CNB(C6F5)3}4] showed minor signs of
decomposition only after 30 min. In both cases the B(C6F5)3
adducts, L2ZrMe(µ-Me)B(C6F5)3, were detected as the only
decomposition products; there was no evidence for CN-bridged
decomposition products, such as L2ZrMe(µ-NC)Ni{CNB-
(C6F5)3}3.
A similar sequence of events is observed in the presence of
AlMe3 which is known to form comparatively stable heterobi-
nuclear cations [Cp2Zr(µ-Me)2AlMe2]+.21 For example, a mix-
ture of Cp′′2ZrMe2, 1, and AlMe3 in a 1:1:1.5-ratio forms the
heterobinuclear complex [Cp′′2Zr(µ-Me)2AlMe2]+[CN{B(C6-
F5)3}2]- (Cp′′ ) 1,3-C5H3(SiMe3)2). The NMR spectrum in
C6D6 at room temperature is broad, but the product is unequivo-
cally characterized in CD2Cl2 at -50 °C.
(21) (a) Bochmann, M.; Lancaster, S. J. Angew. Chem., Int. Ed. Engl.
1994, 33, 1634. (b) Bochmann, M.; Lancaster, S. J J. Organomet. Chem.
1995, 497, 55. (c) For related MeB(C6F5)3- complexes, see: Haselwander,
T.; Beck, S.; Brintzinger, H. H. In Ziegler Catalysts; Fink, G., Mu¨lhaupt,
R., Brintzinger, H. H., Eds.; Springer, Berlin, 1995; p 181. Beck, S.; Prosenc,
M. H.; Brintzinger, H. H. J. Mol. Catal. A: Chem. 1998, 128, 41.
This reaction was repeated on a preparative scale in toluene.
Treatment of a mixture of Cp′′2ZrMe2 and excess AlMe3 with
1 gave an oily precipitate of [Cp′′2Zr(µ-Me)2AlMe2][NC-

Anion Effects in Olefin Polymerization
J. Am. Chem. Soc., Vol. 123, No. 2, 2001 227
metallocene species and may form solvated and solvent-
separated ion pairs.23 As Brintzinger et al. showed recently, these
ion pairs may associate in low-polarity solvents such as benzene
or toluene to give ion quadruplets, at least in the absence of
olefin mononer.24 The importance of ion-pair association has
been underlined further by theoretical modeling studies which
show the displacement of the counteranion from the metal
coordination sphere by ethene as a low-energy pathway that
probably constitutes the largest contribution toward the activa-
tion barrier of the chain growth process.25
Although the expectation of significant anion effects on the
rate of polymerization is therefore reasonable, differences
between chemically closely related anions in high-activity
catalysts can be difficult to assess quantitatively. For example,
early studies on the activity and stereoselectivity of zirconocenes
as a function of the anions [MeB(C6F5)3]-, [PhCH2B(C6F5)3]-,
[B(C6F5)4]-, and [Me-MAO]- had shown that the effects were
readily obscured by differences in reaction exotherms that
proved difficult to control.26 In addition, it is well-known that
the activity of a given catalyst system can vary widely with
reaction conditions, in particular activator:catalyst ratio, po-
lymerization times, and catalyst concentration.27 This is perhaps
best illustrated by Mo¨hring and Coville’s compilation of
literature data for the activity of the “standard” ethene polym-
erization catalyst Cp2ZrCl2/MAO, which cover a range of 4
orders of magnitude.28 We therefore saw the need to arrive at
an experimental protocol that would allow an assessment of
anion and ligand effects in a more quantitative fashion.
Preliminary ethene polymerizations were carried out under
1 bar monomer pressure at 60 °C using (SBI)ZrMe2 in 50 mL
of toluene in the presence of 200 µmol AlBui3 as scavenger,
with a Zr:activator ratio of 1:1 and a stirring rate of 1000 rpm,
to develop a standard set of reaction conditions. With both
[CPh3][B(C6F5)4] and 1, high polymerization activities were
observed. However, in most reactions activator addition led to
the instantaneous formation of a surface polymer film which
was bound to impede monomer uptake from the gas phase, and
mass-transport limitation was therefore suspected.
Figure 3. Molecular structure of Cp′′2ZrMe(µ-NC)B(C6F5)3 6, showing
the atomic numbering scheme. Ellipsoids are drawn at 40% probability
level.
Scheme 3
{B(C6F5)3}2] (Scheme 3). After this mixture was stirred at room
temperature for 30 min, the precipitate dissolved to afford a
clear solution from which pale-yellow crystals were isolated.
The product was identified as the cyanoborate complex, Cp′′2-
ZrMe(µ-NC)B(C6F5)3 6. The structure of 6‚toluene was con-
firmed by X-ray crystallography (Figure 3).
The Zr-CH3 and Zr-N bond lengths of 6 are essentially
identical, 2.254(3) and 2.259(2) Å, respectively, and closely
comparable to the terminal Zr-CH3 distance in Cp′′2ZrMe(µ-
Me)B(C6F5)3 (2.260(4) Å).22 The C-N distance of 1.149(3) Å
is almost identical to the corresponding value in 1, while the
B-C distance in 6 is slightly longer, 1.614(4) Å compared to
1.583(2) Å. The C(14)-Zr-N(1) angle of 102.60(10)° is
marginally wider than the Me-Zr-Me angle in the related
complex Cp′′2ZrMe(µ-Me)B(C6F5)3 (97.1(2)°).
On the basis of the X-ray diffraction data alone it is not
possible to decide unequivocally whether the CN moiety is
N-bound or C-bound to zirconium, although refinement with
the thermal parameters for Zr-N gave slightly better results.
The anion of 1 shows two 11B NMR resonances, a broadened
signal at δ -11.94 for B-N, and a sharp resonance for B-C
at δ -21.68. Complex 6 only shows a sharp signal at δ -19.57,
that is, the boron atom is C-bound, and there is no C/N bond
isomerization.
Polymerization Studies and Anion Effects. As stated in the
Introduction, the nature of the counteranion and its coordinating
ability can exercise a profound effect on catalyst activity. Thus,
even weakly coordinating anions such as [MeB(C6F5)3]- and
[B(C6F5)4]- are involved in dissociation equilibria2a with cationic
This was borne out by the strong apparent dependence of
catalyst activity on [Zr]. Decreasing [Zr] for a (SBI)ZrMe2/
[CPh3][B(C6F5)4] catalyst from 2 × 10-5 mol L-1 to 4 × 10-6
mol L-1 increased the catalyst productivity figure from 33.8 ×
106 g PE (mol Zr)-1 h-1 bar-1 to 168 × 106 g PE (mol Zr)-1
h-1 bar-1 (Figure 4). However, closer inspection showed that
in all reactions almost identical amounts of polymer had been
produced, a clear case of mass-transport limitation due to rapid
monomer consumption from the solution phase. Evidently chain
propagation in these systems is significantly faster than mono-
mer diffusion into the solution, and therefore, none of these
Values should be regarded as a true reflection of the actual
actiVity of this catalyst system.
Reducing the catalyst concentration below ∼4 × 10-6 mol
L-1 in an effort to overcome problems of monomer depletion
proved problematic since very highly diluted stock solutions
(23) Beck, S.; Prosenc, M. H.; Brintzinger, H. H.; Goretzki, R.; Herfert,
N.; Fink, G. J. Mol. Catal. 1996, 111, 67.
(24) Beck, S.; Geyer, A.; Brintzinger, H. H. Chem. Commun. 1999, 2477.
(25) (a) Chan, M. S. W.; Vanka, K.; Pye, C. C.; Ziegler, T. Organome-
tallics 1999, 18, 4624. (b) Vanka, V.; Chan, M. S. W.; Pye, C. C.; Ziegler,
T. Organometallics 2000, 19, 1841.
(26) Lancaster, S. J. Ph.D. Thesis, University of East Anglia, 1995.
(27) (a) Kaminsky, W.; Miri, M.; Sinn, H.; Woldt, R. Macromol. Chem.
Rapid Commun. 1983, 4, 417. (b) Janiak, C.; Versteeg, U.; Lange, K. C.
H.; Weimann, R.; Hahn, E. J. Organomet. Chem. 1995, 501, 219. (c) Janiak,
C.; Lange, K. C. H.; Versteeg, U.; Lentz, D.; Budzelaar, P. H. M. Chem.
Ber. 1996, 129, 1517.
(22) Bochmann, M.; Lancaster, S. J.; Hursthouse, M. B.; Malik, K. M.
A. Organometallics 1994, 13, 2235.
(28) Mo¨hring, P. C.; Coville, N. J. J. Organomet. Chem. 1994, 479, 1.

228 J. Am. Chem. Soc., Vol. 123, No. 2, 2001
Zhou et al.
Table 3. Ethene Polymerizations with Titanium and Zirconium Catalysts as a Function of Anion Structurea
metallocene
(µmol)
activatorb
(µmol)
AlBui3
[µmol]
temp
[°C]
time
[s]
polymer
yield [g]
productivityc
Mw × 10-3
Mw/Mn
Tm [°C]
(SBI)ZrMe2 (1.0)
(SBI)ZrMe2 (0.2)
(SBI)ZrMe2 (1.0)
(SBI)ZrMe2 (0.2)
(SBI)ZrMe2 (1.0)
(SBI)ZrMe2 (0.5)
CGCTiBz2 (0.2)
CGCTiBz2 (0.5)
CGCTiBz2 (1.0)
CGCTiBz2 (1.0)
CGCTiBz2 (1.0)
B(C6F5)4 (1.0)
B(C6F5)4 (0.2)
B(C6F5)4 (1.0)
B(C6F5)4 (0.2)
B(C6F5)3 (1.0)
B(C6F5)3 (0.5)
1 (0.2)
1 (0.5)
1 (1.0)
B(C6F5)4 (1.0)
B(C6F5)3 (1.0)
200
200
200
200
200
200
100
100
100
100
100
60
60
60
60
60
60
60
60
60
60
60
180
30
30
10
180
30
30
30
30
30
30
0.68
0.28
0.28
0.21
0.70
0.29
0.21
0.26
0.37
0.30
0.20
13.6
168
33.8
450
14.0
69.6
126
62.4
44.4
36
n.d.
195.5
2.7
135.4
135.7
n.d.
n.d.
137.3
178
225.5
2.7
3.0
304
254
3.0
3.3
136.9
137.3
138.1
24
b
c
a Conditions: 1 bar ethene pressure, 50 mL toluene, stirring rate 1000 rpm. B(C6F5)4 ) [CPh3][B(C6F5)4]. In 106 g PE/(mol M)‚h‚bar.
Figure 5. Gas-flow profile of an ethene polymerization with (SBI)-
ZrMe2/1/AlBui3 under pressure (5 µmol Zr, 1 µmol 1, 1 mmol Al; 60
°C, initial pressure 7 bar, 3 L toluene), showing (a) injection of the
activator; (b) increase of stirrer speed from 500 to 800 rpm; (c)
termination by methanol injection. After 2 min, gas uptake was impeded
by large amounts of swollen polymer.
Even in ethene polymerizations at 1 bar noticeable reaction
exotherms were observed, and for this reason comparable but
less active catalysts were sought. Earlier work on anion effects
Figure 4. Dependence of catalyst activity and polymer yield in ethene
polymerizations with (SBI)ZrMe2/[CPh3][B(C6F5)4]/AlBui3 (60 °C, 1
bar, 30 s), an illustration of mass-transport limitation.
by Marks et al. had suggested that titanium catalysts of the
“constrained geometry” type, such as Me2Si(C5Me4)(NBut)TiR2
of zirconocene dimethyls are very prone to hydrolysis and tend
(CGCTiR2), gave ethene polymerization activities that were up
to give poorly reproducible results. Ethene polymerizations were
to 2 orders of magnitude lower than those of zirconocenes, for
therefore conducted in a 5 L autoclave under 7 bar ethene. This
example, 0.08 and 1.56 × 106 g PE (mol Ti)-1 h-1 bar-1 for
allowed the maximum concentration of active cationic zirconium
species to be reduced to 6.7 × 10-8 mol L-1. With (SBI)ZrMe2/
1/AlBui3 (5:1:1000) at a start temperature of 60 °C very rapid
polymerization was observed, accompanied by a temperature
increase of ∼40 °C.29 It proved difficult to pump in feed gas
sufficiently fast to maintain the reactor pressure which dropped
within 1 min from 7 to 2 bar. The gas flow diagram (Figure 5)
shows that even under these conditions the reaction is still
diffusion limited, that is, the gas flow is very sensitive to the
stirrer speed. After 4 min the reactor was filled with swollen
polymer, and the reaction was terminated. The lower limit
estimate of catalyst productivity under these conditions was
7.6 × 108 g PE (mol Zr)-1 h-1 bar-1. Assuming 100%
participation of all cationic zirconocene centers in chain growth,
this productivity corresponds to a monomer insertion rate of
over 5 × 104 s-1 averaged over the 4 min period. Initial insertion
rates are presumably much higher. Neither in these autoclave
reactions nor in reactions at 1 bar at lower temperatures were
we able to demonstrate the absence of mass-transport limitations
in the case of ethene polymerizations.
[CGCTiMe][MeB(C6F5)3] and [CGC-HTi][B(C6F5)4], respec-
tively.7d,f We therefore decided to compare our zirconocene
results with the system CGCTiR2/activator/AlBui3.
Ethene polymerizations with (SBI)ZrMe2 and CGCTiBz2
activated with 1, and for comparison, [CPh3][B(C6F5)4] and
B(C6F5)3 are collected in Table 3. In our hands the titanium
catalysts exhibited activities comparable to those of zirconocenes
and were 10-100 times more active than previously reported.
In part this difference may be due to the presence of the AlBui3
scavenger in our system.30 All three activators produced catalysts
of broadly comparable activities which decreased in the order
[CN(B(C6F5)3}2]- g [B(C6F5)4]- > B(C6F5)3. In the titanium
system, too, strictly isothermal conditions proved difficult to
establish.
Propene polymerizations present a different scenario. Some
time ago we reported that Cp*TiMe3/B(C6F5)3 polymerizes
propene to a very high molecular weight elastomeric polymer
with narrow polydispersity.31 In this system the polymer yield
(30) AlBui3 can also act as an alkylating agent, although the extent of
this reaction depends on the transition-metal alkyl. With (SBI)ZrMe2 and
CGCTiMe2 in the presence of excess AlBui3, rapid Me/i-Bu exchange is
observed, while CGCTi(CH2Ph)2 remains essentially unreacted after 5 h at
25 °C (Al:Ti ) 8:1, C6D6).
(29) The power output of 1 µmol catalyst under these conditions is about
4 kW. Internal cooling of the autoclave was not attempted.

Anion Effects in Olefin Polymerization
J. Am. Chem. Soc., Vol. 123, No. 2, 2001 229
Table 4. Propene Polymerizations with Titanium and Zirconium Methyl Complexes as a Function of Anion Structurea
entry
no.
metallocene
(µmol)
activatorb
(µmol)
AlBui3 toluene start temp time ∆Tc polymer
Tm
[µmol] [mL]
[°C]
[min] [°C] yield [g] productivityd Mw × 10-3 Mw/Mn [°C]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
(SBI)ZrMe2 (0.5) 1 (0.5)
(SBI)ZrMe2 (0.5) 1 (0.5)
(SBI)ZrMe2 (1.0) 1 (1.0)
(SBI)ZrMe2 (1.0) 1 (1.0)
100
100
100
100
100
100
100
100
100
100
100
100
5
100
100
100
100
1000
100
100
1000
100
100
100
100
100
100
100
1000
100
1000
100
100
100
100
100
0
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
0.5
0.5
0.5 14
1.0
1.0
0.5
1.0
1.0
10
10
3
30
10
0.5
4
6
0.89
0.88
1.64
0.95
0.89
0.56
0.70
0.58
0
0.19
0.20
5.35
0.08
1.16
1.32
1.29
0.66
1.35
0.25
1.43
0.235
0.787
0.228
167.1
297.5
277.2
160.6
150.4
189.3
118.3
98.0
154.4
150.5
109
95
1.9
2.0
6
0
4
4
0
0
0
1.5
6.5
0
5.5
6.0
0
(SBI)ZrMe2 (1.0) 1 (1.0)
105
112
86.2
3.8
1.9
2.2
(SBI)ZrMe2 (0.5) B(C6F5)4- (0.5)
(SBI)ZrMe2 (0.5) B(C6F5)4- (0.5)
(SBI)ZrMe2 (0.5) B(C6F5)4- (0.5)
(SBI)ZrMe2 (0.5) B(C6F5)3 (0.5)
(SBI)ZrMe2 (1.0) B(C6F5)3 (1.0)
(SBI)ZrMe2 (4.0) B(C6F5)3 (4.0)
(SBI)ZrMe2 (4.0) B(C6F5)3 (4.0)
(SBI)ZrMe2 (4.0) B(C6F5)3 (4.0)
CGCTiMe2 (0.5) 1 (0.5)
135.4
135.4
0
1.60
1.41
3.77
0.17
146.0
144.3
67
925
2.4
3.9
3.7
4.5
100
100
100
100
100
100
100
100
100
100
392.1
e
CGCTiMe2 (0.5) 1 (0.5)
CGCTiMe2 (0.5) 1 (0.5)
1.0
1.0
0.5
1.0
17
30
0.5
223.1
218.0
223.1
228.2
1150
1640
CGCTiMe2 (0.5) B(C6F5)4- (0.5)
CGCTiMe2 (0.5) B(C6F5)4- (0.5)
CGCTiMe2 (10.0) B(C6F5)3 (10.0)
CGCTiMe2 (25.0) B(C6F5)3 (25.0)
(SBI)ZrMe2 (0.5) 4c (0.25)f
(SBI)ZrMe2 (1.0) 4c (0.5)g
(SBI)ZrMe2 (1.0) 5c (0.5)g
4.5
0
0
0
2
0
0
1.24
1.60
56.4
142
98
2.4
2.5
1.0
1.0
47.2
13.7
a Conditions: 1 bar propene pressure, stirring rate 1000 rpm. b B(C6F5)4 ) [CPh3][B(C6F5)4]. c Internal temperature rise at the end of the run. d In
106 g PP/(mol M)‚[C3H6]‚h‚bar. e a-PP, [mm] ) 11.0, [mr] ) 49.6, [rr] ) 39.4%. f In 0.25 mL of toluene/1,2-difluorobenzene (75:25). g In 0.5 mL
of toluene/1,2-difluorobenzene (75:25).
Figure 7. Propene polymerizations with (SBI)ZrMe2/1/AlBui3: de-
pendence of polymer mass and productivity on [Zr] (1 bar, 20 °C,
toluene). For experimental details see Table 4.
Figure 6. Propene polymerization with Cp*TiMe3/B(C6F5)3 (20 °C,
toluene, 1 bar). The linear increase of polymer yield with [Ti] together
with an approximately constant catalyst activity indicates absence of
mass-transport limitation.
Polymerizations were carried out at 20 °C in flame-dried glass
reactors in 100 mL of toluene and terminated typically after 30
s by methanol injection (Table 4). For the system (SBI)ZrMe2/
1/AlBui3 a plot of catalyst productivity as a function of
zirconium concentration shows the desired near-linear increase
of polymer yield with [Zr] over the chosen range of [Zr] )
5-20 × 10-6 mol L-1 (Figure 7). The polymer mass graph
shows slight deviation from linearity, and there is some decrease
of productivity with increasing [Zr]. We ascribe this effect
primarily to the onset of system heterogenization as the product,
isotactic polypropene, begins to precipitate. A similar diagram
for CGCTiMe2/1/AlBui3 indicates that absence of mass-transport
limitation can be assumed up to [Ti] e 2.5 × 10-5 mol L-1
(Figure 8).
increased linearly with [Ti], while the productivity values
remained approximately constant (Figure 6). These parameters
establish that, in contrast to the ethene polymerizations men-
tioned above, mass-transport limitation is absent, an essential
precondition if catalyst productivities are to be related to
variables such as ligand structure and the nature of the anion.
In the following such a regime was adopted for the assessment
of anion effects. To be able to separate the influence of anion
structure from other effects on catalyst activity, such as
heterogenisation of the reaction mixture, catalyst entrapment,
increases in solution viscosity, monomer depletion of the
reaction medium, and irreversible catalyst deactivation, short
and strictly standardized reaction times were employed as far
as possible (typically 30 s).
With (SBI)ZrMe2/1/AlBui3 and [Zr] ) 5 × 10-6 mol L-1
,
productivities were of the order of ∼2.8-3.0 × 108 g PP (mol
Zr)-1 h-1 [C3H6]-1 (Table 4, entries 2-3). Throughout, the
(31) Sassmannshausen, J.; Bochmann, M.; Ro¨sch, J.; Lilge, D. J.
Organomet. Chem. 1997, 548, 23.

230 J. Am. Chem. Soc., Vol. 123, No. 2, 2001
Zhou et al.
13). This, however, led to a further significant reduction in
polymer yield, and it appears, therefore, that the observed data
are primarily a reflection of the comparatively strong coordinat-
ing power of [MeB(C6F5)3]-.
In the case of zirconocene catalysts the polymer molecular
weights are little influenced by the counteranion. By contrast,
CGCTiMe2 catalysts activated with CPh3+ salts give atactic PP
with molecular weights ∼10 times higher than with B(C6F5)3
activation (cf. Table 4, entries 14-18 and 20). With these
relatively open catalysts anion coordination evidently assists
chain termination.
Dianions as Counteranions. For the dianions [M{CNB-
(C6F5)3}4]2- (M ) Ni, Pd) stronger ion-ion interactions and
comparatively reduced catalyst activities could be expected. This
is no doubt reflected in the comparatively poor solubility of
the 2:1 electrolytes [CPh3]2[M{CNB(C6F5)3}4] in neat toluene
at room temperature. These compounds do, however, form
homogeneous solutions on addition of 25 vol % of 1,2-
difluorobenzene; stock solutions of 4c and 5c were therefore
prepared in toluene/difluorobenzene mixtures. Under these
conditions, activation of (SBI)ZrMe2 with 4c led to rapid
propene polymerization, with activities of up to 5.6 × 107 g PP
(mol Zr)-1 h-1 [C3H6]-1 (Table 4, entries 21-23). Although
somewhat lower, these values are comparable with those for
[CPh3][B(C6F5)4] and, once again, are significantly higher than
those obtained with the rather slow B(C6F5)3-based system.33
Chloride-Containing Systems. Some time ago Chien et al.
introduced the system Cp2ZrCl2/AlR3/[CPh3][B(C6F5)4] as an
alternative to the use of preformed metallocene dialkyls as
catalyst precursors.34 In view of the greater stability of highly
diluted stock solutions of metallocene dichlorides, compared
to dialkyls, this system was briefly examined for comparison
(Table 5). Reactions of (SBI)ZrCl2/AlBui3 with 1 gave com-
parable activities to those shown in Table 4, although the data
were less reproducible. The source of these errors was traced
to the pre-alkylation time: the activities of the chloride-
containing system depend markedly on the time the metal
dichloride is allowed to react with AlBui3. While the alkylation
of zirconocene dichlorides by AlR3 is fast,35 in the case of
CGCTiCl2 maximum activity is only reached after a pre-reaction
time of 30 min (Figure 9).
Figure 8. Propene polymerizations with CGCTiMe2/1/AlBui3: de-
pendence of polymer mass and productivity on [Ti] (1 bar, 0 °C,
toluene).
productivities obtained with activator 1 were about 1.3-1.5
times higher than with [B(C6F5)4]- as counteranion. As was
seen in ethene polymerizations, the titanium system CGCTiMe2/
1/AlBui3 gave very similar activities to the zirconocene catalyst
(Table 4, entries 14-18). For isolated [CGC-HTi][B(C6F5)4] a
propene polymerization productivity of 2.12 × 106 g PP (mol
Ti)-1 h-1 bar-1 has been reported, although the higher catalyst
concentration ([Ti] ) 3 × 10-4 mol L-1) and the longer reaction
time employed in this case (5 min) make a direct comparison
difficult.7d
With both 1 and [CPh3][B(C6F5)4] as activators, the reactions
conducted in 100 mL of toluene showed a rise of the internal
reactor temperature of 4-6 °C, that is, on this scale and at these
catalyst concentrations the reactions could not be kept strictly
isothermal. The influence of this temperature effect on activity
was probed in some cases by increasing the solvent volume to
1000 mL. To keep activator mixing and quenching times short
compared to the duration of the polymerization, the reaction
times were extended to 1 min; for comparison, reactions of 1
min duration on a 100 mL scale are also reported. Reducing
the catalyst concentration in this way by a factor of 10 did indeed
eliminate the reaction exotherm, cf. Table 4, entries 4/5, 7/8,
and 15/16. At the same time, the closely comparable activity
figures for these runs are a further confirmation that mass-
transport effects were absent even in the smaller-scale reactions.
The activation of either (SBI)ZrMe2 of CGCTiMe2 with
B(C6F5)3/AlBui3 produces catalysts which are significantly less
active, so that strictly comparable reaction conditions could not
be maintained. To obtain isolable amounts of polymer, it proved
necessary to extend the reaction times to up to 30 min and to
increase the catalyst concentration. The productivities were about
Discussion
The formation of adducts between B(C6F5)3 and strongly
coordinating anions Xn-, for example Xn- ) CN- or [M(CN)4]2-
(M ) Ni, Pd), provides a facile general route for the synthesis
of new bulky anions [X{B(C6F5)3}y]n- in which the negative
charge is delocalized and less than unity per boron center. The
+
compounds form stable NHMe2Ph+ and CPh3 salts and are
suitable as counteranions in metallocene-based polymerization
catalysts.
The reaction of zirconocene dimethyls with 1, 4c, or 5c gives
salts of the binuclear cations, [(L2ZrMe)2(µ-Me)]+ (L2 ) Cp2
or SBI), stabilized by the very bulky anions. Although the anions
+
2 orders of magnitude below those obtained with the CPh3
salts. Here, too, the titanium and zirconium catalysts gave closely
similar results.
+
are stable toward electrophiles such as CPh3 in solution and
There was the possibility that the low activity values might
be caused by the destruction of B(C6F5)3, which is known to
be able to react with AlR3 under alkyl exchange and formation
of potentially deactivating Al-C6F5 species.32 This was tested
by reducing the Al:Zr ratio from 100:1 to 5:1 (Table 4, entry
in the solid for indefinite periods, they react slowly with [(L2-
ZrMe)2(µ-Me)]+ to give, predominantly, L2ZrMe(µ-Me)B-
(C6F5)3 and L2ZrMe(µ-NC)B(C6F5)3. At 20 °C the disappearance
(33) The presence of low concentrations of C6H4F2 (0.1-0.25 mL/100
mL toluene) has no effect on catalytic activity.
(32) (a) Bochmann, M.; Sarsfield, M. J. Organometallics 1998, 17, 5908.
(b) Lee, C. H.; Lee, S. J.; Park, J. W.; Kim, K. H.; Lee, B. Y.; Oh, J. S. J.
Mol. Catal. A: Chem. 1998, 132, 231. (c) Biagrini, P.; Lugli, G.; Abis, L.;
Andreussi, P. (Enichem Elastomeri S.r.l.). Eur. Pat. Appl. EP 0 694 548,
1996.
(34) (a) Chien, J. C. W.; Xu, B. Makromol. Chem., Rapid Commun. 1993,
14, 109. (b) Chien, J. C. W.; Tsai, W. M. Makromol. Chem., Macromol.
Symp. 1993, 66, 141. (c) Chen, Y. X.; Rausch, M. D.; Chien, J. C. W.
Organometallics 1994, 13, 748.
(35) Beck, S.; Brintzinger, H. H. Inorg. Chim. Acta 1998, 270, 376.

Anion Effects in Olefin Polymerization
J. Am. Chem. Soc., Vol. 123, No. 2, 2001 231
Table 5. Propene Polymerizations with Titanium and Zirconium Dichloridesa
entry
no.
metallocene
(µmol)
activatorb
(µmol)
AlBui3
[µmol]
toluene
[mL]
start temp
[°C]
time
[min]
∆Tc
[°C]
polymer
yield [g]
Tm
[°C]
productivityd
1
2
3
4
5
6
7
8
(SBI)ZrCl2 (0.5)
(SBI)ZrCl2 (1.0)
(SBI)ZrCl2 (1.0)
(SBI)ZrCl2 (1.0)
(SBI)ZrCl2 (1.0)
(SBI)ZrCl2 (4.0)
CGCTiCl2 (2.0)
CGCTiCl2 (2.0)
1 (0.5)
1 (1.0)
100
100
100
100
100
100
100e
100f
100
100
100
1000
100
100
100
100
20
20
20
20
20
20
20
20
0.5
0.5
1.0
1.0
10
30
0.5
0.5
6
8
4
0
0
0
0
n.d.
0.86
1.54
0.61
0.63
trace
4.48
0
290.7
260.3
51.4
150.4
B(C6F5)4- (1.0)
B(C6F5)4- (1.0)
B(C6F5)3 (1.0)
B(C6F5)3 (4.0)
1 (2.0)
53.2
3.15
0
47.3
146.4
1 (2.0)
0.56
a Conditions: 1 bar propene pressure, stirring rate 1000 rpm. b B(C6F5)4 ) [CPh3][B(C6F5)4]. c Internal temperature rise at the end of the run. d In
106 g PP/(mol M)‚[C3H6]‚h‚bar. e Alkylation time 0.5 min. f Alkylation time 30 min.
bond can be about 102 times faster than insertion into a M-CH3
bond. These effects serve to maximize the concentration of
active species available for the initiation of the chain growth
process. The fact that we see consistently higher activities with
the (SBI)ZrMe2 and the CGCTiMe2 systems than those mea-
sured previously7d,f is most likely due to generation of higher
concentrations of active species in our system.
Catalysts based on (SBI)ZrMe2/1 /AlBui3 proved to be very
highly active for ethene polymerizations. There was a significant
increase in productivity on increasing the ethene pressure up to
7 bar. Although productivities of up to 7.6 × 108 g PE (mol
Zr)-1 bar-1 h-1 appear to be the highest yet reported, there was
evidence that even at the lowest catalyst concentrations we were
able to handle reproducibly, catalyst productivity was severely
limited by monomer depletion in the solution phase, that is,
the data are subject to mass-transport limitations. There is no
doubt that this activity figure is far from the upper limit for
this catalyst system, and much higher ethene insertion rates than
the 5 × 104 s-1 quoted here should be possible with a more
appropriate reactor design.
Figure 9. Propene polymerization activity of CGCTiCl2/1/AlBui3:
The findings have wider implications. There is much em-
dependence of catalyst productivity on pre-alkylation time. Condi-
phasis on the design of new ligands for ethene polymerization
tions: 2 µmol Ti, Ti:1:Al ) 1:1:50, 100 mL toluene, 1 bar, 20 °C.
catalysts, and activity values are often directly related to ligand
structure.37 However, the productivity of polymerization cata-
of [(Cp2ZrMe)2(µ-Me)]+ is complete after 5 min, giving an
lysts is a function of many variables, of which the nature of the
estimated rate of decomposition of ∼0.003 s-1, while the
ligand may not be the most important. Several scenarios are
decomposition of the bulkier SBI analogue requires several
possible: (a) A catalyst may have an intrinsically high propaga-
hours. From the selective formation of the Zr-NC-B product
tion rate kp but the productivity is low because the activation
6 rather than the Zr-CN-B isomer we conclude that decom-
method is inefficient and only generates very low concentrations
position is initiated by B-N bond dissociation of [CN{B-
of active species [C*]. (b) The activation is efficient but the
(C6F5)3}2]-.
equilibrium between active and dormant species is unfavorable
(e.g. because of tight cation-anion interaction, solvent coor-
dination, etc.). (c) Catalyst activation is efficient and [C*] is
The main purpose of this study was the quantification of anion
effects on the catalytic activity in olefin polymerization systems
with anions of closely similar structures.
high, but a facile decomposition pathway leads to early catalyst
Throughout, a catalyst system based on mixtures of a
deactivation. (d) The catalyst may form high [C*] and give
metallocene and activator 1 in the presence of AlBui3 was used,
intrinsically high kp, but the monomer concentration is insuf-
with an Al:Zr ratio of 200:1 for reactions in 100 mL of toluene,
ficient over the duration of the experiment, or the delivery of
and 2000:1 for reactions on a 1000 mL scale. In our hands the
monomer to the active center is too slow (i.e., mass-transport
addition of AlBui3 allowed us to use lower catalyst concentra-
limitation). (e) As for (d), but the polymer is poorly soluble,
tions without sacrificing reproducibility. Catalyst productivities
begins to precipitate during the early stages of the reaction, and
of repeat reactions were typically reproducible within (10%
thereby removes catalyst from the reaction. (f) Activation is
and in many cases 5% or lower. Apart from acting as a
efficient, the polymerization is not mass-transport limited, and
scavenger, the metallocene reacts with AlBui3 under alkyl
catalyst productivity is a function of the ligand employed.
exchange to give Zr-Bui species. This is beneficial for two
It is clear that although most catalytic experiments tend to
reasons: unlike metal-methyl complexes, these isobutyl species
assume that situation (f) prevails, this is in fact rarely achieved.
are sterically too hindered to form catalytically dormant AlR3
The results shown here confirm that variations in the mode of
adducts of the type [L2M(µ-R)2AlR2]+, and second, as Fink has
catalyst activation and in monomer concentration can produce
shown,36 the rate of monomer insertion into a higher M-alkyl
differences in catalyst activities which are at least as important
as changes in ligand structure. Mindful of these difficulties,
(36) (a) Fink, G.; Zoller, W. Makromol. Chem. 1981, 182, 3265. (b)
Fink, G.; Schnell, D. Angew. Makromol. Chem. 1982, 105, 31. (c) Mynott,
R.; Fink, G.; Fenzl, W. Angew. Makromol. Chem. 1987, 154, 1. (d) Fink,
G.; Fenzl, W.; Mynott, R. Z. Naturforsch., B: Chem. Sci. 1985, 40b, 158.
(37) See, for example: Alt, H. G. J. Chem. Soc., Dalton Trans. 1999,
1703. Alt, H. G.; Ko¨ppl, A. Chem. ReV. 2000, 100, 1205.

232 J. Am. Chem. Soc., Vol. 123, No. 2, 2001
Zhou et al.
Figure 11. Anion effects in propene polymerizations with CGCTiMe2
catalysts. (1 bar, 20 °C, toluene). For experimental details see Table 4.
“intrinsic” activity of the system. Such an extrapolation allows
the quantification of anion and, for a given activator system, of
ligand effects, independent of catalyst concentration. For (SBI)-
ZrMe2/activator/AlBui3 at 20 °C the intrinsic activity values for
[CN{B(C6F5)3}2]- and [B(C6F5)4]- are about 3.2 × 108 and
2 × 108 g PP (mol Zr)-1h-1[C3H6]-1, respectively.
Figure 10. Anion effects in propene polymerizations with (SBI)ZrMe2
catalysts. (1 bar, 20 °C, toluene). For experimental details see Table 4.
+
Gibson et al.38 have suggested a simple classification of
polymerization catalysts, from “very low” (<103 g polymer (mol
cat)-1 h-1 bar-1) to “very high” (>106 g polymer (mol cat)-1
h-1 bar-1), to which even higher categories (108 g polymer (mol
cat)-1 h-1 bar-1) could be added. In view of the findings
discussed above it seems likely that at least those ethene
polymerization catalysts in the “high” classification and above
should be suspected of being mass-transport limited, unless it
can be demonstrated otherwise. A correlation of ligand structure
with catalyst activity in such cases is clearly not very meaning-
ful.
The problem of mass-transport limitation is much reduced
in the case of propene polymerizations. Monitoring the amount
of polymer produced as a function of catalyst concentration is
an easy method for establishing the concentration range where
mass-transport limitation may be assumed to be absent for a
given reaction temperature.
The difference between catalyst activation with CPh3 salts
and with B(C6F5)3 was particularly pronounced in propene
polymerizations. This eliminates the possibility that the observed
activities are due to the in situ formation of L2ZrMe(µ-Me)B-
(C6F5)3 from L2ZrMe2 and 1. Although stoichiometric reactions
had demonstrated an extensive decomposition chemistry of the
cyanoborate anions, these reactions are slow e0.003 s-1. With
observed propene insertion rates of the order of 103 s-1 chain
propagation is g106 times faster than anion decomposition. This
kinetic difference is sufficient to allow thermodynamically labile
anions to be used as components in extremely effective
polymerization systems.
The marked difference between [CN{B(C6F5)3}2]- and
[B(C6F5)4]- was to us initially surprising. The reaction of
zirconocene dimethyls with CPh3+ salts of both anions leads to
identical cationic products [(L2ZrMe)2(µ-Me)]+, and since the
+
reactivity of CPh3 as cation generating agent is unlikely be
affected by the counteranion, these will be formed in equal
concentrations. The weakly coordinating anion [B(C6F5)4]- is
not able to enter the coordination sphere of [(L2ZrMe)2(µ-Me)]+,
and polymerization activity could therefore be expected to
depend simply on the degree of dissociation of [(L2ZrMe)2(µ-
Me)]+ in the presence of monomer to generate [L2ZrMe-
(olefin)]+ + L2ZrMe2. On that basis, a further reduction in anion
nucleophilicity below that of [B(C6F5)4]- might be expected to
involve little change in the transition state and, hence, have no
bearing on the active species concentration or on the rate of
chain growth. The results show, however, that this is not the
case, and even very weakly coordinating anions are intimately
associated with the cations during the olefin insertion step. It
seems likely, therefore, that the binuclear ion pair [(L2ZrMe)2-
(µ-Me)]+‚‚‚X- first produces L2ZrMe2 + [L2ZrMe+‚‚‚X-],
followed by reaction of the mononuclear ion pair with alkene
and displacement of the anion from the coordination sphere of
the metal (Scheme 4). The activation barrier of this process
would of course be lowered by reducing anion nucleophilicity.
We interpret the observed activity enhancement seen with 1 as
a reflection of the more extensive delocalization of the negative
The results show that there is a very significant dependence
of catalyst productivity on the nature of the counteranion, both
in the case of ansa-zirconocenes and constrained geometry
titanium catalysts, as illustrated in Figures 10 and 11, with
catalyst productivities decreasing in the order [CN{B(C6-
F5)3}2]- > [B(C6F5)4]- . [MeB(C6F5)3]-. As in the case of
ethene, the activities of the zirconium and titanium catalysts
are closely comparable. The titanium system produces atactic
polymer and has the advantage of remaining homogeneous,
whereas the i-PP produced with the (SBI)Zr catalyst begins to
precipitate early in the reaction, although with the short reaction
times employed this problem proved not serious.
The system (SBI)ZrMe2/activator/AlBui3 shows an almost
linear dependence of catalyst activity on [Zr] (Figure 10), with
errors of (10% or less even for polymerizations conducted at
[Zr] ) 5 × 10-7 mol L-1 (cf. Table 4, entries 2/3: (3.5%;
4/5: (3.3%, 7/8: (9.4%). Under such conditions it is possible
to extrapolate to zero catalyst concentration to determine the
(38) Britovsek, G. J. P.; Gibson, V. C.; Wass, D. F. Angew. Chem., Int.
Ed. 1999, 38, 429.

Anion Effects in Olefin Polymerization
J. Am. Chem. Soc., Vol. 123, No. 2, 2001 233
Scheme 4
Table 6. Rates of Migratory Monomer Insertions and Activation
Energies for Propene Polymerizations with at 293 Ka
monomer insertion
activator
activityb
frequency/s-1
∆Gq/kJ mol-1
1
297.5
189.3
1.60
1967
1252
10.6
25
53.2
54.3
66.0
63.9
[CPh3][B(C6F5)4]
B(C6F5)3
3.77
a Catalyst (SBI)ZrMe2/activator/AlBui3 1:1:200. b In 106 g PP (mol
Zr)-1 [C3H6]-1 h-1
.
Figure 12. Differences in activation barriers ∆∆Gq for propene
polymerizations with the catalyst system (SBI)ZrMe2/activator/AlBui3
at 20 °C as a function of anion structure [kJ mol-1]. The values for
[MeB(C6F5)3]- were estimated from reactions of 10-30 min duration.
charge in the diborate [CN{B(C6F5)3}2]- which helps to weaken
cation-anion interactions.
The importance of cation-anion association may be gauged
from a consideration of the electrostatic binding energy EC and
the association constant Kass for the process C+ + A- ) C+A-.
According to the Eigen-Denison-Ramsey-Fuoss (EDRF)
model,39 for two monovalent ions in a solvent of low dielectric
constant such as toluene (ꢀ ) 2.379 at 25 °C) at a distance of
8 Å, Kass ) 7.9 × 106 m3 mol-1. The attraction between a
monocation and a dianion, as in [L2ZrMe+‚‚‚Ni{CNB(C6F5)3}42-],
is orders of magnitude higher, Kass (at 8 Å) ) 4.9 × 1019 m3
mol-1. Even allowing for the difficulty of determining the
appropriate inter-ion distance in the case of bulky, nonspherical
anion structures such as [Ni{CNB(C6F5)3}4]2-, it is clear that
dianions are bound considerably more strongly than monoanions,
in line with the reduced catalytic activity of systems activated
with 4c or 5c, compared to 1.
From the observed propene polymerization activities the
differences in activation barriers ∆∆Gq may be calculated as a
function of the counteranion (Table 6). For the [CN{B(C6F5)3}2]-
anion system a monomer insertion rate of k ) 1967 s-1 was
found, corresponding to ∆Gq ) 53.2 kJ mol-1. While less
importance should be attached to the absolute ∆Gq value
obtained by such a simple approximation, the differences
between activation barriers are significant (Figure 12): the
activation energies for the [B(C6F5)4]-- and [Ni{CNB(C6F5)3}4]2--
based systems are higher by 1.1 and 4.1 kJ mol-1, respectively.
The B(C6F5)3-activated catalyst, although not evaluated on
strictly the same time scale, shows a further barrier increase by
6.6-8.6 kJ mol-1. Evidently there is a very significant difference
in activation barrier between the group of three “noncoordinat-
ing” borates, 1, [B(C6F5)4]- and 4c, and zwitterion-forming
[MeB(C6F5)3]-.
anions suitable as counteranions in metallocene-based polym-
erization catalysts. A general protocol has been developed to
define the concentration range for which mass-transport limita-
tion is absent and to quantify the effect of the counteranion on
catalyst activity. Ethene polymerizations were found to be
strongly mass-transport limited for ethene pressures of 1-7 bar,
to the extent that no meaningful correlation can be drawn
between catalyst productivity and ligand or anion effects. By
contrast, propene polymerizations with zirconocene and con-
strained geometry titanium catalysts confirmed that charge
delocalization in cyanoborates gives anions with reduced
nucleophilicity and results in enhanced catalytic activities.
Activation barriers for propene polymerizations increase in
the order [CN{B(C6F5)3}2]- < [B(C6F5)4]- < [Ni{CNB(C6-
F5)3}4]2- , [MeB(C6F5)3]-. In the [CGCTiMe]+X- system, the
polymer molecular weight is strongly anion-dependent. The
observed thermodynamic labilility of cyanoborates in stoichio-
metric reactions is kinetically unimportant under polymerization
reactions, with propagation rates g106 times faster than anion
decomposition.
Experimental Section
General Methods. All manipulations were performed under dini-
trogen using Schlenk techniques. Solvents were distilled under nitrogen
over sodium benzophenone (diethyl ether, THF), sodium (low-sulfur
toluene), sodium-potassium alloy (light petroleum, bp 40-60 °C), or
CaH2 (dichloromethane). NMR solvents were dried over activated 4Å
molecular sieves and degassed by several freeze-thaw cycles. 1,2-
Difluorobenzene was predried over CaH2, stored over 4Å molecular
sieves, and degassed by several freeze-thaw cycles. NMR spectra were
recorded using Bruker ARX250 and DPX300 spectrometers. Chemical
1
Conclusions
shifts are reported in ppm. H NMR spectra were referenced to the
residual solvent protons of the deuterated solvent used; 13C NMR spectra
(75.4 MHz) were referenced internally to the D-coupled 13C resonances
of the NMR solvent. 19F (282.2 MHz) and 11B (96.29 MHz) NMR
spectra were referenced externally to CFCl3 and BF3‚OEt2, respectively.
The compounds B(C6F5)3,9a,b B(C6F5)3‚Et2O,9c [CPh3][B(C6F5)3],21b,41
Adducts between B(C6F5)3 and basic anions, in this case CN-
or [M{CN)4]2-, give rise to new, very weakly coordinating
(39) Gordon, J. E. The Organic Chemistry of Electrolyte Solutions,
Wiley: New York, 1975; p 372ff. The association constant is given by
Kass ) [4πNAr3/3000]eb, where b ) (|z+z-|e2)/4πrꢀꢀ0kT, r ) distance
between ions; ꢀ ) dielectric constant of the solvent.
(40) Samuel, E.; Rausch, M. D. J. Am. Chem. Soc. 1973, 95, 6263.

234 J. Am. Chem. Soc., Vol. 123, No. 2, 2001
Zhou et al.
Cp2ZrMe2,40 (SBI)ZrMe2,22 CGCTiMe2,7d and CGCTi(CH2Ph)27d were
prepared according to literature procedures; other reagents were used
as purchased. Nitrogen, argon, ethene (BOC, 99.5%), and propene
(BOC, 99%) were purified by passing through columns of CaCl2,
supported P2O5 with moisture indicator, and 4Å molecular sieves. The
thermal analysis of polymers was carried out using a Mettler differential
scanning calorimeter (DSC12E) set at a heating rate of 10 °C/min under
N2 atmosphere.
with light petroleum at 5 °C in the presence of traces of acetone; yield
1.68 g (6.8 mmol, 55%). Anal. Calcd for C92H24B4F60N6Ni‚2(CH3)2-
CO: C, 42.0; H, 1.6; N, 3.6. Found: C, 42.2; H, 1.4; N, 3.2. 1H NMR
(300 MHz, 298 K, CD2Cl2) δ 3.28 (s, 6 H, Me), 7.59 (m, 5 H, Ph). 19
F
(282 MHz, 300 K, CD2Cl2) δ -133.68 (d, JFF ) 20.6 Hz, o-F), -160.63
(t, JFF ) 20.6 Hz, p-F), -166.36 (t, JFF ) 20.6 Hz, m-F). Infrared
(Nujol mull) 2239 cm-1 (νCN).
[CPh3]2[Ni{CNB(C6F5)3}4] (4c). To a solution of Ph3CCl (0.58 g,
2.1 mmol) and B(C6F5)3‚Et2O (2.02 g, 3.0 mmol) in CH2Cl2 (50 mL)
was added K2[Ni(CN)4] (0.21 g, 0.75 mmol). The mixture was stirred
for 12 h and filtered. Removal of the solvent in vacuo afforded an
orange-yellow solid which was washed with light petroleum to remove
any free borane. Recrystallization by layering a CH2Cl2 solution of the
product with light petroleum at 5 °C gave 4c‚2CH2Cl2; the crystals
were suitable for single-crystal X-ray diffraction; yield 0.80 g (3.0
mmol, 34%). Anal. Calcd for C114H30B4F60N4Ni‚2CH2Cl2: C, 48.6; H,
Preparation of [CPh3][CN{B(C6F5)3}2] (1). B(C6F5)3‚Et2O (20.1
g, 34 mmol) was combined with 1.10 g (17 mmol) KCN. Diethyl ether
(150 mL) was added and the reaction stirred overnight. The solids
slowly dissolved to give a clear solution after 12 h. The solvent was
removed in vacuo, and the residue was dried to give a colorless foam.
Triphenylchloromethane (5.68 g, 20.4 mmol) was added, followed by
200 mL of dichloromethane. The reaction was stirred for 1 h, filtered,
and concentrated to ∼40 mL. Addition of light petroleum and cooling
to -20 °C produced yellow crystals, yield (two fractions) 10 g (7.7
mmol, 45%). Anal. Calcd for C56H15B2F30N: C, 52.0; H, 1.2; N, 1.1.
1
1.2; N, 2.0. Found: C, 49.0; H, 1.1; N, 2.1. H NMR (250 MHz, 298
K, CD2Cl2) δ 7.69 (d, 6 H, o-H), 7.91 (t, 6 H, J ) 8.1 Hz, m-H), 8.29
(t, 1 H, J ) 7.4 Hz, p-H); 19F NMR (282 MHz, 300 K, CD2Cl2) δ
-134.64 (d, o-F), -159.56 (t, p-F), -165.82 (t, m-F). Infrared (Nujol
mull) 2234 cm-1 (νCN).
K2[Pd{CNB(C6F5)3}4] (5a). To a solution of B(C6F5)3‚Et2O (2.40
g, 4.1 mmol) in CH2Cl2 (50 mL) was added K2[Pd(CN)4] (0.29 g, 1.02
mmol). The mixture was stirred vigorously for 12 h, affording a white
solid which was washed with light petroleum to remove any borane;
yield 1.91 g (0.82 mmol, 80%). The compound was used without further
purification. 19F NMR (282 MHz, 300 K, CD3OD) δ -132.63 to
-130.37 (m, o-F), -158.6 to -157.21 (m, p-F), -164.79 to -163.58
(m, m-F).
[CPh3]2[Pd{CNB(C6F5)3}4] (5c). To a solution of Ph3CCl (0.43 g,
1.54 mmol) in CH2Cl2 (50 mL) was added K2[Pd{CNB(C6F5)3}4] (1.80
g, 0.78 mmol). The mixture was stirred vigorously for 12 h and filtered.
The solvent was removed in vacuo to give an orange-yellow solid which
was washed with light petroleum. Crystals of 5c‚2CH2Cl2 suitable for
single-crystal X-ray diffraction were obtained by layering a CH2Cl2
solution of the product with light petroleum at 5 °C. They are prone to
solvent loss; yield 1.05 g (0.38 mmol, 49%). Anal. Calcd for
1
Found: C, 52.1; H, 1.3; N, 0.8. H NMR (300.13 MHz, CDCl3, 293
K) δ 8.26 (t, 3H, J ) 7.7 Hz, p-H), 7.87 (“t”, 6H, J ) 7.7 Hz, m-H),
7.66 (d, 6H, J ) 7.2 Hz, o-H). 13C{1H} NMR (75.47 MHz, CDCl3,
293 K) δ 210.89 (CPh3), 148.17 (d, JC-F ) 253 Hz, o-C, C6F5), 143.88
(p-C, CPh3), 142.53 (m-C, CPh3), 140.09 (d, JC-F ) 203 Hz, p-C, C6F5),
136.79 (d, JC-F ) 227 Hz, m-C, C6F5), 130.85 (o-C, CPh3). 19F NMR
(282.4 MHz, CDCl3, 293 K) δ -133.29 (d, JF-F ) 21.2 Hz, o-F),
-134.40 (d, JF-F ) 20.4 Hz, o-F), -158.68 (t, JF-F ) 21.6 Hz, p-F),
-158.85 (t, JF-F ) 20 Hz, p-F), -165.14 (t, JF-F ) 16.5 Hz, m-F),
-165.47 (t, JF-F ) 20.9 Hz, m-F). 11B NMR (96.29 MHz, CDCl3, 293
K) δ -11.94 (br, N-B), -21.68 (C-B). Infrared (Nujol mull) 2305
cm-1 (νCN).
Preparation of [CPh3][ClB(C6F5)3] (2). Ph3CCl (0.23 g, 0.84 mmol)
and B(C6F5)3‚Et2O (1.0 g, 1.7 mmol) were dissolved in CH2Cl2 (30
mL). The solution was stirred for 2 h, followed by removal of the
solvent in vacuo. The product was an orange-yellow solid. Crystals
suitable for single-crystal X-ray determination were obtained by layering
a CH2Cl2 solution of the product with light petroleum at 5 °C; yield 0.
27 g (0.35 mmol, 42%). Anal. Calcd for C37H15BClF15: C, 56.2; H,
1.9; Cl, 4.5. Found: C, 56.3; H, 2.0; Cl, 4.5. 1H NMR (300 MHz, 300
K, CD2Cl2) δ 8.26 (t, 3H, J ) 7.5 Hz, p-H), 7.87 (“t”, 6H, J ) 7.9 Hz,
m-H), 7.68 (d, 6H, J ) 8.3 Hz, o-H). 19F (282 MHz, 300 K, CD2Cl2)
δ -130.86 (d, JFF ) 19.7 Hz, o-F), -160.54 (br, p-F), -165.18 (br,
m-F).
C
114H30B4F60N4Pd‚1.5CH2Cl2: C, 48.3; H, 1.2; N, 1.9. Found: C, 47.8;
1
H, 1.4; N, 1.8. H NMR (300 MHz, 300 K, CD2Cl3) δ 8.27 (t, 3 H,
p-H), 7.88 (t, 6 H, m-H), 7.67 (d, 6 H, o-H); 19F NMR (282.4 MHz,
300 K, CD2Cl3) δ -134.87 (d, JFF ) 20.6 Hz, o-F), -160.41 (t, JFF
)
20.6 Hz, m-F), -166.57 (t, JFF ) 20.6 Hz, p-F). 11B NMR (80 MHz,
300 K, CD2Cl2) δ -12.81. Infrared (Nujol mull) 2243 cm-1 (νCN).
Preparation of Cp′′2Zr(Me)NCB(C6F5)3 (6). To a suspension of
0.64 g (0.49 mmol) 1 in 10 mL of toluene was added a solution of
AlMe3 in toluene (4.9 mL, 3.9 mmol), immediately followed by solution
of Cp′′2ZrMe2 (4.9 mL, 0.1 mol/L, 0.49 mmol). There was a rapid color
change from the oily orange trityl salt suspension to a yellow oily
suspension. After being stirred for 30 min the yellow suspension
dissolved to give a pale yellow solution. Cooling to 5 °C overnight
yielded crystals of 6‚toluene suitable for X-ray diffraction; yield 0.3 g
(0.26 mmol, 53%). Anal. Calcd for C42H45BF15NSi4Zr‚C7H8: C, 50.9;
H, 4.6; N, 1.2. Found: C, 50.9; H, 4.5; N, 1.0. 1H NMR (300.13 MHz,
CD2Cl2, 293 K) δ 7.12 (s, 2H, 2-C5H3), 6.35 (t, 2H, J ) 2 Hz, C5H3),
6.28 (t, 2H, J ) 2 Hz, C5H3), 0.54 (s, 3H, Zr-CH3), 0.19 (s, 18H,
Si(CH3)3), 0.13 (s, 18H, Si(CH3)3). 13C NMR (75.47 MHz, CD2Cl2,
293 K) δ 137.02 (2-C5H3), 130.59 (C5H3), 127.60 (C5H3), 120.01
(C5H3), 118.88 (C5H3), 47.58 (Zr-CH3), -0.21 (Si(CH3)3), -0.39 (Si-
(CH3)3). 11B NMR (96.29 MHz, CD2Cl2, 293 K) δ -19.57.
[PhNMe2H][HSO3{OB(C6F5)3}] (3). To a solution of B(C6F5)3‚Et2O
(2.0 g, 3.4 mmol) in CH2Cl2 (50 mL) was added [PhNMe2H][HSO4]
(0.29 g, 0.85 mmol). The solution was stirred vigorously for 12 h,
followed by removal of the solvent in vacuo. The resulting solid was
washed with light petroleum to remove any free borane; yield: 0.39 g
(0.57 mmol, 67%). Anal. Calcd for C26H13BF15NO4S: C, 42.7; H, 1.8;
1
N, 1.9. Found: C, 43.1; H, 2.1; N, 1.8. H NMR (300 MHz, 300 K,
CD3OD) δ 3.31 (s, 12 H, Me), 7.58-7.63 (m, 10 H, Ph); 19F NMR
(282 MHz, 300 K, CD3OD) δ -133.9 (d, JFF ) 20.6 Hz, o-F), -161.24
(t, JFF ) 20.9 Hz, p-F), -166.4 (t, JFF ) 20.6 Hz, m-F). 11B (80 MHz,
300 K, CD3OD) δ -1.99.
K2[Ni{CNB(C6F5)3}4] (4a). To a solution of B(C6F5)3‚Et2O (2.0 g,
3.4 mmol) in CH2Cl2 (50 mL) was added K2[Ni(CN)4] (0.18 g, 0.85
mmol). The mixture was stirred for 12 h. The white solid product was
washed with light petroleum to remove any free borane. The compound
was used without further purification; yield 1.55 g (0.68 mmol, 80%).
19F NMR (282 MHz, 298 K, CD3OD) δ -135.74 (d, JFF ) 19.7 Hz,
o-F), -162.74 (t, JFF ) 19.7 Hz, p-F), -168.19 (t, JFF ) 19.7 Hz,
m-F). 11B NMR (96 MHz, 300 K, CD3OD) δ -2.34. Infrared (Nujol
mull) 2236 cm-1 (νCN).
[PhNMe2H]2[Ni{CNB(C6F5)3}4] (4b). To a solution of [PhNMe2-
H]Cl (1.3 g, 2.1 mmol) in CH2Cl2 (50 mL) was added 4a (2.68 g, 1.05
mmol). The mixture was stirred for 12 h and filtered. The solvent was
removed from the filtrate in vacuo. The white solid residue was washed
with light petroleum. Crystals suitable for single-crystal X-ray deter-
mination were obtained by layering a CH2Cl2 solution of the product
NMR Studies. Samples for studies at room temperature were
prepared from C6D6 stock solutions of the corresponding zirconocene
(0.004 mol/L, Cp2ZrMe2, rac-Me2Si(Ind)2ZrMe2, Cp′′2ZrMe2), and the
activators (0.002 mol/L). Stock solutions were prepared and handled
in a glovebox. NMR samples were prepared by pipetting the required
volumina of the zirconocene and borate solutions using an Eppendorff
pipet, accuracy (2 µL. The total zirconocene concentration ([Zr]tot) in
the NMR tube was kept under ∼3 mmol/L (typically 0.6 mmol/L <
[Zr]tot < 2.5 mmol/L) to prevent the formation of higher aggregates
and the precipitation of ionic species as “orange-red oils”. The solubility
of [CPh3]2[Ni{CNB(C6F5)3}4] in benzene was insufficient, and up to
20 vol % of o-C6H4F2 had to be added; in this way samples of
(41) Bochmann, M.; Lancaster, S. J. J. Organomet. Chem. 1992, 434,
C1.

Anion Effects in Olefin Polymerization
J. Am. Chem. Soc., Vol. 123, No. 2, 2001 235
[Zr]tot ) ∼3 mmol/L were prepared. The samples studied at lower
temperatures in CD2Cl2 were prepared by weighing each component
directly into the NMR tube (glovebox), followed by the addition of
the cooled solvent. The polar solvent allowed concentrations of
Zr-Me side), 0.34 (s, 3H, Me2Si, Me-B side), 0.20 (s, 3H, Me2Si,
Zr-Me side), -0.44 (br, 3H, Me-B), -0.51 (s, 3H, Zr-Me). (SBI)-
ZrMe(µ-NC)B(C6F5)3: 1H NMR (300 MHz, C6D6, 298 K) δ 5.54 (d,
1H, R-C5H2), 5.09 (d, 1H, R-C5H2), 0.39 (s, 3H, Me2Si), 0.28 (s, 3H,
Me2Si), -0.58 (s, 3H, Zr-Me).
1
1
[Zr]tot ) 55 mmol/L. Assignments are based on H, H-1H COSY,
1
13C and H-13C HETCOR spectra.
Cp′′2ZrMe2/1/AlMe3. Into the NMR tube were weighed 15 mg of
Cp′′2ZrMe2 (2.77 × 10-5 mol) and 36 mg of 1 (2.78 × 10-5 mol) . To
this mixture was added 14 µL (2.8 × 10-5 mol) of a toluene solution
of AlMe3 (2 mol/L), followed by 0.5 mL pre-cooled (-78 °C) CD2-
Cl2, [Zr] ) 0.0554 mol/L. On heating to room temperature the binuclear
species proved stable in CD2Cl2 for a few minutes, although decom-
position was complete after ∼10 min. [Cp′′2Zr(µ-Me)2AlMe2][CN-
{B(C6F5)3}2]: 1H NMR (300 MHz, CD2Cl2, 223 K) δ 7.04 (s, 2H,
C5H3), 6.94 (s, 4H, C5H3), 0.46 (s, 6H, µ-Me), 0.26 (s, 36H, SiMe3),
-0.51 (s, 6H, AlMe) (the expected couplings of the Cp-signals are not
resolved under these conditions). 13C NMR (75.5 MHz, CD2Cl2, 223
K) δ 128.6 (4C, C5H3), 125.0 (2C, C5H3), 37.9 (µ-Me), -0.4 (SiMe3),
-6.7 (AlMe) (quaternary Cp carbons not detected).
The System L2ZrMe2/[CPh3][B(C6F5)4]. The reactions of the
zirconocene complexes L2ZrMe2 ) Cp2ZrMe2 and rac-Me2Si(Ind)2ZrMe2
with [CPh3][B(C6F5)4] were studied at various Zr:B ratios. Addition of
L2ZrMe2 to an excess of [CPh3][B(C6F5)4] gives the binuclear zir-
conocene cations [(L2ZrMe)2(µ-Me)+‚‚‚B(C6F5)4-] for both zirconocene
systems in benzene solution at 298 K. The mononuclear product
[L2ZrMe+‚‚‚B(C6F5)4-] is detected after a few minutes.
[(Cp2ZrMe)2(µ-Me)+‚‚‚B(C6F5)4-].21a From C6D6 solutions of Cp2-
ZrMe2 (0.004 mol/L, 220 µL) and [CPh3][B(C6F5)4] (220 µL, 0.002
mol/L), [Zr] ) 0.002 mol/L. [(Cp2ZrMe)2(µ-Me)+‚‚‚B(C6F5)4-] was
detected as the sole product. 1H NMR (300 MHz, C6D6, 298 K) δ 5.56
(s, 20H, C5H5), -0.13 (s, 6H, ZrMe), -1.27 (s, 3H, µ-Me).
[Cp2ZrMe+‚‚‚B(C6F5)4-]: C6D6 solutions of Cp2ZrMe2 (0.004 mol/
L, 100 µL) and Ph3C+B(C6F5)4- (0.002 mol/L, 300 µL) were transferred
into the NMR tube ([Zr] ) 0.001 mol/L). Initially formed [(Cp2ZrMe)2-
(µ-Me)+‚‚‚B(C6F5)4-] reacts with excess trityl borate to give [Cp2-
L2ZrMe2/4c. A mixture of Cp2ZrMe2 in C6D6 (400 µL, 0.004 mol/
L) and 4c in in C6D6/o-C6H4F2 (4:1) (100 µL, 0.004 mol/L) gave
2-
exclusively the ion pair [{(Cp2ZrMe)2(µ-Me)+}2‚‚‚Ni{CNB(C6F5)3}4
]
([Zr] ) 0.0032 mol/L). Due to the higher polarity of the solvent mixture
the signals for the cation show a low-field shift of up to 0.15 ppm
compared to that in pure C6D6. Decomposition of [(Cp2ZrMe)2(µ-Me)]+
was complete after 60 min, while the SBI analogue showed only minor
signs of decomposition after 30 min. In both cases L2ZrMe(µ-Me)B-
(C6F5)3 was formed as the only detectable decomposition product.
[(Cp2ZrMe)2(µ-Me)]2[Ni{CNB(C6F5)3}4]: 1H NMR (300 MHz, C6D6/
o-C6H4F2 24:1, 298 K) δ 5.81 (s, 20H, Cp), 0.06 (s, 6H, Zr-Me), -1.11
1
ZrMe+‚‚‚ B(C6F5)4-]. H NMR (300 MHz, C6D6, 298 K) δ 5.20 (s,
20H, Cp), 0.15 (s, 3H, ZrMe).
[{(SBI)ZrMe}2(µ-Me)+‚‚‚B(C6F5)4-].21a From (SBI)ZrMe2 (220 µL,
0.004 mol/L) and [CPh3][B(C6F5)4] (220 µL, 0.002 mol/L) in C6D6,
[Zr] ) 0.002 mol/L. After a few minutes [{(SBI)ZrMe}2(µ-Me)+‚‚‚
B(C6F5)4-] is observed as the sole product. 1H NMR (300 MHz, C6D6,
298 K) δ -0.96 (s, 6, Zr-Me, rac-like), -1.01 (s, 6, Zr-Me, meso-
like), -2.91 (s, 3, µ-Me, rac-like), -3.10 (s, 3, µ-Me, meso-like). The
overlapping indenyl signals are not listed.
1
(s, 3H, µ-Me). Cp2ZrMe(µ-Me)B(C6F5)3: H NMR (300 MHz, C6D6
/o-C6H4F2 24:1, 298 K) δ 5.42 (s, 10H, Cp), 0.29 (s, 3H, Zr-Me),
0.15 (br, 3H, B-Me). [{(SBI)ZrMe}2(µ-Me)]2[Ni{CNB(C6F5)3}4]:
From (SBI)ZrMe2 in C6D6 (400 µL, 0.004 mol/L) and 4c in in C6D6/
o-C6H4F2 (4:1) (100 µL, 0.004 mol/L), [Zr] ) 0.0032 mol/L. 1H NMR
(300 MHz, C6D6/o-C6H4F2 24:1, 298 K) δ -0.94 (s, 6H, Zr-Me, meso-
like), -0.99 (s, 6H, Zr-Me, rac-like), -2.82 (s, 3H, µ-Me, rac-like),
-3.05 (s, 3H, µ-Me, meso-like).
[(SBI)ZrMe+‚‚‚B(C6F5)4-]. From (SBI)ZrMe2 (100 µL, 0.004 mol/
L) and [CPh3][B(C6F5)4] (500 µL, 0.002 mol/L) in C6D6, [Zr] )
6.67 × 10-4 mol/L. The initially formed [{(SBI)ZrMe}2(µ-Me)+‚‚‚
B(C6F5)4-] reacts slowly to [(SBI)ZrMe+‚‚‚B(C6F5)4-]. The reaction
is complete after 65 min. NMR spectra were measured every 5 min. A
pseudo-first-order rate constant k ) 3 × 10-4 s-1 was determined by
evaluation of the integrals of the two observed zirconocene species.
1H NMR (300 MHz, C6D6, 298 K) δ 5.15 (br, 2, R-C5H2), 0.32 (br, 6,
Si-Me), -0.79 (s, 3, Zr-Me).
Alkene Polymerizations. Normal pressure polymerizations were
performed in a flame-dried glass flask (250 mL for small scale reactions,
2 L for reactions with 1 L solvent volume) equipped with an efficient
magnetic stirrer bar and an internal thermometer. The flask was
evacuated, filled with monomer gas, followed by the required amount
of toluene and triisobutyl aluminum (toluene solution, [Al] ) 0.1 mol/
L). Temperature equilibration was ensured by stirring the mixture on
a water bath of the required temperature for ∼10 min (20 min for 1 L
reactions). The required amount of a stock solution of the metallocene
catalyst precursor in toluene (2 µmol/mL) was then injected using a
gastight syringe with Teflon plunger, followed by the injection of the
activator in toluene (1 µmol/mL). Polymerization times were measured
from that point. The stirrer rate was 1200 rpm. All stock solutions were
prepared freshly prior to polymerization runs and used within 2 h
(usually within 30 min) to ensure consistent results. The polymerization
was stopped by the rapid injection of 2 mL of methanol. The polymer
was precipitated by pouring the contents of the flask into a large volume
of acidified methanol, filtered and dried at 90 °C to constant weight.
13C NMR spectra of polymers were recorded in 1,2-C2D2Cl4 at 110
°C.
The System L2ZrMe2/1. Mixing C6D6 solutions of Cp2ZrMe2 (220
µL, 0.004 mol/L) and 1 (220 µL, 0.002 mol/L) at 25 °C gives [(Cp2-
ZrMe)2(µ-Me)+‚‚‚(C6F5)3BCNB(C6F5)3-] together with decomposition
products ([Zr]tot ) 0.002 mol/L). Decomposition is complete after 5
min. The products Cp2ZrMe(µ-Me)B(C6F5)3 and one other species,
assigned to Cp2ZrMe(µ-NC)B(C6F5)3, were detected. The ion pair
[{(SBI)ZrMe}2(µ-Me)}+‚‚‚(C6F5)3BCNB(C6F5)3-] is more stable in
C6D6. [(Cp2ZrMe)2(µ-Me)+‚‚‚(C6F5)3BCNB(C6F5)3-]: 1H NMR (300
MHz, C6D6, 298 K) δ 5.50 (s, 20H, Cp), -0.14 (s, 6H, Zr-Me), -1.40
(s, 3H, µ-Me). Cp2ZrMe(µ-Me)B(C6F5)3:42 1H NMR (300 MHz, C6D6,
298 K) δ 5.39 (s, 10H, Cp), 0.28 (s, 3H, Zr-Me), 0.1 (br, 3H, Zr-
Me). Cp2ZrMe(µ-NC)B(C6F5)3: 1H NMR (300 MHz, C6D6, 298 K)
δ 5.55 (s, 10H, Cp), 0.26 (s, 3H, Zr-Me). [{(SBI)ZrMe}2(µ-
Me)+‚‚‚(C6F5)3BCNB(C6F5)3-]: From (SBI)ZrMe2 (300 µL, 0.004 mol/
1
L) and 1 (250 µL, 0.002 mol/L) in C6D6, [Zr] ) 0.0022 mol/L. H
NMR (300 MHz, C6D6, 298 K) δ 0.62 (s, 6H, Si-Me, rac-like), 0.59
(s, 6H, Si-Me, meso-like), 0.45 (s, 6H, Si-Me, meso-like), 0.42 (s,
6H, Si-Me, rac-like), -1.00 (s, 6H, Zr-Me, rac-like), -1.06 (s, 6H,
Zr-Me, meso-like), -2.98 (s, 3H, µ-Me, rac-like), -3.13 (s, 3H, µ-Me,
For reactions under pressure, a 5 L Bu¨chi stainless steel autoclave
equipped with a pressure buret and connected to a Bu¨chi gas flow
controller was used. The reactor was heated out in vacuo at 100 °C
(Julabo FP50 heater) for 2 h, allowed to cool, and charged with toluene
(3L) via a wide-bore transfer cannula. The required aliquots of toluene
solutions of AlBui3 and the metallocene were added, the pressure buret
was charged with a toluene solution of the activator, and the autoclave
was pressurized to 7 bar ethene pressure and allowed to equilibrate at
60 °C for 30 min. The buret was pressurized with argon (8 bar), and
the activator injected rapidly with vigorous mechanical stirring. The
reaction was terminated by the injection of 10 mL of methanol via the
pressure buret, the reactor was vented, and the polymer was collected
and worked up as described above.
1
meso-like). H NMR (300 MHz, CD2Cl2, 223 K) δ 1.11 (s, 6H, Si-
Me, rac-like), 1.03 (s, 6H, Si-Me, meso-like), 0.95 (s, 6H, Si-Me,
rac-like), 0.94 (s, 6H, Si-Me, meso-like), -0.94 (s, 6H, Zr-Me, meso-
like), -0.94 (s, 6H, Zr-Me, rac-like), -2.80 (s, 3H, µ-Me, rac-like),
-2.98 (s, 3H, µ-Me, meso-like). Complete decomposition could only
be detected after 24 h at 298K. Mononuclear zirconocene cations were
not detected. (SBI)ZrMe(µ-Me)B(C6F5)3:22 1H NMR (300 MHz, C6D6,
298 K) δ 6.57 (d, 1H, R-C5H2, Zr-Me side), 6.22 (d, 1H, R-C5H2,
Me-B side), 5.66 (d, 1H, R-C5H2, Me-B side), 4.97 (d, 1H, R-C5H2,
(42) Yang, X.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1994, 116,
10015.

236 J. Am. Chem. Soc., Vol. 123, No. 2, 2001
Table 7. Crystal Data for Compounds 1, 2, 4b
Zhou et al.
1
2
4b‚2(CH3)2CO
empirical formula
fw
temp (K)
crystal size (mm)
crystal system
space group
a, Å
C56H15B2F30N
1293.31
150(2)
C37H15BClF15
790.75
150(2)
0.51 × 0.33 × 0.24
monoclinic
P21/c
12.1402(2)
7.91830(10)
33.0914(4)
90
99.3490(7)
90
3138.81(8)
4
1.673
0.241
C92H22B4F60N6Ni‚2(CH3)2CO
2569.26
150(2)
0.6 × 0.36 × 0.18
0.30 × 0.18 × 0.18
triclinic
triclinic
P1h
P1h
13.3342(2)
14.8405(2)
15.8631(2)
116.4370(6)
90.9440(7)
113.6410(5)
2501.10(6)
2
13.1128(3)
13.9811(4)
14.6945(3)
116.1280(13)
93.7390(14)
90.4300(12)
2411.55(10)
1
b, Å
c, Å
R, deg
â, deg
γ, deg
V, (Å3)
Z
D
calcd (g cm-3
)
1.717
0.178
1276
1.769
0.374
1268
µ, mm-1
F(000)
1576
max, min transmissn
θ range, deg
0.9686 and 0.9005
1.79 e θ e 26.00
-16 e h e 16,
-18 e k e 18,
-19 e l e 19
58348
9800 (Rint ) 0.0304)
803
1.034
0.0355
0.9444 and 0.8869
2.27 e θ e 27.42
-15 e h e 15,
-10 e k e 10,
-42 e l e 42
73682
7125 (Rint ) 0.0800)
488
1.031
0.0373
0.9357 and 0.8961
2.11 e θ e 26.00
-16 e h e 16,
-17 e k e 17,
-18 e l e 18
63405
9478 (Rint ) 0.0692)
776
1.022
0.0409
index range
no. of reflections collected
no. of unique reflections, n
no. of parameters, p
goodness of fit on F2, s
R1 [I > 2σ(I)]
wR2 (all data)
0.0972
0.0895
0.1047
weighting parameters a, b
extinction parameter
largest diff, peak, and hole e Å-3
0.0394, 1.0052
0.0048(11)
0.507 and -0.257
0.0429, 1.2710
0.0022(5)
0.429 and -0.325
0.0435, 1.6787
-
0.806 and -0.357
a Definitions: Rint ) (Σ|Fo - Fo (mean)|ΣFo ); S ) ((Σw(Fo - Fc2)2)/(n - p))1/2; wR2 ) ((Σw(Fo - Fc2)2/Σw(Fo )2)1/2; R1 ) (Σ||Fo| - |Fc||/
2
2
2
2
2
2
Σ|Fo|); weighting scheme w ) [σ2(Fo ) + (aP)2 + bP]-1, where P ) [2Fc2 + Max(Fc2, 0)]/3.
2
Table 8. Crystal Data for Compounds 4c, 5c, and 6
4c‚2CH2Cl2
5c‚2CH2Cl2
6‚C7H8
empirical formula
fw
temp (K)
crystal size (mm)
crystal system
space group
a, Å
C114H30B4F60N4Ni‚2CH2Cl2
2867.22
150(2)
0.36 × 0.15 × 0.12
monoclinic
P21/c
12.03400(10)
14.62970(10)
31.8468(3)
90
100.6920(6)
90
5509.41(8)
2
1.728
0.430
C114H30B4F60N4Pd‚2CH2Cl2
2914.92
150(2)
0.69 × 0.42 × 0.38
monoclinic
P21/c
C42H45BF15NSi4Zr‚C7H8
1155.31
150(2)
0.24 × 0.24 × 0.09
triclinic
P1h
12.01790
11.9439(5)
12.9731(5)
18.9530(6)
77.677(2)
b, Å
c, Å
14.7304(3)
31.88720(3)
90
100.2870
90
5551.55(15)
2
1.744
0.419
2864
R, deg
â, deg
74.010(2)
75.461(2)
2699.7(2)
2
1.421
0.378
1180
γ, deg
V, (Å3)
Z
D
calcd (g cm-3
)
µ, mm-1
F(000)
2828
max, min transmissn
θ range, deg
0.9503 and 0.8607
2.40 e θ e 26.00
-14 e h e 14,
-18 e k e 18,
-39 e l e 39
114737
10807 (Rint ) 0.0419)
854
1.006
0.0344
0.9503 and 0.8607
2.40 e θ e 26.00
-14 e h e 14,
-18 e k e 18,
-39 e l e 39
31285
10818 (Rint ) 0.0552)
854
1.026
0.0393
0.9668 and 0.9147
2.12 e θ e 26.00
-13 e h e 14,
-15 e k e 15,
-22 e l e 23
37640
10561 (Rint ) 0.0672)
655
1.102
0.0455
index range
no. of reflections collected
no. of unique reflections, n
no. of parameters, p
goodness of fit on F2, s
R1 [I > 2σ(I)]
wR2 (all data)
0.0857
0.1042
0.1096
weighting parameters a, b
extinction parameter
largest diff, peak, and hole e Å-3
0.0376, 3.1149
0.00045(14)
0.527 and -0.422
0.0487, 3.9774
0.0010(4)
0.602 and -0.709
0.0370, 1.5817
0.0121(10)
0.684 and -0.520
a Definitions: Rint ) (Σ|Fo - Fo (mean)|ΣFo ); S ) ((Σw(Fo - Fc2)2)/(n - p))1/2; wR2 ) ((Σw(Fo - Fc2)2/Σw(Fo )2)1/2; R1 ) (Σ||Fo| - |Fc||/
2
2
2
2
2
2
Σ|Fo|); weighting scheme w ) [σ2(Fo ) + (aP)2 + bP]-1, where P ) [2Fc2 + Max(Fc2, 0)]/3.
2
X-ray Crystallography. In each case a suitable crystal was coated
150 K on a Nonius Kappa CCD area-detector diffractometer. Data
in an inert perfluoropolyether oil and mounted in a nitrogen stream at
collection was performed using Mo KR radiation (λ ) 0.710 73 Å)

Anion Effects in Olefin Polymerization
J. Am. Chem. Soc., Vol. 123, No. 2, 2001 237
with the CCD detector placed 30 mm from the sample via a mixture
of 1° φ and ω scans at different θ and κ settings using the program
COLLECT.43 The raw data were processed to produce conventional
data using the program DENZO-SMN.44 The datasets were corrected
for absorption using the program SORTAV.45 All structures were solved
by heavy-atom methods using SHELXS-9746 and were refined by full-
matrix least squares refinement (on F2) using SHELXL-97.47 All non-
hydrogen atoms were refined with anisotropic displacement parameters.
Hydrogen atoms were constrained to idealized positions. Crystal-
lographic data for compounds 1, 2, 4b‚Me2CO, 4c‚2CH2Cl2, 5c‚2CH2-
Cl2, and 6‚toluene are summarized in Tables 7 and 8.
Acknowledgment. This work was supported by the Engi-
neering and Physical Sciences Research Council and the
University of Leeds.
(43) Collect, data collection software, Nonius, B. V.: Delft, The
Netherlands, 1999.
(44) Otwinowski Z.; Minor, W. In Methods in Enzymology; Carter, C.
W., Jr., Sweet, R. M., Eds.; Academic Press: New York, 1996; pp 276,
307.
(45) Blessing, R. H. Acta Crystallogr., Sect. A 1995, 51, 33.
(46) Sheldrick, G. M. Acta Crystallogr., Sect. A 1990, 46, 467.
(47) Sheldrick, G. M. SHELXL-97, Program for crystal structure refine-
ment. University of Go¨ttingen: Germany, 1997.
Supporting Information Available: Details of crystal
structure determinations for compounds 1, 2, 4b, 4c, 5c, and 6
(PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
JA002820H