Structure, dynamics, and polymerization activity of zirconocenium ion pairs generated with boron-C6F5 compounds and Al 2R6

Article abstract of DOI:10.1021/om200369u

Figure Persented: The activation reaction of the olefin polymerization precatalyst Cp*₂ZrMe₂ with a boron-C ₆F₅ compound (B(C₆F₅)₃, [Ph₃C][B(C₆F₅)₄]) and an aluminum alkyl species (Al₂Me₆, Al₂ⁱBu ₆) is studied by NMR spectroscopy in order to determine the nature of the ion pairs that are formed preferentially. We show that a mixture of ion pairs with general formula Cp*₂Zr(Me)-μ-Me-E(C ₆F₅)_3-xR_x (E = Al, B; x = 0, 1; R = Me, ⁱBu) (1, 2a/b, 4) is generated due to a rapid transfer of pentafluorophenyl groups from boron to aluminum. Therefore, the molecular ratio of the activators determines the final composition of the ion pairs present in solution. When the pentafluorophenyl group transfer is suppressed, the ion pair Cp*₂Zr-(μ-Me)₂-AlMe₂ (5) forms irrespective of the reagent ratio. The high dynamics of these solutions is demonstrated by DNMR studies. Gibbs free energies of activation were determined of 13.6(12) kcal mol^-1 at 298 K for the cocatalyst exchange of ion pair Cp*₂Zr(Me)-μ-Me-Al(C₆F₅) ₂Me (2a) and 13(2) kcal mol^-1 at 298 K for the methyl exchange in Al₂(C₆F₅)_xMe _6-x (x = 0, 1). Due to the stability of the ion pairs generated from the Cp*₂ZrMe₂ precatalyst at temperatures relevant for polymerization, correlations between activities in ethylene polymerization and the nature of the ion pairs can be established. All solutions containing the various ion pairs were found to be catalytically active in ethylene polymerization except that containing the ion pair 2a, which was attributed to the reduced Lewis acidity of the abstractor, as supported by DFT calculations.

Full text of DOI:10.1021/om200369u

ARTICLE
pubs.acs.org/Organometallics
Structure, Dynamics, and Polymerization Activity of Zirconocenium
Ion Pairs Generated with Boron-C6F5 Compounds and Al2R6
Dꢀeborah Mathis, Erik P. A. Couzijn, and Peter Chen*
Laboratorium f€ur Organische Chemie, ETH Z€urich, Wolfgang-Pauli-Strasse 10, CH-8093 Z€urich, Switzerland
S Supporting Information
b
ABSTRACT: The activation reaction of the olefin polymeriza-
tion precatalyst Cp*2ZrMe2 with a boron-C6F5 compound
(B(C6F5)3, [Ph3C][B(C6F5)4]) and an aluminum alkyl species
i
(Al2Me6, Al2 Bu6) is studied by NMR spectroscopy in order to
determine the nature of the ion pairs that are formed preferen-
tially. We show that a mixture of ion pairs with general formula
Cp*2Zr(Me)Àμ-MeÀE(C6F5)3ÀxRx (E = Al, B; x = 0, 1; R =
Me, iBu) (1, 2a/b, 4) is generated due to a rapid transfer of pentafluorophenyl groups from boron to aluminum. Therefore, the
molecular ratio of the activators determines the final composition of the ion pairs present in solution. When the pentafluorophenyl
group transfer is suppressed, the ion pair Cp*2ZrÀ(μ-Me)2ÀAlMe2 (5) forms irrespective of the reagent ratio. The high dynamics
of these solutions is demonstrated by DNMR studies. Gibbs free energies of activation were determined of 13.6(12) kcal molÀ1 at
298 K for the cocatalyst exchange of ion pair Cp*2Zr(Me)Àμ-MeÀAl(C6F5)2Me (2a) and 13(2) kcal molÀ1 at 298 K for the methyl
exchange in Al2(C6F5)xMe6Àx (x = 0, 1). Due to the stability of the ion pairs generated from the Cp*2ZrMe2 precatalyst at
temperatures relevant for polymerization, correlations between activities in ethylene polymerization and the nature of the ion pairs
can be established. All solutions containing the various ion pairs were found to be catalytically active in ethylene polymerization
except that containing the ion pair 2a, which was attributed to the reduced Lewis acidity of the abstractor, as supported by DFT
calculations.
’ INTRODUCTION
well-defined structure, such as boron-C6F5 compounds (e.g.,
tris(pentafluorophenyl)borane (B(C6F5)3) or tetrakis(penta-
fluorophenyl)borate ([Ph3C][B(C6F5)4])).2 The activation of
metallocenes with exclusively a boron-C6F5 compound is the
subject of many reports,8,10À13 but it is well established that the
use of an aluminum alkyl species as co-activator significantly
improves the activity and stability of the catalytic system.14 One
role of the aluminum alkyl species is to scavenge impurities, yet it
may also be involved in the formation of the ion pairs.3
In this work, we investigate the reaction of boron-C6F5
compounds (B(C6F5)3, [Ph3C][B(C6F5)4]) and Al2R6 (R =
Me, iBu)15 with Cp*2ZrMe2 (Cp* = pentamethylcyclopentadienyl)
by 1H and 19F nuclear magnetic resonance (NMR) spectroscopy.
We show that the combination of activators and their molecular
ratios determine the final composition of the catalyst solution,
which in general contains several ion pairs rather than a single
one. Kinetic parameters for the structural rearrangement of the
ion pairs are determined by variable-temperature dynamic NMR
experiments. Furthermore, density functional theory (DFT)
calculations are presented for obtaining geometrical parameters
of the ion pairs. Lastly, the solutions containing the various ion
pairs are tested for their catalytic activity in ethylene polymeri-
zation. The Cp*2ZrMe2 precatalyst presents the advantage that
the ion pairs generated by reaction with the activators are stable
Despite the intensive research relating to the activation of
zirconocene precatalysts for olefin polymerization,1À5 the nature
and molecular ratios of the various ion pairs present in solution
often remain ill-defined when performing a polymerization. Yet,
it is these ion pairs that are responsible for the formation of the
active species. The difficulty in predicting the outcome of a
polymerization is mainly due to the lability of the catalytic system
and its high sensitivity to any variation in reaction conditions.
Even minimal changes in polymerization conditions can have a
substantial influence on the catalytic activity and on the molec-
ular structure of the formed polymer.6 Thus, it is essential
to identify the nature of the ion pairs, the conditions under
which they are generated, and their catalytic activities in olefin
polymerization.
A good activator for metallocene precatalysts should both
generate the cationic active species [L2MR]+ and provide a
weakly coordinating counteranion. The counteranion should
stabilize the highly electrophilic 14-electron active species but
should also be easily displaceable to allow olefin coordination.2
The geometry and thermodynamic nature of the cationÀanion
interaction is of central importance for the catalytic activity, as
reported by Marks and co-workers.7,8 Methylaluminoxane
(MAO) is still the most commonly used activator in industry
because it both generates large, weakly coordinating counter-
anions and has good scavenging properties.9 However, the
mechanism of activation is better studied using reagents with a
Received: May 4, 2011
Published: June 27, 2011
r
2011 American Chemical Society
3834
dx.doi.org/10.1021/om200369u Organometallics 2011, 30, 3834–3843
|

Organometallics
ARTICLE
under conditions close to those for ethylene polymerization,
providing a system uniquely well suited for studying the activa-
tion mechanism. Thus, correlations between the structural para-
meters of the different ion pairs, the strength of the Lewis acid
activator, and the polymerization activities are discussed.
’ EXPERIMENTAL SECTION
All air- and/or water-sensitive compounds were handled under an
inert atmosphere using standard Schlenk and glovebox techniques.
Ethylene (N35 grade, > 99.95%, PanGas), bis(pentamethylcyclopenta-
dienyl)zirconium dichloride (Cp*2ZrCl2, 97%, Acros), tris(pentafluoro-
phenyl)boron (B(C6F5)3, min. 97%, Strem), trityl tetrakis(pentafluoro-
phenyl)borate ([Ph3C][B(C6F5)4], 97%, Acros), trimethylaluminum
solution (TMA, 2.0 M in toluene, Aldrich), triisobutylaluminum solu-
tion (TIBA, 1.1 M in toluene, Acros), and methyllithium solution (1.6 M
in diethyl ether, Acros) were obtained commercially and used without
further purification. Research-grade solvents were distilled from sodium
(hexane, toluene) or Na/K alloy (diethyl ether) under nitrogen.
Toluene-d8 was stirred with Na/K alloy for 24 h and vacuum transferred.
Solutions of trimethylaluminum (0.95 M) and triisobutylaluminum in
toluene-d8 (0.52 M) were prepared from pure trimethylaluminum (min.
98%, ABCR) and pure triisobutylaluminum (Aldrich). Bis(pentamethyl-
cyclopentadienyl)zirconium dimethyl (Cp*2ZrMe2) was synthesized
according to a literature procedure.23
NMR spectra were recorded on Varian Gemini 300 and Bruker ARX
300 instruments (1H, 300 MHz; 19F, 282 MHz). 1H chemical shifts are
reported in ppm relative to tetramethylsilane, using residual solvent
protons as internal standard (toluene-d7: 1H: 2.03 ppm (quintuplet)).
19F chemical shifts are referenced externally to CF3Cl.
DFT calculations were performed using the Gaussian 09 package. The
M06-L24 density functional was used with the SDD basis set and an
automatic density-fitting basis for full optimization of the geometries.
A frequency calculation was performed for each stationary point to
confirm that it is a minimum.
Figure 1. 1H (top) and 19F (bottom) NMR spectra in toluene-d8 at
20 °C of the reaction between B(C6F5)3, Al2Me6, and Cp*2ZrMe2 (ratio
1:2:1), resulting in a mixture of species 1, 2a, Al2(C6F5)xMe6Àx, and
BMe3.
NMR Experiments. For a typical NMR experiment, solutions of
Cp*2ZrMe2 precatalyst (7.5 Â 10À2 M), of B(C6F5)3 or [Ph3C]-
i
[B(C6F5)4] activator, and of Al2Me6 or Al2 Bu6 co-activator (7.5 Â
10À2 M Â equiv((co)activator vs Zr)) in toluene-d8 were prepared in a
glovebox. The activators (0.2 mL of each solution) were mixed in a J.
Young NMR tube for 10 min, after which the precatalyst solution
(0.2 mL) was added, affording a final Zr concentration of 2.5 Â 10À2 M
in 0.6 mL of toluene-d8.
The polymers were filtered, vacuum-dried overnight, and weighed. Polym-
erization activities were derived from the amount of dried material produced
in milligrams per mole of precatalyst per ethylene pressure in bar per hour.
GPC. The molecular weight distributions of the produced polymers
were determined by gel permeation chromatography (PL-GPC 220) at
150 °C using 1,2,4-trichlorobenzene as solvent at a flow rate of 1.0 mL
minÀ1 and with refractive index and viscosimetry detection. Two different
sets of columns were used (2Â PLgel 5 μm MIXED-D and 2Â PLgel 10
μm MIXED-B), and the instrument was calibrated using 1 mg mLÀ1
solutions of monodisperse fractions of atactic polystyrene standards
(Fluka) with a universal calibration. Polyethylene samples with known
molecular masses and polydispersities were measured as reference.
For variable-temperature dynamic NMR (over the range À80 to
80 °C at every 20 °C), the NMR probe was equilibrated for 10 min at the
desired temperature before data acquisition. The spectra were analyzed
with the program iNMR;25 the internal line broadening at each
1
temperature was determined from the H signal of the Cp* ligands,
and the line shapes were simulated to afford experimental exchange rate
constants k. The thermodynamic parameters ΔHq, ΔSq, and ΔGq and
their standard deviations were determined from linear regression
analysis of Eyring plots of ln(k/T) vs 1/T.
’ RESULTS AND DISCUSSION
Polymerization Experiments. Toluene solutions of precatalyst
Cp*2ZrMe2 (7 Â 10À3 M, 5 mL), of B(C6F5)3 or [Ph3C][B(C6F5)4]
I. Characterization of the Zirconocenium Ion Pairs by 1H
(1.4 Â 10À2 M, 10 mL), and of Al2Me6 (3.5 Â 10À2 M, 5 mL) or Al2 Bu6
i
and 19F NMR Spectroscopy. a. Activation of Cp*2ZrMe2 with
(7 Â 10À2 M, 5 mL) were prepared under an argon atmosphere and
used for a series of five polymerization experiments. For each experi-
ment, the required equivalents of each activator relative to 7 Â 10À2
mmol Zr were combined in a 50 mL glass reactor, and toluene was added
to attain a total volume of 5.5 mL. The mixture was aged for 10 min at
40 °C, after which the precatalyst solution (1 mL) was added and the
reaction mixed for 3 min. After preactivation, the reaction was pressur-
ized with ethylene to a controlled total pressure of 2 bar (1 bar Ar and
1 bar ethylene) and quenched after 3 min with HCl/MeOH solution.
1
B(C6F5)3 and Al2R6. H and 19F NMR spectroscopy was used
to determine the outcome of the reaction of B(C6F5)3 with
Al2R6 (R = Me, iBu)15 and Cp*2ZrMe2. In a typical experiment,
B(C6F5)3 and Al2R6 were mixed in toluene-d8 for 10 min, after
which a solution of Cp*2ZrMe2 was added. The experiment was
1
performed with several reactant molecular ratios, and H and
19F NMR spectra were recorded. Typical spectra are shown in
Figures 1 and 2 (see below for further explanation).
3835
dx.doi.org/10.1021/om200369u |Organometallics 2011, 30, 3834–3843

Organometallics
ARTICLE
Scheme 2. The Formation of Various Ion Pairs Depends on
the Ratio between B(C6F5)3 and Al2R6 (R = Me, iBu) in the
Reaction with Cp*2ZrMe2
Table 1. Relative Amounts of the Ion Pairs 1 and 2a Gener-
ated in Solution by Reaction of B(C6F5)3, Al2Me6, and
Cp*2ZrMe2 at Different Molecular Ratiosa
B:Al:Zr
1 (%)
2a (%)
1:1:1
1:2:1
1:5:1
1:10:1
66
51
25
14
34
49
75
86
Figure 2. 1H (top) and 19F (bottom) NMR spectra in toluene-d8 at
20 °C of the reaction between B(C6F5)3, Al2Me6, and Cp*2ZrMe2 (ratio
2:1:1), resulting in a mixture of species 1, 4, and B(C6F5)xMe3Àx
.
a The percentages of ion pairs 1 and 2a were determined by integration
of the corresponding Cp* signals in the 1H NMR spectra (Figure S1).
Scheme 1. Exchange of Aryl and Alkyl Groups between B and
Al Centersa
formed when B(C6F5)3, Al2R6, and Cp*2ZrMe2 are combined
(Scheme 2).
With an excess of Al2R6 vs B(C6F5)3 (ratio B:Al < 1, Scheme 2,
top), all pentafluorophenyl ligands initially bound to boron are
transferred to aluminum, affording a mixture of Al2(C6F5)xR6Àx
(x = 0À6) and BR3. The subsequent reaction with Cp*2ZrMe2
results in a mixture of two methyl-bridged species, namely, the
ion pairs Cp*2Zr(Me)Àμ-MeÀAl(C6F5)3 (1) and Cp*2Zr(Me)À
μ-MeÀAl(C6F5)2R (2a, R = Me; 2b, R = iBu). Figure 1 depicts
the 1H and 19F NMR spectra at 20 °C of the reaction of B(C6F5)3
with Al2Me6 and Cp*2ZrMe2 at a B:Al:Zr ratio of 1:2:1 in
toluene-d8, showing the presence of ion pairs 1 and 2a as well
as residual Al2(C6F5)xMe6Àx and BMe3. The 1H NMR signals for
the bridging and the Zr-terminal methyl groups of 2a shift and
broaden with increasing measurement temperature, indicating
that ion pair 2a can undergo relatively facile dissociation/
recombination via neutral fragments (see below). At 20 °C, the
exchange reaction is fast so that, under the spectroscopic condi-
tions used, the signals coalesce and thus are not visible in the
NMR spectrum. The molecular ratio of ion pairs 1 and 2a
correlates with the initial ratio of the reagents (Table 1 and
Figure S1). Figure S1 shows the stacked 1H NMR spectra for four
reagent molecular ratios (B:Al:Zr = 1:1:1; 1:2:1; 1:5:1, and
1:10:1) at 20 and À80 °C. With increasing amount of Al2Me6
a The composition of the ligands on Al depends on the initial ratio of the
reagents.18
Boron-C6F5 compounds may undergo rapid exchange of the
pentafluorophenyl ligands for alkyl groups of Al2R6.16À22 This
reaction is often classified as a deactivation pathway16 and is fast
in apolar solvents such as toluene but almost totally suppressed in
coordinating solvents such as THF and ether.18 The driving force
for the transmetalation is the formation of AlÀC6F5 bonds,
which are stronger compared to AlÀalkyl bonds than the
corresponding BÀC6F5 vs BÀalkyl bonds,19 leading to the
transfer of all pentafluorophenyl groups to aluminum. Although
the metallocene is not required for the reaction to occur, we
cannot exclude that its presence might influence the reaction
rate.20 The final composition of a mixture of boron-C6F5
compounds and Al2R6 is dependent on the initial ratio of the
reactants, as depicted in Scheme 1. The ratio between B(C6F5)3
and Al2R6 therefore primarily determines which ion pairs are
3836
dx.doi.org/10.1021/om200369u |Organometallics 2011, 30, 3834–3843

Organometallics
ARTICLE
Scheme 3. The Reaction of [Ph3C][B(C6F5)4], Al2Me6 and
Cp*2ZrMe2 Results in the Formation of Ion Pair 5
([Cp*2ZrÀ(μ-Me)2ÀAlMe2][B(C6F5)4])
Figure 3. ORTEP representation of the X-ray crystal structure of ion
pair 1. Ellipsoids are drawn at 30% probability. The crystal structure is
disordered with respect to the Cp* orientation; the major of the two
refined positions is shown (occupation 56%). The hydrogen atoms on
the Cp* ligands are omitted for clarity. Selected distances [Å] and angles
[deg]: Zr1ÀC37 2.246(7), Zr1ÀC2 2.550(6), C2ÀAl3 2.008(6),
C37ÀZr1ÀC2 90.6(2), Zr1ÀC2ÀAl3 176.5(3).
vs Cp*2ZrMe2, the intensity of the Cp* signal of ion pair 2a
(δ 1.45) markedly increases at the expense of that of ion pair 1
(δ 1.40).
i
Using Al2 Bu6 as co-activator, a mixture was obtained contain-
i
ing ion pairs 1 and Cp*2Zr(Me)Àμ-MeÀAl(C6F5)2 Bu (2b),
which has a terminal isobutyl group on Al instead of a methyl
Figure 4. 1H NMR spectrum in toluene-d8 at 20 °C of the reaction of
[Ph3C][B(C6F5)4], Al2Me6, and Cp*2ZrMe2 (ratio 1:2:1), resulting in
the formation of ion pair 5.
1
group as in 2a. H and 19F NMR spectra are presented in
Figure S2.
When an excess of B(C6F5)3 vs Al2R6 is used (ratio B:Al > 1,
Scheme 2, bottom), aryl/alkyl group exchange occurs, but some
pentafluorophenyl ligands remain bound to boron so that a
mixture of B(C6F5)xR3Àx and Al(C6F5)3 is created. The sub-
sequent reaction with Cp*2ZrMe2 results in a mixture of two
methyl-bridged ion pairs, Cp*2Zr(Me)Àμ-MeÀAl(C6F5)3 (1)
and Cp*2Zr(Me)Àμ-MeÀB(C6F5)3 (4), for which Figure 2
depicts the 1H and 19F NMR spectra at 20 °C (reagent ratio B:
Al:Zr = 2:1:1). All chemical shifts are listed in Table S1. Ion pair 1
was structurally characterized by single-crystal X-ray diffraction
(Figure 3). Similar crystal structures have been reported for
L2Zr(Me)Àμ-MeÀAl(C6F5)3 (L = 1,2-(CH3)2C5H3; (CH3)4-
C5H).26,27 The structure of ion pair 1 differs from those reported
earlier in that the bridging methyl group is significantly more
“abstracted” by the aluminum moiety in 1, which may be
attributed to the steric environment caused by the Cp* ligands
(as reflected by the bite angle between the two Cp* planes of
41.4°).
C6F5 groups from boron to aluminum.16 The reaction affords the
same methyl-bridged species as when B(C6F5)3 and Al2 Bu6 are
used (Scheme S1). However, [Ph3C][B(C6F5)4] transfers four
C6F5 groups, while B(C6F5)3 transfers only three.
i
The trityl cation does not react with Al2Me6; that is, no
discoloration of the solution was observed, indicating that no
methide abstraction occurred that would form an [AlR2]+
transient species. Therefore, the borate anion remains unaffected
when [Ph3C][B(C6F5)4] is mixed with Al2Me6.16 Upon addition
of Cp*2ZrMe2, the trityl cation abstracts a methide from Zr to
form the complex [Cp*2ZrÀ(μ-Me)2ÀAlMe2]+[B(C6F5)4]À
(5) (Scheme 3). The 1H NMR spectrum of ion pair 5 is shown
in Figure 4. The cation [Cp*2ZrÀ(μ-Me)2ÀAlMe2]+ is in equilib-
rium with the “free” zirconocenium cation [Cp*2ZrMe]+. When
the amount of Al2Me6 is increased, the equilibrium shifts toward
ion pair 5, which thus decreases the activity of the catalyst (see
below).3,17,31
It is worth mentioning that bridged cations of the form
[L2Zr(Me)Àμ-MeÀ(Me)ZrL2]+ (L = Cp, Me2Si(Ind)2, C2H4-
(Ind)2) were reported by Bochmann,31,32 but that no similar
species was observed under any conditions used in this study.
This might be caused by the high steric demands of the Cp*
ligands, preventing such dimer formation, as has also been
suggested by Brintzinger and co-workers.12
b. Activation of Cp*2ZrMe2 with [Ph3C][B(C6F5)4] and Al2R6.
Several experiments were conducted to study the reaction of
[Ph3C][B(C6F5)4], Al2R6, and Cp*2ZrMe2. The mechanism of
i
activation depends on whether Al2Me6 or Al2 Bu6 is used in
combination with [Ph3C][B(C6F5)4]. Rapid discoloration of the
orange solution was observed when [Ph3C][B(C6F5)4] was
i
mixed with Al2 Bu6, which was attributed to the disappearance
of the trityl cation. The reaction involves hydride abstraction and
isobutene elimination to form transient cationic aluminum
species [AliBu2]+, which reacts further by fast transfer of the
Having identified the ion pairs that are generated under
various conditions, as summarized in Table 2, we now turn to
3837
dx.doi.org/10.1021/om200369u |Organometallics 2011, 30, 3834–3843

Organometallics
ARTICLE
Table 2. Summary of the Ion Pairs Generated by Reaction of Cp*2ZrMe2, a Boron-C6F5 Compound, and Al2R6
Scheme 4. Schematic Reaction Coordinates for the Reorganization Processes of Zirconocenium Ion Pairsad
a Left: anionic exchange pathway (ae) via dissociation of the ZrÀμ-Me bond. Right: cocatalyst exchange pathway (ce) via dissociation of the
μ-MeÀA bond.
study the dynamics of these solutions. The dynamics associated
with ion pair 5 will be treated in another report.33
Cp*2ZrMe2 in a 1:1:1 molecular ratio, and a mixture containing
additionally a residual amount of Al2(C6F5)xMe6Àx (x = 0, 1),
prepared from the same reagents in a 1:2:1 molecular ratio.
Figures 5 and 6 display the experimental and simulated spectra
for both solutions, and Figures 7 and 8 show the Eyring plots
obtained from line shape analysis. Scheme 5 and points 1À3
describe the mechanisms that are proposed for the methyl group
exchanges observed by NMR; all thermodynamic results are
presented in Table 3.
II. Dynamic NMR Study of Solutions Containing Ion Pairs 1
and 2a. As previously described by Marks and co-workers,27,34
the anionic moiety of the ion pair can migrate between the two
available sites on zirconium by two spectroscopically differenti-
able rearrangement processes. The first pathway, called anionic
exchange (ae, Scheme 4, left), implies a transition state in which
the transient [L2ZrMe]+ species dissociates from the anionic
Lewis base. Such a pathway does not lead to coalescence of NMR
signals for catalysts with symmetrical ligands on zirconium (e.g.,
Cp*2ZrMe2). The second process, called cocatalyst exchange (ce,
Scheme 4, right), leads to the migration of the activator A to the
other site on Zr via dissociation of the μ-MeÀA bond and hence
implying fragmentation into neutral species.7,34 Such a pathway
leads to the coalescence of the terminal and bridged methyl
signals of the ion pair.
The NMR spectra of the mixture of ion pairs 1 and 2a exhibit
broadening of several methyl signals due to the structural re-
arrangement processes explained above and due to other methyl
exchange reactions (see below). Two mixtures were subjected to
dynamic NMR experiments: a mixture containing only the ion
pairs 1 and 2a, prepared by reaction of B(C6F5)3, Al2Me6, and
1. Ion pair 2a: The signals of methyl groups A and B of ion
pair 2a broaden and coalesce in the NMR spectra due to
cocatalyst exchange (see above). The rate constant kAB was
used for determining the thermodynamic parameters for
this exchange reaction (i.e., kAB = kce), affording ΔrGq =
13.3(10) and 13.9(12) kcal molÀ1 at 298 K without and
with a residual amount of Al2(C6F5)xMe6Àx (x = 0, 1),
respectively. These Gibbs free energy barriers agree within
their uncertainties.
1
2. Ion pair 1: The H NMR signal for the terminal ZrÀMe
group of 1 remains sharp until at least 80 °C (Figure 9,
methyl Z), suggesting that a cocatalyst exchange reaction
similar to that observed for 2a does not occur for 1. Such a
3838
dx.doi.org/10.1021/om200369u |Organometallics 2011, 30, 3834–3843

Organometallics
ARTICLE
Figure 5. Experimental (left) and simulated (right) temperature-dependent 1H NMR spectra of a mixture of ion pairs 1 and 2a. Peak labels correspond
to the assignments in Scheme 5.
Figure 6. Experimental (left) and simulated (right) temperature-dependent 1H NMR spectra of a mixture of species 1, 2a, and Al2(C6F5)xMe6Àx (x = 0, 1).
Scheme 5. Methyl Group Exchange Reactions for a Solution
Containing Ion Pairs 1 and 2a (A and B) and for a Solution
Containing Ion Pairs 1, 2a, and Al2(C6F5)xMe6Àx (AÀC)
“ce” process would lead to a broadening of both the
bridging and terminal methyl signals of the ion pair. For
1, only a broadening of the bridging methyl signal is
observed (Figure 9, methyl D), which thus must arise from
a pathway involving the dissociation of the ZrÀμ-Me bond.
Simultaneously, a broadening of the terminal methyl on Al
in ion pair 2a (methyl C) is observed. Therefore, we
propose a mechanism in which these methyl groups exchange
by the involvement of a transient species AlMe(C6F5)2 of
low concentration, which is unobservable in the NMR
spectra (Scheme S2). In our simulations, we approximated
this reaction by a direct exchange of methyl group C with D.
Furthermore, as this reaction involves dissociation of the
ZrÀμ-Me bond of 1, we conclude that an anionic exchange
should also occur but cannot be observed due to the
symmetry of 1. In conclusion, for ion pair 1 cleavage of
the μ-MeÀAl bond requires considerably more energy than
the breaking of the ZrÀμ-Me bond.
3. Al2(C6F5)xMe6Àx (x =0, 1): At À80 °C, the spectrum exhibits
six nearly resolved methyl signals for Al2(C6F5)xMe6Àx
(x = 0, 1), which coalesce to a single chemical shift at higher
temperatures. For the simulations we presumed that the
exchange reactions occurring for the methyl groups of
Al2Me6 and Al2(C6F5)Me5 follow similar kinetics. They
were therefore approximated with a single rate constant
3839
dx.doi.org/10.1021/om200369u |Organometallics 2011, 30, 3834–3843

Organometallics
ARTICLE
exchange of the Al2(C6F5)xMe6Àx (x = 0, 1) methyl groups,
we obtained a ΔrGq = 13(2) kcal molÀ1 at 298 K.
DNMR Discussion. Table 3 summarizes our experimental data
and includes literature values for cocatalyst exchange and anionic
exchange of systems closely analogous to ours. For the ion pair
rac-(EBI)Zr(Me)Àμ-MeÀAl(C6F5)3 (EBI = C2H4(Ind)2),
Marks and co-workers27 reported ΔrGqae(298 K) ≈ 21 and
ΔrGqce(298 K) ≈ 20 kcal molÀ1 for “ae” and “ce” processes,
respectively. That we find a lower energy barrier for the cocatalyst
exchange process of 2a is consistent with the fact that Al(C6F5)2Me
is a weaker Lewis acid than Al(C6F5)3. The negative entropy
obtained for the “ce” of 2a suggests that this reaction occurs via
an associative pathway. We propose a transition state where an
incoming Al(C6F5)2Me binds to the vacant methyl on zirconium to
form a doubly activated ion pair. We already suggested above that
Al(C6F5)2Me is present in solution at low concentration. Doubly
activated ion pairs have been reported for several metallocenes
activated with Al(C6F5)3,28 and also spectroscopic evidence has
been reported for the formation of doubly B(C6F5)3-activated
metallocene complexes as intermediates in intermolecular borane
exchange processes.29,30
Figure 7. Eyring plots for the cocatalyst exchange reaction of 2a
(triangles) and for the methyl exchange CD (circles) for the solution
without residual Al2(C6F5)xMe6Àx (x = 0, 1).
Our DNMR data suggest that the anionic exchange process
of 1 involves a lower barrier than a cocatalyst exchange process.
This means that the “ae” process for 1 is probably more facile for
our system than that reported by Marks for rac-(EBI)Zr(Me)À
μ-MeÀAl(C6F5)3, as might be expected from the differences in
the respective crystal structures (see above). This can be attributed
tothe steric congestion ofthe Cp* ligandsthatweakens the cation/
anion attraction and lowers the energy barrier for the dissociation
into ionic fragments.8 In addition, ionic fragmentation will cause
the largest relief of steric congestion at the zirconium center.
The value for the overall methyl exchange processes of
Al2(C6F5)xMe6Àx (x = 0, 1) is slightly higher than that reported
in the literature for methyl exchange in Al2Me6.35 Yamamoto et al.
determined ΔrGq(298 K) = 10.7(2) kcal molÀ1 in cyclopentane
and claimed similar thermodynamics in toluene. This result
is reasonable since we use a single rate constant for simulat-
ing the exchange reaction of both Al2Me6 and Al2(C6F5)Me5.
The electron-withdrawing pentafluorophenyl group on Al
should stabilize the dimeric form of Al2(C6F5)Me5, resulting in
a higher energy barrier for methyl exchange. Furthermore, we
obtained a negative entropy of activation, which suggests that the
reaction proceeds in an associative manner under our conditions.
In the literature, both inter- and intramolecular processes for
methyl group exchange in Al2Me6 have been reported. A detailed
analysis will be presented in future work.33
Figure 8. Eyring plots for the cocatalyst exchange reaction of 2a
(triangles), the methyl exchange CD (circles), the methyl exchange
CE (stars), and the methyl exchange for Al2(C6F5)xMe6Àx (x = 0, 1)
(squares).
III. DFT Calculations. We performed DFT calculations of the
various zirconocenium ion pairs observed in the NMR study to
compare their geometrical parameters. Two additional ion pairs,
3a and 3b, were taken into consideration that were not observed
in the NMR study, having the chemical formula Cp*2Zr(Me)À
μ-MeÀAl(C6F5)Me2 and Cp*2Zr(Me)Àμ-MeÀAl(C6F5)iBu2,
i
Figure 9. 1H NMR spectrum at 80 °C of ion pairs 1 and 2a. A sharp
signal for the terminal ZrÀMe group of 1 is observed at À0.05 ppm.
respectively. Al(C6F5)R2 (R = Me, Bu) is probably not Lewis
acidic enough to form detectable ion pairs 3. Nevertheless, the
comparison of geometrical parameters along the series of ion pairs
1À2À3 is interesting to make from a theoretical point of view.
The geometries of ion pairs 1, 2a, 2b, 3a, 3b, and 4 were
optimized at the M06-L24 level with the SDD basis set and
pseudopotentials on all atoms, and the geometries are displayed
in Figure S3. The overall geometry of ion pair 1 is in good
agreement with the obtained crystal structure (Table S3), which
kEF. We also observed an exchange reaction occurring
between the methyl groups of Al2(C6F5)xMe6Àx (x = 0, 1)
and the methyl resonance C of 2a, which we simulated with a
fourth rate constant, kCE. This reaction probably also involves
the transient AlMe(C6F5)2 species postulated for the exchange
of methyl resonance C with D of 1 (see point 2). For the
3840
dx.doi.org/10.1021/om200369u |Organometallics 2011, 30, 3834–3843

Organometallics
Table 3. Thermodynamic Data (in kcal molÀ1, ΔrGq at 298 K, ΔrSq in cal molÀ1 KÀ1) for the DNMR Experiments in Toluenea
ARTICLE
species
ref
rate constant
pathwayb
ΔrGq
ΔrHq
ΔrSq
2ac
2ad
kAB
kAB
kEF
kCD
kCD
ce
ce
13.9(12)
13.3(10)
13(2)
15.9(8)
16.2(10)
≈21
5.1(8)
5.6(8)
8.1(14)
3.5(6)
0.5(5)
≈16
À29(3)
À26(3)
À16(5)
À42(4)
À53(3)
≈ À15
≈8
Al2(C6F5)xMe6Àx (x = 0, 1) d
exchange reaction of C (2a) with D (1) c
exchange reaction of C (2a) with D (1) d
rac-(EBI)Zr(Me)Àμ-MeÀAl(C6F5)3
27
27
27
27
36
8
ae
ce
ae
ce
ce
ae
ce
ae
ce
ce
≈20
≈22
rac-(EBI)Zr(Me)Àμ-MeÀB(C6F5)3
18.5
14(2)
À15(2)
8(2)
19.6
22(1)
18.4
18.3e
19.7e
14.4f
18.0f
19.8e
(Me2C5H3)2Zr(Me)Àμ-MeÀB(C6F5)3
((SiMe3)2)C5H3)2Zr(Me)Àμ-MeÀB(C6F5)3
Cp*2Zr(Me)Àμ-MeÀB(C6F5)3 (4)
8
8
8
8
g
35
Al2Me6
10.7(2)
15.5(15)
15.9
a Literature values are presented in the second part of the table. b ae and ce stand for anion exchange and cocatalyst exchange, respectively. c Mixture of ion
pairs 1 and 2a. d Mixture of ion pairs 1, 2a, and residual Al2(C6F5)xMe6Àx. e At 80 °C. At 35 °C. g In cyclopentane.
f
Table 4. Comparison of Selected Geometrical Parameters for
the Calculated Structures (M06-L/SDD) of Ion Pairs 1, 2a,
2b, 3a, 3b, and 4 (distances [Å] and angles [deg]; 1À3, E = Al;
4, E = B)
than 2a, which is probably due to the increased steric demands
of the iBu group causing congestion with the Cp* ligands. This
will also facilitate the formation of ionic fragments, which is
important for the activity of the catalyst.
Having studied the geometries of the various ion pairs, we
now investigate the activities in ethylene polymerization of the
different catalytic solutions containing those ion pairs.
IV. Activity in Ethylene Polymerization. In this section we
investigate the catalytic activities of solutions containing the
various ion pairs described above. First we compare the catalytic
systems with B(C6F5)3 as the activator, which results in the
formation of ion pairs 1, 2, and 4. Then, the activities of catalytic
solutions containing ion pair 5 are discussed (see Schemes 2 and
3 for the chemical structures of the ion pairs).
1
2a
2b
3a
3b
4
ZrÀMe
2.259
2.530
2.031
4.547
92.37
2.265
2.477
2.072
4.542
90.01
2.257
2.505
2.065
4.567
93.19
2.257
2.457
2.122
4.579
95.17
2.261
2.470
2.115
4.552
93.51
2.251
2.605
1.673
4.275
91.92
ZrÀμ-Me
μ-MeÀE
distance Zr
E
3 3 3
μ-MeÀZrÀMe
ZrÀμ-MeÀE
171.01 173.22 178.04 178.44 173.69 176.37
From our NMR study it is possible to predict which ion pairs
are present in solution for given initial reactant ratios. In order to
investigate the activities of the various ion pairs in ethylene
polymerization, a series of polymerization experiments was
conducted in which the type and ratios of the reagents were
modified. Two main differences in the conditions used for the
NMR study and the polymerization experiments may influence
the formation and ratios of the ion pairs present in the catalytic
mixture: (1) the concentrations used for the NMR experiments
were higher than for the polymerization experiments (ccat = 25 vs
1 mM) and (2) a minimum of 5 equivalents of Al2R6 relative to
Zr was used to increase the reproducibility of the polymerization
results.
For our polymerization experiments, we applied a preactiva-
tion period of 3 min, during which all reactants were mixed
under an argon atmosphere before the introduction of ethylene.
Thus, formation of the ion pairs studied by NMR spectroscopy
was ensured, and an induction period arising from slow forma-
tion of the active species was prevented. Otherwise, different
ion pairs may form that may also be active in ethylene
polymerization, especially when the conditions prevent transfer
of C6F5 groups, such as when ethylene is added before B-
(C6F5)3 and Cp*2ZrMe2.
therefore supports the utilization of the calculated geometrical
parameters for our discussion.
Three different approaches are used to draw conclusions from
the geometries of the various ion pairs. First, we evaluate the role
of the main-group atom E by comparing ion pairs 1 and 4. Next,
we describe the influence of the number of pentafluorophenyl
groups on Al, i.e., moving in the series from ion pair 1 to 3.
Finally, we look at the influence of the type of alkyl groups (Me or
iBu) by comparing ion pair 2a with 2b and 3a with 3b.
1. Marks et al.27 have reported that B(C6F5)3 has significantly
higher Lewis acidity and methide affinity than does Al-
(C6F5)3. Our results are consistent with their findings, as
the bond between zirconium and the bridging methyl
group is longer in ion pair 4 than in ion pair 1 (ZrÀμ-
Me: 2.605 Å (4) vs 2.530 Å (1)).
2. Along the series of ion pairs 1, 2, and 3, the ZrÀμ-Me bond
length decreases and the μ-MeÀAl bond length increases,
reflecting the decrease in Lewis acidity of the Al(C6F5)n-
R3Àn moiety with increasing n.
3. The terminal alkyl groups on Al affect the geometry of the
ion pair. The largest effect is observed for the ZrÀμ-Me bond
length with values of 2.530, 2.505, and 2.477 Å for 1, 2b, and
2a, respectively. The bond length for 2b is closer to that for 1
3841
dx.doi.org/10.1021/om200369u |Organometallics 2011, 30, 3834–3843

Organometallics
ARTICLE
Table 5. Data of Ethylene Polymerization Catalyzed by
B(C6F5)3, Al2R6 (R=Me, iBu), and Cp*2ZrMe2 (c=1Â 10À3 M)
at 40 °C and 2 bar (1 bar ethylene +1 bar Ar) for 3 min
equilibrium with the loose ion pair Cp*2ZrMe+ [B(C6F5)4]À
3 3 3
(Scheme 3). As previously described,3 we observed that increas-
ing the amount of Al2Me6 decreases the activity of ion pair 5
(Figure S4). Conversely, increasing the amount of [Ph3C][B-
(C6F5)4] slightly improved the catalytic activity (Figure S4).
This has been explained by Bochmann with the “trityl effect”,38,39
according to which the increase in activity is due to the aggrega-
tion of the metallocenium ion pair with [Ph3C][B(C6F5)4].
ratio
ion pair
entry
R
(B:Al:Zr) presenta activityb Mn [Da] PDI Nchains/NZr
1
2
3
4
5
6
7
8
9
Me 10:5:1
iBu 10:5:1
4, 1
79
66
75
97
83
100
1
5405 1.79
4913 1.58
7431 1.52
6234 1.71
9262 2.11
10 060 1.71
0.41
0.43
0.33
0.46
0.21
0.29
4, 1
Me
iBu
5:5:1
5:5:1
1
’ CONCLUSION
1
Me 2.5:5:1
iBu 2.5:5:1
1, 2a
1, 2b
2a, (1)
2b, (1)
In this work, we describe the activation mechanism of a
zirconocene precatalyst with classical reagent combinations,
i.e., with different combinations of a boron pentafluorophenyl
compound and an aluminum alkyl species. Using NMR techni-
ques, we show that different ratios of various ion pairs are formed
depending on the activators and their respective ratios.
Me
iBu
Me
1:5:1
1:5:1
65
10
77
11 861 1.61
2534 1.47
7196 2.29
0.17
0.14
0.23
1:20:1 2a
1:20:1 2b
10 iBu
a From NMR study. b [kgPE mol ZrÀ1 hÀ1 barÀ1].
Most of the reagent combinations studied in this work lead to
a mixture of ion pairs 1, 2a/2b, and 4 (Table 2), which is the
result of transmetalation of pentafluorophenyl groups between
the activators, i.e., from boron to aluminum. Therefore, the ratio
of the formed ion pairs is highly dependent on the initial ratio of
the activators. The catalytic solutions containing ion pairs 1, 2b,
and 4 all actively catalyze the polymerization of ethylene.
Polymerization data indicate that the solutions containing the
ion pair 2a were inactive. Accordingly, DFT calculations suggest
a tighter ion pair for 2a than for 2b. When the transmetalation of
the pentafluorophenyl groups is suppressed, ion pair 5 forms
irrespective of the initial activator ratio. The activity of this
catalyst system is still influenced by the amount of Al2Me6.
Using DNMR techniques, we investigated the dynamics of
two solutions containing ion pairs 1 and 2a in the presence and
absence of Al2(C6F5)xMe6Àx (x = 0, 1). For the cocatalyst
exchange process of 2a, we obtained ΔrGq = 13.6(12) kcal
molÀ1 at 298 K. Concerning ion pair 1, the energy barrier for
cocatalyst exchange, involving breaking of the μ-MeÀAl bond,
lies well above that for anionic exchange, involving breaking of
the ZrÀμ-Me bond. Furthermore, the shapes of the dynamic
NMR spectra reveal that most methyl groups are labile, which
reflects the difficulty of defining the composition of solutions and
the nature of active species when performing a polymerization
experiment.
We note that the correlation between the observed species in
NMR and the activity observed in ethylene polymerization is
delicate because the most abundant species in solution may not
be the major contributor to activity. Furthermore, mass transport
limitation is a common drawback for precise determination
of catalytic activities, especially for ethylene polymerization.3
Nevertheless, we could correlate the activity in ethylene polym-
erization of the different reaction mixtures with the presence of
the ion pairs identified by NMR.
Table 5 lists the results of the polymerization experiments
catalyzed by mixtures of B(C6F5)3, Al2R6, and Cp*2ZrMe2 at
different reagent ratios. The table gives the predicted ion pairs
present in solution (based on the NMR study), the polymeriza-
tion activities, and the polymer molecular structure. The follow-
ing observations can be made:
1. The catalytic mixtures containing ion pairs 1 and 4 were
both active in ethylene polymerization and afforded similar
activities and polymer molecular structures (Table 5, en-
tries 1À4). The comparable results for entries 1 vs 2 and 3
vs 4 are in agreement with the fact that the formation of ion
pairs 1and 4is independent of the alkylaluminum species used.
2. The catalytic solutions containing ion pair 2a showed very
low activity, whereas those containing ion pair 2b were
comparable in activity to solutions of ion pairs 1 and 4
(Table 5, entries 7À10). Therefore, the replacement of the
methyl group on Al by isobutyl in ion pair 2 is sufficient to
improve catalyst activity. This suggests easier ionic disso-
ciation of ion pair 2b as compared to 2a, which is in
agreement with the calculated longer ZrÀμ-Me and shorter
μ-MeÀAl bonds.
To conclude, we show in this work that the transfer of
pentafluorophenyl groups from boron to aluminum, often re-
ferred to as a deactivation pathway,16 may be the origin of the
formation of active catalytic mixtures.
’ ASSOCIATED CONTENT
S
Supporting Information. Additional NMR spectra and
b
3. The number of polymer chains produced relative to the
available Zr centers (Nchains/NZr) was notably lower for the
catalytic solutions containing ion pair 2b (and also 2a) than
for those containing ion pairs 1 and 4 (Table 5). This can be
explained by a lower proportion of active centers generated
by dissociation of ion pair 2b (and 2a), which is a result of
reaction schemes, calculated structures of the various ion pairs,
and polymerization activities of ion pair 5. This material is
available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
i
Corresponding Author
*E-mail: chen@org.chem.ethz.ch.
the lower Lewis acidity of AlÀ(C6F5)2 Bu as compared to
AlÀ(C6F5)3 in 1. The difference in polymerization activ-
ities was not as pronounced as that in the number of poly-
mer chains, probably because of mass-transport limitation.
’ ACKNOWLEDGMENT
The reaction of [Ph3C][B(C6F5)4], Al2Me6, and Cp*2ZrMe2
The authors acknowledge the support from the ETH Z€urich
leads to the formation of the contact ion pair 5, which is in
and the Swiss National Science Foundation. We gratefully
3842
dx.doi.org/10.1021/om200369u |Organometallics 2011, 30, 3834–3843

Organometallics
ARTICLE
acknowledge Dr. Andreas Bach for fruitful discussions, especially
concerning DFT calculations.
(30) Green, M. L. H.; Sassmannshausen, J. Chem. Commun.
1999, 115–116.
(31) Bochmann, M.; Lancaster, S. J. Angew. Chem., Int. Ed. Engl.
1994, 33, 1634–1637.
(32) Lancaster, S. J.; Bochmann, M. J. Organomet. Chem. 2002,
654, 221–223.
(33) Manuscript in preparation.
(34) Deck, P. A.; Beswick, C. L.; A.; Marks, T. J. J. Am. Chem. Soc.
1998, 120, 1772–1784.
(35) Yamamoto, O.; Hayamizu, K.; Yanagisawa, M. J. Organomet.
Chem. 1974, 73, 17–25, and references therein.
(36) Siedle, A. R.; Newmarks, R. A. J. Organomet. Chem. 1995,
497, 119–125.
(37) Liu, Z.; Somsook, E.; White, C. B.; Rosaaen, K. A.; Landis, C. R.
J. Am. Chem. Soc. 2001, 123, 11193–11207.
(38) Song, F.; Cannon, R. D.; Lancaster, S. J.; Bochmann, M. J. Mol.
’ REFERENCES
(1) Bochmann, M. Organometallics 2010, 29, 4711–4740.
(2) Chen, Y.-X.; Marks, T. J. Chem. Rev. 2000, 100, 1391–1434.
(3) Bochmann, M J. Organomet. Chem. 2004, 689, 3982–3998.
(4) Macchioni, A Chem. Rev. 2005, 105, 2039–2073.
(5) Fujita, T.; Makio, H. In Comprehensive Organometallic Chemistry
III, Vol. 11; Elsevier, 2007; pp 692À697, and references therein.
(6) Janiak, C.; Versteeg, U.; Lange, K. C. H.; Weimann, R.; Hahn, E.
J. Organomet. Chem. 1995, 501, 219–234.
(7) Luo, L.; Marks, T. J. Top. Catal. 1999, 7, 97–106.
(8) Yang, X.; Stern, C.; Marks, T. J. J. Am. Chem. Soc. 1994,
116, 10015–10031.
(9) Ystenes, M.; Eilertsen, J. L.; Liu, J.; Ott, M.; Rytter, E.; Stovneng,
J. A. J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 3450–3450.
(10) Henderson, L. D.; Piers, W. E. J. Organomet. Chem. 2007,
692, 4661–4668.
Catal. A 2004, 218, 21–28.
(39) Alonso-Moreno, C.; Lancaster, S. J.; Zuccaccia, C.; Macchioni,
A.; Bochmann, M. J. Am. Chem. Soc. 2007, 129, 9282–9283.
(11) Tritto, I.; Donetti, R.; Sacchi, M. C.; Locatelli, P.; Zannoni, G.
Macromolecules 1997, 30, 1247–1252.
(12) Beck, S.; Prosenc, M.-H.; Brintzinger, H.-H.; Goretzki, R.;
Herfert, N.; Fink, G. J. Mol. Catal. A 1996, 111, 67–79.
(13) Beringhelli, T.; Donghi, D.; Maggioni, D.; D’Alfonso, G. Coord.
Chem. Rev. 2008, 252, 2292–2313.
(14) Brintzinger, H. H.; Fischer, D.; M€ulhaupt, R.; Rieger, B.;
Waymouth, R. M. Angew. Chem., Int. Ed. Engl. 1995, 34, 1143–1170.
(15) In solution, the dimeric species Al2R6 are in equilibrium with
their monomeric form, the exact equilibrium constant depending on R
and on temperature. For consistency and clarity of the data presentation,
we always use the notation Al2R6 (R = Me, iBu), even though Al2 Bu6 is
i
essentially unassociated (monomeric) in solution.
(16) Bochmann, M.; Sarsfield, M. J. Organometallics 1998, 17, 5908–
5912.
(17) Bochmann, M.; Lancaster, S. J. J. Organomet. Chem. 1995,
497, 55–59.
(18) Klosin, J.; Roof, G. R.; Chen, E. Y.-X.; Abboud, K. A. Organo-
metallics 2000, 19, 4684–4686.
(19) M06-L/SDD calculations provided homolytic bond strengths
of 105.9 (MeÀBMe2), 126.1 (C6F5ÀBMe2), 86.9 (MeÀAlMe2), and
109.8 kcal molÀ1 (C6F5ÀAlMe2). Hence, the transfer of a C6F5 moiety
from boron to aluminum in exchange for a methyl group is net
exothermic by about 2.7 kcal molÀ1. In accordance, the reaction
B(C6F5)3 + AlMe3 f BMe3 + Al(C6F5)3 was calculated to be 7.7 kcal
molÀ1 exothermic, which outweighs the energy required to break up 1/2
equiv of dimeric Al2Me6 (6.6 kcal molÀ1).
(20) At least to some extent, the aryl/alkyl group exchange was also
observed when Al2Me6 was added to the preformed zirconocenium/
borate. However, the quality of the resulting NMR spectrum was low,
which was probably due to the high reactivity and sensitivity of the
zirconocenium/borate ion pair when it is present in solution without
alkyl aluminum species.
(21) Janiak, C.; Lassahn, P.-G. Macromol. Symp. 2006, 236, 54–62.
(22) Piers, W. E. Adv. Organomet. Chem. 2005, 52, 49–51.
(23) (a) Manriquez, J. M.; McAlister, D. R.; Sanner, R. D.; Bercaw,
J. E. J. Am. Chem. Soc. 1978, 100, 2716–2724. (b) Miller, F. D.; Sanner,
R. D. Organometallics 1988, 7, 818–825.
(24) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157–167.
(25) iNMR: http://www.inmr.net.
(26) Liu, Z.; Somsook, E.; Landis, C. R. J. Am. Chem. Soc. 2001,
123, 2315–2916.
(27) Stahl, N. G.; Salata, M. R.; Marks, T. J. J. Am. Chem. Soc. 2005,
127, 10898–10909.
(28) Chen, Y.-X; Kruper, W. J.; Roof, G.; Wilson, D. R. J. Am. Chem.
Soc. 2001, 123, 745–746.
(29) Al-Humydi, A.; Garrison, J. C.; Youngs, W. J.; Collins, S.
Organometallics 2005, 24, 193–196.
3843
dx.doi.org/10.1021/om200369u |Organometallics 2011, 30, 3834–3843