Cocatalyst influence in selective oligomerization: Effect on activity, catalyst stability, and 1-hexene/1-octene selectivity in the ethylene trimerization and tetramerization reaction

Article abstract of DOI:10.1021/om070029c

The trimerization and tetramerization of ethylene to 1-hexene and 1-octene with a Cr/PNP/AlEt₃ catalyst system, in combination with a variety of cocatalysts, has been investigated. The cocatalysts B(C₆F ₅)₃ (1), Al(OC₆F₅)₃ (2), [(Et₂O)₂H][Al(OC₆F₅)₄] (3), [Ph₃C][Ta(OC₆F₅)₆] (4), (Et₂O)Al{OCH(C₆F₅)₂}₃ (5), (Et₂O)-Al{OC(CF₃)₃}₃ (6), [Ph₃C][Al{OC(CF₃)₃}₄] (7), [Ph ₃C][AlF{OC(CF₃)₃}₃] (8), [Ph ₃C][{(F₃C)₃CO}₃Al-F-A1{OC(CF ₃)₃}₃] (9), and [Ph₃C][CB] ₁₁H₆Br₆] (10) have been evaluated. The relative selectivity to 1-hexene and 1-octene obtained shows a strong dependence on the nature of the cocatalyst, and a range of selectivities from <5% C₈ (90% C₆) to 72% C₈ have been observed. The stability of several cocatalysts toward AlEt₃ has been studied, and the poor performance of 1 and 2 is linked to degradation of the cocatalyst through ethyl group exchange with AlEt₃. In contrast, the [Al{OC(CF ₃)₃}₄]^- anion in 7 is much more stable and gives rise to a highly active and longer lived catalyst. The overall productivity and selectivity of the catalyst is dependent upon both cocatalyst stability and the nature of the anion present, and a reason for this effect has been suggested. Selectivity control by the cocatalyst has been ascribed to interaction of the anion with the active Cr center.

Full text of DOI:10.1021/om070029c

Organometallics 2007, 26, 2561-2569
2561
Cocatalyst Influence in Selective Oligomerization: Effect on
Activity, Catalyst Stability, and 1-Hexene/1-Octene Selectivity in the
Ethylene Trimerization and Tetramerization Reaction
David S. McGuinness,*,†,§ Adam J. Rucklidge,†,‡ Robert P. Tooze,*,† and
Alexandra M. Z. Slawin|
Sasol Technology UK Ltd., Purdie Building, North Haugh, St. Andrews KY16 9ST, U.K., and School of
Chemistry, UniVersity of St. Andrews, Purdie Building, North Haugh, St. Andrews KY16 9ST, U.K.
ReceiVed January 11, 2007
The trimerization and tetramerization of ethylene to 1-hexene and 1-octene with a Cr/PNP/AlEt3 catalyst
system, in combination with a variety of cocatalysts, has been investigated. The cocatalysts B(C6F5)3 (1),
Al(OC6F5)3 (2), [(Et2O)2H][Al(OC6F5)4] (3), [Ph3C][Ta(OC6F5)6] (4), (Et2O)Al{OCH(C6F5)2}3 (5), (Et2O)-
Al{OC(CF3)3}3 (6), [Ph3C][Al{OC(CF3)3}4] (7), [Ph3C][AlF{OC(CF3)3}3] (8), [Ph3C][{(F3C)3CO}3Al-
F-Al{OC(CF3)3}3] (9), and [Ph3C][CB11H6Br6] (10) have been evaluated. The relative selectivity to
1-hexene and 1-octene obtained shows a strong dependence on the nature of the cocatalyst, and a range
of selectivities from <5% C8 (90% C6) to 72% C8 have been observed. The stability of several cocatalysts
toward AlEt3 has been studied, and the poor performance of 1 and 2 is linked to degradation of the
cocatalyst through ethyl group exchange with AlEt3. In contrast, the [Al{OC(CF3)3}4]- anion in 7 is
much more stable and gives rise to a highly active and longer lived catalyst. The overall productivity and
selectivity of the catalyst is dependent upon both cocatalyst stability and the nature of the anion present,
and a reason for this effect has been suggested. Selectivity control by the cocatalyst has been ascribed
to interaction of the anion with the active Cr center.
the tetramerization of ethylene to 1-octene with Cr complexes
1. Introduction
of PNP ligands I.18-21 These routes largely avoid the production
The commercial oligomerization of ethylene is predominately
carried out using transition-metal catalysts that produce a broad
distribution of linear R-olefins (LAOs). Such distributions do
not closely match present and future market demand, and as
such the development of more selective routes to desired LAOs
is currently of industrial and academic interest. In particular,
1-hexene and 1-octene are in high demand, due to their use as
comonomers for polyethylene production (LLDPE). This has
led to considerable interest in catalysts for the trimerization of
ethylene to 1-hexene (Cr,1-13 Ti,14-16 Ta17), and more recently
of undesirable olefins that conventional full-range oligomer-
ization processes produce, and as such, the first commercial
trimerization process has recently been started.22
(10) Bluhm, M. E.; Walter, O.; Do¨ring, M. J. Organomet. Chem. 2005,
690, 713.
(11) Blann, K.; Bollmann, A.; Dixon, J. T.; Hess, F.; Killian, E.;
Maumela, H.; Morgan, D. H.; Neveling, A.; Otto, S.; Overett, M. Chem.
Commun. 2005, 620.
(12) McGuinness, D. S.; Wasserscheid, P.; Morgan, D. H.; Dixon, J. T.
Organometallics 2005, 24, 552.
(13) McGuinness, D. S.; Brown, D. B.; Tooze, R. P.; Hess, F. M.; Dixon,
J. T.; Slawin, A. M. Z. Organometallics 2006, 25, 3605.
(14) Deckers, P. J. W.; Hessen, B.; Teuben, J. H. Angew. Chem., Int.
Ed. 2001, 40, 2516.
(15) Huang, J.; Wu, T.; Qian, Y. Chem. Commun. 2003, 2816.
(16) Hagen, H.; Kretschmer, W. P.; van Buren, F. R.; Hessen, B.; van
Oeffelen, D. A. J. Mol. Catal. A 2006, 248, 237.
(17) Andes, C.; Harkins, S. B.; Murtuza, S.; Oyler, K.; Sen, A. J. Am.
Chem. Soc. 2001, 123, 7423.
* To whom correspondence should be addressed. E-mail:
bob.tooze@uk.sasol.com.
† Sasol Technology UK Ltd.
‡ Present address: Scientific and Medical Products Ltd., Shirley House,
12 Gatley Rd, Cheadle, Cheshire, SK8 1PY. E-mail: adam@scimed.co.uk.
§ Present address: School of Chemistry, University of Tasmania,
Private Bag 75, Hobart, Tasmania 7001, Australia. E-mail:
david.mcguinness@utas.edu.au.
| University of St. Andrews.
(1) Manyik, R. M.; Walker, W. E.; Wilson, T. P. J. Catal. 1977, 47,
197.
(2) Briggs, J. R. Chem. Commun. 1989, 674.
(3) Reagan, W. K. (Phillips Petroleum Co.) Eur. Patent EP 0417477,
1991.
(4) Wu, F.-J. (Amoco Corp.) U.S. Patent 5811618, 1998.
(5) Carter, A.; Cohen, S. A.; Cooley, N. A.; Murphy, A.; Scutt, J.; Wass,
D. F. Chem. Commun. 2002, 858.
(18) 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.
(6) Morgan, D. H.; Schwikkard, S. L.; Dixon, J. T.; Nair, J. J.; Hunter,
R. AdV. Synth. Catal. 2003, 345, 1.
(7) Mahomed, H.; Bollmann, A.; Dixon, J. T.; Gokul, V.; Griesel, L.;
Grove, C.; Hess, F.; Maumela, H.; Pepler, L. Appl. Catal. A 2003, 255,
355.
(19) Overett, M. J.; Blann, K.; Bollmann, A.; Dixon, J. T.; Hess, F. M.;
Killian, E.; Maumela, H.; Morgan, D. H.; Neveling, A.; Otto, S. Chem.
Commun. 2005, 622.
(20) Walsh, R.; Morgan, D. H.; Bollmann, A.; Dixon, J. T. Appl. Catal.
A 2006, 306, 184.
(8) McGuinness, D. S.; Wasserscheid, P.; Keim, W.; Hu, C.; Englert,
U.; Dixon, J. T.; Grove, C. Chem. Commun. 2003, 334.
(9) McGuinness, D. S.; Wasserscheid, P.; Keim, W.; Morgan, D. H.;
Dixon, J. T.; Bollmann, A.; Maumela, H.; Hess, F.; Englert, U. J. Am. Chem.
Soc. 2003, 125, 5272.
(21) Kuhlmann, S.; Dixon, J. T.; Haumann, M.; Morgan, D. H.; Ofili,
J.; Spuhl, O.; Taccardi, N.; Wasserscheid, P. AdV. Synth. Catal. 2006, 348,
1200.
(22) Dixon, J. T.; Green, M. J.; Hess, F. M.; Morgan, D. H. J. Organomet.
Chem. 2004, 689, 3641.
10.1021/om070029c CCC: $37.00 © 2007 American Chemical Society
Publication on Web 04/10/2007

2562 Organometallics, Vol. 26, No. 10, 2007
McGuinness et al.
required to effect this selectivity shift,19 and intermediate
selectivity is obtained with the pendant alkoxy substitution on
the nitrogen of I.35 Hemilabile coordination of the ether group
has been demonstrated25 and seems to be responsible for the
selectivity switch. A coordinating donor group is not essential,
however, and it has been shown that the introduction of steric
bulk (alkyl groups) in the ortho position leads to increased
1-hexene selectivity at the expense of 1-octene.11 In this case,
the relative ratio of 1-hexene to 1-octene was found to be
dependent on the total amount of steric bulk, as measured by
the number of ortho substituents and the nature of the N
substituent. Furthermore, we have recently demonstrated that
expansion of the chelate ring size in related P∧P-Cr complexes
also leads to a change in C6/C8 selectivity.36 A major advantage
of this catalyst class is thus apparent: a high degree of control
over relative 1-hexene and 1-octene selectivities is available by
careful ligand modification.
Scheme 1. Proposed Mechanism of Ethylene Trimerization
and Tetramerization
The mechanism of ethylene trimerization and tetramerization
is generally thought to follow a metallacyclic route, involving
oxidative addition of two ethylene molecules to the metal
followed by insertion of one (trimerization) or two (tetramer-
ization) further ethylene units to yield higher metallacycles.
Reductive â-hydride transfer can then release the R-olefin and
regenerate the active metal species (Scheme 1). Support for this
mechanism comes from a number of experimental23-26 and
theoretical27-31 investigations. The key to the selectivity of these
systems appears to be the energetically preferred tendency of
these metallacycles to undergo 1-hexene or 1-octene eliminations
rather than further ethylene insertions. This is clearly ligand
dependent, and recent studies have shown that further ethylene
insertion is possible and can lead to “runaway” metallacycle
growth and concomitant unselective Schulz-Flory oligomer-
ization.32,33
The importance of ancillary ligand influence in this mecha-
nism has been demonstrated in a number of studies. For instance,
the pendant arene group in Ti trimerization systems is thought
to moderate its coordination strength throughout the catalytic
cycle,34 while the pyrrolyl ligand of the Phillips catalyst does
likewise by undergoing haptotropic shifts between η1 and η5
coordination.30 Ligand effects are perhaps most pronounced in
trimerization and tetramerization catalysts based on Cr com-
plexes of PNP ligands I. The activity and particularly selectivity
of these systems show a strong dependency on the steric and
coordinative properties of the ligand. The inclusion of o-OMe
groups on arylphosphino derivatives of I leads to a catalyst that
is highly selective to 1-hexene.5 In the absence of ortho
substitution on the aryl group, or with an alkylphosphino ligand
structure, the selectivity shifts to predominately 1-octene.18 It
was subsequently shown that only a single o-OMe group is
The available evidence to date is suggestive of formally
cationic active species in Cr-I-catalyzed trimerization and
tetramerization24,37,38 and trimerization with other systems.13,39,40
This is certainly consistent with the cocatalyst most often
employed being methaluminoxane. The use of MAO as a
cocatalyst in olefin oligomerization and polymerization is
normally thought to implicate a cationic active metal center.41
While MAO is relatively poorly defined, a great deal of work
has been carried out on more well-defined polymerization
systems based upon cation-anion pairs. The most common
systems here incorporate the ubiquitous (perfluoroaryl)borate
anions, which have allowed thorough characterization and
isolation of active polymerization systems.41 These studies have
also highlighted the pronounced effects that the counterion in
these systems can have on catalyst stability, activity, and
stereoselectivity, and it is now realized that the nature of the
metal-anion interaction is a decisive factor in catalyst
performance.42-51 To date, there have been very few investiga-
tions into such cocatalyst effects in selective ethylene oligo-
merization catalysts.13,40
(35) Elowe, P. R.; McCann, C.; Pringle, P. G.; Spitzmesser, S. K.;
Bercaw, J. E. Organometallics 2006, 25, 5255.
(36) Overett, M.; Blann, K.; Bollmann, A.; de Villiers, R.; Dixon, J. T.;
Killian, E.; Maumela, C.; Maumela, H.; McGuinness, D. S.; Morgan, D.
H.; Rucklidge, A.; Slawin, A. M. Z. Submitted for publication.
(37) Schofer, S. J.; Day, M. W.; Henling, L. M.; Labinger, J. A.; Bercaw,
J. E. Organometallics 2006, 25, 2743.
(38) Jabri, A.; Crewdson, P.; Gambarotta, S.; Korobkov, I.; Duchateau,
R. Organometallics 2006, 25, 715.
(39) Ko¨hn, R. D.; Haufe, M.; Kociok-Ko¨hn, G.; Grimm, S.; Wassersc-
heid, P.; Keim, W. Angew. Chem., Int. Ed. 2000, 39, 4337.
(40) Ko¨hn, R. D.; Smith, D.; Mahon, M. F.; Prinz, M.; Mihan, S.; Kociok-
Ko¨hn, G. J. Organomet. Chem. 2003, 683, 200.
(23) Emrich, R.; Heinemann, O.; Jolly, P. W.; Kru¨ger, C.; Verhovnik,
G. P. J. Organometallics 1997, 16, 1511.
(24) Agapie, T.; Schofer, S. J.; Labinger, J. A.; Bercaw, J. E. J. Am.
Chem. Soc. 2004, 126, 1304.
(25) Agapie, T.; Day, M. W.; Henling, L. M.; Labinger, J. A.; Bercaw,
J. E. Organometallics 2006, 25, 2733.
(26) Overett, M. J.; Blann, K.; Bollmann, A.; Dixon, J. T.; Haasbroek,
D.; Killian, E.; Maumela, H.; McGuinness, D. S.; Morgan, D. H. J. Am.
Chem. Soc. 2005, 127, 10723.
(41) Marks, T. J.; Chen, E. Y.-X. Chem. ReV. 2000, 100, 1391.
(42) Chen, Y.-X.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1997,
119, 2582.
(43) Deck, P. A.; Beswick, C. L.; Marks, T. J. J. Am. Chem. Soc. 1998,
120, 1772.
(44) Chen, E. Y.-X.; Kruper, W. J.; Roof, G.; Wilson, D. R. J. Am. Chem.
Soc. 2001, 123, 745.
(27) Blok, A. N. J.; Budzelaar, P. H. M.; Gal, A. W. Organometallics
2003, 22, 2564.
(28) de Bruin, T. J. M.; Magna, L.; Raybaud, P.; Toulhoat, H.
Organometallics 2003, 22, 3404.
(29) Tobisch, S.; Ziegler, T. Organometallics 2003, 22, 5392.
(30) van Rensburg, W. J.; Grove, C.; Steynberg, J. P.; Stark, K. B.;
Huyser, J. J.; Steynberg, P. J. Organometallics 2004, 23, 1207.
(31) Yu, Z.-X.; Houk, K. N. Angew. Chem., Int. Ed. 2003, 42, 808.
(32) Tomov, A. K.; Chirinos, J. J.; Jones, D. J.; Long, R. J.; Gibson, V.
C. J. Am. Chem. Soc. 2005, 127, 10166.
(45) Beck, S.; Lieber, S.; Schaper, F.; Geyer, A.; Brintzinger, H.-H. J.
Am. Chem. Soc. 2001, 123, 1483.
(46) Zhou, J.; Lancaster, S. J.; Walker, D. A.; Beck, S.; Thornton-Pett,
M.; Bochmann, M. J. Am. Chem. Soc. 2001, 123, 223.
(47) Song, F.; Cannon, R. D.; Bochmann, M. Chem. Commun. 2004,
542.
(48) Landis, C. R.; Rosaaen, K. A.; Uddin, J. J. Am. Chem. Soc. 2002,
124, 12062.
(49) Landis, C. R.; Rosaaen, K. A.; Sillars, D. R. J. Am. Chem. Soc.
2003, 125, 1710.
(33) Tomov, A. K.; Chirinos, J. J.; Long, R. J.; Gibson, V. C.; Elsegood,
M. R. J. J. Am. Chem. Soc. 2006, 128, 7704.
(34) Deckers, P. J. W.; Hessen, B.; Teuben, J. H. Organometallics 2002,
21, 5122.
(50) Busico, V.; Castelli, V. V. A.; Aprea, P.; Cipullo, R.; Segre, A.;
Talarico, G.; Vacatello, M. J. Am. Chem. Soc. 2003, 125, 5451.
(51) Al-Humydi, A.; Garrison, J. C.; Youngs, W. J.; Collins, S.
Organometallics 2005, 24, 193.

Ethylene Trimerization and Tetramerization
Organometallics, Vol. 26, No. 10, 2007 2563
Chart 1. Cocatalysts Investigated for Ethylene Trimerization and Tetramerization
Given the dramatic selectivity changes that ligand modifica-
tion can provide, a detailed investigation into alternative
cocatalysts for ethylene trimerization and tetramerization seems
justified. Such an investigation is not just of interest due to
activity and selectivity effects that might arise. The use of excess
MAO has economic implications when it comes to the large-
scale use of such technology. The development of well-defined,
stoichiometric cocatalysts for trimerization and tetramerization,
in analogy to the work that has been done on polymerization
catalysts, is therefore of interest.
For instance, the [Al{OC(CF3)3}4]- anion is estimated to be
one of the most weakly coordinating anions known and contains
a strong Al-O bond that is resistant to alkoxide ion abstraction.
This anion is even reported to be stable in aqueous HNO3.61
The halogenated carborane anion [CB11H6Br6]- was also
investigated, due to the reported robustness of this class of
anions.62 In particular, their stability toward alkylaluminum
reagents63 made them of interest in the context of this study.
The cocatalysts employed in this study are shown in Chart
1. These cocatalysts, which are based predominately upon
fluorinated aluminates, were chosen to contain a range of anion
coordinating strengths, steric protection, and, as will be shown,
cocatalyst stability. Both neutral Lewis acidic species (in situ
anion formation) and also preformed anions incorporating an
alkyl abstracting agent (Ph3C+, H+) are represented. The anions
[Al(OC6F5)4]- and [Ta(OC6F5)6]-,58 [Al{OC(CF3)3}4]-,60 [AlF-
{OC(CF3)3}3]- and [{(CF3)3CO}3Al-F-Al{OC(CF3)3}3]-,64
and [CB11H6Br6]- 65 have been reported previously.
The preparation of Al(OC6F5)3 (2) proceeded smoothly
through addition of an excess of perfluorophenol to AlEt3 in
toluene. A single-crystal X-ray analysis reveals a dimeric
structure in the solid state, in which each Al center is tetrahedral
(Figure 1). In CDCl3 solution compound 2 is only sparingly
soluble and the 19F NMR spectrum reveals many broad and
overlapping signals, possibly due to the dynamic formation of
higher oligomers. The compound is more soluble in DMSO-d6
and displays sharp signals (see Experimental Section), probably
due to formation of a monomeric adduct, (D3C)2SO-Al-
We recently communicated first results in this regard, where-
by a number of Cr-I catalysts were tested under conditions of
activation with B(C6F5)3/AlR3 and [Ph3C][B(C6F5)4]/AlR3.52
This work showed that while liquid fraction selectivities were
similar to MAO activation, the catalyst rapidly deactivated and
a variable, generally high, amount of polyethylene formation
was observed. The rate of deactivation in these systems was
found to increase with the amount of AlEt3 alkylating agent
employed. This led us to suggest degradation of the borane/
borate through alkyl exchange with AlEt3 as the source of cata-
lyst deactivation. Such a deactivation process has been estab-
lished for metallocene/[Ph3C][B(C6F5)4]/AlR3-based polymer-
ization systems,53 and ligand exchange between AlMe3 and
B(C6F5)3 has also been demonstrated.54 If indeed this form of
deactivation is occurring for trimerization and tetramerization
catalyst systems, it implies that a more robust cocatalyst anion
is required in order that highly productive alternatives to MAO
activation might be devised. Confirmation of such a deactivation
process and the development of more stable cocatalysts are
therefore of interest. Herein we report detailed studies into the
effect of varying the cocatalyst stability and anion coordination
strength and the dramatic influence this has on catalyst activity
and selectivity. As a result of these studies, we have developed
highly active tetramerization catalysts without the requirement
for MAO activation and have shown for the first time that the
relative 1-hexene and 1-octene selectivities can be controlled
by cocatalyst modification.55
(52) McGuinness, D. S.; Overett, M.; Tooze, R. P.; Blann, K.; Dixon, J.
T.; Slawin, A. M. Z. Organometallics 2007, 26, 1108.
(53) Bochmann, M.; Sarsfield, M. J. Organometallics 1998, 17, 5908.
(54) Klosin, J.; Roof, G. R.; Chen, E. Y.-X.; Abboud, K. A. Organo-
metallics 2000, 19, 4684.
(55) McGuinness, D. S.; Rucklidge, A. J. (Sasol Technology UK) IPN
PCT/IB2006/0053494, 2006
(56) Krossing, I.; Raabe, I. Angew. Chem., Int. Ed. 2004, 43, 2066.
(57) Krossing, I.; Reisinger, A. Coord. Chem. ReV. 2006, 250, 2721.
(58) Sun, Y.; Metz, M. V.; Stern, C. L.; Marks, T. J. Organometallics
2000, 19, 1625.
(59) Metz, M. V.; Sun, Y.; Stern, C. L.; Marks, T. J. Organometallics
2002, 21, 3691.
(60) Krossing, I.; Brands, H.; Feuerhake, R.; Koenig, S. J. Fluorine Chem.
2001, 112, 83.
(61) Krossing, I. Chem. Eur. J. 2001, 7, 490.
(62) Reed, C. A. Acc. Chem. Res. 1998, 31, 133.
(63) Kim, K.-C.; Reed, C. A.; Long, G. S.; Sen, A. J. Am. Chem. Soc.
2002, 124, 7662.
(64) Krossing, I.; Gonsior, M.; Mueller, L. (University of Karlsruhe) WO
2005/054254, 2005.
(65) Xie, Z.; Jel´ınek, T.; Bau, R.; Reed, C. A. J. Am. Chem. Soc. 1994,
116, 1907.
2. Results and Discussion
2.1. Cocatalysts. In searching for alternative cocatalyst anions
to MAO and perfluoroborates, we became interested in fluori-
nated alkoxymetalate- and aryloxymetalate-based weakly co-
ordinating anions.56,57 Marks has reported activation of metal-
locene polymerization catalysts with [Ph3C][Al(OC6F5)4] and
[Ph3C][Ta(OC6F5)6] and investigated the effect of the anion upon
catalysis.58,59 In particular, we were drawn by reports of
Krossing that bulky fluorinated alkoxy groups on aluminum give
rise to very weakly coordinating and chemically robust anions.60

2564 Organometallics, Vol. 26, No. 10, 2007
McGuinness et al.
Figure 1. Molecular structure of 2. Selected bond lengths (Å) and
angles (deg): Al(1)-O(7), 1.6691(14); Al(1)-O(13), 1.6855(13);
Al(1)-O(1), 1.8424(14); Al(1)-O(1A), 1.8543(14); O(7)-Al(1)-
O(13), 118.68(7); O(7)-Al(1)-O(1), 118.85(7); O(1)-Al(1)-
O(1A), 78.43(7).
Figure 2. Molecular structure of 5. Selected bond lengths (Å) and
angles (deg): Al(1)-O(1), 1.681(5); Al(1)-O(14), 1.712(4);
Al(1)-O(27), 1.699(4); Al(1)-O(40), 1.849(4); O(1)-Al(1)-
O(14), 121.0(3); O(1)-Al(1)-O(27), 111.7(3); O(1)-Al(1)-O(40),
102.6(2).
(OC6F5)3. In contrast, the boron analogue, B(OC6F5)3, is
monomeric both in the solid state and in solution.66 Compound
2 is designated by its empirical formula throughout this paper,
such that 1 equiv of 2 corresponds to 1 equiv of Al in the
catalytic experiments. The solid-state bimetallic system has an
Al‚‚‚Al distance of 2.86 Å. The Al2O2 core is planar, as a
consequence of crystallographic symmetry, while the bridging
OC6F5 groups are essentially coplanar with this ring. The
terminal O-Al-O angles (118.68(7) and 118.85(7)°) are almost
trigonal, while the internal O-Al-O angle is very acute (78.43-
(7)°) in the four-membered ring. The terminal Al-O bonds
(1.6691(14) and 1.6855(13) Å) are somewhat shorter than in
anionic aluminates containing fluorinated alkoxide ligands
(1.70-1.76 Å),60,61 while the bridging Al-O bonds are longer
at 1.8424(14) and 1.8543(14) Å. The same reaction, when
carried out in ether, afforded [(Et2O)2H][Al(OC6F5)4] (3) and
as such represents a convenient route to an aluminate analogue
of Brookhart’s acid [(Et2O)2H][B{C6H3(CF3)2}4].67 Unfortu-
nately, this reaction does not seem to be general, and when the
more bulky alcohols HOC(CF3)3 and HOC(H)(C6F5)2 were
employed, the neutral etherate complexes Al(ORF)3(OEt2) (5
and 6) resulted. Slow crystallization of 5 from petroleum spirits
led to two sets of crystals. Analysis of the minor product showed
it to be the ether-free dimer Al2{OCH(C6F5)2}6, which is
comparable in structure to 2 (see the Supporting Information).
Evidently, some loss of coordinated ether occurs during crystal-
lization, resulting in formation of the dimer. The major set of
crystals corresponds to the tetrahedral ether adduct, as shown
in Figure 2. The Al-alkoxide bonds range in length from 1.681-
(5) to 1.730(4) Å, while the Al-Oether bond is 1.849(4) Å in
length. Compound 6 proved extremely sensitive, and it was not
possible to obtain a meaningful or reproducible elemental
analysis. After a number of attempts, a crystal structure could
be obtained, which is shown in Figure 3. The tetrahedral Al
center of 6 is comparable to that of 5, with Al-alkoxide lengths
of 1.719(9)-1.731(8) Å and an Al-Oether length of 1.856(9)
Å.
Figure 3. Molecular structure of 6. Selected bond lengths (Å) and
angles (deg): Al(1)-O(5), 1.728(10); Al(1)-O(1), 1.731(8);
Al(1)-O(9), 1.719(9); Al(1)-O(13), 1.856(9); O(1)-Al(1)-O(9),
115.9(4); O(1)-Al(1)-O(5), 110.5(4); O(1)-Al(1)-O(13), 103.5(4).
system composed of CrCl3(thf)3 and PNP ligand II in a 1:1.2
ratio. In a typical experiment, a toluene solution of Cr-II was
treated with AlEt3 and stirred for 5 min before addition to the
reactor. Immediately thereafter, a solution of the chosen
cocatalyst, 1-10, was injected and the reactor pressurized with
ethylene. The majority of experiments were carried out at 45
°C. To compare relative C6 and C8 selectivities between
catalysts, the value of the molar ratio of C6/C8 has been reported
for each experiment. The results of these oligomerization runs
are presented in Table 1.
2.2. Ethylene Trimerization and Tetramerization. Through-
out this study, we have employed an in situ formed catalyst
(66) Britovsek, G. J. P.; Ugolotti, J.; White, A. J. P. Organometallics
2005, 24, 1685.
(67) Brookhart, M.; Grant, B.; Volpe, A. F. Organometallics 1992, 11,
3920.
For validation of this experimental setup, an ethylene tet-
ramerization run was conducted with MMAO-3A as activator
at 65 °C (entry 1), confirming that the expected selectivity to

Ethylene Trimerization and Tetramerization
Organometallics, Vol. 26, No. 10, 2007 2565
Table 1. Ethylene Trimerization and Tetramerization with CrCl3(thf)3/II/AlEt3/Cocatalysta
cocat.
(amt (equiv))
amt of Cr
(µmol)
AlR3
(amt (equiv))
T
(°C)
P
(bar)
run time
(min)
PE
(%)
C6 (1-C6)
(%)
C8 (1-C8)
(%)
C6/C8
(molar)
entry
prodb
1
2c
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
MMAO-3A (800)
MMAO-3A (300)
1 (1.3)
1 (1.3)
2 (1.3)
2 (1.3)
2 (1.3)
HOC6F5 (3.0)
3 (1.5)
4 (2.0)
5 (1.5)
6 (1.7)
7 (1.5)
7 (1.5)
7 (1.5)
7 (1.5)
8 (1.5)
9 (1.0)
2.5
20
10
10
30
30
30
30
10
10
10
20
10
10
10
10
10
10
10
10
10
10
65
45
45
45
45
45
45
45
45
45
45
45
45
65
45
45
45
45
45
45
45
45
40
45
50
50
40
40
40
40
40
40
35
40
40
40
40
40
40
35
40
40
35
40
60
30
30
30
30
30
30
30
30
30
30
30
60
60
60
30
60
30
30
30
60
30
1.6
0.1
2.0
25.2
2.3
8.2
23.3 (81.2)
16.0 (66.4)
28.0 (73.0)
21.4 (73.3)
89.9 (95.6)
85.2 (98.5)
86.0 (97.8)
78.3 (97.9)
63.9 (95.0)
46.6 (89.7)
43.8 (86.7)
87.3 (97.0)
16.3 (68.9)
25.0 (82.5)
16.9 (66.3)
28.2 (66.3)
34.6 (80.8)
57.8 (94.7)
44.0 (95.9)
27.9 (85.9)
18.6 (76.3)
22.1 (84.8)
64.4 (99.2)
70.7 (97.9)
67.0 (99.0)
47.6 (96.9)
4.7 (74.4)
3.9 (94.6)
4.0 (96.8)
1.8 (84.4)
23.5 (97.6)
39.6 (98.5)
28.3 (98.2)
6.1 (88.5)
72.2 (99.0)
66.6 (99.2)
72.7 (98.0)
60.1 (96.7)
32.1 (87.7)
30.3 (95.4)
25.5 (98.2)
56.0 (99.2)
68.2 (99.4)
32.0 (97.2)
0.48
0.30
0.56
0.60
25
29
29
58
3.6
1.6
2.1
19
0.30
0.50
0.31
0.62
1.4
550 810
285 100
7 110
587
2 340
5 810
997
3 410
1 680
1 330
760
4 100
125 600
59 560
57 250
1 440
2 280
6 925
12 810
13 730
53 120
1 090
AlEt3 (50)
AlEt3 (100)
AlEt3 (300)
AlEt3 (30)
AlBui3 (100)
AlEt3 (30)
6.4
16.0
12.6
13.7
27.8
1.0
0.41
0.6
1.3
AlEt3 (100)
AlEt3 (100)
AlEt3 (100)
AlEt3 (100)
AlEt3 (100)
AlEt3 (100)
AlEt3 (300)
AlEt3 (600)
AlEt3 (100)
AlEt3 (100)
AlEt3 (100)
AlEt3 (100)
AlEt3 (100)
AlEt3 (100)
3.3
33.3
10.0
21.6
7.9
6.5
45.9
2.5
2.3
0.66
0.36
0.92
9 (1.5)
9 (2.0)
9 (3.0)
10 (1.5)
a Conditions: CrCl3(thf)3/II (1:1.2), toluene (100 mL). b Productivity: g of product/g of Cr. c From ref 18.
Scheme 2
1-hexene and 1-octene is obtained (C6/C8 ) 0.48). A high
productivity was obtained over the 1 h run, although much
higher catalyst activities are achievable with this system under
optimized conditions.36 Entry 2, from previous work,18 shows
that the selectivity obtained at 45 °C is somewhat different, with
more C8 formed at the expense of C6 (C6/C8 ) 0.30).
Full results of activation with perfluoroborane and borate
cocatalysts have been reported;52 herein two catalytic runs with
B(C6F5)3 (1) are included for reference (entries 3 and 4). These
show that a much reduced productivity is observed that is
dependent upon the amount of AlEt3 alkylating agent employed.
While the initial activity is quite high with this system, as judged
by ethylene uptake, a rapid catalyst deactivation occurs early
in the run. This appears to be accelerated by increased AlEt3
loadings beyond 50 equiv. A somewhat higher C6/C8 value is
observed compared to MAO activation (cf. 0.56 for 1 to 0.30
for MMAO-3A at 45 °C), although this was not recognized as
significant previously. In light of the results presented below,
it now seems probable that it is.
The first aluminate cocatalyst to be prepared and tested was
Al(OC6F5)3 (2). To our surprise, the use of this cocatalyst lead
to a remarkable shift in selectivity toward 1-hexene, such that
the C8 content in each case was below 5% (entries 5-7). Within
the C6 fraction, the purity of 1-hexene is high (95.6-98.5%),
and very little cyclic C6 (methylcyclopentane, methylene cy-
clopentane) is formed, in contrast to a normal tetramerization
run with this ligand.26 Overall, the product selectivity is much
like that obtained with a trimerization PNP ligand containing
o-OMe substitution5,19 or ortho steric bulk11 and has not
previously been observed with ligand II. In common with
B(C6F5)3 activation, a rather low productivity is obtained that
seems to stem at least in part from rapid catalyst deactivation
over 5-10 min, this again being dependent upon the amount
of AlEt3 employed. A relatively high amount of polyethylene
is also formed in comparison to activation with MMAO-3A.
There are a number of possible reasons for this dramatic
selectivity shift toward trimerization. A trimerization system
based on aryloxyaluminum compounds together with a Cr source
has been reported previously and probably operates via forma-
tion of an aryloxychromium active species.6 On the basis of
this, it occurred to us that the trimerization activity observed
may be due to a Cr-OC6F5 species, formed via ligand exchange
between Cr and 2. To test for this, a catalytic run was conducted
with CrCl3(thf)3 (30 µmol), HOC6F5 (90 µmol), and AlEt3 (30
equiv), in the absence of ligand II. Although a productivity of
3710 g/g of Cr was obtained, the product was composed of 95%
polyethylene with no selectivity in the liquid fraction. This
clearly shows that the PNP ligand II is essential to trimerization
activity. This is not the say that [OC6F5]- transfer to Cr cannot
or does not occur (see below), but it does not appear to be
responsible for trimerization selectivity. A more likely reason,
it seems, is that activation with 2 gives rise to a more
coordinating anion. The anion or anion-like species initially
formed presumably is [AlEt(OC6F5)3]-. This could coordinate
with the formally cationic Cr center, either via bridging alkyl
coordination, as has been established for [B(C6F5)3R]-, or
through a bridging oxygen donor. Such coordination may mimic
the hemilabile coordination of o-OMe substitution in aryl PNP
ligands, which is known to lead to increased trimerization
selectivity (Scheme 2).
The cocatalyst could also be prepared in situ, by addition of
HOC6F5 to the reactor in place of 2 (entry 8). In this case,
trisubstitution of the aluminum is unlikely, the initial product
prior to reaction with Cr probably being AlEt2(OC6F5). This
increased the C6/C8 value to 58, at the same time producing a
higher amount of polyethylene (16%).
On the basis of the premise that anion coordination strength
can control C6/C8 selectivity, the preformed anionic activator
[H(OEt2)2][Al(OC6F5)4] (3) was expected to lead to increased
tetramerization. This was found to be the case (entry 9), and
23.5% C8 resulted from activation with this cocatalyst (C6/C8
) 3.6). The hexacoordinated Ta aryloxide 4 should be even
less coordinating than 3, and consistent with this, the selectivity
to C8 increased to 39.6% (entry 10; C6/C8 ) 1.6). With both of

2566 Organometallics, Vol. 26, No. 10, 2007
McGuinness et al.
these cocatalysts, however, the same catalyst deactivation as
observed previously was found to occur, and polymer production
was high. Nonetheless, these results confirm the pronounced
selectivity influence that the cocatalyst can have on this system,
which seems most likely to arise from the differing coordination
strengths of the anions.
The performance of cocatalysts 5 and 6 is not easy to predict,
due to coordination of ether to these compounds, which would
buffer the Lewis acidity to an extent even in the presence of
AlEt3, which can act as a scavenger. The very bulky cocatalyst
5 does give rise to a poorly active catalyst (entry 11), with a
C6/C8 ratio of 2.1, suggesting a Cr-anion interaction similar to
that obtained with 4. It is possible that the poor activity and
high relative amount of polymer formed in this case is due to
ineffectual catalyst activation (alkyl abstraction) with this
compound. The etherate 6 gave somewhat better catalyst
activation (entry 12), and in this case a low amount of polymer
was formed. The C6/C8 selectivity (19.0) with this cocatalyst is
again different from that obtained previously and lies somewhere
between those of cocatalysts 2 and 3.
The aluminate anion in cocatalyst 7, [Ph3C][Al{OC(CF3)3}4],
is reported to be one of the most weakly coordinating and
chemically robust anions known.56 As shown in entry 13, these
properties have a marked effect on catalyst activity, stability,
and selectivity when this activator is employed. The catalyst
formed displays a high activity, and this is maintained to an
extent that ethylene uptake continues, albeit slowing, over a 1
h run time. This results in an overall productivity of 125 600
g/g of Cr, which is by far the highest we have seen for a non-
alumoxane-activated tetramerization catalyst. The selectivity is
shifted back toward tetramerization, and the C6/C8 ratio (0.30)
is the same as that obtained with MMAO-3A under comparable
conditions. Importantly, the amount of polymer formed is also
much reduced compared to that for the other cocatalysts tested.
Increasing the temperature to 65 °C with MMAO-3A activation
leads to an increased C6/C8 value (entry 1); thus, a further run
was conducted at 65 °C with cocatalyst 7 (entry 14). It can be
seen that C6/C8 selectivity is almost the same in each run,
although the higher temperature leads to a faster deactivation
with the aluminate cocatalyst. Despite the high activity and
stability at 45 °C and 100 equiv of AlEt3, increasing the loading
of AlEt3 still leads to catalyst deactivation (entries 15 and 16).
This becomes pronounced with 600 equiv of AlEt3, and in this
case the relative amount of trimerization is also increased (C6/
C8 ) 0.62). Evidently, the amount of AlEt3 still has an effect
on the stability of cocatalyst 7, although this effect is much
reduced, and an optimum productivity is achieved with 100-
200 equiv of AlEt3.
Figure 4. Molar C6/C8 ratio and productivity (g/g of Cr) as a
function of amount of cocatalyst 9 (equivalents relative to Cr).
Conditions: 10 µmol of CrCl3(thf)3, 12 µmol of II, 100 AlEt3, 45
°C, 30-60 min.
21). This is more clearly illustrated in Figure 4, which shows
the C6/C8 selectivity and productivity as a function of concentra-
tion of 9 (equivalents relative to Cr). As the amount of 9 is
increased, an increase in catalyst productivity and selectivity
to C8 results. With 3 equiv of 9 the C6/C8 value approaches
that observed with MAO and 7, and a high productivity is
obtained. These observations seem to suggest an anion coor-
dination strength that is dependent upon the concentration of
the cocatalyst. A possible explanation for this lies in the ability
of [{(F3C)3CO}3Al-F-Al{OC(CF3)3}3]- to dissociate, as
shown in reaction 1. This anion is known to break down to
constituent monomers in the presence of donor solvents
-
(FAl(ORF)3 + D f Al(ORF)3),68 and as such this proposal
seems reasonable. A higher concentration of 9 would increase
the relative concentration of the less coordinating dinuclear
anion, and as such the concentration effects observed are
consistent with such a process. Additionally, there is some
precedent for this, whereby Cp2TaMe3 reacts with 1 equiv of
M(C6F5)3 (M ) B, Al) to yield a mixture of [Cp2TaMe2][CH3M-
(C6F5)3] and [Cp2TaMe2][(C6F5)3M-CH3-M(C6F5)3] in dy-
namic equilibrium.70 Addition of excess M(C6F5)3 drives the
equilibrium toward the bridged anion, and in the case of Al-
(C6F5)3 it was possible to isolate pure [Cp2TaMe2][(C6F5)3Al-
Me-Al(C6F5)3]. Such a process has also been invoked to explain
the increased polymerization activity observed when excess
M(C6F5)3 is used to activate metallocene and constrained-
geometry catalysts.44,51
The [FAl{OC(CF3)3}3]- anion64 contained in cocatalyst 8 is
reportedly very stable but somewhat more coordinating than
[Al{OC(CF3)3}4]-.68 As a result, activation with 8 did not lead
to the high productivity achieved with 7, and the product
selectivity was shifted back toward trimerization (C6/C8 ) 1.4,
entry 17).
The fluoride-bridged anion of cocatalyst 9, [{(F3C)3CO}3Al-
F-Al{OC(CF3)3}3]-, has been suggested as the least coordinat-
ing anion known.57,69 We were therefore surprised to find that
this cocatalyst gave a poor activity and more 1-hexene than
1-octene (C6/C8 ) 2.5, entry 17). It was subsequently found,
however, that both activity and selectivity with this cocatalyst
are markedly influenced by the amount employed (entries 19-
The final cocatalyst tested was [Ph3C][CB11H6Br6] (10). The
results obtained with this were disappointing (entry 22), with a
low activity and high polymer content resulting. Overall, the
results were similar to those for activation with 8. It seems
unlikely that this result is due to degradation of the carborane,
given the high stability of this class of anion, especially toward
alkylaluminum.63 In comparison to weakly coordinating anions
(68) Krossing, I. Personal communication, 2005.
(69) Bihlmeier, A.; Gonsior, M.; Raabe, I.; Trapp, N.; Krossing, I. Chem.
Eur. J. 2004, 10, 5041.
(70) Chen, E. Y.-X.; Abboud, K. A. Organometallics 2000, 19, 5541.

Ethylene Trimerization and Tetramerization
Organometallics, Vol. 26, No. 10, 2007 2567
such as [Al{OC(CF3)3}4]-, most carboranes are thought to be
more strongly coordinating to electrophilic cations.56 This is
reflected here in the C6/C8 ratio (0.92) that is obtained with
this cocatalyst. Less coordinating carborane anions, such as
[1-Me-CB11F11]-,71 may be more successful here but have not
been studied as part of this work.
2.3. Cocatalyst Degradation Studies. As outlined in the
Introduction, one of the aims of this study was to investigate
cocatalyst (anion) stability and to evaluate what effect this has
on overall catalyst stability. In previous trimerization and
tetramerization studies with B(C6F5)3 (1),13,52 and in this work,
the amount of AlR3 employed has repeatedly been found to
affect the rate of catalyst deactivation. We have therefore studied
the interaction of selected cocatalysts with excess AlEt3 by NMR
spectroscopy.
Figure 5. Molar C6/C8 ratio as a function of time for catalysis
with 3 equiv of Al(OC6F5)3 (2). Conditions: 10 µmol of CrCl3-
(thf)3, 12 µmol of II, 30 AlEt3, 45 °C, 40 bar.
Treatment of 1 with 30 equiv of AlEt3 at room temperature
in CD2Cl2 results in complete loss of the 19F signals for 1 by
the time the spectrum is run. In their place, three sets of signals
are observed, the major species (88% abundance) having peaks
at -122.0 (o-F), -154.1 (p-F), and -162.4 ppm (m-F). The
1H NMR spectrum cannot be fully interpreted due to overlapping
peaks, but previous72,73 low-temperature studies on AlEt3 (the
Al2Et6 dimer) and mixed Al-alkyl dimers aid in assigning those
new signals which are resolved. The spectrum is consistent with
reaction 2, whereby the terminal methylene protons adjacent to
The stability toward AlEt3 of the two best cocatalysts, 7 and
9, was also investigated by treating metal salts of the anions
with 30 equiv of AlEt3. Li[Al{OC(CF3)3}4] did not react with
AlEt3, either at the time of addition or after sitting overnight.
Thus, the longer catalyst lifetime with this activator can be
rationalized. Ag[{(CF3)3CO}3Al-F-Al{OC(CF3)3}3] is not
quite as robust, and some reaction does take place, as shown
by the appearance of five minor new signals in the 19F NMR
spectrum. The major species, however, remains the starting
compound. It is not possible to say from the NMR data whether
these changes are due to scrambling of the OC(CF3)3 and ethyl
groups or whether they result from an equilibrium involving
the dimer, similar to that suggested in section 2.2.
Degradation of the cocatalyst by this exchange process would
significantly reduce the Lewis acidity of the cocatalyst or
increase the coordinating ability of the anion that is formed.
This could not only reduce catalyst activity but also, on the
basis of the results of section 2.2, should affect the C6/C8
selectivity. Although the rate of degradation of 1 and 2 has not
been quantified, this process appears to occur on a time scale
approximately comparable with the early stages of a tri-/
tetramerization run. For this reason, an ethylene oligomerization
experiment was conducted in which the product was sampled
at intervals over the run and the C6/C8 selectivity monitored.
The tri-/tetramerization of ethylene was carried out at 45 °C
with 30 equiv of AlEt3 and 3 equiv of 1 relative to Cr. The
results are presented in Figure 5, which shows that indeed the
selectivity changes over time. At the start of the run, the catalyst
is predominately a tetramerization system, but it rapidly converts
to a trimerization catalyst. Shortly after activation (20 s), the
C6/C8 ratio is 0.53, which is approaching that obtained with
MAO and 7.
The selectivity to C6 reaches a maximum after 4 min, after
which it slowly reduces. This subsequent slow reduction in C6
selectivity is not expected, as it would seem to suggest formation
of a less coordinating anion in the latter stages of the reaction.
The explanation for this behavior is revealed in Figure 6, which
plots C6 and C8 production over time. While the production of
C8 is reasonably constant (the active species responsible for
tetramerization is stable once AlEt3/2 exchange approaches
equilibrium), the production of C6 starts to tail off after 4 min.
The Cr-anion pair responsible for trimerization undergoes a
more rapid deactivation.
the C6F5 group appear at 0.99 ppm (2H) and the other terminal
methylenes appear at 0.75 ppm (4H). The bridging methylene
signals likely overlap with the CH3 resonances, as found
previously with Al2Et6. The other two sets of 19F signals occur
in 5% and 7% abundance and are very similar in chemical shift
to the corresponding signals for 1. On the basis of this, it seems
likely these species are mixed ethyl-aryl boranes, B(C6F5)xEt3-x
(x ) 1, 2). These results are in line with previous studies on
the rapid exchange between stoichiometric AlMe3 amd 1.54 They
confirm that rapid alkyl-aryl group exchange between 1 and
AlEt3 is occurring, and it seems likely that this process leads to
catalyst deactivation in ethylene tetramerization. This could be
due either to the formation of more coordinating counterions,
such as [(C6F5)AlEt3]-, or Al-mediated exchange of the C6F5
group onto the active Cr center, as Bochmann has shown for
zirconocene polymerization catalysts.53
The same experiment conducted with Al(OC6F5)3 (2) and 30
equiv of AlEt3 showed that all of 2 is likewise consumed after
ca. 10 min, giving a 77:23 mixture of two species by 19F NMR.
The relative abundance of these changes to 91:9 after 2 h. The
terminal methylene and methyl group signals for the major
species are resolved and indicate that the dimer Et2Al(µ-OC6F5)-
(µ-Et)AlEt2 has formed. The minor species observed in the 19
F
spectrum is presumably derived from the Al(OC6F5)2Et unit and
would also be a dimer, Al2(OC6F5)2Et4, although the structure
of this cannot be determined from the NMR data. Again, this
result illustrates that this cocatalyst is susceptible to degradation
in the presence of AlEt3, and this is reflected in the catalytic
results observed.
These combined results provide strong evidence for a
cocatalyst that is changing over time, and an anion coordination
strength that quickly increases at the start of a run. A mechanism
by which this occurs is suggested in Scheme 3. The reason for
the more rapid deactivation of the trimerization active species
may be linked to the greater interaction between Cr and the
(71) Ivanov, S. V.; Rockwell, J. J.; Polyakov, O. G.; Gaudinski, C. M.;
Anderson, O. P.; Solntsev, K. A.; Strauss, S. H. J. Am. Chem. Soc. 1998,
120, 4224.
(72) Ramey, K. C.; O’Brien, J. F.; Hasegawa, I.; Borchert, A. E. J. Phys.
Chem. 1965, 69, 3418.
(73) Yamamoto, O.; Hayamizu, K. J. Phys. Chem. 1968, 72, 822.

2568 Organometallics, Vol. 26, No. 10, 2007
McGuinness et al.
In terms of the mechanism by which coordination of the anion
would suppress C8 formation, it can be speculated that it does
so by hindering ethylene uptake at the Cr-C6 metallacyclic
stage. A number of studies on metallocene polymerization
catalysts have shown that ethylene must first displace the anion
from the coordination sphere prior to insertion,45,48 and the same
condition could apply here. It is interesting that MAO and 7
lead to the same selectivity. It is likely that the anion present
when these activators are employed is near the limit of complete
dissociation, and as such the anion has less effect on ethylene
coordination. Under these conditions solvent coordination could
become the dominant factor.
4. Experimental Section
Figure 6. Mass of C6 and C8 formed as a function of time for
catalysis with 3 equiv of Al(OC6F5)3 (2). Conditions: 10 µmol of
CrCl3(thf)3, 12 µmol of II, 30 AlEt3, 45 °C, 40 bar.
General Comments. All manipulations were carried out using
standard Schlenk techniques or in a nitrogen glovebox, using
solvents purified and dried by standard procedures. [Ph3C][Ta-
(OC6F5)6] (4),59 [Ph3C][Al{OC(CF3)3}4] (7),60 and [Ph3C][CB11H6-
Br6]65 were prepared via literature procedures. [Ph3C][FAl{OC-
(CF3)3}3] (8) and [Ph3C][{(F3C)3CO}3Al-F-Al{OC(CF3)3}3] (9)
were supplied by Prof. Ingo Krossing, while their preparation has
been described.64
Scheme 3
X-ray Crystallography. Data were collected on Rigaku MM007/
CCD diffractometers (Mo KR radiation, confocal optics, λ )
0.710 73 Å) and were corrected for absorption. The structures were
solved by direct methods and refined by full-matrix least-squares
on F2 values of all data.74
anion in this case. Marks has shown that Zr and Ti polymeri-
zation catalysts activated with [Ta(OC6F5)6]- undergo catalyst
degradation that arises from transfer of the OC6F5 group to the
Zr or Ti center.59 An increased interaction between the Cr center
and the aluminate can reasonably be expected to favor this
process. At this stage we have not attempted to characterize
such a Cr-OC6F5 species.
Given these observations, the results in Table 1 must be
reinterpreted to an extent. In the experiments where rapid
catalyst deactivation is observed (with cocatalysts 1-6 and 8),
the C6/C8 ratio reported more likely represents an average over
the run, corresponding to the “average” anion coordination
strength during catalysis.
(a) Al2(OC6F5)6‚(toluene) (2). Crystal data: C36Al2F30O6.C7H8.
Mr ) 1244.45, colorless prism, crystal size 0.15 × 0.15 × 0.15
mm, triclinic, P1h, a ) 10.247(3) Å, b ) 10.451(3) Å, c ) 11.196-
(3) Å, R ) 99.367(6)°, â ) 107.070(4)°, γ ) 104.356(5)°, V )
1073.9(5) Å3, Z ) 1, Dcalcd ) 1.924 Mg m-3, µ ) 0.251 mm-1, T
) 93(2) K, 6752 data (3714 unique, Rint ) 0.0254, 3.28 < θ <
2
2 2
25.34°), Rw ) {∑[w(Fo - Fc2)2]/∑[w(Fo ) ]}1/2 ) 0.0990, con-
2
ventional R ) 0.0383 for F values of reflections with Fo > 2σ-
2
(Fo ) (2981 observed reflections), S ) 0.703 for 381 parameters.
Residual electron density extremes were 0.364 and -0.253 e Å-3
.
(b) (Et2O)Al{OCH(C6F5)2}3 (5). Crystal data: C43H13AlF30O4.
Mr ) 1190.51, colorless platelet, crystal size 0.10 × 0.10 × 0.02
mm, monoclinic, P21/n, a ) 10.7382(6) Å, b ) 46.564(3) Å, c )
17.1024(9) Å, â ) 92.971(3)°, V ) 8540.0(8) Å3, Z ) 8 (two
3. Summary and Conclusion
independent molecules), Dcalcd ) 1.852 Mg m-3, µ ) 0.225 mm-1
,
A variety of new and existing cocatalysts have been evaluated
for the trimerization and tetramerization of ethylene with a Cr-
PNP catalyst system. The stability of selected cocatalysts has
been studied, and the less effective ones have been found to
undergo degradation upon treatment with AlEt3. In contrast, the
use of a stable and noncoordinating anion leads to a high
productivity in the absence of alumoxanes. These studies have
led us to the easily prepared cocatalyst [Ph3C][Al{OC(CF3)3}4]
(7), which due to its high stability and weakly coordinating
nature gives a highly productive tetramerization catalyst.
Herein it has been shown for the first time the remarkable
influence that the cocatalyst can have on relative C6/C8
selectivity, which provides strong evidence for a formally
cationic (or cation-like) active species, in which the coordinating
strength of the anion effects the selectivity. It is a combination
of both coordinating ability and anion stability that controls
activity and selectivity, however, as the two are related in that
poor stability can lead to a more coordinating anion over the
lifetime of a catalytic run. We have previously shown the effect
that ligand modification has on selectivity with this class of
ligand. Herein we have shown that further control over this
versatile catalyst system is available through careful cocatalyst
modification.55
T ) 93(2) K, 55 419 data (13 975 unique, Rint ) 0.0577, 2.23 < θ
< 25.35°), Rw ) {∑[w(Fo - Fc2)2]/∑[w(Fo ) ]}1/2 ) 0.1856,
2
2 2
2
conventional R ) 0.1081 for F values of reflections with Fo
2σ(Fo ) (11 680 observed reflections), S ) 1.174 for 1406 param-
eters. Residual electron density extremes were 0.623 and -0.478
>
2
e Å-3
.
(c) Al2{OCH(C6F5)2}6 (5′). Crystal data: C78H6Al2F60O6. Mr )
2232.79, colorless prism, crystal size 0.13 × 0.13 × 0.10 mm,
monoclinic, P21/n, a ) 14.526(3) Å, b ) 13.990(3) Å, c ) 18.153-
(4) Å, â ) 92.314(10)°, V ) 3685.9(14) Å3, Z ) 2, Dcalcd ) 2.012
Mg m-3, µ ) 0.252 mm-1, T ) 93(2) K, 19 844 data (6412 unique,
2
Rint ) 0.1291, 2.29 < θ < 25.35°), Rw ) {∑[w(Fo - Fc2)2]/∑-
2 2
[w(Fo ) ]}1/2 ) 0.2985, conventional R ) 0.1127 for F values of
2
2
reflections with Fo > 2σ(Fo ) (3730 observed reflections), S )
1.061 for 381 parameters. Residual electron density extremes were
0.916 and -0.514 e Å-3. The molecular structure of this compound
is shown in the Supporting Information.
(d) (Et2O)Al{OC(CF3)3}3 (6). Crystal data: C16H10AlF27O4. Mr
) 806.22, colorless prism, crystal size 0.03 × 0.03 × 0.03 mm,
monoclinic, P21/n, a ) 9.910(5) Å, b ) 16.630(7) Å, c ) 16.280-
(7) Å, â ) 91.16(2)°, V ) 2682(2) Å3 , Z ) 4, Dcalcd ) 1.996 Mg
(74) Sheldrick, G. M. SHELXTL, version 6.1; Bruker AXS, Madison,
WI, 2001.

Ethylene Trimerization and Tetramerization
Organometallics, Vol. 26, No. 10, 2007 2569
m-3, µ ) 0.289 mm-1, T ) 93(2) K, 16 045 data (4750 unique,
portionwise until the starting alcohol was completely consumed
by NMR analysis. The solvent was then removed under vacuum,
the residue taken up in diethyl ether, and the solvent removed again
to yield a white powder. Yield: 5.743 g (91%). Anal. Calcd (found)
2
2 2
Rint ) 0.2077, 1.75 < θ < 25.22°), Rw ) {∑[w(Fo - Fc ) ]/
2 2
∑[w(Fo ) ]}1/2 ) 0.3827, conventional R ) 0.1457 for F values of
2
2
reflections with Fo > 2σ(Fo ) (2016 observed reflections), S )
1
1.076 for 434 parameters. Residual electron density extremes were
0.607 and -0.616 e Å-3
for C43H13O4F30: C, 43.38 (44.14); H, 1.10 (1.50). H NMR (300
.
MHz, CDCl3; δ): 6.65 (s, 3H, OCH); 4.18 (q, J ) 7 Hz, 4H, OCH2-
CH3); 1.32 (t, J ) 7 Hz, 6H, OCH2CH3). 19F NMR (282 MHz,
CDCl3; δ): -144.5 (2F, o-F); -155.2 (2F, m-F); -162.6 (1F, p-F).
13C NMR (75 MHz, CDCl3; δ): 146.5, 143.2, 139.6, 136.2 (CF);
116.8 (ipso-C); 69.6 (OCH2CH3); 61.7 (OCH); 13.3 (OCH2CH3).
The molecular structure of the compound is shown in Figure 2.
(Et2O)Al{OC(CF3)3}3 (6). A solution of 4 mL (29 mmol) of
perfluoro-tert-butyl alcohol in 10 mL of diethyl ether was cooled
to 0 °C and treated with a 1.9 M solution of triethylaluminum (3
mL, 5.7 mmol). After it was stirred overnight, the solution was
heated to 65 °C for a further day. After this solution was cooled,
the solvent was removed under vacuum to give fine needles of
colorless product. Yield: 4.383 g (95%). 1H NMR (300 MHz,
CDCl3; δ): 4.34 (q, J ) 7 Hz, 4H, OCH2CH3); 1.43 (t, J ) 7 Hz,
6H, OCH2CH3). 19F NMR (282 MHz, CDCl3; δ): -75.9 (CF3).
The molecular structure of this compound is shown in Figure 3.
NMR Spectra of 1 + 30 AlEt3 in CD2Cl2. Major species: 19F
NMR (282 MHz; δ) -122.0 (dt, J ) 7 Hz, 15 Hz, o-F), -154.1
(tt, J ) 3 Hz, 14 Hz, p-F), -162.4 (m, m-F); 1H NMR (300 MHz;
δ) 0.75 (“t”, J ) 7 Hz, 4H, AlCH2), 0.99 (“t”, J ) 7 Hz, 2H,
AlCH2). Minor species (7%): 19F NMR (282 MHz; δ) -127.1,
-152.8, -159.2. Minor species (5%): 19F NMR (282 MHz; δ)
-126.3, -152.6, -159.2.
Ethylene Oligomerization. Ethylene oligomerization was carried
out in a 300 mL stainless steel reactor with mechanical stirring.
The oven-dried vessel was purged with N2, followed by ethylene,
and charged with 90 mL of toluene. After heating to 45 °C, a
solution of Cr/II/AlEt3 in 5 mL of toluene was injected, followed
by a 5 mL toluene solution of the cocatalyst. The reactor was
immediately charged with ethylene and maintained at the desired
pressure over the duration of the reaction. At the end of a run, the
reactor was cooled in an ice bath and excess ethylene bled before
addition of an internal standard (nonane, 1000 µL). After quenching
with MeOH followed by 10% HCl, the organic phase was analyzed
by GC, while the solids were filtered, washed, dried, and weighed.
Al(OC6F5)3 (2). A solution of 2 M perfluorophenol (7.0 mL, 14
mmol) in toluene was added to a flask, and a 1.9 M solution of
triethylaluminum (2.0 mL, 3.8 mmol) was added dropwise. The
solution was then heated to 60 °C for 4 h and cooled. The solvent
toluene was reduced under vacuum and 10 mL of petroleum spirits
added. The resulting powder that formed was washed four times
with petroleum spirits (5 mL) and dried under vacuum to afford a
white powder. Yield: 1.694 g (77%) Anal. Calcd (found) for
C18F15O3Al: C, 37.52 (37.34). 19F NMR (282 MHz, DMSO-d6;
δ): -161.2, -162.1, -162.8 (2F, o-F); -169.9 (2F, m-F); 179.0
(1F, p-F). 13C NMR (75 MHz, DMSO-d6; δ): 143.0, 142.4, 139.2,
136.0, 132.8, 129.7 (CF). The molecular structure of the compound
is shown in Figure 1.
NMR Spectra of 2 + 30 AlEt3 in CD2Cl2. Major species: 19F
NMR (282 MHz; δ) -156.9 (d, J ) 22 Hz, o-F), -162.3 (td, J )
1
22 Hz, 4 Hz, m-F), -164.4 (tt, J ) 22 Hz, 4 Hz, p-F); H NMR
[(Et2O)2H][Al(OC6F5)4] (3). At 0 °C a solution of perfluorophe-
nol (7.5 mL, 15 mmol) in 20 mL of diethyl ether was treated
dropwise with 1.9 M triethylaluminum solution (1.7 mL, 3.2 mmol).
After 2 h at room temperature the solvent was removed under
vacuum and the product washed twice with petroleum spirits to
give a white powder. Yield: 1.641 g (56%). Anal. Calcd (found)
for C32H21O6F20Al: C, 42.31 (42.28); H, 2.33 (2.47). 1H NMR (300
MHz, CDCl3; δ): 5.60 (br, 1H, H(OEt2)); 4.39 (q, J ) 7 Hz, 8H,
OCH2CH3); 1.55 (t, J ) 7 Hz, 12H, OCH2CH3). 19F NMR (282
MHz, CDCl3; δ): -163.7 (d, 2F, J ) 22 Hz, o-F); -165.8 (t, 2F,
J ) 22 Hz, m-F); -171.1 (t, 1F, J ) 22 Hz, p-F). 13C NMR (75
MHz, CDCl3; δ): 142.0, 140.0, 138.8, 136.6 (CF); 71.9 (OCH2);
13.9 (OCH2C3).
(300 MHz; δ) 0.83 (t, J ) 12 Hz, 12H, AlCH2CH3), 0.0 (m, AlCH2-
CH3, 8H). Minor species: 19F NMR (282 MHz; δ) -157.1 (d, J )
21 Hz, o-F), -163.3 (m, J ) 21 Hz, m-F), -164.7 (tt, J ) 21, 4
Hz, p-F).
Acknowledgment. We thank Prof. Ingo Krossing for helpful
input on cocatalyst properties and for supplying compounds 8
and 9. We also thank the members of the Sasol Technology
Olefin Transformations group for helpful discussions and
suggestions.
Supporting Information Available: CIF files giving crystal-
lographic data for the structures reported and a figure giving the
molecular structure diagram of Al2{OCH(C6F5)2}6. This material
is available free of charge via the Internet at http://pubs.acs.org.
(Et2O)Al{OCH(C6F5)2}3 (5). At 0 °C a solution of (C6F5)2C-
(H)OH (5.79 g, 15.9 mmol) in 15 mL of diethyl ether was treated
dropwise with 1.9 M AlEt3. The addition of AlEt3 was continued
OM070029C