Novel Cr-PNP complexes as catalysts for the trimerisation of ethylene

Article abstract of DOI:10.1039/b210878j

Cr(III) complexes of tridentate PNP ligands have been prepared and evaluated as catalysts for ethylene trimerisation, with several giving high activity and excellent selectivity towards 1-hexene when activated with methylaluminoxane.

Full text of DOI:10.1039/b210878j

Novel Cr-PNP complexes as catalysts for the trimerisation of ethylene
David S. McGuinness†,*a Peter Wasserscheid,*a Wilhelm Keim,a Chunhua Hu,b Ulli Englert,b John
T. Dixonc and Cronje Grovec
a Institut für Technische Chemie und Makromolekulare Chemie der RWTH Aachen, Worringerweg 1, 52074
Aachen, Germany. E-mail: wasserscheid@itmc.rwth-aachen.de; Fax: 49 (0)241 8022177; Tel: 49 (0)241
8026454
b Institut für Anorganische Chemie der RWTH Aachen, Professor Pirlet Strasse 1, 52074 Aachen, Germany.
E-mail: ullrich.englert@ac.rwth-aachen.de; Fax: 49 (0)241 8092288; Tel: 49 (0)241 8094666
c
Sasol Technology (Pty) Ltd, R&D Division, 1 Klasie Havenga Road, Sasolburg 9570, South Africa.
E-mail: johntho.dixon@sasol.com; Fax: 27 (0)16 960 2826; Tel: 27 (0)16 960 2900
Received (in Cambridge, UK) 6th November 2002, Accepted 10th December 2002
First published as an Advance Article on the web 2nd January 2003
Cr(III) complexes of tridentate PNP ligands have been
prepared and evaluated as catalysts for ethylene trimerisa-
tion, with several giving high activity and excellent selectiv-
ity towards 1-hexene when activated with methylalumi-
noxane.
first to be prepared and tested for ethylene trimerisation, and
when activated with 120 equivalents of MAO was found to give
reasonable activity with excellent selectivity toward 1-hexene
(Table 1, entry 1). Attempts to improve the system centred first
on modification of the phosphorus substitution. Highly basic
but sterically demanding dicyclohexylphosphino substituents
(complex 2) led to a dramatic decrease in activity, with the
product being predominately polymer (entry 2). In contrast,
under comparable conditions basic but compact diethylphos-
phino substituents (complex 3) led to a large increase in activity
with excellent 1-hexene selectivity (entry 3).
When a greater amount of MAO was introduced (680 equiv.)
along with complex 1 activity doubled, however a higher
proportion of polymer and 6% 1-butene was formed at the
expense of hexene selectivity (entry 4). In contrast, at higher
MAO loadings (300–700 equiv.) complex 3 led to a pronounced
increase in activity while high hexene selectivity was main-
tained (entries 5 and 6). The highest activity achieved with
complex 3 was 69 340 turnovers h21 (37400 gProduct per gCr
per h) when activated with 850 equivalents of MAO, however a
somewhat lower hexene selectivity resulted (entry 7). Increas-
ing the amount of MAO beyond this does not significantly
further increase the activity of the complexes. 13C NMR
analysis of polymer samples resulting from runs 2, 3 and 7
While the oligomerisation of ethylene generally gives a broad
range of a-olefins, there is increasing interest in the develop-
ment of catalysts which are able to give greater selectivity
toward more desirable olefins. Of particular interest is the linear
trimerisation of ethylene to 1-hexene, due to the importance of
this co-monomer in the production of polyethylene. Of the
systems known to trimerise ethylene, most are based on
homogeneous Cr catalysts, although several catalysts based on
Ti1 and Ta2 have recently been reported. The most successful of
the Cr based complexes appear to be pyrrolyl-Cr complexes3 as
well as complexes of neutral phosphorus containing ligands,4,5
while neutral tridentate NNN ligands have been applied in the
trimerisation of a-olefins.6 In the course of investigations on
alternative catalysts for ethylene trimerisation, we became
interested in tridentate PNP ligands in combination with Cr as
procatalysts for this reaction. Here we report new complexes of
Cr(III) that, in combination with methylaluminoxane (MAO),
selectively convert ethylene to 1-hexene with TOFs in ethylene
up to ca. 70 000 h21
.
The tridentate PNP ligands I–III were prepared via literature
procedures or via adaptation of these.7 When reacted with
CrCl3(thf)3 the immediate formation of complexes 1–3 was
observed, which were isolated as deep blue-purple powders in
92–100% yield after workup (Scheme 1).‡ Complex 1 was the
† Present address of D. S. McGuinness: Department of Chemistry, Imperial
College London, Exhibition Rd, South Kensington, UK SW7 2AY. Email:
d.mcguinness@ic.ac.uk.
Scheme 1
Table 1 Ethylene trimerisation with complexes 1–3a
1-Hexene
Catalyst
(amount mmol) (equivalents)
MAO
Temperature
(°C)
Run
time (min)
Hexenes
(wt%)
selectivity
(wt%)b
Productivity
(h21 c
Entry
PE (wt%)
)
1
2
3
4
5
6
7
8
9
1 (40.2)
2 (42.3)
3 (50.0)
1 (11.0)
3 (23.0)
3 (10.8)
3 (8.8)
3 (12.0)
3 (13.0)
3 (11.0)
3 (11.8)
3 (11.3)
120
120
100
680
340
690
850
630
600
680
640
660
100
100
100
100
100
100
100
100
50
30
30
30
30
30
30
30
60
60
30
60
30
0.1
85.7
0.2
10.2
0.3
0.7
2.1
0.3
21.9
0.4
98
14
98
83
98
98
94
92
78
97
97
93
99.2
80.0
99.2
99.1
98.9
98.9
99.1
99.2
98.8
99.2
99.3
99.3
8 670
580
17 300
17 620
48 060
58 570
69 340
31 220
2 950
50 770
39 890
49 620
10
11
12
80
80
120
0.3
0.2
a 40 bar ethylene, 25mL toluene. b Selectivity for 1-Hexene as a percentage of total hexenes. c Average turnover frequency of ethylene conversion.
334
CHEM. COMMUN., 2003, 334–335
This journal is © The Royal Society of Chemistry 2003

View Article Online
showed no evidence for 1-hexene incorporation in the polyeth-
ylene chain.
It was noted in all runs conducted at 100 °C with complex 3
that ethylene uptake dramatically decreases towards the end of
a run (30 min), suggesting deactivation of the catalyst. This is
verified by a 1 h run (entry 8), in which the average activity
obtained is roughly half of that for a 30 min run conducted under
the same conditions. At 50 °C the catalytic activity drops
significantly, and a high proportion of polymer is formed (entry
9). At 80 °C the high activity and selectivity of the catalyst is
restored (entry 10). Furthermore, towards the end of this 30 min
run ethylene uptake continues, suggesting a greater catalyst
stability at this temperature. In a 1 h run at 80 °C (entry 11) the
average activity of the catalyst is somewhat lower, however the
total turnover number obtained (39 890) is higher than in all
other runs, reflecting a longer catalyst lifetime at this tem-
perature. To further verify a thermally induced catalyst
deactivation at higher temperatures, when run at 120 °C (entry
12) the catalyst was virtually inactive after only 15 min, a lower
total activity relative to runs at 80–100 °C resulted, and the
formation of higher oligomers occurred which reduced the
hexene selectivity to 93%. These results reveal a high sensitivity
of the catalyst system towards varying reaction conditions, both
in terms of activity and selectivity. Careful optimisation of the
reaction conditions is ongoing.
Crystals of complexes 1 and 2 suitable for X-ray diffraction
studies were grown by slow evaporation of a solution of the
complex in DMSO (1) or acetone (2). The molecular structure
of 1, along with selected bond distances and angles, is shown in
Fig. 1, while that for complex 2 is shown in Fig. 2§ Both
complexes display a slightly distorted octahedral geometry,
with the tridentate ligand coordinated in a meridional fashion.
The chelate bite angles of the PNP ligand in both complexes are
similar [81.08(8), 82.07(8)° (1); 82.54(7), 81.99(7)° (2)], as are
the Cr–P distances, which for the Cr–P1 bond of each complex
are the same within experimental error. The Cr–N distances in
each complex [2.139(3) Å (1); 2.137(3) Å (2)] are likewise
equal within experimental error and are within the range of
Cr(III) amine bond lengths (ca 2.05–2.19 Å).8 The greater steric
demand of the cyclohexyl relative to the phenyl group, which
represents the main difference between the two structures, is
reflected in the larger angles C5–P1–C11 and C17–P2–C23
(Figs. 1 and 2).
Fig. 2 Molecular structure of 2, 30% probability ellipsoids, H atoms omitted
for clarity. Selected bond distances (Å) and angles (°): Cr–PP1 2.4662(10),
Cr–PP2 2.4736(10), Cr–PN 2.137(3), Cr–PCl1 2.2898(10), Cr–PCl2
2.3090(11), Cr–PCl3 2.3616(11), N–PCr–PP1 82.54(7), N–PCr–PP2
81.99(7), P1–PCr–PP2 164.49(4), N–PCr–PCl1 178.96(8), N–PCr–PCl2
86.63(8), N–PCr–PCl3 84.95(8), C5–PP1–PC11 112.51(16), C17–PP2–
PC23 112.19(18).
structures represent only that of the procatalyst, and it is
possible that in the active catalyst the ligand is coordinated in a
manner that is more sensitive to greater steric bulk of the
phosphine, such as a facial coordination geometry. This is under
further investigation. We are also exploring the use of alternate
tridentate ligands, incorporating different donor atoms, such
that the use of phosphines can be avoided. The results of this
work will be published in due course.
D. S. M. and P. W. thank Dr Mike Green and David Morgan
for fruitful discussions and Sasol Technology Ltd for financial
support.
Notes and references
‡ Elemental analysis and mass spectral (+FAB) data for 1–3. 1:
C28H29Cl3CrNP2 calcd. (found) C 56.07 (55.84), N 2.34 (2.14), H 4.87
(5.16)%. m/z 598 [M]+, 563 [M 2 Cl]+. 2: C28H53Cl3CrNP2 calcd. (found)
C 53.89 (54.10), N 2.24 (2.14), H 8.56 (8.54)%. m/z 624 [M]+, 589 [M 2
Cl]+. 3: C12H29NP2Cl3Cr calcd. (found) C 35.36 (35.29), N 3.44 (3.21), H
7.17 (7.49). m/z 352 [M]+, 315 [M 2 Cl]+.
§ 1·DMSO: C30H35Cl3CrNOP2S, M = 677.95, monoclinic, a = 27.667(7),
b = 14.751(4), c = 16.512(4) Å, b = 100.923(7)°, U = 6617(3) Å3, T =
Given the similarities in the molecular structures of 1 and 2 it
is difficult to reconcile the remarkable differences in activity
that each gives. It must be borne in mind however that these
293(2) K, space group C2/c (no. 15), Z = 8, m (Mo-Ka) = 0.772 mm21
,
28626 reflections measured, 6532 unique (R int = 0.0662) which were used
in all calculations. The final R(F) and wR(F2) were 0.0521 (I > 2s(I)) and
0.1512 (all data), respectively. CCDC 179726. 2·acetone:
C31H59Cl3CrNOP2, M
16.2779(11), c = 21.3434(18) Å, b = 90.757(13)°, U = 3592.1(11) Å3, T
= 223(2) K, space group P21/n (no. 14), Z = 4, m (Mo-Ka) = 0.655 mm21
= 682.08, monoclinic, a = 10.340(3), b =
,
10559 reflections measured, 7031 unique (Rint = 0.0494) which were used
in all calculations. The final R(F) and wR(F2) were 0.0520 (I > 2s(I)) and
0.1399 (all data), respectively. CCDC 179727. See http://www.rsc.org/
suppdata/cc/b2/b210878j/ for crystallographic files in CIF or other
electronic format.
1 P. J. W. Deckers, B. Hessen and J. H. Teuben, Angew. Chem., Int. Ed.,
2001, 40, 2516.
2 C. Andes, S. B. Harkins, S. Murtuza, K. Oyler and A. Sen, J. Am. Chem.
Soc., 2001, 123, 7423.
3 M. E. Lashier, EP 0780353A1, 1997 (Phillips Petroleum).
4 F.-J. Wu, US 5811618, EP 0622347B1, 1998 (Amoco).
5 A. Carter, S. A. Cohen, N. A. Cooley, A. Murphey, J. Scutt and D. F.
Wass, Chem. Commun., 2002, 858.
Fig. 1 Molecular structure of 1, 30% probability ellipsoids, H atoms omitted
for clarity. Selected bond distances (Å) and angles (°): Cr–P1 2.4660(12),
Cr–PP2 2.4678(12), Cr–PN 2.139(3), Cr–PCl1 2.2934(11), Cr–PCl2
2.3081(11), Cr–PCl3 2.3480(11), N–PCr–PP1 81.08(8), N–PCr–PP2
82.07(8), P1–PCr–PP2 163.10(4), N–PCr–PCl1 176.74(8), N–PCr–PCl2
87.59(8), N–PCr–PCl3 85.00(8), C5–PP1–PC11 103.42(18), C17–PP2–
PC23 104.40(17).
6 P. Wasserscheid, S. Grimm, R. D. Köhn and M. Haufe, Adv. Synth.
Catal., 2001, 343, 814.
7 A. A. Danopoulos, A. R. Wills and P. G. Edwards, Polyhedron, 1990, 9,
2413.
8 A. Derwahl, F. Wasgestian, D. A. House and W. T. Robinson, Coord.
Chem. Rev., 2001, 211, 45; R. Emrich, O. Heinemann, P. W. Jolly, C.
Krüger and G. P. J. Verhovnik, Organometallics, 1997, 16, 151.
CHEM. COMMUN., 2003, 334–335
335