Selective formation of polycarbonate over cyclic carbonate: Copolymerization of epoxides with carbon dioxide catalyzed by a cobalt(III) complex with a piperidinium end-capping arm

Article abstract of DOI:10.1002/anie.200603132

(Chemical Equation Presented) Sidestepping a cyclic side product: Copolymerization of terminal epoxides with CO₂ was investigated by using a cobalt(III) complex bearing a piperidinium end-capping arm and a piperidinyl arm (see scheme; DME = 1,2-dimethoxyethane). The catalyst system can selectively produce copolymers without contaminant formation of cyclic carbonates even at high conversion of the epoxide (> 99 %).

Full text of DOI:10.1002/anie.200603132

Communications
Polycarbonate Synthesis
DOI: 10.1002/anie.200603132
Selective Formation of Polycarbonate over Cyclic
Carbonate: Copolymerization of Epoxides with
Carbon Dioxide Catalyzed by a Cobalt(III)
Complex with a Piperidinium End-Capping
Arm**
Koji Nakano, Toshihiro Kamada, and Kyoko Nozaki*
[1]
The copolymerization of epoxides with CO2 has been
intensively studied in the last decade as one of the most
promising processes for CO2 utilization.[2–11] The concomitant
production of cyclic carbonate is an incidental problem in the
reaction of terminal epoxides such as propylene oxide (PO).
Recently, the [Co(salcy)OAc] complex (salcyH2 = trans-N,N’-
bis(3,5-di-tert-butylsalicylidene)-1,2-cyclohexanediamine)
was reported to copolymerize PO with CO2 without the
production of cyclic propylene carbonate (PC) in up to
approximately 75% conversion with a turnover frequency
(TOF) of 51 [in moles of repeating unit per molCo per hour
(molru molCoÀ1 hÀ1)].[12] Further enhancement of the catalytic
activity was achieved with anionic cobaltate complexes such
À
+
+
=
as [Co(salcy)Cl(OBzF5)] [PPN] (BzF5 = C( O)C6F5, [PPN]
= bis(triphenylphosphoranylidene)ammonium), which
release the reactive anionic propagating species more readily
than neutral cobalt complexes.[13–15] However, such a high
reactivity also leads to the formation of PC at a higher degree
of conversion (> 50%): the polymerization mixture solidifies
to retard the intermolecular propagation, thus resulting in
relative acceleration of intramolecular degradation, so-called
back-biting.[16] Thus, high catalytic activity has been incom-
patible with the suppression of the formation of cyclic
carbonate. Herein, we report the copolymerization of termi-
nal epoxides with CO2 by using a new cobalt complex that can
selectively produce the copolymer, even at high monomer
conversion.
To achieve a high epoxide conversion without production
of cyclic carbonate, we designed cobaltate complex 1 with two
acetate ligands and a salcy-type ligand (Scheme 1). The key of
the catalyst design is in the piperidinyl and piperidinium arms.
[*] Dr. K. Nakano, T. Kamada, Prof. Dr. K. Nozaki
Department of Chemistry and Biotechnology
Graduate School of Engineering
The University of Tokyo
7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656 (Japan)
Fax: (+81)3-5841-7263
E-mail: nozaki@chembio.t.u-tokyo.ac.jp
[**] This study was supported by a Grant-in-Aid for Scientific Research
on Priority Areas “Advanced Molecular Transformations of Carbon
Resources” from the Ministry of Education, Culture, Sports, Science
and Technology, Japan. We are grateful to Prof. Yoshiaki Nishi-
bayashi and Dr. Yoshihiro Miyake (The University of Tokyo) for high-
resolution (HR) mass-spectrometric measurements.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
7274
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7274 –7277

Angewandte
Chemie
without a significant production of PC (PPC/PC = 99:1;
Table 1, entry 1).[19] Interestingly, high conversion of epoxide
did not result in the formation of PC. The copolymer with a
higher Mn value (23900 gmolÀ1) and narrow polydispersity
Table 1: Copolymerization of epoxides with CO2 catalyzed by cobalt
complex 1.[a]
2/3[b] Mn [gmolÀ1 [c]
]
Mw/Mn
[c]
Entry
R
Solvent t [h] Yield of
2+3 [%][b]
1
2
3
4
5
6
Me
Me
Me
Me
Me DME
Me DME
Me DME
EtDME
–
–
–
–
3
12
116
1
38
99:1 12600
99:1 23900
96:4 83700
90:10 7100
99:1 13200
97:3 26500
1.13
1.14
1.28
1.22
1.10
1.10
1.06
77
79
34
53
6
48 >99
7[d]
8
9
20
48
95
89
95:5
5100
48
89
97:3 31000
98:2 34300
1.12
1.14
Bu DME
[a] Reaction conditions: epoxide (14.3 mmol in entries 1, 2, and 4–9;
47.2 mmol in entry 3), 1 (epoxide/1=2000 in entries 1, 2, and 4; 6500 in
entry 3; 1000 in entries 5–9), CO2 (initial pressure: 1.4 MPa), DME
(1.0 mL in entries 5–9), at 258C (entries 1–3 and 5–9) or 608C (entry 4).
[b] Determined on the basis of 1H NMR spectroscopy of the crude
product by using phenanthrene as an internal standard. [c] Determined
by size-exclusion chromatography analysis using a polystyrene standard.
[d] Methanol (20 equivalents based on 1) was added.
Scheme 1. Catalyst design to suppress the production of cyclic carbon-
ate.
When the two acetate ligands initiate the copolymerization, a
piperidinium arm should control the nucleophilicity of the
propagating species by protonating the anionic propagating
species that is released from the cobaltate center. The
protonated propagating species is not nucleophilic enough
to form cyclic carbonate through back-biting, but it can react
with carbon dioxide[16b,c] or activated epoxide[16a] once it is
deprotonated by one of the two piperidinyl groups.
Complex 1 was synthesized by the reaction of Co(OAc)2
with the corresponding disalicylidenediamine[17] and subse-
quent oxidation in the presence of an excess amount of AcOH
and air, and the product was sufficiently dried under reduced
pressure. The formation of 1 was spectroscopically confirmed
index (PDI) of 1.14 was selectively produced in 76% yield
(TOF(PPC) = 127 molru molCoÀ1 hÀ1
;
Table 1,
entry 2).
Copolymerization at a higher PO/1 ratio of 6500 also gave
the alternating copolymer with a much higher Mn value
(83700 gmolÀ1) with high selectivity (PPC/PC = 96:4; Table 1,
entry 3). The efficient suppression of the back-biting process
allowed high-temperature copolymerization (608C), thereby
giving PPC selectively with higher activity (PPC/PC = 90:10,
TOF(PPC) = 602 molru molCoÀ1 hÀ1; Table 1, entry 4). It is
essential that the piperidinium group is tethered to the
salcy-type ligand framework. In fact, the copolymerization of
PO with CO2 by using the reported [Co(salcy)Cl(OBzF5)]À-
[PPN]+ system[14] in the presence of N-methylpiperidinium
acetate resulted in low selectivity for the copolymer (PO/
Co = 6000, 34% yield of PPC, 37% yield of PC, PPC/PC =
42:58).[20] Thus, the highest level of activity together with
suppression of PC formation and high conversion of PPC
have been accomplished using complex 1.[21]
1
1
as follows.[18] The H NMR spectrum of 1 and the H and
2H NMR spectra of partially deuterated 1, which was
obtained by using an excess mount of CD3COOD, clearly
show that complex 1 contains two acetate ligands. Accord-
ingly, one of the two piperidinyl groups should be protonated.
1
Moreover, the simple pattern of the H NMR spectrum of 1
reflects the symmetrical structure of 1 in solution, thus
indicating that a proton is exchanged rapidly between the two
piperidinyl groups. Thus, the two piperidinyl groups should
contribute equally to the copolymerization.
By using complex 1, high catalytic activity and high
epoxide conversion were accomplished. The copolymeriza-
tion of PO with CO2 (1.4 MPa), catalyzed by complex 1 (PO/
1 = 2000), gave the completely alternating copolymer, poly-
(propylene carbonate) (PPC) in 38% yield (TOF(PPC) =
254 molru molCoÀ1 hÀ1, Mn = 12600 gmolÀ1, Mw/Mn = 1.13)
The complete consumption of epoxide was attained by
adding solvents. Since the polymerization mixture becomes
viscous with increase of conversion, the addition of solvent is
necessary for effective diffusion of the reactant. Copolymer-
izations of PO with CO2 (PO/1 = 1000) in the presence of an
organic solvent such as 1,2-dimethoxyethane (DME), tolu-
ene, or dichloromethane gave the copolymer. Particularly,
employment of DME resulted in the quantitative conversion
with high copolymer selectivity (PPC/PC = 97:3, > 99% con-
version; Table 1, entries 5 and 6). Very recently, a similar
Angew. Chem. Int. Ed. 2006, 45, 7274 –7277
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7275

Communications
solvent effect was reported for the copolymerization with
out in the presence of methanol as a chain-transfer reagent
(Table 1, entry 7). When a large excess of methanol (20 equiv-
alents based on 1) was added, an alternating copolymer with a
methoxy initiating group was obtained as a main component.
The Mn value of the obtained copolymer was estimated to be
5100 gmolÀ1, which agrees with the theoretical Mn value of
4300 gmolÀ1 that was calculated from the copolymer yield and
the monomer/(methanol+acetate) ratio. Furthermore, the
PDI remained narrow (1.06).
In conclusion, the cobaltate complex with a piperidinium
arm was found to copolymerize terminal epoxides with CO2
with high activity and without a significant amount of
associated cyclic carbonate. In particular, complete consump-
tion of the epoxide was accomplished with high copolymer
selectivity. This feature is the key to the successful synthesis of
the block terpolymer.
[Co(salcy)Cl(OBzF5)]À[PPN]+, where the use of DME as a
solvent achieved the high PPC yield of 75% without a
contaminating amount of PC (PPC/PC = 98:2).[22] Under the
same reaction conditions, terminal epoxides with a longer
alkyl chain were also copolymerized with CO2 by using
complex 1 (Table 1, entries 8 and 9). Although the longer
alkyl substituents slightly decreased the catalytic activities,[13]
completely alternating copolymers were obtained in high
yields without formation of a significant amount of cyclic
carbonate.
The successful complete consumption of an epoxide
without production of cyclic carbonate allowed the synthesis
of block polymer. Thus, the block terpolymer was produced
by stepwise addition of the two different epoxides, PO and 1-
hexene oxide, without significant production of cyclic carbo-
nates (Scheme 2). Such a selective synthesis of a block
terpolymer is characteristic of the present catalyst system.
Since the previously reported catalysts hardly achieve the Experimental Section
Representative procedure for the copolymerization of propylene
complete consumption of an epoxide, the polymerization for
a block terpolymer should result in the production of a
tapered one.
oxide with CO2 (Table 1, entry 1). A 50-mL autoclave was charged
with propylene oxide (1.0 mL, 14.3 mmol) and 1 (5.8 mg, 7.2 ”
10À3 mmol) under argon. After CO2 (1.4 MPa) was introduced, the
reaction mixture was stirred at 258C for 3 h. The CO2 pressure was
released, and the polymerization mixture was quenched with
methanol, diluted with CH2Cl2, and transferred into a round-bottom
flask. Phenanthrene (34 mg, 0.19 mmol) was added as an internal
standard to the resulting solution. A small aliquot of the mixture was
taken out and concentrated under reduced pressure by rotary
1
evaporation. Analyses by H NMR spectroscopy and GPC gave the
yields of copolymer (38%) and propylene carbonate (< 0.1%), the
ratio of them (99:1), as well as the molecular weight (12600 gmolÀ1
and molecular-weight distribution (1.13) of the copolymer.
)
Received: August 2, 2006
Published online: October 6, 2006
Keywords: carbon dioxide · cobalt· copolymerization · epoxides ·
.
homogeneous catalysis
Scheme 2. Synthesis of a block terpolymer.
Immortal polymerization is indicated in the present PO/
CO2 copolymerization system.[23] In all experiments, gel
permeation chromatography (GPC) analyses of the obtained
copolymers gave bimodal traces. The Mn values of both
higher- and lower-molecular-weight copolymers increased
linearly in proportion to conversion in the range of 0–99%,
and the PDIs remained narrow (1.10–1.11). Moreover, for
each conversion, the Mn value of the higher-molecular-weight
copolymer was twice as large as that of the lower-molecular-
weight one. This difference can be attributed to the chain
transfer by contaminant water, which works as a bifunctional
initiating group to give a telechelic polymer, as previously
demonstrated by Sugimoto et al.[24] MALDI-TOF mass
spectrometry of the copolymer revealed the formation of
the telechelic copolymer in addition to the copolymer
initiated by the acetate group. These results clearly demon-
strate the reversible and rapid chain transfer of the propagat-
ing species, which is characteristic of successful “immortal”
polymerization.[25] To confirm such an immortal nature of the
present system, copolymerization of PO with CO2 was carried
[1] S. Inoue, H. Koinuma, T. Tsuruta, J. Polym. Sci. Part B 1969, 7,
287 –292.
[2] For recent reviews, see: a) G. W. Coates, D. R. Moore, Angew.
Chem. 2004, 116, 6784 –6806; Angew. Chem. Int. Ed. 2004, 43,
6618 –6639; b) H. Sugimoto, S. Inoue, J. Polym. Sci. Part A 2004,
42, 5561 –5573; c) D. J. Darensbourg, R. M. Mackiewicz, A. L.
Phelps, D. R. Billodeaux, Acc. Chem. Res. 2004, 37, 836 –844;
d) K. Nakano, N. Kosaka, T. Hiyama, K. Nozaki, Dalton Trans.
2003, 4039 –4050; e) M. S. Super, E. J. Beckman, Trends Polym.
Sci. 1997, 5, 236 –240.
[3] D. J. Darensbourg, R. M. Mackiewicz, J. Am. Chem. Soc. 2005,
127, 14026 –14038.
[4] D. J. Darensbourg, A. L. Phelps, Inorg. Chem. 2005, 44, 4622 –
4629.
[5] M. H. Chisholm, Z. P. Zhou, J. Am. Chem. Soc. 2004, 126,
11030 –11039.
[6] H. Sugimoto, H, Ohshima, S. Inoue, J. Polym. Sci. Part A 2003,
41, 3549 –3555.
[7] R. Eberhardt, M. Allmendinger, B. Rieger, Macromol. Rapid
Commun. 2003, 24, 194 –196.
[8] S. D. Allen, D. R. Moore, E. B. Lobkovsky, G. W. Coates, J. Am.
Chem. Soc. 2002, 124, 14284 –14285.
7276
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7274 –7277

Angewandte
Chemie
[9] D. R. Moore, M. Cheng, E. B. Lobkovsky, G. W. Coates, Angew.
Chem. 2002, 114, 2711 –2714; Angew. Chem. Int. Ed. 2002, 41,
2599 –2602.
[10] S. Mang, A. I. Cooper, M. E. Colclough, N. Chauhan, A. B.
Holmes, Macromolecules 2000, 33, 303 –308.
[11] T. Sarbu, T. Styranec, E. J. Beckman, Nature 2000, 405, 165 –168.
[12] Z. Q. Qin, C. M. Thomas, S. Lee, G. W. Coates, Angew. Chem.
2003, 115, 5642 –5645; Angew. Chem. Int. Ed. 2003, 42, 5484 –
5487.
[13] X. B. Lu, L. Shi, Y. M. Wang, R. Zhang, Y. J. Zhang, X. J. Peng,
Z. C. Zhang, B. Li, J. Am. Chem. Soc. 2006, 128, 1664 –1674.
[14] C. T. Cohen, T. Chu, G. W. Coates, J. Am. Chem. Soc. 2005, 127,
10869 –10878.
[15] X. B. Lu, Y. Wang, Angew. Chem. 2004, 116, 3658 –3661; Angew.
Chem. Int. Ed. 2004, 43, 3574 –3577.
[16] a) G. A. Luinstra, G. R. Haas, F. Molnar, V. Bernhart, R.
Eberhardt, B. Rieger, Chem. Eur. J. 2005, 11, 6298 –6314;
b) W. Kuran, T. Listos, Macromol. Chem. Phys. 1994, 195, 1011 –
1015; c) P. Gorecki, W. Kuran, J. Polym. Sci. Polym. Lett. 1985,
23, 299 –304.
[17] a) M. W. Fennie, E. F. DiMauro, E. M. OꢀBrien, V. Annamalai,
M. C. Kozlowski, Tetrahedron 2005, 61, 6249 –6265; b) E. F.
DiMauro, M. C. Kozlowski, J. Am. Chem. Soc. 2002, 124, 12668 –
12669; c) E. F. DiMauro, M. C. Kozlowski, Org. Lett. 2001, 3,
3053 –3056.
[18] See the Supporting Information.
[19] The Mn value is not close to the estimated value. This difference
could be a result of the chain transfer by concomitant water.
[20] The copolymerization using [Co(salcy)Cl(OBzF5)]À[PPN]+ in
the absence of N-methylpiperidinium acetate (PO/Co = 6000,
1.4 MPa of CO2, 3 h) was also carried out, and resulted in a high
catalytic activity (yield of 2 + 3 = 25%, TOF(PPC) =
492 molru molCoÀ1 hÀ1 Mn = 36000 gmolÀ1
, , Mw/Mn = 1.12) and
high selectivity (PPC/PC = 99:1) as reported in reference [14].
The prolonged copolymerization (68 h), however, led to the
partial degradation to PC (yield of 2 + 3 = 84%, PPC/PC =
72:28) and a broader molecular-weight distribution (Mw/Mn =
1.71).
[21] High catalytic activity (500–1000 molru molCoÀ1 hÀ1) was reported
by using cobaltate complexes.[13,14] However, such a system was
found to produce PC in greater than 50% conversion. Other
systems using chromium complexes or zinc complexes also show
high activity of 100–200 molru molCoÀ1 hÀ1 and are classified into a
highly active catalyst system, although a contaminant amount of
PC is produced even at low conversion.[4,8] As mentioned above,
the use of a [Co(salcy)OAc] complex has been the successful
example which achieves both the high catalytic activity
(51 molru molCoÀ1 hÀ1) and suppression of PC formation even at
high monomer conversion.[12]
[22] Although the employment of DME improved the yield of PPC,
the prolonged reaction did not result in a yield higher than 75%;
see: C. T. Cohen, G. W. Coates, J. Polym. Sci. Part A 2006, 44,
5182 –5191.
[23] Immortal polymerization is a living polymerization that involves
rapid, reversible chain transfer; that is, the chain transfer does
not lead to simple chain termination, and the dormant chain end
can repropagate when activated on a catalyst center. Because the
chain-transfer reaction proceeds much faster than the propaga-
tion reaction, polymers are obtained with low PDI values. See
reference [25].
[24] H. Sugimoto, H. Ohtsuka, S. Inoue, J. Polym. Sci. Part A 2005, 43,
4172 –4186.
[25] S. Inoue, J. Polym. Sci. Part A 2000, 38, 2861 –2871.
Angew. Chem. Int. Ed. 2006, 45, 7274 –7277
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7277