Cobalt catalysts for the alternating copolymerization of propylene oxide and carbon dioxide: Combining high activity and selectivity

Article abstract of DOI:10.1021/ja051744l

Synthetic pathways to (salcy)CoX (salcy = N,N′-bis(3,5-di-tert- butylsalicylidene)-1,2-diaminocyclohexane; X = halide or carboxylate) complexes are described. Complexes (R,R)-(salcy)CoCl, (R,R)-(salcy)CoBr, (R,R)-(salcy)CoOAc, and (R,R)-(salcy)CoOBzF

Full text of DOI:10.1021/ja051744l

Published on Web 07/14/2005
Cobalt Catalysts for the Alternating Copolymerization of
Propylene Oxide and Carbon Dioxide: Combining High
Activity and Selectivity
Claire T. Cohen, Tony Chu, and Geoffrey W. Coates*
Contribution from the Department of Chemistry and Chemical Biology, Baker Laboratory,
Cornell UniVersity, Ithaca, New York 14853-1301
Received March 18, 2005; E-mail: gc39@cornell.edu
Abstract: Synthetic pathways to (salcy)CoX (salcy ) N,N ′-bis(3,5-di-tert-butylsalicylidene)-1,2-diaminocy-
clohexane; X ) halide or carboxylate) complexes are described. Complexes (R,R)-(salcy)CoCl, (R,R)-
(salcy)CoBr, (R,R)-(salcy)CoOAc, and (R,R)-(salcy)CoOBzF5 (OBzF5 ) pentafluorobenzoate) are highly
active catalysts for the living, alternating copolymerization of propylene oxide (PO) and CO2, yielding poly-
(propylene carbonate) (PPC) with no detectable byproducts. The PPC generated using these catalyst
systems is highly regioregular and has up to 99% carbonate linkages with a narrow molecular weight
distribution (MWD). Inclusion of the cocatalysts [PPN]Cl or [PPN][OBzF5] ([PPN] ) bis(triphenylphosphine)-
iminium) with complex (R,R)-(salcy)CoCl, (R,R)-(salcy)CoBr, or (R,R)-(salcy)CoOBzF5 results in remarkable
activity enhancement of the copolymerization as well as improved stereoselectivity and regioselectivity
with maximized reactivity at low CO2 pressures. In the case of [PPN]Cl with (R,R)-(salcy)CoOBzF5, an
unprecedented catalytic activity of 620 turnovers per hour is achieved for the copolymerization of rac-PO
and CO2, yielding iso-enriched PPC with 94% head-to-tail connectivity. The stereochemistry of the monomer
and catalyst used in the copolymerization has dramatic effects on catalytic activity and the PPC
microstructure. Using catalyst (R,R)-(salcy)CoBr with (S)-PO/CO2 generates highly regioregular PPC,
whereas using (R)-PO/CO2 with the same catalyst gives an almost completely regiorandom copolymer.
The rac-PO/CO2 copolymerization with catalyst rac-(salcy)CoBr yields syndio-enriched PPC, an unreported
PPC microstructure. In addition, (R,R)-(salcy)CoOBzF5/[PPN]Cl copolymerizes (S)-PO and CO2 with a
turnover frequency of 1100 h-1, an activity surpassing that observed in any previously reported system.
Introduction
The copolymerization of CO2 and alicyclic epoxides with
discrete metal complexes has been well documented.4-10
CO2 is an ideal synthetic feedstock because it is abundant,
inexpensive, and nontoxic. Considerable interest in CO2 activa-
tion with transition metal complexes is directed toward its use
in organic synthesis.1 One growing area in CO2 chemistry is
the development of catalysts for the alternating copolymerization
of CO2 and epoxides to form polycarbonates,2 a field pioneered
by Inoue and co-workers in the late 1960s.3
Aliphatic epoxides, such as propylene oxide (PO), remain a
greater challenge owing to the concomitant production of cyclic
carbonate as shown in Scheme 1.5,6,8,10-14 The generalized
(4) For some examples of alicyclic epoxide/CO2 copolymerzations see: (a)
Moore, D. R.; Cheng, M.; Lobkovsky, E. B.; Coates, G. W. Angew. Chem.,
Int. Ed. 2002, 41, 2599-2602. (b) 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. (c) Cheng, M.; Lobkovsky, E. B.; Coates, G. W.
J. Am. Chem. Soc. 1998, 120, 11018-11019. (d) Darensbourg, D. J.;
Wildeson, J. R.; Yarbrough, J. C.; Reibenspies, J. H. J. Am. Chem. Soc.
2000, 122, 12487-12496. (e) Darensbourg, D. J.; Niezgoda, S. A.; Draper,
J. D.; Reibenspies, J. H. J. Am. Chem. Soc. 1998, 120, 4690-4698. (f)
Darensbourg, D. J.; Holtcamp, M. W. Macromolecules 1995, 28, 7577-
7579. (g) Sugimoto, H.; Ohshima, H.; Inoue, S. J. Polym. Sci, Part A:
Polym. Chem. 2003, 41, 3549-3555. (h) Mang, S.; Cooper, A. I.;
Colclough, M. E.; Chauhan, N.; Holmes, A. B. Macromolecules 2000, 33,
303-308. (i) Sarbu, T.; Beckman, E. J. Macromolecules 1999, 32, 6904-
6912. (j) Super, M.; Berluche, E.; Costello, C.; Beckman, E. Macromol-
ecules 1997, 30, 368-372.
(5) Darensbourg, D. J.; Mackiewicz, R. M.; Rodgers, J. L.; Phelps, A. L. Inorg.
Chem. 2004, 43, 1831-1833.
(6) Darensbourg, D. J.; Yarbrough, J. C.; Ortiz, C.; Fang, C. C. J. Am. Chem.
Soc. 2003, 125, 7586-7591.
(7) Darensbourg, D. J.; Billodeaux, D. R. Inorg. Chem. 2005, 44, 1433-1442.
(8) Aida, T.; Ishikawa, M.; Inoue, S. Macromolecules 1986, 19, 8-13.
(9) Darensbourg, D. J.; Mackiewicz, R. M.; Rodgers, J. L.; Fang, C. C.;
Billodeaux, D. R.; Reibenspies, J. H. Inorg. Chem. 2004, 43, 6024-6034.
(10) Darensbourg, D. J.; Yarbrough, J. C. J. Am. Chem. Soc. 2002, 124, 6335-
6342.
(1) (a) Musie, G.; Wei, M.; Subramaniam, B.; Busch, D. H. Coord. Chem.
ReV. 2001, 219, 789-820. (b) Cooper, A. I. J. Mater. Chem. 2000, 10,
207-234. (c) Bolm, C.; Beckmann, O.; Dabard, O. A. G. Angew. Chem.,
Int. Ed. 1999, 38, 907-909. (d) Yin, X. L.; Moss, J. R. Coord. Chem.
ReV. 1999, 181, 27-59. (e) Gibson, D. H. Chem. ReV. 1996, 96, 2063-
2095. (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.
(2) For reviews on epoxide/CO2 copolymerizations, see: (a) Moore, D. R.;
Coates, G. W. Angew. Chem., Int. Ed. 2004, 43, 6618-6639. (b) Sugimoto,
H.; Inoue, S. J. Polym. Sci., Part A: Polym. Chem. 2004, 42, 5561-5573.
(c) Darensbourg, D. J.; Mackiewicz, R. M.; Phelps, A. L.; Billodeaux, D.
R. Acc. Chem. Res. 2004, 37, 836-844. (d) Darensbourg, D. J.; Holtcamp,
M. W. Coord. Chem. ReV. 1996, 153, 155-174. (e) Super, M. S.; Beckman,
E. J. Trends Polym. Sci. 1997, 5, 236-240. (f) Kuran, W. Prog. Polym.
Sci. 1998, 23, 919-992.
(3) (a) Inoue, S.; Koinuma, H.; Tsuruta, T. J. Polym. Sci., Part B: Polym.
Lett. 1969, 7, 287-292. (b) Inoue, S.; Koinuma, H.; Tsurata, T. Makromol.
Chem. 1969, 130, 210-220. (c) Inoue, S.; Koinuma, H.; Tsurata, T. Polym.
Lett. 1969, 7, 287-292.
9
10.1021/ja051744l CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 10869-10878
10869

A R T I C L E S
Cohen et al.
Scheme 1. Reaction of Propylene Oxide and CO2 to Yield
Poly(propylene carbonate) and Propylene Carbonate
sequence for the PO/CO2 copolymerization involves epoxide
ring opening by a metal carbonate to give a metal alkoxide,
followed by CO2 insertion (Scheme 2). In addition, propylene
carbonate (PC) can be formed through a back-biting reaction
from the propagating alkoxide.11
Figure 1. Catalysts for the copolymerization of propylene oxide and CO2.
Recent advances using â-diiminate zinc or salen-type (salen
) N,N′-bis(salicylidene)-1,2-diaminoethane) chromium catalysts
with Lewis base cocatalysts (Figure 1) have resulted in the
highest reported activities for this reaction but produce signifi-
cant quantities of the byproduct PC.5,6,10-12 In an effort to
improve selectivity for poly(propylene carbonate) (PPC) over
PC, we searched for metal/ligand combinations known to effect
stereo- and regioselective reactions with PO. Jacobsen and co-
workers have reported chiral (salcy)CoIII carboxylates (salcy )
N,N′-bis(3,5-di-tert-butylsalicylidene)-1,2-diaminocyclohex-
ane) for the hydrolytic kinetic resolution of epoxides with
remarkable efficiencies.15 When we applied these catalysts to
PO/CO2 copolymerization, we observed unprecedented selectiv-
ity for highly regioregular PPC.16 Specifically, (R,R)-(salcy)-
CoOAc generated PPC with up to 99% carbonate linkages and
93% head-to-tail (HT) connectivity (Figure 2); however, it did
not yield high number average molecular weights (Mns), narrow
molecular weight distributions (MWDs), or fast reaction rates
characteristic of the zinc- and chromium-based alternatives.5,6,10-12
Figure 2. Regiochemistry of poly(propylene carbonate).
(R,R)-(salcy)CoOAc system, we have furthered catalyst opti-
mization, which has led to the use of organic salt cocatalysts
as well.5,7,8,17,20
Herein, we present our continued studies concerning the PO/
CO2 copolymerization using (salcy)CoX complexes. We explore
the influence of the catalyst initiator (X) on overall performance
and investigate the effects of [PPN]+-based ([PPN] ) bis-
(triphenylphosphine)iminium) ionic cocatalysts on catalyst
activity, product selectivity, stereoselectivity, and regioselec-
tivity. Additionally, we describe the diverse PPC stereo- and
regiochemistries that result from variation of the catalyst and
PO stereochemistry. These optimized systems demonstrate
unprecedented reaction rates and selectivity while producing
PPCs with high molecular weights and narrow MWDs.
The coupling of porphyrin or salen-type metal catalysts with
organic salt cocatalysts has resulted in improved catalytic
activities in a variety of polymerization systems.5,7,8,17-20 The
addition of ammonium salt cocatalysts (n-Bu4NY; Y ) Cl, Br,
I) to (R,R)-(salcy)CoX (X ) OAc, O2CCCl3, O2CCF3, Cl, OTs;
OTs ) tosylate) complexes in the presence of PO and CO2
shows reaction rates that are sensitive to the anionic components
(X and Y) for the generation of PC.21 Although it was originally
thought that ammonium salt cocatalysts limited the (R,R)-
(salcy)CoX catalyzed reaction of PO/CO2 to the production of
PC, PPC was recently achieved through the careful choice of
the complex and cocatalyst.20 Specifically, (R,R)-(salcy)CoOAr
(OAr ) 4-nitrophenoxy, 2,4-dinitrophenoxy, or 2,4,6-trinitro-
phenoxy) derivatives with [n-Bu4N]Y (Y ) Cl, OAc) cocatalysts
were described by Lu and co-workers to afford PPC with
enhanced rates and selectivity. Since our discovery of the active
Results and Discussion
Alternating Copolymerization of Propylene Oxide and
Carbon Dioxide with (salcy)CoX Catalysts. We recently
reported that (R,R)-(salcy)CoOAc catalyzes the alternating
copolymerization of rac- or (S)-PO and CO2 and exhibits
unprecedented selectivity for PPC. Average turnover frequencies
(TOFs), however, were limited to less than 85 h-1 16
. In addition,
(11) Allen, S. D.; Moore, D. R.; Lobkovsky, E. B.; Coates, G. W. J. Am. Chem.
Soc. 2002, 124, 14284-14285.
the MWDs of PPC synthesized by (R,R)-(salcy)CoOAc were
relatively broad (MWD ) 1.2-2.6), and the measured Mn values
of the polymers were lower than the corresponding theoretical
values, results consistent with chain transfer to water throughout
the copolymerization.16 It has been shown that minor alterations
in synthetic conditions and ligand structure can result in large
differences in the catalytic performance of salen-type chromium
complexes for epoxide/CO2 copolymerization.9 Herein, a de-
tailed study of the copolymerization reaction conditions and
nature of the initiating group and cocatalyst structure has
provided for improved molecular weight control, increased
polymerization activity, and mechanistic insight.
(12) Eberhardt, R.; Allmendinger, M.; Rieger, B. Macromol. Rapid Commun.
2003, 24, 194-196.
(13) Ree, M.; Bae, J. Y.; Jung, J. H.; Shin, T. J. J. Polym. Sci., Part A: Polym.
Chem. 1999, 37, 1863-1876.
(14) Chisholm, M. H.; Navarro-Llobet, D.; Zhou, Z. P. Macromolecules 2002,
35, 6494-6504.
(15) Tokunaga, M.; Larrow, J. F.; Kakiuchi, F.; Jacobsen, E. N. Science 1997,
277, 936-938.
(16) Qin, Z. Q.; Thomas, C. M.; Lee, S.; Coates, G. W. Angew. Chem., Int. Ed.
2003, 42, 5484-5487.
(17) Aida, T.; Inoue, S. Acc. Chem. Res. 1996, 29, 39-48.
(18) Aida, T.; Inoue, S. J. Am. Chem. Soc. 1985, 107, 1358-1364.
(19) Aida, T.; Sanuki, K.; Inoue, S. Macromolecules 1985, 18, 1049-1055.
(20) Lu, X. B.; Wang, Y. Angew. Chem., Int. Ed. 2004, 43, 3574-3577.
(21) Lu, X. B.; Liang, B.; Zhang, Y. J.; Tian, Y. Z.; Wang, Y. M.; Bai, C. X.;
Wang, H.; Zhang, R. J. Am. Chem. Soc. 2004, 126, 3732-3733.
9
10870 J. AM. CHEM. SOC. VOL. 127, NO. 31, 2005

Copolymerization of Propylene Oxide and Carbon Dioxide
A R T I C L E S
Scheme 2. Copolymerization of Propylene Oxide and CO2 Using Discrete Metal Alkoxides (R ) OR′) and Carboxylates (R ) Alkyl, Aryl)
Table 1. Effect of Air-Free versus Ambient Reaction Conditions: (R,R)-(salcy)CoX Catalyzed Copolymerization of rac-PO/CO2 (X ) I,
OAc)a
theoretical
d
e
reaction
time
(h)
yieldb
(%)
TOFc
Mn
Mn
head-to-tail
linkagesf (%)
1
e
entry
complex
conditions
(h-
)
(kg/mol)
(kg/mol)
Mw/Mn
1
2
3
4
(R,R)-(salcy)CoI
air-free
ambient
air-free
ambient
5
5
2
2
43
37
30
25
43
37
75
62
21.9
18.9
15.3
12.7
19.6
9.5
15.5
10.4
1.15
1.33
1.16
1.31
79
81
83
83
(R,R)-(salcy)CoI
(R,R)-(salcy)CoOAc
(R,R)-(salcy)CoOAc
a Polymerizations run in neat rac-propylene oxide (PO) with [PO]/[Co] ) 500:1 at 22 °C with 800 psi of CO2. Selectivity for poly(propylene carbonate)
(PPC) over propylene carbonate was >99% in all cases. All product PPC contains g96% carbonate linkages as determined by 1H NMR spectroscopy.
b Based on isolated polymer yield. c Turnover frequency (TOF) ) mol PO ‚ mol Co-1 ‚ h-1 d Theoretical number average molecular weight (Mn) ) TOF
.
‚ h ‚ 102 g/mol. e Determined by gel permeation chromatography calibrated with polystyrene standards in THF. f Determined by 13C NMR spectroscopy.
Scheme 3. Bimetallic (salcy)CoX (X ) Nucleophile) Ring Opening
of Epoxides Proposed by Jacobsen and Co-workers
narrow MWD, it is not a necessary condition for the generation
of PPC.
Jacobsen and co-workers have conducted extensive studies
that support a bimetallic mechanism for the chiral (salcy)CoX
Figure 3. (Salcy)CoX catalysts for the copolymerization of propylene oxide
and CO2.
catalyzed kinetic resolution of terminal epoxides (Scheme 3).22
With these systems, they observed a large reaction rate
dependence on the composition of the axial ligand of the catalyst
We initially carried out a comprehensive study of ligand
(X). These results led us to suspect that the nature of the axial
modification with salen-type cobalt complexes in an effort to
ligand would likewise influence the rate of our PO/CO2
investigate the effects of altering the steric and electronic
environment around the active cobalt center. To our surprise,
the ligand that provided the most active catalyst while maintain-
ing high regiocontrol was the original (salcy) ligand. (Salcy)-
CoX complexes were therefore used for further studies (Figure
3). We speculated that the (salcy)CoX systems would produce
PPC with narrow MWDs by eliminating chain transfer agents
such as water. As such, we thoroughly dried all (salcy)CoX
copolymerizations. The polymerization data of complexes (R,R)-
(salcy)CoX with X ) OAc, pentafluorobenzoate (OBzF5), Cl,
Br, and I are collected in Table 2. Although the behavior of
(R,R)-(salcy)CoOAc and (R,R)-(salcy)CoOBzF5 proved largely
similar (entries 1 and 2), use of the various halides ((R,R)-
(salcy)CoCl, (R,R)-(salcy)CoBr, or (R,R)-(salcy)CoI) resulted
in substantial changes in catalytic activity (entries 3-5). The
order of increasing activity for copolymerization with (R,R)-
complexes before use and performed the rac-PO/CO2 copo-
lymerization under air-free conditions. Table 1 compares the
(salcy)CoX complexes is X ) I < Cl < OAc ≈ OBzF5 < Br,
with the most active catalyst, (R,R)-(salcy)CoBr, producing PPC
rac-PO/CO2 copolymerizations carried out under air-free (entries
with a TOF ) 90 h-1 (entry 4). Although catalyst activity varies
1 and 3) and ambient (entries 2 and 4) conditions using (R,R)-
(salcy)CoI and (R,R)-(salcy)CoOAc catalysts. The PPC gener-
ated under air-free conditions exhibits a measured Mn that agrees
with the applied axial ligand/initiator, we speculate that the
propagation rates in all cases are inherently similar. Furthermore,
we attribute the range of catalyst TOFs to an initiation period
more closely with the theoretical value and has a MWD less
that occurs before polymer propagation. Similar to the bimetallic
than 1.2. In all cases, PPC with g96% carbonate linkages and
g79% HT connectivity was achieved with no detectable PC
byproduct. Although the removal of trace water provides for a
(22) Nielsen, L. P. C.; Stevenson, C. P.; Blackmond, D. G.; Jacobsen, E. N. J.
Am. Chem. Soc. 2004, 126, 1360-1362.
9
J. AM. CHEM. SOC. VOL. 127, NO. 31, 2005 10871

A R T I C L E S
Cohen et al.
Table 2. Effect of Initiating Group: (R,R)-(salcy)CoX Catalyzed
Copolymerization of rac-PO/CO2 (X ) OAc, OBzF5, Cl, Br, I)a
a commercially available, readily dried ionic compound with a
bulky cation and a nucleophilic anion. We initially screened
our catalyst library with [PPN]Cl for the rac-PO/CO2 copo-
lymerization at lower CO2 pressures (200 psi) to look for an
increase in reaction rates. We observed a remarkable improve-
ment in catalyst activities, as well as increased stereo- and
regioselectivities when [PPN]Cl was combined with any of the
(R,R)-(salcy)CoX complexes (Table 3).
By itself, [PPN]Cl demonstrated no activity for the copo-
lymerization of rac-PO and CO2, whereas, in combination
with complex (R,R)-(salcy)CoOBzF5, PPC was generated with
an activity of 520 turnovers per hour (Table 3, entry 2). The
product PPC is nearly completely alternating, with >90% HT
connectivity and is iso-enriched with a high Mn and a narrow
MWD. Complexes (R,R)-(salcy)CoCl and (R,R)-(salcy)CoBr
with [PPN]Cl also exhibit outstanding catalytic activity; how-
ever, selectivity for polymer is reduced with the latter (entries
3 and 4). The copolymerization catalyzed by (R,R)-(salcy)-
CoOAc with [PPN]Cl is significantly slower than that catalyzed
by all other catalyst/[PPN]Cl combinations, and the selectivity
for generating PPC over PC is also somewhat compromised
(entry 1).
To optimize the reactivity of this catalyst system further, the
influence of catalyst and cocatalyst ratio, reaction time, and CO2
pressure on the copolymerization was investigated. The optimal
loading for the rac-PO/CO2 copolymerization with (R,R)-(salcy)-
CoOBzF5 is a [PO]:[[PPN]Cl]:[Co] ratio of 2000:1:1 (Table 4,
entries 1 and 2). Under these conditions, PPC is generated in
31% yield after 1 h and in 52% yield after 2 h, corresponding
to catalyst TOFs of 620 h-1 and 520 h-1, respectively, with
only trace PC byproduct. At extended reaction times, however,
the overall selectivity for polymer is decreased notably (entry
3). We observed that, at approximately 50% conversion, the
polymerization solidifies, and the PPC slowly depolymerizes
in situ to form PC. In an effort to maximize polymer yield, it
is crucial to quench the copolymerization upon reaching
approximately 50% conversion. To achieve PPC at higher %
conversions, the addition of solvent is required, which is a
subject under current investigation. Increasing the loading of
[PPN]Cl to 2 equiv while maintaining 1 equiv of (R,R)-(salcy)-
CoOBzF5 has no noticeable effect on the copolymerization
(entry 4), whereas decreasing the [PPN]Cl content to 0.5 equiv
results in a substantial loss in activity (entry 5). Overall,
decreasing the amount of [PPN]Cl in the copolymerization while
maintaining the same amount of catalyst and the same percent
conversion increases the Mn of the product PPC. In great contrast
to the performance of complex (R,R)-(salcy)CoOBzF5 alone,
the ion-assisted copolymerization is optimized at lower pressures
of CO2 (200 psi), suggesting the possibility of an alternate
d
yieldb TOFc
Mn
head-to-tail
1
d
entry
complex
(%)
(h-
)
(kg/mol) Mw/Mn linkagese (%)
1
2
3
4
5
6
(R,R)-(salcy)CoOAc
(R,R)-(salcy)CoOBzF5
(R,R)-(salcy)CoCl
(R,R)-(salcy)CoBr
(R,R)-(salcy)CoI
(R,R)-(salcy)CoI +
(R,R)-(salcy)CoBr (50:1)
30 75
32 80
26 65
36 90
13 33
28 70
15.5 1.16
14.1 1.22
13.4 1.19
21.0 1.14
10.4 1.17
16.2 1.24
83
82
82
82
85
81
a Polymerizations run in neat rac-propylene oxide (PO) with [PO]/[Co]
) 500:1 at 22 °C with 800 psi of CO2 for 2 h. Selectivity for poly(propylene
carbonate) (PPC) over propylene carbonate was >99% in all cases. All
product PPC contains g92% carbonate linkages as determined by 1H NMR
spectroscopy. b Based on isolated polymer yield. c Turnover frequency
d
(TOF) ) mol PO ‚ mol Co-1 ‚ h-1
.
Determined by gel permeation
chromatography calibrated with polystyrene standards in THF. e Determined
by 13C NMR spectroscopy. [OBzF5] ) pentafluorobenzoate.
mechanism proposed by Jacobsen and co-workers, we suspect
that the rate in which the initiator is delivered from one cobalt
center to ring open a PO bound to a second cobalt center
depends on the Lewis acidity of the catalyst and the nucleo-
philicity of the axial ligand. Despite these initiation rate
differences, all product PPCs are atactic with g92% carbonate
linkages, g82% HT connectivity, and narrow MWDs. In
addition, no detectable PC is observed in all systems.
Because the copolymerization rate exhibits a pronounced
dependence on the nature of the axial ligand, we anticipated
that through combining the slowest catalyst, (R,R)-(salcy)CoI,
with the more active (R,R)-(salcy)CoBr, its copolymerization
activity would improve. This prediction was indeed correct: as
little as 2% of (R,R)-(salcy)CoBr with catalyst (R,R)-(salcy)-
CoI achieved a TOF of 70 h-1 while maintaining an overall
[PO]:[Co] loading of 500:1 (Table 2, entry 6). This copolym-
erization rate is more than double that of the unassisted
(R,R)-(salcy)CoI catalyzed reaction, even though [Br-]:[I-] is
1:50. PPC produced with the mixed catalyst system maintains
92% carbonate linkages and 81% HT connectivity although there
is slight MWD broadening.
Alternating Copolymerization of Propylene Oxide and
Carbon Dioxide with (salcy)CoX Catalysts and Ionic Co-
catalysts. Previous reports have described improved catalytic
activities upon the coupling of organic salt cocatalysts with
porphyrin or salen-type catalysts for epoxide/CO2 copolymer-
ization.5,7,8,17 Recently, Lu and co-workers described outstanding
rates for the copolymerization of PO and CO2 through the
addition of quaternary ammonium salt cocatalysts to (salcy)-
CoX catalysts at low CO2 pressures.20 With the goal of
increasing the catalytic activities of our systems, we indepen-
dently explored the use of [PPN]Cl as a cocatalyst. [PPN]Cl is
Table 3. Effect of Initiating Group: (R,R)-(salcy)CoX Catalyzed Copolymerization of rac-PO/CO2 with Cocatalyst [PPN]Cl (X ) OAc, OBzF5,
Cl, Br)a
e
yieldb
(%)
TOFc
selectivityd
(% PPC)
Mn
head-to-tail
linkagesf (%)
1
e
entry
complex
(h-
)
(kg/mol)
Mw/Mn
1
2
3
4
(R,R)-(salcy)CoOAc
(R,R)-(salcy)CoOBzF5
(R,R)-(salcy)CoCl
(R,R)-(salcy)CoBr
11
52
43
46
110
520
430
460
86
>99
>99
89
47.9
43.0
35.4
33.2
1.15
1.10
1.09
1.09
93
93
95
95
a
Polymerizations run in neat rac-propylene oxide (PO) with [PO]:[[PPN]Cl]:[Co] ) 2000:1:1 at 22 °C with 200 psi of CO2 for 2 h. All product
poly(propylene carbonate) (PPC) contains g98% carbonate linkages as determined by 1H NMR spectroscopy. Based on isolated polymer yield. c Turnover
b
frequency ) mol PO ‚ mol Co-1 ‚ h-1
.
d Selectivity for PPC over propylene carbonate. Determined by gel permeation chromatography calibrated with
e
polystyrene standards in THF. f Determined by 13C NMR spectroscopy. [PPN] ) bis(triphenylphosphine)iminium. [OBzF5] ) pentafluorobenzoate.
9
10872 J. AM. CHEM. SOC. VOL. 127, NO. 31, 2005

Copolymerization of Propylene Oxide and Carbon Dioxide
A R T I C L E S
Table 4. Effect of Cocatalyst [PPN]Cl Concentration and Reaction Time: (R,R)-(salcy)CoOBzF5 Catalyzed Copolymerization of
a
rac-PO/CO2
e
time
‘(h)
yieldb
(%)
TOFc
selectivityd
(% PPC)
Mn
head-to-tail
linkagesf (%)
1
e
entry
[PO]:[[PPN]Cl]:[Co]
(h-
)
(kg/mol)
Mw/Mn
1
2
3
4
5
2000:1:1
2000:1:1
2000:1:1
2000:2:1
2000:0.5:1
1
2
6
2
2
31
52
59
53
36
620
520
200
530
360
99
>99
56
97
>99
26.8
43.0
41.4
33.9
46.3
1.13
1.10
1.36
1.08
1.07
94
93
93
93
94
a Polymerizations run in neat rac-propylene oxide (PO) at 22 °C with 200 psi of CO2. All product poly(propylene carbonate) (PPC) contains g98%
carbonate linkages as determined by 1H NMR spectroscopy. b Based on isolated polymer yield. c Turnover frequency ) mol PO ‚ mol Co-1 ‚ h-1 d Selectivity
.
for PPC over propylene carbonate. e Determined by gel permeation chromatography calibrated with polystyrene standards in THF. f Determined by 13C
NMR spectroscopy. [PPN] ) bis(triphenylphosphine)iminium. [OBzF5] ) pentafluorobenzoate.
Table 5. Effect of Cocatalyst and Initiator: (R,R)-(salcy)CoX Catalyzed Copolymerization of rac-PO/CO2 (X ) OBzF5, Cl)a
e
yieldb
(%)
TOFc
selectivityd
(% PPC)
Mn
head-to-tail
linkagesf (%)
1
e
entry
complex
cocatalyst
(h-
)
(kg/mol)
Mw/Mn
1
2
3
4
5
(R,R)-(salcy)CoOBzF5
(R,R)-(salcy)CoOBzF5
(R,R)-(salcy)CoOBzF5
(R,R)-(salcy)CoOBzF5
(R,R)-(salcy)CoCl
[PPN]Cl
[n-Bu4N]Cl
[PPN][OBzF5]
[PPN][BPh4]
[PPN][OBzF5]
52
18
48
0
520
180
480
0
>99
>99
>99
NA
99
43.0
7.6
51.9
NA
1.10
1.13
1.14
NA
93
94
94
NA
95
46
460
49.0
1.07
a Polymerizations run in neat rac-propylene oxide (PO) with [PO]:[cocatalyst]:[Co] ) 2000:1:1 at 22 °C with 200 psi of CO2 for 2 h. All product
poly(propylene carbonate) (PPC) contains g98% carbonate linkages as determined by 1H NMR spectroscopy. b Based on isolated polymer yield. c Turnover
frequency ) mol PO ‚ mol Co-1 ‚ h-1 d Selectivity for PPC over propylene carbonate. e Determined by gel permeation chromatography calibrated with
.
polystyrene standards in THF. f Determined by 13C NMR spectroscopy. [PPN] ) bis(triphenylphosphine)iminium. [OBzF5] ) pentafluorobenzoate.
mechanism. We suspect that the ionic cocatalyst helps to
stabilize the active species against decomposition to (R,R)-
(salcy)CoII, a reaction that is observed with these catalysts alone
at low CO2 pressures.23 Indeed, with the incorporation of [PPN]-
Cl, PPC can be generated at pressures as low as 50 psi with
only slightly compromising activity and selectivity for polymer.
By altering the composition of the ionic cocatalyst, we
obtained information about how the cation and anion influence
the rac-PO/CO2 copolymerization. Replacement of [PPN]Cl
with [n-Bu4N]Cl results in a notable decrease in activity (Table
5, entries 1 and 2). PPC generated by (R,R)-(salcy)CoOBzF5
with [n-Bu4N]Cl has a low Mn attributable to chain transfer to
water introduced by the ammonium salt. Substitution of [PPN]-
[OBzF5] for [PPN]Cl with complex (R,R)-(salcy)CoOBzF5 has
little effect on the reaction rate (entry 3), but the reaction with
the salt [PPN][BPh4] shows a complete loss in activity (entry
4). In general, a bulky noncoordinating cation and a nucleophilic
anion capable of initiation compose the ideal cocatalyst for the
(R,R)-(salcy)CoOBzF5 catalyzed copolymerization; [PPN]Cl and
[PPN][OBzF5] are most beneficial.
The Mn values of the polymers obtained from the ion-assisted
copolymerizations are reproducible and approach 50 kg/mol,
yet they are equal to approximately half of the theoretical values
based on conversion. We suspect that initiation is not limited
to the original initiating group of the cobalt catalyst but also
involves the anion from the [PPN]+ cocatalyst. To demonstrate
this possibility, PPC was prepared with two different catalyst/
cocatalyst combinations: (R,R)-(salcy)CoOBzF5/[PPN]Cl and
(R,R)-(salcy)CoCl/[PPN][OBzF5]. The isolated polymer from
each system exhibited an Mn of approximately 10 kg/mol with
a narrow MWD and showed UV activity by the gel permeation
chromatography (GPC) UV detector. In addition, the 19F NMR
spectrum of each polymer revealed that OBzF5 end groups were
present in both samples. Each spectrum exhibited 19F resonances
at δ -138 (m), -149 (m), -161 (m) ppm, consistent with a
OBzF5 moiety on the PPC. These data suggest that initiation
can occur from the catalyst precursor as well as the salt
cocatalyst. Furthermore, as there are two feasible initiators per
catalyst and product PPCs exhibit Mn values approximately half
those of the theoretical values with narrow MWDs, we
hypothesize that each cobalt center propagates two polymer
chains.
The addition of a Lewis base cocatalyst to epoxide/CO2
copolymerizations with porphyrin or salen-type metal catalysts
has provided for enhanced catalytic activities in a variety of
systems.5,6,9,10,12,17,24 In several of these cases a proposed
mechanism illustrates that the Lewis base coordinates to the
metal center in the axial site, trans to the propagating species,
resulting in an improved electronic environment for propagation.
In agreement with these schemes, we consider that the axial
ligand trans to the propagating species has a large influence on
the propagation rate. As we suspect that the anion of the ionic
cocatalyst in our systems is capable of initiation, we hypothesize
that two polymer chains can simultaneously propagate from
either side of the Co-salcy plane (Scheme 4), a mechanism
similar to that proposed by Inoue and co-workers using
aluminum porphyrins and ammonium and phosphonium salts.17
Influence of Catalyst and Propylene Oxide Stereochem-
istry on Catalytic Activity and Polymer Microstructure. We
previously reported that catalyst (R,R)-(salcy)CoOAc preferen-
tially enchains (S)-PO over (R)-PO (krel ) 2.8) in rac-PO/CO2
copolymerization.16 Also, substitution of (S)-PO for rac-PO
increases the HT connectivity of the product PPC from 80% to
93%. These initial experiments indicate that the chiralities of
the monomer and the catalyst strongly influence the polymer-
ization activities and the product PPC regiochemistry. Further-
more, the relationship among the catalyst and PO stereochem-
(23) When the PO/CO2 copolymerization is carried out at low CO2 pressures
(<100 psi) in a Fischer-Porter bottle, the reduction of (R,R)-(salcy)CoIII
X
to (R,R)-(salcy)CoII can be observed as a red solid (CoII) precipitates from
the reaction. This reaction is also thought to occur at higher CO2 pressures;
however the CO2 must be vented from the Parr reactor before we are able
to look at the reaction mixture, preventing direct observation.
9
J. AM. CHEM. SOC. VOL. 127, NO. 31, 2005 10873

A R T I C L E S
Cohen et al.
Table 6. (R,R)- and rac-(salcy)CoBr Catalyzed Copolymerization of PO/CO2 Using rac-, (S)-, and (R)-POa
d
yieldb
(%)
TOFc
Mn
head-to-tail
linkagese (%)
1
d
entry
complex
epoxide
(h-
)
(kg/mol)
Mw/Mn
1
2
3
4
5
(R,R)-(salcy)CoBr
(R,R)-(salcy)CoBr
(R,R)-(salcy)CoBr
rac-(salcy)CoBr
rac-(salcy)CoBr
rac-PO
(S)-PO
(R)-PO
(R)-PO
rac-PO
36
49
20
42
41
90
120
50
110
100
21.0
20.1
13.3
26.9
21.8
1.14
1.21
1.16
1.16
1.22
82
93
43
83
84
a Polymerizations run in neat propylene oxide (PO) with [PO]/[Co] ) 500:1 at 22 °C with 800 psi of CO2 for 2 h. Selectivity for poly(propylene carbonate)
(PPC) over propylene carbonate was >99% in all cases. All product PPC contains g92% carbonate linkages as determined by 1H NMR spectroscopy.
b Based on isolated polymer yield. c Turnover frequency (TOF) ) mol PO ‚ mol Co-1 ‚ h-1 d Determined by gel permeation chromatography calibrated
.
with polystyrene standards in THF. e Determined by 13C NMR spectroscopy.
Scheme 4. Proposed Scheme for the Copolymerization of
Propylene Oxide and CO2 with Catalyst (salcy)CoX and Cocatalyst
[PPN]Y; [PPN] ) Bis(triphenylphosphine)iminium, [OBzF5] )
Pentafluorobenzoate
istry and the resultant PPC microstructure provides mechanistic
insight.25 As such, we were eager to pursue this topic with the
(salcy)CoBr and (salcy)CoOBzF5/[PPN]Cl catalyst systems.
Figure 4. Carbonyl region of the 13C NMR spectra (125 MHz, CDCl3) of
poly(propylene carbonate)s generated from the copolymerization of a) rac-
propylene oxide (PO)/CO2, b) (S)-PO/CO2, and c) (R)-PO/CO2 with catalyst
(R,R)-(salcy)CoBr. TT ) tail-to-tail; HT ) head-to-tail; HH ) head-to-
head.
Table 6 lists the copolymerization data for (R,R)-(salcy)CoBr
and rac-(salcy)CoBr with rac-, (S)-, and (R)-PO. As stated
above, the copolymerization of rac-PO and CO2 with catalyst
(R,R)-(salcy)CoBr produces atactic PPC with 82% HT con-
nectivity (entry 1, Figure 4a, and Scheme 5a). Under the same
reaction conditions, but replacing rac-PO with (S)-PO, the
catalytic activity increases from 90 to 120 turnovers per hour,
and the resultant PPC is isotactic with 93% HT connectivity
(entry 2, Figure 4b, and Scheme 5b). This result is consistent
with our earlier work16 and indicates that the (R,R) catalyst
makes fewer regioerrors when only (S)-PO is present. Degrada-
tion of this polymer to PC while conserving all stereocenters14
yielded an (S)-PC:(R)-PC ratio of 97:3 as determined by gas
chromatography (GC). The high (S)-content of the PPC suggests
that transcarbonation and PC reinsertion mechanisms do not
readily occur and that the PO ring opening process predomi-
nately takes place at the methylene or methine carbon with
stereochemical retention. Chisholm and co-workers have previ-
ously shown the possibility of PO ring opening at the methine
carbon with stereochemical retention in related chromium and
aluminum catalyst systems.24 Although a consistent methine (S)-
PO ring opening process with stereochemical retention would
result in PPC with a high (S) content, we believe that the
preferred reaction pathway in our systems is the sterically
favored attack at the methylene carbon.
When (R)-PO is used in place of rac-PO, catalytic activity
decreases to 50 turnovers per hour and the resultant PPC is
almost perfectly regiorandom with only 43% HT connectivity
(Table 6, entry 3, Figure 4c, and Scheme 5c). In this case,
degradation of the polymer to PC yields a (S)-PC:(R)-PC ratio
of 34:66. Assuming PO methine attack inverts stereochemistry,
while methylene attack retains stereochemistry, a HH or TT
linkage in the polymer requires one (S)- and one (R)-stereo-
center. Furthermore, the excess of (R)-PC illustrates that the
HT connectivity of the parent PPC is composed of predomi-
nately (R)-stereocenters. This indicates that generation of an
HH linkage through misinsertion is most often immediately
corrected to generate a TT linkage, followed by the incorporation
of a couple of (R)-HT linkages until the next regioerror occurs.
We proceeded to replace the (R,R)-(salcy)CoBr complex with
rac-(salcy)CoBr in order to obtain further mechanistic insight
concerning propagation with these systems. The copolymeri-
zation of (R)-PO and CO2 with rac-(salcy)CoBr yielded isotactic
PPC with 83% HT connectivity with a catalytic activity of 110
turnovers per hour (Table 6, entry 4, and Scheme 5d). In this
case, we expected two polymers with widely differing Mn values
due to the predicted distinction in activity between the (S,S)-
and (R,R)-(salcy)CoBr catalysts with (R)-PO/CO2. Interestingly,
the product PPC showed a monomodal GPC trace with a MWD
(24) Chisholm, M. H.; Zhou, Z. P. J. Am. Chem. Soc. 2004, 126, 11030-11039.
(25) Byrnes, M. J.; Chisholm, M. H.; Hadad, C. M.; Zhou, Z. P. Macromolecules
2004, 37, 4139-4145.
9
10874 J. AM. CHEM. SOC. VOL. 127, NO. 31, 2005

Copolymerization of Propylene Oxide and Carbon Dioxide
A R T I C L E S
Scheme 5. Effect of Propylene Oxide and (salcy)CoBr Stereochemistry on Poly(propylene carbonate) Microstructure
of 1.16, suggesting that each propagating species can dissociate
from the metal center during the copolymerization and is
influenced by catalysts of both possible chiralities. The degrada-
tion PC product from this polymer has a (S)-PC:(R)-PC ratio
of 9:91, again illustrating that regioerrors are immediately
corrected.
Finally, we investigated the rac-PO/CO2 copolymerization
with catalyst rac-(salcy)CoBr. To our surprise, the product PPC,
with 84% HT connectivity, has a 13C NMR spectrum unlike
that of all other PPC samples (Table 6, entry 5, Figure 5, and
Scheme 5e).26 In this case, the carbonyl resonances correspond-
ing to the four HT triads ([mm], [mr], [rm], and [rr]) did not
show an equal distribution. We initially assigned the most
upfield HT shift in this spectrum as the [mm] triad (Figure 5)
based on the observation that the 13C NMR HT resonance of
isotactic PPC aligns with the most upfield HT resonance of
atactic PPC (Figure 4a,b). Furthermore, the [mr] and [rm] HT
linkages should integrate equally yet show different resonances
due to the directionality in the PPC. We therefore tentatively
assigned the central HT resonances to the [mr]/[rm] triads and
the remaining [rr] triad to the most downfield HT resonance,
while considering that most tactic polymer systems exhibit [mm]
and [rr] resonances that encompass the [mr]/[rm] 13C NMR
resonances.27,28 Based on this scheme, we assigned the most
downfield and upfield resonances in both the TT and HH
portions of the spectrum to [r] and [m] diads, respectively, a
simplification of the complex triad sequences from the many
possible combinations of neighboring stereo- and regiose-
quences. It is also notable that the TT [m] and [r], HT [m] and
[r], and HH [m] and [r] linkages of PPC as well as the HH
[mmm] and [rrr] linkages of oligoether carbonates have been
previously assigned by Chisholm and co-workers24,25 and are
in agreement with our PPC conformational assignments dis-
cussed above.
Based on our 13C NMR assignments, the HT portion of the
PPC generated from rac-PO/CO2 with catalyst rac-(salcy)CoBr
is syndio-enriched, an unreported microstructure for this polymer
system. As the regioregularity in the PPC implicates that the
PO ring-opening event occurs consistently at the less substituted
carbon, we suspect that the PPC syndiotacticity is achieved
through a chain-end control mechanism, with alternate enchain-
ment of (S)- and (R)-PO.29 Specifically, assuming that the
enchainment of (S)- over (R)-PO is favored for catalyst (R,R)-
(26) A 30° shifted squared sinusoidal window function was applied prior to
Fourier transformation in order to enhance resolution.
Figure 5. Carbonyl region of the 13C NMR spectrum (150 MHz, CDCl3)
of poly(propylene carbonate) generated from the copolymerization of rac-
propylene oxide/CO2 with catalyst rac-(salcy)CoBr. TT ) tail-to-tail; HT
) head-to-tail; HH ) head-to-head.
(27) Resconi, L.; Cavallo, L.; Fait, A.; Piemontesi, F. Chem. ReV. 2000, 100,
1253-1345.
(28) Schilling, F. C.; Tonelli, A. E. Macromolecules 1986, 19, 1337-1343.
(29) Ovitt, T. M.; Coates, G. W. J. Am. Chem. Soc. 2002, 124, 1316-1326.
9
J. AM. CHEM. SOC. VOL. 127, NO. 31, 2005 10875

A R T I C L E S
Cohen et al.
Table 7. (R,R)- and rac-(salcy)CoOBzF5 Catalyzed Copolymerization of PO/CO2 with Cocatalyst [PPN]Cl Using rac-, (S)-, and (R)-POa
e
time
(h)
yieldb
(%)
TOFc
selectivityd
(%PPC)
Mn
head-to-tail
linkages (%)
1
e
entry
complex
epoxide
(h-
)
(kg/mol)
Mw/Mn
1
2
3
4
5
(R,R)-(salcy)CoOBzF5
(R,R)-(salcy)CoOBzF5
(R,R)-(salcy)CoOBzF5
rac-(salcy)CoOBzF5
rac-(salcy)CoOBzF5
rac-PO
(S)-PO
(R)-PO
(R)-PO
rac-PO
1
0.5
2
1
1
31
27
21
37
33
620
1100
210
740
660
99
>99
98
>99
>99
26.8
22.2
19.1
34.5
32.7
1.13
1.15
1.12
1.07
1.10
94
96
87
95
95
a
Polymerizations run in neat propylene oxide (PO) with [PO]:[[PPN]Cl]:[Co] ) 2000:1:1 at 22 °C with 200 psi of CO2. All product poly(propylene
carbonate) (PPC) contains g98% carbonate linkages as determined by 1H NMR spectroscopy. b Based on isolated polymer yield. c Turnover frequency
(TOF) ) mol PO ‚ mol Co-1 ‚ h-1. Selectivity for PPC over propylene carbonate. e Determined by gel permeation chromatography calibrated with polystyrene
standards in THF. f Determined by 13C NMR spectroscopy. [PPN] ) bis(triphenylphosphine)iminium. [OBzF5] ) pentafluorobenzoate.
propagating species can dissociate from the metal center during
the copolymerization and is influenced by catalysts of both
possible chiralities.
Using catalyst rac-(salcy)CoOBzF5 with [PPN]Cl for the
copolymerization of rac-PO and CO2 yields atactic PPC with
95% HT connectivity (Table 7, entry 5). Unlike the correspond-
ing results using rac-(salcy)CoBr alone (Table 6, entry 5, and
Figure 5), the PPC microstructures generated by rac-(salcy)-
CoOBzF5/[PPN]Cl or (R,R)-(salcy)CoOBzF5/[PPN]Cl catalyst
systems are essentially the same for the rac-PO/CO2 copolym-
erization. Our experiments with catalysts (salcy)CoX for the
PO/CO2 copolymerization have led us to the following mecha-
nistic assertions. The influence of the initiator on the copolym-
Figure 6. Carbonyl region of the 13C NMR spectrum (125 MHz, CDCl3)
of poly(propylene carbonate) generated from the copolymerization of rac-
propylene oxide/CO2 with catalyst (R,R)-(salcy)CoOBzF5/[PPN]Cl. TT )
tail-to-tail; HT ) head-to-tail; HH ) head-to-head.
erization rate and the effect of combining different (salcy)CoX
catalysts, where a fast initiator can assist a slower one, support
a bimetallic initiation similar to that proposed by Jacobsen and
co-workers for the hydrolytic kinetic resolution of epoxides.22
Additionally, the relationship of the PO and catalyst stereo-
chemistry to the resultant PPC microstructure is consistent with
a scheme in which the propagating species can dissociate from
the metal center during the copolymerization.
(salcy)CoBr and that the carboxylate generated from subsequent
CO2 insertion can migrate between cobalt centers, an (R,R)-
(salcy)CoBr catalyst will ring open an (S)-PO and insert CO2
and then the propagating carboxylate will migrate to an (S,S)-
(salcy)CoBr catalyst before the next PO insertion occurs.
The mechanistic insight obtained from the stereochemical
experiments with (salcy)CoBr led us to pursue similar studies
with the (salcy)CoOBzF5/[PPN]Cl catalyst systems. Table 7
shows the copolymerization data for (R,R)-(salcy)CoOBzF5/
[PPN]Cl and rac-(salcy)CoOBzF5/[PPN]Cl with rac-, (S)-, and
(R)-PO/CO2. Remarkably, the addition of [PPN]Cl to the (R,R)-
(salcy)CoOBzF5 catalyzed rac-PO/CO2 copolymerization not
only increases the catalytic activity but also improves the stereo-
and regioselectivity of the catalyst. Specifically, (R,R)-(salcy)-
CoOBzF5/[PPN]Cl demonstrates an activity of 620 turnovers
per hour for the copolymerization of rac-PO and CO2, affording
iso-enriched PPC with 94% HT connectivity (entry 1, and Figure
6). The substitution of (S)-PO for rac-PO in this copolymeri-
zation results in a near doubling of the reaction rate (TOF )
1100 h-1), affording highly alternating, isotactic PPC with 96%
HT connectivity (entry 2). Alternatively, the substitution of (R)-
PO for rac-PO in the copolymerization results in substantial
rate inhibition in which a TOF of only 210 h-1 is achieved (entry
3). The product PPC from this reaction is isotactic and has 87%
HT connectivity, showing that the regioselectivity of this system
is less dependent on the relative stereochemistry of PO and
catalyst than was observed with (salcy)CoBr.
Our working hypothesis for the PO/CO2 copolymerization
with (salcy)CoX and ionic cocatalysts stems from the observa-
tion that the (salcy)CoX initiator and the cocatalyst anion can
be found as end groups on the resultant PPC. This suggests
that initiation can occur with either of these species, which is
further corroborated by the measured Mn data, which are
approximately half those of the corresponding theoretical values
based on conversion. As previously discussed, we propose a
mechanism in which propagation can occur simultaneously on
either side of the Co-salcy plane (Scheme 4). In addition, our
results support that the propagating species can dissociate from
the metal center during the copolymerization, a mechanism
similar to that of the (salcy)CoBr catalyzed copolymerization
in the absence of cocatalyst.
Conclusions
We have investigated a series of cobalt-based (salcy) catalysts
for the copolymerization of PO and CO2. Through adjustment
of the reaction environment and catalyst optimization we were
able to maximize catalytic activity and selectivity for the
generation of highly alternating, regioregular PPC with con-
trolled molecular weight and no detectable PC byproduct.
Through supplementation of the cobalt catalysts with [PPN]Cl
or [PPN][OBzF5] cocatalysts, we effected dramatic changes in
polymerization behavior, catalytic activity, and regioselectivity.
The PO/CO2 copolymerization with (salcy)CoX (X ) Cl,
OBzF5) and [PPN]Cl or [PPN][OBzF5] effected increased
catalytic activities, enhanced stereo- and regioselectivities, with
The copolymerization of (R)-PO/CO2 with catalyst rac-
(salcy)CoOBzF5 and [PPN]Cl yields isotactic PPC with 95%
HT connectivity (Table 7, entry 4). This polymer shows a
monomodal GPC trace with an MWD of 1.07. Similar to that
discussed above, this result supports a mechanism in which each
9
10876 J. AM. CHEM. SOC. VOL. 127, NO. 31, 2005

Copolymerization of Propylene Oxide and Carbon Dioxide
A R T I C L E S
(R,R)-(salcy)CoCl. This complex was prepared as previously
described.22 Additional characterization: 13C NMR (DMSO-d6, 125
MHz): δ 24.34, 29.51, 30.40, 31.56, 33.51, 35.78, 69.27, 118.58,
128.78, 129.28, 135.86, 141.84, 162.08, 164.68.
maximized reactivity at low CO2 pressures. The copolymeri-
zation of rac-PO and CO2 with (R,R)-(salcy)CoOBzF5 and
[PPN]Cl yielded a TOF of 620 h-1 for iso-enriched PPC with
94% HT connectivity. By varying catalyst and PO stereochem-
istry, we observed pronounced alterations in the resultant PPC
microstructure, as well as changes in catalytic activity. The
copolymerization of rac-PO/CO2 with catalyst rac-(salcy)CoBr
yielded syndio-enriched PPC, an unreported microstructure for
this polymer. The copolymerization of (S)-PO and CO2 with
catalyst (R,R)-(salcy)CoOBzF5 and cocatalyst [PPN]Cl displayed
a TOF ) 1100 h-1 for isotactic PPC. To our knowledge, these
catalyst systems exhibit the highest activities for the copolym-
erization of PO and CO2, as well as maintain excellent selectivity
for polymer with regiocontrol and living behavior. Given the
overall success with these systems for the synthesis of PPC,
our current focus is the study of (salcy)CoX catalyzed copo-
lymerizations in greater mechanistic detail.
(R,R)-(salcy)CoBr and rac-(salcy)CoBr. The procedure for the
synthesis of (R,R)-(salcy)CoCl published by Jacobsen and co-workers22
was applied to the synthesis of complexes (R,R)-(salcy)CoBr and rac-
(salcy)CoBr, with the substitution of NaBr for NaCl. 1H NMR (DMSO-
d6, 500 MHz): δ 1.30 (s, 18H), 1.58 (m, 2H), 1.74 (s, 18H), 1.92 (m,
4
2H), 2.00 (m, 2H), 3.06 (m, 2H), 3.59 (m, 2H), 7.44 (d, J ) 3.0 Hz,
4
2H), 7.47 (d, J ) 3.0 Hz, 2H), 7.83 (s, 2H). 13C NMR (DMSO-d6,
125 MHz): δ 24.32, 29.57, 30.43, 31.55, 33.58, 35.82, 69.32, 118.61,
128.78, 129.28, 135.87, 141.84, 162.11, 164.66. Anal. Calcd for
C36H52N2O2CoBr: C, 63.25; H, 7.67; N, 4.10. Found: C, 63.05; H,
7.69; N, 4.06.
(R,R)-(salcy)CoI. The procedure for the synthesis of (R,R)-(salcy)-
CoCl published by Jacobsen and co-workers22 was applied to the
synthesis of complex (R,R)-(salcy)CoI, with the substitution of NaI for
NaCl. 1H NMR (DMSO-d6, 500 MHz): δ 1.32 (s, 18H) 1.63 (m, 2H),
1.76 (s, 18H), 1.91 (m, 2H), 2.02 (m, 2H), 3.10 (m, 2H), 3.66 (m, 2H),
Experimental Section
General Procedure. All air or water sensitive reactions were carried
out under dry nitrogen using a Braun Labmaster drybox or standard
Schlenk-line techniques. Methylene chloride and diethyl ether were
dried and degassed by passing through a column of activated alumina
and by sparging with dry nitrogen. (S,S)- and (R,R)-N,N′-Bis(3,5-di-
tert-butylsalicylidene)-1,2-diaminocyclohexanecobalt ((S,S)-(salcy)CoII,
(R,R)-(salcy)CoII) were purchased from Aldrich and recrystallized from
methylene chloride and methanol. Bis(triphenylphosphine)iminium
chloride ([PPN]Cl) was purchased from Strem and recrystallized from
dry methylene chloride and diethyl ether under nitrogen before use.
Bis(triphenylphosphine)iminium tetraphenylborate ([PPN][BPh4]) was
prepared following literature procedure.30,31 PO was dried over calcium
hydride and vacuum transferred before use. CO2 (99.998% purity) was
purchased from Airgas and passed over a column of activated 4 Å
molecular sieves. All other reagents were purchased from commercial
sources and used as received. Degradation of PPC to PC was performed
according to literature procedure,14 and the product PC was analyzed
by GC. Gas chromatograms were obtained on a Hewlett-Packard 6890
series gas chromatograph using a beta-DEX 225 chiral capillary column
(30.0 m × 250 µm × 0.25 µm nominal), a flame ionization detector,
and He carrier gas. Varian Mercury (300 MHz), Varian Inova (500
MHz), and Varian Inova (600 MHz) spectrometers were used to record
13C and 1H NMR spectra, which were referenced versus residual
nondeuterated solvent shifts. C6F6 (-162.90 ppm) was used as a
reference for all 19F NMR spectra. GPC analyses were carried out using
a Waters instrument (M515 pump, U6K injector) equipped with a
Waters UV486 and Waters 2410 differential refractive index detector
and four 5 µm PL Gel columns (Polymer Laboratories; 100 Å, 500 Å,
1000 Å, and Mixed C porosities) in series. The GPC columns were
eluted with THF at 40 °C at 1 mL/min and were calibrated using 23
monodisperse polystyrene standards. Elemental analyses were carried
out by Robertson Microlit Laboratories in Madison, NJ.
7.45 (d, 4J ) 2.5 Hz, 2H), 7.50 (d, 4J ) 2.5 Hz, 2H), 7.83 (s, 2H). 13
C
NMR (DMSO-d6, 125 MHz): δ 24.23, 29.54, 30.36, 31.49, 33.47,
35.71, 69.22, 118.59, 128.63, 129.16, 135.82, 141.74, 161.95, 164.49.
Anal. Calcd for C36H52N2O2CoI: C, 59.18; H, 7.17; N, 3.83. Found:
C, 59.14; H, 7.05; N, 3.75.
(R,R)-(salcy)CoOBzF5 and rac-(salcy)CoOBzF5. Recrystallized
(R,R)-(salcy)CoII or rac-(salcy)CoII (1.2 g, 2.0 mmol) and pentafluo-
robenzoic acid (0.42 g, 2.0 mmol) were added to a 50 mL round-
bottomed flask charged with a Teflon stir bar. Toluene (20 mL) was
added to the reaction mixture, and it was stirred open to air at 22 °C
for 12 h. The solvent was removed by rotary evaporation at 22 °C,
and the solid was suspended in 200 mL of pentane and filtered. The
dark green crude material was dried in vacuo and collected in
1
quantitative yield. H NMR (DMSO-d6, 500 MHz): δ 1.30 (s, 18H),
1.59 (m, 2H), 1.74 (s, 18H), 1.90 (m, 2H), 2.00 (m, 2H), 3.07 (m, 2H),
4
4
3.60 (m, 2H), 7.44 (d, J ) 2.5 Hz, 2H), 7.47 (d, J ) 3.0 Hz, 2H),
7.81 (s, 2H).13C NMR (DMSO-d6, 125 MHz): δ 24.39, 29.61, 30.13,
30.42, 31.55, 33.57, 35.83, 69.38, 118.59, 128.78, 129.29, 135.86,
141.83, 162.21, 164.66. Carbons on the phenyl group of pentafluo-
robenzoate were not assigned in the 13C NMR spectrum owing to
complex carbon fluorine splitting patterns.19F NMR (470 MHz, DMSO-
d6): δ -163.32 (m), -162.50 (m), -144.48 (m). Anal. Calcd for
C43H52O4N2F5Co‚H2O: C, 62.01; H, 6.54; N, 3.36. Found: C, 62.25;
H, 6.38; N, 3.42.
[PPN][OBzF5]. NaOH (0.19 g, 4.7 mmol) and pentafluorobenzoic
acid (1.0 g, 4.7 mmol) were added to a 50 mL round-bottomed flask
charged with a Teflon stir bar. Distilled H2O (20 mL) was added to
the reaction mixture, and it was stirred until all was dissolved. The
solution was added to a 250 mL separatory funnel along with [PPN]Cl
(0.40 g, 0.70 mmol) and methylene chloride (40 mL), and the mixture
was shaken vigorously for 10 min. The organic layer was collected
and dried by rotary evaporation to yield crude [PPN][OBzF5] in
quantitative yield. Precipitation from dry methylene chloride and diethyl
Complex Synthesis. Oxidation of the commercially available (R,R)-
(salcy)CoII readily occurs in the presence of a variety of acids to yield
(R,R)-(salcy)CoOAc,15 (R,R)-(salcy)CoOTs (OTs ) tosylate),22 and
(R,R)-(salcy)CoOBzF5. Complex (R,R)-(salcy)CoOTs can be further
modified through metathesis reactions with the desired NaX salt,
affording (R,R)-(salcy)CoX (X ) Cl, Br, I).22 Alternatively, rac-(salcy)-
CoX catalysts are prepared using the same procedures listed above but
starting with a 50:50 mixture of (R,R)-(salcy)CoII and (S,S)-(salcy)-
CoII.
1
ether under N2 at -20 °C afforded a white powder (0.35 g, 67%). H
NMR (CDCl3, 500 MHz): δ 7.39-7.46 (m, 24H), 7.60-7.63 (m,
6H).13C NMR (CDCl3, 125 MHz): δ 116.93, 126.91 (dd, 1JP-C ) 108.0
Hz, 3JP-C ) 1.5 Hz), 129.55 (m), 132.02 (m), 133.88, 137.07 (d of m,
1
1JF-C ) 255.5 Hz), 139.92 (d of m, JF-C ) 250.3 Hz), 143.24 (d of
m, 1JF-C ) 247.3 Hz), 161.21. 19F NMR (470 MHz, CDCl3): δ -164.64
(m), -159.92 (broad s), -142.52 (m). Anal. Calcd for C43H30F5-
NO2P2: C, 68.89; H, 4.03; N, 1.87. Found: C, 69.07; H, 3.95; N, 1.83.
(R,R)-(salcy)CoOAc. The preparation of (R,R)-(salcy)CoOAc has
been described previously by Jacobsen and co-workers;15 however, only
1 equiv of acetic acid was used.
Representative Copolymerization Procedure. A 100 mL Parr
autoclave was heated to 120 °C under vacuum for 4 h, then cooled
(30) Reibenspies, J. H. Z. Kristallogr. 1994, 209, 620-621.
(31) Prakash, H.; Sisler, H. H. Inorg. Chem. 1968, 7, 2200-2203.
9
J. AM. CHEM. SOC. VOL. 127, NO. 31, 2005 10877

A R T I C L E S
Cohen et al.
under vacuum to 22 °C, and moved to a drybox. Complex (R,R)-(salcy)-
CoOBzF5 (11.7 mg, 0.0143 mmol), cocatalyst [PPN]Cl (8.2 mg, 0.014
mmol), and rac-PO (2.00 mL, 28.6 mmol) were placed in a glass sleeve
with a Teflon stir bar inside the Parr autoclave. The autoclave was
pressurized to 200 psi of CO2 and was left to stir at 22 °C for 1 h. The
reactor was vented at 22 °C. A small aliquot of the resultant
precipitated from methanol (30 mL). The polymer was collected and
dried in vacuo to constant weight, and the polymer yield was determined
(0.91 g, 31%).
Acknowledgment. G.W.C. gratefully acknowledges a Pack-
ard Foundation Fellowship in Science and Engineering, and
funding from the NSF (CHE-0243605), the Cornell Center for
Materials Research (supported through the NSF MRSEC
program, DMR-0079992), Sumitomo Chemicals, and the Cor-
nell University Center for Biotechnology, a New York State
Center for Advanced Technology.
1
polymerization mixture was removed from the reactor for H NMR
and GPC analysis. The remaining polymerization mixture was then
dissolved in methylene chloride (5 mL), quenched with 5% HCl solution
in methanol (0.2 mL), and transferred to a preweighed vial. The product
mixture was dried in vacuo to constant weight, and the crude yield
was determined after subtracting out the catalyst weight (0.95 g, 32%).
The product was dissolved in methylene chloride (3 mL) and
JA051744L
9
10878 J. AM. CHEM. SOC. VOL. 127, NO. 31, 2005