Mechanism of the alternating copolymerization of epoxides and CO 2 using β-diiminate zinc catalysts: Evidence for a bimetallic epoxide enchainment

Article abstract of DOI:10.1021/ja030085e

A series of zinc β-diiminate (BDI) complexes and their solid-state structures, solution dynamics, and copolymerization behavior with CO ₂ and cyclohexene oxide (CHO) are reported. Stoichiometric reactions of the copolymerization initiation step

Full text of DOI:10.1021/ja030085e

Published on Web 09/03/2003
Mechanism of the Alternating Copolymerization of Epoxides
and CO2 Using â-Diiminate Zinc Catalysts: Evidence for a
Bimetallic Epoxide Enchainment
David R. Moore, Ming Cheng, Emil B. Lobkovsky, and Geoffrey W. Coates*
Contribution from the Department of Chemistry and Chemical Biology, Baker Laboratory,
Cornell UniVersity, Ithaca, New York 14853-1301
Received February 5, 2003; E-mail: gc39@cornell.edu
Abstract: A series of zinc â-diiminate (BDI) complexes and their solid-state structures, solution dynamics,
and copolymerization behavior with CO2 and cyclohexene oxide (CHO) are reported. Stoichiometric reactions
of the copolymerization initiation steps show that zinc alkoxide and bis(trimethylsilyl)amido complexes insert
CO2, whereas zinc acetates react with CHO. [(BDI-2)ZnOMe]2 [(BDI-2) ) 2-((2,6-diethylphenyl)amido)-4-
((2,6-diethylphenyl)imino)-2-pentene] and (BDI-1)ZnOiPr [(BDI-1) ) 2-((2,6-diisopropylphenyl)amido)-4-
((2,6-diisopropylphenyl)imino)-2-pentene] react with CO2 to form [(BDI-2)Zn(µ-OMe)(µ,η2-O2COMe)Zn(BDI-
2)] and [(BDI-1)Zn(µ,η2-O2COiPr)]2, respectively. (BDI-2)ZnN(SiMe3)2 inserts CO2 and eliminates trimethylsilyl
isocyanate to give [(BDI-2)Zn(µ-OSiMe3)]2. [(BDI-7)Zn(µ-OAc)]2 [(BDI-7) ) 3-cyano-2-((2,6-diethylphenyl)-
amido)-4-((2,6-diethylphenyl)imino)-2-pentene] reacts with 1.0 equiv of CHO to yield [(BDI-7)Zn(µ,η2-OAc)-
(µ,η1-OCyOAc)Zn(BDI-7)]. Under typical polymerization conditions, rate studies on the copolymerization
exhibit no dependence in [CO2], a first-order dependence in [CHO], and orders in [Zn]tot ranging from 1.0
to 1.8 for [(BDI)ZnOAc] complexes. The copolymerizations of CHO (1.98 M in toluene) and 300 psi CO2 at
50 °C using [(BDI-1)ZnOAc] and [(BDI-2)ZnOAc] show orders in [Zn]tot of 1.73 ( 0.06 and 1.02 ( 0.03,
respectively. We propose that two zinc complexes are involved in the transition state of the epoxide ring-
opening event.
tinue to grow.8 A wide variety of recently reported catalytic
systems8-49 convert CO2 and epoxides into polycarbonates and/
or cyclic carbonates. Inoue’s discovery of a heterogeneous
Introduction
At present, the predominant source of carbon for raw
materials used in the chemical industry is fossil fuels; synthetic
plastics account for approximately 7% of worldwide oil and
gas consumption.1 In addition, the current worldwide production
of plastics is approximately 150 million tons per year.1 Because
these resources are expected to be depleted in about 80 years,2
there is significant interest in finding new routes from biore-
newable resources to important chemicals, especially biodegrad-
able and recyclable polymers.3,4
(6) For some other leading references, see: (a) Musie, G.; Wei, M.; Subra-
maniam, B.; Busch, D. H. Coord. Chem. ReV. 2001, 219, 789-820. (b)
Cooper, A. I. J. Mater. Chem. 2000, 10, 207-234. (c) Leitner, W. Comptes
Rendus Acad. Sci. Ser. II C 2000, 3, 595-600. (d) Bolm, C.; Beckmann,
O.; Dabard, O. A. G. Angew. Chem., Int. Ed. 1999, 38, 907-909. (e) Yin,
X. L.; Moss, J. R. Coord. Chem. ReV. 1999, 181, 27-59. (f) Leitner, W.
Coord. Chem. ReV. 1996, 153, 257-284. (g) Leitner, W. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 2207-2221. (h) Jessop, P. G.; Ikariya, T.; Noyori,
R. Chem. ReV. 1995, 95, 259-272.
(7) Gibson, D. H. Chem. ReV. 1996, 96, 2063-2095.
A current focus of our research is the development of catalysts
for the synthesis of polymers from carbon dioxide. Although
CO2 is an ideal synthetic feedstock because it is abundant,
inexpensive, nontoxic, and nonflammable, synthetic chemists
have had limited success in developing efficient catalytic
processes that exploit CO2 as a raw material.5-7 Consequently,
research efforts in the alternating copolymerization of CO2 with
epoxides to give biodegradable aliphatic polycarbonates4 con-
(8) For reviews on epoxide/CO2 copolymerizations, see: (a) Darensbourg, D.
J.; Holtcamp, M. W. Coord. Chem. ReV. 1996, 153, 155-174. (b) Super,
M. S.; Beckman, E. J. Trends Polym. Sci. 1997, 5, 236-240. (c) Rokicki,
A.; Kuran, W. J. Macromol. Sci.-ReV. Macromol. Chem. Phys. 1981, C21,
135-186.
(9) Darensbourg, D. J.; Holtcamp, M. W. Macromolecules 1995, 28, 7577-
7579.
(10) Darensbourg, D. J.; Rainey, P.; Yarbrough, J. C. Inorg. Chem. 2001, 40,
986-993.
(11) Darensbourg, D. J.; Holtcamp, M. W.; Struck, G. E.; Zimmer, M. S.;
Niezgoda, S. A.; Rainey, P.; Robertson, J. B.; Draper, J. D.; Reibenspies,
J. H. J. Am. Chem. Soc. 1999, 121, 107-116.
(12) Darensbourg, D. J.; Wildeson, J. R.; Yarbrough, J. C.; Reibenspies, J. H.
J. Am. Chem. Soc. 2000, 122, 12487-12496.
(1) Stevens, E. S. Green Plastics; Princeton University Press: Princeton, 2002.
(2) Gerngross, T. U.; Slater, S. C. Sci. Am. 2000, 283 (2), 36-41.
(3) Gross, R. A.; Kalra, B. Science 2002, 297, 803-807.
(13) Darensbourg, D. J.; Wildeson, J. R.; Yarbrough, J. C. Inorg. Chem. 2002,
41, 973-980.
(4) Okada, M. Prog. Polym. Sci. 2002, 27, 87-133.
(14) Super, M. S.; Berluche, E.; Costello, C.; Beckman, E. J. Macromolecules
1997, 30, 368-372.
(5) Arakawa, H.; Aresta, M.; Armor, J. N.; Barteau, M. A.; Beckman, E. J.;
Bell, A. T.; Bercaw, J. E.; Creutz, C.; Dinjus, E.; Dixon, D. A.; Domen,
K.; DuBois, D. L.; Eckert, J.; Fujita, E.; Gibson, D. H.; Goddard, W. A.;
Goodman, D. W.; Keller, J.; Kubas, G. J.; Kung, H. H.; Lyons, J. E.;
Manzer, L. E.; Marks, T. J.; Morokuma, K.; Nicholas, K. M.; Periana, R.;
Que, L.; Rostrup-Nielson, J.; Sachtler, W. M. H.; Schmidt, L. D.; Sen, A.;
Somorjai, G. A.; Stair, P. C.; Stults, B. R.; Tumas, W. Chem. ReV. 2001,
101, 953-996.
(15) Sarbu, T.; Beckman, E. J. Macromolecules 1999, 32, 6904-6912.
(16) Sarbu, T.; Styranec, T.; Beckman, E. J. Nature 2000, 405, 165-168.
(17) Darensbourg, D. J.; Lewis, S. J.; Rodgers, J. L.; Yarbrough, J. C. Inorg.
Chem. 2003, 43, 581-589.
(18) Darensbourg, D. J.; Yarbrough, J. C. J. Am. Chem. Soc. 2002, 124, 6335-
6342.
(19) Kruper, W. J.; Dellar, D. V. J. Org. Chem. 1995, 60, 725-727.
9
10.1021/ja030085e CCC: $25.00 © 2003 American Chemical Society
J. AM. CHEM. SOC. 2003, 125, 11911-11924
11911

A R T I C L E S
Moore et al.
ZnEt2/H2O mixture for copolymerization of CO2 and propylene
oxide50 inspired the development of a number of additional
heterogeneous systems.36-45 Discrete zinc phenoxides developed
by Darensbourg9-12 and ZnO/fluorinated carboxylic acid sys-
tems investigated by Beckman14,15 copolymerize cyclohexene
oxide (CHO) and CO2, exhibiting turnover frequencies (TOF)
of approximately 10 h-1. Chromium-based systems, including
chromium porphyrins studied by Kruper19 and Holmes20,21 and
chromium salen complexes reported by Darensbourg,18 give
TOFs of up to 200 h-1. Finally, highly active â-diiminate (BDI)
zinc carboxylates and alkoxides designed in our group afford
TOFs as high as 2300 h-1 for CHO and CO2 copolymerization
(Figure 1).51-56
Figure 1. Highly active [(BDI)Zn(µ-OMe)]2 complexes for the alternating
copolymerization of CHO and CO2.
Scheme 1. â-Diiminate Zinc Complexes as Catalysts for the
Alternating Copolymerization of Cyclohexene Oxide and CO2
Over the past five years, we have published several reports
describing well-defined, single-site BDI zinc complexes as
highly active catalysts for the synthesis of biodegradable
(20) Mang, S.; Cooper, A. I.; Colclough, M. E.; Chauhan, N.; Holmes, A. B.
Macromolecules 2000, 33, 303-308.
(21) Stamp, L. M.; Mang, S. A.; Holmes, A. B.; Knights, K. A.; de Miguel, Y.
R.; McConvey, I. F. Chem. Commun. 2001, 2502-2503.
(22) Eberhardt, R.; Allmendinger, M.; Luinstra, G. A.; Rieger, B. Organome-
tallics 2003, 22, 211-214.
Scheme 2. Catalytic Cycle Illustrating the Alternating Insertion of
CHO and CO2 in the Synthesis of Aliphatic Polycarbonates
(23) Eberhardt, R.; Allmendinger, M.; Rieger, B. Macromol. Rapid Commun.
2003, 24, 194-196.
(24) Kuran, W.; Listos, T. Macromol. Chem. Phys. 1994, 195, 977-984.
(25) Dinger, M. B.; Scott, M. J. Inorg. Chem. 2001, 40, 1029-1036.
(26) Nozaki, K.; Nakano, K.; Hiyama, T. J. Am. Chem. Soc. 1999, 121, 11008-
11009.
(27) Nakano, K.; Nozaki, K.; Hiyama, T. J. Am. Chem. Soc. 2003, 125, 5501-
5510.
(28) Kuran, W.; Listos, T.; Abramczyk, M.; Dawidek, A. J. Macromol. Sci.,
Pure Appl. Chem. 1998, A35, 427-437.
(29) Kim, H. S.; Kim, J. J.; Lee, B. G.; Jung, O. S.; Jang, H. G.; Kang, S. O.
Angew. Chem., Int. Ed. 2000, 39, 4096-4098.
(30) Kim, H. S.; Kim, J. J.; Lee, S. D.; Lah, M. S.; Moon, D.; Jang, H. G.
Chem. Eur. J. 2003, 9, 678-686.
(31) Paddock, R. L.; Nguyen, S. T. J. Am. Chem. Soc. 2001, 123, 11498-11499.
(32) Shen, Y. M.; Duan, W. L.; Shi, M. J. Org. Chem. 2003, 68, 1559-1562.
(33) Yano, T.; Matsui, H.; Koike, T.; Ishiguro, H.; Fujihara, H.; Yoshihara,
M.; Maeshima, T. Chem. Commun. 1997, 1129-1130.
(34) Yamaguchi, K.; Ebitani, K.; Yoshida, T.; Yoshida, H.; Kaneda, K. J. Am.
Chem. Soc. 1999, 121, 4526-4527.
(35) Kawanami, H.; Ikushima, Y. Chem. Commun. 2000, 2089-2090.
(36) Jung, J. H.; Ree, M.; Chang, T. J. Polym. Sci., Part A: Polym. Chem.
1999, 37, 3329-3336.
polymers. These include the production of polycarbonates from
the alternating copolymerization of carbon dioxide and epoxides
(including cyclohexene oxide and propylene oxide) under low
temperatures and mild pressures (Schemes 1 and 2),51-56
heterotactic and syndiotactic poly(lactic acid) from rac-lactide
and meso-lactide, respectively,57,58 and poly(3-hydroxybutyrate)
from â-butyrolactone.59 Using these systems, narrow polydis-
persity indices (PDIs) and number average molecular weights
(Mn) dictated by monomer-to-initiator ratios are observed,
indicative of living polymerizations.60 Additionally, minor
changes in the ligand architecture result in dramatic enhance-
ments in polymerization rates.54,55
Homogeneous catalysts have distinct advantages over their
heterogeneous counterparts.61 Single-site homogeneous catalysts
are typically of the form LnMR, containing a readily modifiable
organic ligand set (Ln), a catalytically active metal center (M),
and a viable initiating group or polymer chain-end (R). These
homogeneous catalysts allow for modifications of (1) ligand
(37) Chen, L. B. Makromol. Chem. Macromol. Symp. 1992, 59, 75-82.
(38) Ree, M.; Bae, J. Y.; Jung, J. H.; Shin, T. J.; Hwang, Y. T.; Chang, T.
Polym. Eng. Sci. 2000, 40, 1542-1552.
(39) Ree, M.; Bae, J. Y.; Jung, J. H.; Shin, T. J. Korea Polym. J. 1999, 7, 333-
349.
(40) Ree, M.; Bae, J. Y.; Jung, J. H.; Shin, T. J. J. Polym. Sci., Part A: Polym.
Chem. 1999, 37, 1863-1876.
(41) Shen, Z. Q.; Chen, X. H.; Zhang, Y. F. Macromol. Chem. Phys. 1994,
195, 2003-2011.
(42) Tan, C. S.; Hsu, T. J. Macromolecules 1997, 30, 3147-3150.
(43) Yang, S. Y.; Fang, X. G.; Chen, L. B. Polym. AdV. Technol. 1996, 7, 605-
608.
(44) Liu, B. Y.; Zhao, X. J.; Wang, X. H.; Wang, F. S. J. Polym. Sci., Part A:
Polym. Chem. 2001, 39, 2751-2754.
(45) Hsu, T. J.; Tan, C. S. Polymer 2001, 42, 5143-5150.
(46) Zhang, M.; Chen, L. B.; Liu, B.; Yan, Z.; Qin, G.; Li, Z. Polym. Bull.
2001, 47, 255-260.
(47) Zhang, M.; Chen, L. B.; Qin, G.; Li, Z. M. Acta Polym. Sin. 2001, 422-
424.
(48) Zhang, M.; Chen, L. B.; Qin, G.; Liu, B.; Yan, Z.; Li, Z. J. Appl. Polym.
Sci. 2003, 87, 1123-1128.
(49) Wang, S. J.; Du, L. C.; Zhao, X. S.; Meng, Y. Z.; Tjong, S. C. J. Appl.
Polym. Sci. 2002, 85, 2327-2334.
(50) Inoue, S.; Koinuma, H.; Tsuruta, T. J. Polym. Sci., Part B: Polym. Lett.
1969, 7, 287-292.
(51) Cheng, M.; Lobkovsky, E. B.; Coates, G. W. J. Am. Chem. Soc. 1998,
120, 11018-11019.
(52) Coates, G. W.; Cheng, M. (Cornell Research Foundation), U.S. Pat.
6133402, 2002 (Chem Abstr. 2000, 132, 152332).
(57) Cheng, M.; Attygalle, A. B.; Lobkovsky, E. B.; Coates, G. W. J. Am. Chem.
Soc. 1999, 121, 11583-11584.
(53) Cheng, M.; Darling, N. A.; Lobkovsky, E. B.; Coates, G. W. Chem.
Commun. 2000, 2007-2008.
(58) Chamberlain, B. M.; Cheng, M.; Moore, D. R.; Ovitt, T. M.; Lobkovsky,
E. B.; Coates, G. W. J. Am. Chem. Soc. 2001, 123, 3229-3238.
(59) Rieth, L. R.; Moore, D. R.; Lobkovsky, E. B.; Coates, G. W. J. Am. Chem.
Soc. 2002, 124, 15239-15248.
(54) Cheng, M.; Moore, D. R.; Reczek, J. J.; Chamberlain, B. M.; Lobkovsky,
E. B.; Coates, G. W. J. Am. Chem. Soc. 2001, 123, 8738-8749.
(55) Moore, D. R.; Cheng, M.; Lobkovsky, E. B.; Coates, G. W. Angew. Chem.,
Int. Ed. 2002, 41, 2599-2602.
(60) Matyjaszewski, K.; Mueller, A. H. E. Polym. Prepr. (American Chemical
Society, DiVision of Polymer Chemistry) 1997, 38 (1), 6-9.
(61) Coates, G. W. Chem. ReV. 2000, 100, 1223-1252.
(56) Allen, S. D.; Moore, D. R.; Lobkovsky, E. B.; Coates, G. W. J. Am. Chem.
Soc. 2002, 124, 14284-14285.
9
11912 J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003

Epoxide−CO Copolymerization Using Zn â-Diiminate
A R T I C L E S
2
syn-syn and syn-anti geometry.62 There are close interactions
between Zn(2) and O(1) (2.497 Å) as well as between Zn(1)
and O(4) (3.049 Å). The molecular structures of [(BDI-4)Zn-
(µ,η2-OAc)]2 and [(BDI-6)Zn(µ,η2-OAc)]2 are dimers isostruc-
tural to [(BDI-2)Zn(µ,η2-OAc)]2 and [(BDI-1)Zn(µ,η2-OAc)]2,
respectively.62 [(BDI-5)Zn(µ,η2-OAc)]2 crystallizes in an anti
fashion (Figure 4). The bridging acetates in [(BDI-5)Zn(µ,η2-
OAc)]2 are between syn-syn and syn-anti geometry, similar to
[(BDI-2)Zn(µ,η2-OAc)]2, Additionally, the Zn‚‚Zn separation
is 4.14 Å, which is intermediate between the 4.24 and 3.94 Å
Zn‚‚Zn separations of [(BDI-1)Zn(µ,η2-OAc)]2 and [(BDI-2)-
Zn(µ,η2-OAc)]2, respectively. [(BDI-7)Zn(µ,η2-OAc)]2 also
crystallizes as a dimer (Figure 5) with a Zn‚‚Zn separation of
3.85 Å, about 0.1 Å shorter than the Zn‚‚Zn separation for
[(BDI-2)Zn(µ,η2-OAc)]2. The bridging acetates are between syn-
syn and anti-syn geometries, comparable to [(BDI-2)Zn(µ,η2-
OAc)]2. Crystal data and structure refinements are provided in
Table 1.
Figure 2. â-Diiminate zinc acetate complexes.
electronics and sterics, (2) labile metal centers, and (3) viable
initiating groups. Mechanistic information can be retrieved from
a system designed with a single-site catalyst, whereas the
presence of multiple active sites makes obtaining this informa-
tion difficult. Therefore, it has been our goal to develop single-
site BDI zinc complexes, which are specifically designed to
possess a permanent ligand on the active metal center, as well
as an initiating group, which serves to mimic the putative
propagating species of the polymer chain.
Our recent findings suggested that the enhanced rates of the
unsymmetrical isopropyl and ethyl/methyl-substituted catalysts55
(see Figure 1) could be informative regarding the mechanism
in the copolymerization of CHO and CO2 (see Scheme 2). The
symmetrically substituted complexes [(BDI-1)ZnOAc] and
[(BDI-2)ZnOAc] (Figure 2) are effective catalysts for the
alternating copolymerization of CHO and CO2. However,
complexes that do not possess adequate steric bulk in the ortho
positions of the N-aryl ring exist as highly unreactive dimeric
species (Figure 3). Conversely, bulky substituents yield pre-
dominantly monomeric complexes in solution that also impede
catalysis. Therefore, we synthesized and investigated a variety
of zinc complexes in the regime of activity between unreactive
tightly bound dimers and monomers. Herein we report solution
studies, stoichiometric enchainment events, copolymerization
data, and rate studies for the alternating copolymerization of
CO2 and CHO. We propose a detailed mechanism, in which
the rate-determining step is a bimetallic epoxide enchainment
event.
Monomer-Dimer Equilibria of Zinc Acetate and Alkox-
ide Complexes. [(BDI-1)ZnOAc] and related zinc complexes
are single-site catalysts and therefore provide an excellent
opportunity to model the dynamic behavior of related catalytic
intermediates in solution. The 1H NMR spectrum of [(BDI-1)-
ZnOAc] in toluene-d8 ([Zn]tot ) 1.0 × 10-2 M) exhibits two
sets of resonances, the relative intensities of which vary with
concentration. The set that becomes more intense as concentra-
tion decreases was assigned to the monomeric species.62 The
monomer predominates in solution at typical copolymerization
conditions (50 °C, [Zn]tot ) 1.0 × 10-2 M). Coordinating
solvents such as tetrahydrofuran (THF) do not significantly
affect the equilibrium, suggesting that the acetate of the
monomer exists in the η2 form. In contrast to [(BDI-1)ZnOAc],
the 1H NMR spectrum of [(BDI-2)ZnOAc] exhibits exclusively
one set of resonances in toluene-d8 ([Zn]tot ) 1.0 × 10-2 M)
from 20 to 100 °C, consistent with only a dimer in solution.
Methine resonances due to dimeric and monomeric species are
observed from 4.6 to 4.8 ppm and 4.9 to 5.0 ppm, respectively.
[(BDI-3)ZnOAc], [(BDI-4)ZnOAc], [(BDI-5)ZnOAc], and [(BDI-
7)ZnOAc] each show one set of peaks in benzene-d6, including
a single methine backbone resonance for [(BDI-3)ZnOAc],
[(BDI-4)ZnOAc], and [(BDI-5)ZnOAc] between 4.6 and 4.8
ppm, consistent with the dimeric complex. However, [(BDI-
4)ZnOAc], [(BDI-5)ZnOAc], and [(BDI-7)ZnOAc] are fluxional
Results and Discussion
Synthesis and Solid-State Structures of Zinc Catalysts.
Previous results showed that subtle ligand modifications produce
dramatic effects on catalytic activity.54,55 For instance, com-
plexes with inadequate steric bulk at the four N-aryl ortho
positions, such as methyl substituents ([(BDI-3)ZnOAc]; Figure
2), fail to afford polymerization activity. However, complexes
containing different N-aryl substituents (e.g., 2,6-diisopropyl-
phenyl/2,6-dimethylphenyl) and an electron-withdrawing cyano
substituent on the ligand generate the most active catalysts for
the alternating copolymerization of CHO and CO2 (see Figure
1).55 To probe the highly sensitive structure-activity relation-
ships in these systems, we synthesized zinc acetate complexes
of varying bulk in the N-aryl positions (see Figure 2).
Reaction of the â-diimine ligands with ZnEt2 affords the
monomeric (BDI)ZnEt complexes,54 which can be cleanly
converted to the zinc acetate analogues in up to 87% yield by
protonating the zinc ethyl moiety with 1.0 equiv of acetic acid
at 0 °C.56 As with [(BDI)ZnOAc] complexes ligated by 1-2,51,54
complexes ligated by 3-7 are dimeric in the solid state. X-ray
analysis of [(BDI-3)Zn(µ-OAc)]2 shows a dimer with four-
coordinate zinc centers containing bridging η2-acetates between
1
dimers on the H NMR time scale, characterized by broader
resonances at 20 °C which become more well defined at
1
60 °C. Similar to that of [(BDI-1)ZnOAc], H NMR spectros-
copy of [(BDI-6)ZnOAc] reveals a monomer/dimer equilibrium
at room temperature ([Zn]tot ) 2.1 × 10-2 M in toluene-d8),
which becomes predominantly monomer at 80 °C.62
The monomer/dimer equilibrium of [(BDI-1)ZnOAc]2, (2
[(BDI-1)ZnOAc] h [(BDI-1)ZnOAc]2), was studied between
1
-80 and 100 °C using H NMR spectroscopy (400 MHz) in
toluene-d8 ([Zn]tot ) 2.2 × 10-2 M; 20 °C, Keq ) 207 M-1
;
a
50 °C, Keq ) 35 M-1). From the plot of ln Keq versus 1/T,62
∆H and ∆S of -10.8 ( 0.1 kcal/mol and -26.3 ( 0.3 eu are
found from the slope and intercept. For this solution, the
complex is predominantly dimer between -80 and -20 °C;
(62) See Supporting Information.
9
J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003 11913

A R T I C L E S
Moore et al.
Figure 3. Tightly bound dimers and monomers show poor catalytic activity for the alternating copolymerization of CHO and CO2. Loosely bound dimers
are highly reactive catalysts (R ) alkyl).
Figure 5. ORTEP drawing of [(BDI-7)Zn(µ,η2-OAc)]2 (non-hydrogen
atoms) with thermal ellipsoids drawn at the 40% probability level. Selected
Figure 4. ORTEP drawing of [(BDI-5)Zn(µ,η2-OAc)]2 (non-hydrogen
bond lengths (Å) and bond angles (deg): Zn(1)-N(1) 1.979(2), Zn(1)-
atoms) with thermal ellipsoids drawn at the 40% probability level. Selected
bond lengths (Å) and bond angles (deg): Zn(1)-N(1) 1.971(2), Zn(1)-
N(2) 1.970 (2), Zn(1)-O(1) 1.962 (1), Zn(1)-O(2A) 1.974(2), O(1)-C(1)
1.252(3), O(2)-C(1) 1.264(3), O(1)-Zn(1)-O(2A) 107.07(7), N(1)-Zn-
N(2) 1.962(2), Zn(1)-O(1) 1.959(2), Zn(1)-O(2A) 1.947(2), O(1)-C(1)
(1)-N(2) 97.00(6), N(1)-Zn(1)-O(1) 111.79(6), N(2)-Zn(1)-O(2A)
1.262(2), O(2)-C(1) 1.233(4), O(1)-Zn(1)-O(2A) 111.63(9), N(1)-Zn-
104.65(7), C(1)-O(1)-Zn(1) 111.9(1), O(1)-C(1)-O(2) 122.5(2).
(1)-N(2) 98.7(1), N(1)-Zn(1)-O(2A) 108.09(9), N(2)-Zn(1)-O(1) 107.45-
(8), C(1)-O(1)-Zn(1) 137.7(2), O(1)-C(1)-O(2) 127.1(2).
complexes exist at least partially as monomeric species in
benzene solution, we explored combined solutions of these
compounds to access the relative stability of dimers of acetate
(eight-membered ring), methoxide (four-membered ring), and
mixed (six-membered ring) species. The relative energies of
from approximately -20 to 80 °C the complex is in both dimer
and monomer forms; and at 100 °C only the monomer is present.
Upon cooling from -40 to -80 °C, a broad shift at 0.8 ppm
that was assigned to the internal methyl groups of the isopropyl
substituents on the ligands coalesces at approximately -60 °C
and reemerges at -80 °C. We believe this is due to site
such species provide useful insight into potential catalytic ground
states during CO2/epoxide polymerizations. Equimolar amounts
of [(BDI-2)Zn(µ-OMe)]2 and [(BDI-2)Zn(µ,η2-OAc)]2 were
exchange between a C2h symmetric dimer species through a D2h
dissolved in toluene at room temperature ([Zn]tot ) 1.2 × 10-1
symmetric dimer intermediate.62
Our previous studies on BDI zinc alkoxides show that (BDI-
1)ZnOiPr is predominantly monomeric in benzene-d6 at room
M). After 1 h of stirring, 1H NMR spectroscopy revealed a new
compound in quantitative yield (Scheme 3). Following crystal-
lization, X-ray diffraction revealed the molecular structure to
temperature but exhibits a monomer/dimer equilibrium at lower
be a mixed dimer of the form [(BDI-2)Zn(µ-OMe)(µ,η2-OAc)-
temperatures.59 Chisholm observed (BDI-1)ZnOtBu and (BDI-
Zn(BDI-2)] (Figure 6; Table 1), suggesting that the six-
1)ZnOSiPh3‚THF to be monomeric in both the solid state and
membered ring mixed dimer exhibits greater stability than do
either of the homodimers. The acetate bridge exhibits a syn-
syn geometry, with a Zn‚‚Zn separation of 3.38 Å.
in solution, and (BDI-1)ZnOSiPh3‚THF shows a temperature-
dependent reversible dissociation of THF.63,64 We have found
that [(BDI-2)Zn(µ-OiPr)]2 and [(BDI-2)Zn(µ-OMe)]2 are both
dimeric at room temperature ([Zn]tot ) 5.6 × 10-2 M), but they
Therefore, as the N-aryl ring substituents on the zinc acetate
complexes become more sterically congested, two general trends
are observed: (1) the Zn‚‚Zn separations in the solid state
increase and (2) the aggregatory state in solution changes from
a tightly bound dimer to a fluxional, loosely bound dimer, and
finally to a monomer/dimer equilibrium (see Table 2; Figure
3).
participate in a slow exchange reaction to produce a mixed dimer
over the course of days.59 Because both methoxide and acetate
(63) Chisholm, M. H.; Gallucci, J.; Phomphrai, K. Inorg. Chem. 2002, 41, 2785-
2794.
(64) Chisholm, M. H.; Huffman, J. C.; Phomphrai, K. J. Chem. Soc., Dalton
Trans. 2001, 222-224.
9
11914 J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003

Epoxide−CO Copolymerization Using Zn â-Diiminate
A R T I C L E S
2
Table 1. Crystal Data and Structure Refinement Parameters
[(BDI-5)-
(µ,η2-OAc)]2
[(BDI-7)Zn-
(µ,η2-OAc)]2
[(BDI-2)Zn(µ-OMe)-
[(BDI-2)Zn(µ-OMe)-
[(BDI-1)Zn-
[(BDI-2)Zn-
(µ-OSiMe3)]2
[(BDI-7)Zn(µ,η2OAc)-
(µ,η1-OCyOAc)Zn[(BDI-7)]
i
(µ,η2-OAc)Zn(BDI-2)] (µ,η2-O COMe)Zn(BDI-2)]
(µ,η2-O COPr)]2
2
2
emp form.
C29H40N2O2Zn C28H35N3O2Zn‚ C53H72N4O3Zn2 C53H72N4O4Zn2‚
C66H96N4O6Zn2‚ C28H42N2OSiZn‚ C124H158N12O10Zn4‚
0.5C7H8
C7H8
C6H6
0.5C7H8
C6H6‚0.5C6H12‚
0.5C4H8O
form. wt
temp [K]
λ [Å]
cryst system
space group
a (Å)
b (Å)
c (Å)
R (deg)
â (deg)
γ (deg)
V [Å3]
514.00
173(2)
0.71073
triclinic
P1h
11.7387(4)
11.8892(4)
12.2015(5)
115.718(2)
105.019(2)
102.599(3)
1370.70(9)
2
557.03
173(2)
0.71073
triclinic
P1h
10.0930(5)
12.9843(7)
13.4344(7)
62.580(1)
74.139(1)
69.807(1)
1452.85(13)
2
943.89
173(2)
0.71073
monoclinic
P21/n
13.375(1)
16.907(1)
22.331(2)
90
95.531(2)
90
5026.2(6)
4
1052.02
173(2)
0.71073
monoclinic
P21/c
13.5383(2)
16.8540(3)
24.8141(1)
90
93.154(1)
90
5653.4(1)
4
1250.32
173(2)
0.71073
monoclinic
C2
17.70(3)
19.44(3)
11.73(2)
90
121.66(3)
90
3435(8)
2
562.16
2394.34
173(2)
0.71073
triclinic
P1h
16.912(4)
17.554(4)
24.490(6)
101.955(8)
101.881(6)
90.558(6)
6950(3)
2
173(2)
0.71073
orthorhombic
Aba2
19.8633(8)
17.2928(8)
17.5338(8)
90
90
90
6022.7(5)
8
1.240
Z
F
calc [Mg/m3]
1.245
0.923
1.273
0.871
1.247
0.999
1.236
0.896
1.209
0.750
1.144
0.739
µ (Mo KR)
[mm-1
F(000)
0.881
]
548
582
2008
2240
1340
2376
2518
cryst size [mm]
0.40 × 0.30
0.30 × 0.15
0.40 × 0.20
0.50 × 0.30
0.30 × 0.20
0.20 × 0.15
0.30 × 0.20
× 0.05
× 0.10
× 0.20
× 0.05
× 0.10
× 0.10
× 0.05
θ range (deg)
index ranges
2.76 to 28.28
2.17 to 33.14
1.51 to 28.37
1.46 to 23.26
2.04 to 18.83
2.32 to 26.37
1.31 to 23.26
-15<)h<)15 -14<)h<)14 -16<)h<)16 -14<)h<)15
-15<)k<)14 -19<)k<)18 -22<)k<)22 -14<)k<)18
-16<)l<)15 -18<)l<)19 -24<)l<)29 -27<)l<)27
10033
5409
-16<)h<)15 -24<)h<)23 -18<)h<)18
-17<)k<)17 -21<)k<)18 -19<)k<)19
-10<)l<)10 -21<)l<)21 -26<)l<)27
7362
2696
data collected
unique data
Rint
19971
9206
0.0302
1.058
31651
11542
0.0498
0.867
24119
8081
0.0792
1.005
18561
6122
0.0465
1.046
34202
19135
0.0476
1.011
0.0272
1.041
0.0606
1.033
GOF
final R indices
[I > 2σ(I)]
R1 ) 0.0458
R1 ) 0.0447
R1 ) 0.0400
R1 ) 0.0527
wR2 ) 0.1121
R1 ) 0.0918
wR2 ) 0.1299
0.564 and
-0.533
R1 ) 0.0607
wR2 ) 0.1321
R1 ) 0.0938
wR2 ) 0.1606
1.357 and
-0.445
R1 ) 0.0409
wR2 ) 0.0903
R1 ) 0.0651
wR2 ) 0.1007
0.428 and
-0.305
R1 ) 0.0698
wR2 ) 0.2048
R1 ) 0.1308
wR2 ) 0.2427
0.951 and
-0.341
wR2 ) 0.1214 wR2 ) 0.1102 wR2 ) 0.0918
R indices (all data) R1 ) 0.0585
R1 ) 0.0618
R1 ) 0.0846
[I > 2σ(I)]
min - Fmax
[e‚Å-3
wR2 ) 0.1279 wR2 ) 0.1179 wR2 ) 0.1115
1.282 and
-0.876
F
0.919 and
-0.659
0.446 and
-0.354
]
Table 2. Copolymerization of CO2 and Cyclohexene Oxide Using â-Diiminate (BDI) Zinc Complexesa
Zn‚‚Zn
separation (Å)b
solution
structure
reaction
time (h)
TOF
(h-1 e
% carbonate
linkages
M (×10-3
)
n
c
d
f
complex
TON
)
(GPC)g
M /M
w
n
(BDI-1)ZnN(SiMe3)2
(BDI-2)ZnN(SiMe3)2
(BDI-1)ZnN(SiMe3)2 +
(BDI-2)ZnN(SiMe3)2
[(BDI-1)ZnOAc]2
[(BDI-2)ZnOAc]2
[(BDI-3)ZnOAc]2
[(BDI-4)ZnOAc]2
[(BDI-5)ZnOAc]2
[(BDI-6)ZnOAc]2
[(BDI-7)ZnOAc]2
NAh
NAh
NAh
M
M
M
0.5
0.5
0.5
172
179
329
345
358
658
94
97
94
25.5
25.6
40.8
1.10
1.16
1.17
i
4.2448(7)
3.941(1)
M/D
D
0.5
0.5
0.5
0.5
0.5
5
180
216
0
311
364
140
306
360
431
0
622
729
28
95
97
15.8
17.3
1.11
1.15
3.8071(6)
4.0948(6)
4.1392(7)
4.1804(9)
3.8486(4)
D
NAh
98
NAh
NAh
Dj
24.9
23.3
1.17
1.15
Dj
99
51
90
M/D
85.8/7.26k
17.9
1.43/1.08k
1.15
Dj
0.33
917
a All reactions were performed in neat cyclohexene oxide (CHO) with [monomer]:[Zn] ) 1000:1 (10 mM in Zn) at 50 °C. b Zn‚‚Zn separations from
X-ray crystallographic studies. See Supporting Information for crystal data and structure refinement. c M ) monomeric species; D ) dimeric species;
M/D ) both monomer and dimer present. Solution dynamic studies performed in C6D6 at room temperature (10 mM in Zn). d TON ) turnover number;
moles of CHO consumed per mole of zinc. e TOF ) turnover frequency; moles of CHO consumed per mole of zinc per hour. f Calculated by integration of
methine resonances in 1H NMR of polymer (CDCl3, 300 MHz). g Determined by gel-permeation chromatography (GPC), calibrated with polystyrene standards
in tetrahydrofuran. h Not applicable. i Equimolar ratios of (BDI-1)ZnN(SiMe3)2 and (BDI-2)ZnN(SiMe3)2. j At room temperature, broad dimeric resonances
are observed, indicating a fluxional process on the NMR time scale. k A bimodal distribution is observed by GPC.
Examination of CHO and CO2 Enchainment Steps Using
Model Compounds. The single-site structure of BDI zinc
complexes allows detailed study of the copolymerization
enchainment steps through stoichiometric reactions with CO2
or CHO. The catalyst must react with one of the two monomers
to effectively initiate the copolymerization (see Scheme 2). As
expected, zinc alkoxides react with CO2 (Scheme 4). For
dioxide per dimeric complex when pressurized with 100 psi
CO2 in toluene at 50 °C. X-ray analysis revealed the molecular
structure to be [(BDI-2)Zn(µ-OMe)(µ,η2-O2COMe)Zn(BDI-2)]
(Figure 7; Table 1), a dimeric compound with a distorted
tetrahedral geometry around zinc and a planar six-membered
ring at the core of the molecule. [(BDI-2)Zn(µ-OMe)(µ,η2-O2-
COMe)Zn(BDI-2)] is isostructural to [(BDI-2)Zn(µ-OMe)-
(µ,η2-OAc)Zn(BDI-2)] and is further evidence supporting the
54
example, [(BDI-2)Zn(µ-OMe)]2 inserted 1 equiv of carbon
9
J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003 11915

A R T I C L E S
Moore et al.
Scheme 3. Solution Behavior of [(BDI-2)]Zn(µ-OMe)]2 and
[(BDI-2)Zn(µ-OAc)]2; Exclusive Product Is
[(BDI-2)Zn(µ-OMe)(µ-OAc)Zn(BDI-2)]
Figure 6. ORTEP drawing of [(BDI-2)Zn(µ-OMe)(µ,η2-OAc)Zn(BDI-2)]
(non-hydrogen atoms) with thermal ellipsoids drawn at the 40% probability
level. Selected bond lengths (Å) and bond angles (deg): Zn(1)-N(1) 1.980-
(2), Zn(1)-N(2) 1.983(2), Zn(1)-O(1) 1.930(2), Zn(1)-O(3) 1.999(2),
C(2)-O(2) 1.251(3), C(2)-O(3) 1.251(3), Zn(2)-O(1) 1.958(2), Zn(2)-
O(2) 1.956(2), N(1)-Zn(1)-N(2) 97.51(9), O(1)-Zn(1)-O(3) 102.18(8),
N(1)-Zn(1)-O(1) 118.07(8), N(1)-Zn(1)-O(3) 106.76(9), Zn(1)-O(1)-
Zn(2) 120.84(9), C(2)-O(3)-Zn(1) 132.6(2), O(2)-C(2)-O(3) 126.2(3),
C(2)-O(2)-Zn(2) 133.8(2).
initiators do not adequately imitate the structure of a growing
polymer chain. Despite this distinction, they are viable initiators
(see Scheme 4). Pressurizing a toluene solution of (BDI-2)-
ZnN(SiMe3)2 at 50 °C with 100 psi CO2 overnight resulted in
the formation of a white precipitate. Analysis of the mother
liquor by 1H NMR spectroscopy showed the presence of
trimethylsilyl isocyanate. After recrystallization, X-ray analysis
revealed a molecular structure corresponding to [(BDI-2)Zn-
(µ-OSiMe3)]2 (Figure 9; Table 1). The structure shows a bridged
siloxide dimer with a distorted tetrahedral geometry around each
zinc. The Zn-O bonds are 1.98 and 2.03 Å, the Zn-N bonds
are 1.96 and 2.05 Å, and the O-Si bond is 1.64 Å. We propose
that CO2 inserts into the zinc amido bond and is followed by
migration of a trimethylsilyl group and loss of trimethylsilyl
isocyanate.53,54,67 We are currently investigating why this
catalytically active zinc siloxide complex does not further react
with CO2 and crystallize as a six- or eight-membered ring
carbonate dimer. Finally, (BDI-1)ZnOiPr and (BDI-1)ZnN-
(SiMe3)2 do not react with CHO, indicating that the BDI zinc
alkoxides and bis(trimethylsilyl)amido complexes react readily
with only CO2.
thermodynamic stability of the six-membered dimeric species
compared to that of the four-membered zinc alkoxide and eight-
membered zinc acetate dimers.
Interestingly, when a toluene solution of the bulkier, mono-
meric [(BDI-1)ZnOiPr] was pressurized with 100 psi CO2 at
50 °C, clear blocks instantly crystallized from solution. X-ray
analysis revealed the molecular structure to be [(BDI-1)Zn(µ,η2-
O2COiPr)]2 (Figure 8; Table 1).65 Thus, CO2 is inserted into
each zinc isopropoxide bond; dimerization through bridging
carbonate groups forms a molecule containing distorted tetra-
hedral zinc centers. The bridging carbonates adopt a syn-syn
geometry, creating a planar eight-membered ring at the core of
the molecule. The Zn‚‚Zn separation is 4.01 Å, which is
considerably shorter than the 4.24 Å Zn‚‚Zn separation observed
in [(BDI-1)Zn(µ,η2-OAc)]2. [(BDI-1)Zn(µ,η2-O2COiPr)]2 is
nearly isostructural to [(BDI-1)Zn(µ,η2-OAc)]2, and we believe
that these related species closely mimic the putative propagating
species of the copolymerization reaction. Additionally, we
propose that CO2 insertion occurs into a monometallic zinc
alkoxide species due to the rapid reaction of CO2 with
monomeric, three-coordinate (BDI-1)ZnOiPr although we cannot
exclude alternate mechanisms.66
Because zinc alkoxides initiate the copolymerization by
reaction with CO2, zinc acetate compounds, mimics for the
putative carbonate species, were expected to complete the
(66) On the basis of their high reactivity, we expect propagating (BDI-1)Zn
alkoxides to be three-coordinate or weakly chelated, four-coordinate
monomeric species. We are currently studying the nature of the zinc
alkoxide intermediates in the polymerization and believe that their
coordination in solution is highly dependent on ligand sterics and
electronics. Furthermore, kinetic studies on the reaction of CO2 with (BDI-
1)ZnOiPr and other BDI zinc alkoxide complexes are underway.
The alkoxides are excellent models for the catalytic species
that enchain CO2; however, the zinc bis(trimethylsilyl)amido
(67) Chisholm has synthesized (BDI-1)ZnO(CO)NiPr2 from (BDI-1)ZnNiPr2 and
CO2 (see ref 63) but found no activity for CO2/epoxide copolymerization,
presumably because rearrangement to the zinc isopropoxide could not occur.
(65) The non-carbon atoms and carbon atoms of [(BDI-1)Zn(µ,η2-O2COiPr)]2
were refined anisotropically and isotropically, respectively.
9
11916 J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003

Epoxide−CO Copolymerization Using Zn â-Diiminate
A R T I C L E S
2
Figure 7. ORTEP drawing of [(BDI-2)Zn(µ-OMe)(µ,η2-O2COMe)Zn(BDI-
2)] (non-hydrogen atoms) with thermal ellipsoids drawn at the 40%
probability level. Selected bond lengths (Å) and bond angles (deg): Zn-
(1)-N(1) 1.976(3), Zn(1)-N(2) 1.965(3), Zn(1)-O(1) 1.947(3), Zn(1)-
O(2) 1.954(3), Zn(2)-O(1) 1.922(3), Zn(2)-O(3) 2.004(3), C(2)-O(4)
1.350(5), C(2)-O(2) 1.250(6), N(1)-Zn(1)-N(2) 98.1(1), O(1)-Zn(1)-
O(2) 103.0(1), N(1)-Zn(1)-O(1) 117.1(1), N(2)-Zn(1)-O(2) 112.5(1),
Zn(1)-O(1)-Zn(2) 122.4(2), C(2)-O(2)-Zn(1) 132.0(3), O(2)-C(2)-
O(3) 128.5(5), C(2)-O(3)-Zn(2) 133.0(3).
Figure 8. ORTEP drawing of [(BDI-1)Zn(µ,η2-O2COiPr)]2 (non-hydrogen
atoms) with thermal ellipsoids drawn at the 40% probability level. Selected
bond lengths (Å) and bond angles (deg): Zn(1)-N(1) 1.97(2), Zn(1)-
N(2) 1.96(2), Zn(1)-O(1) 2.027(8), Zn(1)-O(2) 1.939(8), O(1)-C(30)
1.23(1), O(2)-C(30A) 1.19(1), O(3)-C(30) 1.39(1), O(1)-Zn(1)-O(2)
126.9(3), N(1)-Zn(1)-N(2) 99.8(3), N(1)-Zn(1)-O(2) 109.7(9), N(2)-
Zn(1)-O(1) 102.7(9), C(30)-O(1)-Zn(1) 131.9(8), O(1)-C(30)-O(2A)
137(1).
Scheme 4. Insertion Reactions of CO2
Figure 9. ORTEP drawing of [(BDI-2)Zn(µ-OSiMe3)]2 (non-hydrogen
atoms) with thermal ellipsoids drawn at the 40% probability level. Selected
bond lengths (Å) and bond angles (deg): Zn(1)-N(1) 1.964(8), Zn(1)-
N(2) 2.052(8), Zn(1)-O(1) 1.978(2), Zn(1)-O(1A) 2.026(2), O(1)-Si(1)
1.641(2), O(1)-Zn(1)-O(1A) 83.97(8), N(1)-Zn(1)-N(2) 94.7(1), N(1)-
Zn(1)-O(1) 124.6(3), N(2)-Zn(1)-O(1A) 115.1(3), Zn(1)-O(1)-Zn(1A)
96.03(8).
of zinc acetates with CHO was anticipated to be extremely fast.
By 1H NMR spectroscopy, [(BDI-1)ZnOAc]2 only partially
reacted with excess CHO over the course of days at 50 °C to
give the ring-opened product (Scheme 5). Upon crystallization,
only starting material was isolated. We postulated that the ∆G°
of the reaction might be near zero and decided to study the
microscopic reverse of this process. H NMR spectroscopy
showed that the addition of trans-2-acetoxycyclohexanol to a
solution of (BDI-1)ZnN(SiMe3)2 gave a product with resonances
catalytic cycle by reacting with CHO. CO2 is an unreactive
compound whereas epoxides possess significant ring strain.
Therefore, we initially predicted the rate-determining step in
the copolymerization to be the insertion of CO2 into the zinc
alkoxide bond (Figure 10; Scheme 2). Consequently, the reaction
1
9
J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003 11917

A R T I C L E S
Moore et al.
Ideally, the most efficient catalysts would react with both
CO2 and CHO with low activation barriers and negative ∆G°
values (see Figure 10). On the basis of our extensive copolym-
erization studies with a variety of BDI zinc complexes,55 we
decided to investigate the insertion of CHO into more reactive
zinc acetate compounds. [(BDI-7)ZnOAc]2, a much more active
catalyst than [(BDI-1)ZnOAc]2 (see Table 2 for polymerization
data), reacted with a slight excess of CHO to yield a new product
over the course of several days at room temperature (see Scheme
5). Upon crystallization, the solid-state structure revealed the
mono-CHO inserted product, [(BDI-7)Zn(µ,η2-OAc)(µ,η1-
OCyOAc)Zn(BDI-7)] (Figure 11; Table 1). The dimeric com-
pound exhibits a distorted tetrahedral geometry around both zinc
centers, and the Zn‚‚Zn separation is 3.28 Å. The unreacted
acetate bridge is intermediate between anti-syn and syn-syn
geometries such that the O(1)-C(1) acetate bond is twisted out
of the plane defined by Zn(1), O(3), and Zn(2) by 29.5°. On
the basis of the elongated separation of 3.17 Å between O(4)
and Zn(2) and the lack of distortion to a trigonal-bipyramidal
geometry around zinc, no dative bond is observed between the
reacted acetate moiety and Zn(2). In addition, [(BDI-7)Zn(µ,η2-
OAc)(µ,η1-OCyOAc)Zn(BDI-7)] did not react with additional
equivalents of CHO over the course of 7 days, although it did
react rapidly with CO2 to complete one full cycle in the
alternating copolymerization of CHO and CO2. Therefore, we
Figure 10. Qualitative energy profiles for insertion of CO2 and CHO.
consistent with those of (BDI-1)ZnOCyOAc, which over 2 days
formed [(BDI-1)ZnOAc]2 and CHO. By elimination of CHO,
the microscopic reverse process occurred, indicating the reaction
of CHO with [(BDI-1)ZnOAc]2 is an unexpected equilibrium
process under these conditions (see Figure 10).
Scheme 5. Insertion Reactions of Cyclohexene Oxide (CHO)
9
11918 J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003

Epoxide−CO Copolymerization Using Zn â-Diiminate
A R T I C L E S
Scheme 6. Effect of Catalyst Combinations on Rate
2
copolymerization via formation of mixed (BDI-1) and (BDI-2)
zinc active species.
Figure 11. ORTEP drawing of [(BDI-7)Zn(µ,η2-OAc)(µ,η1-OCyOAc)-
Zn(BDI-7)] (non-hydrogen atoms) with thermal ellipsoids drawn at the 40%
probability level. Selected bond lengths (Å) and bond angles (deg): Zn-
(1)-N(1) 2.006(5), Zn(1)-N(2) 1.993(4), Zn(1)-O(1) 1.977(4), Zn(1)-
O(3) 1.951(3), Zn(2)-O(2) 2.007(4), Zn(2)-O(3) 1.951(3), O(1)-C(1)
1.254(7), O(1)-Zn(1)-O(3) 104.2(2), O(2)-Zn(2)-O(3) 100.4(2), N(1)-
Zn(1)-N(2) 96.1(2), N(1)-Zn(1)-O(3) 118.5(2), N(2)-Zn(1)-O(1) 113.3-
(2), Zn(1)-O(3)-Zn(2) 113.1(2), C(1)-O(1)-Zn(1) 120.7(4), C(1)-O(2)-
Zn(2) 125.7(4).
Table 2 also shows polymerization data for [(BDI)ZnOAc]
complexes using ligands 1-7. [(BDI-3)ZnOAc]2 exhibited no
polymerization activity, which we attribute to the tightly bound
dimeric state of the complex.54 Polymerization activity emerges
as the steric bulk on the N-aryl rings is increased. [(BDI-2)-
ZnOAc]2 copolymerized CHO and CO2 (100 psi) at 50 °C to
give a TOF of 431 h-1 in a 30-min polymerization. [(BDI-4)-
ZnOAc]2, which has para tert-butyl substituents on the 2,6-
diethylaniline moieties, elicited a further augmentation of
activity to 622 h-1. The unsymmetrical [(BDI-5)ZnOAc]2 is the
most active catalyst of the six (Figure 12) as it produced
polycarbonate with a TOF of 729 h-1. As the substituents on
the ortho positions continue to increase in size, the polymeri-
zation activity suffers. [(BDI-1)ZnOAc]2 produced poly(cyclo-
hexene carbonate) with a TOF of 360 h-1. Finally, the addition
of para tert-butyl substituents to the diisopropyl aniline moiety
on [(BDI-6)ZnOAc]2 causes a dramatic decrease in polymeri-
zation activity to only 28 h-1 over 5 h.69 As shown in Figure
12, the polymerization activity peaks with the [(BDI-5)ZnOAc]2
and declines as qualitative steric bulk is decreased or increased
on the zinc catalyst. We propose [(BDI-5)ZnOAc]2 contains
enough steric bulk to be a very reactive dimeric initiator. Both
[(BDI-1)ZnOAc] and [(BDI-6)ZnOAc] show reduced activities
due to excessive steric bulk. Therefore, [(BDI-5)ZnOAc]2
possesses an optimal ligand arrangement to catalyze the
copolymerization of CHO and CO2 effectively.
A proposed mechanism must account for several intriguing
observations: (1) the doubling of activity observed in the
copolymerization by (BDI-1)ZnN(SiMe3)2 and (BDI-2)ZnN-
(SiMe3)2 mixtures; (2) large activity differences for a series of
structurally related compounds; (3) the profound effect of adding
para tert-butyl groups to the catalysts; and (4) the exclusive
insertion of one CHO monomer unit per two zinc centers in
the case of [(BDI-7)ZnOAc]2. We considered that the BDI zinc
centers do not act independently but rather in a cooperative,
bimetallic fashion.55 A bimetallic mechanism satisfies all of
these observations: (1) CO2 insertion into the monomeric (BDI-
1)ZnN(SiMe3)2 and (BDI-2)ZnN(SiMe3)2 can form more acti-
propose that the rate-determining step in the copolymerization
is actually the insertion of CHO into a zinc carbonate bond.
Theoretical molecular orbital studies on the alternating copo-
lymerization mechanism of cyclohexene oxide and CO2 using
monomeric (BDI-1)ZnOMe also predict that the rate-determin-
ing step is insertion of CHO into a zinc carbonate bond.68 In
addition, [(BDI-1)ZnOAc] did not react with CO2. Finally, it
must be noted that like their precursors, the mimics for the
presumed intermediates in the copolymerization, including
[(BDI-2)Zn(µ-OMe)(µ,η2-OAc)Zn(BDI-2)], [(BDI-2)Zn(µ-OMe)-
(µ,η2-O2COMe)Zn(BDI-2)], [(BDI-2)Zn(µ-OSiMe3)]2, and [(BDI-
7)Zn(µ,η2-OAc)(µ,η1-OCyOAc)Zn(BDI-7)] but with the excep-
tion of the insoluble [(BDI-1)Zn(µ,η2-O2COiPr)]2, are highly
active for the alternating copolymerization of CHO and CO2.
Furthermore, they have similar TOFs and yield monodisperse
polycarbonate with low PDIs.
Rate Studies of the Alternating Copolymerization of CHO
and CO2. In Table 2, copolymerization data of monomeric
(BDI)ZnN(SiMe3)2 complexes54 with ligands 1 and 2 are
reported. (BDI-1)ZnN(SiMe3)2 and (BDI-2)ZnN(SiMe3)2 pro-
duced polycarbonate with similar activities (TOF ≈ 350 h-1
)
in a 30-min polymerization. The polycarbonate possessed greater
than 94% carbonate linkages, similar molecular weights (Mn ≈
25,000), and low PDIs (PDI ) 1.10-1.16). Interestingly, when
equimolar amounts of (BDI-1)ZnN(SiMe3)2 and (BDI-2)ZnN-
(SiMe3)2 were simultaneously exposed to cyclohexene oxide
and 100 psi CO2 in a one-pot reaction, such that the reaction
conditions and total zinc concentration were unchanged, a
doubling of activity (TOF ) 658 h-1) resulted (Scheme 6). The
polycarbonate produced possessed 94% carbonate linkages, a
Mn of 40,800 g/mol, and a unimodal PDI of 1.17. This result
suggests that two zinc centers are interacting during the
(69) We are currently investigating the origin of the bimodal GPC traces of the
polymer produced using [(BDI-6)ZnOAc] (see Table 2). Preparatory GPC
of the polymer shows the high molecular weight peak to be predominantly
polyether, whereas the low molecular weight fraction is primarily poly-
carbonate.
(68) Liu, Z. W.; Torrent, M.; Morokuma, K. Organometallics 2002, 21, 1056-
1071.
9
J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003 11919

A R T I C L E S
Moore et al.
Figure 12. Activity versus qualitative ligand sterics for (BDI)ZnOAc catalysts. Copolymerization of neat CHO ([CHO]:[Zn] ) 1000, 100 psi CO2, Trxn
50 °C).
)
vated (BDI-1)Zn/(BDI-2)Zn mixed catalytic species; (2) tightly
bound dimers do not allow coordination of and reaction with
CHO or CO2, whereas sterically encumbered monomeric species
cannot easily access a bimetallic transition state; (3) molecular
modeling suggests that para tert-butyl substituents do impede
access to a bimetallic transition state, but do not affect a
monometallic enchainment; and (4) CHO insertion requires a
dimeric zinc species with two bridging µ,η2-O2CR or µ,η2-O2-
COR groups (R ) alkyl, growing polymer chain) capable of
attacking the epoxide, and [(BDI-7)Zn(µ,η2-OAc)(µ,η1-OCyOAc)-
Zn(BDI-7)] lacks this requirement.
Bimetallic mechanisms have been widely observed in syn-
thetic catalytic processes as well as natural ones. For instance,
aminopeptidases, phosphatases, and phosphotriesterases are just
a few examples of natural enzymatic systems that benefit from
two zinc centers at the active site of the enzyme.70 Scientists
have tried to mimic the successful strategy of nature to harness
the benefits of cooperative bimetallic mechanisms.71,72 Several
organic transformations are presumed to occur via bimetallic
metal centers, including asymmetric aldol condensations,73 ring-
opening of epoxides,74,75 and homopolymerization of epoxides.76
Bi- or multimetallic active species have also been proposed in
epoxide/CO2 coupling reactions.77
Solution dynamics, stoichiometric initiation steps, and co-
polymerization rates of BDI zinc complexes all strongly suggest
that a bimetallic mechanism is indeed in operation in the
alternating copolymerization of CHO and CO2; however, kinetic
studies were performed to directly monitor the copolymeriza-
tion.55 In situ IR78 determines the rate of polycarbonate
(73) For references on asymmetric aldol condensation using bimetallic zinc
compounds, see the following and references within: (a) Kitamura, M.;
Okada, S.; Suga, S.; Noyori, R. J. Am. Chem. Soc. 1989, 111, 4028-4036.
(b) Kitamura, M.; Suga, S.; Niwa, M.; Noyori, R. J. Am. Chem. Soc. 1995,
117, 4832-4842. (c) Kitamura, M.; Suga, S.; Oka, H.; Noyori, R. J. Am.
Chem. Soc. 1998, 120, 9800-9809. (d) Trost, B. M.; Ito, H. J. Am. Chem.
Soc. 2000, 122, 12003-12004. (e) Trost, B. M.; Silcoff, E. R.; Ito, H.
Org. Lett. 2001, 3, 2497-2500. (f) Rasmussen, T.; Norrby, P. O. J. Am.
Chem. Soc. 2001, 123, 2464-2465. (g) Pu, L.; Yu, H. B. Chem. ReV. 2001,
101, 757-824.
(74) For asymmetric ring-opening of epoxides using Cr or Co salen complexes,
see the following and references within: (a) Jacobsen, E. N. Acc. Chem.
Res. 2000, 33, 421-431. (b) Ready, J. M.; Jacobsen, E. N. Angew. Chem.,
Int. Ed. 2002, 41, 1374-1377. (c) Schaus, S. E.; Brandes, B. D.; Larrow,
J. F.; Tokunaga, M.; Hansen, K. B.; Gould, A. E.; Furrow, M. E.; Jacobsen,
E. N. J. Am. Chem. Soc. 2002, 124, 1307-1315. (d) Ready, J. M.; Jacobsen,
E. N. J. Am. Chem. Soc. 2001, 123, 2687-2688. (e) Konsler, R. G.; Karl,
J.; Jacobsen, E. N. J. Am. Chem. Soc. 1998, 120, 10780-10781. (f) Hansen,
K. B.; Leighton, J. L.; Jacobsen, E. N. J. Am. Chem. Soc. 1996, 118, 10924-
10925.
(75) For Zr-catalyzed enantioselective ring-opening of epoxides, see the
following: (a) Nugent, W. A. J. Am. Chem. Soc. 1998, 120, 7139-7140.
(b) McCleland, B. W.; Nugent, W. A.; Finn, M. G. J. Org. Chem. 1998,
63, 6656-6666.
(76) For studies where a bimetallic epoxide enchainment mechanism is proposed,
see the following and references within: (a) Vandenberg, E. J. J. Polym.
Sci. 1960, 47, 486-489. (b) Price, C. C.; Spector, R. J. Am. Chem. Soc.
1966, 88, 4171-4173. (c) Osgan, M.; Teyssie, P. J. Polym. Sci. Part B:
Polym. Lett. 1967, 5, 789-792. (d) Vandenberg, E. J. J. Polym. Sci., Part
A: Polym. Chem. 1986, 24, 1423-1431. (e) Watanabe, Y.; Yasuda, T.;
Aida, T.; Inoue, S. Macromolecules 1992, 25, 1396-1400. (f) Sugimoto,
H.; Kawamura, C.; Kuroki, M.; Aida, T.; Inoue, S. Macromolecules 1994,
27, 2013-2018. (g) Chisholm, M. H.; Navarro-Llobet, D.; Simonsick, W.
J. Macromolecules 2001, 34, 8851-8857. (h) Braune, W.; Okuda, J. Angew.
Chem., Int. Ed. 2003, 42, 64-68. To the best of our knowledge, no kinetic
studies have been performed on these systems. However, Inoue has reported
kinetic studies for the ring-opening polymerization of δ-valerolactone that
illustrate a second-order dependence in aluminum porphyrin alcoholates.
See: Shimasaki, K.; Aida, T.; Inoue, S. Macromolecules 1987, 20, 3076-
3080.
(70) For reviews on cooperative bimetallic catalysis in enzymatic systems, see:
(a) Cricco, J. A.; Orellano, E. G.; Rasia, R. M.; Ceccarelli, E. A.; Vila, A.
J. Coord. Chem. ReV. 1999, 192, 519-535. (b) Wilcox, D. E. Chem. ReV.
1996, 96, 2435-2458. (c) Lipscomb, W. N.; Strater, N. Chem. ReV. 1996,
96, 2375-2433. (d) Coleman, J. E. Curr. Opin. Chem. Biol. 1998, 2, 222-
234. (e) Strater, N.; Lipscomb, W. N.; Klabunde, T.; Krebs, B. Angew.
Chem., Int. Ed. Engl. 1996, 35, 2024-2055. (f) Steinhagen, H.; Helmchen,
G. Angew. Chem., Int. Ed. Engl. 1996, 35, 2339-2342.
(71) For an example of biomimetic, bimetallic catalyst design for lactone
polymerization, see: Williams, C. K.; Brooks, N. R.; Hillmyer, M. A.;
Tolman, W. B. Chem. Commun. 2002, 2132-2133.
(72) For a review on synthetic bimetallic catalysts as enzymatic mimics, see:
van den Beuken, E. K.; Feringa, B. L. Tetrahedron 1998, 54, 12985-
13011.
9
11920 J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003

Epoxide−CO Copolymerization Using Zn â-Diiminate
A R T I C L E S
2
formation through the emerging carbonyl stretch (approximately
1750 cm-1).17,18,79,80 To determine the rate law, we measured
reaction orders in all three components of the polymerization62,81
(eq 1):
[(BDI-1)ZnOAc] (13.9 mM in total Zn) at 30 °C.62 As expected,
the polycarbonate carbonyl stretch intensities increased with
higher CHO concentrations. The initial rates for each run were
plotted versus [CHO], demonstrating a first-order dependence
on monomer concentration.62 Therefore, the rate-determining
step is indeed the insertion of CHO into a growing carbonate
species. The polymerization proceeds as shown in eq 3:
d[P]/dt ) kp[CHO]x[CO2]y[Zn]tot
(1)
z
On the basis of stoichiometric initiation reactions, we believed
CHO insertion would be rate-determining (first-order), whereas
CO2 insertion would be a fast step and would thereby not be
included in the rate equation for the polymerization. The
copolymerization of CHO (1.98 M in toluene) and CO2 from
50 to 200 psi was performed using [(BDI-5)ZnOAc] (9.88 mM
in total Zn) at 30 °C and [(BDI-2)ZnOAc] (15.4 mM in total
Zn) at 50 °C. Zeroth-order kinetics in CO2 monomer were
observed by monitoring initial rates (up to 15% conversion).62
Initial rates for [(BDI-5)ZnOAc] and [(BDI-2)ZnOAc] were
approximately 0.025 and 0.008 absCO/min (the carbonyl absor-
bance, absCO, is directly proportional to [polycarbonate]),
respectively.62,82 Surprisingly, polycarbonate was synthesized
with pressures as low as 20 psi CO2 using [(BDI-5)ZnOAc].
This is an unprecedented result, as previously reported polym-
erization systems require much higher pressures of CO2.83 At
20 psi, the rate of backbiting competes with CO2 insertion into
the growing polymer chain, producing a small amount of trans-
d[P]/dt ) kobs[CHO]1
(3)
where kobs ) kp[Zn]totz. To ascertain the order in total zinc,
[Zn]tot, the copolymerization of 1.98 M CHO in toluene and
300 psi CO2 was completed using [(BDI-1)ZnOAc] (3.87-13.4
mM in total Zn) at 30 °C and 50 °C. By plotting ln kobs versus
ln [Zn]tot, orders in total zinc of 1.39 ( 0.04 and 1.73 ( 0.06
were discovered at 30 and 50 °C (Figure 13), respectively.
Therefore, the copolymerization of CHO and CO2 using [(BDI-
1)ZnOAc] at 50 °C exhibited the following overall rate law:
d[P]/dt ) kp[CHO]1[CO2]0[Zn]tot
(4)
1.73
Furthermore, the copolymerization of 1.98 M CHO in toluene
and 300 psi CO2 was investigated using [(BDI-5)ZnOAc]
(1.04-9.88 mM in total Zn) at 30 and 50 °C. By plotting ln
kobs versus ln [Zn]tot, orders in total zinc of 1.37 ( 0.02 and
1.83 ( 0.04 were discovered at 30 and 50 °C (Figure 13),
respectively. Figure 14 illustrates a plot of initial rate versus
[Zn]tot at 50 °C, which clearly shows a higher order dependence
in [(BDI-5)ZnOR], where R is alkyl, acyl, or polymer chain.
Orders in total zinc that approach 2 are most logically
explained by a bimetallic transition state for epoxide ring
opening and a predominantly monomeric (BDI)ZnO2COR
ground state. We attribute the increase in order for both [(BDI-
1)ZnOAc] and [(BDI-5)ZnOAc] at higher temperatures to a
greater proportion of [(BDI)ZnO2COR] in the monomeric (M)
state (Figure 15). Dynamic solution studies (monomer/dimer
equilibrium) of [(BDI-1)ZnOAc] show an increase in the
concentration of the monomeric form of the complex as the
temperature rises. Hence, if a bimetallic mechanism is operating,
a higher percentage of monomeric ground state would increase
the order in [(BDI)ZnO2COR] at higher temperatures. The rate
law for CHO/CO2 copolymerization is a summation of five
possible elementary steps, as shown in Figure 15. Although there
are multiple possible bimetallic transition structures that ac-
commodate the near second order in total zinc, we currently
favor the one in Figure 15.
Even though rate studies indicate the dominance of a
bimetallic mechanism, we attempted to collect more information
to probe the possibility of a mixture of monometallic and
bimetallic transition states. If a monometallic transition state
exclusively operates, the rate of a catalyst that exists as a dimer
in the ground state would theoretically exhibit an order in zinc
dimer (D) of 0.5. Conversely, if only a bimetallic transition state
occurs, the rate of the same dimeric catalyst would show an
order in zinc dimer (D) of 1.0. Finally, the zinc centers of a
dimeric catalyst may act independently; this will exhibit the
same kinetics. Because [(BDI-2)ZnOAc] is solely dimeric by
1H NMR spectroscopy, even at elevated temperatures of
100 °C, we investigated its kinetics for the alternating copo-
lymerization of CO2 and CHO. The copolymerization of
1.98 M CHO in toluene and 300 psi CO2 was completed using
1
cyclohexenecarbonate (approximately 3% by H NMR spec-
troscopy). Given the zero-order dependence in CO2, the overall
rate law was simplified to eq 2:
z
d[P]/dt ) kp[CHO]x[CO2]0[Zn]tot
(2)
where kp is the propagation rate constant. To determine the order
in epoxide monomer (x), various concentrations of CHO (0.82-
3.29 M in toluene) were copolymerized with 300 psi CO2 using
(77) See refs 12, 22, 24, 27-30, 55 and the following: (a) Kobayashi, M.;
Inoue, S.; Tsuruta, T. Macromolecules 1971, 4, 658-659. (b) Inoue, S.;
Kobayashi, M.; Koinuma, H.; Tsuruta, T. Makromol. Chem. 1972, 155,
61-73. (c) Kobayashi, M.; Inoue, S.; Tsuruta, T. J. Polym. Sci.: Polym.
Chem. Ed. 1973, 11, 2383-2385. (d) Kobayashi, M.; Tang, Y. L.; Tsuruta,
T.; Inoue, S. Makromol. Chem. 1973, 169, 69-81. (e) Kuran, W.;
Pasynkiewicz, S.; Skupinska, J.; Rokicki, A. Makromol. Chem. Macromol.
Chem. Phys. 1976, 177, 11-20. (f) Kuran, W. Appl. Organomet. Chem.
1991, 5, 191-194. (g) Kuran, W.; Listos, T. Macromol. Chem. Phys. 1994,
195, 1011-1015.
(78) ASI Applied Systems, Millersville, MD 21108.
(79) In all kinetic studies, the resultant polycarbonates contain >99% carbonate
linkages as determined by 1H NMR spectroscopy. In addition, the
polycarbonates possess molecular weights based on [CHO]:[Zn] and PDIs
are approximately 1.1, as determined by gel permeation chromatography
(calibrated with polystyrene standards in tetrahydrofuran). It should be noted
that higher levels of carbonate linkages are found in polycarbonate
synthesized in toluene solution versus neat CHO. A similar phenomenon
has been observed in ref 17.
(80) For some examples of in situ IR studies, see: (a) Rutherford, J. L.;
Hoffmann, D.; Collum, D. B. J. Am. Chem. Soc. 2002, 124, 264-271. (b)
Sun, X. F.; Collum, D. B. J. Am. Chem. Soc. 2000, 122, 2452-2458. (c)
Lucht, B. L.; Collum, D. B. J. Am. Chem. Soc. 1995, 117, 9863-9874. (d)
Bernstein, M. P.; Collum, D. B. J. Am. Chem. Soc. 1993, 115, 8008-
8018. (e) Allmendinger, M.; Eberhardt, R.; Luinstra, G.; Rieger, B. J. Am.
Chem. Soc. 2002, 124, 5646-5647.
(81) The concentration of BDI zinc complexes, [(BDI)ZnOR] or [Zn]tot, although
expressed in units of molarity, refers to the concentration of the monomer
unit (normality). However, note that in solution, [Zn]tot ) 2[D] + [M],
where D ) dimer and M ) monomer.
(82) Higher CO2 pressures (approximately 200 psi CO2 and above) cause slightly
decreased initial rates, which we attribute to a dilution of epoxide.
Furthermore, control experiments show lower carbonyl absorbance levels
as CO2 pressure is increased (see Supporting Information). Therefore, we
have adjusted the initial rate values to account for these decreased
absorbance levels.
(83) For leading references employing catalyst systems under high pressures of
CO2 or supercritical CO2, see refs 8-21.
9
J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003 11921

A R T I C L E S
Moore et al.
Figure 13. Copolymerization of CHO (1.98 M in toluene) + CO2 (300 psi) using [(BDI)ZnOAc] catalysts.
order in total zinc of 1.30 ( 0.07 was determined (see Figure
13). We interpret this to mean that a more significant proportion
of monomeric ground state persists with [(BDI-7)ZnOAc] than
with [(BDI-2)ZnOAc]. We propose that electron withdrawal by
the cyano substituent increases the electrophilicity of [(BDI)-
ZnOR] and thereby facilitates the coordination of CHO. The
accumulated rate data demonstrate orders in total zinc from 1.02
to 1.83, which vary due to temperature and steric and electronic
influences of the catalyst. These factors also directly affect
solution behavior and copolymerization activities. As sterics and/
or temperature increase, the proportion of monomeric complex
in solution, and, accordingly, the order in total zinc, rise. As
groups on the ligand framework become too bulky, however,
activity suffers due to a monomeric species that struggles to
access a bimetallic transition state. Catalysts that can attain a
loose dimeric ground state have kinetic advantages over their
monomeric or strongly dimerized counterparts.
Scheme 7 shows the proposed mechanism for the alternating
copolymerization of CHO and CO2 using BDI zinc complexes
based on our copolymerization activities and rate data, dynamic
solution studies, and insight into the enchainment steps. The
zinc alkoxide monomer (A) or dimer (A2) inserts CO2 to yield
either carbonate complexes (B, B2) or an alkoxide/carbonate
dimer (AB), depending on relative sterics and propagating
groups. A2, A, and AB compounds do not react with CHO. The
dimeric eight-membered carbonate dimer (B2) mimics [(BDI)-
ZnOAc]2 compounds, which with adequate bulk (e.g., [(BDI-
1)Zn(µ-OAc)]), participate in a monomer/dimer equilibrium. B2
reacts with CHO in an equilibrium process (e.g., [(BDI-1)Zn-
(µ-OAc)]2) or cleanly (e.g., [(BDI-7)Zn(µ-OAc)]2) to AB, but
B and B2 do not react with CO2. The trans ring-opened product
collapses to AB (e.g., [(BDI-7)Zn(µ,η2-OAc)(µ,η1-OCyOAc)-
Zn(BDI-7)]). AB subsequently reacts with CO2 to complete one
full catalytic cycle. This cycle is continually propagated to make
Figure 14. Initial rates of the copolymerization of CHO (1.98 M in toluene)
+ CO2 (300 psi) using [(BDI-5)ZnOAc] at 50 °C.
[(BDI-2)ZnOAc] (4.01-15.6 mM in total Zn) at 50 °C. By
plotting ln kobs versus ln [Zn]tot, an order in total zinc of 1.02 (
0.03 was determined (see Figure 13). A plot of initial rate versus
[Zn]tot for the polymerization reactions using [(BDI-1)ZnOAc]
and [(BDI-2)ZnOAc] at 50 °C demonstrates an order in total
zinc that is 1 for [(BDI-2)ZnOAc] and greater than 1 for [(BDI-
1)ZnOAc] (Figure 16).
Rate studies were also performed on [(BDI-7)ZnOAc] to
probe the effect of the electron-withdrawing cyano substituent.
The copolymerization of 1.98 M CHO in toluene and 300 psi
CO2 was investigated using [(BDI-7)ZnOAc] (2.06-9.86 mM
in total Zn) at 50 °C. By plotting ln kobs versus ln [Zn]tot, an
9
11922 J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003

Epoxide−CO Copolymerization Using Zn â-Diiminate
A R T I C L E S
2
Figure 15. Potential ground and transition structures for ring opening of CHO, and the corresponding rate equations for five possible elementary steps. Rate
equations (1) and (2) indicate processes that proceed through a monometallic transition state, whereas (3), (4), and (5) show processes that utilize a bimetallic
transition state. The observable ground state associated with (1), (3) , and (5) is a monomeric species, whereas the ground state associated with (2) and (4)
is a dimeric species.
Figure 16. Copolymerization of CHO (1.98 M in toluene) + CO2 (300 psi) using [BDI-1)ZnOAc] and [BDI-2)ZnOAc] catalysts at 50 °C.
a polycarbonate chain possessing a molecular weight determined
by the ratio of [CHO] to [Zn]. With the aid of rate data, we
believe B and B2 both participate as ground state species and
access a bimetallic transition state in the rate-determining step
of the copolymerization (see Scheme 7). We tentatively propose
that one zinc species coordinates and activates the epoxide and
9
J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003 11923

A R T I C L E S
Moore et al.
Scheme 7. Proposed Mechanism of Copolymerization (CHO ) cyclohexene oxide; P ) polymer chain)
the second zinc species delivers the carbonate propagating
species to the back side of the cis-epoxide ring in a concerted
fashion. The ratio of B2 to B in the ground state changes
depending on reaction temperature and steric and electronic
factors of the BDI framework, and consequently the orders in
zinc are altered.
aim to investigate the possibility of a monometallic transition
state and to quantitate the equilibrium constants, thermodynam-
ics, and the activation barriers involved in the insertion processes
of zinc alkoxides with CO2 and zinc acetates with epoxides.
Taking the mechanism into account, we will develop new
catalysts with improved activities and selectivities. We will also
try to understand the importance of the unsymmetrical ligand
geometries in facilitating catalytic activities. Finally, we will
investigate other monomers, such as propylene oxide, to
determine whether a similar bimetallic mechanism is occurring.
Conclusion
We have synthesized a wide array of BDI zinc complexes
that demonstrate excellent activities for the alternating copo-
lymerization of CHO and CO2. Dynamic solution studies reveal
that sterically hindered zinc acetates, including [(BDI-1)Zn(µ-
OAc)]2 and [(BDI-6)Zn(µ-OAc)]2, participate in monomer/dimer
equilibria which demonstrate enthalpic stability toward the
dimer. The sterically less impeded [(BDI-2)Zn(µ-OAc)]2 and
[(BDI-2)Zn(µ-OMe)]2 complexes react to from a six-membered
ring mixed dimeric species, suggesting the stability of the six-
membered ring over the four-membered alkoxides and eight-
membered acetate dimers. Stoichiometric enchainment reactions
reveal BDI zinc acetates react with CHO and zinc alkoxides
react with CO2. In addition, copolymerization behavior suggests
that two zinc centers are interacting in the polymerization. Rate
studies were performed on the copolymerization, showing a
zero-order dependence in CO2, a first-order dependence in CHO,
and orders in total zinc from 1.0 to 1.8. Variations in ligand
sterics, electronics, and temperature play a major role in the
order in total zinc. We propose that a bimetallic mechanism is
in operation, where sterically hindered zinc complexes, such
as [(BDI-1)ZnOAc] and [(BDI-5)ZnOAc], access a bimetallic
transition state through both monomeric and dimeric ground
states. Sterically unencumbered zinc complexes, including
[(BDI-2)ZnOAc], operate via a bimetallic transition state from
a completely dimeric ground state. The rate studies support a
bimetallic mechanism; however, kinetics do not prove a
mechanism. We cannot discount the possibility of minor
contribution from a monometallic mechanism. Future work will
Acknowledgment. We thank Professors David Collum and
Marc Hillmyer for useful discussions on kinetics and Joseph J.
Reczek for the synthesis of 2,6-diethyl-4-tert-butylaniline. This
work was generously supported by the NSF (CHE-0243605).
We also thank the Cornell University Center for Biotechnology,
the New York State Center for Advanced Technology supported
by the New York State Science and Technology Foundation,
and industrial partners for partial support of this research. This
research made use of the Cornell Center for Materials Research
Shared Experimental Facilities, supported through the National
Science Foundation Materials Research Science and Engineering
Centers program (DMR-0079992), and the Cornell Chemistry
Department X-ray Facility (supported by the NSF; CHE-
9700441). G.W.C. gratefully acknowledges an Alfred P. Sloan
Research Fellowship, an Arnold and Mabel Beckman Founda-
tion Young Investigator Award, a Camille Dreyfus Teacher-
Scholar Award, a Packard Foundation Fellowship in Science
and Engineering, and generous support from Eastman Chemical.
D.R.M. is grateful for a Corning Foundation Science Fellowship.
Supporting Information Available: Experimental details,
including synthetic details, crystal structure data for all com-
plexes, rate studies, and equilibria plots. This material is
available free of charge via the Internet at http://pubs.acs.org.
JA030085E
9
11924 J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003