A Mechanistic Investigation of the Polymerization of Ethylene Catalyzed by Neutral Ni(II) Complexes Derived from Bulky Anilinotropone Ligands

Article abstract of DOI:10.1021/ja030634g

An extensive mechanistic investigation has been carried out on ethylene polymerizations catalyzed by neutral Ni(II) catalysts derived from bulky anilinotropone ligands. Complexes and precatalysts prepared include aryl derivatives [(2,6-i-Pr₂C₆H₃)NC ₇H₄O(7-Aryl)Ni(Ph)(PPh₃)] (9, Aryl = phenyl(a), 1-naphthyl(b), p-methoxyphenyl(c), p-trifluoromethylphenyl(d)), alkyl derivatives [[(2,6-ⁱPr₂C₆H₃)NC ₇H₅O]Ni(R)(2,4-lutidine)] (16, R = Et (a), n-Pr (b)) and [[(2,6-ⁱPr₂C₆H₃)NC₇H ₅O]Ni(R)(PPh₃)] (17, R = Et (a), n-Pr (b), n-hexyl (c), i-Pr (d)), and the nickel hydride complex [[(2,6-ⁱPr ₂C₆H₃)NC₇H ₅O]Ni(H)(PPh₃)], 20. Branched polyethylenes are produced at 40-80 °C in toluene with M_n values in the 100-200K range and molecular weight distributions of ca. 1.4-2.2. Branching ranges from 15 to 64 branches/1000 carbons depending on temperature and ethylene pressure. The electron-withdrawing -CF₃ substituent on the 7-aryl group increases activity but has little effect on branching and molecular weight. NMR experiments establish that in the case of the PPh₃-substituted systems, the catalyst rests as an equilibrating mixture of the alkyl phosphine and the alkyl ethylene complexes. At high ethylene pressures, the turnover frequency saturates, indicating that the equilibrium has shifted nearly completely to the alkyl olefin complex. Under these conditions, the barriers to migratory insertion were determined to be ca. 16-17 kcal/mol for 9a, 9c, 9d, and 16a. Extraction of 2,6-lutidine from complexes 16a,b yields highly dynamic β-agostic alkyl complexes [[(2,6-i-Pr₂C₆H ₃)NC₇H₅O]-Ni(Et)] 21 and [[(2,6-i-Pr ₂C₆H₃)NC₇H₅O]Ni(i-Pr)] 22. Free energy barriers to nickel-carbon bond rotation and β-hydride elimination of 11.1 and ca. 17 kcal/mol, respectively, were determined for 22. Themolysis of 17c at 50 °C generates hydride 20 and hexene and occurs by two pathways, one independent of [PPh₃] and one retarded by PPh ₃. At much slower rates, hydride 20 reductively eliminates free ligand, which ultimately generates a bis-ligand complex, 25. Catalyst decay under polymerization conditions was shown to occur by a similar process to generate free ligand and a bis-ligand complex formed by reaction of free ligand with an active catalyst species. The major chain transfer route is a simple P-elimination process, not chain transfer to monomer.

Full text of DOI:10.1021/ja030634g

Published on Web 04/20/2004
A Mechanistic Investigation of the Polymerization of Ethylene
Catalyzed by Neutral Ni(II) Complexes Derived from Bulky
Anilinotropone Ligands
Jason C. Jenkins and Maurice Brookhart*
Contribution from the Department of Chemistry, UniVersity of North Carolina at Chapel Hill,
Chapel Hill, North Carolina 27599
Received November 21, 2003; E-mail: mbrookhart@unc.edu
Abstract: An extensive mechanistic investigation has been carried out on ethylene polymerizations catalyzed
by neutral Ni(II) catalysts derived from bulky anilinotropone ligands. Complexes and precatalysts prepared
include aryl derivatives [(2,6-i-Pr2C6H3)NC7H4O(7-Aryl)Ni(Ph)(PPh3)] (9, Aryl ) phenyl(a), 1-naphthyl(b),
p-methoxyphenyl(c), p-trifluoromethylphenyl(d)), alkyl derivatives [[(2,6-iPr2C6H3)NC7H5O]Ni(R)(2,4-lutidine)]
(16, R ) Et (a), n-Pr (b)) and [[(2,6-iPr2C6H3)NC7H5O]Ni(R)(PPh3)] (17, R ) Et (a), n-Pr (b), n-hexyl (c),
i-Pr (d)), and the nickel hydride complex [[(2,6-iPr2C6H3)NC7H5O]Ni(H)(PPh3)], 20. Branched polyethylenes
are produced at 40-80 °C in toluene with Mn values in the 100-200K range and molecular weight
distributions of ca. 1.4-2.2. Branching ranges from 15 to 64 branches/1000 carbons depending on
temperature and ethylene pressure. The electron-withdrawing -CF3 substituent on the 7-aryl group increases
activity but has little effect on branching and molecular weight. NMR experiments establish that in the case
of the PPh3-substituted systems, the catalyst rests as an equilibrating mixture of the alkyl phosphine and
the alkyl ethylene complexes. At high ethylene pressures, the turnover frequency saturates, indicating that
the equilibrium has shifted nearly completely to the alkyl olefin complex. Under these conditions, the barriers
to migratory insertion were determined to be ca. 16-17 kcal/mol for 9a, 9c, 9d, and 16a. Extraction of
2,6-lutidine from complexes 16a,b yields highly dynamic â-agostic alkyl complexes [[(2,6-i-Pr2C6H3)NC7H5O]-
Ni(Et)] 21 and [[(2,6-i-Pr2C6H3)NC7H5O]Ni(i-Pr)] 22. Free energy barriers to nickel-carbon bond rotation
and â-hydride elimination of 11.1 and ca. 17 kcal/mol, respectively, were determined for 22. Themolysis of
17c at 50 °C generates hydride 20 and hexene and occurs by two pathways, one independent of [PPh3]
and one retarded by PPh3. At much slower rates, hydride 20 reductively eliminates free ligand, which
ultimately generates a bis-ligand complex, 25. Catalyst decay under polymerization conditions was shown
to occur by a similar process to generate free ligand and a bis-ligand complex formed by reaction of free
ligand with an active catalyst species. The major chain transfer route is a simple â-elimination process, not
chain transfer to monomer.
Introduction
Interest in late transition metal olefin polymerization catalysts
has mushroomed over the past decade.1-4 Work in the 1980s
showed that Ni(II) complexes containing modified SHOP
ligands catalyzed ethylene polymerization, but in general, such
systems exhibited modest activities and lifetimes and broad
molecular weight distributions.5-9 A major advance in the area
of late metal catalysis came with the discovery of Ni(II) and
Pd(II) catalysts derived from bulky aryl-substituted R-diimines
of general structure 1.10-16
These systems are not only highly effective ethylene polym-
erization catalysts but also convert R-olefins, trans-1,2-disub-
stituted olefins, and certain cyclic olefins to high-molar mass
The microstructures of the polymers obtained
are quite different than those observed for polymers prepared
materials.2,10-15,17
(1) Boffa, L. S.; Novak, B. M. Chem. ReV. 2000, 100, 1479.
(2) Ittel, S. D.; Johnson, L. K.; Brookhart, M. Chem. ReV. 2000, 100, 1169.
(3) Britovsek, G. J. P.; Gibson, V. C.; Wass, D. F. Angew. Chem., Int. Ed.
1999, 38, 428.
(4) Mecking, S. Coord. Chem. ReV. 2000, 203, 325.
(5) Keim, W.; Hoffman, B.; Lodewick, R.; Peiukert, M.; Schmitt, G. J. Mol.
Catal. 1979, 6, 79.
(10) Johnson, L. K.; Killian, C. M.; Brookhart, M. J. Am. Chem. Soc. 1995,
117, 6414.
(11) Killian, C. M.; Tempel, D. J.; Johnson, L. K.; Brookhart, M. J. Am. Chem.
Soc. 1996, 118, 11664.
(12) Gates, D. P.; Svejda, S. A.; Onate, E.; Killian, C. M.; Johnson, L. K.; White,
P. S.; Brookhart, M. Macromolecules 2000, 33, 2320.
(13) Johnson, L. K.; Mecking, S.; Brookhart, M. J. Am. Chem. Soc. 1996, 118,
267.
(14) Mecking, S.; Johnson, L. K.; Wang, L.; Brookhart, M. J. Am. Chem. Soc.
1998, 120, 888.
(15) Leatherman, M. D.; Brookhart, M. Macromolecules 2001, 34, 2748.
(6) Keim, W. Angew. Chem., Int. Ed. Engl. 1990, 29, 235.
(7) Peuckert, M.; Keim, W. Organometallics 1983, 2, 594.
(8) Ostoya Starzewski, K. A.; Witte, J. Angew. Chem., Int. Ed. Engl. 1985,
24, 599.
(9) Klabunde, U.; Ittel, S. D. J. Mol. Catal. 1987, 41, 123.
9
10.1021/ja030634g CCC: $27.50 © 2004 American Chemical Society
J. AM. CHEM. SOC. 2004, 126, 5827-5842
5827

A R T I C L E S
Jenkins and Brookhart
from early metal catalysts. For example, the polyethylenes
produced range from linear to highly branched to hyperbranched
depending on metal, ligand structure, ethylene pressure, and
temperature.10,12,18,19 Poly(R-olefin)s made from diimine cata-
lysts contain fewer branches than expected and are termed
“chain-straightened”.2,11
of polar monomers, renewed interest arose in developing neutral
(less electrophilic) nickel systems for use with polar monomers
and polar solvents. As a major design feature, most second-
generation neutral Ni catalysts have incorporated a bulky ortho-
substituted aryl imine moiety in analogy with catalysts 1a,b.
Groups from DuPont and Caltech have reported a family of
neutral nickel ethylene polymerization catalysts based on the
deprotonated salicylaldimine ligand in which the imine sub-
stituent is a bulky ortho-disubstituted aryl group.30-33 The
DuPont group has reported use of allyl complexes, 2a, as
initiators where the most active catalysts bear nitro substituents
on the aryl ring, whereas the Caltech group uses the phosphine
complexes, 2b,c as initiators. Catalyst 2c, bearing a bulky
9-anthracenyl substituent was found to be particularly productive
and has been used to copolymerize ethylene and functionalized
norbornenes.31
Extensive mechanistic20-24 and theoretical25-27 investigations
of the diimine catalysts have been reported. The key to forming
a high polymer is the presence of substituents at the ortho
positions of the aryl rings, which places bulk in the axial sites
of the square planar complexes and retards chain transfer. The
unusual branching features observed are a result of migration
of the metal along the polymer chain via â-elimination/
readdition (“chain walking”) prior to insertion. The dynamics
of these processes, the nature of the catalyst resting states, the
barriers to olefin insertions, and other mechanistic details have
been elucidated for a series of diimine Ni and Pd catalysts.22-24
In contrast to early metal catalysts, the late metal systems
are much less oxophilic and therefore potentially more compat-
ible with monomers bearing functional groups. This feature has
attracted considerable interest in that the ability to copolymerize
ethylene with functionalized monomers, especially readily
available vinyl monomers, would allow access to a wide range
of new, inexpensive ethylene copolymers. Limited success has
been achieved in this regard.28 For example, Pd(II) diimine
catalysts were shown to copolymerize ethylene and methyl
acrylate, but formation of a chelate complex involving the
carbonyl group greatly reduces the turnover frequency.14 More
recently, the Ni(II) diimine complexes were also shown to effect
this copolymerization but only at high temperatures, and again
turnover frequencies were low.29 Monomers possessing more
remote functional groups inaccessible to the metal via chain
walking are viable, while other monomers poison the diimine
systems.2,30,31
Recently, we have reported neutral nickel catalysts based on
both 2-anilinotropone, 3,34 and 2-anilinoperinaphthenone, 4.35
High initial turnover frequencies were observed with these
catalysts and substitution of the tropone backbone at the
7-position with bulky aryl substituents resulted in increased
polyethylene productivity at low temperatures. In addition, the
effects of substitution of the N-aryl ring with a variety of alkyl
and aryl groups and halides has been reported.36
In that the electrophilic nature of the cationic Ni and Pd
complexes still presents problems with respect to incorporation
(16) Brookhart, M. S.; Johnson, L. K.; Killian, C. M.; Arthur, S. D.; Feldman,
J.; McCord, E. F.; McLain, S. J.; Kreutzer, K. A.; Nennet, A. M. A.;
Coughlin, E. B.; Ittel, S. D.; Parthasarathy, A.; Tempel, D. J. WO Patent
Application 9623010 to DuPont, 1995.
(17) McLain, S. J.; Feldman, J.; McCord, E. F.; Garner, K. H.; Teasley, M. F.;
Coughlin, E. B.; Sweetman, K. J.; Johnson, L. K.; Brookhart, M.
Macromolecules 1998, 31, 6705.
(18) Cotts, P. M.; Guan, Z.; McCord, E.; McLain, S. Macromolecules 2000,
33, 6945.
(19) Guan, Z.; Cotts, P. M.; McCord, E. F.; McLain, S. J. Science 1999, 283,
2059.
(20) Tempel, D. J.; Brookhart, M. Organometallics 1998, 17, 2290.
(21) Svejda, S. A.; Johnson, L. K.; Brookhart, M. J. Am. Chem. Soc. 1999,
121, 10634.
In contrast to the cationic diimine-based nickel catalysts, there
is relatively little mechanistic information available for these
or any neutral nickel ethylene polymerization systems. In light
of the increasing interest in the neutral systems, it was the goal
of this work to develop a full mechanistic picture of a typical
neutral catalyst system. We report here in-depth studies of
ethylene polymerizations catalyzed by the anilinotropone com-
plexes, 3. Detailed information concerning the chain propagation
process, the barrier to ethylene insertion, the nature and
(22) Tempel, D. J.; Johnson, L. K.; Huff, R. L.; White, P. S.; Brookhart, M. J.
Am. Chem. Soc. 2000, 122, 6686.
(23) Shultz, L. H.; Tempel, D. J.; Brookhart. M. J. Am. Chem. Soc. 2001, 123,
11539.
(24) Leatherman, M. D.; Svejda, S. A.; Johnson, L. K.; Brookhart, M. J. Am.
Chem. Soc. 2003, 125, 3068.
(25) Chan, M. S. W.; Deng, L.; Ziegler, T. Organometallics 2000, 19, 2741.
(26) Deng, L.; Margl, P.; Ziegler, T. J. Am. Chem. Soc. 1997, 119, 1094.
(27) Deng, L.; Woo, T. K.; Cavallo, L.; Margl, P. M.; Ziegler, T. J. Am. Chem.
Soc. 1997, 119, 6177.
(28) Ethylene has been successfully copolymerized with vinyl silanes using both
nickel and palladium diimine catalysts: Johnson, L. K.; McLain, S. J.;
Sweetman, K. J.; Wang, Y.; Bennett, A. M. A.; Wang, L.; McCord, E. F.;
Ionkin, A.; Ittel, S. D.; Radzewich, C. E.; Schiffino, R. S. WO200344066,
2003.
(29) 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.
(30) Wang, C.; Friedrich, S.; Younkin, T. R.; Li, R. T.; Grubbs, R. H.; Bansleben,
D. A.; Day, M. W. Organometallics 1998, 17, 3149.
(31) Younkin, T. R.; Connor, E. F.; Henderson, J. I.; Friedrich, S. K.; Grubbs,
R. H.; Bansleben, D. A. Science 2000, 287, 460.
(32) Johnson, L. K.; Bennett, A. M. A.; Ittel, S. D.; Wang, L.; Parthasarathy,
A.; Hauptman, E.; Simpson, R. D.; Feldman, J.; Coughlin, E. B. WO 98/
30609, 1998.
(33) Connor, E. F.; Younkin, T. R.; Henderson, J. I.; Waltman, A. W.; Grubbs,
R. H. Chem. Commun. 2003, 2272.
(34) Hicks, F. A.; Brookhart, M. Organometallics 2001, 20, 3217.
(35) Jenkins, J. C.; Brookhart, M. Organometallics 2003, 22, 250.
(36) Hicks, F. A.; Jenkins, J. C.; Brookhart, M. Organometallics 2003, 22, 3533.
9
5828 J. AM. CHEM. SOC. VOL. 126, NO. 18, 2004

Polymerization of Ethylene
A R T I C L E S
Scheme 1 a
n
a Reaction conditions: (a) Bu4NOH, CH3I; (b) NBS; (c) ArB(OH)2, Pd(PPh3)4; (d) HCl/MeOH; (e) Tf2O, pyridine; (f) H2NAr, Pd2dba3, BINAP.
Scheme 2
dynamics of the intermediate Ni alkyl complexes, and the chain
transfer and catalyst decay processes are reported.
Ligands 8c,d were converted to the (2-anilino-7-aryltropone)-
nickel(II) complexes 9c,d using the procedure reported for the
preparation of (2-anilino-7-phenyltropone)nickel(PPh3)(Ph) (eq
1).36 The ligands were deprotonated in THF with NaH, and solid
(PPh3)2Ni(Ph)(Cl) was added and the reaction mixture stirred
for 1 h. Following filtration and recrystallization, complexes
9c,d were obtained in 63 and 29% yields, respectively.
Results and Discussion
Syntheses of Catalysts and Intermediates. 2-Anilino-7-
aryltropone Nickel(II)(phenyl)(PPh3) Complexes. Procedures
for the preparation of 2-(2,6-diisopropyl)anilino-7-phenyltropone
(8a) and 2-(2,6-diisopropylanilino-7-(1-naphthyl)tropone (8b)
ligands and their corresponding neutral nickel(II) catalysts, 9a
and 9b, have been reported.36 To investigate electronic effects
in these catalyst systems 2-(2,6-diisopropyl)anilino-7-(p-meth-
oxyphenyl)tropone (8c) and 2-(2,6-diisopropyl)anilino-7-(p-
trifluoromethylphenyl)tropone (8d) were synthesized using
procedures reported for the general preparation of 2-anilino-7-
aryltropone ligands (Scheme 1). Tropolone was converted to
methoxytropone37 and brominated with N-bromosuccinimide to
produce 2-methoxy-7-bromotropone.38 Either (p-methoxyphe-
nyl)boronic acid or (p-trifluoromethylphenyl)boronic acic were
coupled with 2-methoxy-7-bromotropone to yield 2-methoxy-
7-aryltropones 8c and 8d, respectively.39 Demethylation,39
conversion to the triflate with triflic anhydride,40 and coupling
with 2,6-diisopropylaniline under typical Buchwald-Hartwig
coupling conditions produced the 2-anilino-7-aryltropones 8c,d.36
The overall conversion proceeded in significantly better yields
for the methoxy-substituted ligand than the trifluoromethyl-
substituted ligand. The increased solubility of the trifluoro-
methyl-substituted intermediates led to lower yields in steps that
required recrystallization. Ligands 8c,d were characterized by
1H, 13C, and, for 8d, 19F NMR spectroscopy. The spectra differ
little from those reported for 2-anilino-7-phenyltropone, 8a36
(see Experimental Section).
1
Complexes 9c,d were characterized by H, 13C, 31P, and 19F
NMR (9d). Two distinct doublets at 1.2 and 1.0 ppm were
1
observed in the H NMR spectrum for both 9c,d, indicating
restricted rotation about the N-aryl bond, which renders each
methyl group of an isopropyl unit inequivalent. Additional NMR
data for 9c,d are summarized in the Experimental Section.
(2-Anilinotropone)Ni(II)(alkyl)(L) (L ) 2,4 Lutidine,
PPh3) Complexes. Syntheses of nickel alkyl complexes were
developed for several reasons. The nickel alkyl complexes are
intermediates in the ethylene polymerization reaction, they are
precursors to neutral nickel alkyl agostic complexes and are a
source of an anilinotropone nickel(II) complex with a less
strongly coordinated donor ligand, 2,4-lutidine. Warren has
reported the synthesis of a (â-diketiminate)nickel(2,4-lutidine)-
(ethyl) complex 12, which is in equilibrium with the agostic
(37) Tamburlin-Thumin, I.; Crozet, M. P.; Barriere, J. C. Synthesis 1999, 1149.
(38) Takeshita, H.; Mori, A. Synthesis 1986, 578.
(39) Piettre, S. R. et al. J. Med. Chem. 1997, 40, 4208.
(40) Echavarren, A. M.; Stille, J. K. J. Am. Chem. Soc. 1988, 110, 1557.
9
J. AM. CHEM. SOC. VOL. 126, NO. 18, 2004 5829

A R T I C L E S
Jenkins and Brookhart
Scheme 3
ethyl complex, 13 (Scheme 2).41 Upon treatment of 12 with
borontrifluoride diethyl etherate, the nickel-ethyl agostic
complex 13 was isolated and structurally characterized.42,43
Initially, an approach similar to Warren’s was applied to the
synthesis of an (anilinotropone)nickel(2,4-lutidine)(chloride)
complex, 14. Combination of Lu2NiCl2 with the sodium salt of
2-(2,6-diisopropyl)anilinotropone at a range of temperatures
(-78 to 20 °C) led to the predominant formation of the bis-
anilinotropone nickel complex commonly observed as the
primary polymerization decomposition product (see below).34
The lack of steric bulk of 2-anilinotropone compared to the
â-diketiminate ligand likely results in the rapid reaction of 2
equiv of the sodium salt with the nickel halide. To slow the
rate of reaction with the nickel center, 2-anilinotropone (rather
than its sodium salt) was added to Lu2NiCl2 in the presence of
excess triethylamine and produced 14 in high yield. Without
excess triethylamine, 14 decomposes prior to workup. Complex
14 is paramagnetic and was not characterized by NMR
spectroscopy. When 14 was combined with either ethylmag-
nesium chloride or propylmagnesium chloride in diethyl ether
at -78 °C, the desired (anilinotropone)nickel(lutidine)(alkyl)
complexes 16a (ethyl) and 16b (propyl) were formed as red
powders in moderate isolated yields (55-60%, Scheme 3).
Complexes 16a,b are both diamagnetic and were character-
ized by 1H and 13C NMR spectroscopy. The 1H NMR spectrum
of 16a is representative of both complexes (See Experimental
Section for complete data). Restricted rotation about the nickel-
lutidine nitrogen bond renders the methine protons of the
isopropyl groups inequivalent, and they appear as septets at 3.70
and 3.55 ppm. Additionally, the two methylene protons of the
ethyl group are diastereotopic and appear as distinct multiplets
at -0.3 and -0.5 ppm. The methyl resonance of the ethyl group
appears at -0.4 ppm, between the two resonances for the
diastereotopic protons.
conversion to the n-propyl complex 17b. The 1H NMR spectrum
of 17a is representative of this class of compounds. Complex
17a contains a plane of symmetry, and as expected, the
methylene protons of the ethyl complex exhibit a single
resonance at -0.5 ppm, while the methyl protons appear at -0.2
ppm. The methine protons of the isopropyl units are equivalent,
exhibiting a septet resonance at 3.6 ppm. The methyl groups of
the isopropyl unit are inequivalent, appearing as doublets at 1.3
and 1.0 ppm due to slow rotation about the nitrogen-aryl bond.
A nickel-hydride species has been proposed as an intermediate
in the ethylene oligomerization cycle of several neutral nickel
SHOP-type catalysts.31,44 Keim has reported the synthesis of
the nickel hydride complex, 19, by reaction of Ni(COD)2 with
phosphino alcohol 18 and tricyclohexylphosphine in toluene at
-10 °C (eq 2).44 A similar attempt was made to synthesize
(N,O)Ni(PPh3)(H) (20) (N,O ) 2,6-diisopropylanilinotropone)
by reaction of the anilinotropone ligand with Ni(COD)2 and
either PPh3 or PCy3. In both cases, the only product obtained
was the bis-anilinotropone nickel(II) complex (see below).
Successful synthesis of nickel hydride complex 20 was
achieved using a method similar to that employed for the nickel
alkyl complexes 16 and 17. Reaction of (N,O)Ni(PPh3)(Cl), 15,
with excess NaHB(OMe)3 at -78 °C led to clean conversion
to the desired nickel hydride (N,O)Ni(PPh3)(H), 20 (eq 3).
NaHB(OMe)3 was chosen as the hydride source because of the
known lack of reactivity of the B(OMe)3 byproduct. A nickel
hydride resonance was observed at -26.7 ppm with a JPH
coupling constant of 130 Hz, similar to the hydride resonance
for 19 at -25.0 ppm.44 The 31P NMR spectrum of 20 exhibited
a doublet at 31.8 ppm (JPH ) 123 Hz).
Nickel alkyl complexes with a PPh3 donor ligand were
synthesized to model intermediates in the ethylene polymeri-
zation reactions. Complexes 17a-c were obtained as orange
powders in the same manner as the nickel lutidine alkyl
complexes. The (anilinotropone)nickel(PPh3)(isopropyl) com-
plex, 17d, was isolated using techniques to prevent warming
above -25 °C. Filtration and workup at 20 °C led to clean
Chain Propagation Studies. Catalysts 9c,d were synthesized
to investigate electronic effects on ethylene polymerization. The
results of ethylene polymerization with p-methoxyphenyl-9c,
(41) Wiencko, H. L.; Kogut, E.; Warren, T. H. Inorg. Chim. Acta 2003, 345,
199.
(42) Warren, T. H.; Wienko, H. L.; Kogut, E.; Dai, X.; McDermott, J.
International Symposium on Olefin Metathesis, Cambridge, MA, 2001.
(43) Kogut, E.; Zeller, A.; Warren, T. H.; Strassner, T. Personal communication.
(44) Muller, U.; Keim, W.; Kruger, C.; Betz, P. Angew. Chem., Int. Ed. 1989,
28, 1011.
9
5830 J. AM. CHEM. SOC. VOL. 126, NO. 18, 2004

Polymerization of Ethylene
A R T I C L E S
Table 1. Ethylene Polymerization with 9a, 9c, and 9da
entry catalyst T (°C) psig t (min)
TON
M (×10-3
)
M /M branches/1000 C
n
w
n
1
2
3
4
5
6
7
8
9
9c
9c
9c
9a
9a
9a
9d
9d
9d
40 200 10
60 200 10 19 000
80 200 10 59 500
40 200 10
60 200 10 32 000
80 200 10 64 300
40 200 10
60 200 10 27 300
80 200 10 70 000
9100
202
220
100
251
194
87
242
235
102
1.4
1.9
2.1
1.3
2.1
2.2
1.4
1.9
1.8
16
34
64
34
41
76
15
31
64
9600
9400
a Reactions were run in 200 mL of toluene with 6.6 µmol of catalyst.
Scheme 4
phenyl-9a, and p-trifluoromethylphenyl-9d-substituted catalysts
are summarized in Table 1. The polymerizations were performed
in a water-cooled reactor at sufficiently low catalyst loadings
to avoid reaction exotherms greater than 2 °C. Therefore, reliable
comparisons of catalyst activity and lifetime are possible.45 At
40 and 60 °C, no clear trend is evident in overall catalyst
turnover numbers; however, at 60 °C the p-methoxyphenyl-
substituted catalyst 9c is noticeably less active than 9a or 9d
with TONs of 19 000, 32 000, and 27 000, respectively. At 80
°C, the turnover number clearly increases from 59 500 (9c) to
64 300 (9a) to 70 000 (9d) as the electron-withdrawing nature
of the anilinotropone ligand increases. When reaction times are
increased to 1 h, the trend is even more evident, with an increase
in turnover number from 72 000 (9c) to 87 700 (9a) to 110 800
(9d). No significant differences in Mn or molecular weight
distribution are observed for the three catalysts at any of the
polymerization temperatures investigated. As expected,10,12
branching increases with increasing temperature. The branching
numbers for the three catalysts are relatively uniform with
slightly higher values for 9a. Molecular weight distributions of
ca. 2 at 60 and 80 °C indicate that the molecular weights
obtained are chain-transfer limited. At 40 °C, the MWDs of ca.
1.4 indicate that, after 10 min, the chain-transfer-limited Mn is
not reached and that longer reaction times would yield higher
Mn values.
The reactions of ethylene with anilinotropone nickel(II)
complexes 3 and 9a were investigated by in situ NMR
spectroscopic studies at low temperatures. The rate of insertion
of ethylene into the Ni-phenyl bonds of both unsubstituted 3
and phenyl-substituted 9a to form, initially, the corresponding
(N,O)Ni(PPh3)CH2CH2Ph complexes, was determined by moni-
toring the decrease in concentrations of 3 and 9a. In a typical
experiment, the catalyst was dissolved in CD2Cl2 and then
cooled to -78 °C, and ethylene was injected into the NMR
sample, which was then inserted into a precooled NMR probe.
The resonances associated with the methyl and methine groups
of the isopropyl moiety were used to measure relative concen-
trations of the starting complex and single- and double-
(multiple) insertion products. The first-order rate constants, as
illustrated in Scheme 4, were measured as a function of
temperature and ethylene concentration and are summarized in
Table 2.
Table 2. Ethylene Insertion into the Ni-Ph Bond of Catalysts 3
and 9a (Catalyst Concentration of 10.6 mM in CD2Cl2)
T
C H [C H ]
k1(obs)
k1
2
4
2 4
entry
catalyst
(°C) (equiv) (mM) (×105 sec-1
)
(×104 M-1 sec-1
)
1
2
3
4
5
6
3
3
3
3
-10 16
-10 24
170 2.6
254 3.5
85 9.3
1.5
1.4
10.9
0
8
10 14
10 12
-10 12
149 34.4 (3.4)a 23.2 (2.3)a
128
3 + 5 equiv of PPh3
9a
128 3.0
2.5
a Subsequent insertion of ethylene into the nickel-alkyl bond
in entries 1 and 2, insertion rates exhibit a first-order dependence
on [C2H4]. As expected, rates accelerate with temperature,
increasing about an order of magnitude between -10 and 10
°C (compare entries 1 and 4). Insertion is dramatically inhibited
by [PPh3], as shown in entry 5. (Under these conditions the
NMR spectrum of 3 is unchanged, indicating that inhibition is
not due to formation of a five-coordinate bis-phosphine species.)
The first insertion of ethylene into the Ni-aryl bond of 3 is
significantly faster than the second insertion into the Ni-alkyl
bond of (N,O)Ni(CH2CH2Ph)(PPh3), and thus this first insertion
product builds up in high concentrations. Analysis of the kinetics
as two consecutive first-order reactions allows estimation of k2
(entry 4), which is an order of magnitude less than k1. (This
result implies that insertion into the Ni-phenyl bond does not
constitute a slow initiation step.)
The mechanistic scenario shown in Scheme 5 involving a
rapid pre-equilibrium strongly favoring the phosphine complex,
followed by a slow insertion step, is clearly supported by the
observation that the catalyst resting states are exclusively PPh3
complexes and the insertion rates are first-order in [CH2CH2]
and strongly inhibited by PPh3. The overall rate will be
controlled by both the equilibrium constant, Keq, and the
insertion rate constant, kins. Comparing entries 2 and 6 in Table
2 shows that the 7-phenyl-substituted system, 9a, inserts ethylene
at a rate slightly faster than that of unsubstituted 3 (i.e.,
correcting for [CH2CH2], k1(9a)/k1(3) ) 1.8.). Since k1 is
Even at high ethylene concentrations, only PPh3 complexes
are observed (no ethylene complexes are detected). As shown
(45) Typically, more electron-withdrawing ligands lead to higher polymerization
activities for neutral nickel catalysts. See, for example: Cavell, K. J.
Organomet. Chem. 1997, 544, 163. Johnson, L. K. et al. WO 98/30609,
1998.
9
J. AM. CHEM. SOC. VOL. 126, NO. 18, 2004 5831

A R T I C L E S
Jenkins and Brookhart
Scheme 5
Table 4. Ethylene Polymerization with 3 and 9b in the Presence
of PPh3 in 200 mL of Toluene and 7.4 µmol of Catalyst at 200
psig Ethylene
entry
catalyst
T (°C) t (min) TON M (×10-3
)
M /M branches
n
w
n
1
2
3
4
5
6
7
8
9
3
80
80
80
80
10 51 500
10 43 400
30 60 400
30 60 000
92
54
1.8
2.3
72
68
3 + 5 equiv of PPh3
3
3 + 5 equiv of PPh3
67
33
31
2.3
1.9
2.1
2.0
1.3
3.8
1.4
1.9
67
60
58
57
22
9
3 + 50 equiv of PPh3 80
3 + 50 equiv of PPh3 80
15
5600
60 16 400
3 + 50 equiv of PPh3 80 180 18 100
33
9b
9b
40
40
10 11200
60 50 600
250
831
77
Table 3. Exchange of PPh3 by P(p-Tolyl)3 at -14 °C in CD2Cl2
10 9b + 5 equiv of PPh3 40
11 9b + 5 equiv of PPh3 40
10
2700
13
9
60 13 300
382
[Ni]
(mM)
[P(p-tolyl)3]
(mM)
k[PPh3]
(s-1
k
entry
catalyst
)
(M-1 s-1
)
1
2
3
4
3
3
9a
9a
10.6
10.6
9.4
117
318
141
273
6.4 × 10-4
9.3 × 10-4
18.6 × 10-4
5.5 × 10-3
7.9 × 10-3
6.6 × 10-3
6.8 × 10-3
polymerization.46 After 10 min, the turnover number (TON) is
slightly reduced (compare entries 1 and 2); however, after 30
min, the turnover numbers are essentially identical (entries 3
and 4). Under these conditions of high temperature and high
ethylene concentration, the equilibrium (Scheme 5) must very
substantially favor the ethylene complex and be only slightly
shifted to the left by addition of 5 equiv of PPh3. At 50 equiv,
a substantial effect on the equilibrium is observed (entries 5-7).
For example, TONs drop from 51 500 in 10 min with no added
PPh3 to 5600 in 15 min with 50 equiv of added PPh3. These
results also suggest that the addition of PPh3 increases the
lifetime of 3. Typically, in the absence of PPh3 at 80 °C and
200 psig ethylene, TONs increase only slightly when reaction
times are increased from 10 to 60 min (52 400 to 62 100).36
However, with 50 equiv of added PPh3, the TON at 60 min
(16 400) increases 3-fold relative to the TON at 15 min. Further
polymerization (3 h) leads to only slightly increased TONs (entry
7). (Further discussion of catalyst lifetimes appears below.)
The effect of excess PPh3 at lower temperatures was
investigated with 7-naphthyl-substituted catalyst 9b. Due to the
long lifetime of 9b at 40 °C,36 the only effect of added PPh3
will be an effect on the equilibrium between Ni-PPh3 and Ni-
ethylene complexes and not on catalyst lifetime as with 3. In
contrast to the 80 °C polymerizations, at 40 °C, 5 equiv of added
PPh3 has a noticeable effect on activity. After a 60 min run
with no PPh3 added, the overall turnover number for 9b is
50 600 (entry 9). When 5 equiv of PPh3 are added, the turnover
number drops to 13 300 (entry 11). A similar drop in TON is
observed for the 10 min polymerization runs (entries 8 and 10).
These data suggest that at 40 °C, in the presence of 5 equiv of
PPh3, the equilibrium lies substantially in favor of the Ni-PPh3
complex.
25 × 10-4
9.4
proportional to Keqkins, the effect of aryl substitution on the
individual values of Keq and kins cannot be determined from these
data.
Scheme 5 requires exchange of ethylene for PPh3, and since
the ethylene complex cannot be observed, the nature of the
exchange, associative or dissociative, cannot be determined. To
model this process we examined displacement of PPh3 by P(p-
tol)3. Use of high concentrations of P(p-tol)3 shifts the equilib-
rium shown in eq 4 sufficiently far to the right that simple first-
order kinetics are observed.
Table 3 summarizes data for 3 and 9a at -14 °C. Rates are
first order in P(p-tol)3, and displacement is clearly associative.
These results are consistent with the expected mechanism of
ligand exchange at a 16-electron d8 square planar metal center
and suggest that ethylene exchanges with triphenylphospine by
an analogous associative mechanism. The addition of the
sterically bulky phenyl group in the 7-position of 9a has little
effect on the rate of phosphine exchange.
In addition to the activity and lifetime effects of added PPh3,
an influence on polymer molecular weight is also observed. With
excess PPh3, the number average molecular weight drops from
91.5K (no PPh3, entry 1) to 53.9K (5 equiv of PPh3, entry 2) to
33.1K (50 equiv PPh3, entry 5). The decrease in polymer
molecular weight is due to a decrease in the rate of chain
propagation, while the rate of chain transfer remains ap-
proximately constant (this will be explained in more detail
below). In all cases, added PPh3 has little effect on the molecular
weight distribution of the polymers, with all values equal to
ca. 2.0. The very high molecular weight for entry 9 (Mn ) 831K)
Several preparative-scale polymerization reactions were car-
ried out that support and amplify the mechanism suggested by
the NMR spectroscopic studies.
We first examined the effects of added triphenylphosphine
on ethylene polymerization. Standard conditions were employed
(200 psig CH2CH2, 7.4 µmol of catalyst, 200 mL of toluene) in
the presence of a specified amount of PPh3. Exotherms were
less than 2 °C. Reactions were performed with both the
unsubstituted catalyst 3 and the naphthyl-substituted catalyst
9b at 80 and 40°C, respectively. The results are summarized in
Table 4.
(46) Only 3-4 equiv of PPh3 is reported to completely shut down the
salicylaldimine catalysts 2a,b. See: Younkin, T. R.; Connor, E. F.;
Henderson, J. I.; Friedrich, S. K.; Grubbs, R. H.; Bansleben, D. A. Science
2000, 287, 460.
The addition of 5 equiv of added triphenylphosphine to
catalyst 3 at 80 °C and 200 psig ethylene has little effect on the
9
5832 J. AM. CHEM. SOC. VOL. 126, NO. 18, 2004

Polymerization of Ethylene
A R T I C L E S
constants for insertion and ∆Gq were determined to be 91 s-1
and 16.6 kcal/mol, respectively (entry 7). Similar studies led to
insertion barriers of 16.6 kcal/mol for the p-trifluoromethylphe-
nyl-substituted catalyst 9d (entries 15 and 16) and 17.1 kcal/
mol for the p-methoxyphenyl-substituted catalyst 9c (entry 14).
Table 5. Polymerization of Ethylene with 9a, 16a, 9d, and 9b at
Catalyst Loadings of 6.6, 7.5, 6.1, 6.6 µmol, Respectively, in 200
mL of Toluene
T
t
M
n
entry catalyst (°C)
psig
(min)
TON
(×10-3
)
M /M
branches
w
n
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
9a
9a
9a
9a
9a
9a
9a
16a
16a
16a
16a
16a
16a
9c
40
40
40
40
40
40
60
60
60
60
60
60
60
40
40
40
50
100
200
400
600
800
600
100
200
400
600
600
800
600
600
800
20
20
20
20
20
20
10
15
15
15
15
20
15
20
20
20
5500
10 300
16 300
20 500
23 600
22 900
54 400
16 200
35 000
37 800
43 500
55 700
41 700
9250
134
280
356
394
424
431
1.4
1.6
1.6
1.6
1.5
1.6
34
27
11
10
8
To determine the ethylene insertion barrier for the unsubsti-
tuted version of the anilinotropone catalyst, the 2,4-lutidine
complex, (2-anilnotropone)Ni(Lu)(Et), 16a, was employed.
Initial attempts to reach saturation with the parent PPh3-
substituted catalyst 3 were unsuccessful. At 60 °C, where
catalyst lifetime is sufficient for accurate activity determination,
saturation could not be achieved at reasonable pressures (<1000
psig). At 80 °C, the lifetime of 3 is insufficient for accurate
determination of catalyst activity. Therefore, a catalyst with the
less strongly coordinating 2,4-lutidine donor ligand was inves-
tigated at 60 °C. A similar increase in Mn and decrease in
polymer branching number (Table 5, entries 8-13) with
ethylene pressure are observed. Saturation behavior is again
observed with increasing ethylene pressure, and similar turnover
numbers at 600 and 800 psig show that saturation has been
reached. Turnover numbers of 43 500 and 55 700 for 15 and
20 min polymerizations (entries 11, 12) show that no significant
catalyst decomposition occurs in the first 20 min of reaction at
60 °C. These data lead to values for kins of 48 s-1 and for ∆Gq
of 17.0 kcal/mol.
7
24
61
47
33
30
82
159
211
265
245
305
376
524
531
3.1
2.3
2.2
2.1
2.2
1.9
1.5
1.5
1.5
24
10
8
9d
9d
18 900
16 900
7
The small differences in the insertion barriers for the
substituted and unsubstituted catalysts should not be over-
interpreted. Errors in catalyst sample preparation, polymer
collection, and polymerization temperature control could lead
to errors in ∆Gq of up to 0.5 kcal/mol. With that caveat in mind,
the insertion barriers range from 16 to 17 kcal/mol, with no
significant difference between the aryl-substituted and unsub-
stituted catalysts. These values are intermediate between the
values for the cationic diimine nickel catalysts (13.5-14.0 kcal/
mol)21,24 and the cationic diimine palladium catalysts (18.0-
18.5 kcal/mol).23
Figure 1. TON vs P(C2H4) using catalyst 9a at 40 °C.
leads to a viscous GPC sample, which is likely the cause of the
broad molecular weight distribution observed. The polymer
branching data suggest that polymerization temperature and
ethylene pressure, not PPh3 concentration, control the amount
of branching. Similar branching numbers are observed with and
without added PPh3.
In addition to determination of the value for kins, the
equilibrium constant, Keq (Scheme 6), between the nickel-PPh3
and nickel-ethylene complexes can be estimated. Keq can be
calculated from kins, kobs, [PPh3], and [CH2CH2]. The observed
rate constant is equal to the fraction of Ni complex, which sits
as the ethylene complex ([E]/[E] + [P]) times kins. At high
[PPh3], the concentration of [E] will be much less than [P], and
the expression simplifies to kobs ) kins[E]/[P] as shown in
Scheme 6. Further manipulation yields eq 5 for Keq. Since a
Ni-alkyl complex is a better model for the growing polymer
chain than a Ni-aryl complex, kinetics were carried out with
17a in a manner similar to the runs described earlier in Table
2. Excess [PPh3] and [CH2CH2] were added to the solution of
17a, and NMR spectroscopy established [P] . [E] (no ethylene
complex detected). Kinetic measurements of insertion rates were
made at 60 °C to match the temperature at which kins was
measured in the saturation experiments. (Details of kinetic runs
are summarized in the Experimental Section.) At 60 °C, when
[PPh3] ) 334 M and [CH2CH2] ) 153 M, kobs ) 7.8 × 10-4
sec-1. Using the value of kins ) 47 s-1 from the saturation
experiments, Keq can be estimated as 3.7 × 10-5, ∆G° ) 6.8
kcal/mol.47 Arguments based on model systems that suggest this
Keq value is in the expected range are presented in Supporting
Information.
The proposed mechanism predicts that at sufficiently high
ethylene pressures, the dominant resting state should be the
nickel ethylene complex. Under these conditions the rate of
chain growth will be independent of ethylene concentration and
simply equal to the rate of migratory insertion. This saturation
rate behavior was examined for several complexes, and data
are summarized in Table 5. Phenyl-substituted catalyst 9a, has
a long lifetime at 40 °C with only a slight drop in turnover
frequency from 10 to 60 min. Thus, polymerizations carried
out for 20 min at 40 °C with 9a should provide reliable values
of TOFs. The activity of 9a was screened at 40 °C (20 min
runs) from 50 to 800 psig ethylene. Increasing the ethylene
pressure causes an increase in polymer molecular weight from
133.9K (50 psig, entry 1, Table 5) to 431.1K (800 psig, entry
6, Table 5) and a decrease in the total number of branches per
1000 carbon atoms (34, 11, and 7 at 50, 200, and 800 psig,
respectively), consistent with earlier observations. As expected
for saturation behavior, the catalyst turnover number increases
with pressure and gradually levels off between 600 and 800
psig ethylene. A plot illustrating this behavior is shown in Figure
1. At saturation, a turnover number of 23 600 in 20 min (600
psig, entry 5) implies a kins value of 19 s-1 corresponding to a
free energy of activation of 16.1 kcal/mol. At 60 °C, the rate
9
J. AM. CHEM. SOC. VOL. 126, NO. 18, 2004 5833

A R T I C L E S
Jenkins and Brookhart
Scheme 6
Scheme 7
Agostic Complex Formation and Dynamics. Agostic alkyl
nickel complexes are key intermediates whose behavior controls
branching in the R-diimine nickel and palladium catalysts.10,12,23,24
Similar agostic alkyl complexes are also likely key intermediates
for the introduction of branches into the polymers produced with
the anilinotropone neutral nickel catalysts. To observe such
intermediates, we have adopted a method similar to one
developed by Warren,41-43 wherein 2,4-lutidine can be extracted
from neutral nickel alkyl complexes using boron trifluoride
diethyl etherate to yield agostic alkyl species. Results are
summarized in Scheme 7.
The nickel ethyl complex (N,O)Ni(Lu)(Et), 16a, was treated
with a 5-10-fold excess of BF3OEt2 at -78 °C in CD2Cl2. Upon
warming, two separate processes occur. At -10 °C, 16a converts
to an isomer 16ai. Structures of 16a and 16ai are assigned as
shown in Scheme 7 on the expected greater thermodynamic
stability of 16ai; we have no additional evidence for assignments.
The ethyl resonances of 16a shift from bands at -0.3 ppm (one
diastereotopic methylene proton), -0.4 ppm (methyl group),
and -0.5 ppm (one diastereotopic methylene proton) to bands
at 0.25 ppm (one diastereotopic methylene proton), -0.15 ppm
(one diastereotopic methylene proton), and -0.7 ppm (methyl
group). The isomerization is catalyzed by BF3; however, the
precise mechanism by which this occurs is unclear. Upon further
warming to 10 °C, 2,4-lutidine is abstracted by BF3 to yield an
equilibrium mixture of the ethyl agostic complex, 21, and 16ai.
Decomposition was observed at temperatures above 10 °C. The
1
temperature-dependent H NMR spectrum of 21 indicates a
highly dynamic system. At -20 °C, a broad band at -4.8 ppm
that integrates to three protons is observed. This resonance is a
result of rapid rotation about the C-C bond of the ethyl group,
which results in site exchange between the agostic hydrogen
and the two â hydrogens. Upon further cooling to -90 °C,
rotation about the C-C bond is slowed and the agostic proton
is observable as a triplet at -13.3 ppm with a geminal coupling
constant of 16 Hz. Attempted isolation of 21 was unsuccessful.
The nickel n-propyl complex (N,O)Ni(Lu)(Pr), 16b, when
treated with a 5-10-fold excess of BF3OEt2 in CD2Cl2 at -10
°C, isomerizes in a manner similar to that of 16a to yield 16bi.
Complex 16bi is completely converted to the isopropyl agostic
complex 22 upon warming to 0 °C; no 16bi remains. The
intermediate n-propyl agostic complex could not be detected;
its isomerization is fast relative to lutidine abstraction by BF3.
Complex 22 is more stable than its nickel ethyl counterpart,
and no decomposition is observed at temperatures as high as
(47) This analysis assumes that the equilibrium shown in Scheme 6 is fast relative
to insertion. This is supported by the fact that the first insertion is strongly
inhibited by added PPh3 (e.g., see entry 5, Table 2, and kinetics in
Supporting Information).
9
5834 J. AM. CHEM. SOC. VOL. 126, NO. 18, 2004

Polymerization of Ethylene
A R T I C L E S
Scheme 8
35 °C. (A similar trend in stability has been observed with the
cationic R-diimine nickel agostic complexes.24) Complex 22
exhibits dynamic behavior. At 20 °C, the six protons of the
isopropyl agostic moiety are equivalent due to rapid rotation
about the nickel-methine carbon bond and are observed as a
broad singlet at -2.4 ppm. Upon cooling to -40 °C, rotation
about the nickel-carbon bond is slowed and two broad
resonances at -0.6 and -4.3 ppm (equidistant from the -2.4
ppm midpoint), corresponding to the now inequivalent methyl
groups, are observed. Upon further cooling to -90 °C a broad
resonance for the agostic hydrogen is observed at -13.9 ppm.
The isopropyl agostic complex was used to quantitatively
assess the dynamics of nickel-carbon bond rotation and
â-hydride elimination. At the coalescence temperature (-10 °C)
of the two resonances attributed to the inequivalent methyl
Catalyst Decomposition Studies. The products of catalyst
decomposition were isolated from polymerizations carried out
with the unsubstituted anilinotropone catalyst, 3, the 7-phenyl-
substituted anilinotropone catalyst, 9a, and the previously
reported 2-anilinoperinaphthenone catalyst, 4.35 For each po-
lymerization (200 psig, 80 °C), the polymer was precipitated
in methanol and collected by filtration. The filtrate was
concentrated and analyzed for the catalyst decomposition
products. From the polymerization employing the unsusbtituted
anilinotropone catalyst, only the bis-anilinotropone nickel(II)
complex, 25, was isolated. The identity of the bis-ligand
complex was confirmed by previously reported independent
synthesis. A single-crystal X-ray structural analysis confirmed
the geometry about the nickel center with the sterically bulky
nitrogen atoms trans to each other.36 Similar behavior has been
reported for SHOP-type catalysts9,49 and the salicylaldimine-
based catalysts (2a,b).31 From the polymerization reaction
employing the 7-phenyl-substituted catalyst 9a, the bis-anili-
notropone nickel(II) complex 26 and free ligand, 2-anilino-7-
phenyltropone, 8a, were isolated. Finally the anilinoperinaph-
thenone catalyst decomposes exclusively to the free ligand,
2-anilinoperinaphthenone, 24. The catalyst decomposition route
for all three catalysts is likely similar, with the individual
decomposition products coming at different points along the
decomposition route (see below).
Treatment of the complexes 3, 9a, and 4 with the free ligands
results in formation of benzene and the bis-ligand complexes.
An NMR study was used to qualitatively estimate the rates of
these reactions (Table 6). For each experiment, an NMR tube
was charged with the nickel complex and its corresponding
ligand in equimolar ratios. In each case, formation of the bis-
ligand complex was observed via 1H NMR spectroscopy at 75
°C (eq 7). Approximate half-lives are noted in Table 6. The
catalyst/ligand combinations that react the fastest coincide with
the catalysts that decompose to yield the highest ratios of bis-
ligand complex to free ligand during polymerization. As shown
in Table 6, the unsubstituted anilinotropone catalyst 3 and ligand
23 react significantly faster than the anilinoperinaphthenone
catalyst 4 and ligand 24 (t1/2 ) <1 and 26 h, respectively).
The rate of bis-ligand complex formation is governed by a
combination of steric and electronic effects. The steric bulk of
the 7-phenyl-substituted catalyst 9a and ligand 8a slows the
groups, the rate of site exchange was calculated to be 3300 s-1
,
which corresponds to a barrier to nickel-carbon bond rotation
of 11.1 kcal/mol (eq 6). Spin saturation transfer techniques were
used to determine an upper limit for the barrier to â-hydride
elimination. The methine proton resonance at 1.3 ppm was
saturated, and a decrease in the integration of the methyl proton
resonance at -2.4 ppm was observed at 20 °C (Scheme 8).
Quantitative analysis (see Expermental Section) leads to a rate
constant for site exchange of Hb to Ha of 1.1 s-1, ∆Gq ) 17.1
kcal/mol. This value represents an upper limit for the barrier to
â-hydride elimination as shown in the free energy diagram in
Scheme 8. Because the isopropyl agostic complex is more stable
than the n-propyl agostic complex, the highest energy transition
state rate in the exchange of Ha and Hb is likely transition state
2, which lies between the propene hydride complex, 22-o, and
the n-propyl agostic complex. The olefin hydride complex is
expected to lie very close to the transition state;48 thus, the 17.1
kcal/mol barrier is expected to be only slightly greater than the
free energy difference between 22 and ts-1. By comparison,
the nickel-carbon rotation and â-hydride elimination barriers
for the cationic R-diimine nickel catalyst were reported to be
9.0 and 14.0 kcal/mol, respectively.24 The â-elimination barrier
in 13 is reported to be 15.1 kcal/mol.43
(48) Michalak, A.; Ziegler, T. Organometallics 2003, 22, 2069.
(49) Klabunde, U.; Tulip, T. H.; Roe, D. C.; Ittel, S. D. J. Organomet. Chem.
1987, 334, 141.
9
J. AM. CHEM. SOC. VOL. 126, NO. 18, 2004 5835

A R T I C L E S
Jenkins and Brookhart
Table 6. Bis-ligand Formation Half-Lives for Catalysts 3, 4, and
9a and Ligands 23, 24, and 8a in C6D6 at 75 °C
complex, 17c, is converted to a nickel hydride complex and
hexenes (ca. 90% 1-hexene, Scheme 10). The nickel hydride
complex 20 that is formed is identical to independently
synthesized 20 (see above). Further heating leads to the
conversion of the nickel hydride to the bis-anilinotropone nickel
1
complex. No free ligand is observed by H NMR due to the
rapid reaction of the aniliotropone ligand with the nickel-hexyl
or nickel-hydride complex. The rate constants for the conver-
sion of 17c to 20 (k1) and 20 to 25 (k2) at 50 °C are 3.2 × 10-4
and 4.2 × 10-6 sec-1, respectively.
entry
catalyst (concn, mM)
ligand (concn, mM)
product
t1/2
The kinetics of the conversion to both nickel hydride and
bis-ligand complex were also monitored in the presence of added
PPh3. Added PPh3 slightly inhibits the rate of formation of the
nickel-hydride complex and significantly inhibits the rate of
formation of the bis-ligand complex. Table 7 summarizes the
quantitative rate data for the conversion of the nickel-hexyl
complex 17c to the nickel-hydride complex 20 in the presence
of various concentrations of PPh3. A plot of kobs vs 1/[PPh3] is
shown in Figure 3. The plot is linear with an intercept of 1.6 ×
10-4 sec-1. These data establish two routes for â-elimination,
a simple first-order decay independent of PPh3 and a second
route inhibited by PPh3. A plausible mechanism is shown in
Scheme 11, and from Figure 2 and eq 9, k1 is calculated as
1.63 × 10-4 sec-1. A reasonable assumption is that the simple
first-order process occurs through a five-coordinate olefin
hydride intermediate.50 The observation that the major pathway
for â-elimination is a first-order process also explains the drop
in molecular weight with the increase in PPh3 concentration.
With an increase in PPh3 concentration, the rate of chain
propagation will be retarded, while the rate of chain transfer
via â-elimination will be essentially unaffected. Thus, the Rct/
Rprop will increase and Mn values will decrease.
1
2
3
4
5
3 (10.6)
9a (9.4)
4 (10.6)
3 (9.4)
23 (10.6)
8a (9.4)
24 (10.6)
24 (9.4)
23 (9.4)
25
26
27
28
28
<1 h
7-8 h
26 h
>24 h
<1 h
4 (9.4)
rate of the bimolecular reaction relative to the unsubstituted
catalyst 3 and ligand 23 (t1/2 ) 7-8 h and <1 h, respectively).
The N-H proton of the anilinotropone ligand 23 (entry 3, Table
6) is significantly more acidic than the N-H proton of the
anilinoperinaphthenone ligand 24.35 Therefore, the protonation
of the phenyl group to give benzene, PPh3, and bis-ligand
complex is much slower for anilnoperinaphthenone than anili-
notropone (entries 1 and 3). To verify that the acidity of the
ligand is a determining factor in the rate of bis-ligand complex
formation, a set of crossover experiments was conducted.
Anilinoperinaphthenone ligand 24 and anilinotropone catalyst
3 were combined, and the rate of reaction was compared to the
rate of reaction of anilinotropone ligand 23 with anilinoperi-
naphthenone catalyst 4 (entries 4 and 5). As expected, the more
acidic anilinotropone ligand reacted significantly faster than the
anilinoperinaphthenone ligand (t1/2 < 1 h and > 24 h,
respectively), supporting the contention that the acidity of the
N-H proton is a determining factor in the rate of bis-ligand
complex formation.
A decomposition mechanism that accounts for the range of
decomposition products involves the formation of a nickel
hydride complex as an intermediate in the catalytic cycle. The
hydride complex can either insert ethylene and reenter the
propagation cycle or reductively eliminate free ligand, thus
leading to catalyst deactivation (Scheme 9). The distribution of
decomposition products is then determined by the relative rate
of formation of free ligand versus its rate of reaction with the
nickel complex(es) present as part of the propagation cycle. In
the case of anilinoperinaphthenone catalyst, 4, the anilinoperi-
naphthenone ligand 24 reacts so slowly with the propagating
species that catalyst decomposition results only in formation
of free ligand. For anilinotropone catalyst 3, just the opposite
it the case. Free ligand reacts rapidly with the propagating
species, and only bis-ligand complex is observed. The 7-phenyl-
substituted catalyst is intermediate between the two extreme
cases of 4 and 3; thus, both free ligand and bis-ligand complex
are obtained.
k2Keq[S]
rate ) k1[S] +
kobs ) k1 +
(8)
(9)
[PPh3]
k2Keq
[PPh3]
The significant inhibition of bis-ligand complex formation
by excess PPh3 suggests that reductive elimination of the
anilinotropone ligand from the nickel hydride complex occurs
only after PPh3 dissociates from the nickel center. This is
consistent with the observation that excess triphenylphosphine
in the bulk polymerization experiments increases catalyst
lifetime.
Additional Comments on the Propagation Mechanism.
The propagation mechanism shown in Scheme 5 is consistent
with the mechanistic data presented here. The presence of an
equilibrium between the Ni(alkyl)(PPh3) complex and the Ni-
(alkyl)(ethylene) complex is supported by polymerization studies
that show inhibition of catalyst activity at sufficient PPh3
concentrations. The presence of an equilibrium is also supported
by the saturation kinetics observed for polymerizations using
To further probe the catalyst deactivation mechanism, the
thermal stabilities of the parent anilinotropone catalyst 3 and
the nickel-hexyl complex (N,O)Ni(PPh3)(hexyl), 17c, were
investigated. (N,O)Ni(PPh3)(Ph), 3, is stable for hours upon
heating to temperatures above 60 °C in C6D6. Therefore, it is
unlikely that decomposition of the catalyst precursor occurs prior
to initiation. However, upon heating to 50 °C in C6D6, hexyl
(50) A four-coordinate olefin hydride intermediate formed by initial dechelation
of the carbonyl group to yield a three coordinate species cannot be ruled
out. However, one might expect that if a three-coordinate species is required,
the pathway involving loss of PPh3 would be far more favorable than one
involving dechelation of the carbonyl group of the bidentate ligand. We
have shown in earlier work that CO insertion into a Ni(II)-alkyl can occur
via a five-coordinate nickel alkyl carbonyl species that has close analogy
with the proposed five-coordinate olefin hydride species. See: Shultz, C.
S.; DeSimone, J. M.; Brookhart, M. J. Am. Chem. Soc. 2001, 123, 9172.
9
5836 J. AM. CHEM. SOC. VOL. 126, NO. 18, 2004

Polymerization of Ethylene
A R T I C L E S
Scheme 9
Scheme 10
Figure 3. Ethylene insertion barriers and isomerization barriers for A and
B in kcal/mol as calculated by Ziegler.
Table 7. Rate of Conversion of 17c to 20 in the Presence of
Added PPh3 in C6D6 at 50 °C
complex 16a has reached maximum activity and “rests” as the
nickel-ethylene complex at these pressures.
entry
[PPh3] (mM)
kobs (s-1
)
The nature and reason for formation of branched polyethylene
formed by 3 has been discussed previously.36 Variable-temper-
ature and -pressure experiments with 3 suggest the mechanism
depicted in Scheme 12, which mirrors the mechanism suggested
for branching in the diimine catalyst system, 1.10,12 A branch
in the polymer is introduced first by â-hydride elimination of a
growing polymer chain followed by olefin rotation and reinser-
tion with opposite regiochemistry. This process may occur many
times prior to ethylene insertion and therefore lead to methyl
as well as higher branches. As ethylene pressure is increased,
monomer coordination and insertion becomes more favorable
and therefore the unimolecular branching pathway is less
favored. Conversely, as temperature is increased, the unimo-
lecular branching pathway is increasingly favored over ethylene
coordination and insertion and therefore more branching is
observed. The observation that the number of branches depends
on the temperature and pressure of polymerization is consistent
with the observation that the barriers to ethylene insertion and
â-hydride elimination are similar (16-17 and 17.1 kcal/mol).
It is of interest to note that the insertion and â-elimination
barriers are similar for these neutral Ni catalysts (ca. 16-17
kcal/mol), a feature that also applies to the cationic Ni diimine
catalysts (13-14 kcal/mol).24
1
2
3
4
5
6
20.8
31.2
40.6
52.0
156
2.87 × 10-4
2.61 × 10-4
2.54 × 10-4
2.07 × 10-4
1.82 × 10-4
1.63 × 10-4
312
Figure 2. Plot of kobs vs 1/[PPh3] for the conversion of 17c to 20.
9a at 40 °C as ethylene pressure is increased to 800 psig. The
propagation mechanism also implies that a more weakly
coordinating donor ligand should lead to a more active polym-
erization catalyst and saturation of the turnover frequency at
lower ethylene pressures. Complex 16a, containing the more
weakly bound 2,6-lutidine ligand, is significantly more reactive
than the PPh3 analogue, 3, at 600 and 800 psig ethylene at 60
°C (Table 8). The resting state(s) for 3 during polymerization
is still an equlibrium between the two species at 600 and 800
psig (saturation has not been reached), while the lutidine
Two possible isomers exist for nickel alkyl and aryl com-
plexes. For the phosphine complexes, the more stable isomer
is the one in which the alkyl/aryl group is trans to the weaker
donor oxygen as expected on electronic grounds. It is reasonable
to assume that this geometry, A, is also that of the more stable
alkyl olefin complexes. However, since this isomer may readily
interconvert with isomer B, it is unclear from which one
insertion occurs. Ziegler has recently reported DFT-based
9
J. AM. CHEM. SOC. VOL. 126, NO. 18, 2004 5837

A R T I C L E S
Jenkins and Brookhart
Scheme 11
Table 8. Ethylene Polymerization with Catalysts 3 and 16a with a
Catalyst Loading of 6.6 µmol in 200 mL of Toluene at 60 °C
For the (N,O)Ni(R)(PPh3) complexes, the catalyst resting
state(s) are an equilibrium mixture of (N,O)Ni(R)(PPh3) and
(N,O)Ni(R)(C2H4) complexes. At high ethylene pressures, the
equilibrium can be shifted strongly to the side of the ethylene
complex and turnover frequency becomes independent of
ethylene pressure(saturation conditions).
entry
catalyst
P (psig)
TON
1
2
3
4
3
3
16a
16a
600
800
600
800
8600
17 900
43 500
41 700
From mesurement of the turnover frequencies under saturation
conditions, the barriers to migratory insertion in (N,O)Ni(R)-
(C2H4) complexes were determined to lie in the range of 16-
17 kcal/mol, ca. 2-3 kcal/mol greater than the barriers to
insertion in cationic diimine complexes.
calculations on catalyst 3 that suggest that insertion occurs from
the less stable isomer B (ethylene trans to oxygen, Figure 3) to
form A′, an agostic complex with the alkyl group trans to
oxygen.48 Insertion from B is calculated to be ca. 4.4 kcal/mol
more favorable than insertion from the more stable isomer, A.
Ziegler also estimates, on the basis of calculations on the
salicylaldimine-nickel system, that the barrier to isomerization
between A and B is 11.4 kcal/mol.25
Intermediate (N,O)Ni(alkyl) complexes possess a â-agostic
structure and exhibit dynamic behavior similar to the cationic
alkyl diimine complexes. The barrier to â-elimination and
readdition (chain-walking) is ca. 17 kcal/mol. Rotation around
Ni-CR in these complexes occurs with a barrier of 11.1 kcal/
mol. These isomerization processes account for branched
polyethylenes formed from these catalysts and mimic behavior
previously observed with nickel and palladium diimine catalysts.
Thermolysis of the (N,O)Ni(hexyl)(PPh3) complex yields the
nickel hydride complex via â-elimination. Kinetic studies
establish a dual pathway, a (dominant) simple first-order process
and a process inhibited by PPh3 involving reversible loss of
phosphine. This process is a model for chain transfer and clearly
explains why polymer molecular weights decrease with increas-
ing [PPh3]. The rate of propagation is retarded by increasing
[PPh3], while the rate of chain transfer is unchanged, resulting
in an increase in Rct/Rprop. These results argue against a chain
transfer to monomer process, which would exhibit a rate
inhibition equal to that for Rprop with added PPh3
The fact that the molecular weight significantly increases with
ethylene pressure (e.g., for complex 9a Mn ) 134K, 50 psig;
280K, 100 psig; 356K, 200 psig, entries 1-3, Table 5) suggests
that chain transfer to monomer is not the major chain transfer
mechanism. Theoretical calculations have suggested that for
similar systems, the alkyl olefin intermediate in the polymeri-
zation may partition as shown in eq 10 (Scheme 13).26,27
Pathway a represents insertion and propagation, while pathway
b represents chain transfer to monomer. If chain transfer to
monomer is the sole chain transfer mechanism, then the ratio
of rates of chain transfer and propagation, Rct/Rp, should be
independent of ethylene concentration and thus the chain transfer
limited molecular weight should be independent of ethylene
concentration. Since this is not observed, a chain transfer process
that, at least in large part, is not dependent on ethylene
concentration must occur. This is best rationalized by a
conventional â-hydride elimination mechanism, shown in eq
11, that has been observed in the themolysis of the nickel hexyl
complex 17c. Of course, given the qualitative Mn values and
the substantial error in these determinations, it is possible that
some chain transfer occurs via the chain transfer to monomer
pathway.
Reductive elimination of the free ligand occurs from the
hydride, (N,O)Ni(H)(PPh3) and is inhibited by added PPh3. Such
a process, combined with reaction of the free ligand with
propagating nickel species, accounts for the catalyst decay
process and the mixture of catalyst decay products observed,
which includes the free ligands and bis-ligand complexes.
Experimental Section
Summary
General Considerations. Argon was purified by passage through
columns of BASF R3-11 catalyst (Chemalog) and 4 Å molecular sieves.
Toluene and pentane were deoxygenated and dried over a column of
activated alumina. THF was distilled under a nitrogen atmosphere from
sodium benzophenone ketyl prior to use. Triethylamine, triphenylphos-
phine, tri-p-tolylphosphine, and boron trifluoride diethyl etherate were
The studies reported here provide the first in-depth description
of the mechanism of ethylene polymerization catalyzed by
neutral Ni(II) complexes and allow a detailed comparison to
the related cationic Ni diimine systems. Key findings are
summarized below:
9
5838 J. AM. CHEM. SOC. VOL. 126, NO. 18, 2004

Polymerization of Ethylene
A R T I C L E S
Scheme 12
Scheme 13
used as received from Aldrich; (PPh3)2NiCl2 was used as received from
Strem. Polymer-grade ethylene was used as received from Matheson.
Some studies were conducted utilizing Research Plus-grade ethylene
obtained from Matheson; the purity level is 99.9996%. Anilinotropone
catalyst 3,34 anilinotropone 27,51 anilinoperinaphthenone catalyst 4,35
anilinoperinaphthenone 28,35 2-methoxy-7-p-methoxyphenyltropone
5c,52 2-methoxy-7-p-trifluoromethylphenyltropone 5d,52 7-p-methox-
yphenyltropolone 6c,39 7-p-trifluoromethylphenyltropolone 6d,39 2-tri-
flato-7-p-methoxyphenyltropone 7c,40 2-triflato-7-p-trifluoromethylphe-
nyltropone 7d,40 2-methoxytropone,37 2-methoxy-7-bromotropone,38 and
diisopropylaniline were added sequentially, and the reaction mixture
was stirred at 90 °C under argon. After the specified reaction time, the
reaction mixture was cooled to room temperature, diluted with CH2-
Cl2, and filtered through a pad of Celite. The solution was concentrated,
and the residue was chromatographed on silica gel with 15% ethyl
acetate in hexanes as the eluent.
2-(2,6-Diisopropylanilino)-7-p-methoxyphenyltropone, 8c. 2-Tri-
flato-7-p-methoxyphenyltropone (1.40 g, 3.89 mmol) and 880 µL (4.67
mmol) of 2,6-diisopropylaniline were converted to anilinotropone 8c
in 20 h (760 mg, 50%). 1H NMR (CDCl3, 400 MHz): δ 8.7 (bs, 1H);
7.6 (m, 3H); 7.4 (m, 1H); 7.25 (m, 2H); 7.0 (m, 3H); 6.75 (t, J ) 10
Hz, 1H); 6.3 (d, J ) 10 Hz, 1H); 3.85 (s, 3H); 2.9 (sept, J ) 5 Hz,
2H); 1.13 (d, J ) 3 Hz, 6H); 1.10 (d, J ) 3 Hz, 6H). 13C NMR (CDCl3,
100 MHz): δ 174.5, 159.4, 157.3, 147.2, 143.7, 141.2, 139.3, 135.4,
133.1, 131.4, 130.9, 129.4, 129.0, 128.8, 125.9, 124.6, 123.0, 113.9,
110.9, 55.8, 28.9, 25.0, 23.7. Anal. Calcd for C26H29NO2: C, 80.59;
H, 7.54; N, 3.61. Found: C, 80.87; H, 7.40; N, 3.30.
2-(2,6-Diisopropylanilino)-7-p-trifluoromethylphenyltropone, 8d.
2-Triflato-7-p-trifluoromethylphenyltropone (260 mg, 0.65 mmol) and
148 µL (0.78 mmol) of 2,6-diisopropylaniline were converted to
anilinotropone 8d in 20 h (130 mg, 48%). 1H NMR (CDCl3, 400
MHz): δ 8.7 (bs, 1H); 7.7 (s, 4H); 7.5 (d, J ) 7 Hz, 1H); 7.35 (d, J
) 5 Hz, 1H); 7.2 (d, J ) 7 Hz, 2H); 7.0 (t, J ) 5 Hz, 1H); 6.7 (t, J )
5 Hz, 1H); 6.3 (d, J ) 7 Hz, 1H); 2.9 (sept, J ) 5 Hz, 2H); 1.1 (d, J
) 5 Hz, 12H). 19F NMR (CDCl3, 376 MHz): δ -62.9 (s).
Procedure for the Preparation of 2-Anilino-7-aryltroponeNi-
(PPh3)(Ph) Complexes. A flame-dried Schlenk tube was charged with
the appropriate anilinotropone (1.0 equiv) and sodium hydride (5.0
equiv) under an atmosphere of argon. THF (ca. 30 mL) was added,
and the reaction mixture was stirred for 18 h. Then, (PPh3)2Ni(Ph)Cl
(1.0 equiv) was added as a solid, and the reaction mixture was stirred
for 1 h and then filtered through a pad of Celite. The solvent was
removed in vacuo, and the residue was recrystallized from toluene and
pentane.
53
Lu2NiCl2 were prepared according to literature procedures.
All 1H NMR chemical shifts are reported in parts per million (ppm)
downfield from tetramethylsilane referenced to residual solvent 1H
signals. All 13C NMR chemical shifts are reported in ppm relative to
residual 13C of the solvent. 31P NMR shifts are reported relative to H3-
PO4 external standard. 1H NMR spectra of polyethylenes were recorded
in C6D5Br at 120 °C. The formula used to calculate branching was
(CH3/3)/[(CH + CH2 + CH3)/2] x 1000 ) branches per 1000 carbons.
CH3, CH2, and CH refer to the integration obtained for the methyl,
methylene, and methine resonances, respectively. 13C NMR spectra of
polyethylenes were recorded in a 0.05 M Cr(acac)3 solution in
bromobenzene-d5 with 15 wt % polymer. Peaks were assigned as
described by Cotts.18 High-temperature gel permeation chromatography
(GPC) was performed by DuPont (Wilmington, DE) in 1,2,4-trichlo-
robenzene at 135 °C using a Waters HPLC 150C equipped with Shodex
columns. A calibration curve was established with polystyrene stan-
dards, and universal calibration was applied using Mark-Houwink
constants for polyethylene (k ) 4.34 × 10-4; R ) 0.724). Mn values
for low-molecular weight polymer samples were determined by 1H
NMR end group analysis. Elemental analyses were performed by
Atlantic Microlab, Inc., of Norcross, GA.
General Procedure for the Preparation of 7-Aryl-2-anili-
notropones. A flame-dried Schlenk tube was charged with 7-aryl-2-
triflatotropone (1.0 equiv), ground cesium carbonate (1.4 equiv),
Pd2(dba)3 (2.5 mol %), and racemic BINAP (5 mol %) under an
atmosphere of argon. Toluene (2 mL/mmol triflatotropone) and 2,6-
7-p-Methoxyphenyl-2-(2,6-diisopropyl)anilinotroponeNi(PPh3)-
Ph Complex, 9c. Anilinotropone 8c (136 mg, 0.35 mmol) and 244
mg (0.35 mmol) of (PPh3)2Ni(Ph)Cl were converted to the desired
product (176 mg, 63%). 1H NMR (CD2Cl2, 400 MHz): δ 7.3 (m, 9H);
7.1 (m, 7H); 6.8-7.0 (m, 5H); 6.7 (m, 3H); 6.5 (m, 1H); 6.4 (m, 3H);
(51) Hicks, F. A.; Brookhart, M. Org. Lett. 2000, 2, 219.
(52) Suri, S. C.; Nair, V Synthesis 1990, 8, 695.
(53) Machin, D. J.; Sullivan, J. F. J. Chem. Soc. A 1971, 4, 658.
9
J. AM. CHEM. SOC. VOL. 126, NO. 18, 2004 5839

A R T I C L E S
Jenkins and Brookhart
6.2 (m, 1H); 6.1 (m, 2H); 3.7 (s, 3H); 3.5 (sept, J ) 5 Hz, 2H); 1.2 (d,
J ) 5 Hz, 6H); 1.0 (d, J ) 5 Hz, 6H). 13C NMR (CD2Cl2, 100 MHz):
δ 158.1, 144.7, 142.5, 137.7, 135.6, 134.4, 134.3, 132.8, 132.2, 132.0,
1331.7, 131.1, 129.8, 128.1, 128.0, 125.3, 123.4, 121.2, 121.1, 121.0,
113.0, 55.4, 28.8, 25.7, 23.8. 31P NMR (CD2Cl2, 162 MHz): δ 27.0.
Anal. Calcd for C50H48NO2PNi: C, 76.54; H, 6.17; N, 1.78. Found:
C, 76.56; H, 6.18; N, 1.73.
7-p-Trifluoromethylphenyl-2-(2,6-diisopropyl)anilinotroponeNi-
(PPh3)Ph Complex, 9d. Anilinotropone 8d (130 mg, 0.31 mmol) and
212 mg (0.31 mmol) of (PPh3)2Ni(Ph)Cl were converted to the desired
product (82 mg, 29%). 1H NMR (CD2Cl2, 400 MHz): δ 7.3 (m, 10H);
7.1 (m, 10H); 6.95 (m, 3H); 6.8 (m, 1H); 6.7 (d, J ) 10 Hz, 2H); 6.5
(m, 1H); 6.4 (d, J ) 13 Hz, 1H); 6.2 (m, 1H); 6.1 (t, J ) 10 Hz, 2H);
3.55 (sept, J ) 5 Hz, 2H); 1.2 (d, J ) 5 Hz, 6H); 1.0 (d, J ) 5 Hz,
6H). 13C NMR (CD2Cl2, 100 MHz) ) δ 169.9, 147.1, 144.4, 142.3,
137.6, 134.3, 134.2, 132.7, 131.9, 131.4, 131.3, 130.3, 130.0, 128.2,
128.1, 125.4, 125.2, 124.4, 123.5, 121.5, 121.3, 121.1, 28.9, 25.8, 23.7.
31P NMR (CD2Cl2, 162 MHz): δ 27.4 (s). 19F NMR (CD2Cl2, 376
MHz): δ -63.1 (s). Anal. Calcd for C50H45NOPF3Ni: C, 73.00; H,
5.51; N, 1.70 Found: C, 72.43; H, 5.92; N, 1.46.
2-(2,6-Diisopropyl)anilinotroponenickel(2,4-lutidine)chloride, 14.
A flame-dried Schlenk tube was charged with (Lu)2NiCl2 (489 mg,
1.42 mmol) under argon. THF (30 mL) was added, and the mixture
was cooled to 0 °C. Then, NEt3 (1.0 mL, 7.2 mmol) was added followed
by a THF solution (15 mL) of 2-(2,6-diisopropyl)anilinotropone. The
reaction mixture was gradually warmed to room temperature over the
course of 5 h and then filtered through Celite. The solvent was removed
in vacuo, and the residue was recrystallized from toluene and pentane
to afford paramagnetic 14 (445 mg, 65%). Anal. Calcd for C36H31N2-
ONiCl: C, 64.83; H, 6.49; N, 5.81 Found: C, 63.58; H, 6.54; N, 5.81.
2-(2,6-Diisopropyl)anilinotroponenickel(triphenylphosphine)-
chloride, 15. A flame-dried Schlenk tube was charged with (PPh3)2-
NiCl2 (1.0 g, 3.55 mmol) under argon. THF (30 mL) was added, and
the mixture was cooled to 0 °C. Then, NEt3 (2.0 mL, 14.3 mmol) was
added followed by a THF solution (15 mL) of 2-(2,6-diisopropyl)-
anilinotropone. The reaction mixture was gradually warmed to room
temperature over the course of 5 h and then filtered through Celite.
The solvent was removed in vacuo, and the residue was recrystallized
from toluene and pentane to afford paramagnetic 15 (1.98 g, 88%).
Anal. Calcd for C37H37NPONiCl: C, 69.79; H, 5.81; N, 2.20 Found:
C, 70.33; H, 5.96; N, 2.03.
vacuo, and the product was collected (615 mg, 61%). 1H NMR (CD2-
Cl2, 400 MHz): δ 9.1 (d, J ) 6 Hz, 2H); 7.2 (3, 3H); 7.05 (s, 1H); 6.9
(d, J ) 6 Hz, 1H); 6.85 (t, J ) 10 Hz, 1H); 6.6 (t, J ) 10 Hz, 1H); 6.5
(d, J ) 11 Hz, 1H); 6.3 (t, J ) 9 Hz, 1H); 6.0 (d, J ) 12 Hz, 1H); 3.7
(sept, J ) 7 Hz, 1H); 3.55 (m, 4H); 2.3 (s, 3H); 1.4 (vt, J ) 6 Hz,
6H); 1.05 (dd, J ) 7, 10 Hz, 6H); 0.4 (m, 1H); 0.3 (t, J ) 7 Hz, 3H);
0.2 (m, 1H); -0.3 (m, 1H); -0.45 (m, 1H). 13C NMR (CD2Cl2, 100
MHz): δ 179.9, 168.6, 160.5, 150.9, 148.3, 144.0, 142.9, 134.5, 132.6,
125.8, 125.6, 123.9, 122.7, 120.6, 120.0, 119.3, 29.1, 28.4, 26.0, 25.0,
24.2, 23.9, 23.8, 21.0, 16.4, 13.4.
2-(2,6-Diisopropyl)anilinotroponenickel(triphenylphosphine)-
ethyl, 17a. A flame-dried Schlenk tube was charged with (N,O)Ni-
(PPh3)Cl, 15 (180 mg, 0.28 mmol), under argon. Ether (25 mL) was
added, and the solution was cooled to -78 °C. Ethylmagnesium chloride
(2M, 0.36 mmol) was added dropwise. Upon complete addition, the
reaction mixture was stirred at -78 °C for 3 h and then warmed to
room temperature and filtered through florisil. The solvent was removed
in vacuo and 17a was collected (40 mg, 22%). 1H NMR (CD2Cl2, 400
MHz): δ 7.8 (m, 6H); 7.4 (m, 9H); 7.2 (s, 3H); 6.9 (t, J ) 7 Hz, 1H);
6.75 t, J ) 7 Hz, 1H); 6.5 (d, J ) 8 Hz, 1H); 6.4 (t, J ) 7 Hz, 1H);
6.3 (d, J ) 8 Hz, 1H), 3.6 (sept, J ) 5 Hz, 2H); 1.3 (d, J ) 5 Hz, 6H);
1.0 (d, J ) 5 Hz, 6H); -0.2 (t, J ) 5 Hz, 3H); -0.5 (m, 2H). 13C
NMR (CD2Cl2, 100 MHz): δ 143.8, 143.3, 134.9, 134.8, 134.6, 133.5,
132.4, 132.0, 130.2, 128.9, 128.4, 128.3, 125.8, 124.2, 121.4, 120.5,
120.2, 29.0, 28.6, 24.1, 20.4, 18.5. 31P NMR (CD2Cl2, 162 MHz): δ
34.2 (s). Anal. Calcd for C39H42NPONi: C, 74.07; H, 7.01; N, 2.21.
Found: C, 73.78; H, 6.84; N, 2.11.
2-(2,6-Diisopropyl)anilinotroponenickel(triphenylphosphine)(n-
propyl), 17b. A flame-dried Schlenk tube was charged with (N,O)-
Ni(PPh3)Cl, 15 (200 mg, 0.31 mmol), under argon. Ether (25 mL) was
added, and the solution was cooled to -78 °C. Isopropylmagnesium
chloride (2.0 M in ether, 0.38 mmol) was added dropwise, and the
reaction mixture was stirred cold for 3.5 h. The reaction mixture was
filtered through florisil at room temperature, during which time the
product isomerized to the n-propyl isomer, and the solvent was removed
in vacuo to afford 50 mg (25%) of 17b. 1H NMR (CD2Cl2, 400 MHz):
δ 7.8 (m, 6H); 7.4 (m, 9H); 7.2 (m, 3H); 6.9 (t, J ) 10 Hz, 1H); 6.75
(t, J ) 10 Hz, 1H); 6.5 (d, J ) 10 Hz, 1H); 6.4 (t, J ) 10 Hz, 1H); 6.3
(d, J ) 11 Hz, 1H); 3.6 (sept, J ) 7 Hz, 2H); 1.3 (d, J ) 7 Hz, 6H);
1.05 (d, J ) 7 Hz, 6H); 0.5 (m, 2H); -0.2 (t, J ) 7 Hz, 3H); -0.5 (q,
J ) 8 Hz, 2H). 13C NMR (CD2Cl2, 100 MHz): δ 180.0, 168.9, 143.7,
143.2, 134.9, 134.6, 133.2, 132.4, 132.0, 130.2, 128.3, 125.8, 124.1,
121.4, 120.5, 28.7, 25.7, 23.8, 21.9, 16.4, 14.5. 31P NMR (CD2Cl2, 162
MHz): δ 34.2 (s). Anal. Calcd for C40H44NOPNi: C, 74.55; H, 6.88;
N, 2.17. Found: C, 74.15; H, 6.69; N, 1.84.
2-(2,6-Diisopropyl)anilinotroponenickel(2,4-lutidine)ethyl, 16a. A
flame-dried Schlenk tube was charged with (N,O)Ni(Lu)Cl, 14 (1.0 g,
2.08 mmol), under argon. Ether (30 mL) was added, and the solution
was cooled to -78 °C. Ethylmagnesium chloride (2 M, 2.50 mmol)
was added dropwise. Upon complete addition, the reaction mixture was
stirred at -78 °C for 3 h and then warmed to room temperature and
filtered through florisil. The solvent was removed in vacuo, and the
2-(2,6-Diisopropyl)anilinotroponenickel(triphenylphosphine)(n-
hexyl), 17c. A flame-dried Schlenk tube was charged with (N,O)Ni-
(PPh3)Cl, 15 (310 mg, 0.49 mmol), under argon. Ether (25 mL) was
added, and the solution was cooled to -78 °C. Hexylmagnesium
bromide (2.0 M in ether, 0.62 mmol) was added dropwise, and the
reaction mixture was stirred cold for 4 h. The reaction mixture was
filtered through florisil at room temperature, and the solvent was
removed in vacuo to afford 110 mg (33%) of 17d. 1H NMR (CD2Cl2,
400 MHz): δ 7.8 (m, 6H); 7.4 (m, 9H); 7.2 (m, 3H); 6.9 (t, J ) 10
Hz, 1H); 6.75 (t, J ) 9 Hz, 1H); 6.5 (d, J ) 10 Hz, 1H); 6.4 (m, 1H);
6.3 (d, J ) 11 Hz, 1H); 3.6 (sept, J ) 7 Hz, 2H); 1.3 (d, J ) 7 Hz,
6H); 1.0 (d, J ) 7 Hz, 6H); 0.7 (m, 2H); 0.55 (t, J ) 7 Hz, 3H); 0.4
(m, 4H); 0.05 (m, 2H); -0.5 (q, J ) 8 Hz, 2H). 13C NMR (CD2Cl2,
100 MHz): δ 180.0, 168.9, 143.6, 143.1, 135.0, 134.6, 133.2, 132.3,
131.9, 130.2, 128.4, 127.8, 124.1, 121.4, 120.4, 32.6, 31.2, 30.1, 29.1,
25.8,24.3, 23.9, 22.8, 14.2. 31P NMR (CD2Cl2, 162 MHz): δ 34.2 (s).
Anal. Calcd for C43H50NOPNi: C, 75.01; H, 7.61; N, 2.04 Found: C,
73.88; H, 7.38; N, 1.96.
1
product was collected (555 mg, 56%). H NMR (CD2Cl2, 400 MHz):
δ 9.1 (d, J ) 9 Hz, 1H); 7.2 (s, 3H); 7.1 (s, 1H); 6.95 (m, 1H); 6.8 (t,
J ) 7 Hz, 1H); 6.6 (t, J ) 7 Hz, 1H); 6.5 (d, J ) 9 Hz, 1H); 6.3 (t, J
) 7 Hz, 1H); 6.0 (d, J ) 7 Hz, 1H); 3.7 (sept, J ) 5 Hz, 1H); 3.55 (m,
4H); 2.3 (s, 3H); 1.4 (d, J ) 5 Hz, 6H); 1.0 (dd, J ) 5, 5 Hz, 6H);
-0.3 (m, 1H); -0.4 (t, J ) 5 Hz, 3H); -0.5 (m, 1H). 13C NMR (CD2-
Cl2, 100 MHz): δ 179.9, 168.6, 160.4, 150.8, 148.4, 143.9, 143.0, 134.5,
132.7, 125.9, 125.6, 124.0, 122.7, 120.7, 119.9, 119.4, 28.4, 28.3, 25.9,
24.9, 23.9, 21.0, 2.4. Anal. Calcd for C28H36N2ONi: C, 70.76; H, 7.63;
N, 5.89. Anal. Calcd for C28H36N2ONi: C, 70.76; H, 7.63; N, 5.89
Found: C, 69.10; H, 7.45; N, 5.39.
2-(2,6-Diisopropyl)anilinotroponenickel(2,4-lutidine)(n-propyl),
16b. A flame-dried Schlenk tube was charged with (N,O)Ni(Lu)Cl, 14
(1.0 g, 2.08 mmol), under argon. Ether (30 mL) was added, and the
solution was cooled to -78 °C. n-Propylmagnesium chloride (2 M,
2.50 mmol) was added dropwise. Upon complete addition, the reaction
mixture was stirred at -78 °C for 3 h and then warmed to room
temperature and filtered through florisil. The solvent was removed in
2-(2,6-Diisopropyl)anilinotroponenickel(triphenylphosphine)(i-
propyl), 17d. A flame-dried Schlenk tube was charged with 220 mg
(0.35 mmol) of (N,O)Ni(PPh3)Cl, 15, under argon. Ether (30 mL) was
9
5840 J. AM. CHEM. SOC. VOL. 126, NO. 18, 2004

Polymerization of Ethylene
A R T I C L E S
added, and the solution was cooled to -78 °C. Isopropylmagnesium
chloride was added dropwise, and the reaction mixture was stirred for
5 h. The reaction mixture was filtered cold through florisil, and the
solvent was removed in vacuo at -30 °C to yield 80 mg (35%) of
P(p-tolyl)3 in the specified amounts under argon. The tube was sealed
with a septum and cooled to -78 °C. CD2Cl2 (700 µL) was injected,
and the tube was warmed to -14 °C in the NMR probe. Conversion
was observed via 31P NMR spectroscopy as the Ni-PPh3 resonance at
28.6 ppm diminished as the Ni-(p-tolyl)3 resonance at 26.3 ppm grew
in. Also monitored was the appearance of free PPh3 at -5.7 ppm. A
pseudo-first-order rate of exchange was calculated by plotting the natural
log of the concentration of starting material vs time.
General Procedure for Keq Determination. An NMR tube was
charged with 17a (4 mg, 6.3 µmol), PPh3 (137 mg, 523 µmol), and
toluene-d8 (700 µL) under argon and cooled to -78 °C. Ethylene was
injected (4.0 mL, 115 µmol in solution), and the tube was warmed to
60 °C in the NMR probe. Insertion was monitored as with 3 and 9a,
and a pseudo-first-order rate constant was calculated by plotting the
natural log of the mole fraction of 17a vs time, yielding kobs ) 3.9 ×
10-4 s-1. A second experiment with 17a (4 mg, 6.3 µmol), PPh3 (61
mg, 233 µmol), and ethylene (4.0 mL, 107 µmol) gave a kobs of 7.8 ×
10-4. Keq values of 3.7 and 3.6 × 10-5 were calculated from each
experiment.
1
17d. H NMR (CD2Cl2, 400 MHz): δ 7.9 (m, 6H); 7.4 (m, 9H); 7.2
(m, 3H); 6.8 (m, 2H); 6.3 (m, 1H); 6.15 (t, J ) 10 Hz, 2H); 3.7 (sept,
J ) 7 Hz, 2H); 1.3 (d, J ) 7 Hz, 6H); 1.0 (d, J ) 7 Hz, 6H); 0.3 (m,
1H); -0.15 (d, J ) 6 Hz, 6H). 13C NMR (CD2Cl2, 100 MHz): δ 179.1,
168.2, 143.4, 137.2, 135.4, 134.2, 134.0, 132.8, 132.4, 130.7, 129.1,
128.5, 125.8, 123.8, 120.3, 29.0, 28.8, 25.8, 24.9, 23.9, 19.6. 31P NMR
(CD2Cl2, 162 MHz): δ 34.2 (s). The compound decomposes at
temperatures above 0 °C; thus, elemental analysis was not obtained.
2-(2,6-Diisopropyl)anilinotroponenickel(triphenylphosphine)-
(hydride), 20. A flame-dried Schlenk tube was charged with 400 mg
(0.62 mmol) of (N,O)Ni(PPh3)Cl, 15, and 300 mg (2.35 mmol) of
NaHB(OMe)3 under argon. The flask was placed in an ice bath, and
ether (30 mL) was added. The reaction mixture was stirred for 4 h and
filtered through florisil. The solvent was removed in vacuo to afford
1
180 mg (48%) of the hydride product. H NMR (C6D6, 400 MHz): δ
General Procedure for the Reaction of (N,O)Ni(Lu)(R) with
BF3OEt2. An NMR tube was charged with either 16a or 16b and CD2-
Cl2 under argon. The tube was cooled to -78 °C, and 5-10 equiv of
BF3OEt2 was injected. The tube containing 16a and BF3 was warmed
to -10 °C. The following key resonances of 21 were observed: 1H
NMR (CD2Cl2, 400 MHz): δ 2.8 (bs, 3H, BF3-2,4-Lutidine complex),
2.6 (bs, 3H, BF3-2,4-Lutidine complex), 1.4 (m, 12H, isopropyl methyl
groups), -4.8 (bs, 3H, Ni-CH2CH3). At -120 °C in CDCl2F: δ -13.3
(t, J ) 16 Hz, 1H, agostic proton).
Key resonances for 22 at 10 °C: 1H NMR (CD2Cl2): δ 2.8 (bs, 3H,
BF3-2,4-Lutidine complex), 2.6 (bs, 3H, BF3-2,4-Lutidine complex),
1.3 (d, J ) 7 Hz, 6H), 1.0 (d, J ) 7 Hz, 6H), -2.4 (bs, 6H, Ni-iPr
methyls). At -40 °C: δ -0.6 (bs, 3H), -4.3 (bs, 3H). At -90 °C: δ
-13.9 (bs, 1H).
Dynamics of 22, Measured Using Spin Saturation Transfer. The
1.3 ppm resonance (methine H of Ni-iPr) was saturated, and the change
in the integral of the -2.4 ppm resonance (iPr methyl groups) was
observed at 293K. Applying the standard equation k ) (1/T1)[M0/(Msst
- 1)], where M0 ) 2.77, Msst ) 1.76, and T1 ) 0.54 s, yields k ) 1.1
s-1, corresponding to ∆Gq ) 17.1 kcal/mol (293 K).
7.8 (m, 6H); 7.2 (d, J ) 10 Hz, 1H); 7.1 (m, 12 H); 6.8 (t, J ) 10 Hz,
1H); 6.65 (d, J ) 11 Hz, 1H); 6.5 (m, 1H); 6.3 (m, 1H); 3.45 (sept, J
) 7 Hz, 2H); 1.5 (d, J ) 7 Hz, 6H); 1.25 (d, J ) 7 Hz, 6H); -26.7 (d,
JPH ) 130 Hz, 1H). 13C NMR (C6D6, 100 MHz): δ 180.9, 169.3, 149.3,
141.8, 134.8, 134.5, 134.3, 133.6, 130.2, 129.0, 126.0, 124.3, 122.3,
121.8, 118.5, 28.9, 24.7, 24.2. 31P NMR (C6D6, 162 MHz): δ 31.8 (d,
JPH ) 123 Hz). The compound decomposes slowly at room temperature;
thus, elemental analysis was not obtained.
General Procedure for High-Pressure Ethylene Polymerizations.
A 1000 mL Parr autoclave was heated under vacuum up to 110 °C
and then cooled to the desired reaction temperature and backfilled with
ethylene. The autoclave was charged with toluene (185 mL), filled and
evacuated twice with ethylene (200psig), and pressurized with ethylene
to 200 psig. The stirring motor was engaged, and the reactor was
allowed to equilibrate at the desired temperature for approximately 10
min. In a glovebox, a sidearm flask was charged with the catalyst. The
flask was removed from the glovebox and placed on a vacuum line
under Ar. The catalyst was dissolved in 15 mL of toluene and
transferred via cannula into the vented autoclave with the stirring motor
off. The autoclave was sealed and pressurized to the desired level, and
the stirring motor was re-engaged. After the prescribed reaction time,
the stirring motor was stopped, the reactor vented, and the polymer
isolated via precipitation from methanol and dried in a vacuum oven.
This procedure was employed with modifications in reaction time,
temperature, and ethylene pressure. For studies with excess PPh3, both
the catalyst and PPh3 were added in the same 15 mL of toluene.
General Procedure for Monitoring Themolysis of 17c. An NMR
tube was charged with 17c, PPh3 (when appropriate), and C6D6 (700
µL) at room temperature under argon. The tube was warmed to 50 °C
in the NMR probe, and conversion to the nickel hydride complex, 20,
followed by conversion to the bis-ligand complex, 25, was observed
1
by both H and 31P NMR spectroscopy. Complete conversion to 20
was observed prior to conversion of 20 to 25 A first-order rate constant
for each conversion was obtained by comparing the integration of the
isopropyl methine resonances at 4.0, 3.9, and 4.25 ppm for 17c, 20,
and 25, respectively, and plotting the natural log of the starting material
mole fraction vs time. A representative plot may be found in Supporting
Information.
General Procedure for the Reaction of (N,O)Ni(PPh3)(Ph) with
Ethylene. An NMR tube was charged with either catalyst 3 or 9a and
CD2Cl2 (700 µL) under argon and sealed with a septum. The tube was
cooled to -78 °C, and ethylene was injected via syringe (2-5 mL).
The sample was warmed to the desired temperature inside the NMR
probe, and primary insertion was monitored by integration of resonances
associates with the ligand isopropyl groups. For 3, the original
resonances were δ 3.55 (sept, 2H), 1.2 (d, 6H), and 1.0 (d, 6H). The
new resonances associated with the first insertion of ethylene were δ
3.7 (sept, 2H), 1.3 (d, 6H), and 1.1 (d, 6H). Additionally, new
resonances for the methylene units of the nickel-benzyl group were
observed as multiplets at 1.6 and -0.3 ppm. For subsequent insertion,
the methylene benzyl resonances of the singly inserted product were
integrated against the new resonance of the methylene unit R to nickel
at δ -0.5 ppm for the multiply inserted product. A rate of insertion
was calculated by plotting the natural log of the mole fraction of starting
material vs time. A representative example is shown in Supporting
Information.
General Procedure for Bis-Ligand Complex Formation Reac-
tions. A J-Young tube was charged with equimolar amounts of the
specified catalyst and ligand and C6D6 (700 µL) at room temperature.
1
Conversion at 75 °C was observed via H NMR spectroscopy. The
concentrations for each experiment are listed in Table 6. NMR
spectroscopic data for 2536 and 2735 have been previously reported.
26: 1H NMR (C6D6, 400 MHz): δ 7.8-6.1 (m, 16H), 4.3 (sept, J )
7 Hz, 4H), 1.6 (d, J ) 7 Hz, 12H), 1.4 (d, J ) 7 Hz, 12 H).
28: 1H NMR (C6D6, 400 MHz): δ 7.7-6.0 (m, 18H), 4.3 (sept, J )
7 Hz, 2H), 3.3 (sept, J ) 7 Hz, 2H), 1.8 (d, J ) 7 Hz 6H), 1.35 (d, J
) 7 Hz 6H), 1.3 (d, J ) 7 Hz 6H), 1.15 (d, J ) 7 Hz 6H).
Acknowledgment. We thank the National Science Foundation
(CHE0107810) and DuPont for support of this work. J.C.J.
thanks the University of North Carolina for a Venable Fellow-
General Procedure for Determination of Phosphine Exchange
Kinetics. An NMR tube was charged with either catalyst 3 or 9a and
9
J. AM. CHEM. SOC. VOL. 126, NO. 18, 2004 5841

A R T I C L E S
Jenkins and Brookhart
ship. We thank Prof. T. H. Warren for communicating unpub-
lished results.
examples of rate calculations for the insertion of ethylene into
a Ni-Ph bond and for the conversion of the Ni-hexyl complex
17c to the Ni-H complex 20. This material is available free of
charge via the Internet at http://pubs.acs.org.
Supporting Information Available: Qualitative estimation
(for comparison purposes) of the equilibrium constant between
a Ni-PPh3 and a Ni-ethylene complex and representative
JA030634G
9
5842 J. AM. CHEM. SOC. VOL. 126, NO. 18, 2004