Revised interpretation for N-cyclohexylmaleimide polymerization in the presence of an optically active cobalt(II) complex: Polymerization mediated by anionic species formed through monomer-Co(II) complex-O2 interaction

Article abstract of DOI:10.1021/ma025857q

This report revises the interpretations of the reaction mechanism in the polymerization of N-cyclohexylmaleimide (CHMI) in the presence of α,α′-azobis(isobutyronitrile) (AIBN) and (R,R)-N,N′-bis (3,5-di-tert-butylsalicylidene)-1,2-cyclohexanediaminatocobalt(II) (1) [Macromolecules 2000, 33, 1489]. Although we proposed that the reaction was mediated by a radical species and that the interaction between growing polymer radical with 1 induced chirality in the main chain in the earlier report, the results in the present article indicate that the systems reported in the earlier report were most probably contaminated by a small amount of air and that the polymerization proceeds via an anionic species that was produced by the interaction of the monomer, 1, and O₂. The effect of the contamination was overlooked due to the lack of a control experiment using O₂ in the absence of AIBN in our earlier study. In this work, the polymerization of CHMI was performed in the presence and absence of 1, O₂, αα′-azobis-(isobutyronitrile) (AIBN), and protonic additives (methanol and acetic acid). The results indicated the following: (1) the polymerization in a tetrahydrofuran-pyridine mixture occurs in the presence of 1 and a small amount of oxygen even without using AIBN; (2) the reaction does not take place in the presence of 1 alone or in the presence of 1 and AIBN under a strictly controlled N₂ atmosphere; (3) a small amount of AcOH completely inhibits the polymerization with 1 and O₂ while 2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO) does not kill the reaction; and (4) the polymer was likely to possess -OH or -OOH groups at the chain terminals (a telechelic structure). On the basis of these observations, we withdraw our earlier, wrong interpretation that the polymerization is mediated by a radical species and conclude that the polymerization and the observed stereochemistry are effected by an anionic mechanism.

Full text of DOI:10.1021/ma025857q

3498
Macromolecules 2003, 36, 3498-3504
Revised Interpretation for N-Cyclohexylmaleimide Polymerization in the
Presence of an Optically Active Cobalt(II) Complex: Polymerization
Mediated by Anionic Species Formed through Monomer-Co(II)
Complex-O2 Interaction
Ta m a k i Na k a n o,*,† Toh r u Ya d e,† a n d Yosh io Ok a m oto‡
Graduate School of Materials Science, Nara Institute of Science and Technology,
Takayama-cho 8916-5, Ikoma, Nara 630-0101, J apan, and Department of Applied Chemistry,
Graduate School of Engineering, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8603, J apan
Received November 21, 2002; Revised Manuscript Received February 5, 2003
ABSTRACT: This report revises the interpretations of the reaction mechanism in the polymerization of
N-cyclohexylmaleimide (CHMI) in the presence of R,R′-azobis(isobutyronitrile) (AIBN) and (R,R)-N,N′-
bis(3,5-di-tert-butylsalicylidene)-1,2-cyclohexanediaminatocobalt(II) (1) [Macromolecules 2000, 33, 1489].
Although we proposed that the reaction was mediated by a radical species and that the interaction between
growing polymer radical with 1 induced chirality in the main chain in the earlier report, the results in
the present article indicate that the systems reported in the earlier report were most probably
contaminated by a small amount of air and that the polymerization proceeds via an anionic species that
was produced by the interaction of the monomer, 1, and O2. The effect of the contamination was overlooked
due to the lack of a control experiment using O2 in the absence of AIBN in our earlier study. In this
work, the polymerization of CHMI was performed in the presence and absence of 1, O2, R,R′-azobis-
(isobutyronitrile) (AIBN), and protonic additives (methanol and acetic acid). The results indicated the
following: (1) the polymerization in a tetrahydrofuran-pyridine mixture occurs in the presence of 1 and
a small amount of oxygen even without using AIBN; (2) the reaction does not take place in the presence
of 1 alone or in the presence of 1 and AIBN under a strictly controlled N2 atmosphere; (3) a small amount
of AcOH completely inhibits the polymerization with 1 and O2 while 2,2,6,6-tetramethyl-1-piperidinyloxy
(TEMPO) does not kill the reaction; and (4) the polymer was likely to possess -OH or -OOH groups at
the chain terminals (a telechelic structure). On the basis of these observations, we withdraw our earlier,
wrong interpretation that the polymerization is mediated by a radical species and conclude that the
polymerization and the observed stereochemistry are effected by an anionic mechanism.
In tr od u ction
polymerization mechanism involving an anionic growing
species. The anionic mechanism was supported by
copolymerization with methyl methacrylate and methyl
acrylate using conventional anionic initiators. Anionic
polymerization of N-substituted maleimides have been
performed using the complexes of organolithiums2 and
organozincs3 with the purpose of stereoregulation. Struc-
tural comparisons of the polymers obtained in this work
with the polymer synthesized using the n-BuLi-(-)-
sparteine (Sp) complex also supported the anionic
mechanism.
We recently reported that N-cyclohexylmaleimide
(CHMI) afforded an optically active polymer by polym-
erization in a tetrahydrofuran (THF)-pyridine mixture
in the presences of R,R′-azobis(isobutyronitrile) (AIBN)
and (R,R)-N,N′-bis(3,5-di-tert-butylsalicylidene)-1,2-cy-
clohexanediaminatocobalt(II) (1).1 The results were
interpreted by assuming a radical polymerization mech-
anism initiated by AIBN where the proposed interaction
between the growing radical and 1 was responsible for
the chiral induction. However, through the careful
reinvestigations in the present study, we found that the
polymerization was in effect mediated not by a radical
species but by an anionic species formed by the interac-
tion of CHMI, 1, and O2. The polymerization systems
reported in the earlier communication were most prob-
ably contaminated by a small amount of air, and the
contamination was overlooked due to the lack of a
control polymerization using 1 in the presence of O2 in
our earlier study. We herein withdraw the mechanistic
interpretation in the earlier communication, report the
full details of the CHMI polymerizations in the presence
and absence of 1, O2, AIBN, and protonic additives
(methanol and acetic acid), and propose the revised
Exp er im en ta l Section
Ma ter ia ls. CHMI (TCI) was used as obtained; purity
>99.2% (1H NMR), mp 87.6-88.6 °C. Methyl methacrylate
(MMA) and methyl acrylate (MA) (Wako Chemical) were
purified by distillation, dried on CaH2 under a N2 atmosphere,
and distilled under vacuum immediately before use. 2,2,6,6-
Tetramethyl-1-piperidinyloxy (TEMPO) (Aldrich) was used as
obtained. R,R′-Azobis(isobutyronitrile) (AIBN) was recrystal-
lized from ethanol. Tetrahydrofuran (THF) was refluxed over
sodium benzophenone ketyl, distilled and stored on LiAlH4
under a N2 atmosphere, and distilled under vacuum im-
† NAIST.
‡ Nagoya University.
* To whom all correspondence should be addressed: phone
+81-743-72-6081, Fax +81-743-72-6081; e-mail nakano@
ms.aist-nara.ac.jp.
10.1021/ma025857q CCC: $25.00 © 2003 American Chemical Society
Published on Web 04/24/2003

Macromolecules, Vol. 36, No. 10, 2003
Polymerization of RMI Using Co(II) Complex 3499
Ta ble 1. P olym er iza tion of CHMI in THF -P yr id in ea
conversion ratio was 70% by 1H NMR analysis of an aliquot
(ca. 0.1 mL) of the reaction mixture dissolved in CDCl3 (0.6
mL) based on the intensity of the olefinic proton signal (s, 6.6
ppm) using the solvent signal (m, 3.7 ppm) as an internal
reference.
[1]0 [AIBN]0 [TEMPO]0 added temp convb
Mw/
Mn
c
c
run (M)
(M)
(M)
gas (°C) (%)
Mn
1
2
3
4
5
6
7
8
0
0.023
0
0
0
0.10
0
0
0
0.10
60
60
60
60
rt
rt
rt
rt
95
e
1060d 1.98d
0.023 0.023
0.023 0.023
0.023 0.023
0.023
0.023
0.023
0.023
An ion ic P olym er iza t ion Usin g 9-F lu or en yllit h iu m
(F lLi). The glassware setup was the same as that for the
polymerization using 1-O2. A typical procedure is described
for run 3 in Table 4. Fluorene (166 mg, 1.00 mmol) was placed
in a glass ampule, dried under high vacuum for 30 min, and
purged with N2. THF (1.25 mL) was introduced using a syringe
and fluorene was dissolved. To the solution was added n-BuLi
(1.6 M hexane solution, 0.63 mL, 1.0 mmol) at room temper-
ature. The resulting FlLi solution (0.50 M) was used as an
initiator after standing for 10 min at room temperature. CHMI
(500 mg, 2.79 mmol) was placed in a glass ampule, dried under
high vacuum for 30 min, and purged with N2. THF (1.47 mL)
and MA (0.25 mL, 2.8 mmol) were introduced using a syringe,
and the monomer was dissolved. The polymerization was
initiated by adding the FlLi solution (0.11 mL, 0.053 mmol)
to the monomer solution cooled at 0 °C. After 24 h of reaction,
an aliquot (0.1 mL) of reaction mixture was sampled out with
a syringe and dissolved in 0.6 mL of CDCl3, and the solution
O2
O2
33
23
∼0
1970 2.01
2480 2.01
0
0
0
0
O2
air
O2
70 11100 2.42
64
52
6590 2.62
7200 2.65
a
Conditions: [CHMI]0 ) 2.3 M, [pyridine]0 ) 0.35 M, added
b
gas ) 5 mL, time ) 24 h. Determined by 1H NMR analysis of
the reaction mixture. c Determined by SEC (vs polystyrene).
THF-soluble part (51%). e Trace.
d
mediately before use. Toluene was purified by washing with
H2SO4, distilled, dried by mixing with a small amount of
n-BuLi, and distilled under
a vacuum immediately use.
Pyridine (WAKO) was dried on KOH, distilled, and stored on
Molecular Sieve 3A under a N2 atmosphere. Methanol (WAKO)
and acetic acid (WAKO) were degassed under vacuum and
purged with N2 immediately before use. Fluorene (WAKO) was
recrystallized from hexane (mp 117.9-118.3 °C). n-BuLi
(Nacalai) (1.6 M, a hexane solution) was used as obtained. Co-
(OAc)2‚4H2O (WAKO) and (R,R)-(-)-1,2-diaminocyclohexane
(WAKO) were used as obtained. (-)-Spartein (Sp) (Aldrich)
was distilled over CaH2 under a reduced pressure and used
as a dry toluene solution (1.08 M).
1
was subjected to H NMR and SEC analysis.
An ion ic P olym er iza tion Usin g n -Bu Li-Sp . The litera-
ture method2 was followed with modifications. The glassware
setup was the same as that for the polymerization using 1-O2.
To dry toluene (1.7 mL) placed in a glass ampule was added
n-BuLi (1.6 M hexane solution, 0.69 mL, 1.1 mmol) and Sp
(1.1 M toluene solution, 1.2 mL, 1.3 mmol). The resulting
solution (0.30 M) was used as an initiator solution after
standing for 10 min at room temperature. CHMI (1.02 g, 5.68
mmol) was placed in a glass ampule, dried under high vacuum
for 1 h, and purged with argon. Toluene (19.9 mL) was
introduced using a syringe to dissolve the monomer. The
polymerization was initiated by adding the n-BuLi-Sp solu-
tion (0.43 mL, 0.13 mmol) to the monomer solution cooled at
0 °C. After 2 h, the polymerization was quenched by the
addition of methanol (0.2 mL), and the reaction mixture was
poured into methanol (500 mL). The methanol-insoluble part
was collected with a centrifuge and dried under vacuum at 50
°C for 2 h (0.382 g, 38%). Mn 6300, Mw/Mn 2.63.
Syn th esis of Com p lex 1. The literature method4 was
followed with modifications. 3,5-Di-tert-butyl-2-hydroxybenz-
aldehyde (21.7 g, 92.8 mmol) (Aldrich) dissolved in ethanol (340
mL) was mixed with a solution of (R,R)-(-)-1,2-diaminocyclo-
hexane (5.30 g, 46.4 mmol) (WAKO) in ethanol (160 mL). The
mixture was diluted with 200 mL of ethanol and refluxed for
1 h. Removal of solvent afforded (R,R)-N,N′-bis(3,5-di-tert-
butylsalicylidene)-1,2-cyclohexanediamine (25.4 g, 100%) as a
yellow solid. 1H NMR δ: 13.71 (s, 2H), 8.30 (s, 2H), 7.30 (d,
J ) 2.3 Hz, 2H), 6.99 (d, J ) 2.3 Hz, 2H), 3.35-3.29 (m, 2H),
2.0-1.4 (m, 8H), 1.41 (s, 18H), 1.24 (s, 18H). HRMS (ESI).
Calcd for C36H55N2O2 (M + H+): 547.4264. Found: 547.4236.
Co(OAc)2‚4H2O (2.26 g, 9.06 mmol) (WAKO) dissolved in a
mixture of H2O (7 mL) and ethanol (70 mL) was added to a
solution of (R,R)-N,N′-bis(3,5-di-tert-butylsalicylidene)-1,2-cy-
clohexanediamine (4.96 g, 9.06 mL) in toluene (70 mL), and
the reaction mixture was refluxed for 2 h. Solvent was removed
under reduced pressure to give a solid, crude material. The
crude material was dissolved in CHCl3 (30 mL), and hexane
(600 mL) was added to the solution with stirring. This mixture
afforded a deep red solid poduct (1) on standing at -20 °C for
12 h. The product was collected with a centrifuge, washed with
hexane (5 mL), and dried under vacuum. TLC analysis (Silica
Gel 60 F254 (Merck), hexanes-ethyl acetate 9/1) showed a
single spot and indicated the absence of free ligand. Yield 2.18
g (40%). HRMS (ESI). Calcd for C36H52CoN2O2: 603.3361.
Found: 603.3323.
P olym er iza tion Usin g 1-O2. The reactions were carried
out in a glass ampule sealed with a glass-made three-way
stopcock attached to the ampule via a ground joint. The joint
was completely sealed with high-vacuum grease. The ampule
was flame-dried under high vacuum and flushed with N2
immediately before use. A typical procedure is described for
run 6 in Table 1. CHMI (1.000 g, 5.58 mmol) and 1 (33.6 mg,
0.0558 mmol) were placed in a glass ampule, dried under high
vacuum for 30 min at room temperature, and purged with N2.
THF (1.30 mL) and a THF solution of pyridine (8.49 M THF
solution, 0.10 mL, 0.85 mmol) were introduced with a syringe,
and the monomer and 1 were completely dissolved. Oxygen
(purity 99.9%, 5 mL) was bubbled through the solution in a
ca. 3 s period using a syringe. After the reaction mixture was
allowed to stand for 24 h, a part of the reaction mixture (0.60
mL) was poured into methanol (40 mL). The methanol-
insoluble part was collected with a centrifuge and dried under
vacuum at 40 °C for 10 h (0.184 g, 67%). The monomer
1
Mea su r em en ts. The H NMR spectra were recorded on a
J EOL J NM ECP500 or LA400 spectrometer (500 and 400
MHz, respectively, for 1H measurement). CD spectra were
obtained with a J ASCO J -720L spectrometer. SEC analyses
of polymers were performed using a chromatographic system
consisting of a Hitachi L-7100 pump, an L-7420 UV detector
(254 nm), an L-7490 RI detector, and a J ASCO OR990
polarimetric detector equipped with TOSOH TSK gel G3000HHR
(styrene-divinylbenzene gel; particle size 5 µm; molecular
weight range up to 6 × 104) and G6000HHR (styrene-
divinylbenzene gel; particle size 5 µm; molecular weight range
up to 4 × 107) columns connected in series (eluent, THF; flow
rate, 1.0 mL/min). MALDI-mass spectra were taken on a
Voyager DE-STR spectrometer equipped with a N2 laser (337
nm, 3 ns pulse width, frequency up to 20 Hz) under vacuum
(sample chamber pressure 5.5 × 10-7 Torr) using reflector
mode (acceleration voltage 20 000 V). The exciting laser power
was set to 3000. Samples were prepared by mixing a N,N-
dimethylformamide solution of polymer (concentration 10 mg/
mL, 1 µL), aqueous 2,5-dihydroxybenzoic acid (concentration
10 mg/mL, 10 µL) as matrix, and aqueous potassium iodide
(concentration 1 mg/mL, 1 µL) and drying the mixed solution
in a sample well on a gold-plated sample slide under air flow.
ESI MS spectra were recorded using a J EOL J MS-700
spectrometer in the positive measurement mode. TGA analyses
were performed on a SEIKO TG/DTA-6200 system under a
N2 atmosphere at a heating rate of 20 °C/min. Temperature
and heat capacity were calibrated using In as a standard
sample (Tm 156.6 °C, heat capacity 28.59 mJ /mg).
Resu lts a n d Discu ssion
P olym er iza tion Beh a vior a n d P olym er Str u c-
tu r e. In our earlier communication,1 we reported that

3500 Nakano et al.
Ta ble 2. P olym er iza tion of CHMI in THF a
[1]0 [AIBN]0 [TEMPO]0 added temp convb
Macromolecules, Vol. 36, No. 10, 2003
Mw /
Mn
c
c
run (M)
(M)
(M)
gas
(°C) (%)
Mn
1
2
3
4
0
0.023
0
0
0.10
0
60
60
60
rt
>99 1410d 4.11d
0.023 0.023
0.023 0.023
0.023
57 4840 1.57
∼0
0
O2
∼0
a
Conditions: [CHMI]0 ) 2.3 M, added gas ) 5 mL, time ) 24
h. Determined by 1H NMR analysis of the reaction mixture.
b
c Determined by SEC (vs polystyrene). THF-soluble part (51%).
d
CHMI afforded an optically active polymer by polym-
erization in a tetrahydrofuran (THF)-pyridine mixture
using AIBN in the presence of 1. However, through the
repeated reactions using 1 and AIBN in THF-pyridine
under the same conditions as we reported in the
communication, we found that the polymerization had
poor reproducibility with varying polymer yield in the
range of 0-13%. In contrast, the conventional radical
polymerization using AIBN and the anionic polymeri-
zation with n-BuLi-Sp were completely reproducible.
Through reinvestigations into the reaction in the pres-
ence of 1 and AIBN, we noticed that a crude reaction
mixture that had failed in giving a polymer on heating
for 24 h and was left in the ampule with the stopcock
removed in the fume hood at room temperature after
the experiment became gradually viscous due to poly-
mer formation. On the basis of this observation, we
suspected that the inert atmosphere was contaminated
with a small amount of air in the earlier study, and the
contamination played a role in promoting the polymer-
ization. To make this point clear and reveal the nature
of polymerization, the systematic experiments were
carried out as described below.
The results of the polymerization in a THF-pyridine
mixture in the presence and absence of 1, AIBN, O2,
and air are summarized in Table 1. Under the conven-
tional radical polymerization condition with AIBN as
initiator, a polymer was obtained in a high yield (run
1). In contrast, only a trace amount of polymer was
obtained in the presence of 1 (run 2) although we
reported in the eariler communication that an optically
active polymer was produced in 72% yield under the
same conditions. In this reaction, 1 seems to inhibit the
radical polymerization. However, when 5 mL of O2 was
quickly bubbled through the sample solution, a polymer
was obtained in 33% yield (run 3). Further, polymeri-
zation occurred even without AIBN when O2 or air was
injected to the solution of CHMI and 1 at room temper-
ature while 1 alone did not promote polymerization of
CHMI (runs 5-7).
F igu r e 1. 1H NMR spectra of poly(CHMI)s: (A) conventional
radical polymerization with AIBN, (B) anionic polymerization
using n-BuLi-(-)-sparteine at 0 °C, and (C) polymerization
using Co(II) complex 1 and O2 (500 MHz, CDCl3, room
temperature).
obtained in both the presence and absence of 1 where
the former case led to a lower polymer yield (runs 1 and
2). The reaction in the presence of 1 was completely
inhibited by TEMPO, indicating that the polymerization
was mediated by a free-radical species (run 3). In
addition, no polymerization took place in a pure THF
solution containing only CHMI and 1 upon injecting O2
(run 4). Hence, it is assumed that coordination of
pyridine to the cobalt center of 1 is indispensable in the
polymerization involving 1 and O2 in the reaction
mechanism.
There have been reports on the polymerization of
vinyl monomers using a cobalt(II) species and O2 leading
to a polymer containing alternating peroxy (-O-O-)
and vinyl monomeric units in the main chain.5,6 If this
is the case for the CHMI polymerization in the presence
of 1 and O2 in THF-pyridine, the polymer should be
clearly distinguished by spectroscopies from the polymer
prepared with AIBN whose main chain consists only of
vinyl monomeric units, and also it may thermally
decompose through bond-breaking of the peroxy moiety.
Upon these working hypotheses, the structures of the
polymers obtained from runs 1 and 6 of Table 1 were
1
examined by H NMR and thermolysis.
Figure 1 shows the 1H NMR spectra of the two
polymers along with that of the polymer anionically
prepared using the n-BuLi-Sp complex according to the
literature2 (Mn 6300, Mw/Mn 2.63). There was no clear
difference in the NMR spectra of the three polymers
(Figure 1) in the cyclohexyl -CH2- (1-2.4 ppm) and
the main chain -CH and N-CH (2.6-4.6 ppm) regions,
indicating that the polymer prepared using 1-O2 in
THF-pyridine has the same chemical structure as those
of the anionically and radically obtained polymers
having no -O-O- bondings in the main chain. The
slight difference in the peak shape around 2 ppm may
reflect the difference in the tacticity of the polymers.
In the TGA experiments under a N2 atmosphere, the
polymers from runs 1 and 6 of Table 1 indicated very
similar thermolysis profiles (Figure 2). The two poly-
mers decomposed at similar onset temperatures around
ca. 400 °C. In addition, the polymer from run 6 of Table
1 showed no decrease in molecular weight when heated
at 130 °C for 2 h under a N2 flow. These results support
the conclusion from the NMR studies that the polymer
from run 6 of Table 1 does not contain -O-O- bondings
in the main chain.
To obtain information on the nature of growing
species, the effect of TEMPO on the reaction was
investigated. The polymerization in the presence of 1
and O2 and that in the presence of 1, O2, and AIBN were
not inhibited with TEMPO (runs 4 and 8), indicating
that the interaction with O2, 1, and CHMI leads to a
growing species that is not a free radical. A free-radical
growing species would be quickly terminated by TEMPO.
On the basis of the results described so far, it would be
reasonable to conclude that the reaction systems we
reported in the earlier communication was contami-
nated with a small amount of air and that the reaction
was promoted by the interaction of 1, CHMI, and O2 in
the air.
A different polymerization behavior was observed in
a pure THF solution in the absence of pyridine (Table
2). In the polymerization using AIBN, a polymer was

Macromolecules, Vol. 36, No. 10, 2003
Polymerization of RMI Using Co(II) Complex 3501
F igu r e 2. TGA profiles of poly(CHMI)s obtained by radical
polymerization using AIBN (run 1 in Table 1) (A) and by
polymerization using 1 and O2 (run 6 in Table 1) (B).
F igu r e 4. Proposed mechanism of CHMI polymerization
using 1 and O2.
The second series of peaks with mass numbers ) 739
(755), 918 (934), 1098 (1113), 1276 (1292), 1457 (1473),
... may correspond to a poly(CHMI) bearing a CHMI
monomer with the intact vinyl group at the chain
terminal (Mn ) 179 + 179n +1). In this case also, the
mass peaks seem to reflect molecular ions carrying an
Na and a K (mass numbers in parentheses). From the
mass spectral data, the chain reaction may proceed
through chain transfer to monomer to a certain extent.
In contrast to the polymer synthesized using AIBN,
the one synthesized using 1 and O2 showed a series of
group peaks. Each group consisted of three to five peaks
that were separated by a mass number of 16. The
lowest-mass peak in each group seems to correspond to
a polymer molecular ion consisting only of CHMI units
with no terminal group, and the rest may correspond
polymer molecular ions bearing one to four oxygen
atoms. The molecular ions seem to carry a K in this case.
These results indicate that the polymer synthesized
with 1 and O2 contains up to four oxygen atoms per
chain. Because it was shown that there is no -O-O-
bonding in the main chain as described in the preceding
section, the oxygen atoms are assumed to be present at
the chain terminal as an -O-OH or an -OH group.
Therefore, at least, part of the poly(CHMI) synthesized
using 1-O2 is considered to have a telechelic structure
with an -OOH or an -OH groups at the both chain
terminals.
F igu r e 3. MALDI-mass spectra of poly(CHMI)s obtained by
radical polymerization using AIBN (run 1 in Table 1) (A) and
by polymerization using 1 and O2 (run 6 in Table 1) (B). The
spectra were obtained using 2,5-dihydroxybenzoic acid as
matrix.
The structure of the two polymers was further ana-
lyzed by means of MALDI-mass spectrometry (Figure
3). The mass spectrum of the polymer from run 1 of
Table 1 (radical polymerization using AIBN) indicated
two series of peaks corresponding to different types of
structure (Figure 3A). The first series of peaks with
mass numbers ) ca. 809 (827), 988 (1004), 1167 (1183),
1346 (1362), ... may correspond to the structure of a
poly(CHMI) having a 1-cyano-1-methylethyl group at a
chain terminal (Mn ) 68 + 179n). The molecular ions
seem to be accompanied by an Na or a K (mass numbers
in parentheses). This indicates that the 1-cyano-1-
methylethyl radical effectively added to CHMI mono-
mer, leading to a growing radical. The fact that there
were no peaks corresponding to a polymer having two
terminal 1-cyano-1-methylethyl groups demonstrates
that the growing chain is not terminated through
recombination on through primary radical termination.
P r op osed Rea ction Mech a n ism . On the basis of
the findings so far described, we propose the revised
reaction mechanism (Figure 4). The first step of the
series of reactions will be complexation of O2 with 1 (step
A). Such a reaction is known for cobalt(II) complexes
having Schiff base type ligands, and crystal structures
have been shown in some examples.7,8 The next step will
be single-electron transfer from the 1-O2 complex to
CHMI (step B). This type of reaction has been reported
for reduction of a benzoquinone using the Co(CN)5(II)-
O2 complex.9 In step B, it is not clear whether O2 is

3502 Nakano et al.
Macromolecules, Vol. 36, No. 10, 2003
an alternating copolymer of O2 with a vinyl monomer.6
Ta ble 3. Effect of P r oton ic Ad d itive on th e
P olym er iza tion of CHMIa
•
While Co(III)Pc(Py)-O2 was proposed to react with a
•
monomer (M) to give OOM• species and Co(II)Pc(Py)
[1]0 [AIBN]0 protonic added temp convb
Mw/
Mn
c
c
in the oxidative polymerization,6 we assume that the
run (M)
(M)
additive gas
(°C) (%)
Mn
•
Co(III)L(Py)-O2 derived from 1 transfers an electron
1
2
3
4
5
6
0
0
0
0.023
0.023
0.023
0
0
0
none
MeOH
AcOH
none
MeOH
AcOH
60
60
60
rt
rt
rt
95
90
32
1060d 1.98d
1100e 1.85e
1120 1.38
to CHMI in the present reaction systems (step B) based
on the fact that the obtained polymer has a vinyl
homopolymeric structure. The mechanism would change
depending on the monomer: for a more electron-poor
monomer, the electron-transfer mechanism will be more
plausible.
0.023
0.023
0.023
O2
O2
O2
70 11100 2.42
25
f
910 1.18
a
Conditions: [CHMI]0 ) 2.3 M, [pyridine]0 ) 0.35 M, [additive]0
) 0.50 M, added gas ) 5 mL, time ) 24 h, solvent ) THF.
Cop olym er iza tion w ith Acr ylic Mon om er s. Po-
lymerization of methyl acrylate (MA) and methyl meth-
acrylate (MMA) and copolymerization of CHMI with MA
and MMA using 1-O2 were carried out (Table 4). The
results supported the anionic nature of the propagating
species in the CHMI polymerization using 1-O2.
The attempted homopolymerizations of MA and MMA
using 1-O2 under the same conditions as those for run
6 in Table 1 failed in affording a polymer. This may
mean that the second step in the proposed initiation
mechanism in Figure 4 requires a monomer that is
electron-poor enough to accept an electron from 1-O2
complex. Monomers with only one electron-withdrawing
carbonyl group attached to the vinyl group may not be
able to form an anion radical through the interaction
with 1-O2.
In addition, in the attempted copolymerization of
CHMI with MA and MMA using 1-O2, only the ho-
mopolymers of CHMI were produced (runs 5 and 6),
while in the radical copolymerization with AIBN, both
monomers were quantitatively consumed (runs 1 and
2). This polymerization behavior is very similar to that
of the anionic copolymerization using 9-fluorenyllithium
(FlLi), a typical anionic initiator (runs 3 and 4). These
results confirm that the growing species in the CHMI
polymerization with 1-O2 has anionic nature, but its
nucleophilicity is not high enough to polymerize MA and
MMA.
Livin gn ess of th e P olym er iza tion In itia ted w ith
1 a n d O2. The time vs conversion and the conversion
vs Mn plots for the polymerizations under the same
conditions as runs 1 and 6 in Table 1 are shown in
Figure 5. The polymerization initiated with 1 and O2 is
slower than that of the conventional radical polymeri-
zation. A clear difference in the conversion vs Mn profile
was observed for the two systems; Mn gradually de-
creased as the polymerization proceeded in the system
with AIBN while Mn increased with conversion in the
system with 1-O2. The latter system seems to have
living character, and this observation is consistent with
the proposed anionic mechanism.
Determined by 1H NMR analysis of the reaction mixture.
b
d
c Determined by SEC (vs polystyrene). THF-soluble part (51%).
e THF-soluble part (66%). f Trace.
completely freed or it stays within the anion radical
complex structure. Two molecules of the resulting CHMI
anion radical may couple to afford a dianion type species
having two trivalent Co species generated from 1 as a
countercation (step C). This species would add CHMI
monomers, leading to a polymer. The growing species
is expected to have an anionic character, and therefore
it is not quenched with TEMPO.
This anionic reaction mechanism was supported by
the polymerization in the presence of protonic additives
(Table 3). The polymer yield was largely suppressed in
the presence of MeOH, leading to the formation of a
small amount of oligomers, and no polymerization took
place in the presence of acetic acid (runs 4-6) while
MeOH had virtually no effect on the polymerization
using AIBN (run 2). The decrease in polymer yield of
the radical polymerization with AIBN in the presence
of acetic acid may mean that acetic acid suppressed
polymerization through hydrogen radical transfer to the
growing radical (run 3). These results indicate that the
growing species has an anionic character, but its nu-
cleophilicity is lower than other carbanion species such
as those in the anionic polymerization systems of acrylic
monomers or styrene where the reaction is readily
quenched with MeOH.
Uptake of O2 into the polymer chain seems to be a
part of termination step of the polymerization in addi-
tion to protonation of the anionic species (step D). A Co-
O-O-C bonding at the chain terminal will result
through the reaction of the growing species with O2. The
terminal Co-O-O-C group may heterolytically decom-
pose with water in the atmosphere or with MeOH used
in the product purification process to afford a terminal
-O-O-H group that may be converted into a terminal
-OH group through -O-O- bond breaking. These
assumptions are consistent with the mass numbers
observed in Figure 3B.
The reaction in step A in Figure 4 has been proposed
also for the Co(II) phthalocyanine complex-pyridine
(Co(II)Pc(Py))/O2 system which was used as an initiator
for oxidative polymerization of vinyl monomers giving
Ch ir a l In d u ction a n d Ma in -Ch a in Ster eostr u c-
tu r e. The stereochemistry of the present polymerization
systems was compared with that of the conventional
anionic polymerization systems. The main chain of the
Ta ble 4. Cop olym er iza tion of CHMI w ith Meth yl Acr yla te or Meth yl Meth a cr yla tea
CHMI comonomer
c
c
run
comonomer
initiator
solvent
temp (°C)
convb (%) convb (%)
Mn
Mw/Mn
1
2
3
4
5
6
MA
MMA
MA
MMA
MA
MMA
AIBN
AIBN
FlLi
FlLi
1-O2
1-O2
THF-pyridine
THF-pyridine
THF
THF
THF-pyridine
THF-pyridine
60
60
0
0
rt
rt
>99
98
>99
>99
74
98
>99
∼0
d
1840
4320
3420
4540
14700
14700
2.35
6.02
4.10
3.99
3.55
2.53
∼0
74
∼0
a
Conditions: [CHMI]0 ) 1.2 M, [comonomer]0 ) 1.2 M, [initiator]0 ) 0.023 M, [pyridine]0 ) 0.35 M, added gas ) 5 mL, time ) 24 h.
Determined by 1H NMR analysis of the reaction mixture. c Determined by SEC (vs polystyrene). Trace.
b
d

Macromolecules, Vol. 36, No. 10, 2003
Polymerization of RMI Using Co(II) Complex 3503
F igu r e 5. Conversion vs time (A) and Mn vs conversion plots
(B) for CHMI polymerization systems: radical polymerization
using AIBN in THF-pyridine at 60 °C ([CHMI]0 ) 2.3 M,
[AIBN]0 ) 0.023 M) (blue) and polymerization using 1 and O2
in THF-pyridine at room temperature ([CHMI]0 ) 2.3 M, [Co-
(II)]0 ) 0.023 M, O2 ) 5 mL) (red). Products were partially
insoluble in THF in the radical polymerization.
F igu r e 6. CD spectra of poly(CHMI)s prepared by polymer-
ization with 1 and O2 (run 6 in Table 1) (A-1) and by anionic
polymerization with Sp-FlLi (A-2), and those of 1 (B-1) and
(R,R)-N,N′-bis(3,5-di-tert-butylsalicylidene)-1,2-cyclohexanedi-
amine (ligand of 1) (B-2) (THF, room temperature).
polymers may have cis- and trans-monomeric units; the
latter is chiral. Therefore, if a (R,R)-trans- or (S,S)-trans-
monomeric unit is produced in excess, the polymer may
be optically active. The polymers obtained using 1-O2
complex were optically active.
Figure 6 shows the CD spectra of the polymer from
run 6 in Table 1, the anionically obtained polymer, 1,
and (R,R)-N,N′-bis(3,5-di-tert-butylsalicylidene)-1,2-cy-
clohexanediamine (ligand of 1). The first two samples
showed clear CD bands in the imide chromophore region
with a similar spectral shape. Complex 1 and its ligand
showed CD bands with completely different patterns
from those of the polymers, indicating that the chirop-
tical properties of the polymer sample synthesized with
1-O2 does not contain 1 or its ligand residue that may
affect the chiroptical properties. These results demon-
strate that in the polymerization with 1-O2 induced an
excess (R,R)- or (S,S)-configuration in the main chain
similar to the polymerization using the n-BuLi-Sp
complex, a conventional anionic initiator.
The main-chain stereostructure was also examined
by means of 13C NMR spectroscopy (Figure 7). It has
been pointed out that the signals around 40-45 ppm
in 13C NMR spectra of poly(RMI)s based on the main-
chain methine carbons reflect the main-chain configu-
ration (tacticity).3b In the 13C NMR spectra, the polymer
F igu r e 7. 13C NMR spectra of poly(CHMI)s: (A) conventional
radical polymerization with AIBN, (B) anionic polymerization
using n-BuLi-Sp at 0 °C, and (C) polymerization using 1 and
O2 (125 MHz, CDCl3, room temperature).
synthesized using the anionic initiator and that pre-
pared using 1 and O2 indicated a very similar spectral
pattern while the polymer synthesized using AIBN had
a different, broad spectral shape. This supports the idea
that the optically active polymers have a (R,R)- or (S,S)-
configuration in excess while the radically prepared,
optically inactive polymer has an equal amount of (R,R)-

3504 Nakano et al.
Macromolecules, Vol. 36, No. 10, 2003
and (S,S)-configurations. Because the former two poly-
mers and the latter polymer are diastereomeric, they
may lead to different NMR spectral patterns.10 The
observed stereochemical characteristics are consistent
with the revised conclusion that the reaction is mediated
by an anionic species.
We thank Ms. Y. Nishikawa (NAIST) for mass spectro-
metric analyses. We are indebted to Prof. T. Yamada
(Keio University) and Prof. A. Kajiwara (Nara Univer-
sity of Education) for helpful discussions. This work was
supported in part by Grant-in-Aid for Priority Area
Research No. 13128203 “Construction of Dynamic Redox
Systems Based on Nano-Space Control“ from the Min-
istry of Education, Culture, Sports, Science and Tech-
nology of J apan.
Con clu sion s
The findings in this work are summarized as follows:
1. In the absence of O2, complex 1 simply inhibits
radical polymerization with AIBN in THF-pyridine.
2. Interaction of CHMI, 1, and O2 generates an
initiating species probably through single-electron trans-
fer from the 1-O2 complex to CHMI, and the propagat-
ing species has anionic character with relatively low
nucleophilicity.
3. The resulting polymers appear to possess a tele-
chelic structure with -OH or -OOH groups at the chain
terminals introduced probably through termination with
O2.
4. The (R,R)- or (S,S)-trans-configuration of the main
chain is generated in excess by the influence of the
cobalt(III) species with a bulky, chiral ligand that
originates from 1 and acts as a countercation for the
anionic growing species.
Refer en ces a n d Notes
(1) Nakano, T.; Tamada, D.; Miyazaki, J .-i.; Kakiuchi, K.; Oka-
moto, Y. Macromolecules 2000, 33, 1489.
(2) Okamoto, Y.; Nakano, T.; Kobayashi, H.; Hatada, K. Polym.
Bull. (Berlin) 1991, 25, 5.
(3) (a) Oishi, T.; Yamasaki, H.; Fujimoto, M. Polym. J . 1991, 6,
795. (b) Onimura, K.; Tsutsumi, H.; Oishi, T. Macromolecules
1998, 31, 5971.
(4) Larrow, J . F.; J acobsen, F.; Gao, Y.; Hong, Y.; Nie, X.; Zepp,
C. M. J . Org. Chem. 1994, 59, 1939.
(5) (a) Nukui, M.; Yoshiono, K.; Ohkatsu, Y.; Tsuruta, T. Mak-
romol. Chem. 1979, 180, 523. (b) Nukui, M.; Ohkatsu, Y.;
Tsuruta, T. Macromol. Chem. 1982, 183, 1457.
(6) Nanda, A. K.; Kishore, K. Macromolecules 2001, 34, 1558.
(7) (a) Floriani, C.; Calderazzo, F. J . Chem. Soc. A 1969, 946.
(b) Carderazzo, F.; Floriani, C.; Salzmann Inorg. Nucl. Chem.
Lett. 1966, 2, 379. (c) J ones, R. D.; Summerville, D. A.; Basolo,
F. Chem. Rev. 1979, 79, 139.
The proposed anionic mechanism of the present
system involving is in a stark contrast to that in the
living radical polymerization system of acrylates11 or
catalytic chain-transfer systems12 using cobalt(II) por-
phyrin in the absence of O2. Formation of ionic initiating
species through electron transfer has been reported for
the pairs of strongly electron-rich and -poor olefins
involving no metallic species.13 The initiation system
found here may expand the scope of such type of
polymerization because one might possibly be able to
control the tendency of electron transfer by tuning the
complex structure including the internal and external
ligands and the kind of metal and also by choosing an
additive that replaces O2 so that the system can be
applicable to a wider range of monomers.
(8) Niedehoffer, E. C.; Timmons, J . H.; Martell, A. E. Chem. Rev.
1984, 84, 137.
(9) Nishinaga, A.; Tomita, H.; Mastuura, T. Tetrahedron Lett.
1980, 21, 4853.
(10) A conformational effect could also contribute partially to the
chiroptical properties of poly(CHMI). However, the extent of
contribution of chiral conformation is not clear for the present
systems at this point.
(11) (a) Wayland, B. B.; Poszmik, G.; Mukerjee, S. L.; Fryd, M. J .
Am. Chem. Soc. 1994, 116, 7943. (b) Wayland, B. B.; Basickes,
L.; Mukerjee, S. L.; Wei, M.; Fryd, M. Macromolecules 1997,
30, 8109.
(12) (a) Gridnev, A. A.; Ittel, S. D. Chem. Rev. 2001, 101, 3611.
(b) Gridnev, A. J . Polym. Sci., Part A: Polym. Chem. 2000,
38, 1753. (c) Gridnev, A.; Simonsick, W. I., J r.; Ittel. S. D. J .
Polym. Sci., Part A: Polym. Chem. 2000, 38, 1911.
(13) (a) Hall, H. K., J r.; Itoh, T.; Iwatsuki, S.; Padias, A. B.;
Mulvaney, J . E. Macromolecules 1990, 23, 913. (b) Hall, H.
K., J r.; Padias, A. B. Acc. Chem. Res. 1997, 30, 322.
Ack n ow led gm en t. Mr. T. Nishikawa (NAIST) is
acknowledged for his valuable experimental assistances.
MA025857Q