High-temperature catalysts for the production of α-olefins based on iron(II) and iron(III) tridentate bis(imino)pyridine complexes with double pattern of substitution: Ortho-methyl plus meta-aryl

Article abstract of DOI:10.1021/om051069o

Four types of bis(imino)pyridine ligands and their corresponding Fe complexes with the same metaaryl pattern were synthesized for high-temperature α-olefin production (A: 9 and 13; B: 18,19, 20 and 21, 22, 23, 24; C: 27 and 30; and D: 33 and 34). Pattern A (9 and 13) consists of four meta-aryl groups without ortho-methyls. The coordination of FeCl ₂ with the ligand [3,5-bis(4-F-Ph)₂-Ph-N=C(Me)-Py-C(Me)=N- Ph-3,5-bis(4-F-Ph)₂] (9) of type A resulted in the isolation of ion-paired complexes [L₂Fe]²⁺FeCl₄^2- 13. Coordination of ligands 18, 19, and 20 of pattern B with two meta-aryl groups on the same side with two ortho-methyls of the imino aryl group afforded the following 1:1 complexes: [{3-(4-F-Ph)-2-Me-Ph-N=C(Me)-Py-C(Me)=N-Ph-2-Me-3- (4-F-Ph)}FeCl₂](21), [{3-(3,5-bis-(CF₃) ₂-Ph)2-Me-Ph-N=C(Me)-Py-C(Me)=N-Ph-2-Me-3-(3,5-bis-(CF ₃)₂-Ph)}FeCl₂] (22), [{3-(3-Me-thiophene-2-yl)- 2-Me-Ph-N-C(Me)-Py-C(Me)-N-Ph-2-Me-3-(3-Me-thiophene-2-yl)}FeCl₂] (23), and [{3-(3,5-bis-(CF₃)₂-Ph)-2-Me-Ph-N=C(Me)-Py-C(Me) =N-Ph-2-Me-3-(3,5-bis-(CF₃)₂-Ph)}FeCl₃] (24). Reduction of trivalent complex 24 in THF solution yields divalent complex 22. Ligand 27 of pattern C contains two meta-aryl groups on the side opposite two ortho-methyls of the imino aryl group and coordinates with FeCl₂ to form complex [{5-(3,5-bis-(CF₃)₂-Ph)-2-Me-Ph-N=C(Me)-Py- C(Me)=N-Ph-2-Me-5-(3,5-bis-(CF₃)₂-Ph)}FeCl₂] (30). Complex 34 of unsymmetrical pattern D has two meta-aryls in one imino aryl group and two ortho-methyls in the second imino aryl group: [{2,6-dimethyl-4-(3,5-bis-(CF₃)₂-Ph)-2-Me-Ph-N=C(Me)-Py- C(Me)=N-Ph-2-Me-3,5-bis(3,5-bis-(CF₃)₂-Ph)}FeCl ₂]. The introduction of aryl groups into the meta positions of all four types of bis(imino)pyridine ligands was accomplished by palladium-catalyzed Suzuki cross-coupling reaction between aryl boronic acid 5, 16, and 17 and the appropriate meta-bromo-substituted bis(imino)pyridine ligands 4, 15, 25, and 32. Di-tert-butyl(2,2-dimethylpropyl)phosphane (7) and benzyl-di-ferf- butylphosphane (28) were used as ligands for the Suzuki coupling. According to X-ray analysis, there are shortenings and therefore strengthening of the axial Fe-N bond lengths (up to 0.02 A) in complex 22 of type B versus the Fe(II) complex without meta-aryl groups [{o-Me-Ph-N=C(Me)-Py-C(Me)=N-Ph-o-Me}FeCl ₂] (1). Complexes of patterns B and D (21, 22, 23, 24, and 34) afforded very active catalysts for the production of α-olefms with more ideal Schultz-Flory distributions of α-olefins and with higher K values than the parent methyl-substituted Fe(II) complex 1 without weta-aryl groups.

Full text of DOI:10.1021/om051069o

2978
Organometallics 2006, 25, 2978-2992
High-Temperature Catalysts for the Production of r-Olefins Based
on Iron(II) and Iron(III) Tridentate Bis(imino)pyridine Complexes
with Double Pattern of Substitution: ortho-Methyl plus meta-Aryl†
Alex S. Ionkin,* William J. Marshall, Douglas J. Adelman, Barbara Bobik Fones,
Brian M. Fish, and Matthew F. Schiffhauer
DuPont Central Research & DeVelopment, Experimental Station, Wilmington, Delaware 19880-0328
ReceiVed December 15, 2005
Four types of bis(imino)pyridine ligands and their corresponding Fe complexes with the same meta-
aryl pattern were synthesized for high-temperature R-olefin production (A: 9 and 13; B: 18, 19, 20 and
21, 22, 23, 24; C: 27 and 30; and D: 33 and 34). Pattern A (9 and 13) consists of four meta-aryl groups
without ortho-methyls. The coordination of FeCl2 with the ligand [3,5-bis(4-F-Ph)2-Ph-NdC(Me)-Py-
C(Me)dN-Ph-3,5-bis(4-F-Ph)2] (9) of type A resulted in the isolation of ion-paired complexes
[L2Fe]2+FeCl4 13. Coordination of ligands 18, 19, and 20 of pattern B with two meta-aryl groups on
2-
the same side with two ortho-methyls of the imino aryl group afforded the following 1:1 complexes:
[{3-(4-F-Ph)-2-Me-Ph-NdC(Me)-Py-C(Me)dN-Ph-2-Me-3-(4-F-Ph)}FeCl2] (21), [{3-(3,5-bis-(CF3)2-Ph)-
2-Me-Ph-NdC(Me)-Py-C(Me)dN-Ph-2-Me-3-(3,5-bis-(CF3)2-Ph)}FeCl2] (22), [{3-(3-Me-thiophene-2-
yl)-2-Me-Ph-NdC(Me)-Py-C(Me)dN-Ph-2-Me-3-(3-Me-thiophene-2-yl)}FeCl2] (23), and [{3-(3,5-bis-
(CF3)2-Ph)-2-Me-Ph-NdC(Me)-Py-C(Me)dN-Ph-2-Me-3-(3,5-bis-(CF3)2-Ph)}FeCl3] (24). Reduction of
trivalent complex 24 in THF solution yields divalent complex 22. Ligand 27 of pattern C contains two
meta-aryl groups on the side opposite two ortho-methyls of the imino aryl group and coordinates with
FeCl2 to form complex [{5-(3,5-bis-(CF3)2-Ph)-2-Me-Ph-NdC(Me)-Py-C(Me)dN-Ph-2-Me-5-(3,5-bis-
(CF3)2-Ph)}FeCl2] (30). Complex 34 of unsymmetrical pattern D has two meta-aryls in one imino aryl
group and two ortho-methyls in the second imino aryl group: [{2,6-dimethyl-4-(3,5-bis-(CF3)2-Ph)-2-
Me-Ph-NdC(Me)-Py-C(Me)dN-Ph-2-Me-3,5-bis(3,5-bis-(CF3)2-Ph)}FeCl2]. The introduction of aryl
groups into the meta positions of all four types of bis(imino)pyridine ligands was accomplished by
palladium-catalyzed Suzuki cross-coupling reaction between aryl boronic acid 5, 16, and 17 and the
appropriate meta-bromo-substituted bis(imino)pyridine ligands 4, 15, 25, and 32. Di-tert-butyl(2,2-
dimethylpropyl)phosphane (7) and benzyl-di-tert-butylphosphane (28) were used as ligands for the Suzuki
coupling. According to X-ray analysis, there are shortenings and therefore strengthening of the axial
Fe-N bond lengths (up to 0.02 Å) in complex 22 of type B versus the Fe(II) complex without meta-aryl
groups [{o-Me-Ph-NdC(Me)-Py-C(Me)dN-Ph-o-Me}FeCl2] (1). Complexes of patterns B and D (21,
22, 23, 24, and 34) afforded very active catalysts for the production of R-olefins with more ideal Schultz-
Flory distributions of R-olefins and with higher K values than the parent methyl-substituted Fe(II) complex
1 without meta-aryl groups.
tathesis. The oligomerization component of SHOP is based on
the neutral NiII complexes bearing bidentate P∧O ligands. The
SHOP process makes 96-98% R-olefins with 2-4% branched
olefins.2b,c The absence of “chain walking” along the polymer
chain with FeII catalysts is responsible for the high-quality of
the R-olefin products. The corresponding NiII catalysts with
R-diimine ligands, for example, offer only 94% selectivity for
linear R-olefins due to the ability of nickel-based catalysts to
chain walk and therefore to produce substantial amounts of
branched R-olefins.3 Currently, there is also a significant effort
by Sasol to achieve selective production of 1-hexene and
1-octene based on CrIII complexes.4
Introduction
FeII catalysts with tridentate bis(imino)pyridine ligands have
been reported to afford R-olefin oligomers with perfect Schultz-
Flory distributions, exceptional purities (97-99%), and high
productivities.1a-d The results exceed values being reported for
catalysts used in current commercial processes, including the
original Ziegler’s process and the Shell Higher Olefin Process
(SHOP).2 SHOP is a very efficient and flexible combination of
multistep reactions: oligomerization, isomerization, and me-
† Dedicated to Dr. Joel D. Citron, a proponent of iron tridentate chemistry,
on the occasion of his 65th birthday. This is DuPont contribution no. 8667.
* To whom correspondence should be addressed. E-mail:
alex.s.ionkin@usa.dupont.com.
There is an advantage, with regard to toxicity, to using iron
instead of nickel or chromium based catalytic systems.5 The
high catalyst activity of FeII Versipol catalysts also allows once-
(1) For initial patents and publications see: (a) Bennett, A. M. A.
(DuPont) PCT Int. Appl. WO9827124 A1, 1998, 68 pp. (b) Brookhart, M.
S.; Small, B. L. PCT Int. Appl. WO 9902472 A1 1999, 54 pp. (c) Small,
B. L.; Brookhart, M. J. Am. Chem. Soc. 1998, 120, 7143. (d) 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.
For recent reviews see: (e) Ittel, S. D.; Johnson, L. K.; Brookhart, M. Chem.
ReV. 2000, 100, 1169. (f) Britovsek, G. J. P.; Gibson, V. C.; Wass, D. F.
Angew. Chem., Int. Ed. 1999, 38, 428. (g) Mecking, S. Coord. Chem. ReV.
2000, 203, 325. (h) Gibson, V. C.; Spitzmesser, S. K. Chem. ReV. 2003,
103, 283.
(2) (a) Vogt, D. In Applied Homogeneous Catalysis with Organometallic
Compounds; Cornils, B., Herrmann, W. Eds.; VCH Publishers: Weinheim,
1996; Vol. 1, p 245. (b) Lutz, E. F. J. Chem. Educ. 1986, 63, 202. (c)
Vogt, D. In Aqueuos-Phase Organometallic Catalysis; Cornils, B., Her-
rmann, W. A., Eds.; VCH Publishers: Weinheim, 1998; p 541.
(3) Killian, C. M.; Johnson, L. K.; Brookhart, M. Organometallics 1997,
16, 2005.
10.1021/om051069o CCC: $33.50 © 2006 American Chemical Society
Publication on Web 05/11/2006

Production of R-Olefins Based on Iron(II)/(III) Complexes
Organometallics, Vol. 25, No. 12, 2006 2979
Scheme 1
through oligomerization, eliminating the need for catalyst recycle
or removal.1e
An early, leading candidate identified was FeII complex 1,
with a tridentate bis(imino)pyridine ligand bearing only one
methyl group in the ortho positions of the imino aryl groups.1c,e
Four new generic patterns of the substitutions in the ligands
of FeII tridentate complexes (Scheme 1) were reasoned to make
complexes more thermally stable by increasing remotely the
steric protection around the metal with a high degree of
precision. All of them have a similar feature: a meta-aryl group.
Additional placement of a sterically bulky group in vacant ortho
positions, close to the catalytically active center, changes the
process completely from ethylene oligomerization to polymer-
ization.6 Pattern A consists of four meta-aryl groups without
ortho-methyls. Pattern B has two meta-aryl groups on the same
side with two ortho-methyls of the imino aryl group. Pattern C
has meta-aryl groups opposite ortho-methyls of the imino aryl
group. Pattern D has two meta-aryls in one imino aryl group
and two ortho-methyls in the second imino aryl group.
The Suzuki cross-coupling reaction offers a very convenient
and reliable synthetic method for the introduction of a variety
of aryl groups.7 However, a preference was given to the aryl
groups with electron-withdrawing substitutents, (e.g., 3,5-bis-
(trifluoromethyl)phenyl and 4-fluorophenyl) because it has been
shown that such groups can control polymer properties ef-
ficiently in NiII catalytic systems, even when placed relatively
far from the catalytically active center.8
Careful analysis of the performance of complex 1 as a
precatalyst over the commercially desirable temperature range
of 100-120 °C revealed some significant shortcomings. For
example, while complex 1 is extremely productive, its lifetime
at 120 °C is only 3 min. Its high productivity is desirable, but,
for injection into a plug-flow reactor, a less active catalyst with
a longer lifetime would be preferred. This would make genera-
tion of an undesirable “hot-spot” at the injection point less likely
and would reduce the number of injection points needed.
Another concern was the product distribution obtained using
the catalyst derived from complex 1. The oligomer distribution
deviated from Schultz-Flory at the low and high MW ends.
For a general commercial application in which C6 to C16
R-olefins are targeted, it is desirable to minimize the production
of butene, in particular, and of oligomers of >C16. They have
limited commercial value and represent an ethylene yield loss.
These species were instead enhanced by the use of complex 1.
The goal of this work was to synthesize and test catalysts with
several alternative substitution patterns on the imino aryl groups
to see if a more thermally stable candidate with a near ideal
Schultz-Flory distribution of oligomers could be discovered.
In this report, we describe the synthetic approach to the above
complexes and their application as high-temperature catalysts
for the production of R-olefins.
Results and Discussion
Synthesis. Complexes of type A, with four meta-aryl groups
without ortho-methyls, were prepared by the following sequence.
The condensation reaction between commercially available 1-(6-
acetylpyridin-2-yl)ethanone (2) and 3,5-dibromo-4-methylphe-
nylamine afforded bis(imino)pyridine 4 with four bromides in
the meta positions of the aryl imino arms. The palladium-
catalyzed Suzuki cross-coupling reaction between 4-fluorophe-
nylboronic acid (5) and 2,6-bis(1-(3,5-dibromo-4-methyl)-
phenylimino)ethyl)pyridine (4) was used to replace the four
bromides with four aryls to form 2,6-bis(1-(3,5-di(4-fluorophe-
nyl)-4-methyl)phenylimino)ethyl)pyridine (9) (Scheme 2). The
(4) (a) Bollmann, A.; Blann, K.; Dixon, J. T.; Hess, F. M.; Killian, E,;
Maumela, H.; McGuinness, D. S.; Morgan, D. H.; Neveling, A.; Otto, S,;
Overett, M.; Slawin, A. M. Z.; Wasserscheid, P.; Kuhlmann, S. J Am. Chem.
Soc. 2004, 126, 14712. (b) Overett, M. J.; Blann, K.; Bollmann, A.; Dixon,
J. T.; Haasbroek, D.; Killian, E.; Maumela, H.; McGuinness, D. S.; Morgan,
D. H. J Am. Chem. Soc. 2005, 127, 10723.
(5) (a) Agarwal, R. K.; Chandra, A.; Fathima, P. A.; Shanti, T. R.; Hatha,
A. A. M. Pollut. Res. 2005, 24, 247. (b) Frangi, C. Tinctoria 2002, 99, 26.
(c) National Emission Standards for Hazardous Air Pollutants: taconite
iron ore processing; Environmental Protection Agency, Emission Standards
Div., U.S. EPA: Research Triangle Park, NC, 2003; Vol. 68, p 61868. (d)
Buzard, G. S.; Kasprzak, K. S. J. EnViron. Pathol., Toxicol. Oncol. 2000,
19, 179.
(6) (a) Schmid, M.; Eberhardt, R.; Klinga, M.; Leskela, M.; Rieger, B.
Organometallics 2001, 20, 2321. (b) Moody, L. S.; MacKenzie, P. B.;
Killian, C. M.; Lavoie, G. G.; Ponasik, J. A.; Smith, T. W.; Pearson, J. C.;
Barrett, A. G. M. U.S. Pat. Appl. US 2002/0049135 A1. (c) Ionkin, A. S.,
Marshall, W. J. Organometallics 2004, 23, 3276.
(7) Miyaura, N.; Yanagi, T.; Suzuki, A. Synth. Commun. 1981, 11, 513.
(8) Zuideveld, M. A.; Wehrmann, P.; Rohr, C.; Mecking, S. Angew.
Chem., Int. Ed. 2005, 43, 869.

2980 Organometallics, Vol. 25, No. 12, 2006
Scheme 2. Synthesis of the Ligands 4 and 9 of Type A
Ionkin et al.
Scheme 3. Synthesis of Fe Complexes 12 and 13 of Type A
catalytic protocol involved Pd2dba3 (6)/di-tert-butyl-(2,2-di-
methylpropyl)phosphane (7) as the catalyst in the presence of
cesium carbonate with 1,4-dioxane as the solvent. Di-tert-butyl-
(2,2-dimethylpropyl)phosphane (7) was synthesized by the
prolonged refluxing of almost equimolar amounts of di-tert-
butylchlorophosphine (10) and neopentylmagnesium chloride
(11) in THF. Sterical bulky tertiary phosphines, such as 7,
bearing at least two tert-butyl groups, were found among the
most efficient ligands for the palladium-catalyzed Suzuki cross-
coupling reaction.9
Unexpectedly the coordination of Fe(II) chloride with ligands
4 and 9 resulted in the isolation of ion-paired complexes
[L2Fe]2+FeCl4 12 and 13 (Scheme 3).10
2-
The structure of 13 was analyzed by X-ray analysis (Figure
1). The iron atom in the cation of complex 13 is octahedral,
with six bonds to the imino nitrogens. Octahedral complex 13

Production of R-Olefins Based on Iron(II)/(III) Complexes
Organometallics, Vol. 25, No. 12, 2006 2981
bonds by about 10% than high-spin complexes. This is exactly
what is observed in the comparison of the bond lengths of
octahedral 13 with the rest of the bipyramidal complexes in
this article. These observations are consistent with other
published data on iron octahedral complexes containing similar
ligands.11c
The formation of the complexes 12 and 13 with a 1:2 ratio
between metal and ligand can be explained by the lack of ortho
substituents in ligands 4 and 9. Apparently, the presence of even
one methyl group in the ortho position of the imino aryl groups
in bis(imino)pyridine ligands can stabilize the formation of 1:1
complexes (like complex 1) with distorted bipyramidal geom-
etry. These are coordinatively unsaturated and can be precata-
lysts for ethylene oligomerization. The complexes with octa-
hedral geometry, like 13 with a 1:2 ratio, have a coordinatively
saturated and sterically hindered iron atom, which should not
afford an active site for ethylene oligomerization.
Complexes of type B, with two meta-aryl groups on the same
side with two ortho-methyls of the imino aryl group, were
synthesized by the following routes. Condensation between
commercially available 1-(6-acetylpyridin-2-yl)ethanone (2) and
3-bromo-2-methylphenylamine (14) afforded 2,6-bis(1-(2-meth-
yl-3-bromophenylimino)ethyl)pyridine (15) with two bromides
in the meta positions of the aryl imino arms (Scheme 4). The
palladium-catalyzed Suzuki cross-coupling reaction between 2,6-
bis(1-(2-methyl-3-bromophenylimino)ethyl)pyridine (15) and
4-fluorophenylboronic acid (5), 3,5-bis(trifluoromethyl)phenyl-
boronic acid (16), and 5-methyl-2-thiopheneboronic acid (17)
afforded tridentate ligands 18, 19, and 20 of type B, correspond-
ingly (Scheme 4). The same catalytic protocol was used as for
the preparation of ligands 4 and 9 of type A.
Figure 1. ORTEP drawing of the cationic part of iron(2+), bis-
[2,6-bis(1-(3,5-di(4-fluorophenyl)-4-methyl)phenylimino)ethyl)py-
ridine], tetrachloferrate(2-) (13). Thermal ellipsoids are drawn to
the 20% probability level. Hydrogen atoms are omitted for clarity.
has short Fe-N bonds varying from 1.84 to 1.98 Å. These are
shorter by a maximum of 0.4 Å than all trigonal bipyramidal
complexes in this study. Octahedral transition metal complexes
having a d4 to d7 electronic configuration, like complex 13 with
d6 electronic configuration, possess mostly a low-spin electronic
configuration.11a,b Pentacoordinated complexes, such as the
trigonal bipyramidal complexes in this study, possess mostly a
high-spin electronic configuration.1d According to ligand field
theory,11a,b low-spin octahedral complexes always have shorter
Scheme 4. Synthesis of the Ligands 15 and 18-20 of Type B

2982 Organometallics, Vol. 25, No. 12, 2006
Scheme 5. Fe(II) Complexes 21-23 of Type B
Ionkin et al.
Scheme 6. Synthesis of Fe(II) Complex 22 and Fe(III) Complex 24 of Type B
The FeII complexes 21-23 of type B were synthesized by
the reaction of iron(II) chloride with ligands 18-20 in n-butanol
or in THF (Scheme 5).
Complex 24, with a trivalent iron center, was prepared by
the reaction between iron(III) chloride and ligand 19 (Scheme
6). A crystal of 24 suitable for X-ray analysis was grown from
benzene (Figure 2); this confirmed the trivalent state of iron
and the tridentate mode of the coordination of ligand 19.
The trivalent iron complex 24 was not stable in THF,
undergoing reduction to become the divalent iron complex 22
within a week.12 The reducing agent in this process is likely
THF. Abstraction of hydrogen from THF has been found to be
the cause of other such reductions.13 A single crystal suitable
for X-ray analysis of the material obtained from the reduction
of 24 to 22 was grown from methylene chloride (Figure 3).
A type C complex, in which meta-aryl groups are opposite
the ortho-methyls of the imino aryl group, was prepared by the
sequence outlined below. The first reaction was the condensation
between commercially available 1-(6-acetylpyridin-2-yl)etha-
none (2) and 5-bromo-2-methylphenylamine (25) to yield 2,6-
bis(1-(2-methyl-5-bromophenylimino)ethyl)pyridine (26). The
palladium-catalyzed Suzuki cross-coupling reaction between
compound 26 and 3,5-bis(trifluoromethyl)phenylboronic acid
(16) afforded tridenate ligand 27 of type C (Scheme 7). A
different catalytic protocol was used for the preparation of ligand
27. Benzyldi-tert-butylphosphane (28) was employed instead
Figure 2. ORTEP drawing of 2,6-bis(1-(2-methyl-3-(3,5-bis-
(trifluoromethyl)phenyl)phenylimino)ethyl)pyridineiron(III) chloride
(24). Thermal ellipsoids are drawn to the 20% probability level.
Some CF3 groups were split into six fluorine positions due to
disorder. Only three fluorine positions on each CF3 group are
depicted, and all hydrogen atoms are omitted for clarity.
of di-tert-butyl(2,2-dimethylpropyl)phosphane (7) in the cross-
coupling reaction. The sterically bulky phosphane 28 was
obtained by prolonged refluxing of almost equimolar amounts
of di-tert-butylchlorophosphine (10) and of benzylmagnesium
chloride (29) in THF.14

Production of R-Olefins Based on Iron(II)/(III) Complexes
Organometallics, Vol. 25, No. 12, 2006 2983
Scheme 7. Synthesis of Fe Complex 30 of Type C
The reaction of iron(II) chloride with ligand 27 afforded
complex 30 of type C. A crystal of 30 suitable for X-ray was
grown from pentane (Figure 4).
A complex of type D, in which two meta-aryls in one imino
aryl group and two ortho-methyls in the second imino aryl
group, was synthesized by the following reaction sequence. Two
successive condensations, the first between 2,6-diacetylpyridine
and 4-bromo-2,6-dimethylphenylamine yielding compound 31,
followed by a second condensation with 3,5-dibromo-4-methyl-
phenylamine (3), were used to prepare nonsymmetrical ligand
32. Suzuki cross-coupling between bis(imino)pyridine 15 with
three bromides and 3,5-bis(trifluoromethyl)phenylboronic acid
(16) afforded tridentate ligand 33 of type D (Scheme 8). The
Figure 3. ORTEP drawing of 2,6-bis(1-(2-methyl-3-(3,5-bis-
(trifluoromethyl)phenyl)phenylimino)ethyl)pyridineiron(II) chloride
(22). Thermal ellipsoids are drawn to the 50% probability level.
Some CF3 groups were split into six fluorine positions due to
disorder. Only three fluorine positions on each CF3 group are
depicted, and all hydrogen atoms are omitted for clarity.
Figure 4. ORTEP drawing of 2,6-bis(1-(2-methyl-5-(3,5-bis-
(trifluoromethyl)phenyl)phenylimino)ethyl)pyridineiron(II) chloride
(30). Thermal ellipsoids are drawn to the 50% probability level.
Some CF3 groups were split into six fluorine positions due to
disorder. Only three fluorine positions on each CF3 group are
depicted, and all hydrogen atoms are omitted for clarity.

2984 Organometallics, Vol. 25, No. 12, 2006
Scheme 8. Synthesis of Fe Complex 34 of Type D
Ionkin et al.
same catalytic protocol was used as for the preparation of ligands
of types A and B. The reaction between ligand 33 and iron(II)
chloride in THF afforded complex 34 of type D.
Solid-State Structures of Fe Precatalysts. Complexes 22,
30, and 34 have distorted bipyramidal geometries, which are
typical for iron(II) complexes with tridentate bis(imino)pyridine
ligands in a ratio of 1:1.15 The planes of the ortho-methyl-meta-
aryl-substituted arylimino groups of complexes 22 and 30 are
oriented orthogonally to the plane formed by iron and the three
nitrogen atoms. ortho-Methyl-meta-aryl-substituted groups are
oriented in “up-down” conformation relative to each other in
the solid state for 22 and 30. Unexpectedly, complex 1 was
found mostly in the “up-up” conformation, with only 28% of
the symmetrical up-down conformation. One might expect that
the symmetrical “up-down” conformer would be predominant
in all cases. This is evidence that rotation of the aryl group
The structure of complex 34 is shown in Figure 5. A crystal
(10) Chen, Y.; Qian, C.; Sun, J. Organometallics 2003, 22, 1236.
(11) (a) Hauser, A. Top. Curr. Chem. 2004, 233, 49. (b) Hauser, A.;
Jeftic, J.; Romstedt, H.; Hinek, R.; Spiering, H. Coord. Chem. ReV. 1999,
190-192, 471. (c) Scheer, C.; Chautemps, P.; Gautier-Luneau, I.; Pierre,
J.; Serratrice, G. Polyhedron 1996, 15, 219.
(12) Bryliakov, K. P.; Semikolenova, N. V.; Zudin, V. N.; Zakharov,
V. A.; Talsi, E. P. Catal. Commun. 2004, 5, 45.
(13) (a) Moreira, R. F.; Tshuva, E. Y.; Lippard, S. J. Inorg. Chem. 2004,
43, 4427. (b) Li, R.; Smith, R. L.; Kenttamaa, H. I. J. Am. Chem. Soc.
1996, 118, 5056. (c) Malatesta, V.; Ingold, K. U. J. Am. Chem. Soc. 1981,
103, 609.
(14) (a) Johnson, L. K.; McLain, S. J.; Sweetman, K. J.; Wang, Y.;
Bennett, A. M. A.; Wang, L.; McCord, E. F.; Ionkin, A.; Ittel, S. D.;
Radzewich, C. E.; Schiffino, R. S. PCT Int. Appl. WO 2003044066 A2,
2003; 170 pp. (b) Stauffer, S. R.; Hartwig, J. F. J. Am. Chem. Soc. 2003,
125, 6977. (c) Stewart, A. P.; Trippett, S. J. Chem. Soc. C 1970, 9, 1263.
(15) (a) Britovsek, G. J. P.; Mastroianni, S.; Solan, G. A.; Baugh, S. P.
D.; Redshaw, C.; Gibson, V. C.; White, A. J. P.; Williams, D. J.; Elsegood,
M. R. J. Chem. Eur. J. 2000, 6, 2221. (b) Bianchini, C.; Mantovani, G.;
Meli, A.; Migliacci, F.; Zanobini, F.; Laschi, F.; Sommazzi, A. Eur. J. Inorg.
Chem. 2003, 8, 1620.
Figure 5. ORTEP drawing of 4-(3,5-bis(trifluoromethyl)phenyl)-
2,6-dimethylphenyl)-(1-{6-[1-(3,5-(3,5-bis(trifluoromethyl)phenyl)-
4-methylphenylimino)ethyl]pyridin-2-yl}ethylidene)amineiron(II)
chloride (34). Thermal ellipsoids are drawn to the 50% probability
level. Some CF3 groups were split into six fluorine positions due
to disorder. Only three fluorine positions on each CF3 group are
depicted, and all hydrogen atoms are omitted for clarity.
of 34 suitable for X-ray analysis was grown from methylene
chloride.
(9) Christmann, U.; Vilar, R. Angew. Chem., Int. Ed. 2005, 44, 366.

Production of R-Olefins Based on Iron(II)/(III) Complexes
Organometallics, Vol. 25, No. 12, 2006 2985
Table 1. Selected Bond Lengths (Å) and Conformer’s Ratio for 1, 22, 24, 30, 34, and 15
1
22
24
30
34
15
N/A
axial (imino) N-Fe bond
axial (imino) N-Fe bond
basal (central) N-Fe bond
CdN
2.2303(16)
Fe-N1
2.221(2)
Fe-N1
2.221(2)
Fe-N1A
2.133(3)
Fe-N2
1.284(3)
N1-C1
1.284(3)
N3-C7
up-down
2.190(4)
Fe-N3
2.231(2)
Fe-N1
2.209(9)
Fe-N3
2.260(9)
Fe-N1
2.096(9)
Fe-N2
1.283(13)
N1-C1
1.270(13)
N3-C7
N/A
2.2424(17)
Fe-N3
2.193(4)
Fe-N1
2.241(2)
Fe-N3
N/A
N/A
2.0943(16)
Fe-N2
2.121(4)
Fe-N2
2.107(2)
Fe-N2
1.281(2)
N1-C1
1.287(3)
1.260(7)
N1-C1
1.271(7)
N3-C7
up-down
1.288(3)
N1-C1
1.277(3)
N3-C7
up-down
1.294(9)
N2-C6
1.286(8)
N3-C23
up-up
CdN
N3-C7
ratio of up-up to
up-down isomers
72% of up-up and
28% of up-down
Table 2. Estimates of Lifetimes of Catalysts from Ethylene
around the C-N bonds does take place, and this must be
considered in the interpretation of oligomerization results.
The introduction of additional aryl substituents in the meta
positions next to the ortho-methyls (type B, complex 22) resulted
in a shortening of the axial Fe-N bonds by 0.0093 and 0.0214
Å relative to the parent complex 1 (Table 1). This shortening
of Fe-N bonds was not observed in the case of complex 30,
with aryl substituents in 3,5-ortho-meta-substitution (type C).
The axial Fe-N bond lengths for 1 and 30 are essentially
identical. The bond lengths of the imino groups of ligand 15
and its Fe complex 22 are also identical within experimental
error. The bond lengths of the imino moieties of ligand 15 did
not undergo a significant change upon the formation of its Fe
complex 22. Both of them have a typical CdN double bond
length around 1.28 Å.16
Nonsymmetrical complex 34 has the longest and the shortest
axial N-Fe bonds among distorted bipyramidal structures in
this study. The longest [2.260(9) Å] is the bond of the moiety
with two ortho-methyl groups. The steric repulsive interaction
between the ortho-methyl groups and the iron core of the
molecule is likely responsible for this elongation. The moiety
without ortho-methyl groups in 34 has the shortest N-Fe
distance at 2.209(9) Å. meta-Aryl groups in 34 are rotated away
from the main plane of the iminophenyl group with angles
ranging from 35.4° to 59.1°.
Uptake at 120 °C and 700 psig
precatalyst
1
21
22
23
24
30
34
lifetime (min)
2.9
20.0
11.5
20.0
11.2
16.7
20.0
merizations consumed ethylene. The validity of such compari-
sons is a function of the amount of catalyst injected. If too much
is added, a large exotherm will result, and the catalyst is
overheated and so decomposes more rapidly to give a shorter
uptake time. With too small of an amount, the uptake will fall
below the ethylene flowmeter detection limit, and so again an
underestimate of uptake time will result. To address this
problem, the amounts of catalysts added were adjusted such
that the batch temperature stayed within 2 °C of the set point.
While this meant there could be a large range of catalyst charges,
similar amounts of R-olefins were made for all runs and the
ethylene uptake times were a valid measure of the catalyst
lifetimes. In addition, the R-olefin analytical methods were
applied over a comparable concentration range, which is
desirable. A table of lifetime results is shown in Table 2.
Before considering data trends with respect to precatalyst
structure, the main focus is to compare the parent methyl-
substituted FeII complex 1 to all of the other precatalysts with
greater degrees of substitution (types B-D).
With respect to the lifetime data in Table 2, it is seen that
the parent methyl-substituted FeII complex 1 has by far the
shortest lifetime. Otherwise, it appears that having the meta 3,5-
bis(trifluoromethyl)phenyl group on the same side with the
ortho-methyl in 22 of type B gives a shorter lifetime versus
when the meta 3,5-bis(trifluoromethyl)phenyl group is on the
side opposite the ortho-methyl of the imino aryl group, such as
in complex 30 of type C. Also, the iron oxidation state (22 vs
24) appears to have no effect on lifetime.
The geometry of the six-coordinated iron atom in the trivalent
complex 24 may be described as distorted-octahedral. The axial
N-Fe bonds in 24 are shorter by 0.031 Å than corresponding
bonds in the divalent complex 22, reflecting the effects of the
higher oxidation state of the iron in 24.
Oligomerization of Ethylene to r-Olefins by Iron Com-
plexes 1, 21, 22, 23, 24, 30, and 34. Oligomerizations were
run at 700 psig of ethylene in o-xylene at 120 and 130 °C. The
resulting product mixtures were evaluated for soluble R-olefins
via GC and for insolubles via a filtration/gravimetric method.
The distribution of soluble R-olefins and the percent solids of
total R-olefins are important factors for the evaluation of a
potential catalyst. A product distribution, which fits the theoreti-
cal Shultz-Flory distribution, is optimal to minimize ethylene
yield loss. R-Olefins of carbon number greater than 40 make
up the percent solids of total R-olefins. Values for these below
that predicted by the Shultz-Flory distribution were desired.
As noted above, C6 to C16 R-olefins are targeted for a general
commercial application.
A list of precatalyst productivities is in Table 3. The
precatalysts with patterns of double substitution tend to last
longer after an initial uptake but give substantially lower
conversions of ethylene. The parent methyl-substituted FeII
complex 1 is by far the most productive. It is followed by
precatalyst 30 of type C and by precatalyst 34 of type D. All
other precatalysts have about the same productivities. This seems
to indicate that bulky groups in the ortho positions slow access
of ethylene to the metal center of the catalyst.
While most of the precatalysts have half the productivity of
precatalyst 1, they are nonetheless useful: all of these catalysts
are highly productive by R-olefins industry standards, and the
amounts needed are insignificant in terms of R-olefin production
costs. As was mentioned earlier, with a shorter lifetime in a
continuous plug-flow process, there would be a need to install
and maintain more injection points, which could add significant
initial investment and production costs. This means precatalyst
Another important factor is the catalyst productivity measured
as kg R-olefins/g catalyst. This is a function of the catalyst/
ethylene kinetics and catalyst thermal stability. An indication
of the thermal stabilities, lifetimes at a given temperature, was
obtained by measuring the periods of time over which oligo-
(16) De Bruin, B.; Bill, E.; Bothe, E.; Weyhermueller, T.; Wieghardt,
K. Inorg. Chem. 2000, 39, 2936.

2986 Organometallics, Vol. 25, No. 12, 2006
Table 3. r-Olefins from Ethylene Oligomerizations by Iron Complexes 1, 21, 22, 23, 24, and 30a
Ionkin et al.
“K” value
Shultz-Flory
distributionb
precatalyst/
amount µmol
amount cocatalyst,
MMAO, mmol
temperature,
kg of LAO per
g of catalyst
% solids
entry
°C
total LAO c
SFD R2
1
2
3
4
5
6
7
8
9
21/0.457
21/0.610
22/0.336
22/0.224
22/0.090
23/0.121
24/0.107
24/0.053
30/0.112
34/0.08
2.26
1.13
2.26
2.26
0.68
1.13
1.13
1.13
1.13
1.13
1.13
130
120
130
120
120
120
120
120
120
120
120
0.68
0.68
0.67
0.69
0.67
0.68
0.64
0.65
0.57
0.70
0.59
125
54
7.83
8.89
5.10
3.66
7.37
8.81
7.56
9.48
4.39
4.55
3.44
0.9628
0.9923
0.9901
0.9781
0.9920
0.9882
0.9968
0.9968
0.9906
0.9969
0.9862
124
123
160
142
128
128
249
197
458
10
11
1/0.064
a Conditions: solvent: xylenes; pressure: 700 psig. b Determined from GC, using extrapolated values for C-10 and C-12. c Xylenes-insoluble fraction of
R-olefins.
Figure 6. Distribution of R-olefins from C-2 to C-14 for precatalysts 1 (squares) and precatalyst 22 (triangles).
21, 23, or 34 may be a better choice than 1 with regard to
where mp is the weight fraction of the R-olefin and Dp is the
productivity/lifetime considerations.
degree of polymerization of the R-olefin
2-
The ion-pared complex [L2Fe]2+FeCl4 13 of type A was
If the catalyst makes a Shultz-Flory distribution of R-olefins,
a plot of ln(mp/Dp) versus (Dp - 2) should give a linear
relationship with the slope equal to the inverse logarithm of
the growth factor K. An example of such a plot is shown in
Figure 6. The low and high ends of the distribution for
precatalyst 1 deviate more from the line than do those for
precatalyst 22. The low and high end products are of less
commercial value, so the deviations by the parent methyl-
substituted FeII complex 1 yield a less desirable product mix.
found to be inactive in R-olefin oligomerization experiments
within the tested temperature range of 100-130 °C.
The desired R-olefin product distribution depends on the
targeted application. For this work, a K factor in the range 0.6
to 0.7 was desired. The distribution obtained depends on the
propagation and termination rates associated with a given
catalyst. An olefin growth factor K is defined by the equation
The linear regression coefficients (R2) for such plots are
indicators of the distribution ideality. The closer the value is to
1, the more ideal is the product distribution. The right-hand
column in Table 3 contains a list of R2 values. A ranking of R2
values from high to low indicates that precatalyst 1 falls near
the bottom of the list. The nonsymmetric precatalyst 34 is at
the top. Precatalysts 22 and 30, with their different substitution
patterns of the same 3,5-bis(trifluoromethyl)phenyl groups, have
nearly the same value.
K ) RP/(RP + Rt)
where RP is the rate of propagation and Rt is the rate of
termination. This growth factor is used in the following form
of the Schulz-Flory distribution equation:17
ln(mp/Dp) ) ln((1 -K)∧2/(2 - K)) + (Dp - 2) ln K
(17) Elvers, B., et al., Eds. In Ullmann’s Encyclopedia of Industrial
Chemistry; VCH VerlagsgesellschaftmbH: Weinheim, 1989; Vol. A13, pp
243-247 and 275-276.
Growth factors (K) for the precatalysts are also listed in Table
3. Considering the categories mentioned above, the parent

Production of R-Olefins Based on Iron(II)/(III) Complexes
Organometallics, Vol. 25, No. 12, 2006 2987
Table 4. Typical Quality of r-Olefins and Impurities from
the Oligomerization Experiments of the FeII Tridentate
Catalysts with Double Patterns of Substitution
methyl-substituted Fe(II) complex 1 has close to the lowest K
value. This suggests that the multiple substitutions for the others
raise the distributions, meaning that they make, on average,
longer R-olefins. The highest value was obtained by nonsym-
metrical precatalyst 34 of type D. Reaching K’s in the 0.65-
0.70 range at temperatures of 120-130 °C is advantageous
because a desirable product distribution is made over a tem-
perature range where olefin solubility is enhanced to reduce
reactor fouling.
LAO,
wt %
internal olefins,
ppm
branched olefins,
wt %
product
C4
C6
C8
C10
99.98
99.24
98.36
97.61
215
1128
2325
2210
0.64
1.41
2.17
The values for percent solids of total R-olefins are also listed
in Table 3. In a rank of lowest to highest, precatalyst 1 yields
the lowest solids. Precatalysts 22 and 30 have similar values,
suggesting location of the same 3,5-bis(trifluoromethyl)phenyl
groups in different positions had no detectable effect. Nonsym-
metric precatalyst 34 ranks near the top of the list. It should be
kept in mind that the measurements of percent solids are not as
precise as, for example, K values or productivities. Thus, the
data for percent solids are slightly scattered.
the amounts of internal olefins are below the 2500 ppm level.
The purity range slightly exceeds the range reported for the
SHOP process.2b Detailed purity analysis and further compari-
sons with current R-olefins processes are outside the scope of
this study.
As was found by X-ray analysis for the solid state, the
precatalyst 1 forms two conformers: an “up and down”
conformer and an “up-up conformer” (Scheme 9). It is
reasonable to suggest the existence of those conformers in the
solution during the oligomerization. The “up and down”
conformer offers symmetrical steric protection of both axial sides
of iron (below and above the N-Fe(N)-N plane). The “up and
up” conformer protects one axial side only by two ortho-methyl
groups, leaving the other side unprotected. If oligomerization
of ethylene commences from the side protected by two methyl
groups, it should lead to a heavier fraction of R-olefins. If the
oligomerization of ethylene commences from the unprotected
side of the conformer, it should lead to a light fraction of
R-olefins. According to theoretical calculations of the poly-
merization mechanisms on FeII catalysts with tridentate bis-
(imino)pyridine ligands, the chain propagation takes place from
the same axial side of the N-Fe(N)-N plane with backside
approach of ethylene to the alkyl chain.19 As a result, the “up
and up” conformer of 1 will contribute to curvature of the
Shultz-Flory distribution of R-olefins, while the symmetrical
“up and down” conformer should obey the Shultz-Flory
distribution. The possibility of different “interconverting iso-
mers” in a family of nonsymmetrical iron tridentate catalysts
has been proposed in the case of propylene polymerization.20
The importance of the conformeric behavior of these com-
plexes with regards to the steric protection of the axial sides of
the iron center is clear upon comparisons of oligomerization
results of precatalysts 22 and 30, which are isomers. Both of
them have the same electron-withdrawing 3,5-bis(trifluoro-
methyl)phenyl group in meta positions (Scheme 9). However
there is substantial differences in the catalytic performance. The
precatalyst 30 behaves as catalyst 1, giving substantial deviation
from the Shultz-Flory distribution, low K value, and very good
productivity (Table 2). Precatalyst 22 afforded R-olefins with
more ideal Shultz-Flory distribution, higher K value, and
reduced productivity. Thus, steric factors appear to be more
important than electronic ones. Precatalyst 22 likely exists
mostly as an “up and down” conformer in solution, in which
the additional 3,5-bis(trifluoromethyl)phenyl groups in the meta
positions from the same side of the ortho-methyls keep the
arylimino groups in an up and down conformation and restrict
180° rotation of the aryl groups around the C-N bonds. The
conformational behavior of 30 closely resembles that of 1, and
additional increase of steric bulk in 30 resulted in lower
productivity.
From the data in Table 3, it is seen that precatalysts with
double patterns of substitutions have lower productivities, higher
K values, and higher percent solids of total R-olefins. Hindered
ethylene access to the reaction center was proposed earlier as
the explanation of reduced catalytic activity of the bulkier iron
tridentate catalysts.1c Theoretical calculations later confirmed
that ethylene capture is the rate-determining step in these
oligomerizations.19 On the other hand, increasing the steric bulk
above and below the planes of iron and nickel in the Versipol
family of catalysts has been shown to slow chain transfer relative
to chain propagation.1 This, in turn, should lead to the higher
growth factors and higher percent solids of total R-olefins (LAO)
for the precatalysts with double patterns of substitutions. The
higher percent solids of total LAO made for these catalysts,
also seen in Table 3, would support this proposal. As the
distributions are shifted to higher MW products, more insolubles
are made. The losses in productivities are real because the solids
were included in the productivity calculations. In addition, the
double-substituted catalysts have longer lifetimes than precata-
lyst 1, but their productivities are much lower. It appears then
that the bulky second substitutions affected the propagation rates.
Precatalysts 22 and 30 have two large 3,5-electron-withdrawing
bis(trifluoromethyl)phenyl groups on the imino aryls, but the
productivities of 22 versus 30 are very different. This would
suggest that the locations of the bulky groups make the
difference. In 22, the bulky bis(trifluoromethyl)phenyl groups
are adjacent to the ortho-methyl, but in 30 they are opposite.
The latter could give more ready access to the metal center for
ethylene addition and chain transfer. Precatalyst 34 has a
productivity that falls between 22 and 30. Again, the substitution
pattern is very different. The issue is whether this is a
predominantly steric or electronic effect. Precatalysts 21 and
23 have aromatic groups with somewhat different electron-
withdrawing ability and with approximately the same bulk at
the same meta positions as 22, but their K values and
productivities are all nearly the same. This would suggest that
the electron-withdrawing effects are minimal and that the
substitution pattern matters most.
Typical qualities of R-olefins prepared by the Versipol
catalysts with double patterns of substitutions are shown in Table
4. They are in line with previously reported numbers.1c,e The
amount of undesirable branched olefins are below 2 wt %, and
Precatalyst 34 has two ortho-methyl groups closely protecting
both axial sides of iron from one side and two aryl groups in
meta positions protecting the axial sides remotely. It appears
that this combination of steric protection is the best with regard
(18) Bart, S. C.; Lobkovsky, E.; Chirik, P. J. J. Am. Chem. Soc. 2004,
126, 13794.
(19) (a) Deng, L.; Margl, P.; Ziegler, T. J. Am. Chem. Soc. 1999, 121,
6479. (b) Ramos, J.; Cruz, V.; Munos-Escalona, A.; Martinez-Salazar, J.
Polymer 2002, 43, 3635.
(20) Small, B. L.; Brookhart, M. Macromolecules 1999, 32, 2120.

2988 Organometallics, Vol. 25, No. 12, 2006
Ionkin et al.
Table 5. Summary of Crystal Data, Data Collection, and Structural Refinement Parameters for 1, 13, 15, and 22
1
13
15
C39H27F12N3
22
empirical formula
FW
C24H25Cl4FeN3
C94H70Cl4F8Fe2N6
C41H31Cl6F12FeN3
553.12
1689.06
765.64
1062.24
cryst color, form
cryst syst
space group
a (Å)
blue, needle
monoclinic
P2(1)/c
10.4140(5)
black, block
orthorhombic
P2(1)2(1)2
13.482(5)
colorless, block
orthorhombic
Pna2(1)
16.770(2)
green, plate
monoclinic
C2/c
11.5071(4)
b (Å)
15.2465(8)
23.751(8)
15.807(2)
12.4346(5)
c (Å)
15.8488(8)
26.066(11)
27.350(4)
31.4986(13)
R (deg)
90
90
90
90
8347(5)
90
90
90
90
â (deg)
92.3650(10)
90
2514.3(2)
100.3450(10
90
4433.7(3)
γ (deg)
V (Å3)
7250.0(16)
Z
4
4
8
4
density (g/cm3)
1.461
1.042
1136
1.344
0.544
3472
1.403
0.126
3120
1.591
0.787
2136
abs µ (mm-1
F(000)
)
cryst size (mm)
temp (°C)
scan mode
detector
0.32 × 0.06 × 0.02
0.12 × 0.12 × 0.06
0.28 × 0.14 × 0.12
0.34 × 0.32 × 0.07
-100
-100
-100
-100
ω
ω
ω
ω
Bruker-CCD
Bruker-CCD
Bruker-CCD
Bruker-CCD
θ
max (deg)
27.96
37 670
6038
0.0547
22.08
23 505
10 182
0.1818
28.35
45 419
14 682
0.1477
28.46
23 576
5577
0.0356
no. obsvd reflns
no. uniq reflns
Rmerge
no. params
358
1.02
1031
0.924
981
0.968
298
1.039
Sa
R indices [I > 2σ(I)]b
R indices (all data)b
max. diff peak, hole (e/Å3)
wR2 ) 0.075, R1 ) 0.037
wR2 ) 0.084, R1 ) 0.060
0.288, -0.325
wR2 ) 0.220, R1 ) 0.099
wR2 ) 0.302, R1 ) 0.231
0.610, -0.549
wR2 ) 0.153, R1 ) 0.076
wR2 ) 0.223, R1 ) 0.251
0.395, -0.348
wR2 ) 0.131, R1 ) 0.051
wR2 ) 0.142, R1 ) 0.068
0.956, -0.553
2
2
a GooF ) S ) {∑[w(Fo - Fc )2]/(n - p)}1/2, where n is the number of reflections, and p is the total number of refined parameters. b R1 ) ∑||Fo| -
2
2
2
|Fc||/∑|Fo|, wR2 ) {∑[w(Fo - Fc )2]/∑[w(Fo )2]}1/2 (sometimes denoted as Rw2).
Table 6. Summary of Crystal Data, Data Collection, and Structural Refinement Parameters for 24, 30, and 34
24
30
34
empirical formula
fw
C51H39Cl3F12FeN3
1084.05
C39H27Cl2F12FeN3
892.39
C
102H74Cl16F36Fe2N6
2746.57
cryst color, form
cryst syst
space group
a (Å)
gold, plate
monoclinic
P2(1)
16.000(2)
blue, needle
monoclinic
C2/c
32.663(4)
blue, needle
monoclinic
P2(1)/c
20.024(11)
b (Å)
19.310(3)
15.197(2)
21.142(12)
c (Å)
16.689(3)
15.494(2)
29.448(18)
R (deg)
90
90.168(3)
90
90
96.197(2)
90
90
â (deg)
105.241(14)
90
12028(12)
γ (deg)
V (Å3)
5156.2(14)
7646.0(17)
Z
4
8
4
density (g/cm3)
1.396
0.528
2204
1.55
0.627
3600
1.517
0.699
5504
abs µ (mm-1
F(000)
)
cryst size (mm)
temp (°C)
scan mode
detector
0.21 × 0.09 × 0.03
0.420 × 0.080 × 0.030
0.85 × 0.08 × 0.02
-100
-100
-100
ω
ω
ω
Bruker-CCD
Bruker-CCD
Bruker-CCD
θ
max (deg)
26.45
43 238
10 557
0.1054
29.35
23 662
23 680
0
24.34
54 411
17 640
0.1494
no. obsvd reflns
no. uniq reflns
Rmerge
no. params
553
0.92
539
0.944
1469
1.417
Sa
R indices [I > 2σ(I)]b
R indices (all data)b
max. diff peak, hole (e/Å3)
wR2 ) 0.223, R1 ) 0.084
wR2 ) 0.261, R1 ) 0.172
0.538, -0.451
wR2 ) 0.196, R1 ) 0.080
wR2 ) 0.233, R1 ) 0.144
1.371, -0.737
wR2 ) 0.309, R1 ) 0.132
wR2 ) 0.346, R1 ) 0.218
3.611, -0.936
a GooF ) S ) {∑[w(Fo - Fc )2]/(n - p)}1/2, where n is the number of reflections, and p is the total number of refined parameters. b R1 ) ∑||Fo| -
2
2
|Fc||/∑|Fo|, wR2 ) {∑[w(Fo - Fc )2]/∑[w(Fo )2]}1/2 (sometimes denoted as Rw2).
2
2
2
to the linearity of the Shultz-Flory distribution, K value, and
the purity of R-olefins. A conformer with two ortho-methyl
groups simultaneously above and below the N-Fe(N)-N plane
seems to be essential for R-olefin formation with good linearity
and purity. It appears that it is not crucial whether the two ortho-
methyls are from two sides of the aryl imino arms (as in
precatalyst 22 of pattern B) or come from one side (as in
precatalyst 34 of pattern D).
In conclusion several precatalysts were synthesized and tested
for the oligomerization of ethylene to make linear R-olefins.
The goal was to replace complex 1 with a candidate that had a
longer lifetime and a more ideal Shultz-Flory product distribu-

Production of R-Olefins Based on Iron(II)/(III) Complexes
Organometallics, Vol. 25, No. 12, 2006 2989
Scheme 9. Conformers of Complexes 1, 30, and 22 Resulting upon Rotation of the Aryl Moiety around the
Nitrogen-Iron-Nitrogen Coordination Plane
n-butanol were purchased from Aldrich. Complex 1 was prepared
according to the literature.21
tion. A longer lifetime was observed for nearly all candidates
tested. It was also found that the double substitutions of large
aryl groups at the meta positions in addition to methyls at the
ortho positions of the imino aryl groups resulted in higher K
factors, more solids, and a real loss in productivity. Precatalyst
34, of type D, with two meta-aryls in one imino aryl group and
two ortho-methyls in the second imino aryl, had the highest K
factor of all precatalysts and ranked low in percent solids. This
would appear to be a result of the proper combination of steric
and electronic factors and conformeric behavior in this type of
complex. Future theoretical studies may be beneficial in
understanding the correlations between steric and electronic
effects and conformeric behavior created by additional substitu-
tions in tridentate ligands and the parameters of the oligomer-
ization processes, and these are under way in our laboratory.
2,6-Bis(1-(3,5-dibromo-4-methyl)phenylimino)ethyl)pyridine
(4). 1-(6-Acetylpyridin-2-yl)ethanone (2) (4.39 g, 0.027 mol), 15.0
g (0.057 mol) of 3,5-dibromo-4-methylphenylamine (3), 200 mL
of the methanol, and a few crystals of para-toluenesulfonic acid
were stirred for 3 days at room temperature under the flow of
nitrogen. The resultant precipitate was filtered, washed twice with
10 mL of methanol, and dried in a 1 mm vacuum overnight. The
yield of 2,6-bis(1-(3,5-dibromo-4-methyl)phenylimino)ethyl)pyri-
1
dine (4) was 12.1 g (68%) as a pale yellow solid. H NMR (500
MHz, CD2Cl2, TMS): δ 2.40 (s, 6H, Me), 2.58 (s, 6H, Me), 7.10
(s, 4H, Arom-H), 7.63 (t, 3JHH ) 7.8 Hz, 1H, Py-H), 8.30 (d, 3JHH
) 7.8 Hz, 2H, Py-H). 13C NMR (500 MHz, CD2Cl2, (selected
signals)): δ 169.03 (CdN). Anal. Calcd for C23H19Br4N3 (MW
657.03): C, 42.04; H, 2.91; N, 6.40. Found: C, 42.21; H, 3.14; N,
6.48.
Di-tert-butyl(2,2-dimethylpropyl)phosphane (7). Di-tert-bu-
tylchlorophosphine (10) (28.80 g, 0.160 mol), 0.2 mol of neopen-
tylmagnesium chloride (11) in diethyl ether, and 150 mL of THF
were refluxed under argon for 3 days. The reaction mixture was
allowed to cool to RT, and an aqueous solution of ammonium
chloride was added slowly. The organic phase was separated and
dried with magnesium sulfate. After removal of the solvent, the
product was purified by distillation in a vacuum. The yield of di-
tert-butyl(2,2-dimethylpropyl)phosphane was 25.26 g (73%) with
bp 43-47 °C/0.1 mm. 31P NMR (CD2Cl2): +19.76 ppm. 1H NMR
Experimental Section
All air-sensitive compounds were prepared and handled under a
N2/Ar atmosphere using standard Schlenk and inert-atmosphere box
techniques. Anhydrous solvents were used in the reactions. Solvents
were distilled from drying agents or passed through columns under
an argon or nitrogen atmosphere. Anhydrous iron(II) chloride, 1-(6-
acetylpyridin-2-yl)ethanone, 3-bromo-2-methylphenylamine, 3,5-
dibromo-4-methylphenylamine, 5-bromo-2-methylphenylamine, 4-flu-
orophenylboronic acid, 3,5-bis(trifluoromethyl)phenylboronic acid,
5-methyl-2-thiopheneboronic acid, tris(dibenzylideneacetone)di-
palladium(0), cesium carbonate, di-tert-butylchlorophosphine, 2.0
M solution of benzylmagnesium chloride in THF, MMAO, and
(CD2Cl2): 1.17 (s, 9H, Me3C), 1.19 (s, 9H, Me3C), 2.47 (d, 2JPH
)
3.12 Hz, P-CH2-CMe3). Anal. Calcd for C13H29P (MW 216.34):
C, 72.17; H, 13.51; P, 14.32. Found: C, 72.01; H, 13.49; P, 14.08.
(21) (a) Schmidt, R.; Welch, M. B.; Knudsen, R. D.; Gottfried, S.; Alt,
H. G. J. Mol. Catal A: Chem. 2004, 222, 9. (b) Spence, Rupert; Stevens,
Robert Louis. PCT Int. Appl. WO 2005030814, 2005; 20 pp.

2990 Organometallics, Vol. 25, No. 12, 2006
Ionkin et al.
2,6-Bis(1-(3,5-di(4-fluorophenyl)-4-methyl)phenylimino)eth-
yl)pyridine (9). 4-Fluorophenylboronic acid (5) (5.62 g, 0.04 mol),
4.4 g (0.0067 mol) of 2,6-bis(1-(3,5-dibromo-4-methyl)phenylimi-
no)ethyl)pyridine (4), 13.1 g (0.04 mol) of cesium carbonate (8),
0.92 g (0.001 mol) of tris(dibenzylideneacetone)dipalladium(0) (6),
0.52 g (0.0024 mol) of di-tert-butyl(2,2-dimethylpropyl)phosphane
(7), and 50 mL of dioxane were stirred at room temperature for 24
h. The reaction mixture was filtered, and the solvent was removed
under vacuum. The resulting mixture was purified by recrystalli-
zation from 20 mL of methanol. Yield of 2,6-bis(1-(3,5-di(4-
fluorophenyl)-4-methyl)phenylimino)ethyl)pyridine (9) was 1.53 g
(32%) as a light yellow solid. 1H NMR (500 MHz, CD2Cl2, TMS):
δ 2.10 (s, 6H, Me), 2.45 (s, 6H, Me), 6.60 (s, 4H, Arom-H), 7.12
(m, 4H, Arom-H), 7.35 (m, 4H, Arom-H), 7.80 (t, 3JHH ) 7.8 Hz,
(6), 0.35 g (0.0016 mol) of di-tert-butyl(2,2-dimethylpropyl)-
phosphane (7), and 50 mL of dioxane were stirred at room
temperature for 24 h. The reaction mixture was filtered, and the
solvent was removed under vacuum. The resulting mixture was
purified by recrystallization from 20 mL of ethanol. Yield of 2,6-
bis(1-(2-methyl-3-(4-fluorophenyl)phenylimino)ethyl)pyridine (18)
1
was 3.74 g (63%) as a light yellow solid. H NMR (500 MHz,
CD2Cl2, TMS): δ 1.95 (s, 6H, Me), 2.40 (s, 6H, Me), 6.63 (m,
3
2H, Arom-H), 7.40 (m, 14H, Arom-H), 7.91 (t, JHH ) 7.8 Hz,
3
1H, Py-H), 8.40 (d, JHH ) 7.8 Hz, 2H, Py-H). 13C NMR (500
MHz, CD2Cl2, (selected signals)): δ 168.1 (CdN). 19F NMR (CD2-
Cl2): -117.32 (s, 2F). Anal. Calcd for C35H29F2N3 (MW 529.62):
C, 79.37; H, 5.52; N, 7.93. Found: C, 79.52; H, 5.69; N, 8.14.
2,6-Bis(1-(2-methyl-3-(3,5-bis(trifluoromethyl)phenyl)pheny-
limino)ethyl)pyridine (19). 3,5-Bis(trifluoromethyl)phenylboronic
acid (16) (7.44 g, 0.0289 mol), 4.80 g (0.0096 mol) of 2,6-bis(1-
(2-methyl-3-bromophenylimino)ethyl)pyridine (15), 9.40 g (0.0289
mol) of cesium carbonate (8), 0.66 g (0.00072 mol) of tris-
(dibenzylideneacetone)dipalladium(0) (6), 0.38 g (0.0018 mol) of
di-tert-butyl(2,2-dimethylpropyl)phosphane (7), and 50 mL of
dioxane were stirred at room temperature for 24 h. The reaction
mixture was filtered, and the solvent was removed under vacuum.
The resulting mixture was purified by recrystallization from 20 mL
of ethanol. Yield of 2,6-bis(1-(2-methyl-3-(3,5-bis(trifluoromethyl)-
phenyl)phenylimino)ethyl)pyridine (19) was 5.14 g (76%) as a light
3
1H, Py-H), 8.30 (d, JHH ) 7.8 Hz, 2H, Py-H). 13C NMR (500
MHz, CD2Cl2, (selected signals)): δ 168.0 (CdN). 19F NMR (CD2-
Cl2): -117.00 (s, 4F). Anal. Calcd for C47H35F4N3 (MW 717.79):
C, 78.64; H, 4.91; N, 5.85. Found: C, 78.67; H, 5.19; N, 6.02.
Iron(2+), Bis[2,6-bis(1-(3,5-dibromo-4-methyl)phenylimino)-
ethyl)pyridine], Tetrachloferrate(2-) (12). Anhydrous iron(II)
chloride (0.17 g, 0.00137 mol) was dissolved in 40 mL of warm
n-butanol. Then, 1.0 g (0.0015 mol) of 2,6-bis(1-(3,5-dibromo-4-
methyl)phenylimino)ethyl)pyridine (4) was added in one portion
to the reaction mixture. The mixture was kept at 40 °C for 1 h and
then cooled to ambient temperature. The resultant purple precipitate
was filtered and washed twice with 5 mL of pentane and dried at
1 mm vacuum. The yield of bis[2,6-bis(1-(3,5-dibromo-4-methyl)-
phenylimino)ethyl)pyridine]iron(II) chloride (12) was 0.74 g (62%
of starting ligand). Anal. Calcd for C46H38Br8Cl4Fe2N6 (MW
1567.57): C, 35.25; H, 2.44; N, 5.36. Found: C, 35.30; H, 2.50;
N, 5.48. Direct probe MS: calculated top exact mass for C46H38-
Br8Cl4Fe2N6, 1567.40; found exact mass 1567.40.
Iron(2+), Bis[2,6-bis(1-(3,5-di(4-fluorophenyl)-4-methyl)phe-
nylimino)ethyl)pyridine], Tetrachloferrate(2-) (13). Anhydrous
iron(II) chloride (0.17 g, 0.00137 mol) was dissolved in 40 mL of
warm n-butanol. Then, 1.0 g (0.0014 mol) of 2,6-bis(1-(3,5-di(4-
fluorophenyl)-4-methyl)phenylimino)ethyl)pyridine (9) was added
in one portion to the reaction mixture. The mixture was kept at 40
°C for 1 h and then was cooled to ambient temperature. The
resultant purple precipitate was filtered and washed twice with 5
mL of pentane and dried at 1 mm vacuum. The yield of bis[2,6-
bis(1-(3,5-di(4-fluorophenyl)-4-methyl)phenylimino)ethyl)pyridine]-
iron(II) chloride (13) was 0.92 g (78% of starting ligand). Anal.
Calcd for C94H70Cl4F8Fe2N6 (MW 1689.09): C, 66.84; H, 4.18;
N, 4.98. Found: C, 67.03; H, 4.37; N, 5.24. The structure was
proven by X-ray analysis.
1
yellow solid. H NMR (500 MHz, CD2Cl2, TMS): δ 2.05 (s, 6H,
Me), 2.41 (s, 6H, Me), 6.68 (m, 2H, Arom-H), 7.05 (m, 2H, Arom-
H), 7.30 (m, 2H, Arom-H), 7.95 (m, 7H, Arom-H and Py-H), 8.45
3
(d, JHH ) 7.8 Hz, 2H, Py-H). 13C NMR (500 MHz, CD2Cl2,
(selected signals)): δ 167.7 (CdN). 19F NMR (CD2Cl2): -63.54
(s, 12F). Anal. Calcd for C39H27F12N3 (MW 765.63): C, 61.18; H,
3.55; N, 5.49. Found: C, 61.36; H, 3.70; N, 5.52. The structure
was proven by X-ray analysis.
2,6-Bis(1-(2-methyl-3-(3-methylthiophene-2-yl)-phenyl)phe-
nylimino)ethylpyridine (20). 5-Methyl-2-thiopheneboronic acid
(17) (10.0 g, 0.0704 mol), 8.79 g (0.0176 mol) of 2,6-bis(1-(2-
methyl-3-bromophenylimino)ethyl)pyridine (15), 22.95 g (0.0704
mol) of cesium carbonate (8), 1.62 g (0.00177 mol) of tris-
(dibenzylideneacetone)dipalladium(0) (6), 0.91 g (0.0042 mol) of
di-tert-butyl(2,2-dimethylpropyl)phosphane (7), and 50 mL of
dioxane were stirred at room temperature for 24 h. The reaction
mixture was filtered, and the solvent was removed under vacuum.
The resulting mixture was purified by recrystallization from 20 mL
of ethanol. Yield of 2,6-bis(1-(2-methyl-3-(3-methylthiophene-2-
yl)phenyl)phenylimino)ethyl)pyridine (20) was 3.26 g (35%) as a
1
light yellow solid. H NMR (500 MHz, C6D6, TMS): δ 2.12 (s,
6H, Me), 2.32 (s, 6H, Me), 2.39 (s, 6H, Me), 6.50 (s, 2H, Arom-
H), 6.70 (m, 2H, Arom-H), 7.20 (m, 7H, Arom-H), 8.49 (d, JHH
) 7.8 Hz, 2H, Py-H). 13C NMR (500 MHz, C6D6, (selected
signals)): δ 166.6 (CdN). Anal. Calcd for C33H31N3S2 (MW
533.75): C, 74.26; H, 5.85; N, 7.87. Found: C, 74.29; H, 5.91; N,
7.95.
2,6-Bis(1-(2-methyl-3-bromophenylimino)ethyl)pyridine (15).
1-(6-Acetylpyridin-2-yl)ethanone (2) (18.87 g, 0.116 mol), 45.3 g
(0.243 mol) of 3-bromo-2-methylphenylamine (14), 300 mL of the
toluene, and a few crystals of para-toluenesulfonic acid were
refluxed with a Dean-Stark trap for 3 days until the calculated
amount of the water was separated (4.16 mL) under the flow of
nitrogen. The solvent was removed by a rotary evaporator, and the
resultant reaction mixture was recrystallized from 50 mL of ethanol.
The yield of 2,6-bis(1-(2-methyl-3-bromophenylimino)ethyl)pyri-
3
2,6-Bis(1-(2-methyl-3-(4-fluorophenyl)phenylimino)ethyl)-
pyridineiron(II) Chloride (21). Anhydrous iron(II) chloride (0.43
g, 0.0034 mol) was dissolved in 40 mL of warm n-butanol. Then,
2.0 g (0.0038 mol) of 2,6-bis(1-(2-methyl-3-(4-fluorophenyl)-
phenylimino)ethyl)pyridine (18) was added in one portion to the
reaction mixture. The mixture was kept at 40 °C additionally for 1
h and then was cooled to ambient temperature. The resultant blue
precipitate was filtered, washed twice with 20 mL of pentane, and
dried at 1 mm vacuum. Yield of 2,6-bis(1-(2-methyl-3-(4-fluo-
rophenyl)phenylimino)ethyl)pyridineiron(II) chloride (21) was 3.83
g (87%). Anal. Calcd for C35H29Cl2F2FeN3 (MW 656.37): C, 64.05;
H, 4.45; N, 6.40. Found: C, 64.23; H, 4.61; N, 6.48. Direct probe
MS: exact mass for C35H29Cl2F2FeN3 655.11; found 655.11.
1
dine (15) was 47.9 g (83%) as a pale yellow solid. H NMR (500
MHz, THF-D8, TMS): δ 2.15 (s, 6H, Me), 2.33 (s, 6H, Me), 6.67
(m, 2H, Arom-H), 7.35 (m, 4H, Arom-H), 7.90 (t, 3JHH ) 7.8 Hz,
3
1H, Py-H), 8.50 (d, JHH ) 7.8 Hz, 2H, Py-H). 13C NMR (500
MHz, THF-D8, (selected signals)): δ 168.0 (CdN). Anal. Calcd
for C23H21Br2N3 (MW 499.24): C, 55.33; H, 4.24; N, 8.42.
Found: C, 55.40; H, 4.42; N, 8.46.
2,6-Bis(1-(2-methyl-3-(4-fluorophenyl)phenylimino)ethyl)-
pyridine (18). 4-Fluorophenylboronic acid (5) (4.71 g, 0.0337 mol),
5.60 g (0.0112 mol) of 2,6-bis(1-(2-methyl-3-bromophenylimino)-
ethyl)pyridine (15), 10.97 g (0.0337 mol) of cesium carbonate (8),
0.77 g (0.00084 mol) of tris(dibenzylideneacetone)dipalladium(0)
2,6-Bis(1-(2-methyl-3-(3,5-bis(trifluoromethyl)phenyl)phe-
nylimino)ethyl)pyridineiron(II) Chloride (22). Anhydrous iron-

Production of R-Olefins Based on Iron(II)/(III) Complexes
Organometallics, Vol. 25, No. 12, 2006 2991
(II) chloride (0.50 g, 0.0039 mol) was dissolved in 40 mL of warm
n-butanol. Then, 3.0 g (0.0039 mol) of 2,6-bis(1-(2-methyl-3-(3,5-
bis(trifluoromethyl)phenyl)phenylimino)ethyl)pyridine (19) were
added in one portion to the reaction mixture. The mixture was kept
at 40 °C additionally for 1 h and then was cooled to ambient
temperature. The resultant blue precipitate was filtered, washed
twice with 20 mL of pentane, and dried at 1 mm vacuum. Yield of
2,6-bis(1-(2-methyl-3-(3,5-bis(trifluoromethyl)phenyl)phenylimino)-
ethyl)pyridineiron(II) chloride (22) was 3.83 g (87%). Anal. Calcd
for C39H27Cl2F12FeN3 (MW 892.38): C, 52.49; H, 3.05; N, 4.71.
Found: C, 52.57; H, 3.21; N, 4.96. The structure was proven by
X-ray analysis.
7.6 (m, 5H, aromatic protons). Anal. Calcd for C15H25P (MW
236.33): C, 76.23; H, 10.66; P, 13.11. Found: C, 76.15; H, 10.58;
P, 12.87.
2,6-Bis(1-(5-(3,5-bis(trifluoromethyl)phenyl)-6-methylphe-
nylimino)ethyl)pyridine (27). 3,5-Bis(trifluoromethyl)phenylbo-
ronic acid (16) (10.3 g, 0.04 mol), 5.0 g (0.01 mol) of 2,6-bis(1-
(2-methyl-5-bromophenylimino)ethyl)pyridine (26), 12.64 g (0.0388
mol) of cesium carbonate (8), 0.71 g (0.00078 mol) of tris-
(dibenzylideneacetone)dipalladium(0) (6), 0.55 g (0.00233 mol) of
benzyldi-tert-butylphosphane (28), and 50 mL of dioxane were
stirred at room temperature for 24 h. The reaction mixture was
filtered, and the solvent was removed under vacuum. The resulting
mixture was purified by recrystallization from 20 mL of ethanol.
Yield of 2,6-bis(1-(5-(3,5-bis(trifluoromethyl)phenyl)-6-methylphe-
nylimino)ethyl)pyridine (27) was 4.45 g (58%) as a light yellow
solid. 1H NMR (500 MHz, C6D6, TMS): δ 2.10 (s, 6H, Me), 2.32
(s, 6H, Me), 6.80 (s, 2H, Arom-H), 6.85 (m, 2H, Arom-H), 7.11
(m, 2H, Arom-H), 7.30 (m, 2H, Arom-H), 7.40 (t, 3JHH ) 7.8 Hz,
1H, Py-H), 7.60 (s, 2H, Arom-H), 7.80 (s, 4H, Arom-H), 8.50 (d,
3JHH ) 7.8 Hz, 2H, Py-H). 13C NMR (500 MHz, C6D6, (selected
signals)): δ 167.3 (CdN). 19F NMR (C6D6): -63.07 (s, 12F). Anal.
Calcd for C39H27F12N3 (MW 765.63): C, 61.18; H, 3.55; N, 5.49.
Found: C, 61.19; H, 3.50; N, 5.57.
2,6-Bis(1-(2-methyl-3-(3-methylthiophene-2-yl)-phenyl)phe-
nylimino)ethyl)pyridineiron(II) Chloride (23). 2,6-Bis(1-(2-meth-
yl-3-(3-methylthiophene-2-yl)phenyl)phenylimino)ethyl)pyridine (20)
(0.5 g, 0.00094 mol) was dissolved in 30 mL of THF. Then 0.11
g (0.00087 mol) of iron(II) chloride was added in the reaction
mixture in one portion. The resultant blue precipitate was filtered
after 12 h of stirring, washed twice with 20 mL of pentane, and
dried at 1 mm vacuum. Yield of 2,6-bis(1-(2-methyl-3-(3-meth-
ylthiophene-2-yl)phenyl)phenylimino)ethyl)pyridineiron(II) chloride
(23) was 1.34 g (84%). Anal. Calcd for C33H31Cl2FeN3S2 (MW
660.50): C, 60.01; H, 4.73; N, 6.36. Found: C, 60.25; H, 4.90; N,
6.39. Direct probe MS: exact mass for C33H31Cl2FeN3S2: exact
mass 659.07; found exact mass 659.07.
2,6-Bis(1-(5-(3,5-bis(trifluoromethyl)phenyl)-6-methylphe-
nylimino)ethyl)pyridineiron(II) Chloride (30). 2,6-Bis(1-(5-(3,5-
bis(trifluoromethyl)phenyl)-6-methylphenylimino)ethyl)pyridine (27)-
(0.9 g, 0.00118 mol) was dissolved in 30 mL of THF. Iron(II)
chloride (0.13 g, 0.001 mol) was added to the reaction mixture in
one portion. The resultant blue precipitate was filtered after 12 h
of stirring, washed twice with 20 mL of pentane, and dried at 1
mm vacuum. Yield of 2,6-bis(1-(5-(3,5-bis(trifluoromethyl)phenyl)-
6-methylphenylimino)ethyl)pyridineiron(II) chloride (30) was 0.95
g (83%). Anal. Calcd for C39H27Cl2F12FeN3 (MW 892.38): C,
52.49; H, 3.05; N, 4.71. Found: C, 52.54; H, 3.14; N, 4.78. The
structure was proven by X-ray analysis.
2,6-Bis(1-(2-methyl-3-(3,5-bis(trifluoromethyl)phenyl)phen-
ylimino)ethyl)pyridineiron(III) Chloride (24). 2,6-Bis(1-(2-meth-
yl-3-(3,5-bis(trifluoromethyl)phenyl)phenylimino)ethyl)pyridine (19)
(1.0 g, 0.0013 mol) was dissolved in 50 mL of THF. Then, 0.20 g
(0.0012 mol) of anhydrous iron(III) chloride was added in one
portion to the reaction mixture. The mixture was kept stirring for
20 min at ambient temperature. The prolonged stirring caused the
reduction of Fe(III) to Fe(II). The resultant orange precipitate was
filtered, washed twice with 20 mL of pentane, and dried at 1 mm
vacuum. Yield of 2,6-bis(1-(2-methyl-3-(3,5-bis(trifluoromethyl)-
phenyl)phenylimino)ethyl)pyridineiron(III) chloride (24) was 0.88
g (73%). Anal. Calcd for C39H27Cl3F12FeN3 (MW 927.84): C,
50.48; H, 2.93; N, 4.53. Found: C, 50.59; H, 3.20; N, 4.57. Direct
probe MS: exact mass for C39H27Cl3F12FeN3 926.04; found 926.04.
1-{6-[1-(4-Bromo-2,6-dimethylphenylimino)ethyl]pyridin-2-
yl}ethanone (31). 1-(6-Acetylpyridin-2-yl)ethanone (22.5 g, 0.138
mol), 25.0 g (0.125) of 4-bromo-2,6-dimethylphenylamine, and 300
mL of n-propanol with a few crystals of p-toulenesulfonic acid were
stirred at room temperature for 36 h in a 500 mL flask under the
flow of nitrogen. The resultant yellow precipitate was filtered and
washed with 20 mL of methanol. It was then dried at 1 mm vacuum
overnight. The yield of 1-{6-[1-(4-bromo-2,6-dimethylphenylimi-
no)ethyl]pyridin-2-yl}ethanone (31) was 19.08 g (44%) as a yellow
2,6-Bis(1-(2-methyl-5-bromophenylimino)ethyl)pyridine (26).
1-(6-Acetylpyridin-2-yl)ethanone (2) (8.33 g, 0.051 mol), 20.0 g
(0.107 mol) of 5-bromo-2-methylphenylamine (25), and 200 mL
of dry toluene with a few crystals of para-toluenesulfonic acid were
refluxed in a Dean-Stark trap for 3 days until the calculated amount
of water was separated (1.84 mL) under a flow of nitrogen. The
solvent was removed in a rotary evaporator, and the resultant
reaction mixture was recrystallized from 50 mL of ethanol. The
yield of 2,6-bis(1-(2-methyl-5-bromophenylimino)ethyl)pyridine
(26) was 19.88 g (78%) as a pale yellow solid. 1H NMR (500 MHz,
C6D6, TMS): δ 1.90 (s, 6H, Me), 2.12 (s, 6H, Me), 6.50 (m, 2H,
Arom-H), 7.20 (m, 4H, Arom-H), 7.30 (t, 3JHH ) 7.8 Hz, 1H, Py-
H), 8.40 (d, 3JHH ) 7.8 Hz, 2H, Py-H). 13C NMR (500 MHz, C6D6,
(selected signals)): δ 167.4 (CdN). Anal. Calcd for C23H21Br2N3
(MW 499.24): C, 55.33; H, 4.24; N, 8.42. Found: C, 55.48; H,
4.45; N, 8.53.
1
solid. H NMR (500 MHz, CD2Cl2, TMS): δ 1.95 (s, 6H, Me),
2.22 (s, 3H, Me), 2.30 (s, 3H, Me), 6.80 (s, 2H, Arom-H), 7.95 (t,
3
3JHH ) 8.0 Hz, 1H, Py-H), 8.15 (d, JHH ) 8.0 Hz, 1H, Py-H),
8.40 (d, 3JHH ) 8.0 Hz, 1H, Py-H). 13C NMR (500 MHz, CD2Cl2,
TMS (selected signals)): δ 168.4 (CdN), 199.5 (CdO). Anal.
Calcd for C17H17BrN2O (MW 345.23): C, 59.14; H, 4.96; N, 8.11.
Found: C, 59.18; H, 5.07; N, 8.15.
4-Bromo-2,6-dimethylphenyl-(1-{6-[1-(3,5-dibromo-4-meth-
ylphenylimino)ethyl]pyridin-2-yl}ethylidene)amine (32). 1-{6-
[1-(4-Bromo-2,6-dimethylphenylimino)ethyl]pyridin-2-yl}ethan-
one (31) (5.0 g, 0.0145 mol), 4.45 g (0.016 mol) of 3,5-dibromo-
4-methylphenylamine, 100 g of fresh molecular sieves, and 100
mL of toluene were kept at 90 °C for 3 days under the flow of
nitrogen. The solvent was removed by a rotary evaporator, and the
residue was recrystallized from 10 mL of ethanol. The yield of
4-bromo-2,6-dimethylphenyl-(1-{6-[1-(3,5-dibromo-4-methylphe-
nylimino)ethyl]pyridin-2-yl}ethylidene)amine (32) was 5.75 g
Benzyldi-tert-butylphosphane (28). Di-tert-butylchlorophos-
phine (10) (75.0 g, 0.415 mol) and 0.5 mol of a 2.0 M solution of
benzylmagnesium chloride (29) in THF (200 mL) were refluxed
under argon for 2 days. The reaction mixture was allowed to cool
to RT, and an aqueous solution of ammonium chloride was added
slowly. The organic phase was separated and dried with magnesium
sulfate. After removal of the solvent, the product was purified by
distillation in a vacuum. The yield of benzyldi-tert-butylphosphane
(28) was 94.3 g (96%) with bp 56-59 °C/0.1 mm. 31P NMR
(CDCl3): +36.63 ppm. 1H NMR (CDCl3): δ 1.18 (s, 9H, Me3C),
1
(67%) as a yellow solid. H NMR (500 MHz, CD2Cl2, TMS): δ
1.60 (s, 6H, Me), 2.10 (s, 3H, Me), 2.15 (s, 3H, Me), 2.45 (s, 3H,
Me), 6.97 (s, 2H, Arom-H), 7.15 (s, 2H, Arom-H), 7.40 (t, 3JHH
)
3
8.0 Hz, 1H, Py-H), 8.20 (d, JHH ) 8.0 Hz, 1H, Py-H), 8.40 (d,
3JHH ) 8.0 Hz, 1H, Py-H). 13C NMR (500 MHz, CD2Cl2, TMS
(selected signals)): δ 168.5 (CdN), 164.4 (CdN). Anal. Calcd for
2
1.20 (s, 9H, Me3C), 2.90 (d, JPH ) 2.92 Hz, P-CH2-Ph), 7.1-

2992 Organometallics, Vol. 25, No. 12, 2006
Ionkin et al.
C24H22Br3N3 (MW 592.16): C, 48.68; H, 3.74; N, 7.10. Found:
C, 48.69; H, 3.82; N, 7.17.
Zipperclave. Catalyst and cocatalyst are charged separately using
stainless steel injection tubes. The complexes 1, 21, 22, 23, 24, 30,
and 34 are activated by modified methylaluminoxane (MMAO).
The steps for a typical oligomerization follow. The injectors are
charged in a glovebox. The cocatalyst is obtained as a 7 wt %
solution of aluminum in o-xylene and is charged as such into the
injector assembly along with a 10 mL chase of o-xylene. The
catalysts are prepared as suspensions in o-xylene (10 mg/100 mL).
A sample is pulled from a well-stirred suspension and is added to
a 10 mL charge of o-xylene. The injectors are attached to autoclave
ports equipped with dip tubes. Nitrogen is sparged through the loose
fittings at the attachment points prior to making them tight. The
desired charge of o-xylene is then pressured into the autoclave.
The agitator and heater are turned on. When the desired temperature
is reached, the cocatalyst is charged to the clave by blowing ethylene
down through the cocatalyst injector. After a significant pressure
rise is seen in the autoclave to indicate the cocatalyst and chase
solvent have entered, the injector is isolated from the process using
its valves. The pressure controller is then set to 5 psig below the
desired ethylene operating pressure and is put in the automatic mode
to allow it to control the operation of the ethylene addition valve.
When the pressure is 5 psig below the desired operating pressure,
the controller is put into manual mode and the valve is set to 0%
output. When the batch temperature is stable at the desired value,
the catalyst is injected using enough nitrogen such that the reactor
pressure is boosted to the desired pressure. At the same time as
the catalyst injection, the pressure controller is put in the automatic
mode and the oligomerization is underway. The 5 psi boost is
obtained routinely by having a small reservoir between the nitrogen
source and the catalyst injector. A valve is closed between the
nitrogen source and the reservoir prior to injecting the catalyst, so
the same volume of nitrogen is used each time to inject the catalyst
suspension. To stop the oligomerization, the pressure controller is
put into manual, the ethylene valve is closed, and the reactor is
cooled. The reaction time is 1 h.
4-(3,5-Bis(trifluoromethyl)phenyl)-2,6-dimethylphenyl)-(1-{6-
[1-(3,5-(3,5-bis(trifluoromethyl)phenyl)-4-methylphenylimino)-
ethyl]pyridin-2-yl}ethylidene)amine (33). 3,5-Bis(trifluoromethyl)-
phenylboronic acid (16) (9.92 g, 0.0385 mol), 3.80 g (0.0096 mol)
of 4-bromo-2,6-dimethylphenyl-(1-{6-[1-(3,5-dibromo-4-meth-
ylphenylimino)ethyl]pyridin-2-yl}ethylidene)amine (32), 12.54 g
(0.0385 mol) of cesium carbonate (8), 0.88 g (0.00096 mol) of tris-
(dibenzylideneacetone)dipalladium(0) (6), 0.50 g (0.0023 mol) of
di-tert-butyl(2,2-dimethylpropyl)phosphane (7), and 50 mL of
dioxane were stirred at room temperature for 24 h. The reaction
mixture was filtered, and the solvent was removed under vacuum.
The resulting mixture was purified by recrystallization from 20 mL
of ethanol. Yield of 4-(3,5-bis(trifluoromethyl)phenyl)-2,6-dimeth-
ylphenyl)-(1-{6-[1-(3,5-(3,5-bis(trifluoromethyl)phenyl)-4-meth-
ylphenylimino)ethyl]pyridin-2-yl}ethylidene)amine (33) was 5.14
1
g (76%) as a light yellow solid. H NMR (500 MHz, CD2Cl2,
TMS): δ 1.55 (s, 3H, Me), 1.90 (s, 6H, Me), 2.30 (s, 3H, Me),
2.49 (s, 3H, Me), 6.40 (s, 2H, Arom-H), 6.97 (s, 2H, Arom-H),
7.15 (s, 2H, Arom-H), 7.40 (t, 3JHH ) 8.0 Hz, 1H, Py-H), 7.60 (s,
4H, Arom-H), 8.20 (d, 3JHH ) 8.0 Hz, 1H, Py-H), 8.40 (d, 3JHH
)
8.0 Hz, 1H, Py-H). 13C NMR (500 MHz, CD2Cl2, (selected
signals)): δ 168.0 (CdN), 167.2 (CdN). 19F NMR (CD2Cl2) δ
-62.96 (s, 12F), -63.00 (s, 6F). Anal. Calcd for C48H31F18N3 (MW
991.75): C, 58.13; H, 3.15; N, 4.24. Found: C, 58.25; H, 3.18; N,
4.30.
4-(3,5-Bis(trifluoromethyl)phenyl)-2,6-dimethylphenyl)-(1-{6-
[1-(3,5-(3,5-bis(trifluoromethyl)phenyl)-4-methylphenylimino)-
ethyl]pyridin-2-yl}ethylidene)amineiron(II) Chloride (34). 4-(3,5-
Bis(trifluoromethyl)phenyl)-2,6-dimethylphenyl)-(1-{6-[1-(3,5-(3,5-
bis(trifluoromethyl)phenyl)-4-methylphenylimino)ethyl]pyridin-2-
yl}ethylidene)amine (33) (0.5 g, 0.0005 mol) was dissolved in 20
mL of THF. Then 0.064 g (0.0005 mol) of iron(II) chloride was
added to the reaction mixture in one portion. The resultant blue
precipitate was filtered after 12 h of stirring, washed twice with 20
mL of pentane, and dried at 1 mm vacuum. Yield of 4-(3,5-bis-
(trifluoromethyl)phenyl)-2,6-dimethylphenyl)-(1-{6-[1-(3,5-(3,5-bis-
(trifluoromethyl)phenyl)-4-methylphenylimino)ethyl]pyridin-2-yl}-
ethylidene)amineiron(II) chloride (34) was 0.95 g (83%). Anal.
Calcd for C48H31Cl2F18FeN3 (MW 1118.50): C, 51.54; H, 2.79;
N, 3.76. Found: C, 51.57; H, 3.02; N, 3.94. The structure was
proven by X-ray analysis.
Acknowledgment. The authors wish to thank Dr. Jerald
Feldman for proofreading the manuscript.
Supporting Information Available: Crystallographic informa-
tion (CIF file) for compounds 1, 13, 15, 22, 30, and 34. These
materials are available free of charge via the Internet at http://
pubs.acs.org.
General Conditions of the Oligomerizations. Ethylene oligo-
merizations are done in a 1 L stainless steel Autoclave Engineering
OM051069O