Nonchelated alkene and alkyne complexes of d0 zirconocene pentafluorophenyl cations

Article abstract of DOI:10.1021/ja057524p

This paper describes the generation and properties of nonchelated d ⁰ zirconocene-aryl-alkene and alkyne adducts that are stabilized by the presence of β-SiMe₃ substituents on the substrates and the weak nucleophilicity of the -C₆F₅ ligand. The cationic complexes [(C₅H₄R)₂Zr(C₆F ₅)][B(C₆F₅)₄] (4a: R = H, 4b: R = Me) were generated by methide abstraction from (C₅H ₄R)₂Zr(C₆F₅)Me by Ph ₃C⁺. NMR studies show that 4a,b contain an o-CF...Zr dative interaction and probably coordinate a PhCl molecule in PhCl solution. Addition of allyltrimethylsilane (ATMS) to 4a,b in C₆D₅Cl solution at low temperature produces an equilibrium mixture of (C ₅H₄R)₂Zr(C₆F₅)(H ₂C=CHCH₂SiMe₃)⁺ (7a,b), 4a,b, and free ATMS. Similarly, addition of propargyltrimethylsilane (PTMS) to 4a produces an equilibrium mixture of Cp₂Zr(C₆F₅)(HC= CCH₂SiMe₃)⁺ (8a), 4a, and free PTMS. The NMR data for 7a,b,and 8a are consistent with highly unsymmetrical substrate coordination and substantial polarization of the substrate multiple bond with significant positive charge buildup at C_int and negative charge buildup at C_term. PTMS binds to 4a more strongly than ATMS does. The ATMS adducts undergo nondissociative alkene face exchange ("alkene flipping"), i.e., exchange of the (C₅H₄R) ₂Zr(C₆F₅)⁺ unit between the two alkene enantiofaces without decomplexation of the alkene, on the NMR time scale.

Full text of DOI:10.1021/ja057524p

Published on Web 06/07/2006
Nonchelated Alkene and Alkyne Complexes of d0 Zirconocene
Pentafluorophenyl Cations
Edward J. Stoebenau, III and Richard F. Jordan*
Contribution from the Department of Chemistry, The UniVersity of Chicago,
5735 South Ellis AVenue, Chicago, Illinois 60637
Received November 3, 2005; Revised Manuscript Received April 26, 2006; E-mail: rfjordan@uchicago.edu
Abstract: This paper describes the generation and properties of nonchelated d0 zirconocene-aryl-alkene
and alkyne adducts that are stabilized by the presence of â-SiMe3 substituents on the substrates and the
weak nucleophilicity of the -C6F5 ligand. The cationic complexes [(C5H4R)2Zr(C6F5)][B(C6F5)4] (4a: R )
H, 4b: R ) Me) were generated by methide abstraction from (C5H4R)2Zr(C6F5)Me by Ph3C+. NMR studies
show that 4a,b contain an o-CF‚‚‚Zr dative interaction and probably coordinate a PhCl molecule in PhCl
solution. Addition of allyltrimethylsilane (ATMS) to 4a,b in C6D5Cl solution at low temperature produces an
equilibrium mixture of (C5H4R)2Zr(C6F5)(H2CdCHCH2SiMe3)+ (7a,b), 4a,b, and free ATMS. Similarly, addition
of propargyltrimethylsilane (PTMS) to 4a produces an equilibrium mixture of Cp2Zr(C6F5)(HCtCCH2SiMe3)+
(8a), 4a, and free PTMS. The NMR data for 7a,b,and 8a are consistent with highly unsymmetrical substrate
coordination and substantial polarization of the substrate multiple bond with significant positive charge
buildup at Cint and negative charge buildup at Cterm. PTMS binds to 4a more strongly than ATMS does. The
ATMS adducts undergo nondissociative alkene face exchange (“alkene flipping”), i.e., exchange of the
(C5H4R)2Zr(C6F5)+ unit between the two alkene enantiofaces without decomplexation of the alkene, on the
NMR time scale.
Scheme 1
Introduction
Metallocene-catalyzed polymerization of alkenes occurs by
a coordination-insertion mechanism via intermediate metal-
alkyl-alkene species (Scheme 1).1 Studies of discrete d0 metal-
(σ-carbyl)-alkene complexes would be very helpful in under-
standing this important process. However, due to weak alkene
binding and low insertion barriers,2 observation of these
intermediates is exceedingly difficult, and instead, model
complexes have been used to probe their properties.
Chelated d0 metal-alkoxide-alkene complexes,3 chelated d0
metal-alkyl-alkene complexes,4 and nonchelated d0 metal-
alkene complexes without σ-carbyl ligands are known.5 The only
reported nonchelated d0 metal-alkyl-alkene complexes are the
Cp*2YR(alkene) species (A; Cp* ) C5Me5; R ) CH2CH2-
CHMe2, CH2CHMe2, CH2(CH2)4Me, cyclopentyl) described by
Casey,6 and the [(indenyl)2ZrMe][η2-indole‚B(C6F5)3] ion pair
(B) reported by Resconi,7 which are shown in Chart 1. The
yttrium-alkene adducts were not directly observed, but low-
temperature NMR spectra of mixtures of Cp*2YR and alkene
contained one set of exchange-averaged alkene resonances that
are shifted in the same direction from the free alkene positions
as for chelated yttrium-alkene complexes,4a indicating the
reversible formation of Cp*2YR(alkene) adducts. These species
undergo insertion above -100 °C, except when R ) cyclopen-
tyl. Here we describe the generation and properties of nonch-
elated d0 Zr-aryl-alkene and -alkyne complexes.8
(1) (a) Cossee, P. Tetrahedron Lett. 1960, 17, 12. (b) Cossee, P. J. Catal. 1964,
3, 80. (c) Arlman, E. J.; Cossee, P. J. Catal. 1964, 3, 99. (d) Brookhart,
M.; Green, M. L. H. J. Organomet. Chem. 1983, 250, 395. (e) Grubbs, R.
H.; Coates, G. W. Acc. Chem. Res. 1996, 29, 85.
(2) (a) Margl, P.; Deng, L.; Ziegler, T. Top. Catal. 1999, 7, 187. (b) Fan, L.;
Harrison, D.; Woo, T. K.; Ziegler, T. Organometallics 1995, 14, 2018. (c)
Woo, T. K.; Fan, L.; Ziegler, T. Organometallics 1994, 13, 2252. (d) Lanza,
G.; Fragala`, I. L. Top. Catal. 1999, 7, 45. (e) Lauher, J. W.; Hoffmann, R.
J. Am. Chem. Soc. 1976, 98, 1729.
(3) (a) Carpentier, J. F.; Wu, Z.; Lee, C. W.; Stro¨mberg, S.; Christopher, J.
N.; Jordan, R. F. J. Am. Chem. Soc. 2000, 122, 7750. (b) Carpentier, J.-F.;
Maryin, V. P.; Luci, J.; Jordan, R. F. J. Am. Chem. Soc. 2001, 123, 898.
(c) Wu, Z.; Jordan, R. F.; Petersen, J. L. J. Am. Chem. Soc. 1995, 117,
5867.
(4) Leading references: (a) Casey, C. P.; Klein, J. F.; Fagan, M. A. J. Am.
Chem. Soc. 2000, 122, 4320. (b) Casey, C. P.; Carpenetti, D. W., II.
Organometallics 2000, 19, 3970. (c) Brandow, C. G.; Mendiratta, A.;
Bercaw, J. E. Organometallics 2001, 20, 4253. (d) Casey, C. P.; Carpenetti,
D. W., II; Sakurai, H. Organometallics 2001, 20, 4262. (e) Cano, J.; Go´mez-
Sal, P.; Heinz, G.; Mart´ınez, G.; Royo, P. Inorg. Chim. Acta 2003, 345,
15. (f) Mart´ınez, G.; Royo, P. Organometallics 2005, 24, 4782.
(5) (a) Witte, P. T.; Meetsma, A.; Hessen, B.; Budzelaar, P. H. M. J. Am.
Chem. Soc. 1997, 119, 10561. (b) Humphries, M. J.; Douthwaite, R. E.;
Green, M. L. H. J. Chem. Soc., Dalton Trans. 2000, 2952. (c) Kress, J.;
Osborn, J. A. Angew. Chem., Int. Ed, Engl. 1992, 31, 1585.
(6) (a) Casey, C. P.; Tunge, J. A.; Lee, T.-Y.; Fagan, M. A. J. Am. Chem. Soc.
2003, 125, 2641. (b) Casey, C. P.; Lee, T.-Y.; Tunge, J. A.; Carpenetti, D.
W., II. J. Am. Chem. Soc. 2001, 123, 10762.
(7) (a) Bonazza, A.; Cumurati, I.; Guidotti, S.; Mascellari, N.; Resconi, L.
Macromol. Chem. Phys. 2004, 205, 319. (b) For related species see: Kehr,
G.; Fro¨hlich, R.; Wibbeling, B.; Erker, G. Chem. Eur. J. 2000, 6, 258.
(8) Preliminary communication: Stoebenau, E. J., III; Jordan, R. F. J. Am.
Chem. Soc. 2004, 126, 11170.
9
8638
J. AM. CHEM. SOC. 2006, 128, 8638-8650
10.1021/ja057524p CCC: $33.50 © 2006 American Chemical Society

Alkene/Alkyne Coordination to Zr
−(σ
-Carbyl) Complexes
A R T I C L E S
-
Chart 1
Scheme 3. Anion ) B(C6F5)4
Results and Discussion
Targets. For this work, we were interested in systems in
which alkene and alkyne coordination to a d0 Zr-σ-carbyl
complex could be studied in the absence of insertion reactions.
A two-component strategy was used. First, to circumvent the
problems posed by weak substrate binding, the â-Si-substituted
substrates allyltrimethylsilane (ATMS) and propargyltrimeth-
ylsilane (PTMS) were used. These substrates coordinate strongly
to [Cp’2Zr(OtBu)][B(C6F5)4] (1, Cp’ ) C5H4Me) due to â-Si
stabilization of the partial positive charge at Cint of the bound
substrates.8 Second, to inhibit insertion, the electron-deficient,
poorly nucleophilic pentafluorophenyl (C6F5) ligand was used
as the σ-carbyl group.9
H2O or D2O to yield {(C5H4R)2ZrMe}2(µ-O)13 and C6F5H or
C6F5D by selective reaction of the Zr-C6F5 bond. One
explanation for this difference in reactivity is that initial
coordination of H2O to Zr facilitates proton transfer in the latter
reaction.14 Complex 2a is also unreactive with [Ph3C][B(C6F5)4]
and with B(C6F5)3 under these conditions.15
The reaction of 2a, Cp2ZrMe2, and [Ph3C][B(C6F5)4] in a
1:1:2 molar ratio was investigated to determine if Cp2Zr(C6F5)+
could be generated by the methide abstraction and ligand
exchange process shown in Scheme 3. A similar approach was
used for the synthesis of cationic Al alkyls.16 However, this
reaction generates only Cp2ZrMe+ and Ph3CMe (1 equiv), along
with unreacted 2a and [Ph3C][B(C6F5)4]. Exchange of a C6F5
group from 2a to Cp2ZrMe+ to generate Cp2Zr(C6F5)+ and 3a
is not observed.
Neutral (C5H4R)2Zr(C6F5)R Compounds. The pentafluo-
rophenyl compounds Cp2Zr(C6F5)2 (2a, Cp ) C5H5),10 Cp2Zr-
(C6F5)Me (3a),11 and Cp’2Zr(C6F5)Me (3b) were synthesized
by literature methods, as shown in Scheme 2.
Scheme 2
Generation of (C5H4R)2Zr(C6F5)+ Species. The lack of C6F5
abstraction from 2a suggested that selective methide abstraction
from 3a could provide access to Cp2Zr(C6F5)+ species. Indeed,
3a reacts with 1 equiv of [Ph3C][B(C6F5)4] in C6D5Cl at 22 °C
to yield [Cp2Zr(C6F5)][B(C6F5)4] (4a, 97%, Scheme 4) after 2
days. At shorter reaction times, mixtures of 4a, Ph3CMe, Ph3C+,
and the dinuclear species [{Cp2Zr(C6F5)}2(µ-Me)][B(C6F5)4]
(5a)11 are observed. Compound 5a forms by trapping of the
initially generated 4a by 3a (Scheme 4). Compound 4a could
not be isolated, but rather was generated and used in situ.
Similarly, the reaction of 3b with [Ph3C][B(C6F5)4] provides
[Cp’2Zr(C6F5)][B(C6F5)4] (4b, 93%, Scheme 4). Complex 4b
forms within 30 min, and no intermediates are observed in the
reaction.
Reactivity of 2a. The reactions of 2a with standard polym-
erization activators12 were investigated to determine if Zr-C6F5
-
bond protonolysis or C6F5 anion abstraction could provide
access to Cp2Zr(C6F5)+ species. Compound 2a does not react
with [HNPh2Me][B(C6F5)4] in C6D5Cl at 22 °C, even after 4
days. This lack of reactivity was unexpected, as 3a,b react with
The difference in the rate of formation of 4b vs 4a reflects
the difference in stabilities of [{Cp’2Zr(C6F5)}2(µ-Me)][B(C6F5)4]
(5b) and 5a. When 3b is treated with 0.5 equiv of [Ph3C]-
[B(C6F5)4] in C6D5Cl at 22 °C, an equilibrium mixture of 3b,
4b, and 5b is obtained. The NMR signals for all three species
are broad, implying that these species undergo intermolecular
exchange on the NMR time scale. The equilibrium constant for
(9) For M-C6F5 species, especially d0 compounds, see: (a) Chambers, R. D.;
Chivers, T. Organomet. Chem. ReV. 1966, 1, 279. (b) Alonso, P. J.; Favello,
L. R.; Fornie´s, J.; Garc´ıa-Monforte, M. A.; Menjo´n, B. Angew. Chem., Int.
Ed. 2004, 43, 5225. (c) Deacon, G. B.; Forsyth, C. M. Organometallics
2003, 22, 1349. (d) Hauber, S.-O.; Lissner, F.; Deacon, G. B.; Niemeyer,
M. Angew. Chem., Int. Ed. 2005, 44, 5871. (e) Hauber, S.-O.; Lissner, F.;
Deacon, G. B.; Niemeyer, M. Angew. Chem., Int. Ed. 2005, 44, 5871. (f)
Maron, L.; Werkema, E. L.; Perrin, L.; Eisenstein, O.; Andersen, R. A. J.
Am. Chem. Soc. 2005, 127, 279. For M-C6Cl5 species, see: (g) Ara, I.;
Fornie´s, J.; Garc´ıa-Monforte, M. A.; Mart´ın, A.; Menjo´n, B. Chem. Eur.
J. 2004, 10, 4186.
(10) (a) Dioumaev, V. K.; Harrod, J. F. Organometallics 1997, 16, 2798. (b)
Chaudhari, M. A.; Stone, F. G. A. J. Chem. Soc. A 1966, 838. (c) Coe, P.
L.; Stephens, R.; Tatlow, J. C. J. Chem. Soc. 1962, 3227.
(11) Bochmann, M.; Sarsfield, M. J. Organometallics 1998, 17, 5908.
(12) Chen, E. Y.-X.; Marks, T. J. Chem. ReV. 2000, 100, 1391 and references
therein.
formation of 5b from 3b and 4b is K1 ) [5b][4b]-1[3b]-1
)
410(20) M-1 (Scheme 4), as determined by 1H NMR spectros-
copy. In contrast, 5a is formed quantitatiVely from 3a and 0.5
(13) For (Cp2ZrMe)2(µ-O) see: Marsella, J. A.; Folting, K.; Huffman, J. C.;
Caulton, K. G. J. Am. Chem. Soc. 1981, 103, 5596.
(14) Hillhouse, G. L.; Bercaw, J. E. J. Am. Chem. Soc. 1984, 106, 5472.
(15) For use of B(C6F5)3 as a C6F5- abstracting agent, see: Fornie´s, J.; Mart´ın,
A.; Mart´ın, L. F.; Menjo´n, B.; Tsipis, A. Organometallics 2005, 24, 3539.
(16) Korolev, A. V.; Ihara, E.; Guzei, I. A.; Young, V. G., Jr.; Jordan, R. F. J.
Am. Chem. Soc. 2001, 123, 8291.
9
J. AM. CHEM. SOC. VOL. 128, NO. 26, 2006 8639

A R T I C L E S
Stoebenau and Jordan
-
Scheme 4. Anion ) B(C6F5)4
equiv of Ph3C+ under these conditions, implying that K1 is much
larger in this case (>5000 M-1).11 Cation 4a is expected to be
a stronger Lewis acid than 4b due to the weaker donor ability
of the Cp ligands compared to that of the Cp’ ligands.17 The
greater Lewis acidity of 4a vs 4b, and the additional steric
crowding present in 5b vs 5a, may explain the greater stability
of 5a vs 5b.
complexes.18a In contrast, neutral Zr-C6F5 compounds that lack
dative o-CF‚‚‚Zr interactions but undergo slow Zr-C6F5 rotation
typically display small (1-3 ppm) chemical shift differences
between their o-F resonances.18b,19 The 19F NMR spectrum of
4a is similar to that of 4b under these conditions.
Addition of bromobenzene (11 equiv) to a C6D5Cl solution
of 4a results in several changes in the NMR spectra at -38 °C.
The 1H NMR Cp resonance of 4a is shifted downfield by 0.03
ppm, and the p-F signal of 4a is broadened by 15-20 Hz,
compared to spectra in the absence of PhBr. Additionally, the
13C NMR resonances of PhBr are broadened by 2-6 Hz
compared to resonances for Ph3CMe, but are not shifted from
the normal PhBr positions. These effects are ascribed to minor
formation of a PhBr adduct that exchanges rapidly with 4a and
free PhBr and suggest that 4a,b exist as solvent adducts in
halocarbon solution. Zirconium-ClPh and -BrPh complexes
are well characterized.5a,b,20 The structure for 4a,b proposed in
Scheme 4, in which the o-CF‚‚‚Zr interaction occupies the
central coordination site, is based on the solid-state structure
of Cp*2Zr(κ2-C,Cl-o-Cl-phenyl)(NCMe)+.20a
Dynamics of (C5H4R)2Zr(C6F5)+ Species. The 19F NMR
spectrum of 4b at 22 °C contains one very broad o-F resonance,
a sharp p-F triplet resonance, and a slightly broad m-F resonance,
as shown in Figure 1. Two dynamic processes could produce
the observed coalescence of the o-F signals and of the m-F
signals: rotation around the Zr-C6F5 bond of 4b,21 or site
epimerization at Zr (i.e. exchange of the C6F5 and ClPh ligands
between the sides of the zirconocene wedge, probably by
exchange of free and coordinated PhCl), as shown in Scheme
5. Zr-C6F5 bond rotation permutes the sides of the C6F5 ring,
while site epimerization permutes the sides of the C6F5 ring
and the sides of the Cp’ rings.
Compound 4a is stable in C6D5Cl solution for weeks at 22
°C, but decomposes rapidly in CD2Cl2, even at -78 °C (t1/2
<
30 min), to C6F5H, C6F5D, [{Cp2Zr(C6F5)}2(µ-Cl)][B(C6F5)4]
(6a), and other Cp2Zr species. Compound 6a was generated
independently by the reaction of 4a with 0.5 equiv of [Bu3-
NCH2Ph]Cl. Compound 4b exhibits stability similar to that of
4a.
Solution Structure of (C5H4R)2Zr(C6F5)+. The 19F NMR
spectrum of 4b at -38 °C contains two broad o-F resonances
at δ -116.7 and -140.7, a sharp p-F triplet resonance, and
two broad m-F resonances, as shown in Figure 1. The downfield
o-F signal of 4b is in the normal range for Zr-C6F5 compounds
(δ -116 ( 10),11,18 while the upfield o-F signal is 20-25 ppm
upfield from the normal range. These results show that the sides
of the C6F5 ligand are inequivalent and suggest that the o-F
with the upfield resonance is datively coordinated to Zr. Dative
o-CF‚‚‚Zr interactions have been detected in other Zr-C6F5
The 13C{1H} NMR spectrum of 4b at -38 °C in C6D5Cl
contains three broad Cp’ CH resonances (Figure 2), one sharp
Cp’ ipso-C resonance, and one sharp Cp’Me resonance. The
(17) Wieser, U.; Babushkin, D.; Brintzinger, H.-H. Organometallics 2002, 21,
920.
(18) (a) Pindado, G. J.; Lancaster, S. J.; Thornton-Pett, M.; Bochmann, M. J.
Am. Chem. Soc. 1998, 120, 6816. (b) Kraft, B. M.; Jones, W. D. J.
Organomet. Chem. 2002, 658, 132.
Figure 1. Variable temperature 19F NMR spectra (471 MHz) of 4b.
Assignments for -38 °C spectrum: δ -117 (4b o-F), -132 (anion o-F),
-138 (C6F5H o-F), -141 (4b o-F), -150 (4b p-F), -154 (C6F5H), -156
(4b m-F), -159 (4b m-F). The dashed lines highlight the coalescence of
the two o-F and the two m-F resonances of 4b.
(19) Edelbach, E. L.; Rahman, A. K. F.; Lachicotte, R. J.; Jones, W. D.
Organometallics 1999, 18, 3170.
(20) (a) Wu, F.; Dash, A. K.; Jordan, R. F. J. Am. Chem. Soc. 2004, 126, 15360.
(b) Bochmann, M.; Jaggar, A. J.; Nicholls, J. C. Angew. Chem., Int. Ed.
Engl. 1990, 29, 780.
(21) Leblanc, J. C.; Mo¨ıse, C. Org. Magn. Reson. 1980, 14, 157.
9
8640 J. AM. CHEM. SOC. VOL. 128, NO. 26, 2006

Alkene/Alkyne Coordination to Zr
−(σ
-Carbyl) Complexes
A R T I C L E S
Scheme 5
Table 1. NMR Coordination Shifts (∆δ) for the ATMS Ligands in
ZrIV Compoundsa
cmpd
∆δ Cint
∆δ Cterm
∆δ Callylic
∆δ Hint
∆δ Htrans
∆δ Hcis
7a
7b
9
51.2
49.6
32.0
-15.4
-15.2
-20.7
12.2
11.9
8.8
1.89
1.90
1.59
-0.79
-0.72
-0.35
-1.01
-0.85
-0.43
a In C6D5Cl at -38 °C; ∆δ ) δcoord - δfree
.
transferred to a precooled NMR probe with minimal transfer
time (<1 min). This procedure minimized side reactions such
as the dimerization of ATMS (vida infra). The propargyltrim-
ethylsilane (PTMS) complex [Cp2Zr(C6F5)(HCtCCH2SiMe3)]-
[B(C6F5)4] (8a) was generated in an analogous manner. Com-
pounds 7a,b and 8a were characterized by NMR spectroscopy.
These compounds are very thermally sensitive, which precluded
isolation or ESI-MS analysis.
Cp’ CH resonances collapse to two sharp singlets at 22 °C
(Figure 2). The 1H NMR spectrum of 4b at -38 °C in C6D5Cl
contains two broad Cp’ CH resonances, which sharpen at 22
°C. These results show that the sides of the Cp’ ligands are
indeed exchanging rapidly on the NMR time scale, which
implies that site epimerization of 4b is rapid. The site epimer-
ization barrier is ∆Gq ) 11.2(1) kcal/mol at -38 °C, as
estimated from the line broadening of the o-F and m-F
resonances of 4b.
It is also possible that Zr-C6F5 bond rotation occurs
independently of the site epimerization process. If Zr-C6F5 bond
rotation occurred at a rate that is competitive with the site
epimerization rate, the o-F and m-F resonances of 4b would
exhibit greater exchange line broadening than the Cp’ CH
resonances in the slow-exchange region. However, at -38 °C,
the excess line widths of the noncoalesced 13C Cp’ CH
resonances and the 19F o-F and m-F resonances are ap-
proximately equal (ca. 200 Hz). Therefore, Zr-C6F5 bond
rotation in 4b does not occur at a significant rate at this
temperature.
2-Butyne does not coordinate to 4a in C6D5Cl at -38 °C,
even though it coordinates strongly to 1 under these conditions.
Therefore, simple alkenes such as ethylene or propylene are
not expected to bind to 4a, since these substrates bind much
more weakly to 1 compared to 2-butyne. Side reactions
precluded detection of binding of other substrates to 4a,b. For
example, while tert-butyl vinyl ether is expected to bind
strongly, based on the strong coordination of this substrate to
1, it is rapidly polymerized in the presence of 4a, presumably
by a cationic mechanism.
The variable-temperature 19F NMR spectra of 4a are similar
to those of 4b, and it is likely that site epimerization occurs in
this case as well.
Generation of Alkene and Alkyne Complexes of (C5H4R)2-
Zr(C6F5)+. The addition of excess allyltrimethylsilane (ATMS)
to 4a,b results in the formation of equilibrium mixtures of 4a,b,
free ATMS, and the alkene adducts [(C5H4R)2Zr(C6F5)-
(H2CdCHCH2SiMe3)][B(C6F5)4] (7a: R ) H; 7b: R ) Me),
as shown in eq 1. In a typical experiment, ATMS was added
by vacuum transfer to a frozen solution of 4a,b in C6D5Cl in
an NMR tube. The tube was thawed, stored at -40 °C, and
NMR Properties and Solution Structures of (C5H4R)2Zr-
1
(C6F5)(ATMS)+ Species. The H and 13C{1H} NMR spectra
of 7a at -38 °C in C6D5Cl each contain two Cp resonances,
indicative of C1 symmetry as expected due to the chiral center
at Cint of the ATMS ligand. Additionally, these spectra each
contain one set of resonances for the bound ATMS, consistent
with the presence of one rotamer or with rapid rotation around
the Zr-(alkene centroid) axis.22 The 19F NMR spectrum of 7a
contains one p-F resonance and one broad m-F resonance, but
the o-F resonance is broadened into the baseline. This result
shows that rotation around the Zr-C6F5 bond occurs on the
NMR time scale in this species.23 The NMR spectra of 7b are
similar to those of 7a. The -38 °C 19F NMR spectrum of 7b
contains two broad o-F resonances in the normal range for Zr-
C6F5 compounds, consistent with κ1-C6F5 coordination.
The ATMS ligands of 7a,b display characteristic 13C NMR
coordination shifts (∆δ ) δcoord - δfree), which are listed in
Table 1. The Cint resonances of 7a,b are shifted ca. 50 ppm
downfield, and the Cterm resonances are shifted ca. 15 ppm
(22) These ligands probably undergo free rotation around the Zr-(ligand
centroid) axis, based on the observation of fast rotation in Cp’2Zr(OtBu)-
(MetCMe)+ above -59 °C in CD2Cl2. Stoebenau, E. J., III; Jordan, R. F.
J. Am. Chem. Soc. 2003, 125, 3222.
(23) Site epimerization of 7a can be discounted because net exchange of free
and coordinated ATMS is significantly slower than the exchange of the
sides of the C6F5 ring (vida infra).
Figure 2. Partial variable temperature 13C{1H} NMR spectra (126 MHz)
of 4b showing the Cp’ CH resonances. The dashed lines highlight the
coalescence of two of the Cp’ CH resonances. The peaks marked “x” are
due to minor unknown impurities.
9
J. AM. CHEM. SOC. VOL. 128, NO. 26, 2006 8641

A R T I C L E S
Stoebenau and Jordan
Figure 3. Partial 13C{1H} NMR spectrum of an equilibrium mixture of 7a, 4a, and free ATMS in C6D5Cl at -38 °C. ATMS coordination shifts are shown
with dashed arrows. Assignments: δ 187 (7a Cint), 149 (Ph3CMe), 149 (d, anion), 139 (d, anion), 137 (d, anion), 135 (ATMS Cint), 134-126 (large, C6D5Cl
and Ph3CMe), 125 (br, anion), 117 (2 s, 4a and 6a Cp), 116 (br 2 s, 7a Cp), 113 (ATMS Cterm), 98 (7a Cterm).
upfield from the free ATMS resonances. The 13C{1H} NMR
spectrum of 7a is shown in Figure 3. The JCH values for Cint
Table 2. NMR Coordination Shifts (∆δ) and Coupling Constants
(Hz) for the PTMS Ligands in ZrIV Compoundsa
1
b
b
cmpd
∆δ Cint
∆δ Cterm
∆δ Cprop
∆δ Hterm
∆δ Hprop
1JC
2JC
tCH
tCH
(161 Hz) and Cterm (150 Hz) of 7a are not significantly perturbed
by coordination. The Callylic resonances of 7a,b are shifted ca.
12 ppm downfield from the free alkene value.
8ac
11d
62.5
21.7
10.5
-0.9
10.0
7.0
2.57
1.58
0.78
0.48
232
244
34
43
1
a ∆δ ) δcoord - δfree b HCtCCH2R resonance. c In C6D5Cl solution,
.
The H NMR resonances of the ATMS vinyl unit of 7a,b
-38 °C. d In CD2Cl2 solution, -89 °C.
are also perturbed by coordination (Table 1). The Hint resonances
shift 1.9 ppm downfield, and the Htrans and Hcis resonances shift
0.7-1.0 ppm upfield from the free ATMS resonances. Two
broad resonances for the diastereotopic Hallylic hydrogens are
observed for 7a,b, which are shifted downfield by 0.7-1.0 ppm
by coordination. The ATMS nJHH coupling constants for 7a and
7b are virtually unchanged from the free alkene values.
The NMR data for 7a,b are generally similar to data for [Cp’2-
Zr(OtBu)(H2CdCHCH2SiMe3)][B(C6F5)4] (9) and other ZrIV-
alkene complexes studied previously and thus imply the alkene
is bound in an unsymmetrical, η1-fashion (d(Zr-Cint) > d(Zr-
Moreover, as described below, 4a and 7a interconvert on the
NMR time scale, which would also require rapid reversible
insertion of ATMS.
NMR Properties and Solution Structure of Cp2Zr(C6F5)-
1
(HCtCCH2SiMe3)+. The H and 13C{1H} NMR spectra of
PTMS adduct 8a each contain one Cp resonance and one set of
PTMS resonances, consistent with the presence of a single
rotamer in which the PTMS ligand lies in the plane between
the two Cp rings or with fast rotation around the Zr-(alkyne
centroid) axis.22 The 19F NMR spectrum contains two broad
o-F resonances in the normal range for Zr-C6F5 compounds,18
a sharp p-F triplet resonance, and a broad m-F resonance,
consistent with κ1-C6F5 coordination and hindered Zr-C6F5
bond rotation on the NMR time scale.
C
term)), which results in polarization of the CdC bond with
partial positive charge on Cint and partial negative charge on
term, but not significant rehybridization of the alkene carbons.3,8
C
The larger ∆δ values for Cint and Hint for 7a,b compared to
those for 9 suggest that the ATMS CdC bond may be more
polarized in the former species than in the latter, i.e., that there
is a greater contribution of the carbocationic resonance form in
7a,b than in 9. This difference is reasonable because the
(C5H4R)2Zr(κ1-C6F5)+ fragments in 7a,b are expected to be
stronger Lewis acids than the Cp’2Zr(OtBu)+ fragment in 9, due
to the electron-withdrawing nature of the C6F5 group. In fact
the ∆δ values for Cint in 7a,b are similar to that observed for
[(C6F5)3BCH2CHdCH2][SnBu3] (i.e., the B(C6F5)3 adduct of
Bu3SnCH2CHdCH2; ∆δ ) 54), in which the carbocationic-
The NMR data for 8a are compared to those of [Cp’2Zr(Ot-
Bu)(HCtCCH2SiMe3)][B(C6F5)4] (11) in Table 2. For 8a, Cint,
Cprop, and Hterm all show large downfield coordination shifts,
and the ∆δ values for Cint and Hterm are substantially larger than
these values for 11 and other Cp’2Zr(OtBu)(HCtCR)+ com-
pounds. In addition, the coordination shift for Cterm of 8a is
downfield, while for 11 and similar species this coordination
shift is always slightly upfield. Finally, the JCH values of the
alkyne carbons of 8a are 10-15 Hz smaller than for 11. The
C
int resonance form is believed to make a substantial contribu-
JCH values for 8a imply essentially sp hybridization at the alkyne
tion.24
carbons but suggest that there is some bending of the H-CtC
unit. These results are consistent with unsymmetrical PTMS
coordination (d(Zr-Cint) > d(Zr-Cterm); C in Figure 4) and
concomitant polarization of the CtC bond with positive charge
buildup at Cint for both 8a and 11, but imply a greater degree
of polarization in 8a than in 11. In fact, the PTMS ligand of 8a
may possesses moderate vinyl cation character (D and E, Figure
4), as the 13C NMR Cint resonance of 8a (δ 145) is shifted more
than halfway from the free PTMS value (δ 83) to values for
silyl- and ferrocenyl-stabilized vinyl cations (δC 180-200).25
An alternative formulation of 7a,b as alkene insertion
products, i.e., (C5H4R)2Zr{CH2CH(C6F5)CH2SiMe3}+ or
(C5H4R)2Zr{CH(CH2SiMe3)CH2C6F5}+, is ruled out by NMR
1
data and reactivity properties. The JCH values of 7a are
characteristic for sp2 carbons and are too large for insertion
products in which these carbons would be sp3 hybridized. For
1
example, JCH ) 132 Hz for C6F5Me, which is a reasonable
model for Câ of an insertion product. The 1JCH value for CR of
an insertion product would be even lower. Also, addition of
THF to equilibrium mixtures of 7a and 4a at -38 °C results in
fast (<5 min) quantitative formation of [Cp2Zr(C6F5)(THF)]-
[B(C6F5)4] (10a) and free ATMS, which would require that
insertion be reversible on a fast time scale, which is unlikely.
(25) (a) Mu¨ller, T.; Juhasz, M.; Reed, C. A. Angew. Chem, Int. Ed. 2004, 43,
1543. (b) Mu¨ller, T.; Meyer, R.; Lennartz, D.; Siehl, H.-U. Angew. Chem.,
Int. Ed. 2000, 39, 3074. (c) Koch, E.-W.; Siehl, H.-U.; Hanack, M.
Tetrahedron Lett. 1985, 26, 1493. (d) Siehl, H.-U.; Kaufmann, F.-P.;
Apeloig, Y.; Braude, V.; Danovich, D.; Berndt, A.; Stamatis, N. Angew.
Chem., Int. Ed. Engl. 1991, 30, 1479. (e) Mu¨ller, T.; Margraf, D.; Syha,
Y. J. Am. Chem. Soc. 2005, 127, 10852.
(24) Blackwell, J. M.; Piers, W. E.; McDonald, R. J. Am. Chem. Soc. 2002,
124, 1295.
9
8642 J. AM. CHEM. SOC. VOL. 128, NO. 26, 2006

Alkene/Alkyne Coordination to Zr
−(σ
-Carbyl) Complexes
A R T I C L E S
Figure 4. Polarization of the CtC bond in 8a (X ) C6F5) and 11 (X )
OtBu) showing a contribution from â-Si-stabilized vinyl cation canonical
forms (D and E).
Figure 5. van’t Hoff plot for ATMS coordination in 7a in C6D5Cl.
Table 3. Equilibrium Constants (Keq) for Binding of ATMS and
PTMS (L) in (C5H4R)2Zr(X)(L)+a
The Keq value for binding of ATMS in 7b is ca. 3.5 times
lower than that for 7a, consistent with the weaker Lewis acidity
of 4b compared to 4a.17 There is no significant difference in
the Keq values for 7b and 9, even though Cp’2Zr(κ1-C6F5)+ is
expected to be a stronger Lewis acid than Cp’2Zr(OtBu)+.
However, the formation of 7b from 4b requires cleavage of the
dative o-CF‚‚‚Zr interaction and the Zr-ClPh bond, while the
formation of 9 from 1 requires only cleavage of the Zr-ClPh
bond.
The Keq value for binding of PTMS in 8a is 110 times higher
than that for ATMS binding in 7a, consistent with the trend
observed for Cp’2Zr(OtBu)(substrate)+ species, in which alkyne
coordination is normally stronger than alkene coordination. At
[4a]initial ) 0.054 M and [PTMS]initial ) 0.061 M, ca. 91% of
the total Cp2Zr(C6F5)+ exists as 8a at -38 °C in C6D5Cl. The
Keq value for 8a cannot be directly compared to that for 11
since PTMS binds too strongly to Cp’2Zr(OtBu)+ under these
conditions (C6D5Cl, -38 °C) for Keq to be measured in this
case. However, Keq for formation of 11 is estimated to be greater
than 1200 M-1, and so the Keq for formation of 8a is less than
that for 11.
1
cmpd
formula
K
eq (M-
)
8a
7a
9
Cp2Zr(C6F5)(HCtCCH2SiMe3)+
Cp2Zr(C6F5)(H2CdCHCH2SiMe3)+
Cp’2Zr(OtBu)(H2CdCHCH2SiMe3)+
Cp’2Zr(C6F5)(H2CdCHCH2SiMe3)+
910(60)
8.2(1.4)
2.9(7)
7b
2.4(2)
a At -38 °C in C6D5Cl; Keq ) [Zr-L][4]-1[L]-1
.
The greater CtC polarization in 8a vs 11 again reflects the
expected greater Lewis acidity of Cp2Zr(κ1-C6F5)+ vs Cp’2Zr-
(OtBu)+.
The NMR data and reactivity properties of 8a are inconsistent
with the alternative formulation of this species as an insertion
product, i.e., Cp2Zr{C(CH2SiMe3)dCHC6F5}+ or Cp2Zr{CHd
n
C(C6F5)CH2SiMe3}+. The JCH values of 8a are far too large
for an insertion product, in which these carbons would be sp2
hybridized. For example, in 2,3,4,5,6-pentafluorostyrene, a
reasonable model for the insertion products, the 1JCH values are
typical for sp2 carbons (Cterm: 161 Hz, Cint: 163 Hz), and the
2JCH values are not resolvable (<2 Hz). Also, while 8a exhibits
coupling between Cterm and the o-fluorines (JCF ) 8 Hz), such
coupling is not expected in an insertion product since it is not
detectable in 2,3,4,5,6-pentafluorostyrene. Finally, the PTMS
ligand is readily displaced by THF yielding 10a and free PTMS,
which would require that insertion be rapidly reversible.
Thermodynamics of Alkene and Alkyne Binding to
(C5H4R)2Zr(C6F5)+ Species. Equilibrium constants for substrate
binding to 4a,b in eq 1, Keq ) [Zr-L][4]-1[L]-1, were
determined by NMR spectroscopy and are listed in Table 3.26
The equilibrium constant for ATMS coordination in 7a is
Keq ) 8.2(1.4) M-1 at -38 °C in C6D5Cl solution. At [4a]initial
) 0.040 M and [ATMS]initial ) 0.20 M, ca. 59% of the total
Cp2Zr(C6F5)+ exists as 7a under these conditions. This equi-
librium constant is unchanged over the concentration ranges
[4a]initial ) 0.023-0.064 M and [ATMS]initial ) 0.11-2.10 M.
As the temperature is raised, the formation of 7a becomes less
favored, and the equilibrium shifts to 4a and free ATMS. A
van’t Hoff analysis for ATMS coordination to 4a (Figure 5)
gives ∆H° ) -5.3(2) kcal/mol and ∆S° ) -18(1) eu.26 The
∆H° value shows that the Zr-ATMS bond of 7a is ca. 5 kcal/
mol stronger than the sum of the Zr-ClC6D5 and o-CF‚‚‚Zr
bonds of 4a.
Intermolecular Alkene Exchange. The NMR resonances of
an equilibrium mixture of 7a, 4a, and free ATMS in C6D5Cl
solution broaden as the temperature is raised above -38 °C.
1
The H Cp resonance of 7a coalesces with that of 4a at ca.
-12 °C, while the 13C SiMe3 resonance of 7a coalesces with
that of free ATMS at ca. -18 °C. The other resonances of 7a
are extremely broad at 2 °C but are not coalesced with those of
4a or free ATMS at this temperature. Study of this system in
the fast-exchange regime was difficult due to the lopsided
equilibrium above 0 °C. These dynamic effects are consistent
with the exchange of 7a with 4a and of coordinated ATMS
with free ATMS.
Several observations show that the bound ATMS in 7a is
not directly displaced by free ATMS (Scheme 6; direct alkene
exchange can result in retention or inversion of configuration
at Cint). First, the Htrans and p-F NMR signals of 7a show equal
excess line broadening between -38 and +2 °C,27 whereas
direct associative alkene exchange would cause greater broaden-
ing of the former signal than the latter, as Htrans of 7a exchanges
environments in this process while p-F of 7a does not. Second,
the line widths of the Htrans and p-F signals of 7a are independent
of the concentration of free ATMS over the concentration range
(26) If the solvent term is included, the equilibrium constant for eq 1 is Keq’ )
Keq[C6D5Cl], where Keq is defined as in the text. If the solvent concentration
is assumed to be independent of temperature, the value of ∆H° is not
affected, but the entropy term becomes ∆S’° ) ∆S° + R ln[C6D5Cl], where
R ln[C6D5Cl] ≈ 4.5 eu.
(27) The difference in excess line widths of the Htrans and p-F resonances of 7a
showed no trend with temperature and averaged 1.2 Hz between -38 and
+2 °C.
9
J. AM. CHEM. SOC. VOL. 128, NO. 26, 2006 8643

A R T I C L E S
Scheme 6
Stoebenau and Jordan
Figure 7. Partial 1H NMR spectrum (500 MHz) of an equilibrium mixture
of 7a (0.034 M), 4a (0.024 M), and free ATMS (0.17 M) at -38 °C in
C6D5Cl, showing the greater line widths for the Cp signals of 7a at δ 6.06
(ω ) 14 Hz) and 6.00 (ω ) 12 Hz) compared to that of 4a at δ 5.97 (ω )
6 Hz). The minor resonance labeled “x” is due to an unknown impurity.
be due to greater lateral steric crowding in the metallocene
wedge of 7b compared to that in 9, which restricts entry of
solvent.
[ATMS] ) 0.083-0.18 M. These results imply that ATMS is
lost from 7a to give 4a, and free ATMS displaces C6D5Cl from
4a to give 7a (eq 1).
Dynamics of (C5H4R)2Zr(C6F5)(ATMS)+ Complexes. Se-
lective NMR line broadening effects suggested that 7a may
undergo an intramolecular dynamic process in addition to the
intermolecular exchange process in eq 1. First, for the dynamic
equilibrium in eq 1, it is expected that the exchange broadening
of the NMR resonances in the slow-exchange region of 4a will
be greater than that for 7a when [4a] < [7a].29 However, the
1H NMR line widths of the Cp signals of 7a are greater than
the Cp signal of 4a under these conditions (Figure 7). Second,
the Hallylic and Cp resonances of 7a exhibit greater exchange
broadening than the other resonances of 7a in the slow-exchange
region. Third, the excess broadening of the Hallylic and Cp signals
of 7a is independent of [ATMS]. These data imply that 7a
undergoes an intramolecular process that permutes the two
diastereotopic Cp rings, and the two diastereotopic allylic
hydrogens. The simplest such process is nondissociative alkene
enantioface exchange (“alkene flipping”),30 i.e., exchange of the
Cp2Zr(C6F5)+ unit between the two ATMS enantiofaces without
decomplexation of the ATMS ligand, as shown in eq 2.
First-order rate constants for ATMS decomplexation from
7a (k-1 in eq 1) were determined from the excess line
broadening of the Htrans and p-F signals of 7a between -38
and +2 °C. The line broadening data give k-1 ) 5.5(2.5) s-1
1
at -38 °C in C6D5Cl. A H EXSY NMR spectrum under the
same conditions contains cross-peaks between the free and
coordinated ATMS resonances and gives k-1 ) 5.1(1) s-1 28
,
in close agreement with the result from the line broadening
data.
An Eyring analysis for ATMS decomplexation from 7a
q
q
(Figure 6) gives ∆H-1 ) 8.9(6) kcal/mol and ∆S-1 ) -17-
(3) eu. The negative entropy of activation is consistent with
displacement of ATMS from 7a by C6D5Cl in an associative
process. An incipient o-CF‚‚‚Zr interaction in the transition state
may also contribute to the negative entropy of activation.
The ATMS decomplexation rate constant for 7b, k-1 ) 12-
(5) s-1 at -38 °C in C6D5Cl, was determined from the excess
line broadening of the Htrans and p-F NMR signals of 7b. The
excess line widths of these two signals are very similar,
discounting the occurrence of direct alkene exchange between
7b and free ATMS. The slightly faster ATMS decomplexation
rate for 7b vs 7a is consistent with the slightly weaker ATMS
binding in 7b vs 7a (Table 3). Remarkably, ATMS decomplex-
ation from 7b is ca. 10 times slower than for the OtBu analogue
9 (k-1 ) 125(25) s-1 at -38 °C in C6D5Cl), even though the
Keq values are similar for these systems. This difference may
The rate constant for alkene flipping in 7a, determined from
the greater excess line broadening of the downfield Hallylic signal
compared to that of the Htrans signal, is kflip ) 18(1) s-1 (C6D5-
Cl, -38 °C). The EXSY spectrum of an equilibrium mixture
of 7a, 4a, and ATMS (Figure 8) contains cross-peaks between
the two Hallylic resonances of 7a,31 which are larger than those
between the Hallylic resonances of 7a and 4a, confirming that
alkene flipping occurs in 7a. Analysis of the EXSY spectrum
(28) (a) Perrin, C. L.; Dwyer, T. J. Chem. ReV. 1990, 90, 935. (b) Johnston, E.
R.; Dellwo, M. J.; Hendrix, J. J. Magn. Reson. 1986, 66, 399. (c) Perrin,
C. L.; Gipe, R. K. J. Am. Chem. Soc. 1984, 106, 4036.
(29) At equilibrium for eq 1, k1[4a][ATMS] ) k-1[7a] (equal rates). Under slow
NMR exchange conditions, the excess line width due to exchange is ∆ω
) ω - ω0, where ω is the line width at half-height of a given resonance
and ω0 is its line width in the absence of exchange. For 4a, ∆ω4a ) k1-
[ATMS]/π and for 7a, ∆ω7a ) k-1/π. Therefore, if eq 1 is the only exchange
process, at equilibrium, by simple substitution ∆ω4a[4a] ) ∆ω7a[7a]. Thus,
if [4a] < [7a], then ∆ω4a > ∆ω7a, and if the line widths in the absence of
exchange are approximately equal, then ω4a > ω7a
.
(30) The nomenclature of this process is from: Prosenc, M.-H.; Brintzinger,
H.-H. Organometallics 1997, 16, 3889.
Figure 6. Eyring plot for ATMS decomplexation (k-1) from 7a. The plot
of circles (solid line) is from Htrans NMR line broadening data. The plot of
triangles (dashed line) is from p-F NMR line broadening data.
(31) Under the mixing times used, dipole-dipole interactions do not contribute
significantly to the cross-peak volumes, as is evident from the absence of
cross-peaks between Hcis and Htrans of 7a.
9
8644 J. AM. CHEM. SOC. VOL. 128, NO. 26, 2006

Alkene/Alkyne Coordination to Zr
−(σ
-Carbyl) Complexes
A R T I C L E S
Scheme 7. Possible CdC Double Bond Rotation Mechanism for
Alkene Face Exchange
Figure 8. 1H 2D EXSY NMR spectrum of an equilibrium mixture of 7a,
4a, and free ATMS. The Hallylic region is shown. The Hallylic signals of 7a
occur at δ 2.41 (br) and 2.07 (br), while that for free ATMS occurs at δ
1.44. The resonance at δ 2.01 is the Ph3CMe resonance.
line width measurements are precluded by the proximity of more
intense signals. Alkene flipping in 7b is ca. 14 times faster than
in 7a.
Mechanism of Alkene Flipping. Several mechanisms for
alkene flipping in 7a,b can be proposed, based on the framework
of Gladysz.33 Considered here are mechanisms involving double
bond rotation, allyl-type intermediates, or C-H σ-complex
intermediates.34 Mechanisms which involve loss of alkene (eq
1 and Scheme 6) can be rejected since alkene flipping is
significantly faster than alkene decomplexation.
Rotation around the ATMS CdC double bond would lead
to enantioface exchange in 7a,b (Scheme 7), and may be
possible due to the polarized nature of this bond. This process
has been established for CpFe(CO)2(η2-H2CdCHNR2)+ alkene
complexes.35 This process exchanges the two Cp groups and
the two Hallylic hydrogens, and, significantly, also exchanges
Table 4. First-Order Rate Constants and Free Energy Barriers for
ATMS Decomplexation (k-1, ∆G-1q) and Alkene Flipping (kflip
∆Gflipq) for ATMS Complexesa
,
1
1
cmpd
k-1 (s-
)
∆
G- q(kcal/mol)
k
flip (s-
)
∆
Gflipq (kcal/mol)
1
7a (1b)b
7a (EXSY)c
7b (lb)b
5.5(2.5)
5.1(1)
12(5)
12.8(2)
12.9(1)
12.5(2)
11.4(1)
18(1)
12.3(1)
12.2(1)
11.0(1)
-
23.0(3)
290(80)
-
9 (lb)b
125(25)
a In C6D5Cl at -38 °C. b Determined from NMR line broadening data.
c Determined from EXSY NMR spectroscopy.
Htrans with Hcis. However, as noted above, the excess broadening
of the Htrans and p-F signals are the same over a wide temperature
range.27 In addition, 1H EXSY NMR spectra of 7a do not exhibit
cross-peaks between the Hcis and Htrans resonances. Therefore,
Htrans/Hcis exchange does not occur, and double bond rotation
is ruled out for 7a,b.
Deprotonation of an allylic hydrogen or heterolytic Callylic
-
Si bond cleavage of 7a,b would lead to neutral η1-allyl
complexes (Scheme 8). Rotation around the “Cterm-Cint” bond
(i.e. the ZrCH2-CHCHSiMe3 or ZrCH2-CHCH2 bonds) of the
allyl intermediates followed by reformation of the Callylic-H or
Callylic-Si bonds would lead to net enantioface exchange in 7a,b.
However, these mechanisms also result in net CdC bond
rotation and exchange of Htrans with Hcis, and thus can be
rejected. In addition, a base would be required to deprotonate
Figure 9. Free energy diagram comparing alkene flipping and alkene
decomplexation from 7a in C6D5Cl solution at -38 °C.
gives kflip ) 23.0(3) s-1 (C6D5Cl, -38 °C) for 7a, in good
agreement with the value found by NMR line broadening. Thus,
alkene flipping is ca. 4 times faster than alkene decomplexation
in 7a (Table 4). The energetics of alkene flipping and decom-
plexation of 7a are compared in the free energy diagram in
Figure 9.32
Hallylic. Bases are not present under the reaction conditions and
in any case would likely displace ATMS from 7a,b. Also, a
+
transient SiMe3 cation would likely react with other species
in solution.
The NMR results for 7a,b are, however, fully consistent with
alkene enantioface exchange occurring via C-H σ-complex
Compound 7b also undergoes alkene flipping, as is evident
from the greater excess line broadening of the Hallylic NMR
signals compared to the other ATMS resonances for this species.
The rate constant for alkene flipping in 7b is estimated to be
kflip ) 290(80) s-1 (C6D5Cl, -38 °C) from the excess line
broadening of the Hallylic signal of 7b at δ 2.4. More precise
(33) Peng, T.-S.; Gladysz, J. A. J. Am. Chem. Soc. 1992, 114, 4174.
(34) (a) A reversible C6F5 insertion mechanism was also considered. However,
while reversible C6F5 insertion would exchange the Cp rings (due to site
epimerization in the insertion products), it does not exchange the allylic
hydrogens. In addition, the insertion products would be more reactive
towards insertion than 4a itself, and ATMS polymerization would be
observed, which is not the case. (b) Mechanisms considered by Gladysz33
and involving, for example, oxidative addition or M-carbene formation,
are not reasonably possible for 7a,b.
(32) An Eyring analysis for alkene enantioface exchange was precluded by a
narrow viable-temperature range (15 °C) and the small chemical shift
difference of the two 1H NMR Cp resonances of 7a.
(35) (a) Matchett, S. A.; Zhang, G.; Frattarelli, D. Organometallics 2004, 23,
5440. (b) Chang, T. C. T.; Foxman, B. M.; Rosenblum, M.; Stockman, C.
J. Am. Chem. Soc. 1981, 103, 7361.
9
J. AM. CHEM. SOC. VOL. 128, NO. 26, 2006 8645

A R T I C L E S
Stoebenau and Jordan
Scheme 8. Possible Allyl-Intermediate Mechanisms for Alkene
Face Exchange
1,3-enchainment in cyclopentene polymerization has been
explained by this process.37d
Brintzinger proposed that alkene flipping in Cp2Zr(H)-
(H2CdCMe2)+ occurs via a κ2-H2CdCMe2 C-H σ-complex,
based on DFT studies.30 The barrier for this process was
calculated to be 8.1 kcal/mol, which is only slightly lower than
the barriers found for alkene flipping in 7a,b. The calculated
barrier for alkene flipping of coordinated cis-2-butene via a κ2-
MeHC)CHMe σ-complex is much higher (19 kcal/mol).30
Neither the SiMe3 group nor the C6F5 ligand is necessary
per se for alkene flipping to occur. These groups simply
strengthen the metal-alkene interaction and increase the alkene
decomplexation barrier sufficiently to allow alkene flipping to
be detectable and separable from the intermolecular exchange
process in eq 1. Alkene flipping was not detected for 9 or other
Cp’2Zr(OtBu)(alkene)+ complexes at -89 °C in CD2Cl2, or for
9 at -38 °C in C6D5Cl. Alkene flipping is certainly possible in
these cases, but would be masked by fast alkene decomplexation.
Intermolecular Alkyne Exchange. PTMS decomplexation
from 8a is much slower than ATMS decomplexation from 7a,
consistent with the greater binding strength of PTMS vs that of
ATMS (Table 3). No significant NMR line broadening of the
PTMS signals of 8a is observed at -38 °C in C6D5Cl, implying
1
that k-1 e 3 s-1. However, the H EXSY spectrum of an 8a/
4a/PTMS equilibrium mixture (C6D5Cl, -38 °C, τm ) 1.6 s)
contains exchange cross-peaks between the Cp resonances of
8a and 4a and between the free and coordinated alkyne CH2
resonances. This result shows that 8a exchanges with 4a and
free PTMS on the T1 time scale, consistent with eq 1.
Scheme 9. Alkene σ-Complex Intermediate or Transition State
Mechanism for Alkene Face Exchange
Reactivity of Cp2Zr(C6F5)(ATMS)+ and Cp2Zr(C6F5)-
(PTMS)+. When an equilibrium mixture of 4a, 7a, and free
ATMS in C6D5Cl is warmed from -38 to +2 °C over 4 h, 7a
and free ATMS are consumed, and 6,6-dimethyl-4-((trimeth-
ylsilyl)methyl)-6-silahept-1-ene (12), a dimer of ATMS, is
formed. Compound 12 results from a Lewis acid-mediated
dimerization38 of ATMS due to 4a, or possibly trace Ph3C+.39
NMR and GC/MS analysis of the organic products from a 4a/
7a/ATMS mixture maintained at 22 °C for 3 days shows the
presence of dimers and trimers of ATMS. While the structures
of these products were not determined, none contain C6F5
groups. The trimers likely form by a Lewis acid-mediated
allylsilylation of 12.38 In addition, ca. 65% of 4a remains after
this time, but new Cp2Zr species were not observed, and the
fate of the consumed Zr is unknown. There is no evidence for
ATMS insertion in 7a by NMR or GC/MS. Bochmann reported
that solutions of [Cp2Zr(C6F5)][(µ-Me)B(C6F5)3] polymerize
ethylene but noted that this behavior may be due to impurities.11
intermediates (Scheme 9). Steric crowding between the ATMS
and Cp ligands during metal-(alkene centroid) rotation can be
relieved by isomerization of the π-complex to a κ2-H2CdCHR
σ-complex intermediate or transition state, which can then
relax to the π-complex by slippage of the Cp2Zr(C6F5)+ unit
back to either enantioface.36 The weaker π coordination for 7b
compared to that for 7a explains the faster alkene flipping in
7b.
The σ complex mechanism in Scheme 9 is similar to that
proposed by Gladysz for ReI.33 Alkene flipping has also been
proposed to occur by this mechanism in ZrIV-catalyzed olefin
polymerization.30,37 For example, polymerization of isotopically
labeled propylenes gives specifically labeled chain ends that
have been explained in part by alkene flipping,37a-c and trans-
When an equilibrium mixture of 4a, 8a, and free PTMS in
C6D5Cl is maintained at 22 °C for 14 h, 52% of 4a is consumed,
8a and free PTMS are fully consumed, and C6F5H is formed in
31% yield, as determined by NMR spectroscopy. These results
suggest that protonolysis of the Zr-C6F5 bond of 8a by PTMS
(36) Enantioface exchange in 7a,b by dissociation of into 4a,b and ATMS
followed by recombination within a solvent cage, with slower diffusion of
ATMS out of the cage, cannot be definitively excluded.
(37) (a) Sillars, D. R.; Landis, C. R. J. Am. Chem. Soc. 2003, 125, 9894. (b)
Yoder, J. C.; Bercaw, J. E. J. Am. Chem. Soc. 2002, 124, 2548. (c) Leclerc,
M. K.; Brintzinger, H. H. J. Am. Chem. Soc. 1996, 118, 9024. (d) Kelly,
W. M.; Wang, S.; Collins, S. Macromolecules 1997, 30, 3151.
(38) Yeon, S. H.; Lee, B. W.; Yoo, B. R.; Suk, M.-Y.; Jung, I. N. Organome-
tallics 1995, 14, 2361.
(39) Schade, C.; Mayr, H. Makromol. Chem., Rapid Commun. 1988, 9, 477.
9
8646 J. AM. CHEM. SOC. VOL. 128, NO. 26, 2006

Alkene/Alkyne Coordination to Zr
−(σ
-Carbyl) Complexes
A R T I C L E S
occurs to yield C6F5H and the Cp2Zr(CtCCH2SiMe3)+ cation,
which can then catalytically oligomerize PTMS.40 There is no
evidence for PTMS insertion into the Zr-C6F5 bond of 8a.
Reactivity of ATMS with Cp2ZrMe+. The successful
observation of 7a,b suggested that coordination of ATMS to
other, more reactive d0 Zr-carbyl cations may also be detectable
by NMR spectroscopy. Therefore, the reaction of ATMS with
Cp2ZrMe+ was probed. Addition of ATMS to a CD2Cl2 solution
of [Cp2ZrMe][B(C6F5)4] at -78 °C results in an immediate color
change from yellow-orange to orange. NMR spectra of this
mixture at -89 °C show that Cp2ZrMe+ and ATMS are
completely consumed within the combined time periods of 1
min at -78 °C and 10 min at -89 °C. No resonances indicative
of a d0 metal-alkene complex and no NMR line broadening
effects that can be ascribed to fast exchange between free and
coordinated ATMS are observed. Instead, broad resonances for
poly- or oligo(ATMS) are present, implying that ATMS
insertion is fast even at low temperature.41
Experimental Section
General Procedures. All reactions were performed using glovebox
or Schlenk techniques under a purified N2 atmosphere, or on a high
vacuum line. N2 was purified by passage through columns of activated
molecular sieves and Q-5 oxygen scavenger. CD2Cl2 and C6D5Cl were
distilled from P2O5. C6D6 was distilled from Na/benzophenone. Toluene
was dried by passage through columns of activated alumina and BASF
R3-11 oxygen-removal catalyst. Cp2Zr(C6F5)2 (2a),10 Cp2Zr(C6F5)Me
(3a),11 [HNMePh2][B(C6F5)4],43 Cp2ZrMe2,44 and Cp’2ZrMe2 were
45
synthesized by literature methods. Al(C6F5)3‚1.2PhMe was synthesized
by a literature method,11 and the toluene content was determined by
1H NMR spectroscopy with C6Me6 as an internal standard. CAU-
TION: Al(C6F5)3‚PhMe is thermally and shock sensitive! Other
chemical were obtained from standard suppliers. [Ph3C][B(C6F5)4] and
B(C6F5)3 were used as received. Allyltrimethylsilane and tert-butyl vinyl
ether were dried over CaH2. Bromobenzene, 2-butyne, and propargyl-
trimethylsilane were dried over 3 Å molecular sieves. Propyne was
used as received. Elemental analyses were performed by Midwest
Microlab (Indianapolis, IN).
NMR spectra were recorded on Bruker DRX 500 or 400 spectrom-
eters in Teflon-valved NMR tubes at ambient probe temperature unless
Conclusions
1
otherwise noted. H and 13C chemical shifts are reported relative to
The combined use of strongly coordinating â-Si-substituted
alkenes and alkynes, and the poorly nucleophilic C6F5 ligand
enables the generation of stable ZrIV-aryl-alkene and -alkyne
complexes. Both tactics are necessary: non-â-Si substituted
substrates such as 2-butyne do not coordinate to [Cp2Zr(C6F5)]-
[B(C6F5)4] (4a), and Cp2ZrMe+ rapidly oligomerizes or poly-
merizes ATMS even at -78 °C. ATMS and C6D5Cl compete
in an equilibrium for coordination to Cp2Zr(C6F5)+. [Cp2Zr-
(C6F5)(H2CdCHCH2SiMe3)][B(C6F5)4] (7a) is more stable than
[Cp’2Zr(C6F5)(H2CdCHCH2SiMe3)][B(C6F5)4] (7b), mainly due
to Cp electronic effects.17 Compounds 7a,b exhibit divergent
13C NMR chemical shifts for the alkene carbons (Cint shifts
downfield; Cterm shifts upfield) and a low-field 1H NMR
chemical shift for Hint. These data imply that the ATMS ligands
in 7a,b are significantly polarized and are bound in an
unsymmetrical manner (d(Zr-Cint) > d(Zr-Cterm)) as observed
for (C5R5)2Zr(alkoxide-alkene)+ species.3 The similarity of the
NMR data of 7a,b to data for 9 and other Zr-alkoxide-alkene
species shows that Zr-OR π-bonding does not contribute to
the unsymmetrical alkene coordination observed in these
systems.
Compounds 7a,b undergo nondissociative alkene enantioface
exchange, i.e., “alkene flipping,” probably via σ-κ2-H2CdCHCH2-
SiMe3 intermediates or transition states. Similar processes have
been proposed to occur during Zr-catalyzed alkene polymeri-
zation,30,37 and may play a role in stereoselective R-olefin
polymerization by chiral zirconocenes. The free energy barrier
to alkene flipping in 7a,b is similar to the barriers for initiation
and propagation in 1-hexene polymerization by [rac-(EBI)ZrMe]-
[MeB(C6F5)3] (EBI ) 1,2-(1-indenyl)2ethane),42 suggesting that
its occurrence in polymerization is kinetically feasible.
SiMe4 and were referenced to the residual solvent signals. 19F NMR
spectra are reported relative to external CFCl3 and were referenced either
to external neat CFCl3 or the o-F signal of internal C6F5H (δ -138.0,
dd, J ) 22, 9). 11B NMR spectra are referenced to external BF3‚OEt2.
NMR probe temperatures were calibrated by a MeOH thermometer.46
1
Coupling constants are reported in hertz. Where 13C{gated H} NMR
spectra are reported, standard 13C{1H} NMR spectra were also recorded
to assist in interpretation and assignment. For H2CdCHX substrates,
Hcis is the H that is cis to Hint, and Htrans is the H that is trans to Hint.
-
NMR spectra of ionic compounds contain B(C6F5)4 resonances at
the free anion positions. 19F NMR spectra were obtained for all
compounds that contain this anion. NMR spectra of cationic compounds
generated in situ by methide abstraction using Ph3C+ contain resonances
-
for Ph3CMe. The NMR data for B(C6F5)4 and Ph3CMe are given in
the Supporting Information.
The 13C NMR resonances of the C6F5 groups in the compounds
described below were not detected due to low receptivity, overlap with
-
B(C6F5)4 resonances, or exchange line broadening.
Cp’2Zr(C6F5)Me (3b). A colorless solution of Al(C6F5)3‚1.2PhMe
(0.1894 g, 0.297 mmol, 0.335 equiv) in toluene (20 mL) was added to
a colorless solution of Cp’2ZrMe2 (0.2476 g, 0.886 mmol) in toluene
(20 mL) by cannula over 5 min at 22 °C. The mixture turned yellow
during the addition. The mixture was stirred for 15 min at 22 °C. The
color changed to apricot after 10 min. The volatiles were removed under
vacuum, and the residue was dried under vacuum for 2 d, yielding a
pink powder. This material was sublimed (120 °C, 0.1 mTorr), yielding
pure 3b as a pale-yellow powder (0.192 g, 50%). 1H NMR (C6D6): δ
5.74 (q, J ) 2.4, 2H, Cp’ CH), 5.55 (q, J ) 2.5, 2H, Cp’ CH), 5.52 (q,
J ) 2.5, 2H, Cp’ CH), 5.26 (q, J ) 2.4, 2H, Cp’ CH), 1.75 (s, 6H,
Cp’Me), 0.29 (t, JCF ) 4.0, 3H, ZrMe). 19F{1H} NMR (C6D6): δ -113.9
(br d, J ) 24, 2F, o-F), -155.7 (t, J ) 20, 1F, p-F), -161.4 (br m, 2F,
m-F). 13C{1H} NMR (C6D6): δ 125.0 (ipso Cp’), 114.6 (Cp’ CH), 113.7
(Cp’ CH), 108.6 (Cp’ CH), 106.9 (Cp’ CH), 43.7 (t, JCF ) 7, ZrMe),
14.8 (Cp’Me). Anal. Calcd for C19H17F5Zr: C, 52.88; H, 3.97. Found:
C, 52.87; H, 4.04.
(40) (a) Horton, A. D. Chem. Commun. 1992, 185. (b) Akita, M.; Yasuda, H.;
Nakamura, A. Bull. Chem. Soc. Jpn. 1984, 57, 480. (c) den Haan, K. H.;
Wielstra, Y.; Teuben, J. H. Organometallics 1987, 6, 2053. (d) Heeres, H.
J.; Teuben, J. H. Organometallics 1991, 10, 1980.
Generation of [Cp2Zr(C6F5)][B(C6F5)4] (4a). To an NMR tube
charged with Cp2Zr(C6F5)Me (3a, 16.4 mg, 0.0406 mmol) and [Ph3C]-
[B(C6F5)4] (37.7 mg, 0.0409 mmol) was added by vacuum transfer C6D5-
(41) Zirconocene catalysts polymerize ATMS by the same coordination/insertion
mechanism as for other olefins, and exhibit the same metallocene-symmetry/
polymer-tacticity relationships. (a) Resconi, L.; Piemontesi, F.; Franciscono,
G.; Abis, L.; Fiorani, T. J. Am. Chem. Soc. 1992, 114, 1025. (b) Habaue,
S.; Baraki, H.; Okamoto, Y. Macromol. Chem. Phys. 1998, 199, 2211. (c)
Natta, G.; Mazzanti, G.; Longi, P.; Bernardini, F. J. Polym. Sci. 1958, 31,
181.
(43) Tjaden, E. B.; Swenson, D. J.; Jordan, R. F.; Petersen, J. L. Organometallics
1995, 14, 371.
(44) Samuel, E.; Rausch, M. D. J. Am. Chem. Soc. 1973, 95, 6263.
(45) Couturier, S.; Tainturier, G.; Gautheron, B. J. Organomet. Chem. 1980,
195, 291.
(42) Liu, Z.; Somsook, E.; White, C. B.; Rosaaen, K. A.; Landis, C. R. J. Am.
Chem. Soc. 2001, 123, 11193.
(46) Van Geet, A. L. Anal. Chem. 1970, 42, 679.
9
J. AM. CHEM. SOC. VOL. 128, NO. 26, 2006 8647

A R T I C L E S
Stoebenau and Jordan
1
Cl at -196 °C. The tube was warmed to 22 °C and Vigorously shaken
to give an orange-yellow solution. The tube was allowed to sit at 22
°C for 2 d, during which time the solution turned yellow. NMR spectra
showed that 4a (97-99%), Ph3CMe, and trace amounts (1-3%) of
C6F5H and [{Cp2Zr(C6F5)}2(µ-Cl)][B(C6F5)4] (6a, see below) were
6a (72%) and unreacted 4a (23%). Data for 6a: H NMR (C6D5Cl):
1
δ 6.20 (br s, 20H, Cp). H NMR (C6D5Cl, -38 °C): δ 6.18 (s, 20H,
Cp). 19F NMR (C6D5Cl): δ -118.5 (br s, 4F, o-F), -151.6 (t, J ) 19,
2F, p-F), -159.2 (br m, 4F, m-F). 19F NMR (C6D5Cl, -38 °C): δ
-114.6 (br s, 2F, o-F), -121.4 (br s, 2F, o-F), -151.7 (t, J ) 20, 2F,
p-F), -159.7 (v br s, 4F, m-F). 13C{1H} NMR (C6D5Cl): δ 116.6 (Cp).
13C{1H} NMR (C6D5Cl, -38 °C): δ 116.7 (Cp).
present.47 Data for 4a: H NMR (C6D5Cl): δ 6.03 (s, 10H, Cp). H
NMR (C6D5Cl, -38 °C): δ 5.97 (s, 10H, Cp). 19F NMR (C6D5Cl): δ
-129.6 (br s, 2F, o-F), -150.1 (t, J ) 20, 1F, p-F), -157.4 (br m, 2H,
m-F). 19F NMR (C6D5Cl, -38 °C): δ -118.2 (v br s, 1F, o-F), -140.0
(v br s, 1F, o-F), -150.4 (t, J ) 20, 1F, p-F), -157.7 (br s, 2F, m-F).
13C{1H} NMR (C6D5Cl): δ 117.0 (Cp). 13C{1H} NMR (C6D5Cl, -38
°C): δ 117.0 (Cp).
Generation of [Cp’2Zr(C6F5)][B(C6F5)4] (4b). To an NMR tube
charged with 3b (10.1 mg, 0.0234 mmol) and [Ph3C][B(C6F5)4] (21.2
mg, 0.0230 mmol), was added by vacuum transfer C6D5Cl at -196
°C. The tube was warmed to 22 °C and shaken to give an orange-
yellow solution. The tube was allowed to sit at 22 °C for 30 min, during
which time the solution turned yellow. NMR spectra showed that 4b
(93%), Ph3CMe, and small amounts (7-8% each) of C6F5H and an
unknown impurity were present.48 Data for 4b: 1H NMR (C6D5Cl): δ
5.90 (br t, J ) 2, 4H, Cp’ CH), 5.80 (br t, J ) 2, 4H, Cp’ CH), 1.65
(s, 6H, Cp’Me). 1H NMR (C6D5Cl, -38 °C): 5.84 (br s, 4H, Cp’ CH),
5.73 (br s, 4H, Cp’ CH), 1.59 (s, 6H, Cp’Me). 19F NMR (C6D5Cl): δ
-129.3 (v br s, 2F, o-F), -149.7 (t, J ) 19, 1F, p-F), -157.1 (br s,
2F, m-F). 19F NMR (C6D5Cl, -38 C): δ -116.7 (br s, 1F, o-F), -140.7
(br s, 1F, o-F-µ-Zr), -149.9 (t, J ) 20, 1F, p-F), -155.9 (br s, 1F,
m-F), -158.5 (br s, 1F, m-F). 13C{1H} NMR (C6D5Cl): δ 133.6 (ipso
Cp’), 118.8 (Cp’ CH), 114.2 (Cp’ CH), 14.3 (Cp’Me). 13C{1H} NMR
(C6D5Cl, -38 °C): δ 133.6 (ipso Cp’), 119 (v br, Cp’ CH), 118 (v br,
Cp’ CH), 114.0 (br, Cp’ CH), 14.5 (Cp’Me).
1
1
Generation of [Cp2Zr(C6F5)(H2CdCHCH2SiMe3)][B(C6F5)4] (7a).
A solution of [Cp2Zr(C6F5)][B(C6F5)4] (4a, 0.0268 mmol) in C6D5Cl
(0.58 mL) in an NMR tube was cooled to -196 °C, and ATMS (0.13
mmol) was added by vacuum transfer. The tube was warmed to -40
°C and shaken, resulting in an orange-yellow solution. The tube was
placed in an NMR probe that had been precooled to - 38 °C. NMR
spectra showed the presence of 7a (0.029 M), 4a (0.018 M), and free
1
ATMS (0.20 M). Data for 7a: H NMR (C6D5Cl, -38 °C): δ 7.63
(m, 1H, Cint), 6.06 (br s, 5H, Cp), 6.00 (br s, 5H, Cp), 4.06 (d, J )
16.5, 1H, Htrans), 3.84 (br t, J ) 8, 1H, Hcis), 2.41 (br m, 1H, Hallylic),
2.07 (br m, 1H, Hallylic), 0.02 (s, 9H, SiMe3). 19F NMR (C6D5Cl, -38
°C): δ -151.2 (t, J ) 19, 1F, p-F), -158.6 (br s, 2F, m-F). The o-F
resonances were not detected due to line broadening. 13C{gated H}
1
NMR (C6D5Cl, -38 °C): δ 186.5 (dm, 1JCH ) 161, Cint), 116.4 (br d,
1JCH ) 184, Cp), 116.2 (br d, JCH ) 184, Cp), 97.7 (tm, JCH ) 150,
1
1
1
1
C
term), 37.0 (t, JCH ) 129, Callylic), -2.0 (q, JCH ) 120, SiMe3). The
int and Cterm resonances show unresolved coupling to the C6F5 ligand.
C
Generation of [Cp2Zr(C6F5)(THF)][B(C6F5)4] (10a). A solution
of [Cp2Zr(C6F5)][B(C6F5)4] (4a, 0.0235 mmol) in C6D5Cl in an NMR
tube was cooled to -196 °C, and THF (0.216 mmol) was added by
vacuum transfer. The tube was warmed to 22 °C and shaken, resulting
in an orange-yellow solution. The volatiles were removed under vacuum
at 22 °C, and C6D5Cl was added at -196 °C. The tube was warmed to
22 °C and shaken, giving an orange-yellow solution. NMR spectra
revealed the presence of 10a (93%) and Ph3CMe. 1H NMR (C6D5Cl):
δ 6.08 (s, 10H, Cp), 3.67 (br m, 4H, THF), 1.57 (br m, 4H, THF). 19
NMR (C6D5Cl): δ -119.7 (v br s, 2F, o-F), -151.1 (t, J ) 20, 1F,
p-F), -158.5 (m, 2F, m-F). 13C{1H} NMR (C6D5Cl): δ 116.4 (Cp),
81.8 (THF), 25.7 (THF).
Generation of [{Cp’2Zr(C6F5)}2(µ-Me)][B(C6F5)4] (5b). To an
NMR tube charged with 3b (17.0 mg, 0.0394 mmol) and [Ph3C]-
[B(C6F5)4] (17.9 mg, 0.0194 mmol) was added by vacuum transfer C6D5-
Cl (0.60 mL) at -196 °C. The tube was warmed to 22 °C and shaken,
giving a pale-yellow solution. NMR spectra at ambient probe temper-
ature showed that 5b (0.025 M), 3b (0.014 M), 4b (0.0048 M), and
Ph3CMe were present. The signals for the three Zr species were broad
due to chemical exchange. Additional [Ph3C][B(C6F5)4] (7.0 mg) was
added. NMR spectra showed that the concentrations of 5b (0.021 M),
3b (0.0028 M), and 5b (0.018 M) had changed. The signals for the Zr
species again displayed significant NMR line broadening. At -38 °C,
F
Generation of [Cp’2Zr(C6F5)(H2CdCHCH2SiMe3)][B(C6F5)4] (7b).
A solution of 4b (0.0230 mmol) in C6D5Cl (0.58 mL) in an NMR tube
was cooled to -196 °C, and ATMS (0.122 mmol) was added by
vacuum transfer. The tube was warmed to -40 °C and shaken to give
a yellow solution. The tube was placed in an NMR probe that had
been precooled to -38 °C. NMR spectra showed that 7b (0.015 M),
4b (0.025 M), and free ATMS (0.27 M) were present. Data for 7b:
1H NMR (C6D5Cl, -38 °C): δ 7.64 (m, 1H, Hint), 4.13 (d, J ) 16.0,
1H, Htrans), 4.00 (br t, Japparent ) 6, 1H, Hcis), 2.4 (v br s, 1H, Hallylic),
2.1 (v br s, 1H, Hallylic, partially obscured), 1.67 (s, 6H, Cp’Me). The
SiMe3 and Cp’ CH resonances are obscured by resonances of free
ATMS. 19F NMR (C6D5Cl, -38 °C): δ -113.3 (br s, 1F, o-F), -125.6
(br s, 1F, o-F), -150.8 (t, J ) 20, 1F, p-F), -157.8 (br s, 1F, m-F),
-158.7 (br s, 1F, m-F). 13C{1H} NMR (C6D5Cl, -38 °C): δ 184.9
(Cint), 97.9 (Cterm), 36.7 (Callylic), 14.8 (Cp’Me), -0.2 (SiMe3). The Cp’
ipso and CH resonances are broadened into the baseline due to
exchange.
1
only 5b (0.024 M) and 4b (0.018 M) were present. Data for 5b: H
NMR (C6D5Cl): δ 6.00 (br s, 4H, Cp’ CH), 5.96 (br s, 4H, Cp’ CH),
5.90 (br s, 8H, Cp’ CH), 1.81 (br s, 12H, Cp’Me), -0.24 (br s, 3H,
1
µ-Me). H NMR (C6D5Cl, -38 °C): δ 5.99 (t, 8H, Cp’ CH), 5.91
(q, 4H, Cp’ CH), 5.84 (br m, 4H, Cp’ CH), 1.77 (s, 12H, Cp’Me),
-0.27 (s, 3H, µ-Me). 19F NMR (C6D5Cl): δ -116 (v br s, 4F, o-F),
-151.4 (t, J ) 18, 2F, p-F), -158.7 (br s, 4F, m-F). 19F NMR (C6D5-
Cl, -38 °C): δ -112.6 (m, 2F, o-F), -120.6 (m, 2F, o-F), -151.6
(t, J ) 19, 2F, p-F), -158.1 (br m, 2F, m-F), -159.4 (br m, 2F, m-F).
13C{1H} NMR (C6D5Cl): δ 130.5 (ipso Cp’), 118.2 (Cp’ CH), 116.2
(Cp’ CH), 112.7 (Cp’ CH), 112.2 (Cp’ CH), 14.8 (Cp’Me). 13C{1H}
NMR (C6D5Cl, -38 °C): δ 130.4 (ipso Cp’), 117.6 (Cp’ CH), 116.2
(Cp’ CH), 112.4 (Cp’ CH), 112.1 (Cp’ CH), 32.2 (br m, µ-Me), 15.0
(Cp’Me).
Independent Generation of [{Cp2Zr(C6F5)}2(µ-Cl)][B(C6F5)4]
(6a). A solution of [Cp2Zr(C6F5)][B(C6F5)4] (4a, 0.028 mmol) in C6D5-
Cl was prepared in an NMR tube, and [nBu3NCH2Ph]Cl (3.1 mg, 0.0099
mmol) was added. The tube was sealed and shaken at 22 °C, resulting
in an orange-yellow solution. NMR spectra revealed the presence of
Generation of [Cp2Zr(C6F5)(HCtCCH2SiMe3)][B(C6F5)4] (8a).
A solution of [Cp2Zr(C6F5)][B(C6F5)4] (4a, 0.0233 mmol) in C6D5Cl
(0.53 mL) in an NMR tube was cooled to -196 °C, and PTMS (0.027
mmol) was added by vacuum transfer. The tube was warmed to -40
°C and shaken, resulting in a maroon solution. The tube was then placed
in an NMR probe that had been precooled to -38 °C. NMR spectra
revealed the presence of 8a (0.040 M), 4a (0.0044 M), and free PTMS
(0.011 M). Data for 8a: 1H NMR (C6D5Cl, -38 °C): δ 6.04 (s, 10H,
Cp), 4.44 (br s, 1H, tCH), 2.14 (br s, 2H, CH2), 0.11 (s, 9H, SiMe3).
19F NMR (C6D5Cl, -38 °C): δ -114.6 (br s, 1F, o-F), -121.0 (br s,
1F, o-F), -151.5 (t, J ) 20, 1F, p-F), -159.0 (br s, 2F, m-F). 13C-
(47) Trace amounts (1-3%) of 6a were observed in some samples of 4a.
Compound 6a is formed at the beginning of the reaction and does not grow
in at longer reaction times. 6a is probably formed by the reaction of 4a
with alkyl chloride impurities in the solvent or trityl salt, or from Zr-Cl
impurities in 3a.
(48) The unknown impurity is probably {Cp’2Zr(C6F5)}2(µ-Cl)+, analogous to
6a.
1
2
{gated H} NMR (C6D5Cl, -38 °C): δ 145.2 (d, JCtCH ) 34, Cint),
9
8648 J. AM. CHEM. SOC. VOL. 128, NO. 26, 2006

Alkene/Alkyne Coordination to Zr
−(σ
-Carbyl) Complexes
A R T I C L E S
116.1 (d, 1JCH ) 178, Cp), 78.4 (dt, 1JCH ) 232, 4JCF ) 8, Cterm), 16.8
(t, JCH ) 135, CH2), -1.9 (q, JCH ) 121, SiMe3).
determine the thermodynamic parameters and their standard deviations.
The reported values are the averages of the results of these two
experiments.
1
1
Reaction of [Cp2ZrMe][B(C6F5)4] with Allyltrimethylsilane. To
an NMR tube charged with Cp2ZrMe2 (6.0 mg, 0.0239 mmol) and
[Ph3C][B(C6F5)4] (21.4 mg, 0.0232 mmol, 0.972 equiv) was added by
vacuum transfer C6D5Cl at -196 °C. The tube was warmed to 22 °C
(C) Exchange Rate Measurements by NMR Line Broadening.
1
Variable-temperature H and 19F NMR spectroscopies were used to
measure the kinetics of the dynamic processes of 7a. The first-order
rate constant for ATMS decomplexation from 7a (k-1) was determined
by k-1 ) π∆ω, where ∆ω is the excess line width due to exchange of
the Htrans or p-F signals of 7a and is defined by ∆ω ) ω - ω0, where
ω is the actual line width of the Htrans or p-F resonances of 7a and ω0
is the line width of the Ph3CMe or the p-F signal of 6a. All line widths
were found by simulation using gNMR.50 The coupling constants
between Htrans/Hint and p-F/m-F for 7a, and between p-F/m-F of 6a were
found by simulation at all temperatures where these coupling are
resolvable, but were extrapolated at other temperatures. The coupling
between Htrans/Hcis was set to -5.1 Hz, that between Htrans/Hallylic (δ
2.41) was set to -2.4 Hz, and that between Htrans/Hallylic (δ 2.07) was
set to -1.8 Hz. These values were found by simulation at -38 °C and
kept constant with temperature. Eyring plots were used to determine
activation parameters (∆H-1qand ∆S-1q) for ATMS decomplexation
1
and shaken to give an orange-yellow solution. After 4 h, a H NMR
spectrum was taken, which revealed the presence of [Cp2ZrMe]-
[B(C6F5)4] and Ph3CMe in a 1:1 ratio. The volatiles were removed under
vacuum at 22 °C to give a dark oil, and CD2Cl2 was added by vacuum
transfer at -78 °C. The tube was shaken and stored at this temperature,
giving a yellow-orange solution. ATMS (0.118 mmol) was added by
vacuum transfer at -196 °C. The tube was warmed to -78 °C and
shaken, giving an orange solution. The tube was maintained at -78
°C for 1 min and then NMR spectra were recorded at -89 °C. The
NMR spectra revealed that the Cp2ZrMe+ and ATMS were completely
consumed and that resonances for Cp2Zr(Me)(ATMS)+ were absent.
1
Instead, the H NMR spectrum contained broad signals for oligo- or
poly-ATMS at δ 1.7-0.5 and δ -0.07 ppm.41 Numerous unidentified
1
Cp resonances were observed by H and 13C NMR spectroscopy.
1
from 7a using both H and 19F NMR data.49 Reported values are the
Determination of Thermodynamic Parameters and Exchange
Rates. (A) General Variable-Temperature NMR Procedure. NMR
samples were prepared as described above. The NMR tube was stored
at -40 °C, and placed in a precooled NMR probe. The probe was
maintained at a given temperature for 20-30 min to allow the sample
to reach thermal equilibrium, and then NMR experiments were
conducted. The temperature was raised 5 or 10 °C and the procedure
repeated until the desired temperature range had been studied.
(B) Equilibrium Constant Measurements. The equilibrium con-
stant for the reaction:
1
weighted averages between two runs each of H and 19F NMR data
(total of four plots).
(D) EXSY Spectroscopy. Two-dimensional 1H EXSY NMR spectra
of equilibrium mixtures of 7a, 4a, and free ATMS were obtained at
-38 °C using the Bruker pulse program noesytp, with the pulse
sequence d1 - 90° - d0 - 90° - τm - 90° - acquire. The spectral
width consisted of the Hallylic region (δ 2.85-1.08 ppm). The relaxation
delay d1 was set at 6.0 s, which was 4 times the longest T1 of the signals
in this region (determined by an inversion-recovery experiment). The
mixing time τm was set to 50 ms, the acquisition time was set to 1.1 s,
and the 90° pulse width was set to 10.3 µs. The initial d0 value was 3
µs. The pulse sequence was repeated for 68 values of d0, and the F1
dimension was zero-filled to 2048 points. The F2 dimension had 2048
points. The number of scans per experiment was 16. The baseline was
automatically corrected in both dimensions, but the spectrum was not
symmetrized about the diagonal.
4a + H2C)CHCH2SiMe3 a 7a
was defined to be Keq ) [7a][4a]-1[ATMS]-1. The concentrations of
7a, 4a, and free ATMS at equilibrium were determined by 1H and 19
F
NMR. This equilibrium constant was determined to be Keq ) 8.2(1.4)
M-1 (C6D5Cl, -38 °C) from 8 measurements over the concentration
ranges of [4a]initial ) 0.023-0.064 M and [ATMS]initial ) 0.095-0.23
M. This Keq was measured over the temperature range -38 to +2 °C,
in 5 °C intervals.
Integrals of cross-peaks were symmetrical about the diagonal to
within 10%. Due to overlap of the diagonal signal of the coordinated
Hallylic signal at δ 2.07 with that for Ph3CMe (δ 2.01), the integral of
this Hallylic diagonal signal was assumed to be equal to the integral of
the diagonal signal of the coordinated Hallylic signal at δ 2.41. This
assumption does not affect the integrals of any cross-peaks.
The equilibrium constant for the reaction:
4b + H2C)CHCH2SiMe3 a 7b
The EXSY data were analyzed using the procedure described by
Perrin and Dwyer.28 Rate constants were found from the rate constant/
relaxation matrix R:
was defined to be Keq ) [7b][4b]-1[ATMS]-1 and found in a manner
similar to that for 7a. This equilibrium constant was determined to be
Keq ) 2.4(2) M-1 (C6D5Cl, -38 °C) from three measurements over
the concentration ranges of [4b]initial ) 0.029-0.040 M and [ATMS]initial
) 0.015-0.29 M.
Ra kb,a kf,a
k
Rb kf,b
-R )
a,b
The equilibrium constant for the reaction:
ka,f kb,f Rf
4a + HC≡CCH2SiMe3 a 8a
where km,n is the site-to-site rate constant for exchange from site m to
site n, and Rn relates to both kinetic and relaxation data and is not
considered here. Site a is the proton at the Hallylic signal of 7a at δ
2.41, site b is the proton at the Hallylic signal of 7a at δ 2.07, and site
f are the protons at the free Hallylic signal. Matrix R is related to EXSY
integration data by R ) -(1/τm) ln(A) (solved with MATLAB),51 where
A is the normalized spectral intensity matrix as follows:
was defined to be Keq ) [8a][4a]-1[PTMS]-1. The concentrations of
8a, 4a, and free PTMS at equilibrium were determined by H NMR.
The equilibrium constant was determined to be Keq ) 910(60) M-1
(C6D5Cl, -38 °C) by four measurements over concentrations ranges
of [4a]initial ) 0.044-0.054 M and [PTMS]initial ) 0.050-0.067 M.
Thermodynamic parameters (∆H° and ∆S°) for the reaction:
1
Iaa/pa Iab/pb Iaf/pf
4a + H2C)CHCH2SiMe3 a 7a
I /p Ibb/pb Ibf/pf
A )
ba
a
where Keq ) [7a][4a]-1[ATMS]-1 were determined from van’t Hoff
Ifa/pa Ifb/pb Iff/pf
plots for two variable-temperature NMR runs using Minitab49 to
(50) gNMR, v. 4.1.2; Adept Scientific: Letchworth, UK, 2000.
(51) MATLAB, v. 5.0.0.4073; The Math Works, Inc.: Natick, MA, 1996.
(49) Minitab, v. 11.2; Minitab Inc.: State College, PA, 1996.
9
J. AM. CHEM. SOC. VOL. 128, NO. 26, 2006 8649

A R T I C L E S
Stoebenau and Jordan
where Ixy is the integral of the peak between site x and site y is listed
schematically:
7a to give free alkene and 4a, kdissoc ) ka,f ) kb,f; and for alkene
complexation to 4a to give 7a, kassoc,obs ) kassoc[4a] ) 2‚kf,a ) 2‚kf,b.
An EXSY spectrum containing the Hallylic, Hcis, and Htrans regions did
not contain cross-peaks between coordinated Hcis and Htrans, nor between
either of these signals to the coordinated Hallylic signals.
af bf ff
ab bb fb
aa ba fa
Reported values for kdissoc and kflip are the averages from three EXSY
spectra with mixing times of 20, 25, and 50 ms.
and corresponding to the EXSY spectrum in Figure 8, pa is the
population fraction occupying the coordinated Hallylic resonance at δ
2.41, pb is the population fraction occupying the coordinated Hallylic
signal at δ 2.07, and pf is the population fraction occupying the free
Acknowledgment. We thank the NSF (CHE-0212210) for
financial support and the University of Chicago for a William
Rainey Harper fellowship (E.J.S.).
H
allylic position. These population fractions are related to mole fractions
of free ATMS (xfree) and 7a-coordinated ATMS (xcoord) such that pf )
xfree and pa ) pb ) 0.5‚xcoord
Supporting Information Available: Additional experimental
details, data for free substrates (pdf). This material is available
free of charge via the Internet at http://pubs.acs.org.
.
The site-to-site rate constants in R are related to rate constants for
chemical exchange processes as follows: for alkene flipping between
enantiofaces of 7a, kflip ) ka,b ) kb,a; for alkene decomplexation from
JA057524P
9
8650 J. AM. CHEM. SOC. VOL. 128, NO. 26, 2006