Reactions of vinyl acetate and vinyl trifluoroacetate with cationic diimine Pd(II) and Ni(II) alkyl complexes: Identification of problems connected with copolymerizations of these monomers with ethylene

Article abstract of DOI:10.1021/ja045969s

Vinyl acetate (VA) and vinyl trifluoroacetate (VA_f) react with [(NAN)Pd(Me)(L)][X] (M = Pd, Ni, (N∧N) = N,N′-1,2- acenaphthylenediylidene bis(2,6-dimethyl aniline), Ar_f = 3,5-trifluoromethyl phenyl, L = Ar_fCN, Et₂O; X = B(Ar _f)₄^-, SbF₆^-) to form π-adducts 3 and 5 at -40 °C. Binding affinities relative to ethylene have been determined. Migratory insertion occurs in a 2,1 fashion (ΔG ^? = 19.4 kcal/mol, 0 °C for VA, and 17.4 kcal/mol, -40 °C for VA_f) to yield five-membered chelate complexes [(N∧N)Pd(κ²-CH(Kt)(OC(O)-CH₃))]⁺, 4, and [(N∧N)Pd(κ²-CH(Et)(OC(O)CF₃))]⁺, 6. When VA is added to [(N∧N)Ni(CH₃)]⁺, an equilibrium mixture of an η² olefin complex, 8c, and a κ-oxygen complex, 8o, results. Insertion occurs from the η² olefin complex, 8c (ΔG^? = 15.5 kcal/mol, -51 °C), in both a 2,1 and a 1,2 fashion to generate a mixture of five- and six-membered chelates, 9_2,1 and 9_1,2. VA_f inserts into the Ni-CH ₃ bond (-80 °C) to form a five-membered chelate with no detectable intermediate. Thermolysis of the Pd chelates results in β-acetate elimination from 4 (ΔG^? = 25.5 kcal/mol, 60 °C) and β-trifluoroacetate elimination from 6 (ΔG ^? = 20.5 kcal/mol, 10 °C). The five-membered Ni chelate, 9_2,1, is quite stable at room temperature, but the six-membered chelate, 9_1,2, undergoes β-elimination at -34 °C. Treatment of the OAc_f containing Pd chelate 6 with ethylene results in complete opening to the π-complex [(N∧N)Pd(κ²-CH(Et)(OAc _f))(CH₂CH₂)]⁺ (OAc_f = OC(O)CF₃), 18, while reaction of the OAc containing Pd chelate 4 with ethylene establishes an equilibrium between 4 and the open form 16, strongly favoring the closed chelate 4 (ΔH= -4.1 kcal/mol, ΔS = -23 eu, K = 0.009 M^-1 at 25 °C). The open chelates undergo migratory insertion at much slower rates relative to those of the simple (N∧N)Pd(CH₃)(CH₂CH₂)⁺ analogue. These quantitative studies provide an explanation for the behavior of VA and VA_f in attempted copolymerizations with ethylene.

Full text of DOI:10.1021/ja045969s

Published on Web 03/17/2005
Reactions of Vinyl Acetate and Vinyl Trifluoroacetate with
Cationic Diimine Pd(II) and Ni(II) Alkyl Complexes:
Identification of Problems Connected with Copolymerizations
of These Monomers with Ethylene
B. Scott Williams, Mark D. Leatherman, Peter S. White, and Maurice Brookhart*
Contribution from the UniVersity of North Carolina, Department of Chemistry CB 3290,
Venable Hall, Chapel Hill, North Carolina 27599-3290
Received July 6, 2004; E-mail: mbrookhart@unc.edu
Abstract: Vinyl acetate (VA) and vinyl trifluoroacetate (VAf) react with [(N∧N)Pd(Me)(L)][X] (M ) Pd, Ni,
(N∧N) ) N,N′-1,2-acenaphthylenediylidene bis(2,6-dimethyl aniline), Arf ) 3,5-trifluoromethyl phenyl, L )
ArfCN, Et2O; X ) B(Arf)4-, SbF6-) to form π-adducts 3 and 5 at -40 °C. Binding affinities relative to ethylene
have been determined. Migratory insertion occurs in a 2,1 fashion (∆Gq ) 19.4 kcal/mol, 0 °C for VA, and
17.4 kcal/mol, -40 °C for VAf) to yield five-membered chelate complexes [(N∧N)Pd(κ2-CH(Et)(OC(O)-
CH3))]+, 4, and [(N∧N)Pd(κ2-CH(Et)(OC(O)CF3))]+, 6. When VA is added to [(N∧N)Ni(CH3)]+, an equilibrium
mixture of an η2 olefin complex, 8c, and a κ-oxygen complex, 8o, results. Insertion occurs from the η2
olefin complex, 8c (∆Gq ) 15.5 kcal/mol, -51 °C), in both a 2,1 and a 1,2 fashion to generate a mixture
of five- and six-membered chelates, 92,1 and 91,2. VAf inserts into the Ni-CH3 bond (-80 °C) to form a
five-membered chelate with no detectable intermediate. Thermolysis of the Pd chelates results in â-acetate
elimination from 4 (∆Gq ) 25.5 kcal/mol, 60 °C) and â-trifluoroacetate elimination from 6 (∆Gq ) 20.5
kcal/mol, 10 °C). The five-membered Ni chelate, 92,1, is quite stable at room temperature, but the six-
membered chelate, 91,2, undergoes â-elimination at -34 °C. Treatment of the OAcf containing Pd chelate
6 with ethylene results in complete opening to the π-complex [(N∧N)Pd(κ2-CH(Et)(OAcf))(CH2CH2)]+ (OAcf
) OC(O)CF3), 18, while reaction of the OAc containing Pd chelate 4 with ethylene establishes an equilibrium
between 4 and the open form 16, strongly favoring the closed chelate 4 (∆H ) -4.1 kcal/mol, ∆S ) -23
eu, K ) 0.009 M-1 at 25 °C). The open chelates undergo migratory insertion at much slower rates relative
to those of the simple (N∧N)Pd(CH3)(CH2CH2)+ analogue. These quantitative studies provide an explanation
for the behavior of VA and VAf in attempted copolymerizations with ethylene.
Introduction
turnover frequencies in copolymerizations are greatly reduced
relative to ethylene homopolymerization, making use of these
The study of late transition metal olefin polymerization
catalysis has advanced remarkably in recent years.1 A primary
motivating factor for the development of such catalysts is that
the late metals are less oxophilic than early metals and, hence,
more compatible with functionalized olefins. Of particular
interest is the copolymerization of ethylene with polar vinyl
monomers. It has been demonstrated that the palladium diimine
complexes [(N∧N)PdMe(L)][B(Arf)4] ((N∧N) ) N,N′-1,2-
acenaphthylenediylidene bis(2,6-dimethyl aniline), L ) NCArf
(Arf ) 3,5-(CF3)2C6H3), 1; L ) Et2O, 2a) are active catalysts
for the polymerization of ethylene2-4 and are capable of
copolymerizing ethylene and methyl acrylate (MA).4 However,
catalysts commercially impractical. Recently, the DuPont group
has developed similar copolymerizations using an analogous
nickel catalyst operating at high temperature.5 However, the vast
majority of potential polar vinyl comonomers have remained
intractable.1 While some unsaturated species are simply not
incorporated into a copolymer but do not interfere with ethylene
homopolymerization, others (such as vinyl nitriles, vinyl halides,
and vinyl acetate) shut down polymerization activity. In the case
of vinyl nitriles, the nitrile functionality binds directly to the
active polymerization site of palladium, while, for vinyl halides,
â-halide elimination from the growing polymer chain gen-
erates unreactive metal halide dimers.6,7 However, in the case
of vinyl acetate (VA), the mechanism of inhibition has yet to
be explored.
(1) (a) Ittel, S. D.; Johnson L. K.; Brookhart, M. Chem. ReV. 2000, 100, 1169
and references therein. (b) Gibson, V. C.; Spitzmesser, S. K. Chem. ReV.
2003, 103, 283 and references therein.
(2) The designations a and b are used throughout to distinguish identical
-
-
cationic complexes paired with B(Arf)4 and SbF6 counterions, respec-
tively. In cases where only one of the two salts was prepared, a and b are
omitted.
(5) (a) Polym. Mater. Sci. Eng. 2002, 86, 319. (b) Polym. Mater. Sci. Eng.
2002, 86, 320-321
(6) (a) Kang, M.; Sen, A.; Zakharov, L.; Rheingold, A. L. J. Am. Chem. Soc.
2002, 124, 12080. (b) Schultz, L. H.; Brookhart, M. Unpublished results.
(7) (a) Foley, S. R.; Stockland, R. A., Jr.; Shen, H.; Jordan, R. F. J. Am. Chem.
Soc. 2003, 125, 4350. (b) Shen, H.; Jordan, R. F. Organometallics 2003,
125, 4350.
(3) Tempel, D. J.; Johnson, L. K.; Huff, R. L.; White, P. L.; Brookhart, M. J.
Am. Chem. Soc. 2000, 122, 6686.
(4) Mecking, S.; Johnson, L. K.; Wang, L.; Brookhart, M. J. Am. Chem. Soc.
1998, 120, 888.
9
5132
J. AM. CHEM. SOC. 2005, 127, 5132-5146
10.1021/ja045969s CCC: $30.25 © 2005 American Chemical Society

Ethylene Copolymerizations Problems Using Vinyl Acetate
A R T I C L E S
Scheme 1. Formation of VA and VAf Chelates
Understanding the behavior of VA is necessary if catalysts
are to be modified in a manner which may enable ethylene/VA
copolymerization. We report here a detailed study of the
reactions of VA and related unsaturated acetates with Ni and
Pd diimine complexes, which provides fundamental information
regarding insertion barriers of these monomers, their binding
affinities relative to ethylene, the nature and stability of the
insertion products, and the propensity of these inserted species
to incorporate further equivalents of monomer. This body of
information provides significant insight into the problems of
employing vinyl acetate and related comonomers in ethylene
copolymerizations.
[(N∧N)PdMe(OEt2)][SbF6] (2b) reacts in a similar fashion with
vinyl acetate to form 3b (Scheme 1) and subsequently the
chelate 4b. The insertion reaction is first-order in palladium and
displays clean kinetic behavior. Figure 1 shows an Eyring plot
for this insertion reaction over the range -40° to 0 °C in CD2Cl2.
From these data, values of ∆Hq ) 20.2(1.3) kcal/mol and ∆Sq
) 3(5) eu were obtained. At 0 °C, k ) 1.64(2) × 10-3 s-1 and
∆Gq ) 19.4(1) kcal/mol. The corresponding vinyl trifluoro-
acetate (CH2CHO2CCF3, VAf) complex, [(N∧N)PdMe(η2-
C2H3(OAcf))][SbF6] (5, OAcf ) O2CCF3) can be similarly
generated from 2b and VAf. 5 displays similar spectral features
to the vinyl acetate analogue 3, including the upfield shift of
the methylidene olefinic protons (4.18 and 4.66, relative to 5.17
and 4.89 in the free olefin) and the broadening of the signals at
low temperature due to slow rotation. However, the signals of
5 do not decoalesce at accessible temperatures. Upon warming
to -40 °C the VAf moiety inserts into the Pd-C bond with
kobs ) 2.43(1) × 10-4 s-1 (∆Gq ) 17.4(1) kcal/mol) to afford
the chelate complex [(N∧N)PdMe(κ2-CH(OAcf)Et)][SbF6] (6).
At the same temperature, the insertion barrier for vinyl acetate
is calculated to be 19.5 kcal/mol (see above). This is in keeping
with the perception of olefin insertion as a nucleophilic attack
by the palladium alkyl upon the coordinated olefin; electron-
withdrawing groups increase the electrophilicity of the olefinic
group.
When the analogous nickel cation [(N∧N)NiMe(OEt2)]-
[B(Arf)4] (7) is exposed to ca. 2 equiv of VA at -80 °C in
CDCl2F, g90% of the ether is displaced to form a new complex,
[(N∧N)NiMe(C2H3OAc)][B(Arf)4] (8) (Scheme 2). The two
methylidene signals remain broad under these conditions, but
upon lowering the temperature to -120 °C, they separate into
two pairs of signals in a 9:1 ratio (∆G ) 0.7 kcal/mol). The
lesser of these is strongly shielded (δ ) 3.82, 4.06) and is
assigned to the η2-olefin complex, 8c, with a Ni-Me signal at
-0.61 ppm, while the major product (δ ) 4.5, 4.7, partially
obscured by free olefin) is assigned to the O-bound isomer, 8o
with a Ni-Me signal at -0.26 ppm. Ziegler11 has predicted
that the Olefin-bound isomer should be preferred over the O-
bound isomer by over 5 kcal/mol for palladium, yet the O-bound
species is expected to be 0.7 kcal/mol more favorable for nickel.
Both predictions are in excellent agreement with our results.
Upon warming to -51 °C, 8 undergoes insertion (kobs ) 5.5 ×
Results
Low-Temperature Generation of π-Olefin Complexes and
Insertion into the M-Me Bond. Addition of vinyl acetate (VA)
to the cationic palladium complex [(N∧N)PdMe(OEt2)][B(Arf)4]
(2a) at -78 °C results in the quantitative release of ether and
formation of a new complex, [(N∧N)PdMe(η2-(C2H3)O2CCH3)]-
[B(Arf)4] (3a) (Scheme 1). The 1H NMR signals of the
methylidene protons appear upfield (δ ) 4.05, 4.20) of those
for free VA (δ ) 4.60, 4.91), consistent with an η2-olefinic
complex rather than an O-bound complex.4 Similarly, the
methylidene carbon of the olefin moiety appears at 67.8 ppm
in the 13C NMR spectrum (δ ) 98.3 ppm in free VA), fully
consistent with an η2-olefin complex.
Upon cooling to -105 °C in CDCl2F, these signals separate
into two pairs (major δ ) 4.27 (br d, 16 Hz), 3.85 (br s); minor
δ ) 4.17 (br s), 3.70 (br d, 14 Hz)), presumably due to static
rotamers of the olefin complex.8 When the temperature is raised
to -20 °C, the olefin inserts cleanly into the Pd-C bond in a
2,1 fashion, resulting in the formation of a chelate complex,
[(N∧N)PdMe(κ2-CH(OAc)Et)][B(Arf)4] (4a). Goddard et al.
have predicted a strong preference (∆∆Gq ) 3.3 kcal/mol) for
2,1 insertion of vinyl acetate over 1,2 insertion.9 This insertion
reaction is very similar to one reported by Liu et al.,10 in which
vinyl acetate was found to insert into a palladium methyl bond
to form an analogous chelate. The hexafluoroantimonate salt
(8) Similar pairs are observed for coordinated propylene, though the rotation
cannot be completely frozen out in this case.
(9) Phillip, D. M.; Muller, R. P.; Goddard, W. A., III.; Storer, J.; McAdon,
M.; Mullins, M. J. Am. Chem. Soc. 2002, 124, 10198.
(10) (a) Reddy, K. R.; Chen, C.-L.; Liu, Y.-H.; Peng, S.-M.; Chen, J.-T.; Liu,
S.-T. Organometallics 1999, 18, 2574. (b) Reddy, K. R.; Surekha, K.; Lee,
G.-H.; Peng, S.-M.; Chen, J.-T.; Liu, S.-T. Organometallics 2001, 20, 1292.
(11) Michalak, A.; Ziegler, T. Organometallics 2001, 20, 1521.
9
J. AM. CHEM. SOC. VOL. 127, NO. 14, 2005 5133

A R T I C L E S
Williams et al.
Figure 1. Eyring Plot of VA Insertion into Pd-Me Bond of 3b From -40 °C to 0 °C.
Scheme 2. Insertion Mechanism for the Nickel VA Complex 3c
10-4 s-1 for the disappearance of 8 (8c plus 8o)) to form two
species, [(N∧N)Ni(κ2-CH(Et)(OAc))][B(Arf)4] (9(2,1)) and [(N∧N)-
Ni(κ2-CH2CH(Me)(OAc))][B(Arf)4] (9(1,2)) in roughly a 2:1 ratio,
resulting from 2,1 and 1,2 insertion, respectively. Since insertion
presumably proceeds from 8c, kobs ) Keqkins, where Keq
represents the 8c:8o equilibrium (Scheme 2) in which the latter
is disfavored by ca. 0.7 kcal/mol (Keq ) 0.20), and thus kins
can be estimated at ca. 3 × 10-3 s-1 (∆Gq ) 15.5 kcal/mol).12
is rate determining or there is a rapid preequilibrium established
between 7 (favored) and the VAf adduct. The reaction proceeds
much more rapidly in the presence of a higher concentration of
VAf, consistent with either mechanism. A value of kobs ) 1.6
× 10-4 M-1 s-1 was obtained for the overall reaction at -40
°C, corresponding to a ∆Gq ) 17.3(1) kcal/mol. This value is
an upper limit to the true insertion barrier.
Synthesis and Characterization of Chelates from Insertion
of Unsaturated Acetates. Addition of VA to [(N∧N)PdMe-
(NCArf)][B(Arf)4] (1) at 0 °C results in quantitative formation
of [(N∧N)PdMe(κ2-CH(OAc)Et)][B(Arf)4] (4a), but this com-
plex is an oily solid and difficult to obtain in crystalline form.
For this reason, we prepared the analogous SbF6- salt, [(N∧N)-
Pd(κ2-CH(OAc)(Et))][SbF6] (4b), by reaction of (N∧N)PdMeCl
with AgSbF6 in methylene chloride at -78 °C, followed by
In contrast, addition of VAf to 7 at -80 °C results in rapid
formation of [(N∧N)Ni(κ2-CH(Et)(OAcf))][B(Arf)4] (10) with
no intermediate being detectable. Either displacement of ether
(12) This rate represents both 2,1 and 1,2 insertion. By the Curtin-Hammett
principle, the ratio of rates is equal to the ratio of the products, and thus
k2,1 ≈ 2 × 10-3s-1 and k1,2 ≈ 1 × 10-3s-1, corresponding to ∆Gq values
of 15.6 and 15.9 kcal/mol, respectively.
9
5134 J. AM. CHEM. SOC. VOL. 127, NO. 14, 2005

Ethylene Copolymerizations Problems Using Vinyl Acetate
A R T I C L E S
products are typical for R-olefins with these catalysts.14 Since
the pairs of isomers 11(2,1)/(1,2) and 12(2,1)/(1,2) are generated
together and have very similar structures, we were generally
unable to separate these complexes, but pure 11(2,1) (R ) nPr)
was obtained by crystallization from a THF/diethyl ether
solution of 11(2,1) and 11(1,2) and Pasteur separation of the orange
plates (11(2,1)) from the yellow microcrystals (11(2,1)/(1,2) mixture).
The most unusual spectral feature of this class of complexes
is the surprisingly high field chemical shift of the proton and
carbon (5.50 and 105.5 ppm, respectively, for 4a in CD2Cl2) R
to the acetate group. The two protons â to the acetate in 4a are
diastereotopic and appear as complex multiplets at 0.87 and 1.15
ppm, while the γ-methyl group appears as a broad triplet at
1.07 ppm. The products of allyl acetate (11(1,2), R ) iPr; 11(2,1)
,
R ) nPr, Scheme 3) and 3-butenyl acetate insertion (12(1,2), R
) iBu; 12(2,1), R ) nBu, Scheme 4) could be identified by 2D
NMR spectroscopy and exhibit spectral features similar to those
of 4. The related nickel complex [(N∧N)NiMe(CH(OAc)Et)]-
[B(Arf)4] (92,1) was prepared and isolated by a method analogous
to that of 4 and is stable at room temperature.
Figure 2. ORTEP of 4b. Selected Bond Lengths (Å): Pd(1)-N(1),
2.061(6); Pd(1)-N(7), 2.148(6); Pd(1)-O(31), 2.035(5); Pd(1)-C(34),
2.001(7); O(33)-C(34), 1.474(9); C(32)-O(33), 1.281(11); O(31)-C(32),
1.260(10). Selected Angles (deg): N(1)-Pd(1)-N(7), 80.19(22); O(31)-
Pd(1)-C(34), 81.4(3); Pd(1)-C(34)-C(36), 114.4(6); Pd(1)-C(34)-O(33),
106.9(6); N(1)-Pd(1)-C(34), 98.2(3); N(7)-Pd(1)-O(31), 100.05(22).
To prepare chelate complexes which might more easily open
to form monodentate species, we investigated the reaction of
isopropenyl acetate and VAf with (N∧N)PdMeCl in the presence
of AgSbF6 (Scheme 4). In the case of isopropenyl acetate,
exclusive 2,1 insertion afforded the complex [(N∧N)Pd(κ2-
C(OAc)(Me)(Et))][SbF6] (13), which contains a tertiary carbon
bound to palladium. Reaction with VAf at -78 °C, followed
by warming to -20 °C generated the expected chelate [(N∧N)-
Pd(κ2-CH(OAcf)Et)][SbF6] (6, Scheme 1), but it was necessary
to crystallize and isolate this complex at low temperature, as it
decomposes at room temperature in solution. The corresponding
nickel complex was not isolable.
addition of VA and gradual warming to room temperature. X-ray
quality crystals of 4b were grown from slow diffusion of pentane
into a dichloromethane solution. An ORTEP view is shown in
Figure 2. Table 2 (Experimental Section) contains important
structural parameters.
The most interesting feature of this structure is that the formal
carbon-oxygen single (C32sO33) and double (C32sO31) bonds
are very similar in length (1.281(11) and 1.260(10) Å, respec-
tively; ∆ ) 0.021(15) Å), more closely resembling the bond
lengths of a carboxylate anion (1.254(1) Å) than those of an
aliphatic ester (1.19 and 1.34 Å, ∆ ) 0.15 Å).13 Liu et al. have
reported the structural characterization of a similar chelate
species supported by a P∧N ligand, o-(diphenylphosphino)-N-
benzaldimine.10 In their structure, the oxygen is trans to the
stronger donor phosphorus, and the PdsO bond is consequently
0.053(9) Å longer, and the CdO bond is 0.05(2) Å shorter than
the corresponding bonds in 4b. The carboxylate CsO single
bond and PdsC bond lengths are effectively identical in the
two structures.
Thermal Stability of Palladium Chelate Complexes. As
indicated, the trifluoroacetato complex [(N∧N)Pd(κ2-CH(OAcf)-
Et)][SbF6] (6) is unstable at room temperature, generating
propylene and a mixture of palladium decomposition products
(Scheme 5). All of the other chelates (4, 11-13) decompose in
a similar fashion at higher temperatures, affording the corre-
sponding olefin (propylene for 4 and 13, butenes for 11, pentenes
for 12).
The initial palladium-containing decomposition products are
dominated by a highly symmetric species with only a single
signal for the aryl methyl groups of the (N∧N) ligand. These
initial products are tentatively assigned as [(N∧N)Pd(κ2-
O2CR)]+ (R ) CH3, CF3); however, as they are never cleanly
formed under the reaction conditions and themselves decompose
to form mixtures of several species, a complete characterization
has not been pursued. We assume that these decompositions
occur via â-acetate elimination reactions (see below). The key
factor for purposes of understanding catalytic activity is the
stability of the starting material, as decomposition of the catalyst
will obviously result in concomitant loss of catalytic activity,
regardless of the precise nature of these decomposition products.
In all cases, the decomposition was cleanly first-order in
palladium starting material. Table 1 lists first-order rate constants
for decomposition of the various chelate complexes and the
corresponding ∆Gq values. With the exception of [(N∧N)Pd-
(κ2-CH(OAc)iPr)][SbF6] (11(1,2)), [(N∧N)Pd(κ2-CH(OAcf)Et)]-
[SbF6] (6), and the nickel complex [(N∧N)Ni(κ2-CH(OAcf)-
Et)][SbF6] (10) (vide infra), all thermolyses were carried out at
While insertion of VA into the Pd-Me bond of 3 proceeds
exclusively in a 2,1 fashion to afford 4, reaction of allyl acetate
with (N∧N)PdMeCl in the presence of AgSbF6 affords a mixture
of [(N∧N)PdMe(κ2-CH(OAc)iPr)][SbF6] (11(1,2)) (resulting from
1,2 insertion) and [(N∧N)PdMe(κ2-CH(OAc)nPr)][SbF6] (11(2,1)
)
(resulting from 2,1 insertion) in a ratio of roughly 1:2 (Scheme
3). After insertion, the palladium undergoes “chain running”
via â-hydride elimination/ reinsertion reactions, to migrate to
the carbon R to acetate whereupon a chelate forms.
Similarly, reaction of (N∧N)PdMeCl with 3-butenyl acetate
in the presence of AgSbF6 results in a ∼1:2 mixture of [(N∧N)-
PdMe(κ2-CH(OAc)iBu)][SbF6] (12(1,2)) from 1,2-insertion and
[(N∧N)PdMe(κ2-CH(OAc)nBu)][[SbF6] (12(2,1)) from 2,1 inser-
tion (Scheme 4). Similar mixtures of 1,2 and 2,1 insertion
(13) Allen, F. H.; Kennard, O.; Watson, D. G.; Brammer, L.; Orpen, A. G.;
Taylor, R. J. Chem. Soc., Perkin Trans. 2 1987, S1-S19 (follows p 1914).
9
J. AM. CHEM. SOC. VOL. 127, NO. 14, 2005 5135

A R T I C L E S
Williams et al.
Scheme 3. Insertion of Allyl Acetate into the Palladium-Methyl Bond
Scheme 4. Insertion of Other Unsaturated Acetates into the Palladium-Methyl Bond
Scheme 5. Decomposition of Acetate Chelates of Nickel and Palladium
60 °C. Since these latter two complexes were not thermolyzed
at the same temperatures, the free energies of activation are not
directly comparable but are certainly informative in a relative
sense. Complexes 4 and 11-13 are all stable for extended
periods at room temperature in solution and are thus stable under
typical polymerization conditions.
Stability of Nickel Chelate Complexes. Reaction of [(N∧N)-
NiMe(OEt2)][B(Arf)4] with vinyl acetate results in both 2,1 and
1,2 insertion to afford the five-membered chelate [(N∧N)Ni-
(CH(Et)(OAc))][B(Arf)4] (9(2,1)) (Scheme 5) and the six-
membered chelate [(N∧N)Ni(CH2CH(Me)(OAc))][B(Arf)4] (9(1,2)),
respectively. Complex 9(1,2) contains an acetate group â to nickel
and reacts at -34 °C to form propylene and paramagnetic nickel
products (kobs ) 2.87 × 10-4 s-1, ∆Gq ) 17.7 kcal/mol), while
9(2,1) decomposes at much higher temperatures (>70 °C) to
afford similarly intractable products. By way of contrast, VAf
inserts exclusively 2,1 to yield [(N∧N)Ni(CH(Et)(OAcf))]-
[B(Arf)4] (10), which decomposes to form propylene and nickel
products at 40 °C (kobs ) 1.58 × 10-4 s-1, ∆Gq ) 23.8 kcal/
mol). The slightly greater stabilities of the nickel complexes
9
5136 J. AM. CHEM. SOC. VOL. 127, NO. 14, 2005

Ethylene Copolymerizations Problems Using Vinyl Acetate
A R T I C L E S
Scheme 6. Reaction of 4 with Acetonitrile and Subsequent Rearrangement
Table 1. Thermolyses of [(N∧N)M(C(R)(R′)(O2CR′′))][X] in
solution at room temperature results in the formation of the three
related carbon monoxide insertion products, [(N∧N)Pd(κ1-C(O)-
CH(OAc)Et)(NCMe)][SbF6] (15R), [(N∧N)Pd(κ1-C(O)CH(Me)-
(CH2OAc))(NCMe)][SbF6] (15â), and [(N∧N)Pd(κ1-C(O)(CH2)3-
OAc)(NCMe)][SbF6] (15γ, Scheme 6).
Equilibrium Studies of Ligand Binding to [(N∧N)Pd-
(CH(OAc)(Et))] (4b). By variation of the temperature (-110
to -70 °C) of solutions of 4b and acetonitrile in CD2Cl2, the
temperature dependence of the equilibrium binding constant,
Keq, was examined. From these data and the use of the Van 't
Hoff equation, values of ∆H ) -2.1(2) kcal/mol and ∆S )
-9(1) eu were obtained for the binding of acetonitrile (Scheme
7).
C2D2Cl4
T
kobs
104 s-
∆
Gq
(kcal/mol)
1
compound
M
R
R
′
R′′
X-
(
°
C)
(×
)
4a
Pd Et
Pd Et
Pd Et
Ni Et
Pd nPr
Pd iPr
Pd nBu
Pd iBu
Pd Et
H
H
H
H
H
H
H
H
Me B(Arf)4
Me SbF6
CF3 SbF6
CF3 B(Arf)4
Me SbF6
Me SbF6
Me SbF6
Me SbF6
60 1.29(2)
60 1.25(3)
10 8.13(3)
40 1.58(2)
60 1.33(3)
120 2.42(6)
60 1.35(2)
60 0.97(1)
60 3.49(3)
25.5
25.5
20.5
23.8
25.5
29.7
25.5
25.7
24.8
4b
6a
10
11(2,1)
11(1,2)
12(2,1)
12(1,2)
13
Me Me SbF6
a In CD2Cl2.
In a similar manner, methylene chloride solutions of 4b were
charged with varying concentrations of ethylene,15 and the
temperature was varied from -105 to -70 °C. By measurement
of the equilibrium between 4b and the ethylene adduct 16, values
of ∆H ) -4.1(2) kcal/mol and ∆S ) -23(1) eu were obtained
for the binding of ethylene.16
The 1H NMR spectrum of 16 at -40 °C is similar to that of
14R, the product which results from reaction with acetonitrile,
but the resonances are broader. The methine signal of the carbon
chain appears at 4.12 ppm, one of the two diastereotopic
methylene signals appears at 1.63 ppm, and the methyl group
at 0.33 ppm. The other methylene proton is obscured. The bound
relative to the analogous palladium species are likely due to
the stronger chelation to the more oxophilic nickel center.
Reaction of Chelates with Acetonitrile and Ethylene. The
reactions of [(N∧N)Pd(κ2- CH(OAc)Et)][SbF6] (4b) with mono-
dentate ligands were investigated to probe the viability of
forming open chelates of the type [(N∧N)Pd(κ1-CH(OAc)Et)L]-
[SbF6]. When a sample of 4b was dissolved in CD3CN at room
temperature and the solution cooled to -30 °C, three products
were observed, the dominant of which (ca. 93%) displayed broad
resonances for the xylyl and acenaphthyl signals and a broad
doublet at 4.35 ppm for the proton R to acetate. The dia-
stereotopic methylene protons â to acetate appear as complex
multiplets at 1.2 and 1.5 ppm, while the terminal methyl group
γ to acetate appears as a triplet at 0.72 ppm. This species,
[(N∧N)Pd(κ1-CH(OAc)Et)(NCMe)][SbF6] (14R, Scheme 6), was
also generated by addition of acetonitrile to 4b in dichloro-
methane-d2 at -78 °C. 1H COSY data allowed the assignment
of the other two products which result from migration of
palladium along the carbon chain by â-hydride elimination
followed by reinsertion: [(N∧N)Pd(κ1-CH(Me)(CH2OAc))-
(NCMe)][SbF6] (14â) and [(N∧N)Pd((CH2)3OAc)(NCMe)]-
[SbF6] ((14γ) Scheme 6). Addition of carbon monoxide to this
(14) McCord, E. F.; McLain, S. J.; Nelson, L. T. J.; Arthur, S. D.; Coughlin E.
B.; Ittel, S. D.; Johnson, L. K.; Tempel, D.; Killian, C. M.; Brookhart, M.
Macromolecules 2001, 34, 362.
(15) Because ethylene distribution between the headspace and solution varied
with temperature, ethylene concentration at a given temperature was
determined by NMR integration.
(16) Values given for 1 M C2H4 as the standard state. It is surprising that such
a large difference is observed for the ∆S of acetonitrile vs ethylene binding
(-9 eu vs -23 eu). Several factors could conceivably come into play here.
Since acetonitrile binds in a linear fashion it loses fewer degrees of freedom
upon binding than ethylene which binds in an eta-2 fashion. In addition,
ethylene may more extensively restrict the motion of the CH(Et)OAc moiety
than CH3CN would. Unbound CH3CN may organize the solvent more
extensively than free ethylene.
9
J. AM. CHEM. SOC. VOL. 127, NO. 14, 2005 5137

A R T I C L E S
Williams et al.
Scheme 7. Van’t Hoff Studies of Ethylene and Acetonitrile
ethylene protons appear as very broad signals at 4.2, 4.3, and
two overlapping signals at 4.9 ppm.
The trifluoroacetate chelate [(N∧N)Pd(κ2-CH(OAcf)(Et)]-
[SbF6] (6) is much more easily opened. Addition of even small
9
5138 J. AM. CHEM. SOC. VOL. 127, NO. 14, 2005

Ethylene Copolymerizations Problems Using Vinyl Acetate
A R T I C L E S
Scheme 8. Reactions of Trifluoroacetate Chelate (6) with Acetonitrile and Ethylene
Scheme 9. Relative Binding of Olefins to Palladium Diimine Complexes
amounts (6.5 equiv, 200 mM) of acetonitrile at -78 °C results
in quantitative formation of the open species [(N∧N)Pd(κ1-
CH(OAcf)(Et))(NCMe)][SbF6] (17) (Scheme 8).
polyethylene and the polyethylene showed no sign of comono-
mer incorporation. When 4a was exposed to ethylene in an NMR
tube, slow consumption of ethylene was observed, but no
detectable decrease in the concentration of 4a took place,
suggesting that all catalytic activity proceeded from a tiny mole
fraction of the catalyst and thus that initiation is far slower than
propagation. Since the chelate formed from insertion of VAf,
[(N∧N)Pd(OAcf)Et)] (6), is more easily opened by ethylene,
we also attempted copolymerization reactions of ethylene (400
psig) and VAf (0.43 M, 5 vol %) using [(N∧N)PdMe(NCArf)]-
[B(Arf)4] (1) as a precursor in dichloromethane at room
temperature. While an appreciable amount of polyethylene was
produced (11.4 g polyethylene from 0.1 mmol catalyst), it
contained no detectable trifluoroacetate functionalities.
The ethylene adduct [(N∧N)Pd(κ1-CH(OAcf)(Et))(C2H4)]-
[SbF6] (18) is formed in the presence of 2 equiv of ethylene at
-25 °C. Both complexes have NMR spectra much like their
acetate analogues, 14a and 16, with the protons R to trifluoro-
acetate appearing at 4.07 and 4.28 for 17 and 18, respectively.
At -40 °C, both NMR spectra are sharp, with the ethylene
signal for 18 forming a distinctive AA′BB′ pattern at 4.71 ppm.
The ethylene complex 18 is remarkably stable at -25 °C. At
0 °C, the complex very slowly begins to insert ethylene (k ≈
5(4) × 10-5 s-1, ∆Gq ) 21.5(5) kcal/mol from initial rates)
and the growing chain is observable in the 1H NMR spectrum.
The measured insertion rate of 18 is approximate, since the first
insertion (“initiation”) is so much slower than subsequent
insertions. Once a small percentage of the palladium centers
has been so activated, they consume ethylene very rapidly
relative to the disappearance of the starting material, 18.
Maintaining an adequate supply of ethylene in the sample
becomes impractical after very early reaction times, and
measurement of its rate of disappearance is difficult. However,
this very discrepancy in rates in a complex in which the chelate
is easily opened clearly demonstrates the inhibitory effect of
the OAcf group on insertion.17
Relative Binding Studies of Olefins. In this and previous
studies3,4 we have demonstrated that olefin binding and exchange
are rapid relative to insertion, and thus the copolymerizations
are governed by the Curtin-Hammett principle: the relative
incorporation of two olefins (A,B) into the polymer is equal to
KA/BkA/kB, where KA/B is the ratio of binding constants of A
and B, while kA and kB are the two olefin insertion rates.
Generation of the palladium olefin complexes at low temperature
in situ from (N∧N)PdMeI, AgSbF6, and olefin allows the
measurement of the relative binding affinities of two olefins, L
and L′ (Scheme 9). While the relative binding constants of
ethylene and VA are too different to measure directly, pairwise
equilibrium studies with ethylene and p-vinyl anisole as well
as p-vinyl anisole and VA allowed us to determine an equilib-
rium constant KVA/Eth ) KVA/VAnisKVAnis/Eth ) (0.078)(0.011) )
(8.6 × 10-4) at -95 °C (Table 2). This corresponds to a ∆G°
of 2.5 kcal/mol and a KVA/Eth of 0.015 at 25 °C (assuming a
∆S° of ca. 0). A similar procedure determined a relative binding
constant between VA and VAf of KVAf /VA ) ca. 4.9 × 10-3 at
-30 °C and hence a ∆G° of 2.6 kcal/mol. Thus the overall
∆G° between ethylene and VAf is 5.1 kcal/mol, corresponding
to a KVAf /Eth ) 1.8 × 10-4 at 25 °C.
Polymerization Studies. As reported previously,4 MA reacts
with the catalyst precursor [(N∧N)PdMe(NCArf)][B(Arf)4] (1)
to form the isolable complex [(N∧N)Pd(κ2-CH2)3CO2Me]-
[B(Arf)4] (19). This species can be used as a stable, easily
handled precursor for the polymerization of ethylene or for the
copolymerization of ethylene and MA. Given the similarities
between the chelate formed by MA and that formed with vinyl
acetate, we have attempted to use the latter complex [(N∧N)-
Pd(κ2-CH(OAc)Et)][B(Arf)4] (4a) as such a catalyst precursor.
However, at 25 °C and 200 psig of ethylene, over the course of
ca. 12 h, 11 µmol of catalyst produced only 81 mg of
9
J. AM. CHEM. SOC. VOL. 127, NO. 14, 2005 5139

A R T I C L E S
Williams et al.
Table 2. Structural Parameters for the X-ray Diffraction Crystal
Structure of [(N∧N)Pd(CH(Et)(OAc))][SbF6] (4b)
problems, some VA could be incorporated using this palladium
catalyst, particularly if high VA/ethylene ratios were employed.
While we have not carried out a full Eyring study of VAf
insertion, the ∆Gq at -40 °C was determined to be 17.4(1) kcal/
mol (kobs ) 2.43(1) × 10-4s-1). The barrier for ethylene
insertion at this temperature is estimated to be 18.4 kcal/mol,
corresponding to a ∆∆Gq of 1.0 kcal/mol. The binding energy
of VAf is ca. 5.1 kcal less than that of ethylene at -30 °C.
Assuming the differences in these values are temperature
independent, and using the same procedure as above for VA,
one can predict a degree of incorporation of VAf of ca. 0.1% at
equal ethylene and VAf concentrations at 25 °C. This extremely
low level of expected incorporation probably explains the
observation that ethylene, in the presence of VAf, is polymerized
at a rate comparable to the system in the absence of ethylene:
VAf is virtually never incorporated due to its extremely
unfavorable binding affinity and thus will have little or no effect
on the polymerization. It should be stressed that because these
incorporation estimates rely on the assumption that ∆S of
binding and ∆Sq of insertion are roughly the same for all three
olefins, these percentages should be considered approximate.
The rate of VA insertion into the Pd-Me bond of 2 is
surprisingly slow (∆Gq ) 19.3 kcal/mol at 25 °C for VA)
relative to ethylene (∆Gq ) 18.3 kcal/mol, 25 °C). The general
trend which has been observed for nickel and palladium
complexes has been that electron-withdrawing groups, such as
acetate (σp ) 0.31), accelerate insertion;4,6a,18,19 however, in this
case, vinyl acetate inserts more slowly than ethylene. Hammett
studies have been carried out using para-substituted styrenes
with both early20 and late18 transition metals, which have shown
that for 2,1 insertion21 there is an excellent correlation between
σp and the ∆Gq values for insertion, with lower barriers for more
electron-withdrawing substituents. Svensson et al. have calcu-
lated the insertion barriers for a variety of R-substituted olefins.22
These calculations suggested that while, in general, more
electron-rich olefins inserted more slowly, it was far from an
ideal correlation. This is perhaps unsurprising, since an R-sub-
stituted olefin is analogous to an o-substituted arene, for which
Hammett correlations are notoriously problematic for steric
reasons. Our results also suggest that great caution should be
exercised in drawing such correlations, especially in cases where
data sets are very limited.6a
parameter
4b
empirical formula
formula weight
temperature
C35H37Cl4F6N2O2PdSb
1001.63
173.1 K
wavelength
0.710 73 Å
crystal system, space group
unit cell dimensions
triclinic, P1h (No. 2)
a ) 8.8149(9) Å
b ) 12.7183(13) Å
c ) 18.3693(19) Å
R ) 91.953(2)°
â ) 90.798(2)°
γ ) 108.682(2)°
1949.1 Å3
volume
Z, calculated density
absorption coefficient
F (000)
crystal size
index ranges
2, 1.707 Mg/m3
1.49 mm-1 using SADABS
990.60
0.25 × 0.25 × 0.20 mm3
-10 e h e 9
0 e k e 15
-21 e l e 21
reflections collected/unique/
with Inet > 3.0 σ (Inet)
merging R value on intensities
cell dimensions determined by
refinement method
residuals (significant reflections)
residuals (all reflections)
residuals (included reflections)
18479/6826/5499
0.034
4274 reflections, 5 e 2θ e 50
full refinement on Fo
5435, Rf ) 0.057, Rw 0.076
6826, Rf ) 0.073, Rw 0.084
5435, Rf ) 0.057, Rw 0.076,
GoF 2.5632
maximum shift/σ ratio
0.008
minimum density of last D-map
maximum density of last D-map
secondary extinction coefficient
-1.550e/Å3
1.920e/Å3
0.6221 microns (σ ) 0.0801)
Discussion
There are five fundamental features of these reactions that
determine the feasibility of copolymerization of a comonomer
with ethylene:
(1) the relative binding strengths of ethylene and the
comonomer;
(2) the relative insertion rates of ethylene and the comonomer;
(3) if the inserted comonomer forms a chelate, the ease with
which this chelate can be opened to promote subsequent
insertions;
(4) the rate of subsequent ethylene insertion into the metal
carbon bond following comonomer insertion;
(5) the competition of decomposition with chain propagation.
Features 1 and 2: Relative Binding Constants and
Insertion Barriers. We have determined the relative binding
constants of VA and ethylene to be 8.6 × 10-4 at -95 °C (Table
2). This ratio corresponds to a ∆G° of 2.5 kcal/mol and a KVA/Eth
of 0.015 at 25 °C assuming that ∆G is temperature independent.
For palladium complexes containing this ligand, an Eyring
analysis has not been carried out for ethylene insertion, but
assuming a similar ∆Sq to that for VA complex 8c (3(5) eu)
and using the measured ∆Gq at -20 °C (18.4 kcal/mol), a ∆Gq
of ca. 18.3 kcal/mol can be estimated for ethylene insertion at
25 °C, while that for VA is ca. 19.3 kcal/mol. The estimated
difference between the two insertion barriers at 25 °C is
Features 3-4: Chelate Opening and Subsequent Inser-
tion. Three possible reasons why VA blocks olefin polymeri-
zation are shown in Scheme 10. The first is an unfavorable
equilibrium in the opening of the chelate 4 with ethylene (Keq),
(17) It cannot be determined whether species 18 undergoes insertion or if
insertion occurs from a species generated by metal migration down the
chain which moves the OAcf group to a the beta or gamma position and
generates
a species with a much lower barrier to insertion. From
observations summarized in Scheme 6, this method for insertion from 18
appears feasible, but the beta and gamma species are clearly higher in energy
and the overall effect would still be a much higher apparent barrier of 18.
(18) Rix, F.; Brookhart, M.; White, P. S. J. Am. Chem. Soc. 1996, 118, 2436.
(19) Hansch, C.; Leo, A.; Taft, R. W. Chem. ReV. 1991, 91, 165.
(20) Burger, B. J.; Santarsiero, B. D.; Trimmer, M. S.; Bercaw, J. E. J. Am.
Chem. Soc. 1988, 110, 3134.
(21) The exo-insertion of p-substituted styrenes reported in ref 20 for Cp2Nb-
(H)(styrene) complexes is analogous to 2,1 insertion and rates are
accelerated by electron-withdrawing groups. Endo-insertion, corresponding
to 1,2 insertion, showed an opposite trend. In other studies by Halpern and
Okamato (using rhodium),23 Bercaw and Doherty (niobium),24 and
Wolczanski and Strazisar (tantalum),25 more electron rich olefins inserted
more rapidly. In the first and third cases, the overall rates measured were
a combination of both a binding constant and an insertion rate.
therefore ca. 1.0 kcal/mol, corresponding to a rate ratio kVA
/
k
Eth of ca. 0.3. Under equal concentrations of ethylene and VA,
and assuming relative binding affinities of 0.015 (see above),
this would result in a degree of incorporation of roughly 0.5%.
While this is a low percentage, it does imply that, barring other
(22) von Schenk, H.; Stro¨mberg, S.; Zetterberg, K.; Ludwig, M.; A° kermark,
B.; Svensson, M. Organometallics 2001, 20, 2813.
9
5140 J. AM. CHEM. SOC. VOL. 127, NO. 14, 2005

Ethylene Copolymerizations Problems Using Vinyl Acetate
A R T I C L E S
Scheme 10. Opening the Chelate with and Subsequent Insertion of Ethylene
Scheme 11. Comparison of VA and MA Chelates
the second is a slow rate of insertion of ethylene into the Pd-C
bond of 16 (kins), and the third is a decomposition pathway (kdec).
Since insertion of MA forms a similar chelate species (19, also
shown) without shutting down catalysis, comparison between
the chelate complexes generated by VA and MA is germane to
the discussion.
The ease of chelate opening by ethylene can be examined
by measuring the equilibrium reaction of 4b with ethylene over
a range of temperatures. As described above, values of ∆H )
-4.1(2) kcal/mol and ∆S ) -23(1) eu were obtained for the
reaction of 4 with ethylene, while thermodynamic values of ∆H
) -5.7(1) kcal/mol and ∆S ) -27(1) eu have previously been
reported for the opening of 19 (Scheme 11).4,26
The calculations performed by Goddard et al. predict ∆E
values for these reactions of -1.9 kcal/mol and -3.0 kcal/mol
for model complexes of 4 and 19 (the correct relative stabilities),
and the ∆∆E of 1.1 kcal/mol agrees reasonably well with the
experimental value of ∆∆H ) 1.6(2) kcal/mol. Comparison of
the ∆G values for the two systems at 25 °C shows that opening
of 4 is unfavorable (Keq ) 0.009 M-1) but not much less
favorable than opening of the chelate formed from methyl
acrylate (Keq ) 0.02 M-1), since the difference in ∆∆S partially
offsets ∆∆H.
the open chelate. In the case of MA, the open chelate 20
undergoes insertion with a barrier equal to that of insertion into
a normal alkyl group, ∆Gq ) 18.6 kcal/mol (see Scheme 12).
In contrast, complex 16 bearing an alpha acetoxy group has a
much higher barrier to insertion. Unfortunately we have been
unable to measure this specific barrier, but based on the results
of the VAf study, a barrier of at least 21.5 kcal/mol can be
estimated. This value is in line with calculations by Goddard et
al. that have predicted a ∆Eq of 25.1 kcal/mol for this insertion
barrier.4
Thus, the barrier for ethylene insertion into the Pd-C bond
is approximately 3 kcal/mol greater in the case of 18 than 20.
This is perhaps unsurprising given that the R-carbon of 18 bears
a strongly electron-withdrawing substituent, while that of 20
does not. The same holds true of the acetate analogue, 16. It is
well-known that, in alkyl carbonyl complexes, the more
electronegative carbon fragments have a lower migratory
aptitude.27 Such behavior has also been predicted by theoretical
studies, which have linked this trend to the relative amounts of
s- and p-character in the metal-carbon bond of the migrating
group.28 Jordan et al. have recently reported similar behavior
with early transition metal olefin insertion processes.29
A second feature which strongly disfavors propagation in the
VA case relative to MA is the rate of subsequent insertion from
(27) (a) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. Principles
and Applications of Organotransition Metal Chemistry; University Science
Books: Mill Valley, CA, 1987; pp 370-1. (b) Alexander, J. J. In The
Chemistry of the Metal Carbon Bond; Hartley, F. R., Ed.; Wiley: New
York, 1984; Vol 2, p 259, Table 1. (c) Cotton, J. D.; Crisp, G. T.; Daly, V.
A. Inorg. Chim. Acta 1981, 47, 165. (d) Cotton, J. D.; Crisp, G. T.; Latif,
L. Inorg. Chim. Acta 1981, 47, 171. (e) Cotton, J. D.; Dunstan, P. R. Inorg.
Chim. Acta 1984, 88, 223.
(23) Halpern, J.; Okamoto, T. Inorg. Chim. Acta 1984, 89, L53.
(24) Doherty, N. M.; Bercaw, J. E. J. Am. Chem. Soc. 1985, 107, 2670.
(25) Strazisar, S. A.; Wolczanski, P. T. J. Am. Chem. Soc. 2001, 123, 4728
(26) It should be noted that 19 contains 2,6-diisopropylphenyl groups on the
diimine, while the diimine aryls of 4 are 2,6-dimethylphenyl. Typically,
differences between rates and equilibria of palladium complexes of varying
diimine ligands are small, being on the order of a kcal/mol or less.
(28) Axe, F. U.; Marynick, D. S. J. Am. Chem. Soc. 1988, 110, 3728.
(29) Stockland, R. A.; Foley, S. R.; Jordan, R. F. J. Am. Chem. Soc. 2003, 125,
796.
9
J. AM. CHEM. SOC. VOL. 127, NO. 14, 2005 5141

A R T I C L E S
Williams et al.
Scheme 12. Insertion of Ethylene into Various Palladium-Carbon Bonds
Scheme 13. A High-Energy Decomposition Pathway for 11(1,2)
In the case of copolymerization of ethylene and MA, chelate
formation substantially retards the rate of chain growth relative
to ethylene homopolymerization, and this effect can be attributed
solely to an unfavorable equilibrium between the chelate and
the ethylene adduct necessary for chain growth. In the case of
VA, chelate opening by ethylene is slightly less favorable, and
in addition, the subsequent insertion barrier is ca. 3 kcal/mol
greater, implying that VA may retard the polymerization rate
(relative to ethylene homopolymerization) by another factor of
ca. 100. This large rate reduction may well be responsible for
nearly complete deactivation of the catalyst by VA.
Does the VA chelate in the presence of ethylene decompose
competitively with ethylene insertion? Our data cannot answer
this question definitively. The barrier for ethylene insertion in
opened chelate 16 is almost certainly lower (21-22 kcal/mol)
than the barrier to decomposition of chelate 4 (25.5 kcal/mol).
However, under low ethylene pressures the equilibrium lies
strongly in favor of 4, which may result in â-acetate elimination
competitive with chelate opening and insertion. At high ethylene
pressures, the equilibrium can be shifted in favor of open 16,
but the rate of decomposition of 16 via â-acetate elimination is
unknown.
Feature 5: Catalyst Stability. Catalyst instability may play
a role in deactivation by VA and therefore the decomposition
of the chelates 4 and 11-13 was examined (Table 1). The
mechanism of decomposition most likely involves â-hydride
elimination and reinsertion, which places the palladium center
â to the carboxylate moiety (Scheme 5), allowing a formal
â-carboxylate elimination reaction to occur. Such a pathway is
well-established by literature precedent30 and has also been
predicted by Goddard and co-workers utilizing DFT calcula-
tions.9 The difference between the decomposition barrier for
the iPr-substituted chelate 11(1,2) (Table 1, above) and that of
the other chelates is striking. As shown in Scheme 13,
rearrangement in this complex to place palladium â to acetate
situates the palladium at a sterically hindered tertiary carbon
center, thereby accounting for the fact that this intermediate is
formed much more slowly than in the other complexes.
As noted above, when vinyl acetate inserts into the Ni-Me
bond, both 2,1 and 1,2 insertion products are observed. The
1,2 insertion forms a six-membered chelate analogous to the
proposed intermediate in the palladium thermolysis mechanism.
This species is considerably less stable than the five-membered
2,1 insertion product, decomposing at -30° rather than 60 °C
(Scheme 5). This observation strongly suggests a â rather than
R elimination mechanism.
Conclusions
These studies establish quantitative aspects of the reactions
of vinyl acetate(VA) and vinyl trifluoroacetate (VAf) with
[(N∧N)M(CH3)]+ (M ) Pd, Ni, N∧N ) N,N′-1,2-acenaph-
thylenediylidene bis(2,6-dimethyl aniline) and subsequent reac-
tions with ethylene following insertion of these monomers.
Major conclusions from these studies, cast in qualitative terms,
can be summarized as follows:
(1) Binding affinity studies show that, relative to ethylene,
VA binds weakly and VAf binds exceedingly weakly to the
[(N∧N)PdCH3]+ moiety.
(2) VA and VAf insertion into the M-CH3 bond of [(N∧N)-
MCH3]+ results in the formation of chelate complexes. For Pd,
insertion occurs in a 2,1 fashion and yields a five-membered
chelate in which the acetoxy group resides at CR. For Ni,
insertion occurs in both a 2,1 and 1,2 fashion to give a mixture
of five- and six-membered ring chelates.
(3) Thermolysis of the Pd chelates results in â-acetate and
â-trifluoroacetate elimination. Trifluoroacetate eliminates at a
rate considerably greater than that of acetate. The five-membered
Ni acetate chelate is quite stable at room temperature, while
the six-membered chelate from 1,2 insertion of vinyl acetate
eliminates readily at -35 °C.
9
5142 J. AM. CHEM. SOC. VOL. 127, NO. 14, 2005

Ethylene Copolymerizations Problems Using Vinyl Acetate
A R T I C L E S
All other reagents unless otherwise noted were purchased from
commercial sources and used without further purification. n-Butyl
isocyanide, vinyl acetate, and VAf were vacuum transferred prior to
use and dried over 4 Å sieves.
(4) The five-membered chelate formed from VAf insertion
is readily opened by ethylene to form the ethylene adduct
[(N∧N)Pd(κ1-CH(Et)OAcf)(C2H4)]+. In the case of the chelate
from VA insertion, the equilibrium strongly favors the chelate
over the open form [(N∧N)Pd(κ1-CH(Et)OAc)(C2H4)]+.
(5) The insertion barrier of ethylene in the opened form of
the VA and VAf chelates is significantly higher (ca. 3 kcal/
mol) than the barrier of insertion into a simple [(N∧N)Pd-
(alkyl)]+ bond due to the presence of the electron-withdrawing
acetoxy (or trifluoroacetoxy) group at CR. The rate ratio is
estimated to be ca. 1:100 at 25 °C.
(6) Homopolyethylene is generated at good rates when
copolymerization of ethylene and VAf is attempted. This arises
simply because VAf is not incorporated due to the very weak
binding of VAf to the Pd center. VA efficiently quenches catalyst
activity in attempts to copolymerize VA and ethylene. This is
likely due to incorporation of VA to form the chelate which,
due to thermodynamically unfavorable opening with ethylene
and a high barrier to insertion of the low equilibrium concentra-
tion of the opened form, greatly retards chain propagation. Some
catalyst decay through â-acetate elimination cannot be ruled
out.
Kinetic Experiments. Kinetics were performed in all cases in an
NMR probe on a DRX 500 spectrometer maintained at the temperature
in question. Reactions were monitored unless otherwise noted by
integration of a peak of the starting material. A delay of 30 s was used
between pulses for reactions with half-lives greater than 30 min, 10 s
for faster reactions. Reactions were generally followed for 2.5-3.5 half-
lives, with the exception of the two lowest temperatures in the Eyring
study, which were monitored for 0.8 and 0.6 half-lives. Low-temperature
experiments were performed in an oven-dried NMR tube which was
charged with solid reagents in a glovebox and capped with a septum.
Subsequently, the tube was cooled in a -78 °C bath, and solvent and
liquid reagents were added slowly via syringe. High-temperature
experiments were carried out in a J. Young tube, charged with solids
and solvent in the drybox.
Preparation of (N∧N)PdMeI. A Schlenk flask was charged with
(N∧N)PdMeCl (169.2 mg, 0.310 mmol) and sodium iodide (53.6 mg,
0.358 mmol) in the drybox. Dichloromethane (12 mL) and acetone
(15 mL) were added, and the reaction was allowed to stir for several
minutes. The solvent was removed under vacuum, and the solids were
taken up in dry methylene chloride (15 mL) and cannula filtered into
a flame-dried Schlenk flask. The volume was reduced to 5 mL under
vacuum, and methanol (15 mL) was added, with concomitant formation
of precipitate. The solvent was removed under reduced pressure to give
These studies reveal the pitfalls of copolymerization of VA
with ethylene and point to the need to redesign catalysts if such
copolymerizations are to be successful.
1
an iridescent, burgundy crystalline material (158 mg, 80% yield). H
Experimental Section
NMR (400 MHz CDCl3): 8.11, 8.06 (both d, 8 Hz, 1H each, H3 and
H8 of acenaphthyl group), 7.49 (apparent q, 2H, H4 and H7 of
acenaphthyl group), 7.2-7.3 (m, 6H, N-Aryl protons), 6.75, 6.64 (both
d, 8 Hz, 1H each, H4 and H4 of acenaphthyl group), 2.33, 2.33 (6H
General Considerations. All manipulations of air and/or water
sensitive materials were performed under an argon atmosphere using
standard Schlenk or glovebox techniques. Argon was dried by passage
through columns of BASF catalyst (Chemalog) and 4 Å molecular
sieves. Complexes 4 and 6-11 were stored in the glovebox freezer at
-30° to -35 °C, as was [H(OEt2)2][B(Arf)4]. NMR spectra were
obtained using Bruker DRX400 and DRX500 spectrometers. NMR
probe temperatures were obtained by use of a calibrated Omega type
T thermocouple immersed in a 5 mm NMR tube containing anhydrous
methanol and calibrated with baths of known temperatures. Elemental
analyses were performed by Atlantic Microlab, Inc., Norcross GA.
Values for ∆Gq, ∆Hq, and ∆Sq were obtained from first-order reaction
rate data and the Eyring equation. Values for ∆G, ∆H, and ∆S were
obtained from equilibrium constant measurements and the Van 't Hoff
equation. All error estimates were calculated by standard regression
1
each, N-Ar methyl groups), 0.67 (3H, Pd-Me). H NMR (CDCl3):
δ ) 0.60 (s, 3H, Pd-Me), 2.28, 2.29 (s, 12H total, Ar-Me), 6.57,
6.68 (both d, 7.2 Hz, H3 and H8 of acenaphthyl backbone), 7.2-7.3
(6H, m, Ar), 7.45 (2H, m, H4 and H7 of acenaphthyl backbone), 8.05,
8.10 (both d, 8.4 Hz, 1H, H5 and H6 of acenaphthyl backbone). 13C
NMR (CDCl3): δ ) -3.9 (Pd-Me), 18,3, 19.1 (2C each, ArMe),
123.9, 123.9, 126.4, 126.5, 127.3, 127.5, 128.3, 128.8, 128.9, 129.2,
130.8, 131.3, 131.3, 142.8, 144.1, 144.8, 167.7, 171.1. Anal. Calcd
For C29H27IN2Pd: C, 54.69; H, 4.27; N, 4.40. Found: C, 54.18; H,
4.26; N, 4.35.
Preparation of [(N∧N)PdMe(OEt2)][SbF6] (2b). A flame-dried
Schlenk flask was charged with (N∧N)PdMeI (198.0 mg, 0.310 mmol)
and AgSbF6 (108.1 mg, 0.315 mmol) and cooled to -78 °C. Ether
(1 mL) and dichloromethane (10 mL) were added slowly. Immediately,
a flocculent precipitate (AgI) formed. The orange solution was separated
from the precipitate by cannula filtration at -78 °C. Ether (5 mL) was
added, and the solvent was removed in vacuo at -10 °C to form an
orange, tacky oil. Ether (10 mL) was added, and the product was scraped
from the sides and stirred vigorously for ca. 1 h to form a fine, canary
yellow powder. The ether was removed by cannula filtration, and the
solid dried under vacuum at -40 °C for ca. 2 h (195.2 mg, 77% yield).
1H NMR (500 MHz, CD2Cl2): 8.21, 8.18 (both d, 8.4 Hz, 1H each,
H3 and H8 of acenaphthyl group), 7.50 (m, 2H, H4 and H7 of
acenaphthyl group), 7.3-7.4 (m, 6H, N-Aryl protons), 6.85, 6.78 (both
d, 7.2 Hz, 1H each, H5 and H6 of acenaphthyl group), 3.42 (q, 6.8 Hz,
4H, CH3CH2O), 2.33, 2.33 (6H each, N-Ar methyl groups), 1.47 (t, J
) 7.0 Hz, 6H, CH3CH2O), 0.64 (3H, Pd-Me). 13C NMR (125 MHz,
CD2Cl2): δ ) 9.1 (Pd-Me), 16.1 (CH3CH2O, 2C), 17.7 (4C, ArMe),
72.1, (2C, CH3CH2O), 124.8, 124.9, 125.1, 125.2, 127.1 127.7, 128.4,
-
and error propagation techniques. Signals for the B(Arf)4 counterion
are not provided below, as they are invariant from complex to complex
and have been previously reported.2
Materials. All solvents were deoxygenated and dried via passage
through a column of activated alumina.31 Dichlorofluoromethane-d was
prepared following the procedure of Siegel and Anet32 and was distilled
at ambient temperature and pressure to remove traces of chloroform-
d, though it remained a mixture of CDCl2F and CDClF2. Dichloro-
methane-d2 and 1,1,2,2-tetrachloroethane-d2 were dried over P2O5. The
R-diimine ligands33 (COD)PdMeCl,34 (COD)PdMe2,35 (N∧N)PdMeCl,4
[(N∧N)PdMe(OEt2)][B(Arf)4],2 [(N∧N)NiMe(OEt2)][B(Arf)4],36 and
[H(OEt2)2][B(Arf)4]37 were prepared according to literature methods.
(30) (a) Cheng, J. C.-Y.; Daves, G. D., Jr. Organometallics 1986, 5, 1753. (b)
Zhu, G.; Lu X. Organometallics 1995, 14, 4899. (c) Zhang, Z.; Lu, X.;
Xu, Z.; Zhang, Q.; Han, X. Organometallics 2001, 20, 3724.
(31) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.; Timmers,
F. J. Organometallics 1996, 15, 1518-1520.
(32) Siegel, J. S.; Anet, F. A. L. J. Org. Chem. 1988, 53, 2629-2630.
(33) van Koten, G.; Vrieze, K. AdV. Organomet. Chem. 1982, 21, 151-239.
(34) Ru¨lke, R. E.; Ernsting, J. M.; Spek, A. L.; Elsevier, C. J.; van Leeuwen, P.
W. N. M.; Vrieze, K. Inorg. Chem. 1993, 32, 5769.
(36) Svejda, S. A.; Johnson, L. K.; Brookhart, M. J. Am. Chem. Soc. 1999,
(35) Prepared from (COD)PdMeCl34 and MeLi as per the procedure reported
for the synthesis of the analogous Pd complex with 2,6-diisopropyl groups
attached to the nitrogen ligand in ref 2.
121, 10634-10635.
(37) Brookhart, M.; Grant, B.; Volpe, A. F., Jr. Organometallics 1992, 11, 3920-
3922.
9
J. AM. CHEM. SOC. VOL. 127, NO. 14, 2005 5143

A R T I C L E S
Williams et al.
128.6, 129.3, 129.3, 129.5, 130.9, 132.4, 133.1, 142.4, 143.5, 145.4,
168.9, 175.0. The complex is insufficiently stable for elemental analysis.
X-ray Crystal Structure of [(N∧N)Pd(κ2-CH(Et)(OAc))][SbF6]
(4b). X-ray quality crystals of 4b were grown from slow diffusion of
pentane into a solution of methylene chloride, which contained two
methylene chloride molecules in the unit cell. A single crystal was
mounted on a Bruker SMART 1K diffractometer at -100 °C, using
the ω-scan mode. Selected bond lengths and angles are reported in
Figure 2 (Results), while structural parameters are given in Table 2.
There are two methylene chloride molecules of crystallization.
In Situ Formation of Cationic Nickel Olefin Complexes. A
vacuum-dried Wilmad 528-PP screw-cap NMR tube was charged with
7 (13.9 mg, 0.01 mmol) in the drybox and cooled to -80 °C, and
CD2Cl2 (0.70 mL) was added slowly via gastight syringe. The tube
was briefly shaken to dissolve the sample. A solution of vinyl acetate
or VAf in CD2Cl2 (0.10 mL of a 0.20 M solution, ca. 2 equiv) was
added to the tube at -80 °C, and the tube was transferred to the
precooled NMR probe. Displacement of the ether ligand by the acetate
moiety and insertion to form the corresponding chelate species were
In Situ Generation of [(N∧N)PdMe(η2-C2H3OAc)][SbF6] (3b). An
NMR tube was charged in the glovebox with (N∧N)PdMeI (25.3 mg,
39.7 µmol) and AgSbF6 (13.6 mg, 39.6 µmol) and was capped with a
septum. Outside the box, the tube was cooled to -78 °C and CD2Cl2
(650 µL) and VA (5 µL, 50 µmol, 1.4 equiv) were slowly added via
syringe. The tube was vigorously shaken and centrifuged in short
1
intervals so that the temperature could be kept as low as possible. H
(400 MHz, CD2Cl2, -37 °C): δ 0.46 (s, 3H, Pd-Me), 2.16 (s, 3H,
O2CMe), 2.22, 2.24, 2.27, 2.39 (all s, 3H each, NAr-Me), 4.04, 4.17
(br d, 4 Hz and 12 Hz, 1H each, methylidene protons), 6.65, 6.74 (d,
7.2 Hz each, 1H each, H5 and H6 of acenaphthyl group), 7.2 (br,
unresolved from free vinyl acetate, methine proton of vinyl acetate),
7.2-7.40 (m, 6H, N-Ar), 7.57 (two overlapping dd, J ≈ 2 Hz, 5-10
Hz, H4 and H7 of acenaphthyl group), 8.22, 8.25 (d, 8.4 Hz, 8.0 Hz,
H3 and H8 of acenaphthyl group). 13C NMR (100 MHz, CD2Cl2, -37
°C): δ ) 17.1 (br, Pd-CH3), 17.6, 17.8, 17.9, 17.9 (Ar-CH3), 20.5
(CO2CH3), 66.1 (br, methylidene, vinyl acetate moiety), 124.5 (br,
methine carbon, vinyl acetate moiety), 124.4, 124.5, 125.2, 126.1, 126.9,
127.1, 128.1, 128.3, 128.4, 128.5, 129.3, 129.3, 129.5, 129.5, 129.5,
129.7, 130.8, 133.0, 133.9, 141.1, 146.3, 171.0, 175.9, 186.7 (CdO).
The complex is insufficiently stable for elemental analysis, but converts
to 4b (vide infra).
Preparation of [(N∧N)Pd(κ2-CH(Et)(OAc))][B(Arf)4] (4a). A
flame-dried Schlenk flask was charged with [(N∧N)PdMe(NCArf)]-
[B(Arf)4] (1a) (665.7 mg, 0.413 mmol) in the drybox and brought out
to the Schlenk line, where it was lowered to 0 °C. Dry dichloromethane
(10 mL) and vinyl acetate (85 µL, 0.92 mmol) were added via syringe.
The reaction was allowed to stir for ca. 2 h while warming to 20 °C.
The product was purified by column chromatography (Et2O, silica) and
crystallization from dichloromethane/pentane at -30 °C. The product
was a mustard microcrystalline powder (140 mg, 23% yield). NMR
spectra identical to those of 4b (vide infra) except for peaks attributable
to B(Arf)4-. Anal. Calcd for C65H45BF24N2O2Pd (4a): C, 53.50; H, 3.11;
N, 1.92. Found: C, 53.42; H, 3.19; N, 1.96.
Preparation of [(N∧N)Pd(κ2-CH(Et)(OAc))][SbF6] (4b). A flame-
dried Schlenk flask was charged with (N∧N)PdMeCl (790.8 mg, 1.45
mmol) and AgSbF6 (507.2 mg, 1.48 mmol) in the drybox. Dry
dichloromethane (20 mL) and vinyl acetate (2 mL, 22 mmol) were
precooled to -78 °C in a second flame-dried Schlenk flask. The
dichloromethane was added to the solids by cannula transfer. The
reaction was allowed to stir for ca. 1.5 h and was then warmed to -20
°C for another hour. The orange suspension was cannula filtered into
a new flame-dried flask to afford a clear orange solution. The solvent
was removed in vacuo, and the product was extracted with THF (35
mL) and cannula filtered a second time to remove a yellow insoluble
powder. The solvent was reduced to ca. 10 mL under vacuum and
layered with 50 mL of ether. The crystallization solution was allowed
to stand at -30 °C for a week. An orange solid was isolated (902.1
mg, 75% yield). 1H NMR (CDCl3, 400 MHz): δ ) 0.87 (m, 1H,
CH3CHH′CH(OAc)Pd), 1.07 (t, 7 Hz, 3H, CH3CH2CH(OAc)Pd), 1.15
(m, 1H, CH3CHH′CH(OAc)Pd), 2.22 (s, 3H, CO2CH3), 2.27 (s, 3H,
Ar-CH3), 2.41 (s, 3H, Ar-CH3), 2.44 (s, 3H, Ar-CH3), 2.57 (s, 3H,
Ar-CH3), 5.50 (dd, 2 Hz, 5 Hz, 1H, CH3CH2CH(OAc)Pd), 6.66, 6.89
(d, 7 Hz each, 1H each, H3 and H8 of acenaphthyl ligand), 7.3-7.5
(m, 6H, N-Ar signals), 7.60, 7.62 (apparent dt, 8.2 Hz, 2H, H4 and
H7 of acenaphthyl ligand), 8.23, 8.27 (apparent dd, 8 Hz, 2H, H4 and
H5 of acenaphthyl ligand). 13C NMR (CD2Cl2, 125 MHz): δ ) 12.4
(PdCH(OAc)CH2CH3), 17.8 (O2CCH3), 18.4, 18.4, 18.4, 18.5 (Ar-CH3),
29.0 (PdCH(OAc)CH2CH3), 105.5 (Pd-CH), 124.5, 124.7, 125.1, 125.3,
127.7, 127.9, 128.1, 128.6, 128.8, 129.1, 129.1, 129.3, 129.4, 129.5,
129.6, 130.2, 131.3, 133.0, 133.8, 142.3, 143.1, 146.2, 169.5, 175.6
(aryl carbons), 186.6 (CdO). Anal. Calcd for C36H43F6N2O3PdSb (4b‚
Et2O, matches NMR integration): C, 48.37; H, 4.85; N, 3.13. Found:
C, 48.43; H, 4.47; N, 3.26.
1
monitored by variable temperature H NMR.
[(N∧N)Ni(κ2-CH(Et)(OAc))][B(Arf)4] (9(2,1)). This complex was also
synthesized on a preparatory scale, by treating a solution of 7 (125
mg, 0.089 mmol) in CH2Cl2 at -80 °C under Ar with a solution of
vinyl acetate in CH2Cl2 (0.9 mL of a 0.20 M solution, 2 equiv). The
reaction was allowed to warm slowly to 0 °C and stirred for 1 h. All
solvent was removed in vacuo, and the dark purple solid was dried in
vacuo for 1 h at rt, recovered in the drybox, and stored at -35 °C.
Yield of 9(2,1)) ) 76 mg (61%).1H NMR (CD2Cl2, 500 MHz) δ 8.17
(d, J ) 8.5 Hz, 1H, An Hp), 8.14 (d, J ) 8.5 Hz, 1H, An Hp′), 7.70 (s,
8H, BAr4′), 7.56 (t, J ) 7.5 Hz, 1H, An Hm), 7.54 (s, 4H, BAr4′), 7.49
(t, J ) 7.5 Hz, 1H, An Hm′), 7.40-7.26 (m, 6H, ArH), 6.85 (d, J ) 7.0
Hz, 1H, An Ho), 6.55 (d, J ) 7.0 Hz, 1H, An Ho′), 4.92 (dd, J ) 5.0,
2.0 Hz, 1H, NiCH), 2.70, 2.56, 2.37, 2.23 (s, 3H each, ArCH3), 1.99
(s, 3H, NiCH(OC(O)CH3)), 1.34 (m, 1H, NiCH(CHH′CH3)), 1.25 (t, J
) 7.0 Hz, 3H, NiCH(CHH′CH3)), 0.68 (m, 1H, NiCH(CHH′CH3)).
13C{1H} NMR (CD2Cl2, 25 °C, 125 MHz) δ 187.8 (OC(O)CH3), 174.9
and 169.4 (NdCsCdN), 162.1 (q, 1JCB ) 49.6 Hz, BAr′4 Cipso), 147.6,
142.6 and 142.1 (ligand Ar and Ar′ Cipso), 135.2 (BAr′4 Co), 133.4,
133.0, 131.8, 130.7, 130.5, 130.0, 129.9, 129.73, 129.70, 129.52, 129.46,
129.42, 129.2 (2JCF ) 28.5 Hz, BAr′4 Cm), 129.16, 129.0, 128.2, 125.7,
125.5, 125.1, 124.9 (1JCF ) 270.9 Hz, BAr′4 CF3), 124.6, 117.8 (BAr′4
Cp), 28.4 (NiCH(CHH′CH3), 18.7 and 18.5 (Ar CH3 and C′H3), 18.14
(Ar C′′H3), 18.11 (NiCH), 17.8 (OC(O)CH3), 17.7 (Ar C′′′H3), 12.5
(NiCH(CHH′CH3). Anal. Calcd for C65H45BN2O2F24Ni: C, 55.31; H,
3.21; N, 1.98. Found: C, 54.19; H, 3.32; N, 1.93.
[(N∧N)Ni(κ2-CH2CH(CH3)(OC(O)CH3))][BAr4′] (9(1,2)). 1H NMR
(CD2Cl2, -50 °C, 500 MHz) δ 8.11 (apparent t (overlapping d), 2H,
An Hp and Hp′), 7.70 (s, 8H, BAr4′), 7.51 (s, 4H, BAr4′), 7.45 (t, J )
7.5 Hz, 1H, An Hm), 7.33 (t, J ) 7.5 Hz, 1H, An Hm′), 7.31-7.22 (m,
6H, ArH), 6.88 (d, J ) 7.5 Hz, 1H, An Ho), 6.37 (d, J ) 7.5 Hz, 1H,
An Ho′), 3.57 (br m, 1H, NiCHH′CH), 2.42, 2.41, 2.38, 2.26 (s, 3H
each, ArCH3), 1.79 (s, 3H, NiCHH′CH(OC(O)CH3)), 1.21 (d, 3H, J )
7.0 Hz, 1H, NiCHH′CH(CH3)), 1.03 (dd, J ) 11.0, 3.0 Hz, 1H,
NiCHH′), 0.82 (dd, J ) 11, 5 Hz, 1H, NiCHH′). The complex is
insufficiently stable for elemental analysis.
1
[(N∧N)Ni(κ2-CH(Et)(OAcf)] [BAr4′] (10). H NMR (CD2Cl2, 20
°C, 500 MHz) δ 8.20 (d, J ) 8.5 Hz, 1H, An Hp), 8.18 (d, J ) 8.5 Hz,
1H, An Hp′), 7.71 (s, 8H, BAr4′), 7.59 (t, J ) 8.0 Hz, 1H, An Hm),
7.54 (s, 4H, BAr4′), 7.53 (t, J ) 7.5 Hz, 1H, An Hm′), 7.45-7.23 (m,
6H, ArH), 6.91 (d, J ) 7.5 Hz, 1H, An Ho), 6.58 (d, J ) 7.0 Hz, 1H,
An Ho′), 5.38 (dd, J ) 5.0, 2.5 Hz, 1H, NiCH), 2.73, 2.58, 2.39, 2.28
(s, 3H each, ArCH3), 1.43 (m, 1H, NiCH(CHH′CH3)), 1.31 (t, J ) 7.0
Hz, 3H, NiCH(CHH′CH3)), 0.81 (m, 1H, NiCH(CHH′CH3)). The
complex is insufficiently stable for elemental analysis.
In Situ Generation of [(N∧N)PdMe(η2-C2H3OAcf)][SbF6] (5): A
procedure identical to that used for the acetate analogue, 3b , was used
1
to generate 5. H NMR (400 MHz, CD2Cl2, -25 °C): δ 0.47 (s, 3H,
Pd-Me), 2.22, 2.25, 2.25, 2.37 (all s, 3H each, NAr-Me), 4.29, 4.66
9
5144 J. AM. CHEM. SOC. VOL. 127, NO. 14, 2005

Ethylene Copolymerizations Problems Using Vinyl Acetate
A R T I C L E S
(both br, coupling indistinct, 1H each, methylidene protons), 6.63, 6.74
(d, 7.2 Hz each, 1H each, H5 and H6 of acenaphthyl group), 6.9 (br,
unresolved from free vinyl trifluoroacetate, methine proton of vinyl
trifluroacetate) 7.26-7.46 (m, 6H, N-Ar), 7.58 (two overlapping
signals, coupling unresolved, H4 and H7 of acenaphthyl group), 8.23,
8.27 (d, 8.4 Hz each, H3 and H8 of acenaphthyl group). The complex
is insufficiently stable for elemental analysis but converts to 6 (vide
infra).
7.55 (obscured by 11(2,1), identified by COSY and HMBC, N-Ar), 7.54
(obscured, seen by COSY, 1H each, H4 and H7 of acenaphthyl ring),
8.1 (obscured, seen by COSY, 2H, H5 and H6 of acenaphthyl ring).
Anal. Calcd for C34H35N2O2PdSbF6 (11(2,1)): C, 48.28; H, 4.17; N, 3.31.
Found: C, 48.59; H, 4.42; N, 3.19. Anal. Calcd for C34H35N2O2PdSbF6
(11(1,2) and 11(2,1)): C, 48.28; H, 4.17; N, 3.31. Found: C, 48.68; H,
4.53; N, 3.13.
Preparation of [(N∧N)Pd(κ2-CH(R)(OAc))] [SbF6] R ) iBu, nBu,
(12(2,1), 12(1,2)). A flame-dried Schlenk flask was charged with
(N∧N)PdMeCl (220.0 mg, 0.403 mmol) and AgSbF6 (137.7 mg, 0.401
mmol) in the drybox. The flask was brought out to the Schlenk line
and cooled in a -78 °C bath. Dry ether (10 mL) was added, followed
by dry dichloromethane (12 mL) and 3-butenyl acetate (60 µL, 0.88
mmol). After 1 h, the cold bath was removed, and the reaction was
allowed to warm to room temperature over the course of an hour. The
solution was separated from the silver chloride precipitate by cannula
filtration. The solvent was removed in vacuo, and the product
crystallized from THF/ether at -30 °C (orange crystals, 222 mg, 55%
yield). 1H NMR (CDCl3, 400 MHz): 0.40 (d, 6.4 Hz, (CH3)2CHCH2-
CH(OAc)Pd of 12(1,2)), 0.84 (t, 7.0 Hz, CH3CH2CH2CH2CH(OAc)Pd
of 12(2,1)), 1.1-1.2 (m, (CH3)2CHCH2CH(OAc)Pd of 12(1,2) and
CH3CH2CH2CH2CH(OAc)Pd of 12(2,1)), 1.2-1.3 (m, CH3CH2CH2CH2-
CH(OAc)Pd of 12(2,1)), 1.6-1.7 (m, CH3CH2CH2CH2CH(OAc)Pd of
12(2,1) and (CH3)2CHCH2CH(OAc)Pd of 12(1,2)), 2.19 (s, CO2CH3 of
12(1,2)), 2.21 (s, CO2CH3 of 12(2,1)), 2.28, 2.41, 2.42, 2.56 (N-Ar-
CH3 of 12(1,2)), 2.27, 2.41, 2.43, 2.56 (N-Ar-CH3 of 12(2,1)), 5.49 (dd,
2.2 Hz, 6.6 Hz, CH3CH2CH2CH2CH(OAc)Pd of 12(2,1)), 5.54 (br dd,
coupling unresolved, (CH3)2CHCH2CH(OAc)Pd of 12(1,2)), 6.64, 6.87
(d, 7 Hz each, H3 and H8 of acenaphthyl ring of 9), 6.66, 6.88 (d, 7
Hz each, H3 and H8 of acenaphthyl ring of 12(2,1)), 7.25-7.55 (N-Ar
of both 12(1,2)and 12(2,1)), 7.57-7.67 (m, H4 and H7 of acenaphthyl
ring of 12(1,2)and 12(2,1), unambiguously assigned by COSY), 8.24, 8.27
(d, 1H each, 8 Hz, H5 and H6 of acenaphthyl ring of 12(2,1), obscures
Preparation of [(N∧N)Pd(κ2-CH(Et)(O2CCF3))][SbF6] (6). In the
glovebox, a flame-dried Schlenk was charged with (N∧N)PdMeCl
(227.6 mg, 0.417 mmol) and AgSbF6 (148.8 mg, 0.433 mmol). The
flask was brought out to the vacuum line and lowered to -78 °C. Dry
CH2Cl2 (22 mL) was cooled in a separate flame-dried Schlenk flask
and transferred via cannula to the solids. Almost immediately, a yellow,
flocculent precipitate was formed. The reaction was allowed to stir for
ca. 30 min, and VAf (100 µL, 0.86 mmol) was added via gastight
syringe. The reaction was warmed to -20 °C for 30 min and was then
cannula-filtered into another flame-dried Schlenk flask. The resultant
orange solution was kept at -20 °C and evaporated to dryness in vacuo,
yielding a foamy solid. The product was crystallized from CH2Cl2/
diethyl ether at -35 °C to afford brownish clumps which were washed
1
with ether at -20 °C (252 mg, 68% yield). H NMR (CD2Cl2, 500
MHz, 253 K): δ ) 1.05 (br. t, 3H, CH(O2CCF3)CH2CH3); 0.90 and
1.20 (br. m, 1H each, CH(O2CCF3)CH2CH3); 2.17, 2.28, 2.32, 2.40 (s,
3H each, N-aryl-CH3); 5.83 (br. d, 1H, 8.4 Hz, CH(O2CCF3)(Et));
6.65 and 6.80 (d, 1H each, 8 Hz, acenaphthyl H3, H8); 7.3 (m, 10H,
N-Aryl protons (m and p) plus p protons of B(Arf)4-); 7.6 (br s, 8H,
o protons of B(Arf)4-); 8.12 (dd, 2H, 8.0 Hz, 7.6 Hz, acenaphthyl H4,
H7); 8.15 and 8.18 (d, 1H each, 8.4 Hz, 8.8 Hz, acenaphthyl H5, H6).
13C NMR (CD2Cl2, 253 K, 125 MHz): δ ) 11.8 (CH(O2CCF3)
CH2CH3); 17.6, 18.1, 18.2, 18.4 (N-Aryl-CH3); 28.9 (CH(O2CCF3)-
CH2CH3); 114.8 (CH(O2CCF3) CH2CH3); 124.4, 124.6, 128.3, 129.7,
130.7, 130.8, 133.5, 134.5 (aromatic and acenaphthyl C-H carbons);
126.0, 126.2, 128.8, 128.9, 129.2, 129.7, 131.5, 141.9, 142.9, 147.4,
171.0, 177.6 (quaternary aromatic and acenaphthyl carbons). Carbonyl
(O2CCF3) and trifluoromethyl carbons not found. 6b was insufficiently
stable for elemental analysis.
related signals for 9). Anal. Calcd for C35H37N2O2PdSbF6 (12(1,2)
/
12(2,1)): C, 48.89; H, 4.34; N, 3.26. Found: C, 48.29; H, 4.66; N, 3.06.
Preparation of [(N∧N)Pd(κ2-C(Me)(Et)(OAc))][SbF6] (13). A
flame-dried Schlenk flask was charged with (N∧N)PdMeCl (162.7 mg,
0.300 mmol) and AgSbF6 (97.1 mg, 2.83 mmol) in the drybox. Dry
dichloromethane (20 mL) and vinylidene acetate (0.50 mL, 4.5 mmol)
were precooled to -78 °C in a second flame-dried Schlenk flask. The
dichloromethane solution was added to the solids by cannula transfer.
The reaction was allowed to stir for ca. 10 min, and then the cold bath
was removed and the reaction was allowed to warm to room temperature
over the course of ca. 2 h. The solvent was removed in vacuo, and the
product crystallized as brownish crystals from dichloromethane/ether
at -30 °C (190.2 mg, 80% yield). 1H NMR (CD2Cl2, 400 MHz): δ )
0.68 (dq, 1H, 7 Hz, 16 Hz CH3CHH′C(Me)(OAc)Pd), 1.00 (s, 3H,
CH3CHH′C(Me)(OAc)Pd), 1.35 (t, 7.2 Hz, 3H, CH3CH2C(Me)(OAc)-
Pd), 1.50 (dq, 1H, 7 Hz, 16 Hz, CH3CHH′CH(OAc)Pd), 2.20 (s, 3H,
CO2CH3), 2.33 (s, 3H, Ar-CH3), 2.35 (s, 3H, Ar-CH3), 2.38 (s, 3H,
Ar-CH3), 2.44 (s, 3H, Ar-CH3), 6.47, 6.86 (d, 6.8 Hz, 7.2 Hz, 1H
each, H1 and H6 of acenaphthyl ligand), 7.3-7.5 (m, 6H, N-Ar
signals), 7.60 (m, 2H, H2 and H5 of acenaphthyl ligand), 8.23, 8.26
(d, 8 Hz, 8.4 Hz, 1H each, H3 and H4 of acenaphthyl ligand). 13C
NMR (CD2Cl2, 100 MHz): 11.4 (Pd-C(Me)(CH2CH3)), 17.3, 17.6,
17.9, 18.0, (NAr-CH3 and CO2CH3), 26.0 (PdCCH2Me), 35.0 (Pd-
C(Me)(CH2CH3), 118.1 (Pd-C(Me)(Et)(OAc)), 124.5, 125.0, 125.5,
127.7, 127.8, 127.9, 128.8, 129.1, 129.2, 129.3, 129.4, 129.5, 129.6,
129.9, 130.0, 131.3, 132.9, 133.8, 142.4, 143.0, 146.1, 169.1, 175.7,
184.7 (CdO).
Preparation of [(N∧N)Pd(κ2-CH(R)(OAc))] [SbF6] R ) iPr, nPr,
(11(1,2), 11(2,1)). A flame-dried Schlenk flask was charged with
(N∧N)PdMeCl (370.7 mg, 0.680 mmol) and AgSbF6 (237.2 mg, 0.690
mmol) in the drybox. The flask was brought out to the Schlenk line
and cooled in a -78 °C bath. Dry ether (10 mL) was added, followed
by dry dichloromethane (10 mL) and allyl acetate (75 µL, 0.695 mmol).
The reaction warmed to -25 °C over the course of 1.5 h and then
warmed to room temperature with a water bath. The cold bath was
removed, and the reaction allowed to warm to room temperature over
the course of an hour. The solution was separated from the silver
chloride precipitate by cannula filtration. The solvent was removed in
vacuo, and the product crystallized from THF/ether at -30 °C. Two
crystal forms were noted: red-orange plates (114 mg, 20% yield) and
yellow microcrystals (241 mg, 41% yield). These were separated by
the method of Pasteur. The red-orange plates were nearly pure
[(N∧N)Pd(CH(nPr)(OAc))] [SbF6] (11(2,1)), while the yellow crystallites
1
were a mixture of the two isomers. H NMR of 11(2,1) (CDCl3, 400
MHz): δ ) 0.781 (t, 7.2 Hz, 3H, CH3CH2CH2), 1.25 (m, 2H,
CH3CH2CH2), 1.69 (m, 2H, CH3CH2CH2), 2.20 (s, O2CCH3), 2.28,
2.42, 2.43, 2.59 (all s, 3H ea, Ar-CH3), 5.53 (dd, 2 Hz, 6 Hz,
CH(OAc)(nPr)Pd)), 6.67, 6.89 (both d, 7.2 Hz ea, 1H each, H3 and
H8 of acenaphthyl ring), 7.62, 7.58 (both t, each 7.2 Hz, 1H each, H4
and H7 of acenaphthyl ring), 8.23, 8.27 (both d, 8.0, 8.4 Hz, 1H each,
H5 and H6 of acenaphthyl ring). 1H NMR of 11(1,2) (CDCl3, 400 MHz,
many peaks obscured by 11(2,1)): 0.65 (d, 6.4 Hz, 6H, (CH3)2CHCH-
(OAc)Pd), 1.23 (obscured, seen by COSY, (CH3)2CHCH(OAc)Pd), 2.19
(s, 3H, CO2CH3), 2.20, 2.30, 2.41, 2.47, (all s, 3H each, Ar-CH3),
5.45 (obscured, seen by COSY, (CH3)2CHCH(OAc)Pd), 6.66, 6.88 (d,
1H each, 7.4 Hz, 1H each, H3 and H8 of acenaphthyl ring), 7.25-
Preparation of [(N∧N)Pd(κ1-CH(Et)(OAc))(NCCH3)][SbF6] (14R)
and Isomers (14â and 14γ). A J. Young tube (Wilmad 504) was charged
with [(N∧N)Pd(CH(Et)(OAc))][SbF6] (4b) (35.3 mg, 42 µmol) and
1
CD2Cl2 (500 µL). Acetonitrile (100 µL) was added under argon. H
NMR (CD2Cl2, 178 K, 500 MHz): δ ) 0.52 (br. t, 3H, CH(OAc)-
CH2CH3); 1.14 and 1.49 (br. m, 1H each, CH(OAc)CH2CH3); 1.73 (s,
9
J. AM. CHEM. SOC. VOL. 127, NO. 14, 2005 5145

A R T I C L E S
Williams et al.
3H, PdNCCH3); 1.87, 2.13, 2.22, 2.25 (s, 3H each, N-aryl-CH3); 2.37
(s, 3H, O2CCH3); 3.80 (br. d, 1H, 8.4 Hz, CH(OAc)(Et)); 6.51 and
6.94 (d, 1H each, 7.6, 7.2 Hz, acenaphthyl H3, H8); 7.21-7.33 (m,
6H, N-Aryl protons (m and p)); 7.48 and 7.52 (t, 1H each, 8.0 Hz,
7.6 Hz, acenaphthyl H4, H7); 8.15 and 8.18 (d, 1H each, 8.4 Hz, 8.8
Hz, acenaphthyl H5, H6). 13C NMR (CD2Cl2, 178K, 125 MHz): δ )
2.2 (CH3CN); 11.3 (CH(OAc)CH2CH3); 17.4, 17.4, 17.7, 17.9 (N-
Aryl-CH3); 21.0 (O2CCH3); 25.1 (CH(OAc)CH2CH3); 81.7 (CH(OAc)-
CH2CH3); 124.4, 124.9, 126.8, 128.0, 128.3, 128.8, 128.8, 128.9, 129.1,
132.0, 132.8 (aromatic and acenaphthyl C-H carbons); 124.7, 127.0,
127.8, 127.9, 129.7, 130.2, 140.2, 141.7, 145.2, 167.5, 170.3, 174.3
(quaternary aromatic and acenaphthyl carbons) 186.5 (O2CCH3).
CH3CN-Pd obscured by free CH3CN. Upon warming the above
solution of 14R, signals assigned by COSY to 14â and 14γ were
observed. 1H NMR (500 MHz, COSY, 265 K) 14â: 4.02, 3.47
(CH(Me)CH2), 1.60 (CH(Me)CH2), 0.70 (CH(Me)CH2). 14γ: 3.88
(PdCH2CH2CH2OAc), 1.28 (PdCH2CH2CH2OAc), 1.57 (PdCH2CH2-
CH2OAc).
tube charged with CDCl2CDCl2 (ca. 300 µL) while in an argon-filled
glovebox. Kinetic experiments were carried out in the NMR probe
(temperature calibrated with ethylene glycol), and reactions were
monitored by the disappearance of integration intensity of the terminal
methyl signal on the alkyl chain attached to the acetate moiety. For 6,
an NMR tube was charged in the glovebox with 6 and capped with a
septum. Outside the glovebox, the NMR tube was cooled in dry ice/
acetone, and dry CD2Cl2 was added through the septum via gastight
syringe. The tube was placed in the NMR probe while cold (<-40
°C), and the NMR probe was then warmed to -10 °C and the reaction
was monitored as described above. Low-temperature calibration was
performed with a methanol standard in place of ethylene glycol.
Insertion Kintetics (General Procedure). Samples were prepared
as described above for low-temperature kinetics of 6, with either CD2Cl2
or CDCl2F as the solvent. The probe was precooled to -80 °C before
adding the sample and was then warmed to the temperature at which
the kinetic experiment was performed.
Ethylene Insertion into Pd-C Bond of 16. Standard kinetic NMR
procedures were followed with the following modifications. Spectra
were acquired at -40 °C, while the reaction was allowed to proceed
at 0° in an ice bath between spectra. Ethylene (2-10 mL at 25 °C)
was syringed into the solution at varying intervals to maintain a pressure
of ethylene in the tube. During ethylene addition, the tube was
maintained at -78 °C and shaken vigorously in order to adequately
dissolve the ethylene. As ethylene consumption increased in rate as
the reaction progressed, it was only possible to monitor the reaction at
very early reaction times. Propylene was not observed under these
circumstances, and signals between 0.9 and 1.4 ppm grew in with time,
suggesting the growth of a polyethylene chain.
Preparation of [(N∧N)Pd(κ1-C(O)CH(Et)(OAc))(NCCH3)][SbF6]
(15R) and Isomers (15â and 15γ). Initially using the same procedure
described for14R-γ, a sample was subsequently freeze-pump-thawed
1
and placed under an atmosphere of CO to generate 15R-γ. H NMR
(COSY, 500 MHz, 293 K) 15R: 0.70 (PdC(O)CH(OAc)CH2CH3), 1.32,
1.74(PdC(O)CH(OAc)CH2CH3), 4.98 (PdC(O)CH(OAc)CH2CH3).
15â: 0.96 (PdC(O)CH(CH3)CH2OAc), 3.07 (PdC(O)CH(CH3)CH2OAc),
3.94 (PdC(O)CH(CH3)CH2OAc). 15γ: 3.69 (PdC(O)CH2CH2CH2OAc),
1.46 (PdC(O)CH2CH2CH2OAc), 2.82 (PdC(O)CH2CH2CH2OAc).
1
Partial NMR Data for 16-18. H NMR for 16 (158 K, CDCl2F,
500 MHz): All signals extremely broad. 4.94 (2H, CH2CH2), 4.31 (1H,
CHHCH2), 4.22 (1H, CHHCH2), 4.12 (1H, PdCH(OAc)CH2CH3), 1.63
(1H, PdCH(OAc)CHHCH3-other methylene proton obscured), 0.33 (3H,
Acknowledgment. The authors thank Mr. Weijun Liu for his
determination of the relative binding constants of ethylene and
para-vinyl anisole and Professor Olafs Daugulis for many
helpful discussions. The authors also thank the National Science
Foundation(CHE-0107810) and DuPont for financial support
of this work.
1
PdCH(OAc)CH2CH3). H NMR for 17 (233 K, CD2Cl2, 500 MHz):
4.07 (1H, PdCH(OAcf)CH2CH3), 1.71 (1H, PdCH(OAcf)CHHCH3),
1.34 (1H, PdCH(OAc)CHHCH3), 0.67 (3H, PdCH(OAcf)CH2CH3). 1H
NMR for 18 (233 K, CD2Cl2, 500 MHz): 4.71 (4H, AB pattern,
CH2CH2), 4.28 (1H, PdCH(OAcf)CH2CH3), 1.89 (1H, PdCH(OAcf)-
CHHCH3), 1.68 (1H, PdCH(OAc)CHHCH3), 0.35 (3H, PdCH(OAcf)-
CH2CH3).
Thermolysis Experiments (General Procedure). In thermolysis
experiments of 4a, 4b, 11(2,1), 11(1,2), 12(2,1), 12(1,2), and 13, a small
amount of the complex (8-20 mg) was added to a J. Young NMR
Supporting Information Available: Crystallographic data for
4b as a CIF file. This material is available free of charge via
the Internet at http://pubs.acs.org.
JA045969S
9
5146 J. AM. CHEM. SOC. VOL. 127, NO. 14, 2005