Ethylene tri- and tetramerization with borate cocatalysts: Effects on activity, selectivity, and catalyst degradation pathways

Article abstract of DOI:10.1021/om060906z

Chromium-based ethylene tri- and tetramerization catalysts have been evaluated with a combination of trialkylaluminium and borate cocatalyst activation. A remarkable effect on activity and selectivity has been observed and studied, and a cocatalyst-induced degradation pathway has been identified.

Full text of DOI:10.1021/om060906z

1108
Organometallics 2007, 26, 1108-1111
Ethylene Tri- and Tetramerization with Borate Cocatalysts: Effects
on Activity, Selectivity, and Catalyst Degradation Pathways
David S. McGuinness,†,§ Matthew Overett,*,‡ Robert P. Tooze,† Kevin Blann,‡
John T. Dixon,‡ and Alexandra M. Z. Slawin
Sasol Technology UK Ltd, Purdie Building, North Haugh, St. Andrews, KY16 9ST, U.K., Sasol Technology
(Pty) Ltd., 1 Klasie HaVenga Road, Sasolburg 1947, South Africa, and School of Chemistry, UniVersity of
St. Andrews, Purdie Building, St. Andrews, KY16 9ST, U.K.
ReceiVed October 3, 2006
Scheme 1
Summary: Chromium-based ethylene tri- and tetramerization
catalysts haVe been eValuated with a combination of trialkyl-
aluminium and borate cocatalyst actiVation. A remarkable effect
on actiVity and selectiVity has been obserVed and studied, and
a cocatalyst-induced degradation pathway has been identified.
Introduction
In recent years the chromium-catalyzed oligomerization of
ethylene to linear R-olefins (LAOs) has been extensively studied
both in industry and in academia, due in part to increased
demand for LAOs, particularly the co-monomers 1-hexene and
1-octene. Recent work includes the development of increasingly
more active systems for full-range oligomerization,1 a variety
of Cr-based ethylene trimerization systems,2-4 and the first
example of selective ethylene tetramerization to 1-octene.5 The
cocatalyst typically employed in the tetramerization reaction,
MAO, is normally thought to implicate a formally cationic active
species, and as such we were interested in exploring the effect
of more well-defined counterions on the tetramerization reaction.
The use of only alkyl aluminum, AlR3, is ineffectual, presumably
due to the low Lewis acidity of these compounds. In contrast,
activation of olefin polymerization catalysts with combinations
of AlR3 (alkylating agent) and the well-known alkyl-abstracting
agents B(C6F5)3 and Ph3C-B(C6F5)4 (BArF3 and [Ph3C][BArF ])
4
is now well established.6 Herein we report on investigations on
the use of BArF3 and [Ph3C][BArF ] in ethylene tetramerization
4
and the surprising shift in catalyst properties that these cocata-
lysts can produce.
* To whom correspondence should be addressed. E-mail:
matthew.overett@sasol.com.
Results and Discussion
† Sasol Technology UK.
The ligands/complexes tested in this study are shown in
Scheme 1 and were prepared as reported previously5a or via
similar procedures. Complex 5 was prepared such that the effects
of differing ligand sterics on the reaction could be studied (see
below). Crystals of this complex suitable for X-ray analysis were
grown from toluene, and the molecular structure of the complex
is shown in Figure 1. A chloride-bridged dimeric structure is
displayed, as has been found previously for an aryl-substituted
PNP ligand.5a The chromium centers of the binuclear complex
display a distorted octahedral geometry, enforced by the tight
chelate bite angle of the PNP ligand [66.62(5)°]. The Cr-Cr
distance of 3.543(2) Å is indicative of little or no metal-metal
bonding.
‡ Sasol Technology (Pty) Ltd.
University of St. Andrews.
§ Present address: School of Chemistry, University of Tasmania, Private
Bag 75, Hobart, Tasmania 7001, Australia. E-mail: david.mcguinness@
utas.edu.au.
(1) (a) McGuinness, D. S.; Gibson, V. C.; Wass, D. F.; Steed, J. W. J.
Am. Chem. Soc. 2003, 125, 12716. (b) Tomov, A. K.; Chirinos, J. J.; Jones,
D. J.; Long, R. J.; Gibson, V. C. J. Am. Chem. Soc. 2005, 127, 10166. (c)
Tomov, A. K.; Chirinos, J. J.; Long, R. J.; Gibson, V. C.; Elsegood, M. R.
J. J. Am. Chem. Soc. 2006, 128, 7704.
(2) (a) McGuinness, D. S.; Wasserscheid, P.; Keim, W.; Morgan, D. H.;
Dixon, J. T.; Bollmann, A.; Maumela, H.; Hess, F. M.; Englert, U. J. Am.
Chem. Soc. 2003, 125, 5272. (b) McGuinness, D. S.; Brown, D. B.; Tooze,
R. P.; Hess, F. M.; Dixon, J. T.; Slawin, A. M. Z. Organometallics 2006,
25, 3605.
(3) (a) Carter, A.; Cohen, S. A.; Cooley, N. A.; Murphy, A.; Scutt, J.;
Wass, D. F. Chem. Commun. 2002, 858. (b) Agapie, T.; Day, M. W.;
Henling, L. M.; Labinger, J. A.; Bercaw, J. E. Organometallics 2006, 25,
2733.
(4) Blann, K.; Bollmann, A.; Dixon, J. T.; Hess, F. M.; Killian, E.;
Maumela, H.; Morgan, D. H.; Neveling, A.; Otto, S.; Overett, M. Chem.
Commun. 2005, 620.
(5) (a) Bollmann, A.; Blann, K.; Dixon, J. T.; Hess, F. M.; Killian, E.;
Maumela, H.; McGuinness, D. S.; Morgan, D. H; Neveling, A.; Otto, S.;
Overett, M.;. Slawin, A. M. Z.; Wasserscheid, P.; Kuhlmann, S. J. Am.
Chem. Soc. 2004, 126, 14712. (b) Overett, M. J.; Blann, K.; Bollmann, A.;
Dixon, J. T.; Hassbroek, D.; Killian, E.; Maumela, H.; McGuinness, D. S.;
Morgan, D. H. J. Am. Chem. Soc. 2005, 127, 10723. (c) Walsh, R.; Morgan,
D. H.; Bollmann, A.; Dixon, J. T. Appl. Catal. A 2006, 306, 184. (d)
Kuhlmann, S.; Dixon, J. T.; Haumann, M.; Morgan, D. H.; Ofili, J.; Spuhl,
O.; Taccardi, N.; Wasserscheid, P. AdV. Synth. Catal. 2006, 348, 1200.
As shown in Table 1, treatment of chromium/ligand combina-
tions or preformed complexes with trialkylaluminium and
stoichiometric BArF3 or [Ph3C][BArF ] leads to active catalysts
4
for ethylene trimerization and tetramerization. In general,
activation with the boranes leads to greatly lower activities
compared to when MAO is employed, and a greater and
somewhat variable polymer content results. Within the liquid
fraction, the distribution of oligomers is very similar to that
obtained with MAO.5 In particular, tetramerization to 1-octene
(6) (a) Chen, E. Y-X.; Marks, T. J. Chem. ReV. 2000, 100, 1391. (b)
Piers, W. E.; Chivers, T. Chem. Soc. ReV. 1997, 26, 345.
10.1021/om060906z CCC: $37.00 © 2007 American Chemical Society
Publication on Web 01/20/2007

Notes
Organometallics, Vol. 26, No. 4, 2007 1109
although again a greater proportion of polymer was formed. It
seems the selectivity within the oligomeric fraction is relatively
unaffected by the cocatalyst, whether MAO or BArF , and
3
remains controlled by the ligand. This suggests that the
environment at the active Cr center is the same in each case.
As such, the high polymer content may be a consequence of a
separate and distinct chromium species that is formed under
conditions of activation with AlR3/BArF .
3
The cocatalyst [Ph3C][BArF ] also resulted in active tetramer-
4
ization catalysts, again with reduced productivity and, in some
cases, a high proportion of polymer (entries 10 and 11). This
cocatalyst seemed somewhat more tolerant to excess AlEt3 than
BArF was, and the use of 100 equiv of AlEt3 (entry 11) did
3
not result in the same deactivation profile as observed when
using AlR3/BArF (e.g., entry 6).
3
In all of the above runs it was noted that activity was initially
quite high, as judged by ethylene uptake, but a rapid catalyst
deactivation occurred. This was more pronounced when AlMe3
was employed as an alkylating agent or when a large excess of
AlR3 was present. A deactivation pathway caused by trialkyl-
aluminium seems likely. Such catalyst deactivation has been
Figure 1. Molecular structure of complex 5. Selected bond
distances (Å) and angles (deg): Cr(1)-P(1) 2.4152(13); Cr(1)-
P(2) 2.4775(15); Cr(1)-Cl(1) 2.3741(13); Cr(1)-Cl(1A) 2.4191-
(12); Cr(1)-Cl(2) 2.2839(14); Cr(1)-Cl(3) 2.3168(13); P(1)-
N(1)-P(2) 106.32(19); P(1)-Cr(1)-Cl(1) 94.28(5); P(1)-Cr(1)-
Cl(2) 88.43(5); P(2)-Cr(1)-Cl(1A) 99.11(5).
noted for Cp2MR2/[Ph3C][BArF ]/AlR3 polymerization systems,
4
is accompanied by trimerization to 1-hexene along with methyl-
and methylenecyclopentane.
prompting Bochmann to study the interaction of AlR3 with
BArF3 and BArF -.7 This work showed that reaction of AlMe3
4
The use of AlEt3 as an alkylating agent along with BArF
3
with [Ph3C][BArF ] or BArF can lead to degradation of the
4
3
was far more effective than AlMe3 (entries 1 and 2). The
productivity was further improved by lowering the temperature
from 65 °C to 45 °C (entry 3), and all subsequent runs are at
this lower temperature. The amount of AlEt3 employed has a
pronounced effect on productivity, as shown by entries 4-6.
Productivity is highest with between 30 and 50 equiv of AlEt3,
while 100 equiv led to an order of magnitude drop in
productivity. The best result in terms of combined selectivity
and activity was obtained with the catalyst derived from 2/Cr
with 50 equiv of AlEt3 (entry 5), which gave only 2%
polyethylene.
borane compound through alkyl-C6F5 exchange, resulting in
mixtures of BArF3-xMex and AlMe3-xArF (x ) 1-3). While
x
AlArF has been shown to be an effective cocatalyst,8 mixed
3
AlR3-xArFx species are expected to be much less so. It was also
shown that this can lead to catalyst deactivation through transfer
of ArF to the electrophilic catalyst metal center.7 Increasing the
concentration of AlR3 will both accelerate this scrambling
process and push the equilibrium toward increased alkylation
of the borane. The observation that increased amounts of
trialkylaluminium lead to deactivation thus supports such a mode
of catalyst degradation.
The preformed complex 4 was also activated by AlEt3/BArF
3
This reaction likely occurs through dimeric Al-B intermedi-
ates, and as such, it was reasoned that a sterically encumbered
trialkylaluminium reagent may slow the process. AliBu3 was
(entries 7 and 8), although the productivity was considerably
lower than found in comparable in-situ runs. Such a trend was
also observed when MAO activation was employed5 and may
relate to the dimeric and highly insoluble nature of the
complexes. The addition of a large excess of AlEt3, along with
a concomitant increase in the amount of BArF3 to 5 equiv, did
not improve the results (entry 8).
We have previously shown that the introduction of steric bulk
on the PNP ligands leads to a dramatic selectivity shift toward
1-hexene.4 For instance ligand 3 gives over 90% selectivity to
1-hexene under activation with MAO. The same result was
observed when 3/Cr was activated with AlEt3/BArF3 (entry 9),
therefore tested as an alkylating agent with both BArF and
3
[Ph3C][BArF ] (Table 1, entries 12 and 13) and led to the highest
4
productivities in each case. Unfortunately, the high productivity
was also accompanied by very high polymer production (ca.
50%).
Attempts to optimize the system resulted in the amount of
[Ph3C][BArF ] being increased to 5 equiv. To our surprise this
4
led to a remarkable increase in activity and an equally dramatic
change in selectivity from tri-/tetramerization to a Schulz-Flory
Table 1. Ethylene Tri- and Tetramerization with 1-4a
cocatalyst
AlR3
(equiv)
entry
cat
(equiv)
T (°C)
P (bar g)
PE (wt %)
C6 (wt %) (1 - C6)
C8 (wt %) (1 - C8)
prod (g/g Cr)b
1
2
3
4
5
6
7
8
1/Cr
1/Cr
1/Cr
2/Cr
2/Cr
2/Cr
4
AlEt3 (40)
AlMe3 (50)
AlEt3 (40)
AlEt3 (30)
AlEt3 (50)
AlEt3 (100)
AlEt3 (40)
AlEt3 (250)
AlEt3 (40)
AlEt3 (50)
AlEt3 (100)
AliBu3 (30)
AliBu3 (30)
BArF3 (1)
60
60
45
45
45
45
45
45
45
45
45
45
45
45
45
50
50
50
50
50
50
50
40
40
50
50
6.6
17.7
31.3
13.0
2.0
25.2
22.0
7.0
14.7
15.8
trace
50.6
45.1
20.4 (65.1)
26.8 (71.4)
12.8 (44.7)
16.0 (69.0)
28.0 (73.0)
21.4 (73.3)
19.0 (69.0)
21.0 (50.0)
79.3 (99.3)
17.3 (71.4)
19.6 (69.2)
7.7 (66.5)
44.4 (93.4)
51.7 (83.3)
41.2 (95.0)
62.1 (98.3)
67.0 (99.0)
47.6 (96.9)
53.0 (98.0)
40.0 (95.0)
4.8 (100)
62.2 (98.9)
70.8 (98.3)
34.6 (98.6)
39.0 (98.7)
1930
200
5075
6280
7110
587
BArF3 (1)
BArF3 (1)
BArF3 (1)
BArF3 (1)
BArF3 (1)
BArF3 (1)
1930
570
4
BArF3 (5)
9
3/Cr
2/Cr
2/Cr
2/Cr
2/Cr
BArF3 (1)
5580
2595
2540
14 790
10 540
10
11
12
13
Ph3CBArF4 (1)
Ph3CBArF4 (1)
BArF3 (1)
Ph3CBArF4 (1)
8.0 (66.3)
b
a 10 µmol of CrCl3(thf)3, 12 µmol of ligand, 100 mL of toluene as solvent, 30 min. Productivity expressed in g(total product)‚g(Cr)-1
.

1110 Organometallics, Vol. 26, No. 4, 2007
Notes
Table 2. Ethylene Oligomerization with 5 equiv of
Scheme 2
[Ph3C][BArF ]a
4
AlR3
entry cat
(equiv)
PE wt % olig wt %
K
prod (g/g Cr)
1
2
3
4
5
6
7
2/Cr AlEt3 (50)
2/Cr AlEt3 (200)
<0.1
0.4
0.2
1.0
6.3
3.4
1.0
99.9
99.6
99.8
99.0
93.7
96.6
99.0
0.55
0.53
0.54
0.49
0.51
0.52
0.60
59 825
78 850
34 300
38 222
34 410
13 065
25 450
4
4
AlEt3 (200)
AlMe3 (200)
Cr leads to catalysis very similar to the PNP system’s, and there
appears to be very little ligand influence when 5 equiv of [Ph3C]-
[BArF ] is present. A ligand abstraction by excess [Ph3C][BArF ]
3/Cr AlMe3 (200)
5
Cr
AlEt3 (50)
AlEt3 (50)
4
4
a 10 µmol of Cr, 12 µmol of ligand, 45 °C, 40 bar g, 100 mL of toluene,
30 min.
seems likely, and the interaction between the PNP ligands and
[Ph3C][BArF ] was therefore studied by NMR spectroscopy.
4
When a C6D6 solution of 2 (31P: 50 ppm) was treated with
distribution of R-olefins (Table 2).9 For instance, with a
combination of CrCl3(thf)3 and 2, activated with 50 equiv of
an equimolar amount of [Ph3C][BArF ], some precipitation
4
occurs and the singlet gives way to two coupled doublets (J )
65 Hz) at 65 and 47 ppm, along with two weaker resonances at
53 and 22 ppm (the signals are quite weak due to low solubility
of the products). After some hours the signals at 53 and 22
ppm are predominant. After sitting overnight, the C6D6 was
decanted from the precipitate and the product taken up in CDCl3,
in which only the peaks at 53 and 22 ppm are observed. There
AlEt3 and 5 equiv of [Ph3C][BArF ], a productivity of 59 825
4
g/g Cr resulted (Table 2, entry 1). At the start of this run, an
exotherm occurred that took the temperature from 45 °C to 80
°C upon pressurizing with ethylene, the temperature being
brought back to 45 °C over ca. 4 min with internal cooling of
the reactor. Likewise, all experiments listed in Table 2 required
continued cooling throughout the run to maintain the temper-
ature. The shift to Schulz-Flory oligomerization in these
examples was accompanied by a large reduction in the amount
of polymer formed, and the R-content in each olefin fraction
was generally above 99%.
is no change to the BArF - anion, as shown in the 19F NMR
4
spectrum. The 31P, 1H, and 19F NMR are all consistent with the
reaction sequence shown in Scheme 2. Unfortunately, attempts
to isolate this product resulted in a sticky solid that could not
be further purified.
Subsequent experiments showed that this selectivity change
occurs regardless of the amount of trialkylaluminium (entry 2),
whether the catalyst is preformed (entry 3) or whether AlMe3
is employed in place of AlEt3 (entry 4). The sole deciding factor
A further experiment shows that excess [Ph3C][BArF ] can
4
abstract the PNP ligands from an alkylated Cr center. The
methylated complex CrCl2Me(1)(thf) (6) was prepared from
CrCl2Me(thf)3 and ligand 1 in toluene. When this was treated
appears to be the amount of [Ph3C][BArF ] employed. Ad-
4
with 3.5 equiv of [Ph3C][BArF ] in CDCl3, the 1H NMR
4
ditionally, the rapid catalyst deactivation observed in the
tetramerization runs when a single equivalent of cocatalyst was
used was not evident when 5 equiv of this activator was used.
Evidently, a very different catalytic species is forming under
these conditions, and we were intrigued to study this further,
in particular to establish if the olefin distribution could be shifted
to more interesting (higher) K values. The introduction of
increased steric bulk at the metal center has been shown to
increase the molecular weight distribution in both ethylene
oligomerization and polymerization. The more bulky PNP ligand
spectrum clearly shows the expected formation, through alkyl
abstraction, of 1,1,1-triphenylethane. The reaction also cleanly
affords two signals in the 31P NMR spectrum at 37 and 22 ppm
(free ligand 70 ppm), consistent with the iminophosphine.
It is evident from the above that excess [Ph3C][BArF ] can
4
result in ligand loss from Cr. At the very least, this reaction
would destroy the bidentate coordination mode of the ligand
and may even lead to complete dissociation. It seems that the
alkyl group must be abstracted from Cr preferentially, as with
1 equiv of [Ph3C][BArF ] an active tetramerization catalyst
4
3 was therefore tested along with 5 equiv of [Ph3C][BArF ].
4
results. However, with excess trityl cation present, an essentially
unligated Cr species is formed that is highly active for Schulz-
Flory oligomerization.
This ligand was expected to affect the Schulz-Flory distribution,
especially given the pronounced shift in selectivity that occurs
in the tri-/tetramerization reaction when 2 and 3 are compared.
Surprisingly however, 3 led to no change in the K value (entry
5). To further explore this, complex 5, which contains the
compact alkyl-substituted PNP ligand, was also tested. The low
steric demand of this ligand is evident from the crystal structure,
but nonetheless when tested for oligomerization the olefin
distribution was effectively unchanged (entry 6).
These observations are somewhat unexpected if the ligand
mediates oligomerization on Cr and leads to the conclusion that
the ligand may not be involved when a Schulz-Flory distribu-
tion is observed. To test for this, CrCl3(thf)3 was tested without
PNP ligand present, as an approximation of “naked” Cr (entry
7). This led to an uncontrollable exotherm to 110 °C, which
was slowly brought back to 45 °C with cooling of the reactor.
Although the K value obtained was modestly higher than the
PNP-containing catalyst’s, it can be concluded that unligated
Summary and Conclusion
We have shown that fluorinated borane and borate cocatalysts,
in combination with an alkylating agent, give rise to active
trimerization and tetramerization catalysts with Cr/PNP com-
binations. Liquid fraction selectivities are similar to when MAO
is used; however productivities are much reduced and a higher
proportion of polymer is formed. The requirement for known
alkyl-abstracting reagents in catalysis points toward a formally
cationic metal center in the tetramerization reaction.10
A
dramatic shift in activity and selectivity that occurs when excess
[Ph3C][BArF ] is employed is likely due to ligand abstraction
4
from the Cr, rather than a mechanism mediated by the PNP
ligand. This result cautions that careful selection of catalyst/
cocatalyst ratios may be required when highly Lewis acidic
activators are employed in olefin oligomerization and polym-
(7) Bochmann, M.; Sarsfield, M. J. Organometallics 1998, 17, 5908.
(8) Chen, E. Y.-X.; Kruper, W. J.; Roof, G.; Wilson, D. R. J. Am. Chem.
Soc. 2001, 123, 745.
(9) The Schulz-Flory distribution is characterized by the K value; K )
mol(Cn+2)/mol(Cn).
(10) (a) Agapie, T.; Schofer, S. J.; Labinger, J. A.; Bercaw, J. E. J. Am.
Chem. Soc. 2004, 126, 1304. (b) Jabri, A.; Crewdson, P.; Gambarotta, S.;
Korobkov, I.; Duchateau, R. Organometallics 2006, 25, 715.

Notes
Organometallics, Vol. 26, No. 4, 2007 1111
106.366(6)°, V ) 1617.3(11) Å3, Z ) 2, Dcalcd ) 1.501 Mg m-3
;
,
erization reactions. We have also demonstrated that reports on
the development of Cr-based ethylene oligomerization catalysts
must be evaluated with reference to the high intrinsic activity
of unligated Cr for Schulz-Flory oligomerization and a need
to establish genuine ligand effects on catalysis.
Mo KR radiation (confocal optic, λ ) 0.71073 Å), µ ) 1.378 mm-1
T ) 93(2) K, 11 742 data (2887 unique, Rint ) 0.0346, 2.75° < θ
< 25.351°) were collected on a Rigaku MM007/Saturn CCD
diffractometer and were corrected for absorption. The structure was
solved by direct methods and refined by full-matrix least-squares
on F2 values of all data (G. M. Sheldrick, SHELXTL, Bruker AXS
Madison WI, 2001, version 6.1) to give wR ) {∑[w(Fo2 - Fc2)2]/
Experimental Section
2 2
∑[w(Fo ) ]}1/2 ) 0.1110, conventional R ) 0.0502 for F values of
All manipulations were carried out using standard Schlenk
techniques or in a nitrogen glovebox, using solvents purified and
dried by standard procedures. Oligomerization of ethylene was
performed as described previously.2b The preparation of ligands
1-3 has been reported previously.5a Complexes 4 and 5 were
prepared by heating a toluene solution of the ligand and CrCl3-
(thf)3 to 90 °C, as described in ref 5a. Complex 4: Anal. Calcd for
C27H27P2NCrCl3 (found): C, 55.36 (55.19); H, 4.64 (4.87); N, 2.39
(2.29). Complex 5: Anal. Calcd. for C9H23P2NCrCl3 (found): C,
29.57 (29.62); H, 6.34 (6.45); N, 3.83 (3.87). µeff: 3.86 µB/Cr (5.46/
dimer). Complex 6 was prepared from CrCl2Me(thf)3 and ligand 1
in toluene: Anal. Calcd. for C35H36P2ONCrCl2 (found): C, 62.60
(62.37); H, 5.40 (5.23); N, 2.08 (1.95). µeff: 3.59 µB.
2
2
reflections with Fo > 2σ(Fo ) [2486 observed reflections], S )
1.099 for 191 parameters. Disorder in the ethyl arms was resolved
into different orientations of 50% occupancy, except for the C4/
C5 arm, which was not disordered. All P-CH2 distances were
treated with a restraint. Residual electron density extremes were
0.517 and -0.388 e Å-3
.
Acknowledgment. The authors thank the Sasol Olefin
Transformations group for helpful discussions.
Supporting Information Available: Crystallographic data for
5, in CIF format. This material is available free of charge via the
Internet at http://pubs.acs.org.
X-ray Crystallography of 5. C18H46N2P4-Cl6Cr2. M ) 731.15,
blue platelet, crystal size 0.10 × 0.05 × 0.01 mm, monoclinic, P21/
n, a ) 10.443(4) Å, b ) 12.753(3) Å, c ) 12.657(5) Å, â )
OM060906Z