Separation of a diiminopyridine iron(II) complex into rac- and meso- diastereoisomers provides evidence for a dual stereoregulation mechanism in propene polymerization

Article abstract of DOI:10.1039/b810381j

Separation of a diiminopyridine iron(ii) complex into its rac- and meso- diastereoisomers provides for first time the opportunity of observing the enantiomorphic site control competing with the chain-end control mechanism in a non-metallocene catalyst system. The Royal Society of Chemistry.

Full text of DOI:10.1039/b810381j

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Separation of a diiminopyridine iron(II) complex into rac- and
meso- diastereoisomers provides evidence for a dual stereoregulation
mechanism in propene polymerizationw
a
a
a
a
Antonio Rodrıguez-Delgado, Juan Campora,* A. Marcos Naz, Pilar Palma and
´
´
Manuel L. Reyesb
Received (in Cambridge, UK) 18th June 2008, Accepted 13th August 2008
First published as an Advance Article on the web 12th September 2008
DOI: 10.1039/b810381j
Separation of a diiminopyridine iron(II) complex into its rac- and
meso- diastereoisomers provides for first time the opportunity of
observing the enantiomorphic site control competing with the
chain-end control mechanism in a non-metallocene catalyst
system.
should allow the identification of site control effects, even when
they might be minor.6 However resolution of diiminopyridine
iron complexes into stereoisomerically pure rac and meso forms
has never been achieved, and therefore their activities and
selectivities have never been compared. Even the mere existence
of Ni, Fe or Co catalysts in conformationally stable stereo-
isomers in solution counts upon little direct evidence, since their
paramagnetism hinders their study by NMR techniques. In this
communication we identify and resolve the two diastereoiso-
meric forms of complex 1 (Scheme 1), and we compare their
performance as propene polymerization catalysts. We find that,
together with the chain-end isotactic stereoselection, an enan-
tiomorphic metal-site mechanism operates for the C2 diaster-
eoisomer. This demonstrates for the first time that
stereoselective olefin polymerization through rational ligand
design can be achieved with late transition metal, non-
metallocene catalysts.
Stereoselectivity control in olefin polymerization through cata-
lyst design is a spectacular success of homogeneous catalysis.1
Although many catalysts, including heterogeneous Ziegler–
Natta systems, exhibit high stereoselectivity, this is often con-
trolled by the chirality centers of the growing polymer molecule
(chain-end control). Rational design of stereoselective olefin
polymerization catalysts only becomes feasible when the stereo-
chemistry of the monomer insertion is imposed by the symmetry
of the active center environment (enantiomorphic site control).2
The latter is exemplified by the stereoselective polymerization of
propene into isotactic polypropylene by chiral C2 ansa-metallo-
cenes, first developed by Kaminsky and Brintzinger.3 In spite of
the importance of this concept, very few attempts have been
made to extend it beyond the family of metallocene catalysts.4
Catalysts based on imine ligands, such as a-diimines (Ni, Pd)
and 2,6-bis(imino)pyridines (Fe, Co) displaying unsymmetri-
cally substituted aryl groups exist as C2 (chiral racemic mixture)
and Cs (meso) diastereoisomers (Fig. 1) that are formally
analogous to those involved in the metallocene family. Several
authors have examined the stereochemistry of propene poly-
merization by such catalysts.5–8 Pellecchia described the synth-
esis and isolation of configurationally stable a-diimine ligands,
and studied the catalytic performance of the corresponding
nickel complexes.5 Unfortunately, in this case the influence of
the isoselective site control by the chiral C2 configuration of the
diimine ligand is blurred by a concurrent syndio-selective
Complex 1 is readily prepared by the standard procedure
that involves the reaction of hydrated iron(II) chloride and the
corresponding 2,6-diiminopyridine ligand in THF (Scheme 1).
The product precipitates out as an analytically pure blue solid.
Although paramagnetic, iron(II) 2,6-diiminopyridine com-
plexes give rise to useful 1H NMR spectra.9 Samples of
complex 1 in CD2Cl2 display two sets of signals, which indicate
its occurrence in solution as a mixture of isomers. The isomer
ratio gradually changes with time, attaining a 1 : 1 equilibrium
ratio within 1–2 days at room temperature. Kinetic measure-
ments between 296 and 326 K afford the following activation
parameters for relaxation to the equilibrium mixture:
DHz = 25.5(8) kcal molꢀ1, DSz = 5(3) cal molꢀ1
K
ꢀ1, and
DGz = 24.0(8) kcal molꢀ1
.
298
chain-end stereocontrol mechanism.
A chain-end control
mechanism also operates in the propene polymerization by
2,6-diiminopyridine iron complexes, but in contrast with the
above mentioned Ni catalysts, it leads to prevailingly isotactic
products. In principle, comparison of the different degrees of
isotacticity generated by the chiral and achiral diastereoisomers
a
´
Instituto de Investigaciones Quımicas, Consejo Superior de
Investigaciones Cientıficas-Universidad de Sevilla, Avda. Americo
´
Vespucio, 49, 41092 Sevilla, Spain. E-mail: campora@iqq.csic.es;
Fax: +34 954460565; Tel: +34 954489303
´
Repsol-YPF CTR, Ctra. Extremadura NV, Km 18, 28931 Mostoles,
Spain
b
w Electronic supplementary information (ESI) available: Experimental
procedures and characterization data for new compounds. See DOI:
10.1039/b810381j
Fig. 1 Racemic and meso diastereoisomers of non-metallocene poly-
merization catalysts (M0 = Ni, Pd; M = Fe, Co).
ꢁc
This journal is The Royal Society of Chemistry 2008
5230 | Chem. Commun., 2008, 5230–5232

View Article Online
position 4 of the central pyridine ring. We have previously
shown that this kind of modification has little impact on either
the spectroscopic or catalytic properties of Fe or Co diimino-
pyridine complexes, and thus they are very well suited for
comparative purposes.10 Although the higher solubility of 2
hampers the selective precipitation of the kinetic isomer, we
found that fractional crystallization of this compound affords
microcrystalline samples enriched (480%) in a single isomer,
2m. On gentle warming in solution, 2m isomerizes to 2r until a
1 : 1 equilibrium mixture is attained. While the NMR spec-
trum of 2m exhibits a single 6H signal for the CMe2 group at d
2.9 ppm, the two methyls become inequivalent and give rise to
two 3H resonances at d 3.1 and 2.7 ppm in 2r. This observa-
tion proves that 2m is the achiral meso complex; 2r must be the
racemic mixture of chiral diastereoisomers. Brintzinger re-
ported that in some ansa-zirconocenes, the meso form is the
least soluble of the diastereoisomeric pair.11 Since the meso
form 2m displays appreciably lower solubility than 2r, it
becomes very likely that the same relationship would exist in
the case of 1, leading to the conclusion that the kinetic isomer
1r corresponds to the racemic mixture of diastereoisomers.
This conclusion is further supported by the analogous trends
observed in the chemical shifts of homologous NMR signals
for the two diastereoisomeric pairs. In general, whenever the
homologous pairs of signals are well resolved in the spectrum
of a mixture of diastereoisomers, the identity of the high/low
field signals is the same in the 1r/1m and 2r/2m systems. For
example, the signals assigned to the m-H atoms of the N-aryl
groups of 1r give rise to two very close signals at d 15.10 and
15.96 ppm. In the spectra of 1r + 1m mixtures, these signals of
1r are flanked by those corresponding to 1m (d 16.59,
14.37 ppm), giving rise to a conspicuous feature. The same
trend is followed in the 2 system, where the two m-H signals of
the rac diastereoisomer (d 16.14, 14.86 ppm) also come flanked
by the more separated ones of the meso (d 16.29, 14.76 ppm).
Similar relationships have been established for the signals
corresponding to the imino methyls, the isopropyl CH and
Me groups, and the pyridine ring H3 atoms.
Scheme 1
The initial isomeric ratio of the solutions depends on the
origin of the solid sample. Fresh solutions prepared from the
precipitated solid are nearly diastereoselectively pure, as they
contain more than 90% of a single isomer (1r). If this solution
is allowed to equilibrate and then rapidly evaporated, the solid
residue will contain a mixture of the two isomers.
However, when the equilibrium solutions are allowed to
evaporate slowly, the second isomer, 1m, precipitates selec-
tively. These observations indicate that, while 1r is formed
under kinetic control, the precipitation of 1m is favored due to
its lower solubility. The latter proposal can be readily tested by
selective extraction of solid samples containing mixtures of
both diastereoisomers with dichloromethane, which leaves a
solid residue of 1m. The different solubility of the two isomers
provides an efficient method for the preparation of 1m.
Stirring a suspension of a solid diastereomeric mixture of 1
in hexane–dichloromethane for several days causes 1r to
gradually dissolve and equilibrate with 1m, which then
precipitates until all the sample becomes converted.
A comparison of the NMR spectra of 1r and 1m is shown in
Fig. 2. As can be seen, although distinctly recognizable, the
homologous pairs of signals display very similar chemical
shifts and widths, providing no hints at all on the isomers’
identity. In spite of our efforts, we were unable to grow X-ray
quality crystals of any of them. Brookhart and Small found
similar difficulties but reported the crystal structure of a
binuclear mixed valence complex probably arising from the
oxidation of 1, which exhibited the meso configuration.6 In
order to provide the necessary evidence for positive stereo-
chemical identification, we synthesized complex 2, bearing the
prochiral group 2-methyl-2-phenylpropyl (neophyl) bound to
The availability of complex 1 in nearly pure rac and meso
diastereoisomeric forms allowed us to compare their catalytic
performance in propene polymerization. The experiments
were carried out in a Buchi glass reactor with a circulating
¨
cooling bath set to ꢀ10 1C. Monitoring the internal tempera-
ture showed that it never rose above ꢀ5 1C. Extrapolation of
the kinetic parameters indicates that the 1r/1m interconversion
rate is negligible at this temperature, and hence it can be
reasonably assumed that the catalyst is stereochemically stable
under the reaction conditions. Activity and polymer charac-
terization data are given in Table 1. As can be seen, the
experiments have good reproducibility, and thus the averaged
data will be discussed.
The diastereoisomeric forms of 1 exhibit significant perfor-
mance differences. The chiral form is about three times as
active as the meso, and gives rise to slightly higher molecular
weight polypropylene. The two polymers are prevailingly
isotactic, with no detectable content of syndiotactic rrrr
pentads. The relatively high isotactic content (450%) of the
polymer produced by the achiral meso diastereoisomer indi-
cates that, in agreement with previous reports,7 most of the
Fig. 2 1H NMR spectra of complex 1r, directly precipitated from the
reaction mixture (A); 1r + 1m equilibrium mixture (B); 1m (C).
ꢁc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 5230–5232 | 5231

View Article Online
Table 1 Propene polymerization data for catalysts 1m and 1ra
Entry
Catalystb
Activityc
Mw ꢂ 10ꢀ3
Mn/Mw
%m4
%mmrr
%mmrm
mmrm/mmrr
1
2
3
4
Av.
5
6
7
8
1m
1m
1m
1m
1m
1r
1r
1r
1r
1r
118
83
77
86
10.6
4.8
8.3
7.4
7.8
7.5
7.6
7.0
8.4
11.0
8.3
1.6
2.0
1.4
1.9
1.6
1.6
1.4
1.7
1.5
1.1
1.5
51.5
52.6
53.9
53.2
52.8
60.8
66.7
61.6
64.3
63.0
63.3
4.2
4.2
4.0
4.2
4.2
3.7
3.3
3.8
3.7
4.3
3.8
18.6
17.9
17.0
16.3
17.5
10.8
9.8
9.1
9.2
10.9
10.0
4.47
4.26
4.22
3.88
4.21
2.93
3.01
2.39
2.49
2.51
2.65
91
355
367
357
251
354
337
9
Av.
1r
a
b
Reaction conditions: Solvent, toluene, 100 mL, ꢀ6 1C, 2.3 bar, 40–60 min. Catalyst, 7.2 mmol, co-catalyst, MMAO, Al/Fe = 1000.
c
Activity/kg PP (mol Fe)ꢀ1 hꢀ1 barꢀ1
.
stereoselectivity arises from an effective isoselective chain-end
control. However, the polypropylene produced by 1r displays
a distinct excess of isotactic pentads, stemming from a con-
current isospecific enantiomorphous site control mechanism.
This conclusion is further supported by the analysis of the
minor pentad signals in the polymers 13C NMR spectra. Those
produced with 1m display two resonances of similar intensity
attributed to the mmmr and mmrm pentads. This is expected
for stereoblock polymers characteristic of the chain-end
stereoselection mechanism. The higher isotactic content of
the polypropylenes prepared with 1r is mostly due to the
decrease of the frequency of these sequences, while the number
of pentads mmrr and mrrm (approximate intensity ratio 2 : 1),
resulting of isolated stereoerrors characteristic of site control,
decrease only slightly. For the chiral catalyst 1r, the relative
weight of this second kind of propagation ‘‘mistakes’’, esti-
mated from the mmrm to mmrr ratio, becomes nearly double
than that for the achiral 1m, as expected for a significant
contribution of the site control mechanism. The simultaneous
operation of two stereoselection mechanisms has been pre-
viously observed with a metallocenic catalyst, and has led to
polypropylene products whose NMR spectra match very
closely those of our samples.12
are gratefully acknowledged. A.R.-D. thanks the Ministerio de
Educacion y Ciencia for a Ramon y Cajal fellowship,
predoctoral research grant from CSIC
´
´
and A.M.N.
a
(I3P program).
Notes and references
1 (a) W. A. Herrmann and B. Cornills, Angew. Chem., Int. Ed. Engl.,
1997, 36, 1048; (b) Stereoselective Polymerization with Single-Site
Catalysts, ed. L. S. Baugh and J. A. Canich, CRC Press, Boca
Raton, FL, 2008.
2 (a) L. Resconi, L. Cavallo, A. Fait and F. Piemontesi, Chem. Rev.,
2000, 100, 1253; (b) G. W. Coates, J. Chem. Soc., Dalton Trans.,
2002, 467; (c) Metallocene-Based Polyolefins, ed. J. Scheirs and W.
Kaminsky, Wiley, Chichester, UK, 1999, vol. 1.
3 (a) W. Kaminsky, K. Kulper, H.-H. Brintzinger and F. R. W. P.
¨
Wild, Angew. Chem., Int. Ed. Engl., 1985, 24, 507; (b) F. R. W. P.
Wild, M. Wasiucionek, G. Huttner and H.-H. Brintzinger,
J. Organomet. Chem., 1985, 288, 63.
4 M. Lamberti, M. Mazzeo, D. Pappalardo, A. Zambelli and C.
Pellecchia, Macromol. Symp., 2004, 213, 235.
5 D. Pappalardo, M. Mazzeo, S. Antinucci and C. Pellechia, Macro-
molecules, 2000, 33, 9483.
6 B. L. Small and M. Brookhart, Macromolecules, 1999, 32, 2120.
7 C. Pellechia, D. Pappalardo and M. Mazzeo, Macromol. Rapid
Commun., 1998, 19, 651.
In summary, we have shown that the same principles used
for the design of highly stereoselective metallocene polymer-
ization catalysts can also be applied to other systems, such as
the iron(II) polymerization catalysts. The stereoselectivity of
this type of catalysts could be improved by suitable ligand
design. Our results also suggest that diiminopyridine ligands
can be used to impart enantioselectivity to other catalytic
reactions where iron(II) complexes have been shown to be
active, such as hydrogenation and hydrosilation.13
8 H. Zou, F. M. Zhu, Q. Wu, J. Y. Ai and S. A. Lin, J. Polym. Sci.,
Part A: Polym. Chem., 2005, 43, 1325.
9 (a) I. Fernandez, R. J. Trovitch, E. Lobkovsky and P. J. Chirik,
´
Organometallics, 2007, 27, 109; (b) S. C. Bart, E. Lobkovsky, E.
Bill, K. Wieghardt and P. J. Chirik, Inorg. Chem., 2007, 46, 7055.
´
10 J. Campora, M. A. Naz, P. Palma, A. Rodrıguez-Delgado, E.
Alvarez, I. Tritto and L. Boggioni, Eur. J. Inorg. Chem., 2008,
1871.
11 Wiesenfeldt, A. Reinmuth, E. Barsties, E. Kaspar and H.-H.
Brintzinger, J. Organomet. Chem., 1989, 369, 359.
12 Erker, R. Nolte, Y.-H. Tsay and C. Kruger, Angew. Chem., Int. Ed.
Engl., 1989, 28, 628.
¨
Financial support from the DGI (Project CTQ2006-05527/
BQU), European Union (Network of Excellence IDECAT,
contract No NMP3-CT-2005-011730) and Junta de Andalucıa
´
13 (a) S. Bart, E. Lobkovsky and P. J. Chirik, J. Am. Chem. Soc.,
2004, 126, 13794; (b) M. Archer, M. W. Bouwkamp, M.-P. Cortez,
E. Lobkovsky and P. J. Chirik, Organometallics, 2006, 25, 4269.
ꢁc
This journal is The Royal Society of Chemistry 2008
5232 | Chem. Commun., 2008, 5230–5232