Mechanistic studies on the formation of linear polyethylene chain catalyzed by palladium phosphine-sulfonate complexes: Experiment and theoretical studies

Article abstract of DOI:10.1021/ja9047398

Linear polyethylene propagation starting from Pd phosphine-sulfonate complexes, Pd(CH₃)-(L)(Ar₂PC₆H ₄SO₃) (L = 2,6-lutidine, Ar ) o-MeOC₆H ₄ (2a) and L = pyridine, Ar = Ph (2b)), was studied both experimentally and theoretically. Experimentally, highly linear polyethylene was obtained with Pd(CH₃)(L)(Ar₂PC₆H ₄SO₃) complexes 2a and 2b. Formation of a long alkyl-substituted palladium complex (3) was detected as a result of ethylene oligomerization on a palladium center starting from methylpalladium complex. Additionally, well-defined ethyl and propyl complexes (6_Et and 6 _Pr) were synthesized as stable n-alkyl palladium complexes. In spite of the existence of β-hydrogens, the β-hydride elimination to give 1-alkenes was very slow or negligible in all cases. On the other hand, isomerization of 1-hexene in the presence of a methylpalladium/phosphine- sulfonate complex 2a indicated that this catalyst system actually undergoes β-hydride elimination and reinsertion to release internal alkenes. On the theoretical side, the relative energies were calculated for intermediates and transition states for chain-growth, chain-walking, and chain-transfer on the basis of the starting model complex Pd(n-C₃H₇)(pyridine) (o-Me₂PC₆H₄SO₃) (8). First, cis/trans isomerization process via the Berry's pseudorotation was proposed for the Pd/phosphine-sulfonate system. The second oxygen atom of sulfonate group is involved in the isomerization process as the associative ligand, which is one of the most unique natures of the sulfonate group. Chain propagation was suggested to take place from the less stable alkylPd(ethylene) complex 10′ with the TS of 27.4/27.7 ((E+ZPC)/G) kcal/mol. Possible β-hydride elimination was suggested to occur under low concentration of ethylene: the highest-energy transition state to override for β-hydride elimination was either >37.4/25.3 kcal/mol (TS(9-12)) or 29.1/27.4 kcal/mol (TS(8′-9′) to reach 12′). The ethylene insertion to the iso-alkylpalladium species (14′) is allowed via a TS of 28.6/29.1 kcal/mol (TS(14′-15′)), slightly higher in energy than that for the normal-alkylpalladium species (TS(10′-11′)). Easy chain transfer was suggested to proceed from the more stable PdH(olefin) complex 12′ if β-hydride elimination to 12′ does take place. Thus, the production of linear polyethylene with high molecular weight under ethylene pressure suggests that the cis and trans PdH(alkene)(phosphine-sulfonate) complexes (12 and 12′) are merely accessible in the presence of excess amount of ethylene.

Full text of DOI:10.1021/ja9047398

Published on Web 09/11/2009
Mechanistic Studies on the Formation of Linear Polyethylene
Chain Catalyzed by Palladium Phosphine-Sulfonate
Complexes: Experiment and Theoretical Studies
Shusuke Noda,† Akifumi Nakamura,† Takuya Kochi,†,§ Lung Wa Chung,‡
Keiji Morokuma,*,‡ and Kyoko Nozaki*,†
Department of Chemistry and Biotechnology, Graduate School of Engineering, The UniVersity of
Tokyo, 7-3-1 Hongo, Bunkyo-Ku, Tokyo 113-8656, Japan, and Fukui Institute for Fundamental
Chemistry, Kyoto UniVersity, Takano-Nishishiraki-cho, 34-4, Sakyo-ku, Kyoto 606-8103, Japan
Received June 15, 2009; E-mail: nozaki@chembio.t.u-tokyo.ac.jp
Abstract: Linear polyethylene propagation starting from Pd phosphine-sulfonate complexes, Pd(CH3)-
(L)(Ar2PC6H4SO3) (L ) 2,6-lutidine, Ar ) o-MeOC6H4 (2a) and L ) pyridine, Ar ) Ph (2b)), was studied
both experimentally and theoretically. Experimentally, highly linear polyethylene was obtained with
Pd(CH3)(L)(Ar2PC6H4SO3) complexes 2a and 2b. Formation of a long alkyl-substituted palladium complex
(3) was detected as a result of ethylene oligomerization on a palladium center starting from methylpalladium
complex. Additionally, well-defined ethyl and propyl complexes (6Et and 6Pr) were synthesized as stable
n-alkyl palladium complexes. In spite of the existence of ꢀ-hydrogens, the ꢀ-hydride elimination to give
1-alkenes was very slow or negligible in all cases. On the other hand, isomerization of 1-hexene in the
presence of a methylpalladium/phosphine-sulfonate complex 2a indicated that this catalyst system actually
undergoes ꢀ-hydride elimination and reinsertion to release internal alkenes. On the theoretical side, the
relative energies were calculated for intermediates and transition states for chain-growth, chain-walking,
and chain-transfer on the basis of the starting model complex Pd(n-C3H7)(pyridine)(o-Me2PC6H4SO3) (8).
First, cis/trans isomerization process via the Berry’s pseudorotation was proposed for the Pd/
phosphine-sulfonate system. The second oxygen atom of sulfonate group is involved in the isomerization
process as the associative ligand, which is one of the most unique natures of the sulfonate group. Chain
propagation was suggested to take place from the less stable alkylPd(ethylene) complex 10′ with the TS
of 27.4/27.7 ((E+ZPC)/G) kcal/mol. Possible ꢀ-hydride elimination was suggested to occur under low
concentration of ethylene: the highest-energy transition state to override for ꢀ-hydride elimination was either
>37.4/25.3 kcal/mol (TS(9-12)) or 29.1/27.4 kcal/mol (TS(8′-9′) to reach 12′). The ethylene insertion to the
iso-alkylpalladium species (14′) is allowed via a TS of 28.6/29.1 kcal/mol (TS(14′-15′)), slightly higher in
energy than that for the normal-alkylpalladium species (TS(10′-11′)). Easy chain transfer was suggested
to proceed from the more stable PdH(olefin) complex 12′ if ꢀ-hydride elimination to 12′ does take place.
Thus, the production of linear polyethylene with high molecular weight under ethylene pressure suggests
that the cis and trans PdH(alkene)(phosphine-sulfonate) complexes (12 and 12′) are merely accessible
in the presence of excess amount of ethylene.
1996.2 The branched structure was suggested to arise from so-
Introduction
called “chain walking”, that is, rapid ꢀ-hydride elimination from
Group 4 transition metal catalysts play the major role in
industrial production of linear high-density polyethylene (HDPE).
On the other hand, significant efforts have been intensively
devoted in the past decade to developing late transition-metal
catalysts for ethylene polymerization, since they were considered
to be potentially capable of producing copolymers of ethylene
and readily available polar olefins by coordination polymeri-
zation.1 Using R-diimine palladium complexes, Brookhart
reported the coordination copolymerization of ethylene with
methyl acrylate to produce highly branched copolymers in
a normal-alkylmetal intermediate giving a hydridometal-olefin
complex, reinsertion of the olefin to the metal-hydride bond
providing iso-alkylmetal species, and chain growth from the
iso-alkylmetal complex.
In 2002, a group at Shell laboratory reported the synthesis
of linear copolymers of ethylene and alkyl acrylates using in
situ generated Pd catalysts by admixture of palladium precursor
(1) (a) Ittel, S. D.; Johnson, L. K.; Brookhart, M. Chem. ReV. 2000, 100,
1169–1203. (b) Gibson, V. C.; Spitzmesser, S. K. Chem. ReV. 2003,
103, 283–315. (c) Boffa, L. S.; Novak, B. M. Chem. ReV. 2000, 100,
1479–1493. (d) Sen, A.; Borkar, S. J. Organomet. Chem. 2007, 692,
3291–3299.
† The University of Tokyo.
‡ Kyoto University.
§ Present address: Department of Chemistry, Faculty of Technology, Keio
University, 3-14-1 Hiyoshi, Kohoku-ku, Yokohama, Kanagawa 223-8522,
Japan.
(2) (a) Johnson, L. K.; Mecking, S.; Brookhart, M. J. Am. Chem. Soc.
1996, 118, 267–268. (b) Mecking, S.; Johnson, L. K.; Wang, L.;
Brookhart, M. J. Am. Chem. Soc. 1998, 120, 888–899.
9
14088 J. AM. CHEM. SOC. 2009, 131, 14088–14100
10.1021/ja9047398 CCC: $40.75  2009 American Chemical Society

Transition-Metal Catalysts for Ethylene Polymerization
A R T I C L E S
and phosphonium-sulfonate 1a.3 As the copolymer thus
obtained can be considered as a functionalized HDPE, the report
attracted much attention from both academia and industry. Since
then, incorporation of several common polar vinyl monomers
to linear polyethylene were reported for acrylates,4-6 acryloni-
trile,7 vinyl fluoride,8 vinyl ethers,9 and others.10,11 Among the
late transition-metal catalysts, the first-row transition metal such
as Ni, Fe, and Co were usually used to obtain linear polyeth-
ylenes.12 In contrast, few available examples of Pd catalyst can
produce linear polyethylene, and these limited examples require
the presence of activators, such as MAO.13 It should be noted
that the Pd phosphine-sulfonate complexes are active without
using any activators or noncoordinating counterions.
another ꢀ-hydride elimination pathway with lower energy
(TS(9′-12′) in Figure 8) during our preliminary calculations on
the basic reactions of this system. This apparent contradiction
prompted us for further studies on the ethylene polymerization
using Pd phosphine-sulfonate complexes. Here we report our
independent study on the polyethylene synthesis by using
catalyst 2 both experimentally and theoretically. We isolated
and characterized the growing ethylene chain on the isolated
methylpalladium complex 2a. In addition, our detailed theoreti-
cal calculations revealed that the concentration of ethylene may
control the molecular weight and the microstructure of the
polyethylene product.
Results and Discussions
1. Characterization of Alkylpalladium Complexes of Phosphine-
Sulfonate as an Intermediate to Provide Linear Polyethylene.
Formation of linear polyethylene by using palladium complex
derived from phosphonium-sulfonate 1a was recently reported
by several research groups.5-7,10,14 The results we independently
examined with methylpalladium lutidine complex 2a are shown
in Table S-1 in Supporting Information. In addition, we prepared
methylpalladium complex 2b having phenyl groups on phos-
phorus atom and pyridine as the forth ligand. Under ethylene
pressure of 3.0 MPa in 15 h in toluene, highly linear polyeth-
ylene was obtained. Higher catalytic activity and molecular
weight were observed with 2a (15 g ·mmol-1 h-1, Mn 63000)
than 2b (7.3 g ·mmol-1 h-1, Mn 6000). Here we report our
observation of polyethylene chain growth on the palladium
center by NMR spectroscopy, which accords with the coordina-
tion polymerization mechanism. When a solution of complex
2a in CDCl3 was treated with 1.0 atm of ethylene in an NMR
tube, formation of a palladium complex 3 bearing a linear long
alkyl chain was detected (Scheme 1).
In order to elucidate the above-mentioned unique properties
of the catalyst system, the reaction mechanism of ethylene
polymerization catalyzed by the Pd phosphine-sulfonate system
was studied both experimentally5-7,10,14 and theoretically.15 By
using the theoretical approach, Ziegler et al. reported that the
ꢀ-hydride elimination is slower compared to the polymer-chain
growth in the catalyst system.15 On the contrary, occurrence of
chain walking via repetitive ꢀ-hydride elimination and reinser-
tion of the olefin was suggested by Jordan et al. on the basis of
microstructure analysis of the polymers and their observation
of ꢀ-chloride elimination from ω-chloroalkylpalladium species.14
The ꢀ-hydride elimination-reinsertion process was also sug-
gested by Claverie based on their observation of 1-octene
isomerization into internal octenes.10 Concurrently, we found
Scheme 1. Formation of Long Alkyl Complex
(3) Drent, E.; van Dijk, R.; van Ginkel, R.; van Oort, B.; Pugh, R. I.
Chem. Commun. 2002, 744–745.
(4) Kochi, T.; Yoshimura, K.; Nozaki, K. Dalton Trans. 2006, 25–27.
(5) Skupov, K. M.; Marella, P. R.; Simard, M.; Yap, G. P. A.; Allen, N.;
Conner, D.; Goodall, B. L.; Claverie, J. P. Macromol. Rapid Commun.
2007, 28, 2033–2038.
All the peaks of 1H NMR chart were unambiguously
1
characterized by COSY (Figure 1) and H{31P} NMR (Figure
(6) Guironnet, D.; Roesle, P.; Ru¨nzi, T.; Go¨ttker-Schnetmann, I.; Mecking,
S. J. Am. Chem. Soc. 2009, 131, 422–423.
S-1 in Supporting Information). Palladium-lutidine complex
3 was detected as a predominant species accompanied by 3.5%
of the corresponding palladium-ethylene complex, Pd(alkyl)-
(CH2dCH2)(Ar2PC6H4SO3), the values being calculated based
on 1H NMR peaks for coordinating ethylene (δ 4.91-5.02) and
for free lutidine (δ 2.52). The methylene protons at the ꢀ- and
γ-positions of palladium exhibited their peaks with unexpectedly
high-field shift, δ 0.43 and 0.47, respectively (Vide infra). The
long Pd-alkyl complex 3 was reasonably stable even in the
absence of ethylene. No change was detected for complex 3 in
a solid state for several days, but slow decomposition was
observed in a halogenated solvent. In a chloroform or dichlo-
roethane solution, chloropalladium complex 4 was formed as a
result of replacement of the alkyl group by chloride (Scheme
2). This decomposition may have occurred via homolytic
scission of the palladium-carbon bond followed by the abstrac-
tion of a chlorine atom from the halogenated solvent by the
palladium center radical. The structure of complex 4 was
confirmed by X-ray single crystal analysis (Tables S2-S7,
Supporting Information).
(7) Kochi, T.; Noda, S.; Yoshimura, K.; Nozaki, K. J. Am. Chem. Soc.
2007, 129, 8948–8949.
(8) Weng, W.; Shen, Z.; Jordan, R. F. J. Am. Chem. Soc. 2007, 129,
15450–15451.
(9) Luo, S.; Vela, J.; Lief, G. R.; Jordan, R. F. J. Am. Chem. Soc. 2007,
129, 8946–8947.
(10) Skupov, K. M.; Piche, L.; Claverie, J. P. Macromolecules 2008, 41,
2309–2310.
(11) Borkar, S.; Newsham, D. K.; Sen, A. Organometallics 2008, 27, 3331–
3334.
(12) (a) Kuhn, P.; Semeril, D.; Matt, D.; Chetcuti, M. J.; Lutz, P. Dalton
Trans. 2007, 515–528. (b) Younkin, T. R.; Conner, E. F.; Henderson,
J. I.; Friedrich, S. K.; Grubbs, R. H.; Bansleben, D. A. Science 2000,
287, 460–462. (c) Gibson, V. C.; Redshaw, C.; Solan, G. A. Chem.
ReV. 2007, 107, 1745–1776.
(13) (a) Tsuji, S.; Swenson, D. C.; Jordan, R. F. Organometallics 1999,
18, 4758–4764. (b) Li, K. L.; Darkwa, J.; Guzei, I. A.; Mapolie, S. F.
J. Organomet. Chem. 2002, 660, 108–115. (c) Mohlala, M. S.; Guzei,
I. A.; Darkwa, J.; Mapolie, S. F. J. Mol. Catal. A: Chem. 2005, 241,
93–100. (d) Chen, R.; Mapolie, S. F. J. Mol. Catal. A: Chem. 2003,
193, 33–40.
(14) Vela, J.; Lief, G. R.; Shen, Z.; Jordan, R. F. Organometallics 2007,
26, 6624–6635.
(15) Haras, A.; Anderson, G. D. W.; Michalak, A.; Rieger, B.; Ziegler, T.
Organometallics 2006, 25, 4491–4497.
9
J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009 14089

A R T I C L E S
Noda et al.
Scheme 3. Synthesis of Alkyl Complexes 6Et and 6Pr
the R-methylene and δ 0.95 for the ꢀ-methyl. Both R- and
ꢀ-protons coupled with the phosphorus atom as were confirmed
1
by H{31P} NMR (Figure 2). In n-propyl complex 6Pr, the
γ-methyl protons were also observed at higher field, δ 0.17
(Figure 3, bottom). Observation of the n-propyl complex 6Pr is
in sharp contrast to the report by Brookhart showing isomer-
ization of a normal-propylpalladium complex to an iso-
propylpalladium under even -110 °C in the absence of
additional Lewis base (2,6-lutidine in our case).18 In complex
6Pr, not only the R- and ꢀ-methylene protons, but the γ-carbon
coupled with the phosphorus atom (JP-C ) 4 Hz). In both
complexes 6Et and 6Pr, the methyl proton peak for free 2,6-
lutidine (2.5 ppm) was not observed.
Figure 1. Formation of 3 from 2a and ethylene (HH COSY in CDCl3).
Scheme 2. Decomposition of Alkyl Complex
The high stability of the long Pd-alkyl complex 3 encour-
aged us to prepare an alkyl palladium complex with a certain
carbon number. The precursors, dichloropalladium 5 and
chloropalladium(lutidine) 4 were easily prepared following
to Scheme 3. The preparation of ethyl and n-propyl com-
plexes 6Et and 6Pr are achieved by the transmetalation reaction
of trialkylaluminum species with chloropalladium 4.16 The
monoalkyl complexes were isolated as yellow solids in
moderate yields. As shown in the NMR spectra of 6Et, the
peak for the R-methylene protons was detected at δ 1.02 and
that for the ꢀ-methyl protons at δ 0.00 (Figure 2, bottom).
These values are upfield shifted, compared to the peaks reported
for ethylpalladium(phenanthroline)(pyridine)complex:17 δ 2.04 for
1
Figure 2. 1H and H{31P} NMR spectra of 6Et (in CDCl3).
One possible reason for the upfield shift of R-γ protons and
the P-H and P-C couplings in 6Et and 6Pr may be attributed
to shield effect by aromatic rings of the ligand. Another
possibility is an agostic interaction of one of the C-H bonds
with the metal center represented as Scheme 4. In fact, the
contribution of the ꢀ- or γ-agostic alkylpalladium species
accompanied by lutidine dissociation was suggested by the
following experiment. Addition of extra amount of 2,6-lutidine
to a solution of complex 3 generated from 6Pr resulted in the
disappearance of the peak(s) around δ 0.50.
(16) Considering our theoretical studies in this paper, the stability of 6Et
and 6Pr may be explained by the following two possibilities. (1)
ꢀ-hydride elimination from 6Et or 6Pr (corresponds to TS(9-12) and/
or TS(8′-9′)) is slow under the conditions for our NMR studies. (2)
The normal-complex 6Pr is more stable than its iso-alkyl isomer (an
analogue of 13+pyridine) due to the bulkiness of the ligand 1a and
of the lutidine.
(17) Milani, B.; Marson, A.; Scarel, A.; Ernsting, G. M. M.; Elsevier, C. J.
Organometallics 2004, 23, 1974–1977.
(18) (a) Shultz, L. H.; Tempel, D. J.; Brookhart, M. J. Am. Chem. Soc.
2001, 123, 11539–11555. (b) Tempel, D. J.; Johnson, L. K.; Huff,
R. L.; White, P. S.; Brookhart, M. J. Am. Chem. Soc. 2000, 122, 6686–
6700.
9
14090 J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009

Transition-Metal Catalysts for Ethylene Polymerization
A R T I C L E S
1
Figure 3. 1H and H{31P} NMR spectra of 6Pr (in CDCl3).
Figure 4. Isomerization of 1-hexene (1H NMR in C2D4Cl2).
Scheme 4. Dissociation of 2,6-Lutidine from 3 As a Possible
Reason for the Up-Field Shift
Figure 5. Formation of n-hexyl complex 6Hex starting from complex 2a
(31P NMR in C2D4Cl2).
2. Isomerization of 1-Hexene in the Presence of Complex
2a. When a solution of methylpalladium complex 2a (0.010
mmol) in 0.5 mL of C2D4Cl2 was treated with 13 equivalent
amount of 1-hexene in an NMR tube, a terminal olefin was
slowly converted to internal olefins (Figure 4). After 19 h at
60 °C, 63% (7.8 mmol) of the 1-hexene was converted to
internal olefins. Most likely, this transformation was caused
by terminal olefin insertion to a palladium-carbon or
palladium-hydrogen bond followed by ꢀ-hydride elimina-
tion. Thus, it is proven that the ꢀ-hydride elimination does
proceed with this catalyst system, at least in the absence of
ethylene.10 At the same time, 31P NMR spectrum shows
several peaks including 24.5 ppm which corresponds to the
normal-hexylpalladium complex 6Hex (24 µmol, after 19 h)
generated by 1,2-insertion of 1-hexene into a Pd-H bond.
In other words, the normal-alkyl complex 6Hex exists as a
stable species as complex 3. The chloropalladium complex
4 gradually appeared probably via the radical decomposition
process (38 µmol, after 19 h) (Figure 5).
Figure 6. Simplified structure of phosphine-sulfonate palladium complexes.
The propyl complex 6Pr was simplified as 8,21 and the relative
energies (E+ZPC, with zero point energy correction, before /)
and free energies (G, after /) for all the stationary points will
be expressed relative to the complex 8 in the following studies.
Complexes with phenyl groups on the phosphorus atom were
also employed to study some of the key steps (labeled with a
subscript of Ph, such as 8Ph). Complexes with Ph suffix should
be considered as real species because complex 2b did produce
linear polyethylene as was described in the former part of this
3. Results of Theoretical Calculations
In our theoretical studies, density functional theory (B3LYP19/
6-31G* and Lanl2dz20) was employed. As a model for calcula-
tions, the palladium/phosphine-sulfonate complex 2 was sim-
plified as 7 with replacement of the aryl groups on phosphorus
atom by methyl groups and 2,6-lutidine by pyridine (Figure 6).
(19) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652. (b) Lee, C.;
Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785–789.
(21) Additionally, we found that Pd complexes bearing alkylphosphine-
sulfonate ligands are also effective catalysts for the production of
highly linear polyethylene. See Supporting Information.
(20) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270–283.
9
J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009 14091

A R T I C L E S
Noda et al.
studied.23,24 One possible pathway is a simple geometry change
via tetrahedral four-coordinate species (Scheme 5, rotation A).
In fact, we could locate tetrahedral transition state (TS(10-
10′)tetrahedral) for cis/trans isomerization between the alky-
lPd(ethylene) complexes 10 and 10′, which is 19.3/20.3 kcal/
mol above complex 10.
Another possible cis/trans isomerization of the square-planar
palladium complexes is an associative pathway which is
accompanied by ligand association-substitution reactions.23 In
this case, five-coordinate structures can be considered as
transition states for such isomerization. Our trials to locate a
transition state for isomerization of 10 to 10′ involving an
external ethylene as the fifth ligand resulted in a simple
dissociation of ethylene. Instead, for the Pd phosphine-sulfonate
system, we found that the second oxygen atom of the sulfonate
group is inVolVed in the isomerization process as the associatiVe
ligand (Scheme 5, rotation B). For instance, cis/trans isomer-
ization process from cis-alkylPd(ethylene) complex 10 to its
trans-isomer 10′ was facilitated by simultaneous exchange of
oxygen atoms of the SO3 group. The geometry of this transition
state TS(10-10′) resembles square pyramidal where the phos-
phorus atom is located at the apical position. This TS(10-10′)
involving the SO3-rotation is 15.9/16.1 kcal/mol above 10
(Scheme 5), which is lower than the simple rotation TS (19.3/
20.3 kcal/mol) discussed in the previous paragraph. This rotation
pattern (Scheme 5, rotation B) via five-coordinate complex is
recognized as “Berry’s pseudorotation” based on the TS
structure.25 Similar to TS(10-10′), we have successfully found
cis/trans isomerization pathway through TS(9-9′) (Scheme 7)
from 9 to 9′ with exchanging oxygen atoms of sulfonate group.
The agostic hydrogen atom has interaction to the metal center
in TS(9-9′). The geometry of this transition state is also similar
to square pyramidal where the phosphorus atom is located at
the apical position. As another possibility for isomerization from
9 to 9′, one may consider the dissociation-isomerization
pathway, that is a loss of agostic C-H interaction and
isomerization from one T-shape structure to another.23 Never-
theless, we could not find such processes between the alkylpal-
ladium complexes 9 and 9′.
Figure 7. Optimized structure of 2a. The hydrogen atoms are omitted for
clarity.
Table 1. Selected Bond Lengths and Angles of 2a Determined by
the X-ray Crystallography and Calculations (in parentheses)
bond
length [Å]
angle
[deg]
Pd-C
Pd-N
Pd-O
Pd-P
2.134 (2.056)
2.134 (2.181)
2.159 (2.190)
2.234 (2.311)
C-Pd-N
N-Pd-O
O-Pd-P
P-Pd-C
C-Pd-O
P-Pd-N
88.66 (91.71)
89.99 (84.77)
95.00 (93.54)
86.44 (89.99)
178.53 (175.65)
173.29 (178.19)
Table 2. Substituents Effect on the Relative Energies of Some
Key Intermediates and Transition States (kcal/mol)
with Me
E+ZPE
H
G
with Ph
E+ZPE
H
G
8
0.0
0.0 0.0 8Ph
0.0
0.0 0.0
TS(8′-9′)
TS(9′-10′)
29.1 29.2 27.4 TS(8′Ph-9′Ph
33.0 32.8 30.9 TS(9′Ph-10′Ph
)
28.6 28.9 26.0
32.2 32.0 30.3
26.8 25.7 26.1
42.2 41.8 28.9
36.4 35.9 24.5
30.7 30.4 18.4
28.0 26.8 28.8
42.9 42.9 18.6
26.0 25.5 25.8
)
TS(10′-11′) 27.4 26.1 27.7 TS(10′Ph-11′Ph
)
TS(9-9′)
TS(9-12)
TS(9′-12′)
44.5 44.1 31.7 TS(9Ph-9′Ph
no TS found TS(9Ph-12Ph
31.5 31.0 19.9 TS(9′Ph-12′Ph
)
)
)
TS(14′-15′) 28.6 27.4 29.1 TS(14′Ph-15′Ph
19′ 43.0 42.8 19.4 19′Ph
TS(12′-17′) 24.6 24.1 24.2 TS(12′Ph-17′Ph
)
)
paper. The optimized structure of 2a was in reasonable
agreement with the X-ray crystal structure,7 as shown in Figure
7 and Table 1.
The processes we studied include cis/trans isomerization of
palladium phosphine-sulfonate complexes (section 3.1), chain-
propagation reaction to form linear polyethylene (section 3.2),
ꢀ-hydride elimination reactions (section 3.3), chain-propagation
reaction to form branched-polyethylene (section 3.4), chain
transfer reaction to release 1-alkene (section 3.5), and calcula-
tions for representative intermediates and transition states with
the real system (section 3.6). The representative data of the real
system relative to 8Ph are summarized in Table 2 in section 3.1.
The overall calculated energy profile for sections 3.1-3.4 is
summarized in Figure 8 for perspective. After the calculation
results in sections 3.1-3.6, we will discuss the overall reaction,
especially focusing on Routes 1-3 in Figure 8, in order to clarify
why Pd phosphine-sulfonate catalysts can produce highly linear
polyethylene, selectively.
3.1. cis/trans Isomerization of Four-Coordinate and ꢀ-Ago-
stic Three-Coordinate Pd Phosphine-Sulfonate Complexes
and Their Stability. Due to the unsymmetric character of the
phosphine-sulfonate bidentate ligand, cis/trans isomerization
should be considered for ꢀ-agostic cis-alkyl complex 9 as well
as for four-coordinate complexes such as cis-alkylPd(pyridine)
complex 8 and cis-alkylPd(ethylene) complex 10.22 Here in this
study, we have revealed that the sulfonate group plays a unique
role to facilitate the isomerization. The cis/trans isomerization
mechanisms of square-planar palladium complexes were well
The cis/trans isomerization from 8 to 8′ was also searched.
In this case, we found that the reaction takes place in two steps
(Scheme 6) with an intermediate (8med), as was confirmed by
intrinsic reaction coordinate (IRC) calculations. In the course
of 8 f TS(8-8med) f 8med f TS(8med-8′) f 8′, the oxygen
atoms of the SO3 group exchanged. For both 8 and 9, TSs via
simple rotation via tetrahedral four-coordinate species were also
calculated giving higher barriers than the TSs for isomerization
with SO3 rotation.
Isomerization pathway via dissociation of the sulfonate group
from the metal center is less probable. Dissociation of the SO3
(22) Although cis/trans isomerization reactions of a series of Pd phosphine-
sulfonate complexes were proposed by Ziegler et al. in reference 15
the structures of the TSs were not fully reported. See also: (a) Haras,
A.; Michalak, A.; Rieger, B.; Ziegler, T. J. Am. Chem. Soc. 2005,
127, 8765–8774. (b) Haras, A.; Michalak, A.; Rieger, B.; Ziegler, T.
Organometallics 2006, 25, 946–953.
(23) Anderson, G. K.; Cross, R. J. Chem. Soc. ReV. 1980, 9, 185–215.
(24) For some examples of the studies on cis/trans isomerization reactions
of square-planar Pd complexes, see: (a) Deubel, D. V.; Ziegler, T.
Organometallics 2002, 21, 4432–4441. (b) Frankcombe, K. E.; Cavell,
K. J.; Yates, B. F.; Knott, R. B. Organometallics 1997, 16, 3199–
3206. (c) Casares, J. A.; Espinet, P. Inorg. Chem. 1997, 36, 5428–
5431. (d) Casares, J. A.; Espinet, P. Inorg. Chem. 1998, 37, 2096.
(25) Ugi, I.; Marquard., D.; Klusacek, H.; Gillespi., P.; Ramirez, F. Acc.
Chem. Res. 1971, 4, 288–296. (a) See also references 24c and d.
9
14092 J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009

Transition-Metal Catalysts for Ethylene Polymerization
A R T I C L E S
Figure 8. Energy profile for chain propagation to give linear and branched alkyl chain, and ꢀ-hydride elimination from alkyl Pd species. The Gibbs free
energies (in kcal/mol) relative to 8 are given. Bold: Route 1, Blue: Route 2, Red: Route 3 in Discussion section (Scheme 11).
Scheme 5. Two Possible Routes for Isomerization of Pd
Phosphine-Sulfonate Complexes
in agreement with the X-ray analyses of the reported Pd
phosphine-sulfonate complexes.4-7,14 This preference can be
attributed to the strong trans influence of alkyl group, hydrogen
atom, and phosphine moiety which dislike being trans to each
other.14,22
3.2. Chain-Propagation Reaction to Form Linear Polyeth-
ylene. Chain-propagation reaction proceeds via Cossee mech-
anism including olefin insertion into the metal-carbon bond.
Four-membered ring transition states for ethylene insertion were
calculated for two isomeric ethylene complexes 10 and 10′ to
produce 11 and 11′, respectively. Possible intermediates and
transition states for the ethylene insertion to the carbon-palladium
bond starting from the propyl complex 8 are drawn in Scheme
6. First, replacement of pyridine by ethylene takes place by an
associative mechanism via TS(8-10) (5.6/17.5 kcal/mol). As
described in the previous section, cis/trans isomerization is
possible from 10 to 10′. Although the isomer 10 is lower in
energy than the other isomer 10′, the transition state TS(10-11)
for ethylene insertion is higher than the other transition state
TS(10′-11′) by 11.4/11.2 kcal/mol. Lower energy of TS(10′-
11′) can be attributed to the stronger trans effect of the
phosphine moiety than the sulfonate oxygen. Because olefin
insertion is recognized as the migration of alkyl group to the
ethylene, the strong trans effect of the phosphorus atom should
enhance the migrating ability of the alkyl group leading to the
rapid insertion. Accordingly, the propyl group in 10 undergoes
isomerization to form the more reactive isomer 10′ with the
alkyl group trans to the phosphine via TS(10-10′).26,27 Although
group from the complex 8 provided a three-coordinate inter-
mediate at quite high energy level (8-SO3away, 31.7/29.7 kcal/
mol from 8, see Supporting Information). The rigid structure
of the phenylene backbone of the bidentate ligand seems to
prevent dissociation of the SO3 group.
These isomerization processes TS(8-8′), TS(9-9′),and TS(10-
10′) require rather similar energy barriers from their reactants
8, 9, and 10. However, the relative energies of these TSs based
on the complex 8 are different because relative stability of the
reactants 8, 9, and 10 are not equal. Therefore, some isomer-
ization can proceed, while the others cannot, as will be discussed
below. Other cis/trans isomerizations, namely, TS(12-12′),
TS(13-13′), and TS(14-14′) (Scheme 8) were also found to occur
with rotation of the sulfonate group. These TSs will appear in
the following sections without detailed discussions. For more
structural information, see Supporting Information.
(26) Similar results were obtained in the case of other olefin polymerization
catalysts bearing unsymmetric bidentate ligands. For example,(a) Chan,
M. S. W.; Deng, L. Q.; Ziegler, T. Organometallics 2000, 19, 2741–
2750. (b) Michalak, A.; Ziegler, T. Organometallics 2003, 22, 2069–
2079. (c) See also references 15, 22a,b, and 24a.
(27) More detailed analysis on ethylene insertion regarding the nature of
the unsymmetric character of this ligand is described in Supporting
Information.
For all pairs of cis/trans isomers in the intermediates, it is
apparent from Figure 8 that the alkyl group or hydrogen atom
located trans to oxygen is favored over their isomers. This is
9
J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009 14093

A R T I C L E S
Noda et al.
Scheme 6. Propagation Reaction for Linear Alkyl Complex (E+ZPE/G, kcal/mol)
8 f 8med f 8′ f 10’ is another possible route from 8 to 10′,
we omit this discussion because 8 f 10 f TS(10-10′) f 10′
is lower in energy than TS(10′-11′), anyway. Thus, processes
via 8 f 10 f TS(10-10′) f 10′ f TS(10′-11′) f 11′ complete
the propagation reaction.15 Immediate association of pyridine
to 11′ without barrier regenerates an alkyl pyridine complex
8next. In addition to TS(10′-11′) for the model complex, the
transition state TS(10′Ph-11′Ph) was calculated for the realistic
complex bearing phenyl groups on the phosphorus atom showing
a relative energy of 26.8/26.1 kcal/mol from the starting propyl
complex, 8Ph (Table 2).
3.3. ꢀ-Hydride Elimination Reactions. For the ꢀ-hydride
elimination process, two pathways based on cis/trans isomers
are considered (Scheme 7). The ꢀ-agostic cis-alkyl complex 9
undergoes ꢀ-hydride elimination to give olefin hydride complex
12 via transition state TS(9-12). The CdC bond of the
coordinated olefin in the complex 12 is perpendicular to the
square-planar plane. The transition state TS(9-12) could not
be located due to its very late nature and low reverse barrier:
The relative energy of the ꢀ-elimination product 12 is 37.4/
25.3 kcal/mol relative to 8, indicating that the relative energy
of TS(9-12) is equal or slightly higher than this value (Scheme
7 and Figure 8). In addition, when the two methyl groups on
the phosphorus atom are replaced by phenyl groups, the
transition state TS(9Ph-12Ph) was located with the energy of 36.4/
24.4 kcal/mol relative to the starting complex 8Ph (Table 2).
The critical issue is that the ꢀ-hydride elimination barrier for 9
or 9Ph is reasonably high, at least comparable to the ethylene
insertion TS, TS(10′-11′) (27.4/27.7 kcal/mol) or to TS calcu-
lated for the diphenylphosphino system TS(10′Ph-11′Ph) (26.8/
26.1 kcal/mol, see Table 2).
It is of interest to note that the other isomer, ꢀ-agostic trans-
alkyl complex 9′ would undergo much lower-energy ꢀ-hydride
elimination via transition state TS(9′-12′) if 9′ exists. The
relative energy for TS(9′-12′) from 8 is 31.5/19.9 kcal/mol which
is much lower than the other isomeric TS(9-12) by at least 5.9/
5.4 kcal/mol. The lower-energy TS(9′-12′) is the result of the
stronger trans effect of the phosphorus atom than the sulfonate
oxygen. The relative energy of the favorable ꢀ-hydride elimina-
tion TS is not strongly affected by the size of the ligand
(TS(9′Ph-12′Ph): 30.7/18.4 kcal/mol). It should be noted, how-
eVer, the reactant 9′ is less accessible due to the high-energy
TSs to be oVercome, such as TS(8′-9′) (29.1/27.4 kcal/mol),
TS(9-9′) (44.5/31.7 kcal/mol), or TS(9′-10′) (29.1/27.4 kcal/
mol). In addition, the route 9 f 12 f 12′ f 9′ is also less
(28) Relative electronic energy of TS(8′-9′) is lower than that of 9′. This
indicates the existence of another lower-energy intermediate before
9′. By IRC calculation from TS(8′-9′), an intermediates which can be
recognized as “9′ + pyridine” was obtained. In fact, the relative
electronic energy of “9′ + pyridine” was lower than that of 9′;
however, the Gibbs free energy was higher than that of 9′.
9
14094 J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009

Transition-Metal Catalysts for Ethylene Polymerization
A R T I C L E S
Scheme 7. ꢀ-Hydride Elimination and Re-insertion Reactions (E+ZPE/G, kcal/mol)28
probable because cis/trans isomerization of 12 and 12′ needs
45.6/33.3 kcal/mol from 8.
alkyl complex (27.4/27.7 kcal/mol for TS(10′-11′)) by 1.2/1.4
kcal/mol; this is even higher by 1.2/2.7 kcal/mol by using the
more realistic catalyst via TS(14′Ph-15′Ph) and TS(10′Ph -11′Ph)
(Table 2). The energy difference value of 1.2/2.7 kcal/mol is
similar to the value reported for the cationic R-diimine
system.29,30
3.5. Chain-Transfer Reaction to Release 1-Alkene. If ꢀ-hy-
dride elimination reactions from 9 to 12 or from 9′ to 12′
does take place, the chain-transfer reaction, which gives a
short oligomer, should be considered. In the late transition
metal catalyzed olefin polymerization, the chain transfer
reactions via (i) direct ꢀ-hydride transfer, (ii) associative
olefin displacement, and (iii) dissociative olefin displacement
have been proposed.1a,29
3.4. Chain-Propagation Reaction to Form Branched Polyeth-
ylene. After ꢀ-hydride elimination, the coordinating olefin in
12 can reinsert into the palladium hydride bond. If the reinsertion
takes place with the opposite regiochemistry to the direction of
9, branched alkyl complex 13 would be given via TS(12-13)
(Scheme 8 and Figure 8). Although the coordinated propylene
is required to rotate from perpendicular to parallel to the square
planar plane in 12, we assume that the rotation of propylene
takes place with a low barrier prior to the processes 12 f TS(9-
12) f 9 or 12 f TS(12-13) f 13.29 The relative energy for
TS(12-13) was calculated to be 38.9/27.7 kcal/mol from starting
8. The transition state connecting complexes 12′ and 13′, TS(12-
13′), was located at 29.8/18.2 kcal/mol.
The branched alkyl complex 13 and its isomer 13′ can bind
ethylene to form the branched alkyl olefin complexes 14 and
14′, and then migratory insertion occurs via transition states
TS(14-15) and TS(14′-15′), respectively (Scheme 8). The
transition state TS(14-15) is shown to be higher in energy (40.5/
40.3 kcal/mol) than TS(14′-15′) (28.6/29.1 kcal/mol). The
relative energy for olefin insertion into the branched alkyl
complex via TS(14′-15′) is slightly higher than that into linear
The possibilities of (i) direct ꢀ-hydride transfer reactions can
be excluded (Scheme 9). From the propylpalladium(ethylene)
complex 10, the direct ꢀ-hydride transfer reaction requires 56.8/
57.0 kcal/mol to produce 16. Similarly, the isomeric complex
10′ requires 60.6/60.8 kcal/mol to undergo the direct ꢀ-hydride
transfer. These TSs are so high in energy that these pathways
are impossible to proceed.
(30) One may assume that the steric demand would elevate in the ethylene
(29) (a) Musaev, D. G.; Svensson, M.; Morokuma, K.; Stromberg, S.;
Zetterberg, K.; Siegbahn, P. E. M. Organometallics 1997, 16, 1933–
1945. (b) Froese, R. D. J.; Musaev, D. G.; Morokuma, K. J. Am. Chem.
Soc. 1998, 120, 1581–1587.
insertion TS for the iso-alkylpalladium when the longer alkyl chain is
adopted. However, the difference in TS(10′Ph-11′Ph) and TS(14′Ph
15′Ph) was quite similar if 8nextPh was used as a starting structure
instead of 8Ph
-
.
9
J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009 14095

A R T I C L E S
Noda et al.
Scheme 8. Re-insertion of the Coordinating Propylene and Subsequent Ethylene Insertion into iso-Alkylpalladium Species (E+ZPE/G,
kcal/mol)
Scheme 9. Direct ꢀ-Hydride Transfer Reaction (E+ZPC/G, kcal/
mol)
Scheme 10. Associative and Dissociative Olefin Displacement As
Possible Chain Transfer Processes (E+ZPC/G, kcal/mol)
Associative (ii) or dissociative (iii) olefin displacement seems
to happen, but only occasionally (Scheme 10). For the associa-
tive olefin displacement from 12 to 17, the transition state
TS(12-17) could not be located. However, a five-coordinate
intermediate29 18 was found with a relative energy of 35.9/
35.5 kcal/mol, suggesting that this route is less favorable.
Dissociation of the olefin from 12 requires at least a relative
energy of 69.5/45.5 kcal/mol or higher to give three-coordinate
intermediate 19. Thus, neither associative nor dissociative
pathway for the olefin displacement from 12 can take place.
On the other hand, if the PdH(olefin) complex 12′, the trans
isomer of 12, could be generated, both associative (ii) and
dissociative (iii) pathways may be possible to give ethylene
complex 17′. The relative energy of a TS from 12′ to 17′ for
the associative pathway was calculated to be 24.6/24.2 kcal/
mol, which is lower than that for the olefin insertion via TS(10′-
11′) (27.4/27.7 kcal/mol) by 2.8/3.5 kcal/mol. The preference
toward the associative olefin displacement for the realistic model
via TS(12′Ph-17′Ph) with respect to the olefin insertion via
TS(10′Ph-11′Ph) is reduced to 0.8/0.3 kcal/mol (Table 2). The
three-coordinate PdH complex 19′ was also found to be at the
relative energy of 43.0/19.4 kcal/mol, which is partly driven
by a large entropy gain in the gas phase. Assuming that the
dissociation of the olefin from 12′ to 19′ proceeds without
barrier, the dissociative pathway is likely to take place and result
in chain-transfer reaction, but it could be partly suppressed by
excess amount of ethylene (Vide infra). In summary, once the
9
14096 J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009

Transition-Metal Catalysts for Ethylene Polymerization
A R T I C L E S
palladium hydride complex 12′ is produced, it may undergo
the chain-transfer reaction via either associative or dissociative
mechanism.
ꢀ-agostic trans-alkyl complex 9′, such as pyridine dissociation
TS(8′-9′) (29.1/27.4 kcal/mol), cis-trans isomerization of
ꢀ-agostic cis-alkyl complex 9 TS(9-9′) (44.5/31.7 kcal/mol),
and ethylene dissociation TS(10′-9′) (33.0/30.9 kcal/mol). It is
notable that the reactions such as ethylene dissociation or alkyl-
isomerization require such high energies. The high barriers to
reach 9′ can be attributed to the fact that complex 9′ does have
high energy. When the isomer pairs of 8/8′, 9/9′, and 10/10′
are compared, the isomers 8′, 9′, and 10′ having propyl-groups
trans to the phosphine are higher in ∆G than the isomers 8, 9,
and 10 having alkyl groups at trans to the sulfonate by 10.7,
10.4, and 9.4 kcal/mol, respectively (Figure 8). The difference
between the three pairs is rather small. Thus, the high energy
of 9′ can be explained by the high-energy nature of the ꢀ-agostic
alkyl complex compared to that of the four-coordinate species.
In section 3.4, we found that the ethylene insertion to the
iso-alkylpalladium species is allowed via a TS (TS(14′-15′),
28.6/29.1 kcal/mol) slightly higher in energy than that for the
normal alkylpalladium species. The result nicely explains the
fact that methyl branches but not any longer branches were
detected in the polymerization of ethylene.13
In section 3.5, easy chain transfer was suggested to proceed
from the more stable PdH(olefin) complex 12′, if 12′ exists.
According to the results discussed above, the barriers for chain
propagation and ꢀ-hydride elimination are comparable. Fur-
thermore, after ꢀ-hydride elimination, both ethylene insertion
into the iso-alkylpalladium complex (branch formation) and
chain transfer may be possible. Thus, careful comparison with
experiments is necessary to interpret these calculation results
correctly.
3.6. Calculations for Representative Intermediates and
Transition States with the Realistic System. Some of the critical
intermediates and transition states were calculated with phenyl
groups on the phosphorus atom. The energy values relative to
the starting propyl complex 8Ph were summarized in Table 2.
In addition to the values ∆E+ZPE and ∆G, ∆H is also listed.
It should be emphasized that the differences in relative energies
between the simplified ligand (Me groups on phosphorus) and
the real ligand (Ph groups on phosphorus) are negligible. As
we previously mentioned in this manuscript, complex 2b with
the ligand bearing Ph groups on the phosphorus atom and
pyridine as the fourth ligand produced highly linear polyeth-
ylene(Vide supra).21
Furthermore, instead of the combination of 6-31G* and
Lanl2dz, some important intermediates and transition states were
also tested by larger triple-ꢁ basis sets plus polarization and
diffuse functions, 6-311+G** and SDD+f (see Supporting
Information).31 In addition, solvent effect was also considered
by PCM calculation using toluene as a solvent which is actually
employed in polymerization experiments (see Supporting In-
formation). According to the results with larger basis sets and
PCM calculation, some relative energies were changed; how-
ever, these differences did not alter the interpretation of the
calculation discussed below.
3.7. Discussions on the Theoretical Studies Sections 3.1-3.6.
The chain-growth, chain-walking, and chain-transfer mecha-
nisms were studied starting from a model complex 8. In section
3.1, a unique cis/trans isomerization process was proposed for
the Pd phosphine-sulfonate system, via the so-called “Berry’s
pseudorotation”. The second oxygen atom of sulfonate group
is involved in the isomerization process as the associative ligand
to lower the transition state compared to a simple rotation
mechanism. The isomerization was accompanied by simulta-
neous exchange of oxygen atoms of the SO3 group at the
transition state. We believe that this is one of the most
characteristic features of a sulfonate group, which differentiate
the phosphine-sulfonate ligand from other anionic bitentate
ligands. For all pairs of cis/trans isomers in the intermediates,
the alkyl group or hydrogen atom located trans to oxygen is
favored over their isomers. This preference can be attributed to
the strong trans influence of alkyl group, hydrogen atom, and
phosphine moiety which dislike being trans to each other.
In section 3.2, chain propagation was suggested to take place
from the less stable cis-alkylPd(ethylene) complex 10′ rather
than its trans-isomer 10 via a route, cis-alkylPd(pyridine) 8 f
10 f TS(10-10′) f 10′ f TS(10′-11′) f ꢀ-agostic cis-alkyl
complex 11′ with the highest-energy TS(10′-11′) of 27.4/27.7
kcal/mol. Our result on this propagation step is in good
accordance to the previous study by Ziegler using QM/MM,
BP86 for QM part.14
Experimentally, the following facts are reported.
(1) Linear polyethylene with high molecular weight can be
obtained under high pressure of ethylene and reducing ethylene
pressure results in lowering the molecular weight of the
polymer.14
(2) Reducing ethylene pressure resulted in increasing the Me-
branch formation.14
(3) In the absence of ethylene, isomerization of 1-hexene into
internal alkenes did occur.
These facts can be interpreted as below using our calculation
results.
The molecular weight and the microstructure of the products
are qualitatively controlled by the balance of Routes 1-3 in
Scheme 11 (the numbers are relative energies to 8 in Gibbs
free energy). Route 1 deals with ethylene insertion into the
normal-alkylpalladium complex which leads to the linear chain
propagation. Routes 2 and 3 are the most probable pathways
for ꢀ-hydride elimination to form the PdH(olefin) complexes
12 and 12′. Once 12 or 12′ are formed, branched polyethylene
can be produced. In addition, formation of 12′ leads to the chain
transfer reaction to release low molecular weight product. Thus,
the production of linear polyethylene with high molecular weight
under high ethylene pressure suggests that 12 and 12′ are merely
accessible in the presence of excess amount of ethylene. In fact,
among rate equations of these reactions (Routes 1-3), only
Route 1, insertion rate, is accelerated by ethylene concentration
term (see more details for Supporting Information). Therefore,
we conclude that the Route 1 predominates under the sufficient
ethylene pressure while Routes 2 and 3 can occur at the shortage
of ethylene.
In Section 3.3, possible ꢀ-hydride elimination from ꢀ-agostic
cis-alkyl complex 9 to Pd(H)(olefin) 12 or the reaction of
isomers, from 9′ to 12′, were studied. In addition to the TS(9-
12) proposed by Ziegler, here we reported TS(9′-12′) with much
lower relative energy. For the ꢀ-hydride elimination from 9 to
12, TS(9-12) (>37.4/25.3 kcal/mol) should be overcome. In
contrast, higher barriers should be overridden in order to reach
Although Ziegler reported that the suppression of ꢀ-hydride
elimination is the sole reason for the linearity of the polyeth-
ylene,15 they did not consider Route 3 which is indispensable
(31) (a) Andrae, D.; Haeussermann, U.; Dolg, M.; Stoll, H.; Preuss, H.
Theor. Chim. Acta 1990, 77, 123–141. (b) Martin, J. M. L.;
Sundermann, A. J. Chem. Phys. 2001, 114, 3408–3420.
9
J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009 14097

A R T I C L E S
Noda et al.
Scheme 11. Most Probable Pathways for Ethylene Insertion and ꢀ-H Elimination (∆G, based on 8)a
a
Introductions of the rate equations are shown in Supporting Information.
for understanding the mechanism of ethylene polymerization
catalyzed by Pd phosphine-sulfonate complexes. Moreover, by
considering the ethylene concentration, we could precisely
interpret the theoretical calculation results in agreement with
the experiments.
ethylene pressure, the olefin propagation dominates over ꢀ-hy-
dride elimination while ꢀ-hydride elimination eventually pro-
ceeds at the shortage of ethylene.
5. Experimental Section
General Methods. All manipulations were carried out using the
standard Schlenk technique under argon purified by passing through
a hot column packed with BASF catalyst R3-11. NMR spectra were
recorded on a JEOL JNM-ECP500 (1H: 500 MHz, 13C: 126 MHz,
31P: 202 MHz with digital resolution of 0.239, 0.960, 4.33 Hz,
respectively) or a JEOL JNM-ECS400 (1H: 400 MHz, 13C: 101
MHz, 31P: 162 MHz with digital resolution of 0.09125, 0.767, 3.46
Hz, respectively) NMR spectrometer. Size exclusion chromatog-
raphy (SEC) analyses at 145 °C were carried out with a Tosoh
instrument (HLC-8121) equipped with three columns (Tosoh
TSKgel GMHhr-H(20)HT) by eluting the columns with o-dichlo-
robenzene at 1 mL/min. Differential scanning calorimetry was
performed on a Mettler DSC 30 instrument. Samples were heated
from 50 to 300 °C at a rate of 10 °C/min, and melting points were
determined from the data obtained during the second heating run.
Fast atom bombardment mass spectrometry (FAB-MS) was carried
out on a JEOL JMS-700 spectrometer using PEG calibration
and NBA matrix solvent. Dichloromethane, toluene, THF, and
hexanes were purified by the method of Pangborn et al.32 2-{Di(2-
methoxyphenyl)phosphino}benzenesulfonic acid (1a),33 2-(diphe-
nylphosphino)benzenesulfonic acid (1b),34 and {(o-(o-MeOC6H4)2P)-
C6H4SO3}PdMe(2,6-Me2C5H3N) (2a)7 were prepared according to
literature procedures.
4. Conclusions
Linear polyethylene propagation on palladium/phosphine-
sulfonate was studied both experimentally and theoretically.
Experimentally, highly linear polyethylene was obtained with
2a and 2b. Formation of a long alkyl-substituted palladium
complex 3 was detected as a result of ethylene oligomerization
on a palladium center starting from methylpalladium complex
2a. Additionally, well-defined ethyl and propyl complexes were
synthesized as stable n-alkyl palladium complexes. In spite of
the existence of ꢀ-hydrogens, the ꢀ-hydride elimination to give
1-alkenes was very slow or negligible in all cases. On the other
hand, our observed isomerization of 1-hexene in the presence
of the complex 2a indicated that this catalyst system actually
undergoes ꢀ-hydride elimination and reinsertion to release
internal alkenes, while formation of normal-alkylpalladium
species was detected during the isomerization process.
The chain-growth, chain-walking, and chain-transfer mech-
anisms were also studied, theoretically. A unique cis/trans
isomerization process was proposed for the Pd phosphine-
sulfonate system, via the so-called “Berry’s pseudorotation” by
the aid of the second oxygen atom of sulfonate group as the
associative ligand. The barriers for chain propagation and
ꢀ-hydride elimination were calculated to be comparable. Once
the ꢀ-hydride elimination takes place, both ethylene insertion
into the iso-alkylpalladium complex (branch formation) and
chain transfer would be possible. Thus, in the presence of
(32) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Organometallics 1996, 15, 1518–1520.
(33) Drent, E.; van Dijk, R.; van Ginkel, R.; van Oort, B.; Pugh, R. I.
Chem. Commun. 2002, 964–965.
(34) Hearley, A. K.; Nowack, R. A. J.; Rieger, B. Organometallics 2005,
24, 2755–2763.
9
14098 J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009

Transition-Metal Catalysts for Ethylene Polymerization
A R T I C L E S
Polymerization of Ethylene. To a 50 mL autoclave containing
Pd complex (0.010 mmol), additive (if any), and a stir bar was
transferred 2.5 mL of toluene under argon atomosphere. The mix-
ture was stirred at rt for 10 min and charged with ethylene. The
autoclave was heated, and the mixture was stirred. After the
reaction, MeOH was added to the cooled contents of the autoclave.
Precipitated materials were collected by filtration and washed
several times with MeOH. The remaining solid was dried under
vacuum at 80 °C to afford polyethylene, which was analyzed
without further purification. The number of branches per 1000
carbons was determined by 13C NMR spectroscopy using inverse-
gated decoupling.35 SEC analysis was performed in o-dichloroben-
zene at 120 °C.
CDCl3): δ 0.7 (d, J ) 5 Hz), 125.0 (s), 128.4 (d, J ) 49 Hz),
128.6 (s), 128.7 (d, J ) 12 Hz), 129.8 (d, J ) 7 Hz), 129.9 (d, J
) 55 Hz), 130.9 (d, J ) 3 Hz), 131.1 (d, J ) 3 Hz), 134.2 (d, J )
12 Hz), 134.5 (d, J ) 2 Hz), 138.3 (br), 149.3 (d, J ) 13 Hz),
150.2 (br); 31P NMR (202 MHz, CDCl3): δ 28.7. The white powder
included CH2Cl2 (7 mol %) and absorbed water (8 mol %), which
is confirmed by 1H NMR. Anal. Calcd for C24H22NO3PPdS + 7%
CH2Cl2 + 8% H2O: C, 52.63; H, 4.09; N, 2.55. found: C, 52.63;
H, 4.17; N, 2.56.
Synthesis of [iPr2EtNH][{(o-(o-MeOC6H4)2P)C6H4SO3}PdCl2]
(5). To a solution of 0.614 g of 1a (1.53 mmol) in 20 mL of CH2Cl2
was added 2.8 mL (16 mmol) of iPr2EtN, and the mixture was stirred
for 10 min at rt. PdCl2(cod) (0.461 g, 1.61 mmol) was added to a
mixture and stirred for 1.5 h at rt. The resulting solution was filtered
through Celite, concentrated and reprecipitated with hexane. The
residual solid was washed with hexane and ether. The volatiles were
removed in vacuo to afford a yellow powder (407 mg, 57.4% yield).
mp: 129 °C dec; IR (KBr) cm-1: 2986, 1588, 1477, 1273, 1161,
Formation of 3 by Reaction of 2a with Ethylene. A solution
of 6.3 mg of 2a (0.010 mmol) in 0.5 mL of CDCl3 was frozen at
-78 °C. The NMR tube was removed from the cool bath, and 1
atm of ethylene was quickly introduced to the NMR tube. The
mixture was gradually warmed to melt, and the tube was transferred
1
to the NMR probe. The reaction was monitored at 30 °C by H
1
and 31P NMR. Formation of 3 was confirmed by 1H, 31P, and
1020, 988, 762; H NMR (400 MHz, CDCl3): δ 1.47 (d, J ) 6.6
1
1H{31P} NMR, and HH-COSY. H NMR (500 MHz, CDCl3): δ
Hz, 6H, HNCH(CH3)2), 1.55-1.64 (m, 9H, HNCH2CH3,
HNCH(CH3)2), 3.19-3.28 (m, 2H, HNCH2CH3), 3.67 (s, 6H,
CH3OC6H4), 3.83-3.92 (m, 2H, HNCH(CH3)2), 6.88 (dd, J ) 8.4,
5.0 Hz, 2H), 7.04 (dd, J ) 7.3, 7.3 Hz, 2H), 7.28-7.36 (m, 2H),
7.44-7.49 (m, 1H), 7.52 (dd, J ) 7.8, 7.8 Hz, 2H), 7.98 (dd, J )
7.4, 4.4 Hz, 1H), 8.12 (br, 2H), 9.24 (br, 1H, NH); 13C NMR (126
MHz, CDCl3) δ 12.0 (s, HNCH2CH3), 17.6 (s, HNCH(CH3)2), 18.9
(s, HNCH(CH3)2), 42.1 (s, HNCH2CH3), 54.0 (s, HNCH(CH3)2),
55.7 (s, CH3OC6H4), 111.5 (d, J ) 5 Hz), 120.7 (d, J ) 13 Hz),
126.7 (d, J ) 9 Hz), 127.8 (d, J ) 33 Hz), 128.7 (d, J ) 8 Hz),
130.2 (d, J ) 2 Hz), 130.4 (d, J ) 3 Hz), 133.9 (d, J ) 2 Hz),
134.6 (s), 134.7 (d, J ) 15 Hz), 138.6 (br d), 160.3 (s); 31P NMR
(202 MHz, CDCl3): δ 3.93; HRMS-FAB (m/z): [M + iPr2EtNH]+
calcd for C36H58Cl2N2O5PPdS, 837.2166; found, 837.2247.
0.35-0.52 (m, 4H), 0.67 (tt, J ) 7.4, 7.4 Hz, 2H), 0.82-0.88 (m,
2H), 0.88 (t, J ) 7.0 Hz, 3H), 0.93-0.98 (m, 2H), 1.00 (tt, J )
7.6, 7.6 Hz, 2H), 1.07-1.13 (m, 2H), 1.15-1.20 (m, 4H), 1.25 (s),
1.27-1.31 (m, 2H), 3.21 (s, 6H), 3.59 (s, 6H), 6.90 (dd, J ) 4.4,
8.2 Hz, 2H), 7.06 (dd, J ) 7.4, 7.4 Hz, 2H), 7.11 (d, J ) 7.8 Hz,
2H), 7.24 (dddd, J ) 7.6, 7.6, 1.1, 1.1 Hz, 1H), 7.38 (dddd, J )
7.6, 7.6, 1.1, 1.1 Hz, 1H), 7.45 (ddd, J ) 11.2, 7.8, 1.1 Hz, 1H),
7.48-7.51 (m, 2H), 7.57 (t, J ) 7.7 Hz, 1H), 7.89 (br s, 2H), 8.12
(ddd, J ) 7.8, 4.8, 1.1 Hz, 1H). 31P NMR (202 MHz, CDCl3): δ
23.5. 1H{31P} NMR showed that in the alkyl region of the 1H NMR
spectrum, 31P-decoupling essentially only changed the shape of the
multiplet at 0.93-0.98 ppm to a triplet (J ) 8.0 Hz) at 0.96 ppm
1
in the alkyl region of the H NMR spectrum. Therefore, this peak
was assigned to the signal of the methylene directly bonded to the
palladium. The HH-COSY spectra revealed the connectivity of
methylenes not giving signals at 1.25 ppm and the signals of
terminal ethyl group of the n-alkyl chain. n-Alkyl group is assigned
to be located at the cis-position to the phosphorus atom due to strong
trans-influence of alkyl group and phosphine and structural similar-
ity to 2a.
Synthesis of [{(o-(o-MeOC6H4)2P)C6H4SO3}PdCl(2,6-Me2C5-
H3N)] (4). To a solution of 0.496 mg of 5 (0.714 mmol) in 30 mL
of CH2Cl2 were added 0.744 mg of NaBArF4(0.840 mmol) and
97.8 µL (0.867 mmol) of 2,6-lutidine, and the mixture was stirred
for 5 h at rt. The solution was filtered through Celite. The solvent
was removed in vacuo, and the resulting solid was washed with
hexane and methanol. The precipitate was filtered and dried in vacuo
to afford a yellow solid (270 mg, 59.3% yield). mp: 162 °C dec;
IR (KBr) cm-1: 1589, 1466, 1283, 1163, 982, 768.; 1H NMR (500
MHz, CDCl3): δ 3.38 (s, 6H, CH3 of lutidine), 3.65 (s, 6H,
CH3OC6H4), 6.93 (dd, J ) 8.3, 5.1 Hz, 2H), 7.06-7.14 (m, 4H),
7.30-7.37 (m, 2H), 7.47-7.60 (m, 4H), 8.02 (br, 1H), 8.06 (dd, J
) 6.3, 6.3 Hz, 2H).; 13C NMR (101 MHz, CDCl3): δ 25.5 (s, CH3
of lutidine), 55.2 (s, CH3C6H4), 111.0 (d, J ) 5 Hz), 113.8 (d, J )
65 Hz), 120.9 (d, J ) 13 Hz), 122.9 (d, J ) 4 Hz), 126.6 (d, J )
55 Hz), 127.3 (d, J ) 9 Hz), 128.7 (d, J ) 9 Hz), 130.7 (d, J ) 2
Hz), 134.1 (d, J ) 2 Hz), 134.3 (d, J ) 2 Hz), 138.1 (d, J ) 12
Hz), 138.6 (s), 146.1 (d, J ) 13 Hz), 159.0 (s), 160.1 (s); 31P NMR
(162 MHz, CDCl3): δ 3.15. These NMR data are consistent with
that of the crystals obtained by the reaction of 3 with CDCl3 (Vide
supra). See Supporting Information for X-ray analysis of 4.
Palladium complex with a long alkyl chain 3 can also be
generated by performing the reaction in a Schlenk flask. To a
solution of 19 mg of 2a (0.030 mmol) in 2.5 mL of CHCl3 was
1
introduced 1 atm of ethylene. The reaction was monitored by H
and 31P NMR by taking a small fraction of the solution, evaporating
the solvent, and dissolving the residue in CDCl3. Formation of 3
1
was confirmed by H and 31P NMR (>90% conversion after 200
min). No free ethylene or coordinated ethylene was observed by
1H NMR.
Synthesis of [{o-(Ph2P)C6H4SO3}PdMe(py)] (2b). A solution
of 2-(diphenylphosphino)benzenesulfonic acid (1b) (0.342 g, 1.00
mmol) and (tmeda)PdMe2 (0.252 g, 1.00 mmol) in CH2Cl2 (15 mL)
was cooled to -78 °C and stirred for 15 min. Pyridine (0.41 mL,
5.0 mmol) was added dropwise. The mixture was stirred at -78
°C for 30 min, warmed to room temperature, and stirred for 80
min. The resulting solution was concentrated to ∼5 mL and added
dropwise to 60 mL of hexane. The precipitate was collected by
filtration and washed with hexane. The compound was purified by
recrystallization from CH2Cl2 to afford white powder (209 mg,
38.5% yield); mp: 162 °C dec; IR (KBr) cm-1: 3059, 1437, 1269,
Generation of [{(o-(o-MeOC6H4)2P)C6H4SO3}PdEt(2,6-Me2C5-
H3N)] (6Et). To a solution of 75.0 mg of 4 (0.10 mmol) in 15 mL
of toluene was added 1.2 equiv of triethylaluminium hexane
solution, and the mixture was stirred for 3 h at rt. The solution
was filtered through Celite. The solvent was removed in vacuo and
extracted with hexane. The precipitate was filtered and dried in
vacuo to afford 6Et as a yellowish white powder (22 mg) containing
1
1167, 1113, 1001, 745, 673; H NMR (500 MHz, CDCl3): δ 0.50
(d, J ) 2.5 Hz, 3H), 7.04 (ddd, J ) 1.1, 7.8, 10.1 Hz, 1H), 7.35
(ddd, J ) 1.1, 1.1, 7.7 Hz, 1H), 7.42-7.55 (m, 9 H), 7.58-7.65
(m, 4H), 7.85 (dd, J ) 7.6, 7.6 Hz, 1H), 8.28 (ddd, J ) 1.4, 4.4,
7.8 Hz, 1H), 8.79 (d, J ) 4.8 Hz, 2H); 13C NMR (101 MHz,
1
11 mol % of 4. 6Et: H NMR (500 MHz, CDCl3): δ 0.00 (td, J )
7.5, 4.6 Hz, 3H), 1.02 (dq, J ) 14.4, 7.2 Hz, 2H), 3.20 (s, 6H),
3.58 (s, 6H), 6.90 (dd, J ) 8.1, 4.3 Hz, 2H), 7.05 (dd, J ) 7.3, 7.3
Hz, 2H), 7.10 (d, J ) 7.6 Hz, 2H), 7.22-7.27 (m, 1H), 7.32-7.45
(m, 2H), 7.49 (dd, J ) 7.9, 7.9 Hz, 2H), 7.55 (dd, J ) 7.6, 7.6 Hz,
1H) 7.85 (br, 2H), 8.12 (ddd, 7.7, 7.9, 1.0 Hz, 1H); 13C NMR (126
(35) Cotts, P. M.; Guan, Z.; McCord, E.; McLain, S. Macromolecules 2000,
33, 6945–6952.
9
J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009 14099

A R T I C L E S
Noda et al.
MHz, CDCl3): δ 12.1, 16.1, 26.0, 54.9, 114.3, 121.1, 128.1, 129.8,
130.1, 132.6, 133.1, 136.4, 157.7; 31P NMR (202 MHz, CDCl3): δ
23.0.
optimized by using the Gaussian 03 program,36 and the B3LYP19
method with combined basis sets (Lanl2dz basis sets and effective
core potential for the palladium atom20 and 6-31G(d) basis sets
for the other atoms). Harmonic vibration frequency calculations at
the same level were performed to verify all stationary points as
local minimum (with no imaginary frequency) or transition state
(with one imaginary frequency). IRC calculations37 were also
performed.
Generation of [{(o-(o-MeOC6H4)2P)C6H4SO3}PdPr(2,6-Me2C5-
H3N)] (6Pr). To a solution of 75.0 mg of 4 (0.10 mmol) in 15 mL
of toluene was added 0.30 equiv of tripropylaluminium toluene
solution, and the mixture was stirred for 1 h at rt. The solution
was filtered through Celite. The solvent was removed in vacuo and
extracted with hexane. The precipitate was filtered and dried in
vacuo to afford 6Pr as a white powder (55 mg) containing 8 mol
Acknowledgment. We are grateful to Assoc. Prof. Y. Nish-
ibayashi and Dr. Y. Miyake at the University of Tokyo for high-
resolution FAB-MS analysis of 5. Computer resources for theo-
retical calculation were mainly provided by the Research Center
for Computational Science in the National Institutes of Natural
Sciences. This work was supported by the Global COE Program
for Chemistry Innovation. A.N. is grateful to the Japan Society for
the Promotion of Science (JSPS) for a Research Fellowship for
Young Scientists. LWC acknowledges the Fukui Institute Fellow-
ship. This work is in part supported by Japan Science and
Technology Agency (JST) with a Core Research for Evolutional
Science and Technology (CREST) grant in the Area of High
Performance Computing for Multiscale and Multiphysics Phenomena.
1
% of 4. 6Pr: H NMR (500 MHz, CDCl3): δ 0.17 (t, J ) 7.1 Hz,
3H), 0.43-0.54 (m, 2H), 0.93 (dt, J ) 11.0, 5.1 Hz, 2H), 3.20 (s,
6H), 3.57 (s, 6H), 6.90 (dd., J ) 8.2, 4.4 Hz, 2H), 7.06 (dd, J )
7.3, 7.3 Hz, 2H), 7.11 (d, J ) 7.8 Hz, 2H), 7.21-7.28 (m, 1H),
7.38 (dd, J ) 7.6, 7.6 Hz, 1H), 7.45 (dd, J ) 11.2, 8.0 Hz, 1H),
7.50 (dd, J ) 7.7, 7.7 Hz, 2H), 7.56 (dd, J ) 7.7, 7.7 Hz, 1H),
7.90 (br, 2H), 8.13 (ddd, J ) 7.6, 4.8, 0.9 Hz, 1H); 13C NMR (126
MHz, CDCl3): δ 15.9 (JP-C ) 4 Hz), 21.1, 24.4, 114.4, 121.0, 129.8,
130.1, 133.1, 136.0, 158.9; 31P NMR (202 MHz, CDCl3): δ 23.2.
Lutidine addition to 3. Complex 3 could be also prepared from
6
Pr with ethylene in the same procedure as from 2a. To a solution
of 4 (13 mg, 0.016 mmol) in toluene (10 mL), was added 0.20
equiv of tripropylaluminium (3.5 µL of 0.1 M toluene solution).
After stirring under Ar for 20 min, the solution was filtered through
a pad of Celite. The solvents were removed in Vacuo to give a
crude mixture of 6Pr. Propyl complex 6Pr thus obtained was
dissolved in 0.5 mL of CDCl3 and treated with ethylene under
atmospheric pressure for 9 h at room temperature to afford 3. The
Supporting Information Available: Complete citation of
reference 36, results of ethylene polymerization (Table S-1),
NMR charts for 2b, 3, 4, 5, 6E, 6Pr (Figures S1-S10, S-13-15),
NMR studies on the reaction of 3 with lutidine (Figure S-11),
expansion of Figure 4 (Figure S-12), crystal data for 4 (Tables
S2-S8 and Figure S-16), optimized structures for all intermedi-
ates with simplified ligand (Figure S-17), Cartesian coordinates
of optimized species (Table S-7). This material is available free
of charge via the Internet at http://pubs.acs.org.
formation of 3 was confirmed by H and 31P NMR spectroscopy.
1
Upon addition of 1.9 µL of 2,6-lutidine to this solution, the peak
1
at 0.05 ppm in H NMR chart disappeared as demonstrated in
Supporting Information.
Isomerization experiment. To a solution of complex 2a (0.020
mmol) in 0.5 mL of CDCl3 was added 13 equivalent amounts of
1-hexene via syringe. The tube was sealed with screw cap, stored
in an NMR probe under 60 °C, and monitored by NMR. After 19 h
at 60 °C, 63% (7.8 mmol) of the 1-hexene converted to internal
olefins as shown in Supporting Information.
JA9047398
(36) Frisch, M. J.; et al. Gaussian 03, Revision C02; Gaussian, Inc.:
Wallingford, CT, 2007.
(37) (a) Gonzalez, C.; Schlegel, H. B. J. Chem. Phys. 1989, 90, 2154–
2161. (b) Gonzalez, C.; Schlegel, H. B. J. Phys. Chem. 1990, 94, 5523–
5527.
Theoretical Methods. All geometries of the reactants, intermedi-
ates, transition states, and products of the reactions were fully
9
14100 J. AM. CHEM. SOC. VOL. 131, NO. 39, 2009