Ethylene oligomerization using first-row transition metal complexes featuring heterocyclic variants of bis(imino)pyridine ligands

Article abstract of DOI:10.1021/om900280j

This report examines the replacement of the imine and pyridine functionalities of the ubiquitous bis(imino)pyridine ligand with various heterocycles. The synthesis of a new class of ligand based around thiazole is described; 2,4-bis[1-(arylimino)ethyl]thi

Full text of DOI:10.1021/om900280j

4852 Organometallics 2009, 28, 4852–4867
DOI: 10.1021/om900280j
Ethylene Oligomerization Using First-Row Transition Metal Complexes
Featuring Heterocyclic Variants of Bis(imino)pyridine Ligands
Kenny Tenza,*,† Martin J. Hanton,† and Alexandra M. Z. Slawin‡
‡
†Sasol Technology (U.K.) Ltd, Purdie Building, North Haugh, St Andrews, KY16 9ST, U.K., and School of
Chemistry, University of St Andrews, North Haugh, St Andrews, KY16 9ST, U.K.
Received April 13, 2009
This report examines the replacement of the imine and pyridine functionalities of the ubiquitous
bis(imino)pyridine ligand with various heterocycles. The synthesis of a new class of ligand based
around thiazole is described; 2,4-bis[1-(arylimino)ethyl]thiazole (aryl=Ph, 1a; Dipp, 1b) and 2,5-
bis[1-(arylimino)ethyl]thiazole (aryl=Ph, 1c; Dipp, 1d) have been prepared in good yield and fully
characterized. The coordination chemistry of these ligands with chromium, iron, and cobalt is
explored, and the potential of these complexes as ethylene oligomerization initiators is assessed. The
chromium complex 2a shows an extremely unusual alternating distribution of higher R-olefin
products, which has been previously observed on only one occasion. Both series of products, C4n
and C4nþ2, show Schulz-Flory behavior but with distinctly different k values. The new ligand 2,5-
bis[1-(phenylimino)ethyl]-1-methylpyrrole (1e) is reported along with the attempted synthesis
of some corresponding iron complexes. The complexation of chromium by 2,5-bis-
(phenyliminomethyl)thiophene (1f) is also described, and this material was screened for ethylene
oligomerization activity, detailed studies indicating that the ligand may be labile under catalytic
conditions. A number of other known heterocyclic ligands incorporating pyrazolyl and benzimida-
zole functionalities have also been explored with iron, and for the first time their potential to facilitate
ethylene oligomerization was assessed. All complexes have been tested via activation with MMAO-
3A and AlEt3/[Ph3C][Al(OtBuF)4].
Introduction
including pyrimidine (A),4 pyrazine (B),5 triazine (C),4 pyr-
role (D),6 carbazole (E),4,7 furan (F),4 or thiophene (G).4
While the six-membered N-heterocyclic permutations have
generally still yielded complexes that initiated oligomeriza-
tion catalysis with reasonable activity (A and B), the five-
membered heterocycle variants have been less successful.1,4
Indeed subtle changes to the ligand structure appear to have
a dramatic effect upon catalysis, as well as the ability of the
ligand to ligate iron or cobalt.4,6 It is noted that in the case of
bis(imino)pyrrole ligands (D) a bidentate coordination mode
is observed with zirconium, chromium, iron, and cobalt,
rather than the expected tridentate coordination, rationa-
lized as a result of the smaller reach of the pyrrolide-based
ligand.1a,6,8 Further evidence to support this logic can be
drawn from the ability of the reported bis(imino)carbazole
ligands (E), which have an extended imine donor arm, to
ligate iron,4,7 although the resulting complexes were inactive
for ethylene oligomerization when combined with MAO.1c
Non-metallocene-based oligomerization initiators have
come of age over the past decade, with bis(imino)pyridine
complexes being in the vanguard of this assault.1 Since the
initial reports of ethylene polymerization with catalysts
based upon ligands of this type,2 there has been a significant
focus on developing alternative variants of the bis-
(imino)pyridine ligand with varying degrees of success.1c,3
One avenue of exploration has been the replacement of the
central pyridine ring with alternative heterocycles (Scheme 1)
*Corresponding author E-mail: kenny.tenza@sasol.com.
(1) (a) Britovsek, G. J. P.; Gibson, V. C.; Wass, D. F. Angew. Chem.,
Int. Ed. 1999, 38, 428. (b) Gibson, V. C.; Spitzmesser, S. K. Chem. Rev.
2003, 103, 283. (c) Gibson, V. C.; Redshaw, C.; Solan, G. A. Chem. Rev.
2007, 107, 1745.
(2) Small, B. L.; Brookhart, M. J. Am. Chem. Soc. 1998, 120, 7143. (b)
Small, B. L.; Brookhart, M.; Bennett, A. M. A. J. Am. Chem. Soc. 1998, 120,
4049. (c) Bennett, A. M. A. (Dupont) WO 98/27124, 1998; Chem. Abstr.
1998, 129, 122973x. (d) Bennett, A. M. A. CHEMTECH 1999 (July), 24-
28. (e) Britovsek, G. J. P.; Gibson, V. C.; Kimberley, B. S.; Maddox, P. J.;
McTavish, S. J.; Solan, G. A.; White, A. J. P.; Williams, D. J. Chem.
Commun. 1998, 849. (f) Britovsek, G. J. P.; Bruce, M.; Gibson, V. C.;
Kimberley, B. S.; Maddox, P. J.; Mastroianni, S.; McTavish, S. J.; Redshaw,
(4) Britovsek, G. J. P.; Gibson, V. C.; Hoarau, O. D.; Spitzmesser,
S. K.; White, A. J. P.; Williams, D. J. Inorg. Chem. 2003, 42, 3454.
(5) (a) Beaufort, L.; Benvenuti, F.; Noels, A. F. J. Mol. Catal. A:
Chem. 2006, 260, 210. (b) Beaufort, L.; Benvenuti, F.; Noels, A. F. J. Mol.
Catal. A: Chem. 2006, 260, 215.
€
C.; Solan, G. A.; Stromberg, S.; White, A. J. P.; Williams, D. J. J. Am. Chem.
Soc. 1999, 121, 8728. (g) Britovsek, G. J. P.; Dorer, B. A.; Gibson, V. C.;
Kimberley, B. S.; Solan, G. A. WO patent 99/12981 (BP Chemicals Ltd);
Chem. Abstr. 1999, 130, 252793.
(3) (a) Small, B. L.; Rios, R.; Fernandez, E. R.; Carney, M. J.
Organometallics 2007, 26, 1744. (b) Schmiege, B. M.; Carney, M. J.; Small,
B. L.; Gerlach, D. L.; Halfen, J. A. Dalton Trans. 2007, 2547.
(6) Dawson, D. M.; Walker, D. A.; Thornton-Pett, M.; Bochmann,
M. J. Chem. Soc., Dalton Trans. 2000, 459.
(7) Gibson, V. C.; Spitzmesser, S. K.; White, A. J. P.; Williams, D. J.
Dalton Trans. 2003, 2718.
(8) Gibson, V. C.; Maddox, P. J.; Newton, C.; Redshaw, C.; Solan,
G. A.; White, A. J. P.; Williams, D. J. Chem. Commun. 1998, 1651.
r
pubs.acs.org/Organometallics
Published on Web 07/27/2009
2009 American Chemical Society

Article
Organometallics, Vol. 28, No. 16, 2009 4853
Scheme 1. The Different Heterocycles and Their Complexes That Have Been Used to Replace Pyridine in Variations of the
Bis(imino)pyridine Ligand
Scheme 2. Donor Arm-Modified Bis(imino)pyridine Complexes
Previously Reported
Furthermore, in the case of the triazine (C), furan (F), and
thiophene (G) variants, the ligand fails to ligate iron or cobalt.4
An alternative strategy for modifying the bis-
(imino)pyridine scaffold has been to target the imine-donor
arms, and again the use of heterocyclic moieties has been
examined in the form of pyridine (H and I),9 phenanthroline
(J),10 imidazolylidene (carbenes) (K),11 furan (L),12 thio-
phene (M),12 oxazoline (N),13 and benzimidazole (O þ P)14
functionalities (Scheme 2). The pyridyl (H þ I) and phenan-
throlyl (J) derivatives, when bound to iron, serve as highly
active oligomerization initiators when activated with methy-
lalumoxane (MAO),9,10 contrasting with the imidazolylidene
derivatives (K), which while exceptionally active in combina-
tion with chromium,11 have not been reported as success-
ful oligomerization catalysts when bound to iron.1c Cobalt
complexes of the furan (L) derivatives show low activity for
the dimerization of ethylene to butenes, while the thiophene-
based analogues (M) oligomerize ethylene with moderate
activity.12 Oxazoline-based ligands (N) have been described
in combination with iron and ruthenium for ethylene
polymerization after activation with MAO, but low activities
were observed.15,16 In the case of benzimidazole-substituted
pyridines, ligands of the type bis(benzimidazole)pyridine (O)
have been applied only to chromium, where in combination
with MAO or diethylaluminum chloride (DEAC) they acted
as highly effective initiators for ethylene oligomerization.14a
Variants of the bis(imino)pyridine ligand where only one
imino arm has been replaced with benzimidazole (P) have
been reported bound to iron, cobalt, and nickel, all of which
successfully initiated ethylene oligomerization.14b-14d
While the use of heterocycle-based variations of the bis-
(imino)pyridine ligand has clearly been widely explored,
there appeared to us several rational alternatives that have
not yet been considered. For example, while replacement of
pyridine with a pyrrole ring allows a different ring size to be
examined, it also introduces the complicating factor of a sp3-
hybridized-N as compared to the sp2-N in pyridine. Two
options exist: the N-Me variant can be used to ensure a
neutral N-donor, but the N-H offers the opportunity of an
amide. The latter of these options has been the avenue chosen
by others.6 In order to maintain the sp2-hybridized-N while
moving to a five-membered heterocycle, we envisaged the
replacement of the central pyridine ring with a thiazole. A
further point of interest with such a motif is the option of
N- or S-binding via the thiazole segment depending upon
(9) Britovsek, G. J. P.; Baugh, S. P. D.; Hoarau, O.; Gibson, V. C.; Wass,
D. F.; White, A. J. P.; Williams, D. J. Inorg. Chim. Acta 2003, 345, 279.
(10) (a) Sun, W.-H.; Jie, S.; Zhang, S.; Zhang, W.; Song, Y.; Ma, H.;
€
Chen, J.; Wedeking, K.; Frohlich, R. Organometallics 2006, 25, 666. (b)
Pelletier, J. D. A.; Champouret, Y. D. M.; Cadarso, J.; Clowes, L.; Gaete, M.;
Singh, K.; Thanarajasingham, V.; Solan, G. A. J. Organomet. Chem. 2006,
691, 4114.
(11) (a) McGuinness, D. S.; Gibson, V. C.; Wass, D. F.; Steed, J. W.
J. Am. Chem. Soc. 2003, 125, 12716. (b) Danopoulos, A. A.; Tsoureas, N.;
Wright, J. A.; Light, M. E. Organometallics 2004, 23, 166.
(12) (a) Bianchini, C.; Mantovani, G.; Meli, A.; Migliacci, F.; Laschi,
F. Organometallics 2003, 22, 2545. (b) Bianchini, C.; Giambastiani, G.;
Mantovani, G.; Meli, A.; Mimeau, D. J. Organomet. Chem. 2004, 689,
1356. (c) Bianchini, C.; Gatteschi, D.; Giambastiani, G.; Rios, I. G.; Ienco,
A.; Laschi, F.; Mealli, C.; Meli, A.; Sorace, L.; Toti, A.; Vizza, F. Organo-
metallics 2007, 26, 726.
(13) (a) Nomura, K.; Warit, S.; Imanishi, Y. Macromolecules 1999, 32,
4732. (b) Nomura, K.; Warit, S.; Imanishi, Y. Bull. Chem. Soc. Jpn. 2000,73, 599.
(14) (a) Zhang, W.; Sun, W.-H.; Zhang, S.; Hou, J.; Wedeking, K.;
€
Schultz, S.; Frohlich, R.; Song, H. Organometallics 2006, 25, 1961. (b)
Hao, P.; Zhang, S.; Sun, W.-H.; Shi, Q.; Adewuyi, S.; Lu, X.; Li, P.
Organometallics 2007, 26, 2439. (c) Sun, W.-H.; Hao, P.; Zhang, S.; Shi,
Q.; Zuo, W.; Tang, X. Organometallics 2007, 26, 2720. (d) Xiao, L.; Gao, R.;
Zhang, M.; Li, Y.; Cao, X.; Sun, W.-H. Organometallics 2009, 28, 2225.
(15) Nomura, K.; Warit, S.; Imanishi, Y. Bull. Chem. Soc. Jpn. 2000,
73, 599.
(16) Nomura, K.; Warit, S.; Imanishi, Y. Macromolecules 1999, 32, 4732.

4854 Organometallics, Vol. 28, No. 16, 2009
Tenza et al.
how the imine moieties are distributed (Scheme 3, 1a,b versus
1c,d). Herein we report the synthesis of ligands of this type
with phenyl and 2,6-diisopropylphenyl (Dipp) substituents
at the imine nitrogen, explore their coordination chemistry
with chromium, iron, and cobalt centers, and study their use
as initiators for ethylene oligomerization. The oligomeriza-
tion catalysis has been performed using MMAO-3A17 as
activator alongside the triethylaluminum (TEA)/trityl
tetrakis(perfluoro-tert-butoxy)aluminate (TA) activator
combination for comparison, which has been recently re-
ported by us for the effective activation of bis(imino)pyridine
complexes of the first-row transition metals.18 We also report
the synthesis of a new 1-methylpyrrole-based bis(imine)
ligand (Scheme 3, 1e) and describe the attempted formation
of iron complexes. While the failure of bis(imino)thiophene
ligands (Scheme 3, 1f) to bind iron centers has been pre-
viously reported (vide supra), we report here for the first time
the ability of these compounds to ligate chromium and
examine the resulting complexes as initiators for ethylene
oligomerization.
Scheme 3. Synthesis of Five-Membered Heterocycle-Based
Bis(imine) Ligands
It is noteworthy that all of the most active iron initiators
based upon the bis(imino)pyridine ligand scaffold and deri-
vatives thereof feature at least one imine donor arm. Thus we
choose to explore diaza-heterocycles further in the form of
bis(pyrazole)pyridine ligands (Scheme 4, 1g and 1h) and
bis(benzimidazole)pyridine ligands (Scheme 4, 1i and 1j),
where the imine donor is incorporated into the heterocycle.
While both of these ligand types are known,19-22 and 1g,23
1i,21a,21c and 1j21a,21c have been reported in combination with
iron, the precise iron complexes prepared either have not
been relevant precursors for this type of catalysis or their
potential to initiate oligomerization catalysis has not been
assessed. Herein we report the synthesis of the iron-
(II)bromide complexes of these ligands and probe the cata-
lytic potential of these toward ethylene oligomerization.
These materials were subsequently condensed with the re-
levant aniline in the presence of titanium tetrachloride to
yield the corresponding bis(imino) products (Scheme 3) as
yellow solids (1a, 96%; 1b, 40%; 1c, 81%; 1d, 33%).25 The
1H and 13C{1H} NMR spectra of these materials were as
expected, with resonances for the carbons of the imine
moieties at ∼δ 160 ppm. The slight upfield shift observed
for the resonances associated with the methyl groups, as
compared to the ketone precursors, was as anticipated,
whereas the resonances associated with the thiazole moiety
remained largely unaffected by the conversion from diketone
to diketimine.
Results and Discussion
The preparation of 2,5-bis(acetyl)-1-methylpyrrole has
been reported previously but requires the use of autoclaves
at high temperature;26 thus an alternative method was
employed here using trifluoroacetic anhydride to accelerate
the reaction.27 Condensation of aniline with the diketone
yielded the desired 2,5-bis[1-(phenylimino)ethyl]-1-methyl-
pyrrole (1e) in good yield (Scheme 3). The synthesis of a
number of bis(imino)thiophene compounds has been
reported previously,4,28 including the phenylimino derivative
used here.28e-28g The solvent-free mixture of the two pre-
cursors reacted rapidly to furnish the desired ligand (1f) in
excellent yield (Scheme 3).
Synthesis of Ligands. The thiazole precursors 2,4-diace-
tylthiazole and 2,5-diacetylthiazole were prepared according
to the method of Aitken et al. via a three-step synthesis.24
(17) MMAO-3A is a modified alumoxane cocatalyst sourced from
Akzo Nobel with the general formula [-Al(R)-O-]n, where R=Me (75%),
iBu (25%).
(18) (a) Hanton, M. J.; Tenza, K. Organometallics 2008, 27, 5712. (b)
Tooze, R. P.; Hanton, M. J.; Tenza, K. (Sasol Technology) WO2008038173,
2008.
(19) Jameson, D. L.; Goldsby, K. A. J. Org. Chem. 1990, 55, 4992.
(20) Brien, K. A.; Garner, C. M.; Pinney, K. G. Tetrahedron 2006, 62,
3663.
The pyrazolyl ligand 2,6-bis(3,5-dimethylpyrazol-1-yl)-
pyridine (1g) has been widely reported and was prepared via a
modification of the literature procedure,19 whereas the 2,6-bis-
(3,5-diphenylpyrazol-1-yl)pyridine ligand (1h) has been reported
(21) (a) Addison, A. W.; Burman, S.; Wahlgren, C. G.; Rajan, O. A.;
Rowe, T. M.; Sinn, E. J. Chem. Soc., Dalton Trans. 1987, 2621. (b) Piguet,
C.; Bocquet, B.; Muler, E.; Williams, A. F. Helv. Chim. Acta 1989, 72, 323.
(c) Wang, X.; Wang, S.; Li, L.; Sundberg, E. B.; Paolo Gacho, G. Inorg.
Chem. 2003, 42, 7799.
(22) Other examples of ethylene oligomerization using complexes
with nitrogen-bridged bis(pyrazolyl) ligands have been reported, but are
far less closely related to the bis(imino)pyridine scaffold and have not
been demonstrated with iron, but instead chromium and nickel, metals
with a much higher predisposition toward ethylene oligomerization. (a)
(25) The 2,6-diisopropylimine variants, 1b and 1d, were isolated in
lower yield due to problems associated with their purification, specifi-
cally the removal of the DippNH3Cl byproduct.
€
Ajellal, N.; Kuhn, M. C. A.; Boff, A. D. G.; Horner, M.; Thomas, C. M.;
(26) Harsanyi, M. C.; Norris, R. K. J. Org. Chem. 1987, 52, 2209.
ꢀ
(27) Song, C.; Knight, D. W.; Whatton (nee Fagan), M. A. Tetra-
Carpentier, J.-F.; Casagrande, O. L.Jr. Organometallics 2006, 25, 1212.
(b) Junges, F.; Kuhn, M. C. A.; dos Santos, A. H. D. P.; Rabello, C. R. K.;
Thomas, C. M.; Carpentier, J.-F.; Casagrande, O. L.Jr. Organometallics
2007, 26, 4010. (c) de Oliveira, L. L.; Campedelli, R. R.; Kuhn, M. C. A.;
Carpentier, J.-F.; Casagrande, O. L.Jr. J. Mol. Catal. A 2008, 288, 58.
(23) Calderazzo, F.; Englert, U.; Hu, C.; Marchetti, F.; Pampaloni, G.;
Passarelli, V.; Romano, A.; Santi, R. Inorg. Chim. Acta 2003, 344, 197.
(24) Aitken, K. M.; Aitken, R. A. Tetrahedron 2008, 64, 4384.
hedron Lett. 2004, 10, 133.
(28) (a) Skene, W. G.; Dufresne, S. Acta Crystallogr. E 2006, E62,
1116. (b) Skene, W. G.; Dufresne, S. Polym. Prepr. 2004, 45, 728. (c) Skene,
W. G.; Dufresne, S. Org. Lett. 2004, 6, 2949. (d) Xe, W.; Wang, Q.; Wurz,
R. P. (Nova Chemicals) EP1046647 A2, 2000. (e) Lumbroso, H.; Pastour, P.
Compt. Rend. 1965, 261, 1279. (f) Vaysse, M.; Pastour, P. Compt. Rend.
1964, 259, 2657. (g) Sone, C. Bull. Chem. Soc. Jpn. 1964, 37, 1197.

Article
Organometallics, Vol. 28, No. 16, 2009 4855
Scheme 4. Bis(diaza-heterocyclic)pyridine Ligands Employed in
This Study
and 3c) in high yields as dark green-brown solids (Scheme 6).
The magnetic moments observed were lower than that
typically seen for planar tridentate ligands bound to iron(II)
(3a, μeff=3.96 μB; 3c, μeff=3.52 μB).4,30,31 FAB-MS analysis
revealed a [M - Br]þ fragment for 3c, but no tractable signal
for 3a. The ability of both forms of the bis(imino)thiazole
ligand to bind iron is most interesting when considered in
light of previous studies. In the case of 3a, where the thiazole
moiety is bound via the N atom, the ligand is unequivocally
tridentate (vide infra), contrasting with the bis(imino)pyrrole
analogues and illustrating that replacement of pyridine
with a five-membered N-heterocycle is possible but that
maintenance of the N donor in a sp2-hybridized form is
essential. Finally, the ability of 1c, with the thiazole moiety
predisposed for S-binding, to complex iron when bis-
(imino)thiophene fails to do so was most unexpected.
Scheme 5. Synthesis of Chromium Complexes
Crystals of 3a suitable for study by X-ray diffraction
were grown from prolonged cooling of a DCM/toluene
solution. The structure obtained clearly shows a tridentate
binding mode via the three N donors for the 2,4-bis-
[1-(phenylimino)ethyl]thiazole ligand (Figure 1). The thia-
zole moiety of the ligand exhibits disorder as a consequence
of the crystallographic C2 symmetry, and so no analysis of
the metric parameters for this fragment can be made; simi-
larly there is also a disordered toluene solvate molecule in the
structure. The observation of a tridentate binding mode for
ligand 1a with FeBr2 is particularly interesting when com-
pared to the bidentate binding mode observed exclusively for
the bis(imino)pyrrole ligands (vide supra),6 a difference we
assign to the sp2-hybridized N in a thiazole as compared to
the sp3-N in a pyrrole ring. This serves to illustrate
that replacement of the central six-membered N-heterocycle
in the bis(imino)pyridine ligand with a five-membered
N-heterocycle is possible in a way that maintains the
tridentate binding mode of the ligand. Interestingly the
on only one previous occasion via a slightly different route from
that employed here.20 The ligand 2,6-bis(1-methyl-2-benzimida-
zolyl)pyridine (1j) has also been reported previously, being
prepared from the reaction of pyridine-2,6-dicarboxylic acid
with 1-methyl-1,2-diaminobenzene,21 while the synthesis em-
ployed herein was via the N-methylation of 2,6-bis(2-benzimi-
dazolyl)pyridine.
Fe-Ncentral heterocycle bond length observed in 3a of 2.050(3) A
˚
is slightly shorter then that found for the examples of the
bis(imino)pyridine ligand bound to FeX2 (X = Cl, Br): in
these cases it is 0.04-0.07 A shorter, while the Fe-Nimine bond
˚
30,32
˚
˚
length of 2.3046(18) A is 0.05-0.10 A longer.
These
comparisons indicate that the iron center sits closer to the
central heterocycle but that the distance to the imine arms is
longer, presumably due to an inherently larger exocyclic angle
{C(7)-C(8)-S(8) or C(7A)-C(8A)-C(10A)} for ligands of
this type, as compared to those with a six-membered central
Synthesis of Complexes. The chromium complexes, 2a-d,
were isolated as green solids in near quantitative yield from
the reaction between tris(tetrahydrofuran)chromium(III)
chloride and the ligands (1a-d) at elevated temperature in
toluene or chlorobenzene (Scheme 5). Elemental analysis
was consistent with the proposed empirical formula, and
FAB-MS analysis revealed the [M - Cl]þ fragment in all
cases. The magnetic moments measured for 2a-d are con-
sistent with an octahedral, high-spin chromium(III) envir-
onment (2a, μeff=3.83 μB; 2b, μeff=3.82 μB; 2c, μeff=3.59 μB;
2d, μeff=3.88 μB; OH, Cr(III), S=3/2, μeff(obs)=3.7-3.9 μB).
The reaction of 1f under the same conditions as for 2a-d
gave the corresponding chromium complex, 2f, as an orange
powder in 93% yield (μeff = 3.45 μB). The ability of 1f to
successfully bind chromium, given the failure to ligate iron, is
surprising and remains without good explanation, the ionic
6
˚
ring. The imine bond length observed of 1.290(3) A is equi-
pollent with that seen in the bis(imino)pyridine analogues and
is indicative of conjugation between the imine and the central
heterocycle.30
Attempts to ligate iron with ligand 1e and 1f all failed,
specifically, heating toluene and THF solutions of the ligand
with iron(II) bromide and bis(acetonitrile)iron(II) triflate.
(30) Britovsek, G. J. P.; Gibson, V. C. G.; Kimberley, B. S.; Maddox,
P. J.; McTavish, S. J.; Solan, G. A.; White, A. J. P.; Williams, D. J. Chem.
Commun. 1998, 849.
˚
˚
(31) (a) Bianchini, C.; Giambastiani, G.; Guerrero, I. R.; Meli, A.;
Passaglia, E.; Gragnoli, T. Organometallics 2004, 23, 6087. (b) Bianchini,
C.; Mantovani, G.; Meli, A.; Migliacci, F.; Zanobini, F.; Laschi, F.;
Sommazzi, A. Eur. J. Inorg. Chem. 2003, 1620. (c) Barbaro, P.; Bianchini,
C.; Giambastiani, G.; Guerreo Rios, I.; Meli, A.; Oberhauser, W.; Segarra
A. M.; Sorace, L.; Toti, A. Organometallics 2007, 26, 4639. (d) Bianchini,
C.; Giambastiani, G.; Guerreo Rios, I.; Meli, A.; Oberhauser, W.; Sorace, L.;
Toti, A. Organometallics 2007, 26, 5066.
radii of Cr(III) [0.62 A] and Fe(II) [0.64-0.78 A] being rather
similar.29 Elemental analysis was as expected, but the
FAB-MS analysis revealed only a very weak signal for the
[M-Cl]þ fragment.
Heating toluene solutions of the ligands, 1a and 1c, and
iron(II) bromide at 60 °C furnished the iron complexes (3a
(32) Tellmann, K. P.; Gibson, V. C. G.; White, A. J. P.; Williams
(29) Shannon, R. D. Acta Crystallogr. 1976, A32, 751.
D. J. Organometallics 2005, 24, 280.

4856 Organometallics, Vol. 28, No. 16, 2009
Scheme 6. Synthesis of Iron and Cobalt Complexes
Tenza et al.
In the case of 1e, presumably in-plane coordination of the
ligand is blocked due to the presence of the N-methyl group,
precluding even the bidentate coordination mode seen for
the pyrrolide derivatives.6 In the case 1f, the failure to ligate
iron corresponds with previous reports where the 2,5-
bis[(2,6-diisopropylphenyl)iminomethyl]thiophene variant
failed to ligate iron.4 Heating THF solutions of the ligands
1g-j and iron(II) bromide at 60 °C furnished the iron
complexes in high yields as tan (3g), ochre (3h), and purple
(3i-j) colored solids.
Crystals of 3g grown from a MeOH solution layered with
petroleum ether 40-60 facilitated an X-ray diffraction
study. The geometry at iron can be considered as a strongly
distorted trigonal bipyramid, with the ligand 1g bound in a
tridentate fashion, as expected, occupying the two axial and
one equatorial position (Figure 2). An examination of the
metric parameters for 3g reveals it to be essentially isostruc-
tural with the molecular structure determined for the corre-
sponding FeCl2 complex.23
analysis and FAB-MS gave results consistent with the pro-
posed formula, and magnetic moments consistent with a
high-spin cobalt(II) five-coordinate center were observed
(4a, μeff=4.44 μB; 4c, μeff=4.81 μB),33 corresponding with
that expected for a planar tridentate ligand bound to a cobalt
dichloride moiety.4,31b,32 Crystals of 4a grown from a DCM
solution layered with hexane facilitated an X-ray diffraction
analysis. As can be seen from Figure 3, ligand 1a is again
bound in a planar tridentate coordination mode via the two
imine arms and the sp2-hybridized thiazole N (sum of angles
at N = 359.9°). This structure does not exhibit the C2
crystallographic symmetry observed for 3a, and so the
various metric parameters of the thiazole ring may be mean-
ingfully examined. A comparison of the N(8)-C(8) and
N(8)-C(18) bond lengths reveals that the first, formally a
double bond, is shorter than the latter, which is formally a
single bond; however, the difference in these bond lengths is
(33) (a) Morassi, R.; Bertini, I.; Sacconi, L. Coord. Chem. Rev. 1973,
11, 343. (b) Sacconi, L.; Morassi, R.; Midollini, S. J. Chem. Soc. A 1968,
1510. (c) Sacconi, L. J. Chem. Soc. A 1970, 248. (d) Morassi, R; Sacconi, L.
J. Chem. Soc. A 1971, 492.
The cobalt complexes, 4a and 4c, were isolated as bright
green solids in excellent yield from the reaction of the ligands
1a and 1c and cobalt(II)chloride in THF at 60 °C. Elemental

Article
Organometallics, Vol. 28, No. 16, 2009 4857
Figure 3. X-ray molecular structure of 4a. The molecule is
disordered, and the major (75%) orientation is illustrated. The
DCM solvate molecule has been omitted for clarity. Selected
Figure 1. X-ray molecular structure of 3a. The complex has a
crystallographic 2-fold axis running through the Fe-N(8) axis;
the C3NS ring is thus disordered, and only one of the 50%
occupancy models is illustrated. The toluene solvate molecule has
˚
bond lengths (A) and angles (deg): Co(1)-Cl(1) 2.246(2), Co(1)-
Cl(2) 2.233(2), Co(1)-N(8) 2.001(5), Co(1)-N(1) 2.395(5), Co-
(1)-N(11) 2.305(5), N(8)-C(8) 1.311(9), N(8)-C(18) 1.344(8),
S(8)-C(8) 1.680(7), S(8)-C(10) 1.598(5), C(10)-C(18) 1.576(8),
N(1)-C(7) 1.289(8), (N11)-C(17) 1.277(9); N(8)-Co(1)-N(1)
71.31(19), N(8)-Co(1)-N(11) 74.1(2), N(1)-Co(1)-N(11)
145.18(19), N(8)-Co(1)-Cl(1) 125.44(17), N(8)-Co(1)-Cl(2)
118.25(17), Co(1)-N(8)-C(8) 124.6(5), Co(1)-N(8)-C(18)
121.2(4), C(8)-N(8)-C(18) 114.1(6), Cl(2)-Co(1)-Cl(1)
115.78(8).
˚
been omitted for clarity. Selected bond lengths (A) and angles
(deg): Fe(1)-Br(1) 2.4303(4), Fe(1)-Br(1A) 2.4303(4), Fe(1)-
N(8) 2.050(3), Fe(1)-N(1) 2.3046(18), Fe(1)-N(1A) 2.3047(18),
N(1)-C(7) 1.290(3); N(8)-Fe(1)-N(1) 71.46(5), N(8)-Fe(1)-
N(1A) 71.46(5), N(1)-Fe(1)-N(1A) 142.93(9), N(8)-Fe(1)-Br-
(1A) 121.805(12), N(1)-Fe(1)-Br(1A) 103.05(5), N(1A)-Fe-
(1)-Br(1A) 96.27(5), N(8)-Fe(1)-Br(1) 121.803(12), N(1)-
Fe(1)-Br(1) 96.27(5), N(1A)-Fe(1)-Br(1) 103.05(5), Br(1A)-
Fe(1)-Br(1) 116.39(2), C(7)-N(1)-Fe(1) 117.23(15).
considered as indicative of conjugation between the imine
functionalities and the central thiazole ring. Indeed, an
examination of the structure reveals that the entire ligand
scaffold (thiazole and imine moieties) and iron center are
coplanar, with the imine substituents sitting out of plane
by 72° and 51°.
Ethylene Oligomerization Results. All of the complexes
(10 μmol of precatalyst) were screened for ethylene oligo-
merization with modified methylalumoxane (MMAO-3A)17
and AlEt3/[Ph3C][Al(OtBuF)4] (TEA/TA). The chromium
complexes of the bis(imino)thiazole, 2a-d, and bis-
(imino)thiophene, 2f, ligands facilitated ethylene oligomer-
ization with moderate activity with either activation method
(Table 1, entries 1-25) and were thus selected for further
study. In contrast the iron and cobalt complexes of the
bis(imino)thiazole ligand were less successful (Table 1,
entries 26-33), showing low productivities and short life-
times, with the heaviest oligomers observed being C8; in all
but one case there was no concomitant polymer formation.
The bis(pyrazole)pyridine and bis(benzimidazole)pyridine
complexes of iron, 3g-j, all initiated ethylene oligomeriza-
tion (Table 2), but the productivities were exceptionally low,
the catalyst lifetime was very short, and no material heavier
than C6 was produced. In all cases only liquid product was
formed, no polymer being evident. Some of the catalysts
3g-j were totally selective toward dimerization, and the
selectivity to R-olefins was high in all cases, but this is often
seen with poorly active catalysts and is somewhat artificial.
A closer examination of the chromium complexes begin-
ning with 2a reveals that this complex, when activated with
MMAO-3A, is primarily a polymerization system (Table 1,
entry 1). A repeat of this test (Table 1, entry 2), but under
more concentrated conditions and higher temperature (30 °C
vs 20 °C), induces a further shift toward polymerization and
achieves a higher productivity and longer catalyst lifetime,
Figure 2. X-ray molecular structure of 3g. The half-weight
methanol solvate molecule has been omitted for clarity. Selected
bond lengths (A) and angles (deg): Fe(1)-N(9) 2.147(5), Fe(1)-
N(15) 2.174(5), Fe(1)-N(1) 2.178(5), Fe(1)-Br(2) 2.4317(11),
Fe(1)-Br(1) 2.4935(12), N(1)-N(2) 1.377(6), N(14)-N(15)
1.371(7); N(9)-Fe(1)-N(15) 72.84(18), N(9)-Fe(1)-N(1)
72.68(17), N(15)-Fe(1)-N(1) 145.06(19), N(9)-Fe(1)-Br(2)
121.32(12), N(9)-Fe(1)-Br(1) 122.19(12), Br(2)-Fe(1)-Br(1)
116.48(4).
˚
˚
only 0.03 A, while the C(18)-C(10) bond length is very long
for a C(sp2)-C(sp2) double bond. Furthermore, as with the
structure of 3a, the imine bond lengths [N(1)-C(7) and
N(11)-C(17)] and [C(7)-C(8) and C(17)-C(18)] can be

4858 Organometallics, Vol. 28, No. 16, 2009
Tenza et al.
Table 1. Ethylene Oligomerization Results Obtained with Bis(imino)thiazole- and Bis(imino)thiophene-Based Catalystsa
based
on Al
based on transition
metal
total
C4
C6
C8
C10
C12
cat
cocat b
(equiv)
T
PE
liq product
(1-C4) C14þ
{wt %} {wt %} {wt %} {wt %} {wt %} {wt %}
(1-C6)
(1-C8)
(1-C10
)
(1-C12
)
entry {μmol}
{min} prodc actd
prode
actf
{%}
{%}
{g}
kg
1
2
2a (10) MMAO (500)
16
54
110 430
140 160
56 700 212 610
70 670 78 840
68.8
31.2 15.91 0.63
5.7 19.83 0.73
89.7
(99.1)
65.8
6.6
(37.8)
9.2
1.1
(59.8)
4.0
0.7
(46.1)
3.1
0.4
1.5
(63.7)
2ah,i (10) MMAO
94.3
63.1
52.5
80.2
77.2
3.4 14.5
(500)
(98.4)
(62.2)
7.4
(89.0)
9.3
(88.5)
7.7
(72.3)
8.4 59.0
3
4
5
6
2a (10) TEA/TA
16
29
41
510 1920
830 1750
600 880
51 320 192 430
83 110 174 980
120 400 175 770
146 410 732 070
36.9 14.40 0.64, 0.90 8.3
(100/1.5)
(99.6)
(83.5)
13.2
(84.7)
9.7
(84.2)
5.9
(90.8)
5.2 32.1
2ai (2.5) TEA/TA
47.5
19.8
5.83 0.56, 0.96 34.0
(100/1.5)
(99.5)
(87.2)
6.7
(83.5)
6.4
(83.7)
4.7
(89.4)
5.1 35.1
2ai (2.5) TEA/TA
8.44 0.57, 0.95 42.1
(100/1.5)
(99.6)
(79.1)
12.7
(79.6)
12.4
(82.8)
9.4
(90.6)
8.8 42.8
2ai,j (2.5) TEA/TA
12 1460 7320
22.8 10.26 0.65, 0.81 13.9
(100/1.5)
(98.8)
(97.5)
11.2
(97.2)
5.4
(98.0)
2.7
(98.6)
1.4
7
8
2ai (2.5) TEA/TB (100/1.5) 35
800 1370
550 2870
79 980 136 720
54 930 286 570
54.1
69.3
45.9
30.7
5.61 0.53, 0.56 76.0
(99.5)
3.3
0.7
(92.4)
2.7
(96.0)
0.3
(95.9) (>99.9)
0.2 0.2
2ai (2.5) TEA/ B(C6F5)3
12
3.85 0.53, 0.59 95.8
(100/1.5)
(99.8)
(80.4) (>99.9) (>99.9) (>99.9)
9
2b (10) MMAO (500)
3
3
6
<10 150
80 1650
3790 75 760
7770 164 510
8920 89 170
20 590 112 290
28 440 56 890
45 250 125 690
98.1
77.8
89.7
64.3
76.0
84.1
1.9
22.2
10.4
35.7
24.0
15.9
1.06
-
100
0
0
0
0
0
(100)
10 2b (10) TEA/TA (100/1.5)
11 2c (10) MMAO (500)
2.18 0.70
2.50 0.57
3.72 0.51
3.99 0.54
6.35 0.55
30.0
21.8
(91.9)
30.8
12.0
(83.3)
12.7
5.9
(83.7)
4.0
3.8 26.5
(81.7)
2.2
(89.7)
45.5
20 180
4.9
4.2
9.2
(99.4)
52.7
(92.3)
14.7
(89.7)
12.6
(80.6)
9.3
(52.9)
6.4
12 2c (10) TEA/TA (100/1.5) 11
13 2ci (5) TEA/TA (100/1.5) 30
210 1120
280 570
450 1260
(86.6)
33.7
(88.8)
30.0
(88.1)
16.2
(63.7)
7.4
(79.5)
3.5
(98.3)
32.1
(97.9)
26.0
(97.0)
16.0
(97.9) (>99.9)
14 2ch,i (5) TEA/TA
(100/1.5)
22
8.6
4.8 12.5
(98.9)
82.5
(98.4)
17.5
(97.1)
0
(97.9)
0
(97.1)
15 2d (10) MMAO (500)
16 2d (10) TEA/TA (100/1.5)
17 2f (10) MMAO (500)
18 2fh,i (5) MMAO (500)
19 2f (10) TEA/TA (100/1.5)
4
5
<10
30
1160 15 870
2580 30 990
86.7
72.1
22.8
11.0
12.3
25.8
24.2
2.3
13.3
27.9
77.2
89.0
87.7
74.2
75.8
0.33
0.73 0.70
0.16
-
0
0
(93.5)
21.3
(80.3)
19.7
30 310
15.2
(89.2)
0
11.0
(91.3)
0
8.7 24.1
(92.2)
(68.7)
98.2
(84.5)
1.8
2
<10
20
20
30
290
8670
-
0
0
(96.5) (>99.9)
32
5
9050 17 030
9890 114 830
1.27 0.43
2.77 0.43
1.43 0.39
0.52 0.39
40.0
(99.3)
36.8
28.0
(93.7)
29.9
15.0
(96.9)
16.8
8.1
(93.1)
8.3
4.6
(83.5)
4.0
4.4
4.2
2.5
2.8
9.7
100 1150
100 310
80 220
(>99.9)
41.4
(96.9)
30.3
(96.2)
15.1
(96.7)
7.4
(97.0)
3.2
20 2fi (5)
TEA/TA (100/1.5) 20
10 190 30 580
7470 22 410
(98.0)
41.3
(98.2)
30.0
(97.1)
14.9
(85.0)
7.6
(82.7)
3.3
21 2fi (2.5) TEA/TA (100/1.5) 20
(96.4)
26.0
(97.0)
27.1
(94.5)
18.3
(77.8)
11.8
(72.5)
7.1
22 2fh,i (5) TEA/TA (100/1.5) 170 2410 850
23 2fi,k (5) TEA/TA (100/1.5) 106 1570 880
24 2fh,i (5) TEA/TB (100/1.5) 198 2910 880
240 880 84 920
156 470 88 880
291 200 88 240
38 090 58 610
97.7 33.79 0.51
99.1 21.95 0.55
98.0 40.85 0.53
(99.6)
23.3
(95.5)
24.4
(93.8)
18.0
(82.4)
12.5
(76.4)
0.9
8.2 13.6
(81.9)
8.0 11.7
(73.1)
(99.7)
24.5
(96.5)
25.6
(94.5)
18.0
(86.1)
12.1
2.0
(99.6)
7.2
(95.5)
6.6
(93.7)
6.8
(82.0)
6.5
25 2fh,i (5) TEA/ B(C6F5)3
39
380 590
96.5
3.5
5.34 0.89
7.7 65.3
(100/1.5)
(78.7)
90.4
(41.2)
7.4
(24.6)
2.2
(7.9)
0
(8.4)
26 3a (10) MMAO (500)
27 3a (10) TEA/TA (100/1.5)
28 3c (10) MMAO (500)
29 3c (10) TEA/TA (100/1.5)
30 4a (10) MMAO (500)
31 4a (10) TEA/TA (100/1.5)
2
2
3
2
2
5
<10
<10
<10
<10
<10
<10
80
70
20
30
30
5
1270 38 010
0
0
0
0
0
0
100
0.36
0.07
0.14
0.02
0.15
0.01
-
-
-
-
-
-
0
0
0
0
0
0
0
0
0
0
0
0
(35.8)
99.4
(92.4) (100)
230
510 10 250
80 2521
550 16 470
50 550
6920
100
100
100
100
100
0.6
0
0
0
0
0
0
0
0
(90.3) (100)
78.5
(64.1)
81.9
21.5
(99.3)
18.1
(71.4)
20.8
(97.3)
0
(80.4)
73.8
5.4
(94.3)
0
(89.3)
100
(65.5)

Article
Organometallics, Vol. 28, No. 16, 2009 4859
Table 1. Continued
based
on Al
based on transition
metal
total
C4
C6
C8
C10
C12
cat
entry {μmol}
cocat b
(equiv)
T
PE
liq product
(1-C4) C14þ
{wt %} {wt %} {wt %} {wt %} {wt %} {wt %}
(1-C6)
(1-C8)
(1-C10
)
(1-C12
)
{min} prodc actd
prode
actf
{%}
{%}
{g}
0.20
0.84
kg
32 4c (10) MMAO (500)
3
1
2
<10
30
700 13 900
2980 178 670
0
100
-
-
72.3
(84.2)
100
17.5
(0)
0
10.2
(0)
0
0
0
0
0
0
33 4c (10) TEA/TA (100/1.5)
34 Crh,l (5) TEA/TA (100/1.5)
30 1790
200 5860
830 2000
99.4
2.8
0.7
0
(61.4)
35.9
19 520 585 480
83 400 200 170
327 170 109 057
18 965 568 950
97.2
2.74 0.41
31.1
(97.8)
28.9
17.3
(96.2)
17.0
8.2
(97.1)
9.1
3.9
(95.4)
4.7
3.5
4.6
8.9
3.2
(97.7)
35.7
35 Crh,l (5) TEA/TA (100/1.5) 25
1.9
98.1 11.70 0.43
99.3 45.89 0.52
(99.9)
26.8
(98.2)
27.8
(98.6)
18.3
(94.3)
11.4
(91.0)
6.7
36 Crh,l (5) TEA/TA (100/1.5) 180 3270 1090
37 Crl (5) TEA/TA (100/1.5)
190 5690
0.75
48.9
(99.7)
36.7
(96.2)
32.6
(93.9)
17.0
(85.0)
7.2
(79.9)
3.3
2
51.1
2.66 0.41
(95.3)
(96.1)
(93.2)
(95.5)
(87.9)
a General conditions: toluene (120 mL), p(ethene)=8 bar, T=20 °C. b MMAO = MMAO-3A; TEA/TA = Et3Al/[Ph3C][Al(OtBuF)4]; TEA/TB=
Et3Al/[Ph3C][B(C6F5)4]. c (mol ethylene) (mol Al)-1 d (mol ethylene) (mol Al)-1 h-1 e (mol ethylene) (mol TM)-1. f (mol ethylene) (mol TM)-1 h-1 g k=
.
.
.
MolCnþ2/MolCn; averaged k value over the carbon ranges C6-C20, if fractions up to or over C20 were present, otherwise averaged over C6-Cmax. For 2a, two
k values are quoted kC4nþ2/C4n and kC4n/C4nþ2, respectively. h T=30 °C. i Toluene (50 mL). j 2a and TEA were stirred for 30 s in a Schlenk; then TA was added
and the mixture stirred for a further 30 s, before the whole volume was withdrawn and added to the autoclave. k T=40 °C. l Precatalyst=(THF)3CrCl3.
Table 2. Ethylene Oligomerization Results Obtained with Bis(pyrazole)pyridine and Bis(benzimidazole)pyridine Iron Complexes 3g-ja
based on
based on Al transition metal
total
product
{g}
cocat b
(equiv)
T
PE liq
{%} {%}
C4 (1-C4)
{wt %}
C6 (1-C6)
{wt %}
C8þ
{wt %}
entry cat
{min} prod c act d prod e act f
1
2
3
4
5
6
7
8
3g
3g
3h
3h
3i
MMAO (500)
6
1
4
4
2
2
3
1
<10
<10
<10
<10
<10
<10
<10
<10
<10
10
150
20
1510
760
0
0
0
0
0
0
0
0
100
0.09
81.0
(93.8)
96.3
(73.2)
100
(>99.9)
100
(66.0)
100
(>99.9)
100
(87.1)
84.9
(96.8)
100
(85.3)
19.0
(84.6)
3.7
0
0
0
0
0
0
0
0
TEA/TA (100/1.5)
MMAO (500)
100
100
100
100
100
100
100
0.01
0.07
0.01
0.02
0.01
0.03
0.01
(>99.9)
<10
<10
<10
<10
<10
10
120
20
1840
250
0
0
0
0
TEA/TA (100/1.5)
MMAO (500)
40
1290
240
3i
TEA/TA (100/1.5)
MMAO (500)
20
3j
40
880
15.1
(63.6)
0
3j
TEA/TA (100/1.5)
20
660
a General conditions: 20 μmol of precatalyst complex, toluene (120 mL), p(ethene) = 8 bar, T = 20 °C. b MMAO = MMAO-3A; TEA/TA =
Et3Al/[Ph3C][Al(OtBuF)4]. c (mol ethylene) (mol Al)-1 d (mol ethylene) (mol Al)-1 h-1 e (mol ethylene) (mol TM)-1. f (mol ethylene) (mol TM)-1 h-1
.
.
.
but a lower activity than previously. In both cases a normal
Schulz-Flory distribution of oligomers is observed within
the liquid fraction (kC6-C20=0.63 and 0.73). Catalysis using
2a activated with TEA/TA (Table 1, entry 3) gave a catalyst
with the same lifetime as, and very similar productivity to
that when activated with MMAO-3A. Furthermore, a very
similar selectivity between polymer and liquid oligomers was
observed; however the distribution of the oligomers is strik-
ingly different. From the GC-FID trace recorded it can be
clearly seen that two apparently overlapping distributions
are discerned (Figure 4).
A distribution of higher R-olefins such as this has only
been reported on one previous occasion, also with a Cr-based
initiator featuring a tridentate N,N,N-donor ligand, and
again a similar split between polymer and liquid products
was observed.34 Previously it was shown that the C4n series
(C8, C12, C16, etc.) followed a Schulz-Flory distribution,
while the C4nþ2 series (C10, C14, C18, etc.) deviated from this,
exhibiting what appears to be a Poisson distribution. In
contrast, the results obtained with 2a indicate that both the
C4n (C4, C8, C12, C16, etc.) and C4nþ2 (C6, C10, C14, C18, etc.)
series show a Schulz-Flory distribution (Figure 5). On this
occasion a number of different k values can be specified as
follows (Figure 5). First, a pair of k values may be defined as
proposed by Gibson et al.,34 one for each of the two
distributions kC4n (i.e., kC8/C4, kC12/C8, etc.) and kC4nþ2 (i.e.,
kC10/C6, kC14/C10, etc.), which give a measure of the type of
distribution within each separate series.35,36 Second, we pro-
pose another pair of k values that compares the extent of the
(35) As the two series appear potentially independent of each other
[one could be a Poisson distribution and the other a Schulz-Flory
distribution, as was seen by Gibson et al., or as observed here], both
series could be of the same class (Schulz-Flory), but with different rates
of decay as you move to higher carbon numbers.
(34) Tomov, A. K.; Chirinos, J. J.; Long, R. J.; Gibson, V. C.;
Elsegood, M. R. J. J. Am. Chem. Soc. 2006, 128, 7704–7705.
(36) Flory, P. J. J. Am. Chem. Soc. 1940, 62, 1561.

4860 Organometallics, Vol. 28, No. 16, 2009
Tenza et al.
Figure 4. GC-FID trace recorded for catalysis using 2a activated with TEA/TA (Table 1, entry 6).
Figure 5. Plot of mol % of each carbon fraction for Table 1, entry 6, annotated to show the specification of different k values.
alternation between the two series, kC4nþ2/C4n (i.e., kC6/C4
,
overheated from 20 to 38 °C,37 which likely explains the
higher polymer formation in this run. Notably, both entries
4 and 5 again showed the unusual distribution of olefins;
however both produced a significantly disproportionate
amount of C4 and a slightly disproportionate amount of C6.
Due to the changes in catalyst loading and solvent volume,
the aluminum concentration in entries 4 and 5 was roughly
half that for entry 3. Thus, the increased C4 and C6
formation is postulated to result from incomplete or slower
alkylation of the catalyst, giving a situation where some of
the chromium is producing light oligomers with a normal
Schulz-Flory distribution (small k value), while the re-
mainder is fully activated, producing the alternating dis-
tribution observed. In order to solve this, a test was
performed where the catalyst was preactivated under con-
kC10/C8, etc.) and kC4n/C4nþ2 (i.e., kC8/C6, kC12/C10, etc.). Indeed,
a chart of the kC4nþ2/C4n and kC4n/C4nþ2 Schulz-Flory
constants for the fractions clearly reveals the extent of the
alternation (Figure 6 and ESI), while the various k values
specified are tabulated for each run with 2a in Table 3. As can
be seen from Tables 1 and 3, the C4n and C4nþ2 series both show
Schulz-Flory distributions that decay at almost the same rate
(kC4n and kC4nþ2), but start at different initial levels, leading to
the alternation observed, which is then quantified by kC4nþ2/
C4n and kC4n/C4nþ2 (Figure 6). It is noted that in all cases the
kC4nþ2 series decays slightly more slowly than the kC4n series.
Catalysis with 2a and TEA/TA was subsequently re-
peated at a lower catalyst loading (2.5 μmol) and in half
the solvent volume (Table 1, entries 4 and 5). A similar
activity was observed to the test at high precatalyst loading,
but the productivities observed were increased. In the case
of entry 5, temperature control was lost and the reaction
(37) In all other runs reported rigorous temperature control ensured a
maximum deviation of 3 °C from the set temperature.

Article
Organometallics, Vol. 28, No. 16, 2009 4861
Figure 6. Bar chart of Schulz-Flory constants for each fraction for Table 1, entry 6.
showed similar selectivity between solid and liquid products
to when TEA/TA was used, the selectivity within the
liquid product was more heavily biased toward the lighter
fractions.
Table 3. Tabulation of the Various k Values Specified for
Catalysis with 2aa
cocatb
entry
cat
(equiv)
kC4n kC4nþ2 kC4nþ2/C4n kC4n/C4nþ2
Catalysis results for 2b reveal a higher level of polymer
formation as compared to liquid oligomer than for 2a, as is
expected due to the increased steric bulk at the imine
substituents. However, in contrast to 2a, 2b shows essentially
no liquid fraction with MMAO-3A (Table 1, entry 9) and a
normal Schulz-Flory distribution of olefins within the
liquid fraction when activated with TEA/TA (Table 1, entry
10). The selectivity within each carbon fraction is typically
80-90% linear R-olefin, with the remainder being comprised
approximately equally of linear internal olefins and
branched olefins. Considering precatalyst 2c, which now
features the thiazole in an S-bound configuration, a lower
productivity and activity is seen in comparison to 2a under
similar conditions, and whether activated with MMAO-3A
or TEA/TA a normal Schulz-Flory distribution of higher R-
olefins is observed (Table 1, entries 11-14). Precatalyst 2d
showed a significant drop in productivity and activity com-
pared to either 2b or 2c and did not show such a dramatic
shift toward polymer formation compared to 2c, as 2b
compared to 2a.
The observation of the unusual alternating product dis-
tribution with precatalyst 2a in conjunction with TEA and a
weakly coordinating anion, but not with MMAO-3A or any
of precatalysts 2b-d, clearly suggests that a very specific
environment at the chromium center is required. An expla-
nation for related unusual behavior was proposed by Gibson
et al., who suggested that if a metallocyclic mechanism is in
action, then the growing metallocycle could occupy two
distinct sites; the nonplanarity of the central nitrogen donor
in their ligand is cited as giving rise to two such differentiated
conformers.34 Indeed, if the rates of insertion into, or
elimination from, these two conformers varied and inter-
1
2
3
4
5
6
2ac (10) TEA/TA (100/1.5)
2a (2.5) TEA/TA (100/1.5)
2a (2.5) TEA/TA (100/1.5)
2ad (2.5) TEA/TA (100/1.5)
2a (2.5) TEA/TB (100/1.5)
0.59 0.63
0.56 0.59
0.56 0.62
0.53 0.55
0.37 0.40
0.64
0.56
0.57
0.65
0.53
0.53
0.90
0.96
0.95
0.81
0.56
0.59
2a (2.5) TEA/B(C6F5)3 (100/1.5) 0.41 0.47
a General conditions: toluene (50 mL), p(ethene)=8 bar, T=20 °C.
b TEA/TA = Et3Al/[Ph3C][Al(OtBuF)4]; TEA/TB = Et3Al/[Ph3C][B-
(C6F5)4]. c Toluene (120 mL). d 2a and TEA were stirred for 30 s in a
Schlenk; then TA was added and the mixture stirred for a further 30 s,
before the whole volume was withdrawn and added to the autoclave.
centrated conditions in a Schlenk, before being transferred
to the autoclave (Table 1, entry 6). This gave a catalyst that
showed the highest productivity and activity seen for 2a
and restored the consistency of the alternating olefin
distribution without disproportionately large C4 and C6
fractions. It can be seen from Table 1 that the selectivity for
the linear R-olefin within each carbon fraction is
high, the remaining mass in each fraction arising from
branched olefins, presumably due to secondary incorpora-
tion of primary products (R-olefins), and in the case of
entry 6 (see also Figure 4) never drops below 97% for any
fraction.
The observation that a single precatalyst gives this unusual
distribution of products when activated with TEA/TA but
not with MMAO-3A prompted us to consider the behavior
of other activation vectors. Thus activation using
AlEt3/[Ph3C][B(C6F5)4] (TEA/TB) and TEA/B(C6F5)3 was
examined (Table 1, entries 7 and 8), and again the unusual
product distribution was noted, although it was less pro-
nounced than with TEA/TA, and a disproportionately
large amount of C4 was formed. Although both catalysts

4862 Organometallics, Vol. 28, No. 16, 2009
Tenza et al.
Scheme 7. Mechanism Based on Two Differentiated Propagat-
ing Sitesa
the possibility of exchange between the two metallocycle
conformers is most likely energetically possible, but the im-
portant consideration is whether this process is compe-
titive with termination and/or propagation. To probe this,
a statistical model was developed accounting for this exchange,
and again a reasonable correlation with experiment could
be achieved, but the relative rate of exchange between
conformers (kE,A and kE,B) was very small compared to the
relative rates of termination and propagation, suggesting that
for 2a/TEA/TA this process is quite minor (full details of this
more complex model are including in the Supporting In-
formation).
It is noted, however, that many unsymmetrical oligomer-
ization precatalysts have been reported previously and only
that of Gibson et al. and 2a herein show this behavior,
suggesting that a very specific environment at chromium
must exist for such a pathway to be followed. Crucially, while
an unsymmetrical chromium center is most likely necessary,
it is in itself not the only criteria for this behavior. This
statement is supported by the observations that 2a has an
unsymmetrical chromium center whether activated by
MMAO-3A or TEA/TA, yet the latter shows the alternat-
ing distribution, while the former does not. With catalyst
2b, where the phenyl substituents on the imine moieties
have been increased in steric bulk to 2,6-diisopropylphe-
nyl, the liquid fraction also no longer shows the alternating
distribution of products. One could speculate that the
extra bulk around the metal center now forces all propaga-
tion down a single pathway, the other conformer now
becoming energetically inaccessible. Likewise, with pre-
catalysts 2c,d, which are also unsymmetrical but with
sulfur bound in place of nitrogen, the delicate balance
required for this alternating propagation is presumably
disrupted.
Initial screening of 2f with MMAO-3A revealed a poorly
productive and active catalyst that predominantly produced
liquid oligomers with a Schulz-Flory distribution (Table 1;
entry 17, 10 μmol; entry 18, 5 μmol); the formation of both
liquid and polymer products under ethylene oligomerization
conditions with a single precatalyst is commonly observed.18
However, when activated with TEA/TA under similar con-
ditions (Table 1; entry 19, 10 μmol; entry 20, 5 μmol; entry 21,
2.5 μmol), the activity, if not the productivity, was signifi-
cantly increased. A further test was undertaken at the slightly
higher temperature of 30 °C (Table 1, entry 22), and this
introduced a step change in performance: a nearly 3-fold
increase in activity was observed as compared to the analo-
gous test at 20 °C, and the lifetime of the catalyst appeared
greatly enhanced, leading to a 23-fold increase in productiv-
ity. Notably, the level of polymer formation also dropped to
2.3%, and there was a shift in k value from 0.39 at 20 °C to
0.51 at 30 °C. A further test at 40 °C (Table 1, entry 23)
showed no further enhancement in activity and a shorter
catalyst lifetime, although still greatly improved over 20 °C.
The level of polymer formation dropped even further to
0.9% and the k value increased slightly to 0.55. In all cases,
the selectivity within each carbon fraction was respectively
high to the linear R-olefin, with almost no linear internal
olefin being observed, the remaining mass being accounted
for by branched products, most likely formed through the
secondary incorporation of primary product (R-olefin)
(Figure 9).
a Favored pathways are indicated by solid arrows, disfavored by
dashed arrows.
conversion between conformers was energetically possible,
then two different product distributions could result. Here,
this type of mechanistic explanation is possible, especially as
the ligand of 2a is clearly unsymmetrical, something that
could thus also give rise to differentiated sites at the metal
center. We propose a similar mechanism for the alternating
distributions observed with 2a (Scheme 7) and have also
undertaken some statistical modeling to test the validity of
the mechanism with the data from Table 1, entries 3 and 6.
The correlation between the experimental and modeled
product distributions is shown graphically for Table 1,
entries 3 and 6, in Figures 7 and 8. A good fit with the
experimental data was achieved with similar parameters in
both cases (see Supporting Information for full details).
There are similarities and differences between the model
invoked herein and that devised by Gibson et al.34 In both
cases one conformer of the initially formed metallacyclopen-
tane must be preferential to the other, but whereas this
preference had to be exclusively for one conformer (A) in
the previous work, herein a more modest ∼70:30 split (A:B)
was required. Considering the rate of termination in both
cases, this occurs more preferentially from the conformer
type (A) that dominates the initial metallacyclopentane
intermediate, but again whereas in the previous work this
must be an exclusive preference, in this case elimination from
either form of the conformer is possible, just to different
extents with kT,A dominating over kT,B. Interestingly,
whereas propagation from either conformer was equally
likely in the previous work, herein, propagation is favored
relative to termination to a greater extent for conformer B
compared conformer A. Most significantly, the distribution
observed with 2a can be adequately modeled without the
need to invoke interconversion between the two metallocycle
conformers, in contrast to the previous case.38 Nonetheless,
(38) Indeed, when such modeling is studied carefully, the need to
invoke interconversion between the two conformers is only a necessity to
describe the situation when the two distributions are of a different type,
i.e., one Schulz-Flory and one Poisson.
The significant enhancement in activity and productivity
upon moving from 20 to 30 °C caused us to speculate with

Article
Organometallics, Vol. 28, No. 16, 2009 4863
Figure 7. Graphical representation of the correlation between the experimental and modeled distributions for the C4n and C4nþ2 series
for Table 1, entry 3.
Figure 8. Graphical representation of the correlation between the experimental and modeled distributions for the C4n and C4nþ2 series
for Table 1, entry 6.
regard to the underlying reason. One explanation is im-
proved solubility of the catalyst species, but given the dilute
nature of the catalytic conditions and the step change in
performance this seems unlikely, especially given that a
homogeneous solution was observed upon opening the
autoclave after runs with 2f at all temperatures. Another
explanation that bears some consideration is that of ligand
lability; if the increase in temperature caused one of the
ligand arms to become highly labile, or indeed the entire
ligand to dissociate, then possibly a more active catalyst may
result. In order to probe this hypothesis, catalysis was
undertaken with (THF)3CrCl3 in order to examine ethylene
oligomerization catalysis under these conditions with unli-
gated chromium (Table 1, entries 34-37).39 A comparison of
entry 36 with entry 22, catalysis with (THF)3CrCl3 and 2f
under identical conditions for equal durations, reveals a very
similar selectivity within the liquid fraction (k=0.51 vs 0.52)
and also a similar split between liquid product and polymer
formation. Furthermore a similar activity and productivity
are achieved. However, it was noted from the rate of ethylene
uptake that 2f gave a catalyst that was very stable with time,
whereas (THF)3CrCl3 had a very high initial rate, which
decayed with time. This later observation is supported by
comparison of entry 36 (180 min duration) with entries 34
and 35 (stopped after 2 and 25 min); it can be clearly seen that
a high initial activity is achieved but not sustained. Further-
more, given the poor performance of 2f at 20 °C, a run was
undertaken with (THF)3CrCl3 under identical conditions
(Table 1, entry 37). Interestingly, although a much higher
level of polymer formation was observed with (THF)3CrCl3
at 20 °C as compared to 30 °C, mirroring the observation
with 2f, and the liquid fraction selectivity was similar at 20 °C
to that with 2f (k=0.39 vs 0.41), it is noted that the activity
and productivity are much higher with (THF)3CrCl3 than 2f
at 20 °C (even accounting for the shorter run duration).
Taking into account all of the above observations at both
temperatures, it seems reasonable to postulate that under the
ethylene oligomerization conditions employed the active
catalytic species may well be an unligated chromium species.
(39) The assumption is made that after addition of TEA the three
THF ligands will be sequestered by the alkylaluminium reagent. Indeed,
the ability of the alkylaluminium reagents to readily sequester THF has
been shown previously: Derouault, J; Granger, P.; Forel, M. T. Inorg.
Chem. 1977, 16, 3214.

4864 Organometallics, Vol. 28, No. 16, 2009
Tenza et al.
Figure 9. Portion of representative GC-FID trace for precatalyst 2f activated with TEA/TA.
Thus, the improved activity and productivity at 30 °C may
arise from a more complete dissociation of the ligand from
2f. However, it is interesting to note an overall more stable
catalyst when 2f is employed as compared to (THF)3CrCl3,
suggesting that perhaps a ready equilibrium exists between
free and bound ligand 1f, thus stabilizing the active chro-
mium species. One further piece of evidence regarding the
possible ligand lability was the extremely weak signal ob-
tained by FAB-MS for [M - Cl]þ as compared to the other
chromium complexes reported herein.
To further explore catalysis with 2f, activation with TEA/
TB and TEA/B(C6F5)3 was examined under optimum
conditions (50 mL of solvent, 5 μmol of precatalyst, 30
°C). Both activation vectors successfully initiated catalysis,
with TB performing almost exactly the same as TA in all
respects, while the borane gave significantly lower produc-
tivity and activity and curiously showed an inversion in
selectivity between solid and liquid products, producing
96.5% polymer.
catalyst species may in fact be unligated chromium. The
ability of 3a,c to oligomerize ethylene serves to demons-
trate successfully for the first time the replacement of the
six-membered N-heterocycle ring of the bis(imino)pyridine
ligand with a five-membered N-heterocycle while maintain-
ing an active oligomerization catalyst with iron. The ability
of 2a to oligomerize ethylene to produce an alternating
distribution of higher R-olefins with two distinct Schulz-
Flory constants has been observed on only one previous
occasion to our knowledge, and then a slightly different
behavior was observed from the case seen here. Mechan-
istically, the unsymmetrical nature of the ligand is proposed
to give rise to differentiated sites at the metal center, which
lead to this unusual product distribution. Clearly, a deeper
understanding of the pathways that lead to this product
distribution could lead to the development of new selective
oligomerization technologies, and studies are in progress to
probe the mechanism of catalysis by 2a more thoroughly.
Experimental Section
Conclusions
Transition metal halide salts and trityl tetrakis-
(pentafluorophenyl)borate (TB) were purchased from Strem,
all organic precursors {including 2,6-bis(2-benzimi-
dazolyl)pyridine} from Aldrich and MMAO-3A,17 and TEA
from Akzo Nobel; all materials were used as received. Trityl
tetrakis(perfluoro-tert-butoxy)aluminate (TA),40 2,6-bis(3,5-di-
methylpyrazol-1-yl)pyridine (1g),19 2,6-bis(3,5-diphenylpyra-
zol-1-yl)pyridine (1h),20 2,5-bis(acetyl)-1-methylpyrrole,26,27
and 2,5-bis(carboxaldehyde)thiophene41 were prepared by lit-
erature procedures or modifications thereof. Solvents were
procured from Aldrich, purified using a solvent purification
system, and deoxygenated prior to use. All operations were
conducted under inert atmosphere using an argon-filled glove-
box and standard Schlenk techniques. Elemental analyses were
A new class of ligand has been reported in the form of
bis(imino)thiazoles and their coordination chemistry ex-
plored with chromium, iron, and cobalt. Considered in the
context of previous work with bis(imino)pyrrolide ligands
and that with bis(imino)-1-methylpyrrole reported herein,
when designing new variants of the bis(imino)pyridine
ligand, the importance of maintaining an sp2-hybridized N
is shown to be crucial when moving away from pyridine to
other sizes of N-heterocyclic rings. Despite the failure in
ligating iron, the ability of bis(imino)thiophene to complex
chromium has also been demonstrated.
All of the complexes prepared have been screened for
ethylene oligomerization in combination with MMAO-3A
and TEA/TA, with the chromium complexes of 2a-d and 2f
showing notable productivity and activity, although in the
case of 2f, further investigations revealed that the active
(40) Krossing, I.; Brands, H.; Feuerhake, R.; Koenig, S. J. Fluorine
Chem. 2001, 112, 83.
(41) Chmielewski, P. J.; Latos-Grazynski, L.; Oldstead, M. M.;
Balch, A. L. Chem.;Eur. J. 1997, 3, 268.

Article
Organometallics, Vol. 28, No. 16, 2009 4865
performed by the Science Technical Support Unit of London
Metropolitan University. Mass spectrometric analysis (FABþ)
was conducted by the Department of Chemistry Mass Spectro-
metry Service, University of Leicester. NMR spectra were
collected on a Bruker Avance 400 spectrometer and chemical
shifts referenced to the solvent. All 13C{1H} NMR spectra were
collected as DEPT135 experiments to assist in assignment.
Magnetic moments were determined using a Gouy balance
and/or the Evans NMR method.42
cold methanol to give the desired product as a yellow solid (1.800 g,
96%). Anal. Calcd for C19H17N3S (found): C, 71.44 (71.59);
H, 5.36 (5.43); N, 10.04 (10.14). H NMR (400.1 MHz, CDCl3,
1
300 K) δH (ppm): 2.32 (3H, s, CH3), 2.35 (3H, s, CH3),
6.80-6.83 (4H, m, N-C6H5), 7.06-7.11 (2H, m, N-C6H5), 7.28-
7.34 (4H, m, N-C6H5), 8.17 (1H, s, CH-5). 13C{1H} NMR (100.6
MHz, CDCl3, 300 K) δC (ppm): 17.0 (s, CH3), 18.5 (s, CH3), 119.9
(s, N-C6H5), 120.1 (s, N-C6H5), 124.2 (s, C-5), 124.6 (s, N-C6H5),
124.8 (s, N-C6H5), 129.2 (s, N-C6H5), 149.7 (s, N-ipso-C6H5),
156.0 (s, C-4), 160.1 (s, CdN), 161.8 (s, CdN), 169.7 (s, C-2). EI
mass spectrum, m/z: 319 [M]þ, 304 [M - Me]þ, 242 [M - Ph]þ.
2,4-Bis[1-({2,6-diisopropylphenyl}imino)ethyl]thiazole (1b). TiCl4
(2.7 mL, 2.66 mmol; 1.0 M solution in toluene) was added to a
solution of 2,4-diacetylthiazole (0.429 g, 2.54 mmol) and DippNH2
(3.4 mL, 3.152 g, 17.8 mmol) in DCM (60 mL) at RT. The resulting
brown-colored suspension was stirred at RT for 18 h. The solvent
volume was reduced to ∼20 mL, and diethyl ether (200 mL) added
to precipitate the titanium salts. The mixture was filtered to leave a
yellow solution, and the volatiles were removed in vacuo to give an
impure yellow solid. Repeated extractions with MeOH/cyclohex-
ane-hexanes, combination of the hexane fractions, washing with
MeOH, and removal of solvent in vacuo gave the product as a
yellow powder (0.492 g, 40%). Anal. Calcd for C311H41N3S (found):
C, 76.34 (76.27); H, 8.47 (8.55); N, 8.62 (8.56). H NMR (400.1
MHz, CDCl3, 300 K) δH (ppm): 1.17-1.21 (24H, m, (CH3)2CH),
2.27 (3H, s, CH3), 2.29 (3H, s, CH3), 2.80 (4H, sept, 3JHH=6.7 Hz,
(CH3)2CH), 7.12-7.21 (6H, m, N-{2,6-(Pri)2-C6H3}), 8.31 (1H, s,
CH-5). 13C{1H} NMR (100.6 MHz, CDCl3, 300 K) δC (ppm): 17.7
(s, CH3), 19.2 (s, CH3), 23.0 (s, (CH3)2CH), 23.2 (s, (CH3)2CH),
23.3 (s, (CH3)2CH), 23.4 (s, (CH3)2CH), 28.5 (s, (CH3)2CH), 28.7
(s, (CH3)2CH), 123.2 (s, N-{2,6-(Pri)2-C6H3}), 123.3 (s, N-{2,-
6-(Pri)2-C6H3}), 123.5 (s, N-{2,6-(Pri)2-C6H3}), 123.9 (s, N-{2,-
6-(Pri)2-C6H3}), 124.5 (s, C-5), 136.2 (s, N-{2,6-(Pri)2-C6H3}),
136.4 (s, N-{2,6-(Pri)2-C6H3}), 145.2 (s, N-{2,6-(Pri)2-ipso-
C6H3}), 146.2 (s, ipso-C N-{2,6-(Pri)2-ipso-C6H3}), 157.3 (s, C-4),
161.8 (s, CdN), 162.0 (s, CdN), 169.4 (C-2). EI mass spectrum,
m/z: 487 [M]þ, 742 [M - Me]þ, 444 [M - iPr]þ.
2,5-Bis[1-(phenylimino)ethyl]thiazole (1c). The same proce-
dure as for 1a was used, except with 2,5-diacetylthiazole (0.800
g, 4.73 mmol), PhNH2 (3.0 mL, 3.083 g, 33.1 mmol), and TiCl4
(4.7 mL, 4.73 mmol; 1.0 M solution in toluene) in DCM
(100 mL). This gave the product as a yellow solid (1.22 g,
81%). Anal. Calcd for C19H17N3S (found): C, 71.44 (71.33);
H, 5.36 (5.61); N, 10.04 (10.15). 1H NMR (400.1 MHz, CDCl3,
300 K) δH (ppm): 2.27 (3H, s, CH3), 2.38 (3H, s, CH3), 6.80-
6.89 (4H, m, N-C6H5), 7.12-7.14 (2H, m, N-C6H5), 7.33-7.38
(4H, m, N-C6H5), 8.18 (1H, s, CH-4). 13C{1H} NMR (100.6
MHz, CDCl3, 300 K) δC (ppm): 16.8 (s, CH3), 18.1 (s, CH3),
120.1 (s, N-C6H5), 120.2 (s, N-C6H5), 124.5 (s, N-C6H5), 124.9
(s, N-C6H5), 129.4 (s, N-C6H5), 143.9 (s, C-4), 145.3 (s, N-ipso-
C6H5), 149.9 and 150.3 (s, N-ipso-C6H5 and C-5), 159.3 (s,
CdN), 161.9 (s, CdN), 173.1 (s, C-2). EI mass spectrum, m/z:
319 [M]þ, 304 [M-Me]þ, 242 [M - Ph]þ.
Catalysis was performed in 250 mL volume Buchi miniclaves
equipped with stainless steel vessels with integral thermal-fluid
jackets, internal cooling coils, and mechanical mixing via gas-
entraining stirrers. Ethylene (Grade 4.5) was supplied by Linde
and passed through oxygen and moisture scrubbing columns
prior to use; ethylene flow was measured using a Siemens Sitrans
F C Massflo system (Mass 6000-Mass 2100) and the data were
logged. All catalytic tests were allowed to run until ethylene
uptake had ceased. Gas phase sample: GC-FID analysis was
performed using an Agilent Technologies 6890N GC system
equipped with a 250 μL gas sample loop and GS GasPro column
(30 m ꢀ 0.32 mm) using hydrogen as carrier gas. Liquid phase
sample: GC-FID analysis was performed using an Agilent
Technologies 6850N GC system equipped with a PONA column
(50 m ꢀ 0.20 mm ꢀ 0.50 μm) using hydrogen as carrier gas.
Hydrogenative GC-FID analysis was performed on liquid phase
samples (to confirm the assignment of branched and linear
products) using an Agilent Technologies 6890N GC system
equipped with an inlet line packed with hydrogenating catalyst
(Pt on Chromosorb W at 200 °C) and PONA column (50 m ꢀ
0.20 mm ꢀ 0.50 μm).43 GC-MS analysis (to confirm the assign-
ment of branched and linear products) was performed on an
Agilent Technologies 6890N GC system equipped with a PONA
column (50 m ꢀ 0.20 mm ꢀ 0.50 μm), coupled to an Agilent
Technologies 5973N MSD mass spectrometric instrument with
EI source.
2,5-Bis(acetyl)-1-methylpyrrole. 1-Methyl-2-acetylpyrrole (10.00 g,
81.2 mmol), trifluoroacetic anhydride (25.00 g, 119 mmol), and
acetic acid (7.30 g, 122 mmol) were stirred at RT for 6 weeks. After
this time a saturated aqueous solution of sodium bicarbonate was
added dropwise over 20 min until the mixture was alkaline to litmus
(note: quick addition results in significant frothing). The organics
were extracted into ethyl acetate (6 ꢀ 70 mL), dried over MgSO4, and
filtered, and solvent was removed in vacuo. The resultant black
viscous oil was triturated with hot distilled water (3ꢀ 40 mL) for
2 h. The aqueous extracts were combined and extracted with ethyl
acetate (4 ꢀ 50 mL). The combined organic fractions were dried over
MgSO4, filtered, and concentrated in vacuo. The resulting light brown
solid was dissolved in acetone and purified on silica gel eluting with
diethyl ether/petroleum ether 40-60 (3:1). After removal of solvent
the product was isolated as a very pale brown solid (3.432 g, 26%).
Anal. Calcd for C9H11NO2 (found): C, 65.44 (65.48); H, 6.71 (6.61);
N, 8.48 (8.38). 1H NMR (400.1 MHz, CDCl3, 300 K) δH (ppm):2.51
(6H, s, CH3), 4.24 (3H, s, N-CH3), 6.89 (2H, s, CH-3 and CH-4).
13C{1H} NMR (100.6 MHz, CDCl3,300K) δC (ppm):28.5(s, CH3),
35.4 (s, N-CH3), 117.4 (s, C-3 and C-4), 135.0 (s, C-2 and C-5), 190.5
(s, CdO).
2,5-Bis[1-({2,6-diisopropylphenyl}imino)ethyl]thiazole (1d). The
same procedure as for 1b was used, but with 2,5-diacetylthiazole
(0.266 g, 1.57 mmol), DippNH2 (2.1 mL, 1.948 g, 11.0 mmol), and
TiCl4 (1.7 mL, 1.65 mmol; 1.0 M solution in toluene) in DCM
(50 mL). This gave the product as a brown oil (0.251 g, 33%).
Anal. Calcd for C31H41N3S (found): C, 76.34 (76.21); H, 8.47
(8.56); N, 8.62 (8.89). 1H NMR (400.1 MHz, CDCl3, 300 K) δH
(ppm): 1.17-1.20 (24H, m, (CH3)2CH), 2.16 (3H, s, CH3), 2.25
(3H, s, CH3), 2.80 (4H, d, 3JHH=6.2 Hz, (CH3)2CH), 7.11-7.21
(6H, m, N-{2,6-(Pri)2-C6H3}), 8.24 (1H, s, CH-4). 13C{1H} NMR
(100.6 MHz, CDCl3, 300 K) δC (ppm): 17.5 (s, CH3), 18.5 (s,
CH3), 23.0 (s, (CH3)2CH), 23.1 (s, (CH3)2CH), 23.3 (s, (CH3)2-
CH), 23.4(s, (CH3)2CH), 28.6(s, (CH3)2CH), 28.7 (s, (CH3)2CH),
123.2 (br s, N-{2,6-(Pri)2-C6H3}), 124.2 (s, N-{2,6-(Pri)2-C6H3}),
124.5 (s, N-{2,6-(Pri)2-C6H3}), 136.1 (s, N-{2,6-(Pri)2-C6H3}),
136.4 (s, N-{2,6-(Pri)2-C6H3}), 143.5 (s, C-4), 145.0 (s, N-{2,-
6-(Pri)2-ipso-C6H3}), 145.2 (s, ipso-C N-{2,6-(Pri)2-ipso-C6H3}),
2,4-Bis[1-(phenylimino)ethyl]thiazole (1a). TiCl4 (5.9 mL, 5.91
mmol; 1.0 M solution in toluene) was added to a solution of 2,4-
diacetylthiazole (1.000 g, 5.91 mmol) and PhNH2 (3.8 mL, 3.856
g, 41.4 mmol) in DCM (100 mL) at RT. The resulting brown-
colored suspension was stirred at RT for 6 h, and then diethyl
ether (200 mL) added to precipitate the titanium salts. The
mixture was filtered to leave a yellow solution, and the volatiles
were removed in vacuo to give a yellow oil. This was treated with
(42) (a) Evans, D. F. J. Chem. Soc. 1959, 2003. (b) Loliger, J.;
Scheffold, R. J. Chem. Educ. 1972, 49, 646. (c) Sur, S. K. J. Magn. Reson.
1989, 82, 169.
(43) Kuzmenko, T. E.; Samusenko, A. L.; Uralets, V. P.; Golovnya,
R. V. J. High Res. Chrom. 1979, 2, 43.

4866 Organometallics, Vol. 28, No. 16, 2009
Tenza et al.
145.4 (s, C-5), 159.2 (s, CdN), 162.0 (s, CdN), 172.7 (C-2). EI
mass spectrum, m/z: 487 [M]þ, 742 [M - Me]þ, 444 [M-Pri]þ.
2,5-Bis[1-(phenylimino)ethyl]-1-methylpyrrole (1e). TiCl4
(0.73 mL, 0.73 mmol; 1.0 M solution in toluene) was added
dropwise to a flask containing the 2,5-bis(acetyl)-1-methylpyr-
role (60 mg, 0.36 mmol) and PhNH2 (0.33 mL, 339 mg, 3.64
mmol) in DCM (10 mL) at 0 °C. After addition was complete,
the ice bath was removed and the reaction stirred at RT for 5 h
to leave a yellow solution with a white precipitate. Diethyl ether
(20 mL) was added to ensure full precipitation of titanium salts,
and the solution filtered. The solvent was removed under
reduced pressure to leave the product as a yellow solid, which
was washed with cold MeOH (∼5 mL), dried in vacuo, and then
washed with diethyl ether (2 ꢀ 10 mL) and dried in vacuo to leave
the product as a yellow crystalline solid (101 mg, 88%). Anal.
Calcd for C21H21N3 (found): C, 79.97 (79.73); H, 6.71 (6.81); N,
13.32 (13.29). 1H NMR (400.1 MHz, CDCl3, 300 K) δH (ppm):
2.21 (6H, s, CH3), 4.39 (3H, s, N-CH3), 6.73 (2H, s, CH-3, CH-
4), 6.79-6.81 (4H, m, N-C6H5), 7.07-7.12 (2H, m, N-C6H5),
7.34-7.39 (4H, m, N-C6H5). 13C{1H} NMR (100.6 MHz,
CDCl3, 300 K) δC (ppm): 26.0 (s, CH3), 120.0 (s, C-3, C-4),
123.7 (s, N-C6H5), 127.1 (s, N-C6H5), 129.1 (s, N-C6H5), 138.6
(s, C-2, C-5), 153.6 (s, N-ipso-C6H5), 156.6 (s, CdN). EI mass
spectrum, m/z: 315 [M]þ, 300 [M-Me]þ, 238 [M - Ph]þ.
2,5-Bis(phenyliminomethyl)thiophene (1f). PhNH2 (4.391 g,
47.2 mmol) was added to a flask containing the 2,5-bis-
(carboxaldehyde)thiophene (3.000 g, 21.4 mmol) at RT. This
resulted in an exothermic reaction with liberation of water vapor
over 1 min to give a hard yellow solid. The solid was crushed and
washed with methanol (4 ꢀ 50 mL), then with DCM (1 ꢀ 10
mL), in which it was sparingly soluble. After drying in vacuo a
bright yellow solid was recovered (6.010 g, 97%). Anal. Calcd
for C18H14N2S (found): C, 74.45 (74.37); H, 4.86 (4.78); N, 9.65
(9.74). 1H NMR (400.1 MHz, CDCl3, 300 K) δH (ppm): 7.23-
7.28 (6H, m, N-C6H5), 7.38-7.43 (4H, m, N-C6H5), 7.50 (2H, s,
CH-3, CH-4), 8.58 (1H, s, NdCH). 13C{1H} NMR (100.6 MHz,
CDCl3, 300 K) δC (ppm): 121.3 (s, C-3, C-4), 126.8 (s, N-C6H5),
129.4 (s, N-C6H5), 131.9 (s, N-C6H5), 146.5 (s, N-ipso-C6H5),
151.1 (s, C-2, C-5), 152.5 (NdCH).
mmol) in DMF (40 mL) at RT. After 30 min MeI (0.80 mL,
1.822 g, 12.84 mmol) was added dropwise, and the mixture was
stirred at RT for 16 h and then quenched with ethanol (1.0 mL).
Distilled water was added and an extraction performed using
ethyl acetate (3 ꢀ 100 mL). The combined organic fractions were
dried (MgSO4), and solvent was removed in vacuo to leave a
white solid (2.121 g, 97%). Anal. Calcd for C21H17N5 (found):
C, 74.32 (74.30); H, 5.05 (5.11); N, 20.63 (20.58). 1H NMR
(400.1 MHz, CDCl3, 300 K) δH (ppm): 4.24 (6H, s, CH3), 7.33-
7.49 (6H, m, CH), 7.86-7.90 (2H, m, CH), 8.06 (1H, t, 3JHH
=
3
7.9 Hz, CH), 8.43 (2H, d, JHH = 7.9 Hz, CH). 13C{1H}
NMR (100.6 MHz, CDCl3, 300 K) δC (ppm): 32.7 (s, CH3),
110.2 (s, CH), 120.3 (s, CH), 123.2 (s, CH), 123.9 (s, CH), 125.6
(s, CH), 137.3 (s, C), 138.3 (s, CH), 142.5 (s, C), 149.6 (s, C),
150.3 (s, C).
2,4-Bis[1-(phenylimino)ethyl]thiazole Chromium Trichloride
(2a). A toluene (60 mL) solution of 1a (100 mg, 0.31 mmol)
and (THF)3CrCl3 (117 mg, 0.31 mmol) was stirred and heated at
60 °C overnight. A light green solid precipitated. Toluene was
removed by filtration, the solid washed with petroleum ether 40-60
(3 ꢀ 20 mL), and the material dried in vacuo to obtain the product
as a pale green solid (145 mg, 98%). Anal. Calcd for C19H17N3-
SCrCl3 (found): C, 47.76 (47.59); H, 3.59 (3.70); N, 8.79 (8.66).
Magnetic moment: μeff 3.83 μB. MS (FABþ) m/z: 441 [M-Cl]þ.
2,4-Bis[1-({2,6-diisopropylphenyl}imino)ethyl]thiazole Chro-
mium Trichloride (2b). A PhCl solution (40 mL) of 1b (381 mg,
0.78 mmol) and (THF)3CrCl3 (284 mg, 0.76 mmol) was stirred
and heated at 60 °C for 4 h. Some green solid precipitated. The
volume of PhCl was reduced to ∼10 mL in vacuo and petroleum
ether 40-60 (40 mL) added to fully precipitate the product. The
solidwas recovered by filtration, washed with petroleum ether 40-
60 (3 ꢀ 20 mL), and dried in vacuo to obtain the product as a green-
colored powder (481 mg, 98%). Anal. Calcd for C31H41N3SCrCl3
(found): C, 57.63 (57.74); H, 6.40 (6.50); N, 6.50 (6.42). Magnetic
moment: μeff 3.82 μB. MS (FABþ) m/z: 609 [M - Cl]þ.
2,5-Bis[1-(phenylimino)ethyl]thiazole Chromium Trichloride
(2c). The same procedure as for 2a instead using 1c (120 mg,
0.38 mmol), (THF)3CrCl3 (114 mg, 0.38 mmol), and toluene
(20 mL) gave a bright olive-green solid (177 mg, 99%). Anal.
Calcd for C19H17N3SCrCl3 (found): C, 47.76 (47.75); H, 3.59
(3.61); N, 8.79 (8.76). Magnetic moment: μeff 3.59 μB. MS
(FABþ) m/z: 441 [M-Cl]þ.
2,5-Bis[1-({2,6-diisopropylphenyl}imino)ethyl]thiazole Chro-
mium Trichloride (2d). A PhCl solution (40 mL) of 1d (247 mg,
0.51 mmol) and (THF)3CrCl3 (172 mg, 0.46 mmol) was stirred
and heated at 60 °C for 4 h. A brown solution resulted. The
volume of PhCl was reduced to ∼5 mL in vacuo and petroleum
ether 40-60 (40 mL) added to fully precipitate the product. The
solid was recovered by filtration, washed with petroleum ether
40-60 (2 ꢀ 20 mL), and dried in vacuo to obtain the product as a
brown powder (197 mg, 66%). Anal. Calcd for C31H41N3SCrCl3
(found): C, 57.63 (57.86); H, 6.40 (6.59); N, 6.50 (6.71). Mag-
netic moment: μeff 3.88 μB. MS (FABþ) m/z: 609 [M-Cl]þ.
2,5-Bis(phenyliminomethyl)thiophene Chromium Trichloride (2f).
The same procedure as for2a instead using 1d (150 mg, 0.52 mmol),
(THF)3CrCl3 (195 mg, 0.52 mmol), and toluene (50 mL) gave an
orange solid (217 mg, 93%). Anal. Calcd for C18H14N2SCrCl3
(found): C, 48.18 (48.29); H, 3.14 (3.25); N, 6.24 (6.16). Magnetic
moment: μeff 3.45 μB. MS (FABþ) m/z: 412 [M-Cl]þ.
2,6-Bis(3,5-dimethylpyrazol-1-yl)pyridine (1g). 3,5-Dimethyl-
pyrazole (1.370 g, 14.4 mmol) was treated with K (0.560 g, 14.4
mmol) in diglyme (200 mL) at 70 °C for 16 h (until all the metal
had been consumed). 2,6-Dibromopyridine (1.682 g, 7.1 mmol)
was added, and the mixture heated to 110 °C for 3 days. Water
was added to quench the reaction and the mixture concentrated
in vacuo; more water was added when necessary to help remove
all the solvent. The residue was dissolved in methanol, and then
water was added, resulting in a precipitate, which was recovered
by filtration. The solid was dissolved in DCM, hexane added,
and the DCM allowed to evaporate slowly to give the product as
white crystals (1.063 g, 56%). Anal. Calcd for C15H17N5
1
(found): C, 67.39 (67.46); H, 6.41 (6.41); N, 26.20 (26.09). H
NMR (400.1 MHz, CDCl3, 300 K) δH (ppm): 2.29 (6H, s, CH3),
2.58 (6H, s, CH3), 5.99 (2H, s, CH), 7.66-7.69 (2H, m, C5H3N),
7.85-7.89 (1H, m, C5H3N). 13C{1H} NMR (100.6 MHz, CDCl3,
300 K) δC (ppm): 13.4 (s, CH3), 14.0 (s, CH3), 108.9 (s, CH), 113.6
(s, CH), 140.4 (s, CH), 141.0 (s, C), 150.0 (s, C), 151.4 (s, C).
2,6-Bis(3,5-diphenylpyrazol-1-yl)pyridine (1h). The same pro-
cedure as for 1g instead using 3,5-diphenylpyrazole (2.000 g,
9.08 mmol), K (0.354 g, 9.08 mmol), and 2,6-dibromopyridine
(1.070 g, 4.54 mmol) in diglyme (200 mL) gave the product as
yellow crystals (1.590 g, 68%). Anal. Calcd for C35H25N5
2,4-Bis[1-(phenylimino)ethyl]thiazole Iron Dibromide (3a). A
toluene (80 mL) suspension of 1a (200 mg, 0.63 mmol) and
FeBr2 (135 mg, 0.63 mmol) was stirred and heated at 60 °C
overnight, resulting in a dark green precipitate. The solvent was
removed under reduced pressure, and the solid was washed with
petroleum ether 40-60 (3 ꢀ 50 mL) and dried in vacuo to leave
the product as a dark orange-brown solid (331 mg, 99%). Anal.
Calcd for C19H17N3SFeBr2 (found): C, 42.65 (42.75); H, 3.20
(3.32); N, 7.85 (7.79). Magnetic moment: μeff 3.96 μB
1
(found): C, 81.53 (81.47); H, 4.89 (4.86); N, 13.58 (13.62). H
NMR (400.1 MHz, CDCl3, 300 K) δH (ppm): 7.29-7.34 (15H,
m, ArH), 7.41-7.69 (10H, m, ArH). 13C{1H} NMR (100.6
MHz, CDCl3, 300 K) δC (ppm): 100.0 (s, CH), 125.6 (s, CH),
128.2 (s, CH), 128.7 (s, CH), 128.8 (s, CH), 131.2 (s, C).
2,6-Bis(1-methyl-2-benzimidazolyl)pyridine (1j). 2,6-Bis(2-
benzimidazolyl)pyridine (2.000 g, 6.42 mmol) in DMF (60
mL) was slowly added to a suspension of NaH (0.308 g, 12.84
2,5-Bis[1-(phenylimino)ethyl]thiazole Iron Dibromide (3c). The
same procedure as for 3a instead using 1c (200 mg, 0.63 mmol),

Article
Organometallics, Vol. 28, No. 16, 2009 4867
FeBr2 (135 mg, 0.63 mmol), and toluene (70 mL) gave a dark
green-brown solid (312 mg, 93%). Anal. Calcd for C19H17-
N3SFeBr2 (found): C, 42.65 (42.52); H, 3.20 (3.26); N, 7.85
(7.74). Magnetic moment: μeff 3.52 μB. MS (FABþ) m/z: 454
[M-Br]þ.
the reactor. The vessel was vigorously stirred and the tempera-
ture maintained at the desired temperature. MMAO-3A was
then added to the reactor, the vessel immediately charged with 8
bar of ethylene, and the pressure kept constant throughout the
reaction by the continuous addition of ethylene, which was
monitored via a flow-meter. Once ethylene uptake had ceased,
the gas supply was closed and the reactor cooled to 5 °C.
The reactor was carefully vented, with a portion of the vent
gas being directly fed to a GC-FID instrument equipped with
gas-sampling loop. The reactor contents were treated sequen-
tially with 1000 μL of nonane (GC internal standard), MeOH,
and 10% HCl. A sample of the organic phase was taken for GC-
FID analysis. Any solid formed was collected, washed repeat-
edly with 10% HCl and EtOH, dried overnight, and weighed.
Typical Catalytic Run with TEA/TA. An autoclave was
heated to 90 °C under vacuum for 30 min. After back-filling
with Ar, the precatalyst as a solution in toluene (100 mL) and
TA (solution in toluene) were added to the reactor. The vessel
was vigorously stirred and the temperature maintained at the
desired temperature. AlEt3 (solution in toluene) was then added
to the reactor, the vessel immediately charged with 8 bar of
ethylene, and the pressure kept constant throughout the reac-
tion by the continuous addition of ethylene, which was mon-
itored via a flow-meter. Once ethylene uptake had ceased, the
gas supply was closed and the reactor cooled to 5 °C. The reactor
was carefully vented, with a portion of the vent gas being directly
fed to a GC-FID instrument equipped with gas-sampling loop.
The reactor contents were treated sequentially with 1000 μL of
nonane (GC internal standard), MeOH, and 10% HCl. A
sample of the organic phase was taken for GC-FID analysis.
Any solid formed was collected, washed repeatedly with 10%
HCl and EtOH, dried overnight, and weighed.
Crystallographic Data. X-ray diffraction studies were per-
formed using Mo KR radiation. Data for 3a and 3g were
collected at 93K using a Rigaku MM007/Saturn/Mercury dif-
fractometer (confocal optics Mo KR radiation), while data for
4a were collected at 125 K using the St Andrews robotic
diffractometer (Rigaku sealed tube, shine monochromator,
Saturn724 CCD with robotic sample changing and auto-
centering). Intensity data were collected using narrow width ω
steps accumulating area detector frames spanning at least a
hemisphere of reciprocal space for all structures (data were
integrated using the CrystalClear program). All data were
corrected for Lorentz, polarization, and long-term intensity
fluctuations. Absorption effects were corrected on the basis of
multiple equivalent reflections or by semiempirical methods.
Structures were solved by direct methods and refined by full-
matrix least-squares against F2 (SHELXTL).45 All hydrogen
atoms were assigned riding isotropic displacement parameters
and constrained to idealized geometries.
2,6-Bis(3,5-dimethylpyrazol-1-yl)pyridine Iron Dibromide
44
(3g).
A THF (100 mL) suspension of FeBr2 (1.760 g, 8.16
mmol) and 1g (2.180 g, 8.16 mmol) was stirred at 60 °C for 6 h.
After cooling to RT the resulting tan-colored precipitate was
recovered by filtration, washed with diethyl ether (3 ꢀ 20 mL),
and dried in vacuo to leave a tan-colored powder (3.102 g, 79%).
Anal. Calcd for C15H17N5FeBr2 (found): C, 37.30 (37.19); H,
3.55 (3.52); N, 14.50 (14.36). Magnetic moment: μeff 6.46 μB. MS
(FABþ) m/z: 483 [M]þ, 402 [M - Br]þ.
2,6-Bis(3,5-diphenylpyrazol-1-yl)pyridine Iron Dibromide (3h).
A THF (100 mL) suspension of FeBr2 (0.421 g, 1.95 mmol) and
1h (1.000 g, 1.95 mmol) was stirred at 60 °C for 16 h. After
cooling to RT no precipitate resulted, so the THF was removed
in vacuo and the material dissolved in toluene (150 mL) and
heated to reflux for 16 h. After cooling, the volume was reduced
in vacuo; then excess petroleum ether 40-60 was added, pre-
cipitating an ochre-colored powder. The solid was recovered by
filtration, washed with petroleum ether 40-60 (3 ꢀ 20 mL), and
dried in vacuo to leave an ochre-colored powder (1.216 g, 85%).
Anal. Calcd for C35H25N5FeBr2 (found): C, 57.49 (57.43); H,
3.45 (3.37); N, 9.38 (9.26). Magnetic moment: μeff 4.96 μB.
2,6-Bis(2-benzimidazolyl)pyridine Iron Dibromide (3i). 1i
(1.000 g, 3.21 mmol) and FeBr2 (623 mg, 3.21 mmol) in THF
(50 mL) were stirred at RT for 1 h, giving a purple suspension,
which was then heated to 60 °C for 4 h. After cooling to RT, the
solid was recovered by filtration, washed with diethyl ether (3 ꢀ
20 mL), and dried in vacuo to leave a purple powder (1.58 g, 93%).
Anal. Calcd for C19H13N5FeBr2 (found): C, 43.30 (43.53); H,
2.49 (2.77); N, 13.29 (13.51). Magnetic moment: μeff 2.90 μB. MS
(FABþ) m/z: 446 [M - Br]þ.
2,6-Bis(1-methyl-2-benzimidazolyl)pyridine Iron Dibromide
(3j). 1j (200 mg, 0.59 mmol) and FeBr2 (127 mg, 0.59 mmol) in
THF (50 mL) were heated to 60 °C for 4 h. After cooling to RT,
the solid was recovered by filtration, washed with THF (1 ꢀ 20
mL), then diethyl ether (3 ꢀ 20 mL), and dried in vacuo to leave a
gray-purple powder (301 mg, 92%). Anal. Calcd for C21H17-
N5FeBr2 (found): C, 45.44 (45.60); H, 3.09 (3.14); N, 12.62
(12.56). Magnetic moment: μeff 5.50 μB. MS (FABþ) m/z: 474
[M - Br]þ.
2,4-Bis[1-(phenylimino)ethyl]thiazole Cobalt Dichloride (4a).
A THF (30 mL) solution of 1a (150 mg, 0.469 mmol) was added
to CoCl2 (61 mg, 0.469 mmol) and heated to 60 °C for 6 h. After
cooling, hexane (100 mL) was added and the precipitate
collected via filtration, washed with diethyl ether (2 ꢀ 20 mL),
and dried in vacuo to leave the product as a bright green powder
(201 mg, 96%). Anal. Calcd for C19H17N3SCoCl2 (found): C,
50.80 (50.90); H, 3.81 (3.95); N, 9.35 (9.18). Magnetic moment:
eff 4.44 μB. MS (FABþ) m/z: 413 [M-Cl]þ.
2,5-Bis[1-(phenylimino)ethyl]thiazole Cobalt Dichloride (4c).
Acknowledgment. The authors wish to thank Sasol
Technology UK Ltd. and Sasol Technology (Pty) Ltd.
for permission to publish this work.
μ
The same procedure as for 4a instead using 1c (150 mg, 0.469
mmol), CoCl2 (61 mg, 0.469 mmol), and THF (30 mL) gave an
olive-green solid (206 mg, 98%). Anal. Calcd for C19H17N3-
SCoCl2 (found): C, 50.80 (50.92); H, 3.81 (3.90); N, 9.35 (9.22).
Magnetic moment: μeff 4.81 μB. MS (FABþ) m/z: 413 [M-Cl]þ.
Typical Catalytic Run with MMAO. An autoclave was heated
to 90 °C under vacuum for 30 min. After back-filling with Ar,
the precatalyst as a solution in toluene (100 mL) was added to
Supporting Information Available: Crystallographic data in
CIF format for the structures reported; numbering schemes for
the heterocyclic rings; full selectivity data (up to C36); and
charts of k values and weight % of each fraction for every run
with 2a and TEA/TA, TEA/TB, or TEA/B(C6F5)3. This ma-
terial is available free of charge via the Internet at http://pubs.
acs.org.
(44) The FeCl2 and FeI2 analogues of this complex have been
reported previously;23 the FeCl2 complex was prepared in a similar
manner to the FeBr2 complex prepared here.
(45) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112.