Mechanistic studies of nickel(II) alkyl agostic cations and alkyl ethylene complexes: Investigations of chain propagation and isomerization in (α-diimine)Ni(II)-catalyzed ethylene polymerization

Article abstract of DOI:10.1021/ja021071w

The synthesis of a series of (α-diimine)NiR₂ (R = Et, ⁿPr) complexes via Grignard alkylation of the corresponding (α-diimine)NiBr₂ precursors is presented. Protonation of these species by the oxonium acid [H(OEt₂)₂]⁺[BAr ′₄]^- at low temperatures yields cationic Ni(II) β-agostic alkyl complexes which model relevant intermediates present in nickel-catalyzed olefin polymerization reactions. The highly dynamic nature of these agostic alkyl cations is quantitatively addressed using NMR line broadening techniques. Trapping of these complexes with ethylene provides cationic Ni alkyl ethylene species, which are used to determine rates of ethylene insertion into primary and secondary carbon centers. The Ni agostic alkyl cations are also trapped by CH₃CN and Me₂S to yield Ni(R)(L)⁺ (L = CH₃CN, Me₂S) complexes, and the dynamic behavior of these species in the presence of varied [L] is discussed. The kinetic data obtained from these experiments are used to present an overall picture of the ethylene polymerization mechanism for (α-diimine)-Ni catalysts, including effects of reaction temperature and ethylene pressure on catalyst activity, polyethylene branching, and polymer architecture. Detailed comparisons of these systems to the previously presented analogous palladium catalysts are made.

Full text of DOI:10.1021/ja021071w

Published on Web 02/14/2003
Mechanistic Studies of Nickel(II) Alkyl Agostic Cations and
Alkyl Ethylene Complexes: Investigations of Chain
Propagation and Isomerization in (r-diimine)Ni(II)-Catalyzed
Ethylene Polymerization
Mark D. Leatherman, Steven A. Svejda, Lynda K. Johnson, and Maurice Brookhart*
Contribution from the Department of Chemistry, UniVersity of North Carolina at Chapel Hill,
CB# 3290 Venable Hall, Chapel Hill, North Carolina 27599-3290
Received August 12, 2002 ; E-mail: mbrookhart@unc.edu
n
Abstract: The synthesis of a series of (R-diimine)NiR2 (R ) Et, Pr) complexes via Grignard alkylation of
the corresponding (R-diimine)NiBr2 precursors is presented. Protonation of these species by the oxonium
acid [H(OEt2)2]+[BAr′4]- at low temperatures yields cationic Ni(II) â-agostic alkyl complexes which model
relevant intermediates present in nickel-catalyzed olefin polymerization reactions. The highly dynamic nature
of these agostic alkyl cations is quantitatively addressed using NMR line broadening techniques. Trapping
of these complexes with ethylene provides cationic Ni alkyl ethylene species, which are used to determine
rates of ethylene insertion into primary and secondary carbon centers. The Ni agostic alkyl cations are
also trapped by CH3CN and Me2S to yield Ni(R)(L)+ (L ) CH3CN, Me2S) complexes, and the dynamic
behavior of these species in the presence of varied [L] is discussed. The kinetic data obtained from these
experiments are used to present an overall picture of the ethylene polymerization mechanism for (R-diimine)-
Ni catalysts, including effects of reaction temperature and ethylene pressure on catalyst activity, polyethylene
branching, and polymer architecture. Detailed comparisons of these systems to the previously presented
analogous palladium catalysts are made.
Introduction
(II)- and Pd(II)-R-diimine systems that exhibit several unique
features, including the copolymerization of ethylene and func-
Significant recent research has been directed toward the
design and application of catalyst systems based on late
transition metals capable of effecting the polymerization of
olefins to high molar mass homo- and copolymers.1-12 Reports
from these laboratories13,14 and DuPont15-19 have explored Ni-
tionalized olefins,20-23 the oligomerization of ethylene and
R-olefins,24-26 the living polymerization of ethylene and R-ole-
fins,27,28 and the homopolymerization of cyclic16 and internal
acyclic olefins.29
The microstructure of polyethylene produced by nickel(II)-
and palladium(II)-diimine catalysts varies from strictly linear
(1) Keim, W.; Appel, R.; Storeck, A.; Kruger, C.; Goddard, R. Angew. Chem.,
Int. Ed. Engl. 1981, 20, 116-117.
(16) McLain, S. J.; Feldman, J.; McCord, E. F.; Gardner, K. H.; Teasley, M.
F.; Coughlin, E. B.; Sweetman, K. J.; Johnson, L. K.; Brookhart, M.
Macromolecules 1998, 31, 6705-6707.
(2) Mo¨hring, V. M.; Fink, G. Angew. Chem., Int. Ed. Engl. 1985, 24, 1001-
1003.
(3) Ostoja Starzewski, K. A.; Witte, J. Angew. Chem. 1985, 97, 610-612.
(4) Klabunde, U.; Ittel, S. D. J. Mol. Catal. 1987, 41, 123-134.
(5) Brookhart, M.; DeSimone, J. M.; Grant, B. E.; Tanner, M. J. Macromol-
ecules 1995, 28, 5378-5380.
(17) Guan, Z.; Cotts, P. M.; McCord, E. F.; McLain, S. J. Science 1999, 283,
2059-2062.
(18) Cotts, P. M.; Guan, Z.; McCord, E. F.; McLain, S. J. Macromolecules 2000,
33, 6945-6952.
(6) Small, B. L.; Brookhart, M.; Bennett, A. M. A. J. Am. Chem. Soc. 1998,
120, 4049-4050.
(19) McCord, E. F.; McLain, S. J.; Nelson, L. T. J.; Arthur, S. D.; Coughlin, E.
B.; Ittel, S. D.; Johnson, L. K.; Tempel, D. J.; Killian, C. M.; Brookhart,
M. Macromolecules 2001, 34, 362-371.
(7) Small, B. L.; Brookhart, M. Macromolecules 1999, 32, 2120-2130.
(8) Britovsek, G. J. P.; Gibson, V. C.; Kimberly, B. S.; Maddox, P. J.;
McTravish, S. J.; Solan, G. A.; White, A. J. P.; Williams, D. J. Chem.
Commun. 1998, 849-850.
(20) Johnson, L. K.; Mecking, S.; Brookhart, M. J. Am. Chem. Soc. 1996, 118,
267-268.
(21) Mecking, S.; Johnson, L. K.; Wang, L.; Brookhart, M. J. Am. Chem. Soc.
1998, 120, 888-899.
(9) Britovsek, G. J. P.; Gibson, V. C.; Wass, D. F. Angew. Chem., Int. Ed.
1999, 38, 429-447.
(22) Johnson, L.; Bennett, A.; Dobbs, K.; Hauptman, E.; Ionkin, A.; Ittel, S.;
McCord, E.; McLain, S.; Radzewich, C.; Yin, Z.; Wang, L.; Wang, Y.;
Brookhart, M. Polym. Mater. Sci. Eng. 2002, 86, 319.
(23) McLain, S. J.; Sweetman, K. J.; Johnson, L. K.; McCord, E. F. Polym.
Mater. Sci. Eng. 2002, 86, 320-321.
(10) Younkin, T. R.; Connor, E. F.; Henderson, J. I.; Friedrich, S. K.; Grubbs,
R. H.; Bansleben, D. A. Science 2000, 287, 460-462.
(11) Hicks, F. A.; Brookhart, M. Organometallics 2001, 20, 3217-3219.
(12) Ittel, S. D.; Johnson, L. K.; Brookhart, M. Chem. ReV. 2000, 100, 1169-
1203.
(24) Killian, C. M.; Johnson, L. K.; Brookhart, M. Organometallics 1997, 16,
2005-2007.
(25) Tempel, D. J. Ph.D. Dissertation, University of North Carolina at Chapel
Hill, 1998.
(26) Svejda, S. A.; Brookhart, M. Organometallics 1999, 18, 65-74.
(27) Killian, C. M.; Tempel, D. J.; Johnson, L. K.; Brookhart, M. J. Am. Chem.
Soc. 1996, 118, 11664-11665.
(28) Gottfried, A. C.; Brookhart, M. Macromolecules 2001, 34, 1140-1142.
(29) Leatherman, M. D.; Brookhart, M. Macromolecules 2001, 34, 2748-2750.
(13) Johnson, L. K.; Killian, C. M.; Brookhart, M. J. Am. Chem. Soc. 1995,
117, 6414-6415.
(14) Gates, D. P.; Svejda, S. A.; On˜ate, E.; Killian, C. M.; Johnson, L. K.; White,
P. S.; Brookhart, M. Macromolecules 2000, 33, 2320-2334.
(15) Brookhart, M. S.; Johnson, L. K.; Killian, C. M.; Arthur, S. D.; Feldman,
J.; McCord, E. F.; McLain, S. J.; Kreutzer, K. A.; Bennett, A. M. A.;
Coughlin, E. B.; Ittel, S. D.; Parthasarathy, A.; Tempel, D. J. WO Patent
Application 9623010 to DuPont, April 3, 1995.
9
3068
J. AM. CHEM. SOC. 2003, 125, 3068-3081
10.1021/ja021071w CCC: $25.00 © 2003 American Chemical Society

Nickel(II) Alkyl Agostic Cations
A R T I C L E S
Scheme 1. Proposed Mechanism of Ethylene Polymerization
Scheme 2. Isomerization Equilibria in (R-diimine)Pd Complexes
to highly branched, while the microstructures of poly(R-olefins)
produced by these same catalyst systems are characterized by
fewer branches than are expected from 1,2-monomer enchain-
ment (chain-straightening).13,14 The degree of branching in the
polyethylene and the polymer architecture can depend on a
number of factors: the metal, the steric bulk of the diimine
ligand, and the reaction conditions. Polyethylenes produced by
palladium(II)-diimine catalysts are generally highly branched
(e.g., >70 methyls/1000 carbons) regardless of ligand structure
or reaction conditions, although the precise polymer architecture
varies from linear polyethylene with short chain branches,
obtained at high ethylene pressures, to a more hyperbranched
structure at lower ethylene concentrations.17 In contrast, poly-
ethylenes prepared from nickel(II)-diimine catalysts display
overall branching numbers that are quite variable. Polymeriza-
tions run at low temperatures and high ethylene pressure using
nickel complexes bearing relatively nonbulky diimine ligands
yield nearly linear polyethylene, while higher temperatures and/
or lower ethylene concentrations in combination with nickel
complexes bearing sterically demanding diimine ligands result
in more highly branched polyethylene.12-14,30 In the case of
nickel-catalyzed polymerization of R-olefins, the degree of
chain-straightening observed in the resulting poly(R-olefin)s is
strongly affected by diimine ligand symmetry and reaction
conditions.25 Palladium catalysts generally produce more chain-
straightened poly(R-olefin) products than their nickel analogues,
but unlike the nickel systems, the degree of branching seen in
poly(R-olefin)s produced by the palladium complexes is rela-
tively independent of ligand structure and reaction conditions.
Mechanistic studies of ethylene polymerizations catalyzed by
both nickel(II)- and palladium(II)-diimine complexes have led
to the proposed polymerization mechanism shown in Scheme
1.13,25,31-35 Following initiation, the mechanism is characterized
by chain propagation, chain isomerization, and chain transfer
steps. To date, the chain propagation step is the best-studied
portion of the overall mechanism. Low-temperature NMR
studies of ethylene polymerization have established the catalyst
resting state for both metals under these conditions as the alkyl
ethylene complexes, implying turnover-limiting migratory inser-
tion of ethylene. Activation barriers to ethylene insertion have
been measured for both palladium and nickel complexes. As
previously reported, the activation barriers to ethylene insertion
in the chain propagation step are 4-5 kcal/mol lower for the
nickel complexes as compared to their palladium congeners.32
Mechanistic studies of the chain isomerization process have
focused on (R-diimine)Pd(II) complexes. As established by these
studies, alkyl chain isomerization occurs in the â-agostic species,
which can be formed either after olefin insertion from the metal-
(alkyl)olefin resting state or by ethylene loss from the resting
state. The barrier to interconversion of the resulting â-agostic
species (via â-H elimination, olefin rotation, and reinsertion)
is quite low (<11 kcal/mol), and the rate of the interconversion
is much faster than the rate of trapping of the â-agostic species
by ethylene. Low-temperature NMR studies of these (R-
diimine)Pd(II) â-agostic complexes have further established that
the activation barriers to olefin insertion (propagation) are ca.
17-18 kcal/mol and are significantly higher than the activation
barriers for chain isomerizations occurring from these resting
states.25,33,34 This feature is illustrated in the simplified version
of chain isomerization shown in Scheme 2. Low-temperature
NMR studies have shown that the (R-diimine)Pd(iPr)(C2H4)+
and (R-diimine)Pd(nPr)(C2H4)+ complexes completely equili-
brate via ethylene loss and isomerization of the agostic species
prior to insertion. That chain isomerization occurs rapidly
relative to insertion even from the catalyst resting state explains
both the highly branched polyethylenes produced by the (R-
diimine)palladium systems and the independence of the total
(30) Killian, C. M. Ph.D. Dissertation, University of North Carolina at Chapel
Hill, 1996.
(31) Tempel, D. J.; Brookhart, M. Organometallics 1998, 17, 2290-2296.
(32) Svejda, S. A.; Johnson, L. K.; Brookhart, M. J. Am. Chem. Soc. 1999,
121, 10634-10635.
(33) Tempel, D. J.; Johnson, L. K.; Huff, R. L.; White, P. S.; Brookhart, M. J.
Am. Chem. Soc. 2000, 122, 6686-6700.
(34) Shultz, L. H.; Tempel, D. J.; Brookhart, M. J. Am. Chem. Soc. 2001, 123,
11539-11555.
(35) Shultz, L. H.; Brookhart, M. Organometallics 2001, 20, 3975-3982.
9
J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003 3069

A R T I C L E S
Leatherman et al.
Scheme 3. Synthesis of Pd â-Agostic Isopropyl Complexes
palladium complexes, the clean, quantitative formation of the
methyl ethylene complex shown in Scheme 3 is straightforward.
Excess monomer can be easily purged from the solution with
argon. Migratory insertion of ethylene occurs when the solution
is warmed to 0 °C. The isopropyl â-agostic complex 1b is the
observed reaction product, which presumably forms following
rapid isomerization of the initially formed n-propyl â-agostic
complex 1a. When the solution is cooled to -110 °C, the agostic
hydrogens of complexes 1a and 1b are clearly observed in the
1H NMR spectra.
Attempts to form â-agostic nickel complexes through inser-
tion of 1 equiv of ethylene in nickel methyl ethylene complexes
were unsuccessful (Scheme 4). The addition of 1 equiv of
ethylene to ether adduct 2 at -110 °C led to very slow formation
of methyl ethylene complex 3. Even at -110 °C, multiple
insertions of ethylene occurred prior to complete ether displace-
ment, resulting in a complex mixture of the starting ether adduct,
the desired methyl ethylene compound 3, and various insertion
products. Use of a large excess of ethylene to hasten the
formation of the methyl ethylene complex 3 from ether adduct
2, followed by an argon purge to remove the excess ethylene,
also failed. Multiple insertions of ethylene again occurred,
possibly due to slight warming of the solution by the argon
purge gas. The sterically bulkier diisopropyl ether ligand of
[(ArNdC(An)-C(An)dNAr)Ni(CH3)(OiPr2)]+[BAr4′]- (Ar )
2,6-C6H3Me2), 4, was anticipated to undergo more facile
displacement by ethylene to form methyl ethylene complex 3.
Unfortunately, numerous attempts to prepare 4 failed to yield
the desired product; this compound appears to be unstable even
at -78 °C.
An alternative route to the desired â-agostic nickel complexes
based upon protonolysis of an (R-diimine)NiPr2 complex with
[H(OEt2)2]+[BAr′4]- was found to be successful. Alkylation of
(R-diimine)NiBr2 complexes 5 and 6 with 2 equiv of n-
propylmagnesium bromide in diethyl ether at -78 °C produced
the corresponding (R-diimine)NiPr2 complexes 7 and 8 (Scheme
5). Use of ethylmagnesium bromide in this synthesis produces
the corresponding (R-diimine)NiEt2 complexes 9 and 10 (Scheme
5). Diethyl ether solutions of all complexes are very dark blue
in color, while the complexes are dark purple in the solid state.
The solids appear to be stable indefinitely when stored at
temperatures below -15 °C in a drybox, but are thermally
sensitive in solution. Complexes 7, 8 and 9, 10 undergo
reductive elimination to yield hexane or butane, respectively,
degree of polymer branching on ethylene pressure demonstrated
by palladium catalysts. However, since the rate of trapping is
first-order in ethylene, the average distance that the Pd center
“walks” along the polymer chain prior to insertion is affected
by monomer concentration, resulting in varying polymer topolo-
gies depending on olefin pressure.17
Thus far, the mechanism of chain growth and chain isomer-
ization has been studied in some detail for the palladium
systems. In contrast, difficult synthesis and manipulation of well-
defined (R-diimine)nickel(II) complexes has, until recently,
hindered mechanistic investigations of these systems. However,
recent advances in the preparation and purification of such
complexes have facilitated mechanistic studies of olefin insertion
and chain growth processes in these systems.32 In this paper,
an account of our studies of cationic (R-diimine)Ni(II) â-agostic
complexes is presented. Some implications of these studies
concerning the mechanism of chain isomerization during (R-
diimine)nickel(II)-catalyzed olefin polymerizations are discussed
and compared to Pd(II) analogues.
Results
A. Synthesis and Reactivity of (r-Diimine)Ni(II) â-Agostic
n-Propyl and Isopropyl Complexes. Recent work in our
laboratory with (R-diimine)Pd(II) â-agostic complexes has
established that chain isomerization occurs rapidly relative to
chain propagation.25,33,34 In these experiments, palladium â-ago-
stic complexes 1a and 1b were prepared and isolated on a
preparative scale using the conditions shown in Scheme 3. Since
the barrier to ethylene insertion is relatively high in these
Scheme 4. Attempted Synthesis of Ni â-Agostic Complexes through Single Ethylene Insertion
9
3070 J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003

Nickel(II) Alkyl Agostic Cations
A R T I C L E S
Scheme 5. Synthesis of (R-diimine)Nickel Di-n-propyl and Diethyl
Complexes
Scheme 7. Dynamic Processes Observed in â-Agostic Isopropyl
Complex i-11
Scheme 6. Synthesis of Ni â-Agostic Complexes i-11/n-11 via
Protonolysis of Nickel Dipropyl Complex 7
chemically inequivalent and appear as broad multiplets at 0.26
and 0.11 ppm. The nonagostic methyl group resonance appears
as a broad singlet at -0.09 ppm. The resonance due to the
isopropyl methine hydrogen (δ 2.00) is a complex multiplet at
temperatures below -120 °C, but appears as a septet at
temperatures above -80 °C.
Variable-temperature 1H NMR studies reveal that three
distinct dynamic processes occur in complex i-11: in-place
methyl rotation, isopropyl methyl exchange, and methyl/methine
hydrogen exchange. These processes are shown in Scheme 7.
Two of these processes, in-place methyl rotation and isopropyl
methyl exchange, are relatively low-energy processes as com-
pared to methyl/methine hydrogen exchange. As previously
mentioned, at temperatures below -120 °C, the proton NMR
resonance of the â-agostic hydrogen of complex i-11 is observed
as a triplet. Above -120 °C this signal quickly broadens, as do
the agostic and, at slightly higher temperatures, the nonagostic
isopropyl methyl resonances. This dynamic behavior is indica-
tive of facile hydrogen exchange through an in-place rotation
of the agostic methyl group, along with exchange of the agostic
methyl group with the free methyl group via rotation of the
isopropyl group about the Ni-C bond (Scheme 7).31 Quantita-
tive dynamic NMR spectroscopic analysis allows for the
determination of the energy barriers for these processes. Using
the slow exchange approximation, in-place methyl rotation
occurs with a ∆Gq of 8.0 kcal/mol at -99 °C, and the barrier
to isopropyl methyl exchange is slightly higher, 9.0 kcal/mol
at -77 °C. Coalescence of these six protons is observed as a
broad resonance at δ -2.1 ppm at temperatures above -50 °C.
Over the temperature range of -140 to -40 °C, the isopropyl
methine resonance displays no discernible line broadening; only
above -30 °C does this resonance begin to broaden. These
observations indicate that the barrier to exchange of the
isopropyl methine hydrogen with the isopropyl methyl hydro-
gens (interconversion of i-11 and i-11′, Scheme 7) is consider-
ably higher than the barriers to in-place methyl rotation and
when their CD2Cl2 solutions are warmed above -20 °C. These
complexes are extremely air-sensitive and immediately decom-
pose to form unknown products when exposed to the air.
Reactions of the nickel diethyl complexes 9 and 10 will be
discussed in the following section.
Protonolysis of nickel dipropyl complex 7 with 1 equiv of
[H(OEt2)2]+[BAr′4]- in diethyl ether at approximately -100 °C
yields propane and a mixture of two â-agostic species (Scheme
6). The mixture consists of a 20:1 equilibrium ratio of â-agostic
isopropyl complex i-11 and â-agostic n-propyl complex n-11.
DFT calculations are in agreement with the observed relative
stabilities of i-11 and n-11 and suggest that the origin of the
difference is electronic rather than steric in nature.36,37 The
n-propyl complex n-11 is presumably the initial product formed
immediately following the loss of propane after protonation of
7. Complex n-11 then isomerizes to the â-agostic isopropyl
complex i-11. The structure of the â-agostic isopropyl complex
i-11 was confirmed by low-temperature NMR experiments (1H
and 1H-1H COSY) obtained in CDCl2F at -130 °C. The proton
NMR resonance of the â-agostic hydrogen of i-11 is observed
2
as a triplet (δ -12.5, JHH ) 19 Hz), while the â-agostic
hydrogen of the n-propyl complex n-11 appears as a doublet
2
(δ -13.0, JHH ) 19 Hz) (see Supporting Information for the
1H NMR spectrum of the agostic region). The other two
hydrogens on the agostic methyl group (complex i-11) are
(36) Musaev, D. G.; Froese, R. D. J.; Svensson, M.; Morokuma, K. J. Am. Chem.
Soc. 1997, 119, 367-374.
(37) Deng, L.; Woo, T. K.; Cavallo, L.; Margl, P. M.; Ziegler, T. J. Am. Chem.
Soc. 1997, 119, 6177-6186.
9
J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003 3071

A R T I C L E S
Leatherman et al.
Scheme 8. Trapping of â-Agostic Species i-11/n-11 with Ethylene
at -130 °C
isopropyl methyl exchange. Line broadening analysis establishes
a ∆Gq of ca. 13.9 kcal/mol at 5 °C for this process. The overall
dynamic behavior exhibited by this agostic nickel complex is
similar to that reported for the analogous, well-studied â-agostic
isopropyl palladium complex.31 In the Pd case, the barrier to
â-H elimination/reinsertion (measured for the Pd ethyl agostic
cation as 6.8 kcal/mol at -120 °C)38 is lower than the barrier
for either in-place rotation (9.1 kcal/mol at -90 °C) or methyl
exchange (9.6 kcal/mol at -90 °C), indicating that the â-H
elimination is likely not the rate-determining step in the chain
isomerization. In contrast, the comparatively large ∆Gq for
scrambling of the methine hydrogen in i-11 supports the â-H
elimination/reinsertion step as rate-determining for the Ni system
(additional evidence for this supposition comes from study of
the agostic ethyl complex, presented in the following section).
ether is present. With 7 equiv of diethyl ether present, a roughly
1:1 ratio of the â-agostic isopropyl complex/diethyl ether adduct
is observed (Keq ca. 0.15 M-1).
Ligand steric effects on the dynamic processes in nickel
propyl agostic cations were examined using the Ni(iPr) agostic
complex 12 [(ArNdC(An)-C(An)dNAr)Ni(CH(CH2-µ-H)-
(CH3))]+[BAr4′]- (Ar ) 2,6-C6H3 Pr2), obtained via protonolysis
i
of 8 as previously described. This reaction furnishes the
isopropyl agostic species exclusively, as no characteristic
resonances for the n-propyl group are observed in the agostic
or alkyl regions of the low-temperature static 1H NMR spectrum
(-120 °C, CDCl2F).39 The agostic resonance for this complex
The addition of 5 equiv of ethylene at -130 °C to the mixture
of â-agostic complexes i-11 and n-11 (produced in situ by
protonolysis of nickel dipropyl complex 7) results in trapping
of the agostic complexes as their corresponding alkyl ethylene
complexes i-13 and n-13 in a 20:1 ratio (Scheme 8). The
structure of each complex was confirmed by low-temperature
1H NMR and 1H-1H COSY NMR spectroscopy. Key 1H NMR
resonances for i-13 (CDCl2F/CDClF2, -130 °C) include bound
ethylene signals at 3.96 and 3.53 ppm (apparent d, 2H each),
the isopropyl methine resonance at 1.47 ppm (m, 1H), and the
isopropyl methyl resonance at 0.40 ppm (br s, 6H). Diagnostic
resonances for n-13 (CDCl2F/CDClF2, -100 °C) include the
bound ethylene resonances at 4.17 and 3.31 ppm (br d, 2H each)
and the n-propyl group resonances at 1.22 ppm (m, 2H,
NiCH2CH2CH3), 0.61 ppm (t, J ) 6.9 Hz, 3H, NiCH2CH2CH3),
and 0.48 ppm (t, J ) 7.5 Hz, 2H, NiCH2CH2CH3). At -130
°C, no insertion of ethylene into the Ni-alkyl bond of either
complex is observed.
When the protonolysis of nickel dipropyl complex 7 with 1
equiv of [H(OEt2)2]+[BAr′4]- is carried out in the presence of
ethylene at -130 °C, the n-propyl ethylene complex n-13 is
the only product observed (Scheme 9). This result supports the
initial formation of the â-agostic n-propyl complex n-11
immediately following protonolysis. In the absence of a trapping
ligand such as ethylene, the â-agostic n-propyl complex n-11
isomerizes to the â-agostic isopropyl complex i-11. In the
presence of ethylene, the initially formed â-agostic n-propyl
complex n-11 is trapped by ethylene faster than it isomerizes
via â-elimination/readdition, forming n-13. Isomerization of
n-13 to i-13 is not observed at temperatures between -130 and
-90 °C, characterized by the lack of the distinctive isopropyl
methyl doublet upfield of all other Ni-alkyl resonances.
However, ethylene insertion does occur at -100 °C, and
subsequent chain growth to give the linear alkyl species 14 is
observed (Scheme 9).
again appears as a triplet (δ -12.5, JHH ) 18 Hz). H-1H
COSY experiments allow assignment of the two nonagostic
methyl protons of the agostic methyl group (δ 0.67, br m), the
methine proton (δ 2.16, m), and the nonagostic methyl group
(δ -0.22, br d). Complex 12 displays variable-temperature
dynamic behavior similar to i-11, including scrambling of the
three protons of the agostic methyl group (∆Gq ) 8.4 kcal/mol
at -90 °C) and Ni-CR bond rotation to equilibrate all six methyl
protons (∆Gq ) 9.4 kcal/mol at -75 °C). The methine resonance
is a sharp septet (2JHH ) 5.2 Hz) above -70 °C and broadens
due to â-H elimination/reinsertion above -10 °C (∆Gq ) 13.4
kcal/mol). Comparison of these barriers to those determined for
i-11 shows that increased steric bulk around the metal center
slightly increases the barriers to CR-Câ and Ni-CR bond
rotations while slightly decreasing the â-H elimination/reinser-
tion rate.
2
1
When ca. 7 equiv of diethyl ether relative to complex i-11 is
present in a CDCl2F NMR solution, new resonances in the
proton NMR spectrum are observed. At -130 °C, previously
unobserved peaks appear at 6.28 ppm (br), 3.95 ppm (br m),
and 0.24 ppm (d, J ) 6.7 Hz). These signals are consistent with
a diethyl ether adduct. The resonance appearing at 6.28 ppm is
due to the acenaphthene backbone ortho hydrogens, the
resonance at 3.95 ppm is likely a methylene resonance of bound
diethyl ether, and the doublet at 0.24 ppm is consistent with
the isopropyl methyl group (eq 1). The equilibrium shown in
eq 1 lies in favor of the â-agostic isopropyl complex i-11, as
no diethyl ether adduct is observed when only 1 equiv of free
(38) In the 1H NMR spectra, the resonance for the methine hydrogen in the
Pd(isopropyl)agostic complex is obscured by ligand resonances; thus
quantitative analysis of its dynamic behavior cannot be determined.
Therefore, the barrier obtained in the similar Pd ethyl agostic species is
used as an approximation in this analysis.
(39) Calculations predict the iPr agostic complex to be thermodynamically
favored even in the presence of bulky ligands: Deng, L.; Woo, T. K.;
Cavallo, L.; Margl, P. M.; Ziegler, T. J. Am. Chem. Soc. 1997, 119, 6177-
6186. A small amount of the nPr agostic analogue of 12 may be present
upon protonolysis of 8 (as seen in i-11/n-11) with obscured resonances in
the 1H NMR spectrum.
At temperatures above -100 °C, migratory insertion of
ethylene in complex i-13 is observed (Scheme 10). As the
9
3072 J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003

Nickel(II) Alkyl Agostic Cations
A R T I C L E S
Scheme 9. Protonolysis of 7 in the Presence of 5 equiv of C2H4 at -130 °C and Subsequent Reactivity of n-Propyl Ethylene Complex n-13
Scheme 10. Ethylene Insertion into the Ni-Isopropyl Bond of
Complex i-13
Studies of similar Pd alkyl ethylene complexes have estab-
lished that Pd n-propyl ethylene and Pd isopropyl ethylene
complexes equilibrate prior to insertion via reversible ethylene
loss and facile â-elimination/reinsertion.25,33 Attempts to probe
the possible existence and nature of such an equilibrium between
the analogous Ni complexes i-13 and n-13 have shown no
observable ethylene dissociation or alkyl chain isomerization
in i-13 up to -80 °C, but are complicated by the relatively low
barrier to olefin insertion. However, the use of small, nucleo-
philic, noninserting ligands such as acetonitrile and dimethyl
sulfide to trap the free agostic alkyl cation provides insight into
the alkyl chain isomerization behavior of these complexes.
Introduction of a solution of 2 equiv of NCCH3 in CD2Cl2 to a
solution of i-11/n-11 (20:1) in CD2Cl2 in an NMR tube at -80
°C furnishes the Ni(isopropyl)(NCCH3) complex i-16 and Ni-
(n-propyl)(NCCH3) complex n-16 in a 20:1 kinetic ratio
(consistent with the equilibrium ratio of i-11:n-11) and 1 equiv
of free NCCH3 (Scheme 11). This ratio is unchanged upon
warming the sample to temperatures up to -10 °C; at higher
temperatures i-16 isomerizes slowly to n-16, until an equilibrium
mixture is reached (Keq ) 3, see Scheme 11). The dependence
of this isomerization rate on ligand concentration was deter-
mined,40 and the results are shown in Table 1. An approximately
inverse dependence of isomerization rate on [NCCH3] was
observed over a 5-fold concentration range. This behavior shows
that isomerization occurs via reversible dissociation of CH3CN
from i-16 followed by rate-determining isomerization of i-11
to n-11 and rapid trapping by CH3CN to yield n-16. A free
energy diagram illustrating this process is shown in Figure 1.
Similar effects are observed using dimethyl sulfide as the
trapping ligand. Addition of 1 equiv of Me2S to an NMR sample
of i-11/n-11 at -80 °C yields a mixture of i-17 and n-17 (n-
1
isopropyl methyl H NMR resonance disappears (δ 0.39, d, J
) 5.9 Hz, -80 °C), new resonances appear at δ 0.96 (br m,
-80 °C) and δ 0.64 (d, J ) 6.1 Hz, -80 °C), which are
consistent with the isopropyl methine and methyl groups,
respectively, of the single insertion product 15 (Scheme 10).
Isomerization of i-13 to n-13 prior to the initial ethylene
insertion does not occur, since none of the characteristic
resonances for n-13 or its single insertion product are observed.
Also, the appearance of the previously unobserved doublet at
0.64 ppm and resonances corresponding to the R- and â-me-
thylene protons of 15 (δ 0.47, t, J ) 6.4 Hz and 1.07, br q, J
) 6.4 Hz) with the coincidental disappearance of the isopropyl
methyl doublet in i-13 strongly suggests that migratory insertion
of ethylene into the Ni-2° alkyl bond of i-13 is occurring much
faster than isomerization of i-13 to n-13. The ethylene insertion
barriers in both i-13 and n-13 are discussed below.
(40) Espenson, J. H. Chemical Kinetics and Reaction Mechanisms; McGraw-
Hill: New York, 1981.
9
J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003 3073

A R T I C L E S
Leatherman et al.
Scheme 11. Isomerization Processes in Ni(propyl)(L) Cations (L ) NCCH3, SMe2)
complexes 18 and 19 would provide additional insight into the
mechanism of chain isomerization. Spencer and co-workers have
examined ligand substituent effects on dynamic processes in
similar bisphosphine platinum42,43 and nickel and palladium44
ethyl agostic cations, including the isolation and X-ray structural
characterization of a [(dbpe)Ni(C2H5)][BF4] (dbpe ) bis(di-tert-
butylphosphino)ethane) complex. Rapid CR-Câ bond rotation
averages the agostic methyl hydrogens above -100 °C, but â-H
elimination/reinsertion to equilibrate the two carbons and five
protons is not observed below 27 °C.
Figure 1. Free energy diagram for the isomerization of i-16 to n-16
Table 1. Isomerization Rates of Ni(isopropyl)L Complexes i-16
and i-17
As mentioned in the previous section, protonolysis of nickel
dipropyl complex 7 with [H(OEt2)2]+[BAr′4]- (Scheme 6)
proved to be a convenient synthetic route to the â-agostic
isopropyl complex i-11. Using the same strategy, protonolysis
of nickel diethyl complex 9 with [H(OEt2)2]+[BAr′4]- was
considered the best approach to the synthesis of â-agostic ethyl
complex 18. The synthesis of 9 was described in the previous
section and is shown in Scheme 5. The combination of 9 and
[H(OEt2)2]+[BAr′4]- at -130 °C in CDCl2F resulted in no
1
immediate reaction, as determined by H NMR. The sample
was slowly warmed, and no appreciable protonation was
observed until the temperature reached -90 °C. At -90 °C,
the resonances of nickel diethyl complex 9 slowly disappeared,
and at least two new nickel complexes were observed. The major
compound present in this solution is the cationic nickel ethyl
17/i-17 ) 0.33), which begins to isomerize at -50 °C (Scheme
11). Complete conversion to the linear species n-17 is observed.
The isomerization rate of the dimethyl sulfide complex is also
inhibited by excess ligand, but significant deviation from an
inverse relationship between kisom and [SMe2] is observed (see
last two entries in Table 1).41
1
diethyl ether adduct 20 (Scheme 12). A H-1H COSY NMR
(41) This deviation suggests a mechanistic regime in which the barrier to trapping
of i-11 is comparable to the isomerization barrier (i-11 to n-11), and thus
the loss of SMe2 cannot be treated as fast and reversible relative to the
isomerization step. The much lower temperature of isomerization of the
SMe2 complex and trapping of i-11/n-11 by SMe2 in a nonkinetic ratio
supports this contention.
B. (r-Diimine)Ni(II) â-Agostic Ethyl Complexes. The
dynamics of the â-agostic isopropyl complex i-8 discussed above
shed some light on basic processes that occur during chain
isomerization reactions of (R-diimine)Ni complexes. The an-
ticipated dynamic behavior of the similar â-agostic ethyl
(42) Carr, N.; Mole, L.; Orpen, A. G.; Spencer, J. L. J. Chem. Soc., Dalton
Trans. 1992, 2653-2662.
(43) Mole, L.; Spencer, J. L.; Carr, N.; Orpen, A. G. Organometallics 1991,
10, 49-52.
(44) Conroy-Lewis, F. M.; Mole, L.; Redhouse, A. D.; Litster, S. A.; Spencer,
J. L. J. Chem. Soc., Chem. Commun. 1991, 1601-1603.
9
3074 J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003

Nickel(II) Alkyl Agostic Cations
A R T I C L E S
Scheme 12. Protonation of Nickel Diethyl Complex 9 with
[H(OEt2)2]+[BAr′4]-
CR protons provides a barrier of 14.0 kcal/mol at 6 °C for the
â-hydride elimination-olefin rotation-reinsertion process.
As is the case for the Ni(propyl) agostic systems, ligand steric
effects can be ascertained through the synthesis of Ni diethyl
complex 10 (Scheme 5) and the analogous agostic cation 19.
Complex 10 was obtained with minor residual diethyl ether
present (<0.25 equiv) and, upon protonation with [H(OiPr2)2]+-
[BAr′4]-, furnishes exclusively the ethyl agostic cation 19 in
situ. The lack of diethyl ether coordination to the nickel center
in this case is likely due to steric effects of the bulkier ortho-
isopropyl groups on the ligand aryl rings. At -130 °C in
CDCl2F, 19 is not fluxional and exhibits distinct resonances
for the two nonagostic protons of the methyl group (δ 0.35, m)
and the agostic proton (δ -13.8, t, J ) 19 Hz). The Ni-CH2
resonances at 1.43 ppm were identified by significant coupling
to the nonagostic methyl protons and weaker coupling to the
1
agostic proton in the H-1H COSY spectrum, but were not
observed directly due to the coincident shift of one of the ligand
isopropyl methyl groups. As observed in complex 18, the
resonances of the agostic and nonagostic methyl protons of 19
begin to broaden above -105 °C and provide a ∆Gq ) 8.6
kcal/mol at -90 °C for the CR-Câ bond rotation. Coalescence
of these resonances is again observed (δ -4.2 ppm) at temp-
eratures above -45 °C, followed by broadening due to exchange
of all five ethyl protons through â-H elimination and reinsertion.
However, since the resonance for the CR protons is obscured
by the methyl resonances of the ligand isopropyl groups, no
quantitative analysis was feasible. An approximation for the â-H
elimination-rotation-reinsertion process in 19 can be obtained
from the broad coalesced methyl resonance at -4.2 ppm, which
agrees with the barrier determined in 18 (ca. 14.3 kcal/mol at
30 °C in 19 vs 14.0 ( 0.1 kcal/mol at 6.4 °C in 18).
C. Ethylene Insertion in (r-diimine)Ni(II) â-Agostic Pro-
pyl and Ethyl Complexes. Ethylene insertion into the cationic
(ArNdC(An)-C(An)dNAr)Ni(CH3)(C2H4)+ complexes 3 (Ar
) 2,6-C6H3(Me)2) and 21 (Ar ) 2,6-C6H3(iPr)2) provides a
convenient method for determination of the energy barriers for
insertion into Ni-Me species as well as subsequent ethylene
insertion.32 The latter barrier is (theoretically) a weighted average
of all ethylene insertions possible during the polymerization
(e.g., insertion into Ni-1° C, 2° C, and 3° C bonds). However,
the accessibility of the agostic complexes and ethylene-trapped
species described in the previous sections allows measurement
of the individual insertion barriers for each specific case with
both ligands to give a more complete analysis of the relative
energies of insertion versus isomerization.
spectrum taken at -100 °C indicated that the Ni ethyl
resonances for complex 20 appear at 1.23 (m) and -0.21 ppm
(t). With respect to the minor species, no agostic resonance was
1
observed (-12 to -14 ppm), but the H-1H COSY NMR
spectrum revealed a correlation between a broad resonance at
1.56 ppm (tentatively NiCH2) and a multiplet at 0.18 ppm
(agostic methyl), consistent with 18. A more definitive identi-
fication of diethyl ether coordination to a similar cationic
palladium ethyl species has recently been obtained.35
Recent work with palladium analogues of 18 have revealed
that diethyl ether competes with the agostic hydrogen for the
vacant metal coordination site. However, substitution of steri-
cally bulkier diisopropyl ether for diethyl ether resulted in clean
formation of the palladium â-agostic ethyl complex, with no
evidence for a palladium ethyl diisopropyl ether complex. Using
this same approach, formation of the ether adduct can also be
suppressed in the nickel systems. Substitution of [H(OiPr2)2]+-
[BAr′4]- for [H(OEt2)2]+[BAr′4]- under the synthetic conditions
shown in Scheme 12 results in the formation of 18 as the major
reaction product. The agostic hydrogen for complex 18 appears
1
at -13.7 ppm (br t, J ≈ 20 Hz, CDCl2F, -120 °C) in the H
NMR spectrum and is coupled to the two nonagostic protons
of the methyl group at 0.39 ppm and weakly to the two
methylene protons at 1.28 ppm. Some of the diethyl ether adduct
20 was also observed. Complete removal of diethyl ether from
complex 9 has proven difficult due to the extreme thermal
sensitivity of 9. Therefore, the diethyl ether adduct 20 presum-
ably forms as a consequence of the residual diethyl ether
competing with the agostic hydrogen for the vacant coordination
site.
As shown in Scheme 10, the (2,6-Me2)Ar(R-diimine)Ni(n-
propyl)ethylene complex n-13, generated by protonation of the
dipropyl complex in the presence of ethylene, exhibits migratory
insertion without observable isomerization at temperatures above
-100 °C. The energy barrier for this process was determined
to be 13.4 kcal/mol at -92 °C. Isopropyl ethylene complex i-13
undergoes ethylene insertion with a barrier of 14.2 kcal/mol at
-82 °C. Thus, there is a distinct difference between barriers to
ethylene insertion into a primary and secondary carbon center.
Comparisons to previously determined barriers will be addressed
in the following section.
Two distinct fluxional processes in 18 can be quantified using
dynamic NMR analysis. At temperatures above -100 °C, the
resonances of the agostic and nonagostic methyl protons begin
to broaden, indicating facile rotation about the CR-Câ bond
and scrambling of the three protons. The barrier for this process
was determined based on broadening of both resonances for
the agostic and nonagostic protons, and both sets of data provide
∆Gq ) 8.7 kcal/mol at -85 °C. At more elevated temperatures
(ca. -40 °C), the resonances for the three Câ protons coalesce
at ca. δ -3.0 ppm. This resonance sharpens with increasing
temperature, then begins to broaden at 10 °C due to exchange
with the protons on CR. Quantification of the broadening of the
Migratory insertion rates for complexes with the more
sterically demanding (2,6-iPr)Ar(R-diimine) ligand were deter-
mined in a similar fashion. However, the agostic ethyl complex
9
J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003 3075

A R T I C L E S
Leatherman et al.
Scheme 13. Generation and Subsequent Insertion of
Scheme 14. Generic Model for Metal Migration (Chain Walking)
Ni(ethyl)ethylene Cation 22 and Ni(isopropyl)ethylene Cation 23
Table 2. Migratory Insertion Barriers of Ni Alkyl Ethylene
Complexesa
q
complex
k (temp, °C)
∆G , kcal/mol
3
1.2 × 10-3 s-1 (-80.9)
2.3 × 10-4 s-1 (-92.2)
2.5 × 10-4 s-1 (-81.6)
3.1 × 10-4 s-1 (-92.9)
7.6 × 10-4 s-1 (-93.5)
2.2 × 10-4 s-1 (-88.5)
13.6
13.4
14.2
13.3
12.9
13.7
n-13
i-13
21
22
23
a Errors in ∆Gq measurements are (0.1 kcal/mol (see Experimental
Section).
19 was employed for measurement of olefin insertion into a
primary carbon since it does not exhibit ether binding, which
could be competitive with association of ethylene. This complex
also precludes any complications due to alkyl isomerization prior
to olefin insertion. Introduction of 5 equiv of C2H4 to a sample
of 19 generated in situ in an NMR tube at -130 °C yields the
Ni(ethyl)ethylene cation 22 quantitatively (Scheme 13). The
characteristic alkyl resonances are observed at 0.85 ppm for
the methylene protons and 0.65 ppm for the terminal methyl
group. Olefin insertion in this species occurs with a ∆Gq of
12.9 kcal/mol at -94 °C. As in the previous case for the ortho-
methyl ligand, treatment of the Ni(isopropyl) agostic cation 12
with excess ethylene at -130 °C generates the Ni(isopropyl)-
ethylene complex 23. This species exhibits behavior similar to
i-13 in that no isomerization to the linear alkyl-ethylene cation
is observed prior to insertion, and this insertion barrier (∆Gq )
13.7 kcal/mol at -89 °C) is significantly greater than that for
the linear species.
The influence of the ligand structure and the nature of the
migrating alkyl group on insertion barriers can be discerned
from the data summarized in Table 2. The primary alkyl groups
(n-propyl for n-13 and ethyl for 22) exhibit barriers ca. 0.8 kcal/
mol less than the corresponding barriers for the migration of
the secondary isopropyl group (compare n-13 to i-13 and 22 to
23). Ethylene insertion barriers for the bulkier diisopropyl(aryl)-
substituted complexes 22 and 23 are ca. 0.5 kcal/mol lower than
their dimethyl(aryl) counterparts n-13 and i-13 for insertions
into both primary and secondary carbon centers. The lower
barriers for the diisopropyl(aryl)-substituted systems account for
the somewhat higher activity of these more hindered catalysts.
These same trends are evident for the Pd complexes and are
best rationalized by proposing that the steric interactions in the
alkyl olefin ground states are more severe than in the transition
states where the Ni-alkyl bond is lengthened, and thus the
ground-state/transition-state energy difference is less for the
bulkier diisopropyl(aryl)-substituted systems. Theoretical cal-
culations have fairly accurately modeled the insertion barriers
in these systems and the observed substituent effects.36,37,45,46
Discussion
The mechanistic studies described here for (R-diimine)Ni
catalysts, coupled with previous mechanistic studies of the Pd
analogues, provide a quantitative comparison of the dynamics
and reactivity of key intermediates in the ethylene polymeri-
zation reactions. The mechanistic data account for the differing
rates of chain propagation and polymer microstructures observed
for the two systems. The discussion below highlights the key
differences between the Pd and Ni systems.
Insertion Barriers. Ethylene insertion barriers for the nickel
alkyl ethylene complexes are 4-5 kcal/mol lower than their
palladium congeners, accounting for the significantly greater
turnover frequency observed in the nickel systems. With a
barrier difference of 4-5 kcal/mol, turnover frequencies of Ni
systems should be ca. 104 times greater than turnover frequencies
for Pd analogues at 25 °C. This is in line with bulk polymer-
ization results.13
Metal Migration along the Polyethylene Chain (Chain
Walking). Migration of the metal along the alkyl chain occurs
by the processes shown in Scheme 14. The metal can move
between C1 and C2 (24 to 25) via â-elimination and readdition.47
(45) Deng, L.; Margl, P. M.; Ziegler, T. J. Am. Chem. Soc. 1997, 119, 1094-
1100.
(46) Froese, R. D. J.; Musaev, D. G.; Morokuma, K. J. Am. Chem. Soc. 1998,
120, 1581-1587.
9
3076 J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003

Nickel(II) Alkyl Agostic Cations
A R T I C L E S
Scheme 15. Mechanism for Formation of Branched Polyethylenes
Movement from C1 to C3 (24 to 27) requires breaking the H2
agostic interaction, rotation about M-C1, formation of an
agostic interaction with H3 (24 to 26), followed by â-elimina-
tion/readdition (26 to 27). In the case of palladium, â-elimination
is the lower energy barrier (∆Gq ≈ 7 kcal/mol). The barrier to
rotation about Pd-C1 is estimated as ca. 9-10 kcal/mol and is
thus the barrier that controls the rate of metal migration along
the chain. In the case of nickel, the Ni-C1 rotational barrier is
comparable to the Pd barrier (ca. 9.0-9.5 kcal/mol), but the
â-elimination barrier of 14.0 kcal/mol is higher than either Pd
barrier and controls the rate of chain migration of Ni. This
substantial difference in barriers of ca. 4-5 kcal/mol, resulting
in a much slower migration of Ni along the chain, clearly has
an impact on the extent of branching in the polyethylenes and
is discussed in more detail below.
involves sequential insertion of two ethylene units into a
M-CH3 bond to form either a linear or branched C5 chain.
In the case of Pd, ethylene complexation is reversible and
complexes 30 and 32 are in rapid equilibrium prior to migratory
insertion.33 The fraction of branched product produced is
controlled by the equilibrium ratio of 32/30 times the ratio of
rate constants k5/k4 (Curtin-Hammett kinetics). Translating this
result to Scheme 1 implies that Pd can migrate over a large
number of carbons prior to insertion, and since migration
through tertiary carbon centers is facile,34 highly branched
polymers containing branches-on-branches (hyperbranched struc-
tures) are formed. At very high ethylene pressures the equilib-
rium ratio [(diimine)Pd(R)(C2H4)+]/[(diimine)PdR+ + C2H4]can
be shifted so far toward the alkyl ethylene complex that the
number of carbons, on average, over which Pd can migrate prior
to insertion can be significantly reduced. As shown by Guan et
al.,17 this results in a change of polymer architecture from a
hyperbranched structure to a more linear structure with fewer
branches-on-branches even though the total number of branches
per 1000 carbons remains essentially constant.
The Ni systems follow a similar mechanistic scheme, but, as
noted, branching is much reduced and branching density is
sensitive to temperature and ethylene pressure. At the low
temperatures used here for mechanistic work, virtually no
branching is observed. Referring to Scheme 15, no branched
C5 product is formed as experimentally demonstrated above.
This arises for two reasons. First, the â-elimination barrier, 29
to 31, is much higher than in the case of Pd, and trapping of 29
by ethylene is thus fast relative to formation of 31. Second,
once formed, the low insertion barrier of 30 ensures that loss
of ethylene to re-form 29 (and potentially equilibrate 30 and
32) does not occur (k-1 < k4). These arguments are further
supported by the observation that isopropyl ethylene complex
32 (formed by trapping of 31) does not form the isomeric and
more stable n-propyl ethylene complex 30 prior to insertion.
As the temperature is increased, the first-order rate of
â-elimination will increase faster than the second-order ethylene
trapping rate and thus favor increased branching (in Scheme
15, k2/k1[C2H4] increases with temperature). In addition, the ratio
of (diimine)NiR+/(diimine)Ni(R)(C2H4)+ will increase with
increasing temperature, favoring increased chain walking relative
to insertion. Indeed, in Ni systems the rate of chain growth at
higher temperatures shows some dependence on [C2H4] and
suggests that the catalyst resting state is a mix of (diimine)Ni-
(R)(C2H4)+ and (diimine)Ni(R)+. Furthermore, we have shown
that for the bulkier olefin propylene the resting state is the
When chain walking occurs from the alkyl ethylene com-
plexes, a question arises as to whether these isomerization
reactions occur through a five-coordinate intermediate or via
reversible ethylene loss to yield the highly dynamic metal alkyl
agostic complexes. For palladium complexes, the mechanism
was clearly shown to be ethylene loss followed by isomerization
of the agostic Pd alkyl species.33,34 Since nickel is more prone
to form five-coordinate species, confirmation of the ligand loss
mechanism was sought. Indeed, as with the Pd systems, this
mechanism was verified in the case of (diimine)NiR(L)+ (L )
CH3CN, Me2S) by demonstrating that chain isomerization is
retarded by added L, an observation inconsistent with a
nondissociative, five-coordinate intermediate. Since ethylene and
CH3CN have similar binding affinities, it can be reasonably
assumed that a dissociative mechanism applies in the case of L
) C2H4.
Formation of Branched Polyethylenes. At similar reaction
temperatures and ethylene pressures, palladium diimine catalysts
yield much more highly branched polymers than their nickel
analogues.12,13 In contrast to Pd catalysts, the branching density
in Ni systems is quite sensitive to temperature and ethylene
pressure. The data reported here coupled with earlier mechanistic
studies of the Pd systems provide a clear explanation for the
differences in behavior of the Ni and Pd catalysts. The general
mechanism for formation of branched polymers was shown
earlier in Scheme 1, but to simplify discussion of this scheme,
a simple model system is shown below in Scheme 15, which
(47) Density functional calculations suggest a concerted â-H elimination/olefin
rotation/reinsertion process rather than a stepwise mechanism: Deng, L.;
Margl, P.; Ziegler, T. J. Am. Chem. Soc. 1997, 119, 1094-1100.
9
J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003 3077

A R T I C L E S
Leatherman et al.
equipped with a magnetic stir bar and flame-dried under vacuum. The
flask was charged with (diimine)NiBr2 5 (1.00 g, 1.65 mmol) in the
drybox and sealed with a rubber septum that had been previously dried
overnight at 100 °C. Dry diethyl ether (30 mL) was added to the flask
via syringe. The resulting red-brown slurry was stirred at room
temperature for 15 min, then cooled to -78 °C under argon. After 15
min, PrMgBr (1.65 mL of a 2.0 M Et2O solution, 3.30 mmol) was
slowly added dropwise. The reaction slurry immediately changed to a
very dark blue color. After 3 h at -78 °C, the stirring was ceased and
the reaction mixture allowed to settle for 10 min. A jacketed filter frit
was charged with Florisil (2 cm wide by 1 cm deep column), attached
to a Schlenk flask, and the entire assembly was flame-dried under
vacuum. Upon cooling, the filter frit assembly was back-filled with
argon, and the Florisil column and Schlenk flask were cooled to -78
°C. Dry diethyl ether (10 mL) was added to the top of the Florisil
column, the receiver flask was vented to an argon-purged oil bubbler,
and the reaction mixture was quickly transferred to the filter frit using
a 12-gauge cannula. It is critical to transfer the solution as fast as
possible; reductive elimination of hexane occurs if the solution warms
appreciably in the cannula. The entire solution was forced through the
Florisil as fast as possible under argon head pressure. The filtered
solution was filtered a second time through a filter cannula into another
flame-dried Schlenk flask cooled to -78 °C, and the solvent was
removed under vacuum at ca. -40 °C. The flask was quickly taken
into the drybox, and the dark solid purple product was recovered (610
mg, 69%). This manipulation of the product was performed at room
temperature, but as quickly as practical. The product is stable when
stored below -15 °C in the drybox, but decomposes within an hour at
room temperature. CD2Cl2 solutions of the product decompose in
minutes at room temperature, turning from the characteristic blue color
of the desired nickel dipropyl complex to a purple color. The structure
of the complex(es) formed at this point are unknown. Exposure of the
purple solution to air results in the immediate formation of a black
heterogeneous solution. In addition, the isolated nickel dipropyl complex
is extremely air-sensitive and decomposes immediately upon exposure
to oxygen, forming a black solid. The product was characterized by
agostic Ni alkyl complex, even at very low temperatures.
Decreasing ethylene pressure increases the extent of branching
for obvious reasons. The rate of chain walking relative to the
rate of trapping (k1[C2H4] vs k2) and the ratio of (diimine)Ni-
(R)+ to (diimine)Ni(R)(C2H4)+ is increased as ethylene pressure
is decreased.
Increased bulk of the diimine ligand (e.g., the diisopropyl-
(aryl)- vs the dimethyl(aryl)-substituted ligand) favors increased
branching. This is likely due to a decreased trapping rate of the
more sterically hindered alkyl complex relative to the rate of
chain walking. Theoretical calculations support decreased eth-
ylene trapping rates of complexes containing sterically bulky
ligands.48
Experimental Section
General Methods. All manipulations of air- and/or water-sensitive
compounds were performed using standard high-vacuum or Schlenk
techniques. Argon was purified by passage through columns of BASF
R3-11 catalyst (Chemalog) and 4 Å molecular sieves. Solid organo-
metallic compounds were transferred in argon-filled Vacuum Atmo-
spheres or MBraun dryboxes. 1H and 13C NMR spectra were recorded
on Bruker Avance 300, 400, or 500 MHz spectrometers. Chemical shifts
are reported relative to residual CHDCl2 (δ 5.32 for 1H), CHCl2F (1H,
1
doublet centered at δ 7.47, J ) 50 Hz), CHClF2 (δ ) 7.16 for H),
CD2Cl2 (δ 53.80 for 13C), or CDCl3 in CDCl2F solutions (δ 77.00 for
13C).
Materials. Diethyl ether, pentane, and methylene chloride were
purified using procedures recently reported by Pangborn et al.49
Polymer-grade ethylene was purchased from National Specialty Gases
and used without further purification. The R-diimine ligands (ArNd
C(An)-C(An)dNAr, An,An ) acenaphthyl ) 1,8-naphthdiyl) and
corresponding (R-diimine)nickel(II) bromide complexes were prepared
according to modified literature procedures.13,50-52 Nickel methyl ether
adduct 2 was prepared as described previously.32 (DME)NiBr2,53
[H(OEt2)2]+[BAr′4]- (BAr′4- ) B(3,5-C6H3(CF3)2)4-),54 and [H(OiPr2)2]+-
[BAr′4]-55 were prepared according to literature procedures. Diisopropyl
ether was distilled from potassium/benzophenone ketyl under argon.
Dichlorofluoromethane-d was prepared according to the literature56 and
was purified by stirring over silica gel to remove silane byproducts,57
dried over CaH2, vacuum transferred, and degassed by repeated freeze-
pump-thaw cycles. Dichloromethane-d2 was dried over P2O5, vacuum
transferred, degassed by repeated freeze-pump-thaw cycles, and stored
over activated 4 Å molecular sieves. Acenaphthenequinone, 2,6-
dimethylaniline, 2,6-diisopropylaniline, EtMgBr, and PrMgBr were
purchased from Aldrich Chemical Co. and used as received. Anhydrous
acetonitrile and dimethyl sulfide were purchased from Aldrich Chemical
Co. and used as received. NaBAr′4 was purchased from Boulder
Scientific.
low-temperature H NMR, 13C NMR, H-1H COSY, and 13C DEPT
NMR experiments. 1H NMR (CD2Cl2, -80 °C, 300 MHz): δ 8.31 (d,
J ) 8.3 Hz, 2H, An Hp), 7.47-7.37 (m, 6H, ArH), 7.06 (apparent t
(overlapping doublets), 2H, An Hm), 6.65 (d, J ) 7.1 Hz, 2H, An Ho),
2.30 (br t, 4H, NiCH2), 2.22 (s, 12H, ArCH3), 0.53 (br s, 10H,
coincidental overlap of NiCH2CH2CH3 and NiCH2CH2CH3). 13C NMR
(CD2Cl2, -80 °C, 100 MHz): δ 162.7 (NdC-CdN), 149.2, 136.7,
133.4, 133.2, 130.1, 127.84, 127.78, 125.9, 125.4, 118.7, 23.3 (NiCH2
or NiCH2CH2), 18.06 (ArCH3), 18.01 (NiCH2CH2CH3), 15.9 (NiCH2
or NiCH2CH2).
1
1
(ArNdC(An)-C(An)dNAr)Ni(CH2CH2CH3)2 (Ar ) 2,6-C6H3-
(iPr)2), 8. This compound was prepared using the same procedure
described above for 7. (Diimine)NiBr2 6 (0.750 g, 1.04 mmol) and
PrMgBr (1.15 mL of a 2.0 M Et2O solution, 2.30 mmol) were used as
starting materials. Using the workup described above, 291 mg (43%)
of the purple solid product was recovered. Et2O and CH2Cl2 solutions
of the product are dark blue in color. Solutions of the product
decompose at temperatures above -20 °C, producing hexane and
unidentified products; the solution is purple and homogeneous. Exposure
of the purple solution to air results in immediate decomposition to a
black heterogeneous solution. The product was characterized by low-
temperature 1H, 13C, 1H-1H COSY, and 13C DEPT NMR experiments.
1H NMR (CD2Cl2, -60 °C, 400 MHz): δ 8.27 (d, J ) 8.0, 2H, An
Hp), 7.60 (br m, 2H, Ar Hp), 7.50 (br d, 4H, Ar Hm), 7.03 (apparent t
(overlapping doublets), 2H, An Hm), 6.49 (d, J ) 7.2, 2H, An Ho),
3.34 (br m, 4H, ArCHMeMe′), 2.86 (q, J ) 7.2, 4H, NiCH2CH2CH3),
1.36 (d, J ) 6.4, 12H, ArCHMeMe′), 0.71 (d, J ) 6.4, 12H,
ArCHMeMe′), 0.57 (t, J ) 6.4, 6H, NiCH2CH2CH3), 0.27 (br m, 4H,
NiCH2CH2CH3). 13C NMR (CD2Cl2, -60 °C, 100 MHz): δ 162.4 (Nd
C-CdN), 146.9, 138.6, 136.2, 134.1, 133.6, 130.1, 126.3, 125.7, 123.5,
Synthesis of Nickel Dialkyl Complexes. (ArNdC(An)-C(An)d
NAr)Ni(CH2CH2CH3)2 (Ar ) 2,6-C6H3(Me)2), 7. A Schlenk flask was
(48) Woo, T. K.; Blo¨chl, P. E.; Ziegler, T. J. Phys. Chem. A 2000, 104, 121-
129.
(49) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.; Timmers,
F. J. Organometallics 1996, 15, 1518-1520.
(50) van Asselt, R.; Elsevier, C. J.; Smeets, W. J. J.; Spek, A. L.; Benedix, R.
Recl. TraV. Chim. Pays-Bas 1994, 113, 88-98.
(51) Tom Dieck, H.; Svoboda, M.; Grieser, T. Z. Naturforsch. 1981, 36b, 823-
832.
(52) Svoboda, M.; Tom Dieck, H. J. Organomet. Chem. 1980, 191, 321-328.
(53) Ward, L. G. L. Inorg. Synth. 1971, 13, 154-164.
(54) Brookhart, M.; Grant, B.; Volpe, A. F. Organometallics 1992, 11, 3920-
3922.
(55) White, P. A.; Calabrese, J.; Theopold, K. H. Organometallics 1996, 15,
5473-5475.
(56) Siegel, J. S.; Anet, F. A. J. Org. Chem. 1988, 53, 2629-2630.
(57) A small amount of Me2SiF2 was occasionally observed as a byproduct of
the fluorination reaction. Formation of this product can be suppressed
through minimal use of silicone grease on glass joints used in the preparation
of CDCl2F.
9
3078 J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003

Nickel(II) Alkyl Agostic Cations
A R T I C L E S
1
119.7, 28.1 (ArCHMeMe′), 24.0 (NiCH2CH2CH3 or NiCH2CH2CH3),
23.3 (ArCHMeMe′), 23.0 (ArCHMeMe′), 17.7 (NiCH2CH2CH3), 16.7
(NiCH2CH2CH3 or NiCH2CH2CH3).
n-propyl species n-11. H NMR (CDCl2F, -130 °C, 400 MHz): δ
8.12 (apparent t (overlapping doublets), 2H, An Hp and Hp′), 7.85 (s,
8H, BAr4′), 7.55-7.27 (m, 8H, ArH), 7.53 (s, 4H, BAr4′), 6.89 (d, J )
7.3 Hz, 1H, An Ho), 6.74 (d, J ) 7.3 Hz, 1H, An Ho′), 2.43, 2.41,
2.37, 2.25 (all s, 3H each, ArCH3, ArCH3′, ArCH3′′, ArCH3′′′), 2.00
(m, 1H, NiCH), 0.26 (br m, 1H, CHH′(µ-H)), 0.11 (br m, 1H, CHH′-
(ArNdC(An)-C(An)dNAr)Ni(CH2CH3)2 (Ar ) 2,6-C6H3(Me)2),
9. This compound was prepared using the same procedure described
above for 7. (Diimine)NiBr2 5 (1.00 g, 1.65 mmol) and EtMgBr (1.65
mL of a 2.0 M Et2O solution, 3.30 mmol) were used as starting
materials. Using the workup described above, 610 mg (69%) of the
purple solid product was recovered. Et2O and CH2Cl2 solutions of the
product are dark blue in color. Solutions of the product decompose at
temperatures above -20 °C, producing butane and unidentified
products; the solution is purple and homogeneous. Exposure of the
purple solution to air results in immediate decomposition to a black
heterogeneous solution. The product was characterized by low-
temperature 1H, 13C, 1H-1H COSY, and 13C DEPT NMR experiments.
1H NMR (CD2Cl2, -80 °C, 400 MHz): δ 8.28 (d, J ) 8.0 Hz, 2H, An
Hp), 7.46-7.40 (m, 6H, ArH), 7.07 (apparent t (overlapping doublets),
2H, An Hm), 6.59 (d, J ) 7.0 Hz, 2H, An Ho), 2.23 (s, 12H, ArCH3),
2.12 (q, J ) 7.6 Hz, 4H, NiCH2), 0.22 (t, J ) 7.6 Hz, 6 H, NiCH2CH3).
13C NMR (CD2Cl2, -80 °C, 100 MHz): δ 163.5 (NdC-CdN), 148.8,
136.9, 133.0, 132.9, 130.0, 127.9 (2 overlapping signals), 126.1, 125.5,
118.9, 18.0 (ArCH3), 14.1 (NiCH2CH3), 6.4 (NiCH2).
(ArNdC(An)-C(An)dNAr)Ni(CH2CH3)2 (Ar ) 2,6-C6H3(iPr)2),
10. This compound was prepared using the same procedure described
above for 7. (Diimine)NiBr2 6 (0.750 g, 1.04 mmol) and EtMgBr (1.15
mL of a 2.0 M Et2O solution, 2.30 mmol) were used as starting
materials. Using the workup described above, 439 mg (68%) of the
purple solid product was recovered. Et2O and CH2Cl2 solutions of the
product are dark blue in color. Solutions of the product decompose at
temperatures above -20 °C, producing butane and unidentified
products; the solution is purple and homogeneous. Exposure of the
purple solution to air results in immediate decomposition to a black
heterogeneous solution. The product was characterized by low-
temperature 1H, 13C, 1H-1H COSY, and 13C DEPT NMR experiments.
1H NMR (CD2Cl2, -60 °C, 400 MHz): δ 8.27 (d, J ) 8.0, 2H, An
Hp), 7.61 (apparent t (overlapping doublets), 2H, Ar Hp), 7.49 (d, J )
10.4, 4H, Ar Hm), 7.04 (apparent t (overlapping doublets), 2H, An Hm),
6.51 (d, J ) 6.8, 2H, An Ho), 3.28 (septet, J ) 8.8, 4H, ArCHMeMe′),
2.67 (q, J ) 7.2, 4H, NiCH2CH3), 1.34 (d, J ) 5.6, 12H, ArCHMeMe′),
0.73 (d, J ) 6.0, 12H, ArCHMeMe′), 0.06 (t, 6H, NiCH2CH3). 13C
NMR (CD2Cl2, -60 °C, 100 MHz): δ 163.2 (NdC-CdN), 147.0,
138.6, 136.4, 133.9, 133.6, 130.1, 126.4, 125.9, 123.6, 119.8, 28.4
(ArCHMeMe′), 23.5 (ArCHMeMe′), 23.1 (ArCHMeMe′), 15.3 (NiC-
H2CH3), 6.3 (NiCH2CH3).
2
(µ-H)), -0.09 (br s, 3H, CH3), -12.49 (t, JHH ) 19 Hz, 1H, µ-H).
n-propyl â-agostic species n-11: δ -13.0 (d, JHH ) 19 Hz).
[(ArNdC(An)-C(An)dNAr)Ni(CH(CH2-µ-H)(CH3))]+[BAr4′]-
2
i
(Ar ) 2,6-C6H3 Pr2), 12. A flame-dried Schlenk tube was charged with
the nickel dipropyl complex 8 (100 mg, 0.155 mmol) and [H(OEt2)2]+-
[BAr′4]- (185 mg, 0.183 mmol) in the drybox. The tube was sealed
and cooled to -80 °C under Ar, and CH2Cl2 (20 mL) was added slowly
on the sides of the tube. The dark blue-black, homogeneous reaction
solution was stirred at -80 °C for 30 min, then allowed to warm slowly
to -20 °C. All solvent was removed in vacuo at -20 °C, and the
maroon solid was dried in vacuo at 0 °C for 2 h, recovered in the dry-
box (138 mg, 61%), and stored at -35 °C. The product was charac-
terized by low-temperature 1H NMR and 1H-1H COSY NMR. 1H NMR
(CDCl2F, -120 °C, 500 MHz): δ 8.08 (apparent t (overlapping doub-
lets), 2H, An Hp and Hp′), 7.80 (s, 8H, BAr4′), 7.55-7.42 (m, 8H, ArH),
7.53 (s, 4H, BAr4′), 6.83 (d, J ) 5.6 Hz, 1H, An Ho), 6.78 (d, J ) 6.0
Hz, 1H, An Ho′), 3.47, 3.38, 3.18, 3.09 (all br septets, 1H each, Ar-
CH(CH3)2, ArCH′(CH3)2, ArCH′′(CH3)2, ArCH′′′(CH3)2), 2.16 (septet,
J ) 5.2 Hz, 1H, NiCH), 1.56, 1.50, 1.44, 1.40, 1.14, 1.10, 1.04, 0.90
(all d, 3H each, ArCH(CH3)2), 0.67 (br m, 2H, CHH′(µ-H) and CHH′-
2
(µ-H)), -0.22 (br d, 3H, Ni(CHCH3)), -12.49 (t, JHH ) 18 Hz, 1H,
µ-H).
Rates of Alkyl Isomerization and Migratory Insertion. Rates for
alkyl isomerization of Ni(â-agostic alkyl) cations were measured by
1H NMR line broadening techniques using the slow exchange ap-
proximation at various temperatures. NMR probe temperatures were
calibrated using an Omega type T thermocouple immersed in anhydrous
methanol in a 5 mm NMR tube. Activation parameters (∆Gq) were
determined using the Eyring equation assuming an estimate of 10%
error in k and (0.1 °C in temperature. Rates for alkyl isomerization of
Ni(iPr)L complexes were determined by monitoring the loss of the
1
NiCH(CH3)2 resonances in the H NMR spectra over time, as well as
the appearance of the NiCH2CH2CH3 resonance, using the BAr′4 p-H
signal as an internal reference. The natural logarithm of the number of
equivalents of the NiCH(CH3)2 complex was plotted versus time (first-
order treatment) to obtain kinetic plots (see Supporting Information).
Rates for ethylene insertion were determined by monitoring the loss
of the NiCH(CH3)2, NiCH2CH2CH3, or NiCH2CH3 resonance as
appropriate, using the BAr′4 p-H signal as an internal reference. The
natural logarithm of the starting alkyl ethylene complex resonance was
plotted versus time (first-order treatment) to obtain kinetic plots (see
Supporting Information).
Synthesis of â-Agostic Isopropyl Complexes. [(ArNdC(An)-C-
(An)dNAr)Ni(CH(CH2-µ-H)(CH3))]+[BAr4′]- (Ar ) 2,6-C6H3Me2),
i-11. Two Schlenk tubes were equipped with stir bars, fitted with rubber
septa, flame-dried, and taken into a drybox. The nickel dipropyl complex
7 (101 mg, 0.19 mmol) was added to one tube, and [H(OEt2)2]+[BAr′4]-
(202 mg, 0.20 mmol) was added to the other. The tubes were sealed,
placed on a Schlenk line, and cooled to -78 °C. After 5 min, 10 mL
of dry CH2Cl2 was slowly added to each tube at -78 °C. The solutions
were stirred to fully dissolve the solids (15 min), and the oxonium
acid solution was then cooled to -100 °C (pentane/liquid N2 slurry).
The solution of the nickel dipropyl complex 7 was quickly added to
the stirred solution of the oxonium acid via cannula. The reaction was
initially the deep blue color of the nickel dipropyl complex. The reaction
mixture was allowed to warm to -78 °C over 30 min and slowly turned
a darker midnight-blue color. The reaction was stirred at -78 °C for
2 h, during which time no apparent color change was observed. The
reaction was allowed to warm slowly to -30 °C, and all solvent was
removed in vacuo at -30 °C. The maroon-black solid was dried in
vacuo at -20 °C for 2 h, recovered in the drybox (221 mg, 86%), and
stored at -35 °C. The product was characterized by low-temperature
Synthesis of Cationic Nickel Alkyl(L) Complexes (L ) NCCH3,
S(CH3)2). [(ArNdC(An)-C(An)dNAr)NiCH(CH3)2)(NCCH3)]+-
[BAr4′]- (Ar ) 2,6-C6H3Me2), i-16. â-Agostic complex i-11 (13.9 mg,
0.0103 mmol) was quickly added to an NMR tube in the drybox. The
tube was sealed with a septum, wrapped with Parafilm, and cooled to
-80 °C. CD2Cl2 (0.6 mL) was added via gastight syringe, followed by
a solution of NCCH3 in CD2Cl2 (0.11 mL of a 0.187 M solution, 2
equiv of NCCH3). The tube was shaken briefly to dissolve the solids,
resulting in an immediate color change of the solution from the maroon
of the free agostic cation to deep blue. The structure of i-16 was
confirmed by H and H-1H COSY experiments. H NMR (CD2Cl2,
-60 °C, 400 MHz): δ 8.13 (apparent t (overlapping doublets), 2H,
An Hp and Hp′), 7.70 (s, 8H, BAr4′), 7.53 (s, 4H, BAr4′), 7.48-7.28
(m, 8H, ArH), 6.92 (d, J ) 7.2 Hz, 1H, An Ho), 6.38 (d, J ) 6.8 Hz,
1H, An Ho′), 2.39 (s, 6H, ArCH3), 2.33 (s, 6H, ArCH3′), 1.80 (br s,
6H, NCCH3), 1.45 (m, 1H, Ni-CH), 0.27 (d, J ) 6.0 Hz, 6H, CH-
(CH3)2).
1
1
1
1
1H NMR and H-1H COSY NMR. At -120 °C, the product appears
to be a 20:1 mixture of the â-agostic isopropyl species i-11/â-agostic
9
J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003 3079

A R T I C L E S
Leatherman et al.
[(ArNdC(An)-C(An)dNAr)Ni(CH2CH2CH3)(NCCH3)]+-
[BAr4′]- (Ar ) 2,6-C6H3Me2), n-16. This complex is observed in situ
upon isomerization of i-16 at temperatures above -10 °C. An
equilibrium between the two complexes is slowly established, with a
final ratio of 1:3 i-16/n-16 (-10 °C). The structure of n-16 was
(d, J ) 7.5 Hz, 1H, An Ho), 6.70 (d, J ) 8.0 Hz, 1H, An Ho′), 1.45,
1.39, 1.25, 0.99 (all br d, 6H each, Ar(CH(CH3))), 0.35 (br m, 2H,
NiCH2CH2(µ-H)), -13.8 (t, J ) 19.0 Hz, 1H, NiCH2CH2(µ-H)). The
NiCH2CH2(µ-H) resonance is obscured by ligand Ar(CH(CH3)2) signals,
but was identified in the 1H-1H COSY spectrum (δ 1.42) via coupling
to both the agostic and nonagostic protons of the terminal methyl group.
Ethane, diisopropyl ether, and a small amount of residual [H(OiPr2)2]+-
[BAr′4]- are also present in the sample.
confirmed by H and H-1H COSY experiments. H NMR (CD2Cl2,
-60 °C, 400 MHz): δ 8.09 (apparent t (overlapping doublets), 2H,
An Hp and Hp′), 7.71 (s, 8H, BAr4′), 7.51 (s, 4H, BAr4′), 7.48-7.25
(m, 8H, ArH), 6.91 (d, J ) 7.2 Hz, 1H, An Ho), 6.31 (d, J ) 7.2 Hz,
1H, An Ho′), 2.34 (s, 6H, ArCH3), 2.26 (s, 6H, ArCH3′), 1.91 (br s,
6H, NCCH3), 1.00 (t, 2H, Ni-CH2), 0.69 (br t, 3H, NiCH2CH2CH3),
0.62 (m, 2H, NiCH2CH2CH3).
1
1
1
Synthesis of Alkyl Olefin Complexes. [(ArNdC(An)-C(An)d
NAr)NiCH(CH3)2)(C2H4)]+[BAr4′]- (Ar ) 2,6-C6H3Me2), i-13. â-A-
gostic isopropyl complex i-11 (∼13 mg, 10 µmol) was placed in an
NMR tube in the drybox. The tube was quickly capped with a rubber
septum, wrapped with Parafilm, and cooled to -130 °C. A solution of
CDCl2F/CDClF2 (1:1, ∼700 µL) was added to the tube via cannula.
Ethylene (0.5 mL, ca. 2 equiv) was added to the solution via gastight
syringe at -130 °C. The tube was briefly agitated to dissolve the
sample, and the resulting solution was red and homogeneous. The
sample was then transferred to a NMR probe precooled to -130 °C.
[(ArNdC(An)-C(An)dNAr)NiCH(CH3)2)(S(CH3)2)]+[BAr4′]- (Ar
) 2,6-C6H3Me2), i-17. â-Agostic complex i-11 (10.4 mg, 0.0077 mmol)
was quickly added to an NMR tube in the drybox. The tube was sealed
with a septum, wrapped with Parafilm, and cooled to -80 °C. CD2Cl2
(0.65 mL) was added via gastight syringe, followed by a solution of
SMe2 in CD2Cl2 (0.05 mL of a 0.154 M solution, 1 equiv of SMe2).
The tube was shaken briefly to dissolve the solids, resulting in an
immediate color change of the solution from the maroon of the free
agostic cation to emerald green. The structure of i-17 was confirmed
1
1
The structure of i-13 was confirmed by H and H-1H COSY NMR
experiments. No agostic hydrogen resonances were observed at -130
1
°C. H NMR (CDCl2F/CDClF2 (1:1), -130 °C, 500 MHz): δ 8.09
1
1
1
by H and H-1H COSY experiments. H NMR (CD2Cl2, -60 °C,
400 MHz): δ 8.10 (apparent t (overlapping doublets), 2H, An Hp and
Hp′), 7.71 (s, 8H, BAr4′), 7.51 (s, 4H, BAr4′), 7.46-7.27 (m, 8H, ArH),
6.62 (d, J ) 7.2 Hz, 1H, An Ho), 6.36 (d, J ) 7.2 Hz, 1H, An Ho′),
2.42 (s, 6H, ArCH3), 2.33 (s, 6H, ArCH3′), 1.86 (br s, 6H, S(CH3)2),
1.52 (m, 1H, Ni-CH), -0.07 (d, J ) 6.0 Hz, 6H, CH(CH3)2).
[(ArNdC(An)-C(An)dNAr)Ni(CH2CH2CH3)(S(CH3)2)]+-
[BAr4′]- (Ar ) 2,6-C6H3Me2), n-17. This complex exists in a mixture
with the isopropyl complex i-17 upon SMe2 trapping of i-11 below
-45 °C; at higher temperatures, the isopropyl species i-17 isomerizes
completely to yield n-17, confirmed by 1H and 1H-1H COSY
(apparent t (overlapping doublets), 2H, An Hp and Hp′), 7.83 (s, 8H,
BAr4′), 7.49 (s, 4H, BAr4′), 7.51-7.30 (m, 8H, ArH), 6.70 (apparent
t (overlapping doublets), 2H, An Ho and Ho′), 3.96 (d, J ) 14 Hz, 2H,
bound CHH′dCHH′), 3.53 (d, J ) 14 Hz, 2H, bound CHH′dCHH′),
2.28 (s, 6H, ArCH3), 2.23 (s, 6H, ArCH3′), 1.47 (m, 1H, Ni-CH),
0.40 (br s, 6H, CH(CH3)2).
[(ArNdC(An)-C(An)dNAr)Ni(CH2CH2CH3)(C2H4)]+[BAr4′]- (Ar
) 2,6-C6H3Me2), n-13. Nickel dipropyl complex 8 (6.5 mg, 1.2 X 10-5
mol) and [H(OEt2)2]+[BAr′4]- (11.6 mg, 1.2 × 10-5 mol) were
combined as solids in an NMR tube in the drybox. The tube was quickly
sealed with a rubber septum, wrapped with Parafilm, and cooled to
-130 °C. Ethylene (1.0 mL, ∼4 equiv) was added to the tube via
gastight syringe. The tube was briefly immersed in liquid nitrogen,
and a mixture of CDCl2F/CDClF2 (1:1) was added via cannula. Initially,
a near-black solution was formed. The NMR tube was placed in the
NMR probe, which had been precooled to -130 °C. The presence of
the nickel n-propyl group was confirmed by low-temperature 1H NMR
and 1H-1H COSY NMR. No agostic hydrogen resonances were
observed at -130 °C. 1H NMR (CDCl2F/CDClF2 (1:1), -100 °C, 500
MHz): δ 8.11 (d, J ) 8.3 Hz, 1H, An Hp), 8.07 (d, J ) 8.3 Hz, 1H,
An Hp′), 7.84 (s, 8H, BAr4′), 7.56 (s, 4H, BAr4′), 7.54-7.29 (m, 8H,
ArH), 6.78 (d, J ) 7.3 Hz, 1H, An Ho), 6.71 (d, J ) 7.3 Hz, 1H, An
Ho′), 5.47 (very br, free C2H4), 4.17 (br d, 2H, bound CHH′dCHH′),
3.31 (br d, 2H, bound CHH′dCHH′), 2.38 (s, 6H, ArCH3), 2.29 (s,
6H, Ar′CH3), 1.22 (m, 2H, NiCH2CH2CH3), 0.61 (t, J ) 6.9 Hz, 3H,
NiCH2CH2CH3), 0.48 (t, J ) 7.5 Hz, 2H, NiCH2CH2CH3). Propane,
diethyl ether, and a small amount of residual [H(OEt2)2]+[BAr′4]- are
also present in the sample. The sample was observed to be a red
homogeneous solution upon removal from the NMR probe.
1
experiments. H NMR (CD2Cl2, -20 °C, 400 MHz): δ 8.14 (d, J )
8.0 Hz, 1H, An Hp), 8.11 (d, J ) 8.4 Hz, 1H, Hp′), 7.70 (s, 8H, BAr4′),
7.53 (s, 4H, BAr4′), 7.50-7.27 (m, 8H, ArH), 6.65 (d, J ) 7.2 Hz, 1H,
An Ho), 6.51 (d, J ) 7.2 Hz, 1H, An Ho′), 2.35 (s, 6H, ArCH3), 2.33
(s, 6H, ArCH3′), 1.93 (br s, 6H, S(CH3)2), 1.01 (m, 2H, Ni-CH2CH2-
CH3), 0.61 (br t, 3H, NiCH2CH2CH3), 0.47 (t, J ) 7.2 Hz, 2H, NiCH2-
CH2CH3).
Synthesis of â-Agostic Ethyl Complexes. [(ArNdC(An)-C(An)d
NAr)Ni(CH2(CH2-µ-H))]+[BAr4′]- (Ar ) 2,6-C6H3Me2), 18. Nickel
diethyl complex 9 (5.3 mg, 10.5 µmol) and [H(OiPr2)2][BAr′4] (11.2
mg, 10.5 µmol) were placed in a screw-cap NMR tube in the drybox.
A solution of CDCl2F (∼700 µL) was added to the tube via cannula at
-80 °C. The tube was briefly agitated to dissolve the sample, and the
resulting solution was dark maroon and homogeneous. The sample was
then transferred to a NMR probe precooled to -130 °C. The structure
of 18 was confirmed by 1H and 1H-1H COSY NMR experiments. 1H
NMR (CDCl2F, -120 °C, 400 MHz): δ 8.08 (apparent t (overlapping
doublets), 2H, An Hp and Hp′), 7.82 (s, 8H, BAr4′), 7.55 (s, 4H, BAr4′),
7.42-7.35 (m, 8H, ArH), 6.90 (d, J ) 6.8 Hz, 1H, An Ho), 6.71 (d, J
) 6.8 Hz, 1H, An Ho′), 2.41 (s, 6 H, ArCH3), 2.37 (s, 6 H, ArCH′3),
1.22 (br m, 2 H, NiCH2CH2(µ-H)), 0.41 (br m, 2H, NiCH2CH2(µ-H)),
-13.7 (br t, 1H, NiCH2CH2(µ-H)). Ethane and diisopropyl ether are
also present in the sample.
[(ArNdC(An)-C(An)dNAr)Ni(CH2CH3)(C2H4)]+[BAr4′]- (Ar )
i
2,6-C6H3 Pr2), 22. Nickel diethyl complex 13 (6.0 mg, 9.7 µmol) and
[H(OiPr2)2][BAr′4] (14.3 mg, 13.4 µmol) were placed in a screw-cap
NMR tube in the drybox. A solution of CDCl2F (∼700 µL) was added
to the tube via cannula at -80 °C. The tube was briefly agitated to
dissolve the sample, and the resulting solution was dark maroon and
homogeneous. Ethylene (2.4 mL, ca. 10 equiv) was added to the solution
via gastight syringe at -130 °C. The sample was then transferred to a
NMR probe precooled to -130 °C. The structure of 22 was confirmed
by 1H and 1H-1H COSY NMR experiments. No agostic hydrogen
[(ArNdC(An)-C(An)dNAr)Ni(CH2(CH2-µ-H))]+[BAr4′]- (Ar )
i
2,6-C6H3 Pr2), 19. Nickel diethyl complex 10 (7.5 mg, 12.1 µmol) and
[H(OiPr2)2][BAr′4] (14.6 mg, 13.6 µmol) were placed in a screw-cap
NMR tube in the drybox. A solution of CDCl2F (∼700 µL) was added
to the tube via cannula at -80 °C. The tube was briefly agitated to
dissolve the sample, and the resulting solution was dark maroon and
homogeneous. The sample was then transferred to a NMR probe
1
resonances were observed at -130 °C. H NMR (CDCl2F, -100 °C,
500 MHz): δ 8.02 (apparent t (overlapping doublets), 2H, An Hp and
Hp′), 7.80 (s, 8H, BAr4′), 7.54 (s, 4H, BAr4′), 7.50-7.42 (m, 8H, ArH),
6.57 (d, J ) 6.5 Hz, 1H, An Ho), 6.51 (d, J ) 6.5 Hz, 1H, An Ho′),
5.42 (very br, free C2H4), 3.95 (br d, 2H, bound CHH′dCHH′), 3.69
(br d, 2H, bound CHH′dCHH′), 3.32 (br m, 2H, ArCH(CH3)2), 3.07
1
precooled to -130 °C. The structure of 19 was confirmed by H and
1
1H-1H COSY NMR experiments. H NMR (CDCl2F, -130 °C, 500
MHz): δ 8.05 (apparent t (overlapping doublets), 2H, An Hp and Hp′),
7.70 (s, 8H, BAr4′), 7.52-7.42 (m, 8H, ArH), 7.51 (s, 4H, BAr4′), 6.84
9
3080 J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003

Nickel(II) Alkyl Agostic Cations
A R T I C L E S
(br septet, 2H, ArCH′(CH3)2), 1.42 (br d, 6H, ArCH(CH3)2), 1.33 (br
d, 6H, ArCH(CH′3)2), 0.93 (br, 12H, ArCH(CH′′3)2 and ArCH(CH′′′3)2),
0.85 (br m, 2H, NiCH2CH3), 0.65 (br t, 3H, NiCH2CH3). Ethane,
diisopropyl ether, and a small amount of residual [H(OiPr2)2]+[BAr′4]-
are also present in the sample.
bound CHH′dCHH′), 3.60 (d, J ) 14.5 Hz, 2H, bound CHH′dCHH′),
3.17 (br, 2H, ArCH(CH3)2), 2.80 (br, 2H, ArCH′(CH3)2), 1.38 (br d,
6H, ArCH(CH3)2), 1.29 (br d, 6H, ArCH(CH′3)2), 0.91 (br, 12H, ArCH-
(CH′′3)2 and ArCH(CH′′′3)2), 1.82 (m, 1H, Ni-CH), 0.42 (br d, 6H,
NiCH(CH3)2). Propane, diethyl ether, and a small amount of residual
[H(OEt2)2]+[BAr′4]- are also present in the sample.
[(ArNdC(An)-C(An)dNAr)NiCH(CH3)2)(C2H4)]+[BAr4′]- (Ar
i
) 2,6-C6H3 Pr2), 23. â-Agostic isopropyl complex 12 (9.0 mg, 6.1
Acknowledgment. The authors wish to thank the NSF (Grant
CHE-0107810) and DuPont for funding. S.A.S. acknowledges
the United States Air Force Institute of Technology Civilian
Institutions Program (AFIT/CI) for sponsorship.
µmol) was placed in an NMR tube in the drybox. The tube was quickly
capped with a rubber septum, wrapped with Parafilm, and cooled to
-130 °C. A solution of CDCl2F (∼700 µL) was added to the tube via
cannula. Ethylene (1.5 mL, ca. 10 equiv) was added to the solution via
gastight syringe at -130 °C. The tube was briefly agitated to dissolve
the sample, and the resulting solution was red and homogeneous. The
sample was then transferred to a NMR probe precooled to -130 °C.
Supporting Information Available: NMR spectra of repre-
sentative (R-diimine)Ni(dialkyl) complexes, Ni(â-agostic alkyl)
cations, Ni(alkyl)ethylene species, and Ni(alkyl)L complexes
and graphs of kinetic data for ethylene insertion and alkyl
isomerization. This material is available free of charge via the
Internet at http://pubs.acs.org.
1
1
The structure of 23 was confirmed by H and H-1H COSY NMR
experiments. No agostic hydrogen resonances were observed at -130
°C. 1H NMR (CDCl2F, -100 °C, 500 MHz): δ 8.02 (apparent t
(overlapping doublets), 2H, An Hp and Hp′), 7.78 (s, 8H, BAr4′), 7.54
(s, 4H, BAr4′), 7.49-7.42 (m, 8H, ArH), 6.49 (br d, 1H, An Ho), 6.45
′
(br d, 1H, An Ho ), 5.43 (very br, free C2H4), 4.16 (d, J ) 13.5 Hz, 2H,
JA021071W
9
J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003 3081