Steric activation of chelate catalysts: Efficient polyketone catalysts based on four-membered palladium(II) diphosphine chelates

Article abstract of DOI:10.1039/b010063n

Palladium(II) complexes of ligands of the type Ar₂PCH₂-PAr₂ and Ar₂PN(Me)PAr₂ (Ar = ortho-substituted phenyl group) are very efficient catalysts for copolymerisation of CO and C₂H₄

Full text of DOI:10.1039/b010063n

View Article Online / Journal Homepage / Table of Contents for this issue
Steric activation of chelate catalysts: efficient polyketone catalysts based on
four-membered palladium(II) diphosphine chelates
Stephen J. Dossett,*a Amy Gillon,b A. Guy Orpen,b James S. Fleming,b Paul G. Pringle,*b Duncan F. Wassa
and Matthew D. Jonesa
a BP, Chertsey Road, Sunbury-on-Thames, Middlesex, UK TW16 7LN
b School of Chemistry, University of Bristol, Cantock’s Close, Bristol, UK BS8 1TS.
E-mail: paul.pringle@bristol.ac.uk
Received (in Cambridge, UK) 18th December 2000, Accepted 6th February 2001
First published as an Advance Article on the web 26th March 2001
the complexes of unsubstituted ligand 1a.3 The catalyst
sensitivity to chelate ring size might be rationalised in terms of
bite angle effects.4 However, we show here that the activity of
4-membered Pd–P–C–P chelates is dramatically increased by
the presence of bulky ortho substituents on the aryl groups and
that polymerisation rates comparable to, or even exceeding, the
best commercial catalysts are obtained with 4-membered Pd–P–
N–P chelates.5,6
Palladium(II) complexes of ligands of the type Ar2PCH2-
PAr2 and Ar2PN(Me)PAr2 (Ar = ortho-substituted phenyl
group) are very efficient catalysts for copolymerisation of
CO and C2H4.
The perfectly alternating polyketone made by palladium(II)-
catalysed copolymerisation of CO and alkenes [eqn. (1)] has
The diphosphine ligands 3b–e were made from bis(di-
(1)
chlorophosphino)methane7 according to eqn. (2) and fully
numerous potential applications.1 Several types of palladium(II
)
chelates have been reported to be catalysts1,2 but diphosphine
chelates outperform all others in terms of activity and
selectivity. Drent et al.1 have shown that the CO/C2H4
copolymerisation catalytic activity of palladium(II) complexes
of diphosphines of the type Ar2P(CH2)nPAr2 is a sensitive
function of the backbone chain length, n, and the aryl group, Ar.
(2)
characterised. The catalysts were tested for the production of
ethene/propene/CO terpolymer over a 3 h period at 50 barg and
70 °C in dichloromethane using tris(pentafluorophenyl)borane
promoter. In Table 1, the polymer yield is a measure of the
productivity of the catalysts over 3 h. The polymerisation rate
and catalyst half-life, as determined by fitting first order curves
to the cumulative gas uptake profiles, are also given in Table 1.
The borane activation method was reported recently and
involves the transfer of a pentafluorophenyl group from boron
to palladium to form a cationic Pd–C6F5 complex which is the
catalytically active species.6 The catalysts were formed by the
reaction of palladium acetate and ligand in situ or using pre-
formed palladium acetate complexes. The results obtained in
terms of rate and productivity are the same (within the accuracy
of our measurements) for both of these methods, indicating that
the same catalytic species is formed in both cases (see Table 1,
entries 3 and 4).
Thus it was shown that the 6-membered chelate derived from 1a
was an order of magnitude more active than the 5-membered
analogue from 2a and that the 4-membered chelate from 3a was
essentially inactive. It was also disclosed that the complex of
1b, containing ortho-methoxy substituents, gave a polymer of
significantly greater molecular weight than that obtained with
Polydispersities are consistently around the value of 2, as
expected for a single site catalyst, and termonomer incorpora-
Table 1 Copolymerisation of CO and C2H4 results
Rateb/
Half-lifeb/
C3H6 incorp.
(%)
Entry
Liganda
Yield/g
g gPd21 h21 min
Mw
58000
81000
78000
71000
98000
278000
Mn
24000
31000
42000
42000
50000
107000
PDI
c
c
1
2
3
4
5
3ad
3bd
3c
3cd
3dd
3ed
3ee
5a
0.2
1.5
2.8
3.0
5.0
25.1
0.3
0.4
2.4
2.6
1.9
1.7
2.0
2.6
6.0
6.0
6.6
6.3
6.5
5.0
1029
2020
2185
2776
9396
44
38
35
53
6
7
165
c
c
c
17000
13000
1.3
7.1
c
f
f
f
f
8
9
10
11
12
13
5b
5c
5d
5e
5ee
12.0
7.7
10.4
17.0
0.2
4800
3454
4029
160
144
152
103000
282000
639000
888000
20000
43000
113000
179000
355000
9300
2.4
2.5
3.6
2.5
2.1
3.6g
3.6g
3.7g
3.7g
4.9g
11517
134
c
c
a Polymerisation conditions: 50 barg C2H4/CO, 20.0 g C3H6, CH2Cl2 diluent, 70 °C for 3 h using B(C6F5)3 promoter (0.2 mmol). Catalyst solution formed
in situ with Pd(OAc)2 (0.01 mmol) and ligand (0.01 mmol). b Determined by fitting first order curve to cumulative gas uptake profile. c Rate too low to be
determined. d Conditions as above, except [Pd(OAc)2(ligand)] (0.01mmol) complex was pre-formed. e Conditions as above, except methanol diluent and
HBF4·OMe2 promoter (0.05 mmol). f Only oligomeric products were formed. g 12.0 g Propene used.
DOI: 10.1039/B010063n
Chem. Commun., 2001, 699–700
This journal is © The Royal Society of Chemistry 2001
699

View Article Online
tion is largely unaffected by ligand structure. However, other
aspects of the polymerisations are very sensitive to the structure
of the ligand. Generally, polymerisation rate, catalyst stability
and polymer molecular weight all increase with increasing
steric bulk of the aryl ortho substituent (entries 1–6). These
effects are very pronounced with the o-isopropylphenyl deriva-
tive 3e (entry 6) where, compared with the o-ethylphenyl
derivative 3d (entry 5), the productivity has increased 5-fold
and the molecular weight has doubled. Polymerisation under the
methanolic conditions described by Drent1 gave only a very low
rate (entry 7), showing the importance of the activation method
used. Indeed, even dppm (3a) produces polymer under our
conditions, albeit in very low yield (entry 1). Derivative 3b
(entry 2), bearing polar methoxy substituents, gives some
improvement over the performance of 3a; however, productiv-
ity is less than one third that of the sterically similar ortho ethyl
compound 3d.
observed with 4e are associated with the presence of the bulky
ortho substituents. Thus we tentatively assign the fluxionality of
4e to interconversion of symmetrical and unsymmetrical
rotamers arising from restricted rotation around the aryl–P
bond.
The success of bulky diarylphosphinomethane ligands
prompted us to investigate other one-atom backbone diph-
osphines in which similar steric effects may be observed.
Amino-bridged diphosphines offered such a suitably versatile
ligand architecture. Although the chemistry of the simple
ligands based on this structural motif has been studied,8 the
synthesis of bulky derivatives 5b–e and their use in catalysis has
not been explored. The methylamino-bridged diphosphines
(5a–e) were made according to eqn. (3) and fully charac-
terised.
(3)
The correlation between the steric bulk of the ortho
substituents and the catalytic performance of the 4-membered
chelates led us to investigate further the source of this steric
activation by carrying out solid state and solution structural
studies of model chelates of general structure 4 (M = Pd, X =
Cl, OAc; M = Pt, X = Cl). Most enlightening to the catalysis
The palladium(II) complexes of ligands 5a–e were tested for
polymerisation activity as before (entries 8–13) and the results
largely parallel those obtained with (3a–e). Thus, polymer-
isation rate, catalyst stability and polymer molecular weight all
increase with increasing steric bulk of the aryl ortho substituent
(entries 8–13) and methanolic conditions give only very low
activity (entry 13). In this case, the ortho-methoxy derivative 5b
(entry 9) shows comparable productivity to the ethyl derivative
5d, albeit to give a lower molecular weight material. The N1-
backbone chelates are consistently superior to their C1-
backbone analogues and the polymerisation rate of the most
active catalyst in this family, 5e, exceeds that of even dppp (1a)
under our conditions.6 The molecular weight of the polymer
produced by this system is also extremely high compared with
other polyketone catalysts.1,3,6
discussed here is the structure of the chelate complex 4e (M =
Pt, X = Cl), a model for the very active catalyst system derived
from 3e. The crystal structure of 4e was determined (as its
dichloromethane solvate) and is shown in Fig. 1.† It shows a flat
4-membered chelate ring with, as a consequence, isoclinal aryl
groups, two of which have ortho-isopropyl groups orientated so
as to block the axial sites at the metal. The NMR spectra for 4e
show that this species is fluxional. The 31P NMR spectrum of 4e
in CDCl3 at +20 °C is a singlet at 264.5 ppm [1J(PtP), 3130 Hz]
but this signal broadens as the temperature is lowered and at
260 °C, there appears to be a 1+1 mixture of two species which
Our studies have shown that 4-membered palladium(II
)
diphosphine chelates, hitherto considered to be inefficient
polyketone catalysts, are in fact very active when sterically
demanding derivatives are used.
1
give rise to an AB pattern [dA 265.4, J(PtPA) 3045 Hz, dB
Notes and references
1
1
266.1, J(PtPB) 3337 Hz, J(PAPB) 48 Hz] and a singlet [dC
268.4, 1J(PtP) 3195 Hz]. The 1H NMR signals for the
isopropylgroups of 4e are broad at ambient temperatures and at
260 °C this signal is resolved into overlapping complex
† Crystal data for 4b·CH2Cl2: C38H48Cl4P2Pt, M = 903.59, monoclinic,
space group P21/c (no. 14), a = 12.942(2), b = 12.338(2), c = 23.711(7)
Å, b = 98.64(2)°, U = 3743.1(15) Å3, Z = 4, m = 4.146 mm21, T = 173
K, 8581 unique data, R1 = 0.0220.
CCDC 155957. See http://www.rsc.org/suppdata/cc/b0/b010063n/ for
crystallographic files in .cif or other electronic format.
1
multiplets. The H and 31P NMR spectra for the parent dppm
complex 4a (M = Pt, X = Cl) are invariant with temperature
and therefore the changes in the NMR spectra with temperature
1 E. Drent and P. H. M. Budzelaar, Chem. Rev., 1996, 96, 663 and
references therein; E. Drent, J. A. M. vanBroekhoven and P. H. M.
Budzelaar, Recl. Trav. Chim., 1996, 115, 263; W. P. Mul, H. Oosterbeek,
G. A. Beitel, G. J. Kramer and E. Drent, Angew. Chem., Int. Ed., 2000, 39,
1848; E. Drent, Eur. Pat. Appl.,1984, EP121965 (to Shell).
2 S. Doherty, G. R. Eastham, R. P. Tooze, T. H. Scanlan, D. Williams,
M. R. J. Elsegood and W. Clegg, Organometallics, 1999, 18, 3558; B.
Milani and G. Mestroni, Comments on Inorg. Chem., 1999, 20, 301; K.
Nozaki, T. Hiyama, S. Kacker and I. T. Horvath, Organometallics, 2000,
19, 2031; C. Gambs, S. Chaloupka, G. Consiglio and A. Togni, Angew.
Chem., Int. Ed., 2000, 39, 2486; K. Vrieze, J. H. Groen, J. G. P. Delis,
C. J. Elsevier and P. W. N. M. vanLeeuwen, New. J. Chem., 1997, 21,
807.
3 J. A. M. van Broekhoven and R. L Wife, Eur. Pat. Appl., 1988, EP257663
(to Shell).
4 P. W. N. M van Leeuwen and P. Diekes, J. Chem Soc., Dalton Trans.,
1999, 1519.
5 For first disclosure of these catalysts see: S. J. Dossett, World Pat. Appl.,
1997, 97/37765 (to BP Chemicals); S. J. Dossett, World Pat Appl., 2000,
00/06299 (to BP Chemicals); S. J. Dossett J. S. Fleming and P. G. Pringle,
World Pat. Appl., 2000, 00/03803 (to BP Chemicals).
6 G. K. Barlow, J. D. Boyle, N. A. Cooley, T. Ghaffar and D. F. Wass,
Organometallics, 2000,19, 1470 and references therein.
7 S. Heitkamp, H. Sommer and O. Stelzer, Inorg. Synth., 1989, 27, 120.
8 M. S. Balakrishna, V. Sreenivasa Reddy, S. S. Krishnamurthy, J. F.
Nixon and J. C. T. R. Burkett St. Laurent, Coord. Chem. Rev., 1994, 129,
1.
Fig. 1 Molecular structure and numbering scheme for 4b. All but tertiary
isopropyl hydrogen atoms have been omitted for clarity. Important
molecular bond lengths (Å) and angles (°): Pt(1)–P(1) 2.2301(9), Pt(1)–P(2)
2.2202(9), Pt(1)–Cl(1) 2.3627(9), Pt(1)–Cl(2) 2.3465(9); P(1)–Pt(1)–P(2)
75.96(3), P(1)–C(1)–P(2) 94.54(13).
700
Chem. Commun., 2001, 699–700