Same ligand, different metals: Diiodo-salan complexes of the group 4 triad in isospecific polymerization of 1-hexene and propylene

Article abstract of DOI:10.1021/ma902626e

The synthesis of a salan ligand featuring diiodo-phenolate rings, the dibenzyl complexes of the ligand and the polymerization of 1-hexene and propylene by the complexes was reported. The zirconium complex led to the most active catalyst with an activity of 3800g mmol^-1 h^-1, the hafnium and titanium complexes exhibited lower activities of 1500g mmol ^-1 h^-1. GPC analysis showed that the Zr catalyst led to a low molecular weight poly(1-hexene) of M_w=1400 and PDI=2.0. The Zr catalyst led to poly-(1-hexene) of 33%, which is low, but higher than that obtained from the corresponding dibromo-Salan-Zr complex. Chain-end analysis of the low-M_w polypropylene derived from the Zr catalyst revealed vinylidene and n-propyl groups as major chain ends, supporting a regular 1,2-insertion prior to a hydride-transfer termination. Several polypropylene samples produced by the titanium catalyst were analyzed by differential scanning calorimetry, which showed the melting transitions in the temperatures of 120.9-123.8°C.

Full text of DOI:10.1021/ma902626e

Macromolecules 2010, 43, 1689–1691 1689
DOI: 10.1021/ma902626e
Scheme 1. Synthesis of the Ligand, the Dibenzyl Complexes, and the
Polymers Described in This Work
Same Ligand, Different Metals: Diiodo-Salan
Complexes of the Group 4 Triad in Isospecific
Polymerization of 1-Hexene and Propylene
Ad Cohen,† Jacob Kopilov,† Marina Lamberti,‡
Vincenzo Venditto,‡ and Moshe Kol*,†
†School of Chemistry, Raymond and Beverly Sackler Faculty of
Exact Sciences, Tel Aviv University, Ramat Aviv, Tel Aviv 69978,
‡
ꢀ
Israel, and Dipartimento di Chimica, Universita di Salerno, Via
Ponte don Melillo, I-84084 Fisciano (SA), Italy
Received November 27, 2009
Revised Manuscript Received January 20, 2010
The design of catalysts that will enable the living polymeriza-
tion of propylene with high activities and isoselectivities under
ambient conditions remains a challenge in current polymer
chemistry.1 Such polymerization may give rise to specialty
polymers like elastomers, obtained by block copolymerization
of ethylene and propylene by sequential monomer addition.2,3
Since metallocene catalysts do not lead to living polymerization
under ambient temperatures, the search for such catalysts lies
in the regime of group 4 metal non-metallocenes.4 In the past
decade, several noteworthy systems were introduced that led to
isoselective polymerization of high olefins such as 1-hexene.
While the degree of the polymers isotacticity is determined by the
ability of the ligand(s) bound to the metal center to direct the
incoming olefin via a preferred enantioface, it is also strongly
affected by the bulk of the olefin, with higher olefins commonly
giving rise to more stereoregular polymers. Thus, the Salan-,5
the Cp*/amidinate-,6 and the amidomethylpyrrolidepyridine-
zirconium7 catalysts that led to almost perfectly isotactic poly-
(1-hexene) gave reduced tacticities in polypropylene of [mmmm]
of 79%,8 71%,9,10 and 73%,7 respectively. A similar trend was
reported for the binaphthyl-bridged Salen-zirconium catalyst
(81f11% [mmmm]).11
Zirconium-Salan complexes were found to lead to isoselective
polymerization of 1-hexene only if the ligands included bulky
ortho-substituents on the phenolate rings.5 Their activity was low.
Replacing the bulky alkyl groups with electron-withdrawing
chloro or bromo groups led to higher activities but to stereo-
irregular poly(1-hexene). We reasoned that binding the same
ligands to a smaller metal atom, i.e., titanium, may intensify their
effective bulk, leading to increased degree of isotacticity, which
was indeed found, but was not high.12 In this work we describe
the synthesis of a Salan ligand featuring diiodo-phenolate rings,
the dibenzyl complexes of this ligand with the group 4 triad
metals, and the polymerization of 1-hexene and propylene by
these complexes. The polymerization of propylene by titanium-
Salan complexes was not reported previously, and the activity of
such a catalyst, revealed herein, is quite remarkable.
derivatives of the group 4 triad to give the corresponding
LigMBn2 (M=Zr, Hf, Ti) complexes as yellow, white, and brown
solids, respectively, that could be stored at -30 °C as solids. The
zirconium and hafnium complexes were relatively stable in
toluene solution at RT, whereas the titanium complex decom-
posed readily. 1H NMR characterization indicated that all
complexes formed as single isomers of C2 symmetry, signifying
a fac-fac ligand wrapping and cis relationship between the two
benzyl groups typical of Salan ligands and essential for olefin
polymerization catalysis (Scheme 1).
Upon activation with tris(pentafluorophenyl)borane all com-
plexes led to active polymerization catalysts of neat 1-hexene
at RT, as apparent from the polymerization mixtures becom-
ing viscous within a few minutes. As expected, the zirconium
complex led to the most active catalyst with an activity of 3800 g
mmol-1 h-1; the hafnium and titanium complexes exhibited
somewhat lower activities of ca. 1500 g mmol-1 h-1. While the
lower activity of the hafnium relative to the zirconium catalyst is
in line with previous findings, the activity of the titanium
catalyst is high when taking into account the instability of the
dibenzyl complex LigTiBn2 in solution. Evidently, the active
cationic species is stable enough to sustain a long-lasting
polymerization. GPC analysis revealed that the Zr catalyst led
to a low molecular weight poly(1-hexene) of Mw=14 000 and
PDI=2.0, signifying a nonliving single-site catalyst. The Hf and
Ti catalysts led to high molecular weight polymers of 320 000
(PDI =1.7) and 410 000 (PDI =1.4), respectively. The mole-
cular weight trend is consistent with the reduced termination/
propagation tendencies of these metals.3,12 Pentad analysis
by 13C NMR revealed that the polymer samples were isotactic
to different extents (Figure 1a). The Zr catalyst led to poly-
(1-hexene) of [mmmm] of 33%, which is low, but higher than
that obtained from the corresponding dibromo-Salan-Zr com-
plex (atactic);12 i.e., the increase in the size of halo substituent
from bromo to iodo was sufficient to induce an apparent
isotacticity.15 The Hf and Ti catalysts led to higher isotacticities
of [mmmm] of 50% and 87%, respectively, as expected. Pentad
overlap in the samples of lower isotacticities hampered the
determination of stereochemical control mechanism. However,
the presence of an isolated mrrm pentad as the only impurity
(besides the mmmr and mmrr overlapping the mmmm peak)16 in
the titanium derived highly isotactic polymer gives a strong
support for an enantiomorphic site control mechanism. We
propose that a similar mechanism operates for the zirconium
and hafnium complexes.
The attempted synthesis of the target ligand precursor by
the Mannich condensation13 between N,N0-dimethylethylenedia-
mine, formaldehyde, and 2,4-diiodophenol failed. We therefore
reverted to an alternative reaction14 between this diamine and
the bromomethyl derivative of that phenol (obtained by borohy-
dride reduction of diiodosalicylaldehyde followed by HBr treat-
ment; see the Supporting Information) that gave the ligand
precursor LigH2 in 56% yield. LigH2 reacted with the tetrabenzyl
*Corresponding author. E-mail: moshekol@post.tau.ac.il.
r 2010 American Chemical Society
Published on Web 01/25/2010
pubs.acs.org/Macromolecules

1690 Macromolecules, Vol. 43, No. 4, 2010
Communication
Figure 3. Mn (9) and Mw/Mn (2) versus polymerization time for
polypropylene samples obtained from LigTiBn2 at ambient conditions
(1 bar, 27 °C).
catalyst implied that chain transfer to aluminum is not an
important termination mechanism for this catalyst.
Next, we attempted the polymerization of propylene under
more uniform conditions, aiming to address the activity and
possible living character of the titanium catalyst. The polymer-
ization was done by passing of propylene (1 bar, [C3H6] = 0.66
M)18 at 27 °C in toluene solution with MAO as cocatalyst. Since
LigTiBn2 is not stable in solution, we formed it in situ by mixing
of LigH2 and TiBn4 in toluene under propylene atmosphere for
5 min prior to adding the MAO. Samples were taken at given
intervals until the reaction mixture was observed to be hetero-
geneous (after 50 min). These results are represented in Figure 3.
With Mn of 157 000, a very narrow PDI of 1.12, and activity
of 390 gPP mmolCAT-1 h-1 M-1 after 20 min of polymerization,
this catalyst features a rare combination of high activity and
controlled character.19 A slight broadening of PDI’s after this
period might be due to polymer precipitation. The final molecular
weight after 50 min was Mn=240000 (PDI=1.26). These Mn
and PDI values represent the most controlled polymerization of
propylene by Salan complexes reported to date.3 The controlled
character is unusual considering the polymerization temperature,
the low monomer concentration, the lack of chain-transfer
quencher, and the instability of the precatalyst in solution. The
activity of this catalyst in propylene polymerization is the highest
activity reported for propylene polymerization by Salan com-
plexes.8 Relative to the original Salan-zirconium catalyst,5,8 the
Salan-titanium catalyst reported herein is more than 100 times
faster and leads to polypropylene of much higher molecular
weight while exhibiting a slightly better isoselectivity ([mmmm]
of 82% was found for the polypropylene prepared at 1 bar).20
Several polypropylene samples produced under various con-
ditions (see the Supporting Information) by the titanium cata-
lyst were analyzed by differential scanning calorimetry. They
showed (second heating cycle) melting transitions in the tem-
peratures of 120.9-123.8 °C, which are consistent with the
[mmmm] of 80-83%. Yet, the heat of fusion for these samples
of 64.2-72.0 J g-1 is higher than polypropylene samples of
similar tacticities prepared with metallocene catalysts21 and
possibly indicates a higher degree of crystallinity. The “as
produced” samples (having [rr] of ca. 3.4-4.0%) were also
analyzed by X-ray powder diffraction and were found to feature
very similar crystallinity and include only the R-form of the
isotactic polypropylene.22 We are currently studying the effect of
isothermal crystallization at different temperatures on the
crystal forms of these polymers.
Figure 1. 13C NMR spectra of carbons 3, 2, and 4 of the poly(1-hexene)
samples (a) and methyl region of polypropylene samples (b) produced
by Zr (bottom), Hf (middle), and Ti (top) Salan catalysts.
Figure 2. Polypropylene samples prepared by the Zr-, Hf-, and
Ti-Salan catalysts (left to right).
We found that solid MAO was a suitable cocatalyst for the
polymerization of propylene by these halo-containing Salan
complexes, without the need for a chain transfer quencher.3
Preliminary experiments included the polymerization of neat
propylene by its cryogenic condensation into a reactor containing
the solid dibenzyl complex and MAO, slow thawing, and stirring
at RT for 16 h. While this setup does not allow activity measure-
ments, it is noteworthy that a considerable portion of the
propylene was consumed by all catalysts. With low-to-imperfect
isoselectivities in 1-hexene polymerizations, we expected these
catalysts to exhibit considerably lower isoselectivities in polym-
erization of the smaller monomer. Indeed, the zirconium and
hafnium catalysts led to practically atactic polymers having
[mmmm] of 6% and 10%, respectively. However, in contrast to
the gradual trend observed for the 1-hexene polymerization, the
titanium catalyst led to polypropylene whose isotacticity was
almost the same as that of the poly(1-hexene): [mmmm] of 80%
([rr] = 4.0%, Figure 1b). A clear [mmmr]:[mmrr]:[mrrm] ratio of
2:2:1 indicates an enantiomorphic site control mechanism.
The molecular weights were strongly dependent on the metal
employed: 4000, 37 000, and 1 050 000 for the Zr, Hf, and Ti
catalysts, respectively. The physical attributes of the polymer
samples were consistent with their tacticities and molecular
weights: oils of medium and high viscosity and a white crystalline
solid, respectively (Figure 2). 13C NMR revealed that the polymer
samples were highly regioregular. Chain-end analysis of the low-
Mw polypropylene derived from the Zr catalyst revealed vinyli-
dene and n-propyl groups as major chain ends, supporting a
regular 1,2-insertion prior to a hydride-transfer termination
event. Very low intensity peaks of an isobutyl end group indicate
that chain transfer to aluminum is a very minor termination
process.17 The ultrahigh-Mw polymer obtained with the Ti
In conclusion, we have found that a Salan-titanium complex
leads to a rare combination of activity, controlled character, and
isoselectivity in propylene polymerization at RT, with the com-
mon MAO as cocatalyst. This work demonstrates that a sub-
stantial decrease in isoselectivity on going from higher R-olefin

Communication
Macromolecules, Vol. 43, No. 4, 2010 1691
polymerization to propylene polymerization is not a general rule,
that unstable precatalysts may lead to very well-behaved polym-
erization catalysts, and that titanium may well be the metal of
choice in Salan-based polymerization catalysis. We are currently
investigating the activity of this and related catalysts in copoly-
merizations.
(11) Lamberti, M.; Consolmagno, M.; Mazzeo, M.; Pellecchia, C.
Macromol. Rapid Commun. 2005, 26, 1866.
(12) Segal, S.; Goldberg, I.; Kol, M. Organometallics 2005, 24, 200.
(13) Tshuva, E. Y.; Gendziuk, N.; Kol, M. Tetrahedron Lett. 2001, 42,
6405.
(14) Cohen, A.; Kopilov, J.; Goldberg, I.; Kol, M. Organometallics
2009, 28, 1391.
(15) For a recent report on a highly active, isoselective, and nonliving
[OSSO]-Zr catalyst for 1-hexene polymerization, see: Ishii, A.;
Toda, T.; Nakata, N.; Matsuo, T. J. Am. Chem. Soc. 2009, 131,
13566.
(16) Asakura, T.; Demura, M.; Nishiyama, Y. Macromolecules 1991,
24, 2334.
(17) Resconi, L.; Camurati, I.; Sudmeijer, O. Top. Catal. 1999, 7, 145.
(18) Lin, S.; Tagge, C. D.; Waymouth, R. M.; Nele, M.; Collins, S.;
Pinto, J. C. J. Am. Chem. Soc. 2000, 122, 11275.
Acknowledgment. We thank the Israel Science Foundation
for funding. We thank Dr. Maria Grazia Napoli for GPC
analysis.
Supporting Information Available: Experimental details for
the synthesis of the ligand and complexes, polymerization de-
tails, and polymer characterization. This material is available
free of charge via the Internet at http://pubs.acs.org.
(19) (a) For a catalyst discovered by high throughput screening and
exhibiting a (nonliving) higher activity at 90 °C, and leading to
polypropylene having a melting point of 141 °C, see: Boussie,
T. R.; Diamond, G. M.; Goh, C.; Hall, K. A.; LaPointe, A. M.;
Leclerc, M. K.; Murphy, V.; Shoemaker, J. A. W.; Turner, H.;
Rosen, R. K.; Stevens, J. C.; Alfano, F.; Busico, V.; Cipullo, R.;
Talarico, G. Angew. Chem., Int. Ed. 2006, 45, 3278. (b) For the living
polymerization of 1-hexene by a related catalyst, see: Domski, G. J.;
Lobkovsky, E. B.; Coates, G. W. Macromolecules 2007, 40, 3510.
(20) The highest degree of isoselectivity reported for propylene polym-
erization at 0 °C by bis(phenoxyketimine)titanium complexes was
[mmmm] = 73%. See ref 2b.
(21) (a) De Rosa, C.; Auriemma, F.; Di Capua, A.; Resconi, L.;
Guidotti, S.; Camurati, I.; Nifant’ev, I. E.; Laishevtsev, I. P.
J. Am. Chem. Soc. 2004, 126, 17040. (b) Corradini, P.; Auriemma,
F.; De Rosa, C. Acc. Chem. Res. 2006, 39, 314.
(22) A preliminary study of the crystallization behavior of these poly-
propylene samples suggests that during a slow crystallization at a
higher temperature the γ form develops as well. For example, for a
polypropylene sample (having an [rr] of 4%) crystallized at 75 °C
for 13 h, the relative amount of the γ form with respect to the R form
(evaluated from the intensity of the (117)γ and (130)R reflections)21a
was found to be 46%. See the Supporting Information.
References and Notes
(1) Domski, G. J.; Rose, J. M.; Coates, G. W.; Bolig, A. D.; Brookhart,
M. Prog. Polym. Sci. 2007, 32, 30.
(2) (a) Mason, A. F.; Coates, G. W. J. Am. Chem. Soc. 2004, 126,
16326. (b) Edson, J. B.; Wang, Z.; Kramer, E. J.; Coates, G. W. J. Am.
Chem. Soc. 2008, 130, 4968.
(3) Cipullo., R.; Busico, V.; Fraldi, N.; Pellecchia, R.; Talarico, G.
Macromolecules 2009, 42, 3869.
(4) Half-metallocenes are regarded as non-metallocenes in this work.
(5) Tshuva, E. Y.; Goldberg, I.; Kol, M. J. Am. Chem. Soc. 2000, 122,
10706.
(6) Jayaratne, K. C.; Sita, L. R. J. Am. Chem. Soc. 2000, 122, 958.
(7) Annunziata, L.; Pappalardo, D.; Tedesco, C.; Pellecchia, C.
Macromolecules 2009, 42, 5572.
(8) Busico, V.; Cipullo, R.; Pellecchia, R.; Ronca, S.; Roviello, G.;
Talarico, G. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 15321.
(9) Harney, M. B.; Zhang, Y.; Sita, L. R. Angew. Chem., Int. Ed. 2006,
45, 2400.
(10) Busico, V.; Carbonniere, P.; Cipullo, R.; Pellecchia, R.; Severn,
J. R.; Talarico, G. Macromol. Rapid Commun. 2007, 28, 1128.