Structure-activity relationship of palladium phosphanesulfonates: Toward highly active palladium-based polymerization catalysts

Article abstract of DOI:10.1002/chem.201103694

Palladium phosphanesulfonate [R₂P(C₆H ₄-o-SO₃)PdMeL] catalysts permit the copolymerization of an exceptional large number of functional olefins with ethylene. However, these catalysts usually have reduced activi

Full text of DOI:10.1002/chem.201103694

FULL PAPER
DOI: 10.1002/chem.201103694
Structure–Activity Relationship of Palladium Phosphanesulfonates: Toward
Highly Active Palladium-Based Polymerization Catalysts
Laurence Piche, Jean-Christophe Daigle, Gregor Rehse, and Jerome P. Claverie*[a]
Abstract: Palladium phosphanesulfo-
nate [R2P(C6H4-o-SO3)PdMeL] cata-
lysts permit the copolymerization of an
exceptional large number of functional
olefins with ethylene. However, these
catalysts usually have reduced activity.
We here have conducted a systematic
study on the influence of the phos-
phane substituent, R, on activity and
molecular weight. Phosphanes with
strong s-donating character are shown
to lead to the most active catalysts.
Thus, the catalyst based on phosphane
bis-tert-butyl-phosphanyl-benzenesul-
fonic acid (R=tBu) exhibits unprece-
dented high activity, rapidly polymeriz-
ing ethylene at room temperature to
yield a linear polymer of high molecu-
lar weight (Mw =116000 gmolÀ1). The
influence of the R group on the cata-
lyst ability to incorporate methyl acry-
late is also investigated.
Keywords: acrylate · catalytic poly-
merization · palladium · phosphane
ligands · structure–activity relation-
ships
Introduction
by the inherent moderate activity of the Drent-type catalyst
and the high cost of palladium. To overcome this problem,
several strategies were investigated, such as the replacement
of palladium by nickel[5] or ruthenium[6] (resulting in cata-
lysts that are unable to insert functional monomers), or by
changing the ligand scaffold. For example, activity increases
with less donating L ligands (Scheme 1, pyridine[4i] <lutidi-
ne[4a] <DMSO[7] ꢀno ligand[8]–coordination by a sulfonate
O lone pair). We have demonstrated that under typical poly-
merization conditions (T=858C, P=10–20 atm, cat=
86 mmolLÀ1) the proportion of pyridine-bound catalyst to
ethylene-bound catalyst is approximately 98:2, thus remov-
ing the pyridine completely should result at best in a 50 fold
The evolution of olefin polymerization catalysis since Zie-
glerꢀs discovery in 1953 has involved a prolific coupling of
polymer science with organometallic chemistry. However,
there are still no commercially viable catalysts for the con-
trolled copolymerization of simple olefins with polar func-
tional monomers.[1] Currently, commercial processes for the
copolymerization of ethylene with polar monomers employ
free radical processes which require extreme pressures and
afford little or no control over polymer architecture (tactici-
ty or crystallinity, blockiness, molecular weight and distribu-
tion thereof) and thus limit the range of materials perfor-
mance available. Starting from 1995, Brookhart et al.[2] re-
ported that cationic nickel and palladium diimines were able
to insert functional monomers such as acrylates. Due to a
chain walking mechanism the polymer was branched, with
functionalities located at the branch extremities. Thus, al-
though these catalysts demonstrate the feasibility of the cat-
alytic polymerization of polar olefins, they cannot be used
to exert a control over polymer architecture. In 2002, Drent
et al.[3] reported that palladium aryl sulfonates permit the
main-chain insertion of acrylates in polyethylene. This semi-
nal report provoked a flurry of activity as these catalysts
offer unique versatility, allowing main-chain insertion of an
unprecedented number of functional monomers.[1b,4] Howev-
er, further exploitation of these novel materials is hampered
[a] L. Piche, J.-C. Daigle, G. Rehse, Prof. J. P. Claverie
Quebec Center for Functional Materials–NanoQAM
Department of Chemistry, UQAM
Succ Centre Ville–CP 8888, Montreal, QC, H3C3P8 (Canada)
Fax : (+1)514-987-4054
E-mail: claverie.jerome@uqam.ca
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201103694.
Scheme 1. Ligands and catalysts prepared in this study.
Chem. Eur. J. 2012, 18, 3277 – 3285
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3277

improvement in activity.[9] In practice, the improvement may
not be that large, as the catalyst is never base-free (even the
toluene solvent can coordinate). For example, Guironnet
et al.[7] have demonstrated that, under similar conditions, the
activity of Drent catalysts bound to DMSO and to pyridine
are 1150 and 620 gPE mmolPdÀ1 hÀ1, respectively. Solvent ef-
fects,[10] the possibility to coordinate the axial position of the
complex,[11]as well as the aggregation state (in the case of di-
sulfonate catalysts[12]) are also factors that influence activity
and molecular weights. To our knowledge, no in-depth study
exists on the effect of the phosphane substituent (R group,
Scheme 1) on activity.
We here present such a structure–activity relationship that
has led us to the discovery of two very active polymerization
catalysts (2g-lut and 2h-lut); one of them (2g-lut) being so
active that it catalyzes the polymerization of ethylene at
room temperature to form a linear polymer, whereas the
other one (2h-lut) is able to form copolymers of methyl ac-
rylate with ethylene while combining high activity, higher
molecular weight and high polar monomer incorporation.
By elaborating this structure–activity relationship, our goal
did not merely consist in presenting a series of novel ethyl-
ene polymerization catalysts, but it mostly aimed at estab-
lishing a guide for the rational design of highly active Pd
catalysts, which are crucially needed for the preparation of
functional polyolefins by catalytic copolymerization of ethyl-
ene with polar monomers on a large scale.
Figure 1. ORTEP view of 2d-lut. Hydrogen atoms are omitted. Selected
bond lengths [ꢃ] and angles [8]: Pd1–C1 2.100, Pd1–P1 2.222, Pd1–N1
2.128, Pd1–O1 2.161, S1–O1 1.436, S1–O2 1.449, S1–O3 1.481; C1-Pd1-N1
90.39, O1-Pd1-P1 93.61, C1-Pd1-P1 89.93, N1-Pd1-O1 86.06.
quite similar to those of other [(PO)PdMe(L)] complexes
with the palladium atom in a square-planar environment
and the Me group is trans with respect to the sulfonate
group. In this study, the influence of the phosphane substitu-
ent on catalytic activity is scrutinized. This influence should
À
be easily described by comparing the Pd P bond lengths for
each catalyst (Table 1). Their ranking is unexpected as the
Results and Discussion
Preparation of the ligands and the catalysts: For the sake of
clarity, we have adopted the notation Nx-L, in which N is a
number indicating a ligand (N=1) or a well-defined Pd
complex (N=2), x is a letter designing the phosphane sub-
stitution pattern (Scheme 1), and L indicates which ligand is
bound to the catalyst (e.g., pyr for pyridine or lut for luti-
dine). Palladium complexes 2a,[4i] 2b,[13] 2c,[4f] 2f,[4i] 2i[13], and
ligand 1h[14] were already described in the literature, but li-
gands 1d, 1e and 1g, and catalysts 2b-lut, 2d-lut, 2g-lut, 2h-
lut were prepared for the first time for this study. We pur-
posely decided to concentrate our attention to lutidine com-
plexes (or pyridine complexes for catalysts already report-
ed), because we found that DMSO complexes (e.g., 2h-
DMSO) were difficult to isolate in a pure form due to the
difficulty of evaporating DMSO.[7] Ligand 1e was also pre-
pared here for the first time, but in line with its electron-
À
À
Table 1. Comparison of experimental Pd P and Pd N lengths [ꢃ] of pal-
ladium phosphanesulfonate complexes.
À
À
Catalyst
Pd P [ꢃ]
Pd N [ꢃ]
X-ray
source
2a-pyr
2a-lut
2b-pyr
2c-lut
2d-lut
2 f-pyr
2.232
2.234
2.229
2.234
2.222
2.231
2.103
2.134
2.110
2.138
2.128
2.131
Ref. [15]
Ref. [4a]
Ref. [13]
Ref. [4f]
this paper
Ref. [4i]
most s-donating phosphanes (1c and 1 f) do not generate
the complexes with the shortest Pd P bonds, possibly indi-
cating that steric factors are also at stake. However, the dif-
ference in bond lengths between each complex is small. The
À
À
withdrawing character, it failed to react with [PdMeCl
G
Pd P bond is constrained by the presence of the Pd-O-S-C-
C-P chelate and a change in its length probably generates
important ring strain in the chelate.
or with [PdMe2A(tmeda)] to afford the corresponding Pd
N
complex. All ligands and catalysts were extensively charac-
terized by multidimensional NMR spectroscopy (1H, 13C,
31P), mass spectrometry and elemental analysis. In
Scheme 1, ligands are shown under their expected zwitter-
ionic form but, as no JACTHNUGRTENUNG(P,H) coupling was observed, a neu-
tral form could be present.
Experimental mechanistic and kinetic studies: Displacement
of lutidine by pyridine: When 2h-lut (c=1.3ꢂ10À2 molLÀ1 in
CDCl3) is contacted with pyridine (0.45 equiv), lutidine is
displaced quantitatively by pyridine, as expected for steric
À
Suitable crystals for XRD structure determination were
obtained for complex 2d (Figure 1). The overall structure is
reasons (shorter Pd N bonds for related pyridine complexes
(except for complex 2 f-pyr; see Table 1). This ligand substi-
3278
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 3277 – 3285

Toward Highly Active Palladium-Based Polymerization Catalysts
FULL PAPER
tution reaction could be monitored at 48C over a period of
1
2 h either by following by H NMR the disappearance of Me
signals in bound lutidine (two singlets at d=3.12 and
3.10 ppm) and the appearance of free lutidine (one singlet
at d=2.51 ppm), or by 31P NMR (peaks at d=41.6 and
45.8 ppm for 2h-lut and 2h-pyr, respectively). These data
were fitted by a second-order kinetics (k=0.139 LmolÀ1 sÀ1,
see the Supporting Information) as we[9] and Conley et al.[16]
have shown that ligand exchange occurs through an associa-
tive bimolecular pathway for palladium phosphanesulfo-
nates. This relatively low value of k contrasts with the value
of 2.3ꢂ105 LmolÀ1 s reported by us for pyridine exchange at
48C for catalyst 2a-pyr.[9] For 2a-pyr, free pyridine is in
rapid exchange with bound pyridine at room temperature
1
(coalescence at À508C), as observed by H NMR spectros-
copy, whereas for catalyst 2h-pyr, free pyridine and bound
pyridine are in the slow exchange regime at 508C. For cata-
lyst 2 f-pyr, coalescence is not yet reached at room tempera-
ture in CDCl3, and coalescence between free and bound pyr-
idine occurs at 608C in C2D2Cl4 (the Supporting Informa-
tion). Notwithstanding the exact values of pyridine exchange
rate for each catalyst, these data indicate that these rates
drastically change with the nature of the phosphane (strong
kinetic trans influence). Importantly, steric factors should
also be considered to explain the dissociate rate of lutidine
or pyridine.
Scheme 2. Survey of the reactivity of 2h-lut.
When no fluorinated borane is used, catalysts 2a-pyr (C=
4.9ꢂ10À2 molLÀ1 in CD2Cl2), 2 f-pyr (C=3.7ꢂ10À2 molLÀ1
in CDCl3) and 2h-lut (C=3.1ꢂ10À2 molLÀ1 in CDCl3) react
very slowly with ethylene (P=50 psi) at room temperature
(growing chain observable after several days). By contrast,
2g-lut (C=2.3ꢂ10À2 molLÀ1 in CDCl3) does rapidly insert
ethylene at room temperature (complete insertion of ethyl-
ene within several minutes) and even at 48C. In all cases,
the growing chain could be observed and characterized by
1H NMR spectroscopy (see the Supporting Information): all
attributions are in clear agreement with those published by
Noda et al.[17]
Removal of lutidine by BACHTUNRGTNE(UGN C6F5)3: Since it is well established
that L ligands such as pyridine, and to a lesser extent luti-
dine,[7] inhibit the reactivity of the catalyst by blocking its
cis-coordination site, we endeavored to remove lutidine with
the Lewis acid B
or CDCl3 formation of a dimeric Pd species (2h-dim) and
(C6F5)3·lutidine occurs over a period of a week ([B-
ACHTUNGTRENNUNG
(C6F5)3.[8a] At room temperature in CD2Cl2
BACHTUNGTRENNUNG
A
in CD2Cl2). However, the reaction occurs as soon as ethyl-
ene or methyl acrylate (MA) is introduced (Scheme 2). In
this case, a mixture of 2h-lut, 2h-dim and 3 or 2h-eth de-
pending whether MA or ethylene is used is formed (for the
sake of simplicity, Scheme 2 does not show the reaction of
Methyl acrylate insertion: When 2h-lut is contacted with
methyl acrylate (MA) at 353 K, resonances for 3, 3-mult and
4 appear to the expense of the 2h-lut Pd–Me resonance (d=
2h-lut with B
ACHTUNGTRENNUNG
ion to 2h-lut+B
G
À0.05 ppm, J
guishable by an upfield triplet assigned to Pd CH-
(H,H)=7.1 Hz), where-
ACHTUNGTRENUN(NG P,H)=3.09 Hz). Compound 3 was easily distin-
À
placement from Pd occurs much faster when ethylene or
MA is present, in conformity with the fact that such reaction
proceeds through an associative pathway. A Lewis base such
as ethylene or MA is needed to displace lutidine, which is
(CO2Me)CH2CH3 (d=À0.09 ppm, J
ACHTUNGTRENNUNG
as compound 4 exhibited a characteristic upfield doublet of
À
doublet assigned to Pd CH
(CO2Me)CH3 (d=0.10 ppm, J-
then trapped by B
E
(C6F5)3
(H,H)=7.0 Hz, J
R
and ethylene, an important broadening of ethylene and luti-
dine resonances in H NMR spectra suggests that dynamic
of the non-aromatic region for compounds 3, 3-mult and 4
1
1
was based on careful examination of H NMR spectrum (31P
exchange processes take place for ethylene and for lutidine
(between BACHTUNGTRENNUNG(C6F5)3·lutidine and bound lutidine), indicating
coupled and decoupled) and COSY spectra. Irradiation of
the 31P peak at d=39.7 ppm (compound 4) results in the
simplification of the doublet of doublet corresponding to
À
that reactions marked as A and B in Scheme 2 are equili-
brated. Importantly, at no point during this study have we
been able to unequivocally characterize a complex with MA
or ethylene coordinated to Pd, even at temperatures as low
as À408C. Only lutidine containing species or dimeric spe-
cies were observed.
Pd CHACTHNUTGRNEUGN(CO2Me)CH3 into a doublet, indicating that this
CH3 group is coupled to a single CH group and to the phos-
phane (Figure 2). By contrast, the terminal methyl group in
compound 3 is not coupled to phosphorous, but coupled to
two enantiotopic CH2 protons (d=0.4 ppm and 1.3 ppm) in
Chem. Eur. J. 2012, 18, 3277 – 3285
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3279

J. P. Claverie et al.
Figure 3. Concentration of 2h-lut, 3, 3-mult and 4 against time.
À
tion, and therefore after 3. Since no Pd H could be detected
1
(either by 31P or H NMR spectroscopy), we have inferred
À
that the insertion of the acrylate in the Pd H bond is
rapid.[18] To determine the kinetic constants k1, kmult, and
kbH, we have assumed identical rate constants (kmult) for in-
sertion of the acrylate into 3, into 4 or even into 3-mult.
This assumption is reasonable because structural differences
between these complexes are two carbons away from the
À
carbon atom involved into the C C making step. With this
assumption in hand, k1 and kmult were found to be 0.12 and
0.48 LmolÀ1 minÀ1, respectively. Importantly, these kinetic
constants are apparent kinetic constants that contain terms
relative to the pre-equilibrium between the lutidine complex
and the acrylate. The fact that kmult is four times greater
than k1 can be explained by the more facile lutidine dissoci-
ation from 3 relative to 2h-lut, due to the greater bulk of
the substituent cis to lutidine. Interestingly, kbH was found to
be 0.088 minÀ1: at 353 K, the half lifetime of a chain is t=
8 min. During this period, complex 3 or 4 have inserted an
average kmult·[MA]·tꢀ0.25 monomers. Thus, in Scheme 3,
n=1.25, which is consistent with the fact that trans-
(MeO2C)CH=CHCH3 (corresponding to n=0) is detected
in the kinetic experiment performed with low MA concen-
ACHTUNGTRENNUNG
Figure 2. 1H and 1H{31P} NMR spectra of 2h-lut, 3 and 4 (in [D2]TCE).
a of Pd-CHACHTUNGTRENNUNG(CO2Me). These as-
signments are consistent with
those found by Johnson et al.[2b]
for palladium diimine catalysts,
by Noda et al.[17] for ethylene
inserted into aryl phosphanesul-
fonate catalysts, and by Guiron-
net et al. for MA inserted into
aryl phosphanesulfonate cata-
lysts.[7,8b] These assignations are
also supported by the evolution
over time of the concentration
of each species (Figure 3). In
the first 15 min, [2h-lut] de-
creases and [3] increases rapidly
to then reach
a maximum,
whereas [4] only increases later,
which is consistent with reac-
tion Scheme 3, whereby 4 is
generated after b-H elimina- Scheme 3. Reaction of 2h-lut with MA.
3280
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 3277 – 3285

Toward Highly Active Palladium-Based Polymerization Catalysts
FULL PAPER
tration. Longer chain lengths should be obtained at higher
MA concentration (see below).
CF3 fragment on the diimine leads to a significant activity
increase, thus hinting that electron-deficient ligands result in
more active catalysts.[20] In a previous study,[13] we demon-
strated that catalysts based on phenantryl, naphtyl, and an-
thracenyl moieties are less active than catalyst 2a.The differ-
ence in phosphane basicities between phenyl, naphtyl, an-
thracenyl, and phenantryl-based ligands is expected to be
small in comparison to the large range of basicities assessed
in this study. In our case, it is necessary to clarify whether
phosphane s-bonding ability is
Polymerization of ethylene: All catalysts were tested under
similar conditions for the polymerization of ethylene; the re-
sults of which are reported in Table 2. The pentafluorophen-
yl-substituted ligand 1e was tested using a catalytic system
generated in situ in the presence of PdACTHNUTRGENUN(G dba)2. The polymeri-
zation activity is very low (3 gPE mmolPdÀ1 hÀ1) and the reac-
truly at stake. The stretching
frequency of trans CO in
Table 2. Polymerization of ethylene (T=858C, P=300 psi, in 200 mL toluene).
[b]
Entry Catalyst
N
t
Yield Act[a]
Actmax
Tm
[8C]
Mn
Mw/Mn
LNi(CO)3 is often used as a
measurement of L-donating
ability, however, stretch fre-
quencies for phosphanes report-
ed here have not been reported.
Thus we have correlated our ac-
tivity data with phosphane de-
scriptors listed in the ligand
knowledge base for monoden-
tate P donors (LKB-P).[21] We
[min]
A
]
[gPEmmolPdÀ1 hÀ1
]
[gPEmmolPdÀ1 hÀ1
]
G
[ꢂ103 gmolÀ1
]
1
2
3
4
2a-pyr
2b-pyr
2c-lut
2d-lut
1e
2 f-lut
2g-lut
2g-lut
2h-lut
2i-pyr
10.0
10.8
9.1
6.1
23
10.0
6.4
6.4
6.4
15
60
60
90
60
60
60
30
15
60
60
210
1027
1611
170
3
1040
275
1430
4900
191
129.0 18.2
128.2 9.3
131.3 12.4
2.8
1.7
2.0
1.5
2.2
3.9
3.9
2.6
2.3
1.4
11.1
22.0
1.03
0.06
10.4
128.9
4.9
0.53
5[c]
6
71
nd[d]
1820
4600
25000
18000
<100
133.5 227
132.4
135.8 44.7
129.8 14.6
128.7
7
3.60 1125
6.6
8[e]
9
0.3
30.0
0.71
188
4714
48
10
5
[a] Average activity. [b] Maximum instantaneous activity recorded after ethylene dissolution and temperature have compared each of our bi-
stabilization. [c] Catalytic system generated in situ with [PdACHTUNGRTENUNG(dba)2]. [d] Not determined due to the low amount
of sample. [e] T=308C.
dentate ligands with the corre-
sponding trialkyl or triarylphos-
phane: for example, for ligand
tion only yielded trace amounts of a very low molecular
weight product (Mn =530 gmolÀ1). With the exception of
catalyst 2g-lut, the optimal temperature range to maximize
both PE yield and activity is in the interval 75–958C. By
contrast, the di-tert butyl-substituted catalyst 2g-lut is active
over a wider range of temperatures, being able to rapidly
polymerize ethylene at 308C (Table 2, entry 8). At this tem-
perature, the resulting polyethylene is highly linear, as
shown by a high Tm of 135.88C with a high molecular weight
of Mn =44,700 gmolÀ1 (being only surpassed by 2 f-pyr). Al-
though peak activity of catalyst 2g-lut at 308C is higher than
all other ones at 858C, the catalyst decomposes rapidly,
whereas other catalysts have lifetimes exceeding one hour
(and greater than a day in certain cases; for example, see
catalyst 2 f-pyr). Peak activity of catalyst 2g-lut at 858C is
likely underestimated due to significant catalyst decomposi-
tion before temperature stabilization. Our experimental
setup does not permit to accurately record activities below
room temperature (due to the lack of temperature stabiliza-
tion), but it was found that 2g-lut is capable of polymerizing
ethylene at 48C in a NMR experiment. Comparing the re-
sults of Table 2, it appears that electron-poor phosphanes
(1e, 1d, and 1i) yield the least active catalysts, whereas elec-
tron-rich phosphanes (1c, 1 f, and 1g) generate more active
catalysts. Few complete structure–activity relationships for
metal late-transition polymerization catalysts have been re-
ported in the literature. In the case of cationic Pd-diimines,
Popeneyet al. have demonstrated that electron-rich ligands
yield somewhat more active catalysts (one order of magni-
tude between smaller and larger TOF).[19] However, Gates
et al. have also shown that replacing a CH3 fragment by a
1e, descriptor values of PACTHNGUTERN(NUG C6F5)3 were used. The LKB-P lists
348 P-ligands for which 26 descriptors are reported. Some of
these descriptors (e.g., He8 steric) capture the influence of
sterics,[22] whereas others (BE(B), Q
ACTHNUTRGNEUNG(B fragm.), or DA-P-A)
correlate with the s-bonding-ability of the ligand. Most de-
scriptors (structural and thermodynamic parameters of Pd,
Pt, Au, and borane complexes) are influenced by several pa-
rameters at the same time (e.g., sterics, s- and p bonding).
Figure 4 illustrates the relationship of the logarithm of the
polymerization activity to the descriptor having the lowest
(A) and the highest (B) correlation factor r2. There is no
correlation (r2 ꢀ0) between activity and the descriptor
He8 steric. The He8 steric descriptor is the interaction
energy between the ligand in its groundstate conformation
and a ring cluster of 8 He atoms. It has been shown to corre-
late well with the Tolman cone angle,[21] and it is an accepta-
ble descriptor to assess the influence of steric factors. The
BE(B) descriptor describes the bond energy between BH3
and the ligand (in kcalmolÀ1). Since BH3 is a small molecule
(not prone to be influenced by sterics) and since its d orbi-
tals are vacant and high-lying (no p bonding), this descriptor
provides an accurate measurement of the s-donating charac-
ter of the phosphane. This descriptor is also correlated to
the phosphane bulk, which influences the phosphorus hy-
bridization and therefore the donating ability of the phos-
phane. Interestingly, our activity data do not correlate favor-
ably with descriptors involving Pd complexes (0.02<r2 <
0.54). Such descriptors include bond dissociation energy of
À
À
the phosphane ligand from LPdCl3-, P Pd and trans Cl Pd
bond lengths in this same complex. Bonding in LPdCl3- is of
course influenced by s donation and p acidity of the phos-
Chem. Eur. J. 2012, 18, 3277 – 3285
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3281

J. P. Claverie et al.
points in comparison to the number of descriptors. The mo-
lecular weight data also did not correlate with Tolman cone
angles:[23] for example, P
ACTHUNRGTNENUG(tBu)3 and PAHCTUNGTREN(NUNG C6F5)3 have sensibly
the same cone angle (182 and 1848) but polymers produced
with 1e and 1g have very different molecular weights. In
the past, we also have demonstrated using polyACHTNUTRGNE(UNG arylated)
phosphanes (Ar=naphtyl, phenantryl and anthracenyl) that
molecular weight does not increase when the bulk of the
ligand increases.[13] We believe that the failure to correlate
our molecular weight data with simple descriptors is likely
due to the lack of descriptor to assess steric bulk specifically
located in the axial faces of the complex (by opposition to
steric bulk of the whole ligand that is captured by the cone
angle). In this sense, one can anticipate that complex 2 f,
which has one large 2,6-(OMe)2C6H3 substituent extending
directly above one of the Pd axial position (see the X-ray
data of reference. [4i]) should yield very high molecular
weight (as observed in Table 2). Accordingly, the Nickel
complex reported by Perrotin et al., which is also bearing
the substituent 2,6-(OMe)2C6H3 on one axial position, yields
linear polyethylene with molecular weights up to
400000 gmolÀ1.[5b]
Copolymerization of ethylene with methyl acrylate: The
most active catalysts for ethylene polymerization were also
found to be capable of inserting MA (Table 3). The influ-
ence of MA concentration, ethylene pressure and tempera-
ture on activity and MA incorporation has already been dis-
cussed previously.[4i,7,9] These exhaustive data were obtained
with ligands 1a, 1b, and 1 f, and they have not all been in-
cluded in Table 3. In brief, to reach a high MA incorpora-
tion, it is necessary to use the lowest ethylene pressure and
the highest MA concentration possible during the reaction.
Figure 4. LnACHTUNGTRENNUNG ) versus He8 steric (A) and
(activity) (in gPE mmolPdÀ1 hÀ1
BE(B) descriptors. Values of the descriptors of each ligand can be found
in references [21] and [22]. Lutidine complexes: c, d, f, g, and h, pyridine
complexes: a, b, i; no base: e.
phane. To conclude this section, there is a one-to-one corre-
spondence between the activity of Pd phosphanesulfonates
and the s-donating ability of the phosphane moiety. This is
also supported by common-sense scrutiny of Table 2: the ac-
tivities of alkylated 1h and
1c phosphanes are signifi-
Table 3. Copolymerization of ethylene with methyl acrylate.
cantly greater than those of
aryl phosphanes and even
more than those of fluorinat-
ed aryl phosphanes. How-
ever, it should also be men-
tioned that only seven exper-
imental data were exploited
for the correlation (data for
1 f and for 1h being current-
ly unavailable in the LKB-
P), and a significantly larger
set of experiments will be
necessary to generate a more
significant correlation. Com-
paring our molecular weight
data with LKB-P did not
yield readily interpretable
[b]
Entry Catalyst
n
cMA
[molLÀ1
PE
A
T
t
A
Yield
Act max
Mn
Mw/ MA
Mn
[b]
A
E
]
[min] [g polymer] [gPEmmolPdÀ1 hÀ1
]
[ꢂ103
gmolÀ1
ACHTUNGTRENNUNG
[mol[c] %]
]
1
2
3
4
5
6
7
8
2a-pyr
2a-pyr
2b-lut
2b-lut
2d-lut
2 f-pyr
2h-lut
2h-lut
2h-lut
2g-lut
2g-lut
27.5
27.5
21.0
14.1
24.1
4.0
24.3
36.5
91.0
10.9
20.0
0.11
0.46
0.60
0.11
0.60
1.5
1.10
2.75
4.80
1.10
1.1
100 100 120
60 100 120
100 100 120
110 100
100
100
100 100
100
0
100
100
0.65
0.10
0.18
1.1
0.01
1.4
0.16
0.52
0.12
0.09
0.2
27
27
11
120
<1
62
43
50
–
8.3
65
8.9
4.9
3.5
3.2
2.7
2.6
9.6
9.0
2.2
2.2
2.0
2.1
60
90
[a]
[a]
[a]
85
85 120
90
95 150
95 240
85
65
–
–
–
41.2
12.3
11.5
0.6[d]
6.3
2.5
3.0
2.1 23.0
2.0 28.0
9
10
11
–
2.2
2.7
100
2.1
3.3
60
15
2.4
[a] Not determined due to the low amount of sample. [b] Determined by GPC at 1608C in TCB. [c] Determined
by H NMR at 1408C in TCE-d2. [d] Determined by NMR spectroscopy, by using the attribution published in ref-
1
erence [7].
results, one of the possible reasons being that the phosphane
yielding the highest molecular weight (1 f) is not available in
the data set. A principal component analysis (to correlate
molecular weights with a subset of descriptors) was not
meaningful due to the limited number of experimental
In general, the activity of catalysts in copolymerization
parallels the activity in homopolymerization: 2h-lut proving
to be able to insert large amounts of MA with higher activi-
ties than other catalysts. In fact, this catalyst is so active that
it is able to homooligomerize MA. The MA oligomer after
3282
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 3277 – 3285

Toward Highly Active Palladium-Based Polymerization Catalysts
FULL PAPER
ymers were determined by gel permeation chromatography (GPC) using
a Viscotek HT GPC equipped with triple detection operating at 1408C.
For lower molecular weights, the light scattering detector was not used.
The eluent was 1,2,4-trichlorobenzene, and separation was performed on
purification (to remove BHT, the radical inhibitor added to
prevent radical homopolymerization of MA) and characteri-
zation by H NMR spectroscopy (using the assignations re-
1
ported by Guironnet et al.[7]) was found to have an average
degree of polymerization of 7 MA. Interestingly, in compari-
son to catalyst 1a-DMSO, which is also able to incorporate
high amounts of MA into polyethylene, catalyst 2h-lut
yields polymers of higher molecular weight (compare en-
tries 1 and 2 to entries 6, 7, and 8).
three PolymerLabs Mixed BACTHNUTRGNE(NUG -LS) columns. The dn/dc of pure linear poly-
ethylene was found to be 0.106 mLgÀ1 at this temperature. The dn/dc of
several copolymers was also determined. A linear extrapolation between
the dn/dc of the copolymer and the weight% composition in acrylate was
then performed, leading to the determination of an extrapolated dn/dc of
a poly(methyl acrylate) (0.2179 mLgÀ1). The dn/dc of any copolymer was
then calculated as the weighted-average of the dn/dc of pure PE and dn/
dc of pure poly(methyl acrylate). Differential scanning calorimetry meas-
Although 2g-lut was able to copolymerize ethylene and
MA at 658C, only low MA incorporations were reached. In-
urements (DSC) of solid samples were performed on
a DSC823e
(TOPEM modulation) equipped with an FRS5 sample cell, a sample
robot, a Julabo FT400 intracooler and an HRS7 sensor from Mettler
Toledo. Samples were heated from 20 to 1608C at a rate of 58CminÀ1
and data were analyzed with STAR software. All reported values are for
samples which have been slowly cooled from the melt at a rate of
58CminÀ1. Electrospray mass spectra (ESI-MS) of organic compounds in
acetonitrile were recorded on an Agilent 6210 LC-MSD TOF mass spec-
trometer. Elemental analysis was performed by the elemental analysis
laboratory, Department of Chemistry, University of Montreal.
À
sertion of acrylate in the Pd Me bond of 2g-lut (c=2.3ꢂ
10À2 molLÀ1) does not occur in CDCl3 at room temperature.
It is conceivable that for 2g-lut, insertion of MA is more dif-
ficult than insertion of ethylene because of the steric conges-
tion imparted by the two tert-butyl groups. This was also
previously reported by us for catalyst 2 f-pyr.[4i]
Polymerizations: Polymerizations of ethylene and copolymerizations of
ethylene with methyl acrylate were carried out in a stainless steel reactor
(450 mL, Parr). Catalyst, toluene, and comonomer were added to a
Schlenk flask in a nitrogen-filled glovebox. The reactor, which was first
dried and kept under nitrogen, was loaded with the toluene solution by
cannula transfer from the Schlenk flask under nitrogen. The reactor was
then sealed, pressurized with ethylene, stirred, and heated. The polymeri-
zations were performed at constant pressure in the feed reactor and in-
stantaneous activities were calculated from the rate of ethylene consump-
tion, which was monitored by the decrease of the ethylene pressure in
the feed tank. The first five minutes of the data were first discarded
(pressure drop due to ethylene dissolution). Then, only data for which
the temperature was constant (+/À38C) were kept for interpretation
(catalyst loading for each polymerization was adapted to avoid large exo-
therms), and an activity-versus-time curve was traced. In Tables 2 and 3,
Activity max corresponds to the maximum of this curve. It is usually lo-
cated approx 10 min after polymerization start. Average activity is calcu-
lated from the weight of collected polymer and reaction time. Once the
reaction was over, the reactor was cooled down to room temperature and
slowly depressurized. The polymers were precipitated in four volumes of
methanol, collected by centrifugation or filtered, washed with methanol,
and dried under vacuum.
Conclusion
Herein we have illustrated the influence of s-donating char-
acter of the phosphane ligand on the catalytic polymeri-
zation activity of phosphane Pd sulfonates. Thus, the best
donor phosphane ligand, bearing two tert-butyl substituents
exhibits an unprecedented activity at room temperature
whereas all other catalysts function at 80–1008C. Further-
more, highly active catalyst 2g-lut, bearing one phenyl and
one tert-butyl substituent is sufficiently active to incorporate
MA in large amounts, and even to homopolymerize MA.
The discovery of these new catalysts should facilitate the
preparation and the study of functional materials based on
polyolefins, which so far was only prepared in low yield due
to the low activity of the Pd catalysts.
Homopolymerization of methyl acrylate with catalyst 2h-lut (entry 9 in
Table 3) was performed by dissolving catalyst (50.2 mg; 91 mmol) and
BHT (radical inhibitor; 125 mg) in toluene (6 mL) and MA (4 mL) in a
Schlenck tube under nitrogen. This solution was heated for 4 h at 958C.
The solution was then opened to air, cooled, and filtered over diatoma-
ceous silica (to remove BHT). The solution was then roto-evaporated
and the solid was dried in vacuo. Yield: 125 mg.
Experimental Section
General considerations: All manipulations were performed under an
inert atmosphere by using standard Schlenk techniques. Solvents were
degassed with N2 and dried over activated molecular sieves using a
MBraun solvent purification system. Benzene sulfonic acid was pur-
chased from Sigma Aldrich and dried by azeotropic distillation with ben-
Catalyst 2b-lut: [PdMe2ACTHNUTRGNE(NUG TMEDA)] (0.208 g, 0.50 mmol) and ligand 1b
zene. BisACHTUNGTRENNUNG(3,5-di(trifluoromethyl)phenyl)chlorophosphane, di-tert-butyl-
(0.172 g, 0.50 mmol) were dissolved in dry THF (10 mL) under inert at-
mosphere and the resulting solution was stirred for 60 min. Lutidine
(0.10 g, 0.93 mmol) was then added followed by stirring for another
60 min. After adding Et2O (15 mL), the white precipitate was collected,
washed twice with cold Et2O (15 mL) and dried under vacuum. Yield=
0.215 g (75%). 1H NMR (CDCl3): d=8.24 (dd, J=7.4, 4.3 Hz, 1H, C-
chlorophosphane, chloro(tert-butyl)phenylphosphane and n-butyllithium
in hexane were purchased from Sigma Aldrich and used as received. Di-
methyl(N,N,N’,N’-tetramethylethylenediamine)-palladium(II),
[PdMe2-
ACHTUNGTRENNUNG
mers were purified by sparging them with argon and passing them over a
bed of inhibitor-remover resin (Aldrich) as acrylic monomers are usually
inhibited with quinones that interfere with the catalyst. The monomers
were then spiked with tert-butyl catechol (0.25% wt/wt) to prevent spon-
taneous radical polymerization of the acrylate during the polymerization
process.
Instrumentation: 1H, 13C, 19F, and 31P NMR spectra were recorded on a
Varian Inova 600 MHz spectrometer or a BrukerUltrashield 300 MHz at
ambient temperature except for the polymers, which were analyzed in
[D4]1,1,2,2-tetrachloroethane at 1158C. 1H and 13C NMR chemical shifts
were referenced to the solvent signal. All spectra were recorded in 1H de-
coupled 31P NMR mode. The molecular weight distributions of homopol-
AHCTUNGTRENNUNG
H
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
3.16 (s, 6H, CH3lutidine), 0.23 ppm (d, J=3.3 Hz, 3H, Pd-CH3).
31P NMR (CDCl3): d=25.6 ppm (s).
Ligand 1d: Bis-3,5-di(trifluoromethyl)phenyl)phosphanobenzenesulfonic
acid: To a solution of dry benzenesulfonic acid (0.32 g, 2.03 mmol) in
THF (15 mL) was added nBuLi 2.5m in hexanes (1.70 mL, 4.25 mmol) at
08C. After stirring for 1 h at room temperature, the solution was added
dropwise to a solution of bisACTHNUGRTNEUNG(3,5-di(trifluoromethyl)phenyl)chlorophos-
Chem. Eur. J. 2012, 18, 3277 – 3285
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3283

J. P. Claverie et al.
phane (1.00 g, 2.03 mmol) in THF (15 mL) at 08C. After stirring for 2 h
at room temperature, the solvent was removed in vacuo leaving a purple
solid. The solid was dissolved in dichloromethane (30 mL) and extracted
with acidic water (0.5 mL of concentrated HCl in 20 mL of water) and
then twice with water (15 mL). The organic phase was dried over MgSO4,
the solution was filtered and the organic solvent was removed in vacuo.
The resulting pale yellow crystals were dried in vacuo. Yield=0.85 g
Ligand 1g: di-tert-butylphosphanobenzenesulfonic acid: To a solution of
dry benzenesulfonic acid (0.87 g, 5.5 mmol) in THF (15 mL) was added
nBuLi (2.5m) in hexanes (4.5 mL, 11.25 mmol) at 08C. After stirring for
1 h at room temperature, the solution was added dropwise to a solution
of di-tert-butylchlorophosphane (1.0 g, 5.5 mmol) in THF (15 mL) at 08C.
The reaction mixture was heated to 678C for 48 h and then cooled to
room temperature. The solvent was removed in vacuo leaving a pale
brown solid. The solid was dissolved in dichloromethane (30 mL) and ex-
tracted with acidic water (0.5 mL of conc. HCl in 20 mL of water) and
then twice with water (15 mL). The organic phase was dried over MgSO4,
the solution was filtered and the organic solvent was removed in vacuo.
The desired product was recrystallised from dichloromethane and ether
to afford 0.98 g (60%) of a white solid. 1H NMR (CDCl3): d=8.36 (dd,
J=7.6, 4.0 Hz, 1H), 7.68 (t, J=8.7 Hz, 1H), 7.57 (t, J=7.1 Hz, 1H), 7.45
(s, 1H), 1.49 ppm (d, J=17.0 Hz, 18H); 13C NMR (CDCl3): d=152.7 (d,
(68%). 1H NMR (CDCl3): d=7.82 (s, 1H, C(P)-C
2H, paraditrifluoromethyl phenyl), 7.54 (d, J=5.3 Hz, 4H,
orthoditrifluoromethyl phenyl), 7.29 (t, J=11.7 Hz, 1H, C(P)-C(SO3)=
ACHTUNGTREN(NNUG SO3)=CH-), 7.74 (s,
H
H
ACHTUNGTRENNUNG
CH-CH=), 6.98 (m, 1H, C(P)-CH-), 6.90 ppm (dd, J=6.70, 3.54 Hz, 1H,
C(P)-CH=CH-). 13C NMR ([D6]DMSO): d=153.2 (s), 153.1 (s), 142.7 (s),
134.2 (s), 133.1 (d, J=21.3 Hz), 130.6 (s), 130.4 (d, J=5.2 Hz), 130.2 (d,
J=5.4 Hz), 129.9 (s), 129.6 (s), 127.3 (d, J=4.8 Hz), 124.1 (s), 122.4 (s),
122.4 (s), 122.3 ppm (s); 31P NMR (CDCl3): d=À8.3 ppm (s); 19F NMR
(CDCl3): d=7.0 ppm (s). HRMS (ESI): m/z calcd for C22H11F12O3PS:
613.9975 [M+]; found: 613.9974.
J
A
J
A
JACHTUNGTRENNUNG
A
R
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
Catalyst 2d-lut: [PdMe2ACHTUNGTRENNUNG(TMEDA)] (0.113 g, 0.44 mmol) and ligand 1d
35.0 ppm (s); HRMS (ESI): m/z calcd for C14H23O3PS: 302.1106 [M+];
(0.270 g, 0.44 mmol) were dissolved in dry THF (10 mL) under an inert
atmosphere and the resulting solution was stirred for 60 min at room
temperature. Lutidine (0.054 g, 0.50 mmol) was then added followed by
stirring for another 60 min. After adding Et2O (15 mL), the white precipi-
tate was collected, washed with Et2O and dried under vacuum. Yield=
found: 302.1101.
Catalyst 2g-lut: [PdMe2ACTHNUTRGNE(NUG TMEDA)] (0.152 g, 0.60 mmol) and ligand 1g
(0.181 g, 0.60 mmol)[14] were dissolved in dry THF (10 mL) under an inert
atmosphere and the resulting solution was stirred for 60 min at room
temperature. Lutidine (0.075 g, 0.70 mmol) was then added followed by
stirring for another 60 min. After adding Et2O (15 mL), the white precipi-
1
0.25 g (68%). H NMR (CDCl3): d=8.28 (ddd, J=7.7, 4.9, 1.0 Hz, 1H, C-
A
ACHTUNGTREN(NUGN CF3)-), 8.07 (s, 2H, C-
tate was collected, washed with Et2O and dried under vacuum. Yield=
G
E
ACHTUNGTRENNUNG
1
0.20 g (63%). H NMR (CDCl3): d=8.15 (dd, J=7.8, 3.8 Hz 1H, C
N
(t, J=7.7 Hz, 1H, Hparalutidine), 7.20 (d, J=7.7 Hz, 2H, Hmetalutidine),
7.02 (m, 1H, CP-CH=CH), 3.13 (s, 6H, CH3 of lutidine), 0.17 ppm (d, J-
CH=), 7.97 (t, J=7.4 Hz, 1H), 7.64 (t, J=7.7 Hz, 1H), 7.51 (t, J=
7.6 Hz, 1H), 7.46 (t, J=7.6 Hz, 1H, Hparalutidine), 7.19 (d, J=7.7 Hz, 2H,
Hmetalutidine), 3.19 (s, 6H, CH3 of lutidine), 1.52 (d, J=14.4 Hz),
ACHTUNGTRENNUNG
(P,H)=3.9 Hz, 3H, Pd-CH3). 13C NMR (CDCl3): d=158.5 (s), 149.7 (d, J-
T
ACHTUNGTRENNUNG
0.59 ppm (d, J
(s), 150.7 (d, J
2.0 Hz), 128.8 (d, J
(P,C)=4.9 Hz), 123.4 (d, J
(d, J
(P,H)=1.2 Hz, 3H, Pd-CH3). 13C NMR (CD2Cl2): d=159.4
ACHTUNGTRENNUNG
E
G
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG(P,C)=
AHCTUNGTRENNUNG
A
ACHTUNGTRENNUNG
119.9 (s), 26.2 (s, CH3 of lutidine), À3.6 ppm (d, J
ACHTUNGTRENNUNG
31P NMR (CDCl3): d=27.3 ppm (s); 19F NMR (CDCl3): d=À63.5 ppm
(s); elemental analysis calcd (%) for C30H22F12NO3PPdS: N, 1.66; C,
42.80; H, 2.63; S, 3.81; found: N, 1.60; C, 42.89; H, 2.28; S, 3.63.
E
A
ACHTUNGTRENNUNG
(P,C)=5.6 Hz), 26.7 (s, CH3 of lutidine), À9.5 ppm (d, J
ACHTUNGTRENNUNG
1.5 Hz). 31P NMR (CD2Cl2): d=53.2 ppm (s);elemental analysis calcd
(%) for C22H34NO3PPdS: N, 2.64; C, 49.86; H, 6.47; S, 6.05; found: N,
2.61; C, 49.62; H, 6.72; S, 5.33.
Ligand 1e: Bis-pentafluorophenylphosphanobenzenesulfonic acid: Fluo-
rinated aryl groups are known to undergo explosive reactions in the pres-
ence of alkyl lithiums.[25]. Thus, PCl3 was reacted with a Grignard reagent
to prevent the use of an alkyl lithium to introduce the two first substitu-
ents of the phosphane. The third substituent was introduced with aortho-
lithiated salt of benzenesulfonic acid. Although we never encountered
any safety problem with this experiment, it should not be undertaken on
a large scale. A solution of C6F5Br (3.087 g, 12.5 mmol) in dry diethyl
ether (20 mL) was added dropwise to a suspension of Mg turnings
(0.33 g, 13.5 mmol) activated previously by a few grains of I2. After the
addition was completed, the dark brown solution was heated at 358C for
2 h and then allowed to cool to room temperature. Phosphorus trichlor-
ide (0.858 g, 6.25 mmol) in dry diethyl ether (30 mL) was added by a can-
nula. The solution (A) was stirred for 3 h. In another Schlenk flask,
nBuLi (2.5m) in hexanes (5 mL, 12.5 mmol) was added to a solution of
dry benzenesulfonic acid (0.988 g, 6.25 mmol) in dry diethyl ether
(50 mL) at 08C. The solution (B) was stirred for 1 h at room temperature.
Solution A was then transferred into solution B by cannula and the grey
mixture was stirred overnight. The solvent was then removed in vacuo
leaving a grayish solid. The solid was dissolved in dichloromethane
(30 mL) and extracted with acidic water (0.5 mL of concentrated HCl in
20 mL of degassed water) and then twice with water (15 mL). The organ-
ic phase was dried over MgSO4, the solution was filtered and the organic
solvent was removed in vacuo, leaving a brown pasty solid. The product
was recrystallized from methanol to afford 0.95 g (43%) of the title com-
pound as a brown solid. 1H NMR (CDCl3): d=7.96 (t, J=6.0 Hz, 1H, C-
Catalyst 2h-lut: [PdMe2ACTHNUTRGNE(NUG TMEDA)] (0.113 g, 0.44 mmol) and ligand 1h
(0.142 g, 0.44 mmol) were dissolved in dry THF (10 mL) under an inert
atmosphere and the resulting solution was stirred for 60 min at room
temperature. Lutidine (0.054 g, 0.50 mmol) was then added followed by
stirring for another 60 min. After adding Et2O (15 mL), the white precipi-
tate was collected, washed with Et2O and dried under vacuum. Yield=
0.215 g (88%). 1H NMR (CDCl3): d=8.31 (ddd, J=8.0, 4.3, 1.3 Hz, 1H,
CACHTUNRGTEN(UNG SO3)-CH=), 7.72 (m, 2H, Horthophenyl), 7.57 (t, J=7.7 Hz, 1H,
À
À
H
paralutidine), 7.50 (tt, J=8.0, 1.4 Hz, 1H, CACHTGNUTERNNU(G SO3) CP=CH CH), 7.39
(m, 3H, Hmeta+paraphenyl), 7.10 (dd, J=7.7, 3.6 Hz, 2H, Hmetalutidine),
À
7.05 (dd, J=9.0, 8.2, 1.2 Hz, 1H, CP CH=CH), 3.12 (s, 3H, CH3 of luti-
dine) and 3.11 (s, 3H, CH3 of lutidine), 1.56 (d, J=15.23 Hz, 9H, CH3 of
tBu), À0.05 (d, J
159.4 (s), 158.4 (s), 138.1 (s), 136.4 (s) 133.1 (d, J
(P,C)=2.2 Hz), 129.5 (d, (P,C)=2.3 Hz), 129.1 (d,
128.4 (d, J(P,C)=5.9 Hz), 128.2 (d, J(P,C)=9.8 Hz), 126.5 (s), 126.2 (s),
122.7 (d, J(P,C)=3.3 Hz), 122.1 (d, J(P,C)=3.0 Hz), 68.0 (s), 35.9 (s), 36.1
(s), 29.1 (d, J
(P,H)=3.09 Hz, 3H, Pd-CH3). 13C NMR (CDCl3): d=
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
J
A
J
A
JACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
(CDCl3): d=41.6 (s);elemental analysis calcd (%) for C24H30NO3PPdS:
N, 2.55; C, 52.41; H, 5.50; S, 5.83; found: N, 2.52; C, 52.57; H, 5.57; S,
5.29.
Catalyst 2h-dim: In a nitrogen-filled glovebox, a NMR tube was charged
with 2h-lut (0.025 g, 0.0455 mmol) and BACHTNUTRGNE(UNG C6F5)3 (0.024 g, 0.0470 mmol),
and CD2Cl2 (1 mL) was added. The tube was agitated at room tempera-
ture for two weeks. The 1H and 31P NMR spectrum showed that 2h-lut
was completely consumed and that 2h-dim was formed. 1H NMR
À
À
À
À
ACHTUNGTRENNUNG(SO3) CH), 7,37 (t, J=7.6 Hz, 1H, CP CH=CH), 7.23 (m, 1H, CP CH=
CH), 7.09 ppm (m, 1H, CACTHNUTRGNEUNG
(SO3) CH=CH). 13C NMR (CDCl3): d=149.2,
À
147.5, 144.8, 143.7, 143.1, 142.0, 139.2, 137.4 ppm. 31P NMR (CDCl3): d=
À45.8 ppm (s); 19F NMR (CDCl3): d=À129.1, À151.9, À161.7 ppm;
HRMS (ESI): m/z calcd for C18H5F10O3PS: 521.9537 [M+]; found:
521.9539.
(CD2Cl2): d=8.16 (m, 1H, CACTHUNGTRENNU(G SO3) CH=), 7.62 (m, 2H, Horthophenyl),
À
À
7.41 (m, 1H, CACTHUNGTRENUNG(SO3) CP=CH CH), 7.24 (m, 3H, Hmeta+paraphenyl), 6.7
À
À
(m, 1H, CP CH=CH), 1.51 (9H, CH3 of tBu), 0.35 ppm (s, 3H, Pd
CH3). 31P NMR (CD2Cl2): d=50.5 ppm (s).
3284
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 3277 – 3285

Toward Highly Active Palladium-Based Polymerization Catalysts
FULL PAPER
3, 3-mult and 4: In a nitrogen filled glovebox, a NMR tube was charged
with 2h-lut (12 mg, 22 mmol). TCE-d2 (1.1 g) was then added followed by
methyl acrylate (10 mL, 0.11 mmol). The tube was sealed and then heated
at 808C in the NMR probe. 1 H, 1H COSY and 31P spectrum were ac-
quired at regular intervals over an 8 h period.
131, 14606–14607; g) C. Bouilhac, T. Runzi, S. Mecking, Macromole-
cules 2010, 43, 3589–3590; h) S. Ito, M. Kanazawa, J. I. Kuroda, Y.
Okumura, K. Nozaki, J. Am. Chem. Soc. 2011, 133, 1232–1235;
i) K. M. Skupov, P. R. Marella, M. Simard, G. P. A. Yap, N. Allen, D.
Conner, B. L. Goodall, J. P. Claverie, Macromol. Rapid Commun.
2007, 28, 2033–2038; j) J. C. Daigle, L. Piche, J. P. Claverie, Macro-
molecules 2011, 44, 1760–1762; k) V. A. Kryuchkov, J. C. Daigle,
K. M. Skupov, J. P. Claverie, F. M. Winnik, J. Am. Chem. Soc. 2010,
132, 15573–15579.
3: 1H NMR ([D2]TCE, 600 MHz, RT, excluding Ar resonances) d=1.91
(dd, J=9.7, 3.2 Hz, Pd-CH
(CO2Me)CH2); À0.09 ppm (t, J=7.1 Hz, 3H, Pd-CH
31P NMR (CDCl3): d=30.4 ppm (s).
3-mult: 1H NMR ([D2]TCE, 600 MHz, RT, excluding Ar resonances) d=
2.05 (m, 1H, Pd-CH(CO2Me)CH2), 0.64 ppm (dd, J=6.8, 3.0 Hz, 2H, Pd-
CH(CO2Me)CH2.
4: 1H NMR ([D2]TCE, 600 MHz, RT, excluding Ar resonances) d=2.25
(dd, J=9.5, 7.1 Hz, 1H, Pd-CH(CO2Me)CH3); 0.10 ppm (dd, J=7.0 Hz,
5.2 Hz, 3H, Pd-CH
(CO2Me)CH3); 31P NMR (CDCl3): d=39.7 ppm (s).
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
[5] a) R. J. Nowack, A. K. Hearley, B. Rieger, Z. Anorg. Allg. Chem.
2005, 631, 2775–2781; b) P. Perrotin, J. S. J. McCahill, G. Wu, S. L.
Scott, Chem. Commun. 2011, 47, 6948–6950; c) X. Zhou, S. Bon-
temps, R. F. Jordan, Organometallics 2008, 27, 4821–4824; d) D.
Guironnet, T. Runzi, I. Gottker-Schnetmann, S. Mecking, Chem.
Commun. 2008, 4965–4967; e) S. Noda, T. Kochi, K. Nozaki, Orga-
nometallics 2009, 28, 656–658.
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
Formation of 2h-eth from 2 h-lut: In a nitrogen-filled glovebox, a high-
pressure NMR tube from Wilmad Glass was charged with 2h-lut (0.016 g,
0.029 mmol) and CDCl3 (1.2 mL) was added. The sealed tube was then
frozen in liquid nitrogen and connected to a vacuum pump. It was then
let to reach room temperature and then pressurized with ethylene
(3 atm). The tube was then heated at 508C for 4 h. After this time, poly-
ethylene was visible in the tube (see the Supporting Information). Ex-
treme care should be applied when manipulating a NMR tube under
high pressure. The tube should be manipulated with adequate protection
equipment (thick leather gloves and safety glasses).
[6] L. Piche, J. C. Daigle, J. P. Claverie, Chem. Commun. 2011, 47,
7836–7838.
[7] D. Guironnet, P. Roesle, T. Runzi, I. Gottker-Schnetmann, S. Meck-
ing, J. Am. Chem. Soc. 2009, 131, 422–423.
[8] a) J. Vela, G. R. Lief, Z. Shen, R. F. Jordan, Organometallics 2007,
26, 6624–6635; b) D. Guironnet, L. Caporaso, B. Neuwald, I.
Goettker-Schnetmann, L. Cavallo, S. Mecking, J. Am. Chem. Soc.
2010, 132, 4418–4426.
[9] K. M. Skupov, J. Hobbs, P. R. Marella, D. Conner, S. Golisz, B. L.
Goodall, J. P. Claverie, Macromolecules 2009, 42, 6953–6963.
[10] T. M. J. Anselment, C. E. Anderson, B. Rieger, M. B. Boeddinghaus,
T. F. Fꢄssler, Dalton Trans. 2011, 40, 8304–8313.
[11] T. M. J. Anselment, C. Wichmann, C. E. Anderson, E. Herdtweck,
B. Rieger, Organometallics 2011, 30, 6602–6611.
[12] Z. L. Shen, R. F. Jordan, J. Am. Chem. Soc. 2010, 132, 52–53.
[13] L. Piche, J. C. Daigle, R. Poli, J. P. Claverie, Eur. J. Inorg. Chem.
2010, 2010, 4595–4601.
[14] B. Sundararaju, Z. Tang, M. Achard, G. V. M. Sharma, L. Toupet, C.
Bruneau, Adv. Synth. Catal. 2010, 352, 3141–3146.
[15] D. K. Newsham, S. Borkar, A. Sen, D. M. Conner, B. L. Goodall,
Organometallics 2007, 26, 3636–3638.
Formation of 2g-eth from 2 g-lut: In a nitrogen-filled glovebox, a high
pressure NMR tube from Wilmad Glass was charged with 2g-lut (0.016 g,
0.029 mmol) and of CDCl3 (1.2 mL) was added. The sealed tube was then
frozen in liquid nitrogen and connected to a vacuum pump. It was then
let to reach room temperature and then pressurized with ethylene
(3 atm). After a few minutes, polyethylene was visible in the tube.
CCDC-844018 (2d-lut) contains the supplementary crystallographic data
for this paper. These data can be obtained free of charge from The Cam-
bridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_re-
quest/cif.
[16] M. P. Conley, R. F. Jordan, Angew. Chem. 2011, 123, 3828–3830;
Angew. Chem. Int. Ed. 2011, 50, 3744–3746.
[17] S. Noda, A. Nakamura, T. Kochi, L. W. Chung, K. Morokuma, K.
Nozaki, J. Am. Chem. Soc. 2009, 131, 14088–14100.
Acknowledgements
[18] It is well known that chain-transfer can occur by two possible
modes, either associative displacement from the olefin hydride inter-
mediate requiring b-hydride-elimination before ethylene coordina-
tion or direct b-hydride-transfer to monomer from the alkyl-olefin
complex (see Ref. [2d]). The fact that we have indicated a Pd hy-
dride in Scheme 3 should only be taken as an indication of the
chain-transfer reaction and not as an indication that one or the
other mechanism is predominant.
[19] C. Popeney, Z. Guan, Organometallics 2005, 24, 1145–1155.
[20] D. P. Gates, S. A. Svejda, E. Onate, C. M. Killian, L. K. Johnson,
P. S. White, M. Brookhart, Macromolecules 2000, 33, 2320–2334.
[21] J. S. Jover, N. Fey, J. N. Harvey, G. C. Lloyd-Jones, A. G. Orpen,
G. J. J. Owen-Smith, P. Murray, D. R. J. Hose, R. Osborne, M.
Purdie, Organometallics 2010, 29, 6245–6258.
[22] N. Fey, A. C. Tsipis, S. E. Harris, J. N. Harvey, A. G. Orpen, R. A.
Mansson, Chem. Eur. J. 2006, 12, 291–302.
[23] C. A. Tolman, Chem. Rev. 1977, 77, 313–348.
[24] W. De Graaf, J. Boersma, N. Smeets, J. J. Wilberth, A. L. Spek, G.
Van Koten, Organometallics 1989, 8, 2907–2917.
[25] http://pubs.acs.org/cen/safety/20050228.html.
This work was supported by NSERC (Discovery program). LP and JCD
kindly thank NSERC for graduate fellowships. We wish to thank Mr
Simard, Mrs Ohlund, Mr Kryuchkov and Mr LeBlanc for assistance with
collection of X-ray diffraction, mass spectrometry, and GPC data.
[1] a) L. S. Boffa, B. M. Novak, Chem. Rev. 2000, 100, 1479–1493; b) A.
Nakamura, S. Ito, K. Nozaki, Chem. Rev. 2009, 109, 5215–5244.
[2] a) L. K. Johnson, C. M. Killian, M. Brookhart, J. Am. Chem. Soc.
1995, 117, 6414–6415; b) L. K. Johnson, S. Mecking, M. Brookhart,
J. Am. Chem. Soc. 1996, 118, 267–268; c) S. Mecking, L. K. Johnson,
L. Wang, M. Brookhart, J. Am. Chem. Soc. 1998, 120, 888–899;
d) S. D. Ittel, L. K. Johnson, M. Brookhart, Chem. Rev. 2000, 100,
1169–1203.
[3] E. Drent, R. van Dijk, R. van Ginkel, B. van Oort, R. I. Pugh,
Chem. Commun. 2002, 744–745.
[4] a) T. Kochi, S. Noda, K. Yoshimura, K. Nozaki, J. Am. Chem. Soc.
2007, 129, 8948–8949; b) S. Luo, J. Vela, G. R. Lief, R. F. Jordan, J.
Am. Chem. Soc. 2007, 129, 8946; c) W. Weng, Z. Shen, R. F. Jordan,
J. Am. Chem. Soc. 2007, 129, 15450; d) S. Borkar, D. K. Newsham,
A. Sen, Organometallics 2008, 27, 3331–3334; e) K. M. Skupov, L.
Piche, J. P. Claverie, Macromolecules 2008, 41, 2309–2310; f) S. Ito,
K. Munakata, A. Nakamura, K. Nozaki, J. Am. Chem. Soc. 2009,
Received: November 23, 2011
Published online: February 13, 2012
Chem. Eur. J. 2012, 18, 3277 – 3285
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3285