Development of an arylthiobismuthine cocatalyst in organobismuthine- mediated living radical polymerization: Applications for synthesis of ultrahigh molecular weight polystyrenes and polyacrylates

Article abstract of DOI:10.1021/ja8092899

Diphenyl(2,6-dimesitylphenylthio)bismuthine (1a) serves as an excellent cocatalyst in organo-bismuthine-mediated living radical polymerization (BIRP). Both low and high molecular weight polystyrenes and poly(butyl acrylate)s (PBAs) with controlled molecul

Full text of DOI:10.1021/ja8092899

Published on Web 01/22/2009
Development of an Arylthiobismuthine Cocatalyst in
Organobismuthine-Mediated Living Radical Polymerization:
Applications for Synthesis of Ultrahigh Molecular Weight
Polystyrenes and Polyacrylates
Eiichi Kayahara and Shigeru Yamago*
Institute for Chemical Research, Kyoto UniVersity, Uji, Kyoto 611-0011, Japan
Received November 28, 2008; E-mail: yamago@scl.kyoto-u.ac.jp
Abstract: Diphenyl(2,6-dimesitylphenylthio)bismuthine (1a) serves as an excellent cocatalyst in organo-
bismuthine-mediated living radical polymerization (BIRP). Both low and high molecular weight polystyrenes
and poly(butyl acrylate)s (PBAs) with controlled molecular weights and low polydispersity indexes (PDIs)
were synthesized by the addition of a catalytic amount of 1a to an organobismuthine chain-transfer agent,
methyl 2-dimethylbismuthanyl-2-methylpropionate (3). The number-average molecular weight (Mn) of the
resulting polymers increases linearly with the monomer/3 ratio. Structurally well-defined polystyrenes with
Mn’s in the range from 1.0 × 104 to 2.0 × 105 and PDIs of 1.07-1.15 as well as PBAs with Mn’s in the
range from 1.2 × 104 to 2.8 × 106 and PDIs of 1.06-1.43 were successfully prepared under mild thermal
conditions. Control experiments suggested that 1a reversibly reacts with the polymer-end radical to generate
an organobismuthine dormant species and 2,6-dimesitylphenylthiyl radical (2a). This reaction avoids the
occurrence of chain termination reactions involving the polymer-end radicals and avoids undesired loss of
the bismuthanyl polymer end group. The bulky 2,6-dimesitylphenyl group attached to the sulfur atom may
prevent the addition of thiyl radicals to the vinyl monomers to generate new polymer chains.
ganobismuthine-mediated17 LRPs (TERP, SBRP, and BIRP,
respectively), which have proven to be powerful methods.18,19
Introduction
Living radical polymerization (LRP) has become an indis-
pensable method for the synthesis of well-defined, advanced
polymeric materials because it allows the polymerization of a
variety of functional vinyl monomers to give products with well-
controlled molecular weights and molecular weight distribu-
tions.1-3 LRPs that have been widely used include nitroxide-
mediated radical polymerization,4 atom-transfer radical polym-
erization (ATRP),5-7 and reversible addition-fragmentation
chain-transfer radical polymerization (RAFT).8,9 We have also
developed organotellurium-,10-13 organostibine-,14-16 and or-
New variants of LRP have also emerged, such as cobalt-
mediated polymerization,20-26 single-electron-transfer (SET)
LRP,27-30 titanium-catalyzed polymerization,31 and reversible
chain-transfer catalyzed polymerization.32,33
(13) Sugihara, Y.; Kagawa, Y.; Yamago, S.; Okubo, M. Macromolecules
2007, 40, 9208–9211.
(14) Yamago, S.; Ray, B.; Iida, K.; Yoshida, J.; Tada, T.; Yoshizawa, K.;
Kwak, Y.; Goto, A.; Fukuda, T. J. Am. Chem. Soc. 2004, 126, 13908–
13909.
(15) Ray, B.; Kotani, M.; Yamago, S. Macromolecules 2006, 39, 5259–
5265.
(1) Handbook of Radical Polymerization; Matyjaszewski, K., Davis, T. P.,
Eds.; Wiley-Interscience: New York, 2002.
(2) Moad, G.; Solomon, D. H. The Chemistry of Radical Polymerization;
Elsevier: Amsterdam, 2006.
(16) Yamago, S.; Yamada, T.; Togai, M.; Ukai, Y.; Kayahara, E.; Pan, N.
Chem.sEur. J. 2009, 15, 1018–1029.
(17) Yamago, S.; Kayahara, E.; Kotani, M.; Ray, B.; Kwak, Y.; Goto, A.;
Fukuda, T. Angew. Chem., Int. Ed. 2007, 46, 1304–1306.
(18) Yamago, S. J. Polym. Sci., Part A: Polym. Chem. 2005, 44, 1–12.
(19) Yamago, S. Proc. Jpn. Acad., Ser. B 2005, 81, 117–128.
(20) Gridnev, A. A.; Ittel, S. D. Chem. ReV. 2001, 101, 3611–3660.
(21) Wayland, B. B.; Poszmik, G.; Mukerjee, S. L. J. Am. Chem. Soc. 1994,
116, 7943–7944.
(3) Braunecker, W. A.; Matyjaszewski, K. Prog. Polym. Sci. 2007, 32,
93–146.
(4) Hawker, C. J.; Bosman, A. W.; Harth, E. Chem. ReV. 2001, 101, 3661–
3688.
(5) Matyjaszewski, K.; Xia, J. Chem. ReV. 2001, 101, 2921–2990.
(6) Ouchi, M.; Terashima, T.; Sawamoto, M. Acc. Chem. Res. 2008, 41,
1120–1132.
(22) Debuigne, A.; Caille, J.-R.; Je´roˆme, R. Angew. Chem., Int. Ed. 2005,
44, 1101–1104.
(7) Kamigaito, M.; Ando, T.; Sawamoto, M. Chem. ReV. 2001, 101, 3689–
3746.
(23) Debuigne, A.; Caille, J.-R.; Je´roˆme, R. Macromolecules 2005, 38,
5452–5458.
(8) Moad, G.; Rizzardo, E.; Thang, S. H. Polymer 2008, 49, 1079–1131.
(9) Lowe, A. B.; McCormick, C. L. Prog. Polym. Sci. 2007, 32, 283–
351.
(24) Debuigne, A.; Caille, J.-R.; Willet, N.; Je´roˆme, R. Macromolecules
2005, 38, 9488–9496.
(25) Bryaskova, R.; Detrembleur, C.; Debuigne, A.; Je´roˆme, R. Macro-
molecules 2006, 39, 8263–8268.
(10) Yamago, S.; Iida, K.; Yoshida, J. J. Am. Chem. Soc. 2002, 124, 2874–
2875.
(26) Debuigne, A.; Willet, N.; Je´roˆme, R.; Detrembleur, C. Macromolecules
2007, 40, 7111–7118.
(11) Yamago, S.; Iida, K.; Yoshida, J. J. Am. Chem. Soc. 2002, 124, 13666–
13667.
(27) Percec, V.; Popov, A. V.; Ramirez-Castillo, E.; Monteiro, M.; Barboiu,
B.; Weichold, O.; Asandei, A. D.; Mitchell, C. M. J. Am. Chem. Soc.
2002, 124, 4940–4941.
(12) Yusa, S.; Yamago, S.; Sugahara, M.; Morikawa, S.; Yamamoto, T.;
Morishima, Y. Macromolecules 2007, 40, 5907–5915.
9
2508 J. AM. CHEM. SOC. 2009, 131, 2508–2513
10.1021/ja8092899 CCC: $40.75
2009 American Chemical Society

Development of an Arylthiobismuthine Cocatalyst
A R T I C L E S
Scheme 1. Mechanisms of Activation/Deactivation of Polymer-End
Radical P · by (A) Ditelluride (TeR)2 or Distibine (SbR2)2 and (b)
Organotellurium (P-TeR), Organostibine (P-SbR2), or
Organobismuthine (P-BiR2) Dormant Species
Despite these developments, a significant drawback of LRP
is the synthesis of high molecular weight polymers. Because
the polymer-end radical is always subject to irreversible
termination reactions,34 dead polymers accumulate in the
reaction mixture as the targeted molecular weight increases.
Nevertheless, a few examples of the synthesis of high molecular
weight polyacrylates and polymethacrylates have been reported,
with number-average molecular weights (Mn’s) sometimes
exceeding 1 × 106, which usually corresponds to a degree of
polymerization of more than 104. For example, ultrahigh
molecular weight poly(methyl methacrylate), with Mn’s in the
range (1.3-3.6) × 106 and low polydispersity indexes (PDIs)
of less than 1.3 have been prepared by RAFT35 and ATRP36,37
under high-pressure conditions. However, this method requires
special apparatus and is hence synthetically unattractive. The
only examples under ambient conditions reported to date are
the syntheses of poly[2-(dimethylamino)ethyl methacrylate] with
Mn ) 8.5 × 105 by ATRP,38 poly(methyl acrylate) with Mn )
1.4 × 106 by SET LRP using a copper catalyst,28 and poly(butyl
methacrylate) with Mn ) 1.0 × 106 by ATRP under miniemul-
sion conditions.39 The synthesis of high molecular weight
polystyrene is more difficult than that of polymethacrylates
because of the existence of autoinitiation and the slow propaga-
tion rate. The synthesis of polystyrene by RAFT with an Mn of
2 × 105 and a PDI of 1.12 has been reported, but it was
performed under high-pressure conditions.40 Ultrahigh molecular
weight polymethacrylates and polystyrenes with Mn’s exceeding
107 were synthesized by plasma-initiated polymerization, but
its synthetic efficiency and generality were quite limited.41
Therefore, the development of a new method that allows the
synthesis of high molecular weight polymers under ambient
conditions is desirable but also a significant challenge.
be suitable for the synthesis of high molecular weight polymers
in a controlled manner.
We have already reported that the addition of ditellurides11
and distibines16 effectively increases the PDI control that can
be achieved in TERP and SBRP, respectively, for the polym-
erization of styrene and methyl methacrylate (MMA). Kinetic
studies of the effect of ditelluride revealed that it acts as a
capping reagent for the polymer-end radical (P·) through a
homolytic substitution reaction, forming a dormant species
P-TeR (Scheme 1a).42 The liberated tellanyl radical (RTe ·) is
essentially inert toward the monomers but reacts with the
organotellurium dormant species to regenerate the polymer-end
radical and the ditelluride. In contrast, deactivation of the
polymer-end radical to the dormant species in the absence of
ditelluride proceeds by a homolytic substitution reaction with
the organotellurium dormant species, that is, a degenerative
transfer reaction (Scheme 1b).43,44 In the case of the methyl-
tellanyl group (X ) TeMe), the rate of deactivation by the
ditelluride is ∼100 times faster than that by the organotellurium
dormant species in the polymerization of styrene, and this is
the origin of the increased PDI control achieved by the addition
of ditellurides. We envisioned that faster deactivation would
decrease the occurrence of undesired termination reactions of
the polymer-end radicals and that this type of diheteroatom
compound is thus potentially useful as a cocatalyst for the
synthesis of high molecular weight polymers.
During the course of developing new LRP methods, we found
that BIRP allows better control of PDI than TERP and SBRP
for the synthesis of low molecular weight polymers (Mn < 1 ×
105). However, when the target molecular weight of the
polymers synthesized by BIRP was increased, we observed an
apparent loss of PDI control. We also observed the precipitation
of black particles, presumably bismuth metal, indicating the loss
of the organobismuthine polymer end group. Therefore, if the
loss of this polymer end group could be avoided, BIRP would
By analogy with the effects of ditellurides and distibines,
dibismuthines would be a formidable cocatalyst for BIRP.
However, these compounds are thermally labile. For example,
tetramethyldibismuthine decomposes even at room tempera-
ture,45 and thus, dibismuthines cannot be used as cocatalysts in
this process. We focused on the arylthiobismuthines 1 as
potential cocatalysts (Scheme 2). Barton et al.46 have already
reported that tris(phenylthio)bismuthine generates a benzenethiyl
radical upon reaction with carbon-centered radicals. Therefore,
we expected that 1 should react with polymer-end radicals to
form an organobismuthine dormant species and an arylthiyl
radical 2. We have recently found that arylthiyl radicals are
highly reactive toward organobismuthines, generating carbon-
centered radicals through homolytic substitution reactions.47
Therefore, radical 2 should react with the organobismuthine
(28) Percec, V.; Guliashvili, T.; Ladislaw, J. S.; Wistrand, A.; Stjemdahl,
A.; Sienkowska, M. J.; Monteiro, M.; Sahoo, S. J. Am. Chem. Soc.
2006, 128, 14156–14165.
(29) Lligadas, G.; Rosen, B. M.; Monteiro, M. J.; Percec, V. Macromol-
ecules 2008, 41, 8360–8364.
(30) Lligadas, G.; Rosen, B. M.; Bell, C. A.; Monteiro, M. J.; Percec, V.
Macromolecules 2008, 41, 8365–8371.
(31) Asandei, A. D.; Moran, I. W. J. Am. Chem. Soc. 2004, 126, 15932–
15933.
(32) Goto, A.; Zushi, H.; Hirai, N.; Wakada, T.; Tsujii, Y.; Fukuda, T.
J. Am. Chem. Soc. 2007, 129, 13347–13354.
(33) Goto, A.; Hirai, N.; Wakada, T.; Nagasawa, K.; Tsujii, Y.; Fukuda,
T. Macromolecules 2008, 41, 6261–6264.
(34) Fukuda, T.; Terauchi, T.; Goto, A.; Ohno, K.; Tsujii, Y.; Miyamoto,
M.; Kobatake, S.; Yamada, B. Macromolecules 1996, 29, 6393–6398.
(35) Rzayev, J.; Penelle, J. Angew. Chem., Int. Ed. 2004, 43, 1691–1694.
(36) Kwiatkowski, P.; Jurczak, J.; Pietrasik, J.; Jakubowski, W.; Mueller,
L.; Matyjaszewski, K. Macromolecules 2008, 41, 1067–1069.
(37) Arita, T.; Kayama, Y.; Ohno, K.; Tsujii, Y.; Fukuda, T. Polymer 2008,
49, 2426–2429.
(42) Kwak, Y.; Tezuka, M.; Goto, A.; Fukuda, T.; Yamago, S. Macro-
molecules 2007, 40, 1881–1885.
(43) Goto, A.; Kwak, Y.; Fukuda, T.; Yamago, S.; Iida, K.; Nakajima,
M.; Yoshida, J. J. Am. Chem. Soc. 2003, 125, 8720–8721.
(44) Kwak, Y.; Goto, A.; Fukuda, T.; Kobayashi, Y.; Yamago, S.
Macromolecules 2006, 39, 4671–4679.
(38) Mao, B. W.; Gan, L. H.; Gan, Y. Y. Polymer 2006, 47, 3017–3020.
(39) Simms, R. W.; Cunningham, M. F. Macromolecules 2007, 40, 860–
866.
(45) Ashe, A. J., III; Ludwig, E. G., Jr.; Oleksyszyn, J. Organometallics
1983, 2, 1859–1866.
(40) Arita, T.; Buback, M.; Janssen, O.; Vana, P. Macromol. Rapid
Commun. 2004, 25, 1376–1381.
(46) Barton, D. H. R.; Bridon, D.; Zard, S. Z. Tetrahedron 1989, 45, 2615–
2626.
(41) Simionescu, C. I.; Simionescu, B. C. Pure Appl. Chem. 1984, 56, 427–
438.
(47) Yamago, S.; Matsumoto, A. J. Org. Chem. 2008, 73, 7300–7304.
9
J. AM. CHEM. SOC. VOL. 131, NO. 7, 2009 2509

A R T I C L E S
Kayahara and Yamago
Scheme 3. Synthesis of Arylthiobismuthine Derivatives
Scheme 2. Hypothetical Reaction of Arylthiobismuthines with a
Polymer-End Radical P ·
Table 1. Effect of Arylthiobismuthines 1 on Polymerization of
Styrenea
dormant species to regenerate the polymer-end radical (P ·) and
1. Thiyl radicals are also reactive toward alkenes,48 and the
degree of control over the LRP would diminish if 2 were to
react with the monomers. Therefore, we decided to use an
arylthiyl radical bearing the sterically bulky mesityl group (Mes
) 2,4,6-Me3C6H2) at both the 2 and 6 positions. We chose this
substituent because of its well-known steric shielding effect as
well as the availability of 2,6-dimesitylbenzenethiol, the precur-
sor of 1a (R1 ) Ph, R2 ) Mes).49,50
We report here on the synthesis of arylthiobismuthine
cocatalysts 1 and their effect on BIRP. We show that BIRP
becomes more controllable in the presence of a catalytic amount
of 1a, allowing the synthesis of high molecular weight
polystyrenes (Mn ≈ 2 × 105) and ultrahigh molecular weight
polyacrylates (Mn ≈ 3 × 106) with well-controlled molecular
weight distributions, starting from monofunctional organobis-
muthine chain-transfer agents such as methyl 2-dimethylbis-
muthanyl-2-methylpropionate (3)17 (eq 1):
equiv of
styrene
time conv.
(%)b
c
run
cocatalyst (equiv)
(h)
Mn,theor
Mn,exp
PDIc
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
100
200
500
700
1500
100
1a (0.2)
1a (0.5)
1a (1.0)
1b (0.2)
none
(Ph2Bi)2 (0.2)
1a (0.2)
1a (0.5)
1a (1.0)
none
1a (0.2)
1a (0.2)
1a (0.2)
1a (0.2)
1a (0.2)
1b (0.2)
1b (0.2)
1a (0.2)
84
84
84
84
84
84
96
96
96
96
24
32
60
65
90
24
60
84
98 102 000 103 000 1.11
83
82
88
96
75
86 400
85 400
91 500
99 800
78 100
85 500 1.13
80 500 1.17
90 200 1.27
81 900 1.21
69 200 1.31
94 196 000 198 000 1.15
94 167 000 147 000 1.15
81 169 000 130 000 1.22
55 115 000 107 000 1.34
96
99
10 000
20 600
51 600
72 800
10 900 1.07
24 700 1.09
55 500 1.10
79 500 1.11
99
100
96 150 000 139 000 1.13
100
100
86
10 400
52 000
89 600
13 600 1.13
54 700 1.18
84 600 1.13
500
1000
a A mixture of 3, 1 (0-1.0 equiv), AIBN (0.2 equiv), and styrene
(100-2000 equiv) was heated under a nitrogen atmosphere. b Monomer
conversion was determined by 1H NMR. c Number-average molecular
weight (Mn) and polydispersity index (PDI ) Mw/Mn) were obtained by
size exclusion chromatography, which was calibrated with polystyrene
standards, using CHCl3 as the eluent.
This is the first example in which the synthetic utility of
thiobismuthines in controlling radical reactions and radical
polymerization has been demonstrated. On the basis of several
control experiments, we also propose the mechanism for the
action of 1 shown in Scheme 2. This new mechanism for
deactivation of polymer-end radicals in addition to the conven-
tional degenerative transfer of organobismuthines (Scheme 1b,
X ) BiR2) should be responsible for the successful synthesis
of ultrahigh molecular weight polymers.
amorphous solid, some impurities could not be completely
removed. Therefore, we used 1c only for mechanistic studies,
because the degree of control achievable in LRP is sometimes
strongly influenced by small amounts of an impurity.
Effect of 1 on Polymerization of Styrene by BIRP. We first
examined the effects of 1a on the polymerization of styrene
using organobismuthine 3 as a chain-transfer agent (Table 1).
When a mixture of 3, 2,2′-azobis(isobutyronitrile) (AIBN) (0.2
equiv), 1a (0.2 equiv), and styrene (1000 equiv) was heated at
60 °C under a nitrogen atmosphere in a glovebox, polystyrene
formed with quantitative monomer conversion, as judged by
1H NMR spectroscopy (Table 1, run 1). The slightly yellowish
crude product (Figure 1a) was dissolved in chloroform, and
precipitation from methanol gave polystyrene as a white powder.
Analysis by gel permeation chromatography (GPC) indicated
the formation of highly controlled polystyrene with an Mn of
103 000, which is close to the value calculated from the
monomer/3 ratio (Mn,theor ) 102 000), and a low PDI of 1.11.
When the amount of 1a was increased to 0.5 and 1.0 equiv, we
also obtained highly controlled polystyrenes with low PDIs
(1.13-1.14) as colorless products (Table 1, runs 2 and 3).
However, the degree of control achieved, as indicated by the
Mn and PDI values, was slightly lower than that attained in the
presence of 0.2 equiv of 1a.
Results and Discussion
Synthesis of Arylthiobismuthines 1. 2,6-Dimesitylbenzeneth-
iol50 was treated with n-butyllithium (1.1 equiv) in THF at -78
°C followed by diphenylbismuthanyl bromide51 (Scheme 3). The
desired thiobismuthine 1a was isolated in 69% yield by silica
gel chromatography followed by recrystallization as a yellow
crystal that is stable toward air and moisture. Diphenyl(phe-
nylthio)bismuthine (1b), which has no sterically bulky substit-
uents on the arylthiyl group, was prepared in 75% yield starting
from benzenethiol. Synthesis of dimethyl(2,6-dimesitylphe-
nylthio)bismuthine (1c) was also carried out. Though 1c was
obtained as the major component in the form of an air-sensitive
(48) Sulfur-Centered Radicals; Bertrand, M. P., Ferreri, C., Eds.; Wiley-
VCH: Weinheim, Germany, 2001; Vol. 2.
The addition of 0.2 equiv of 1b also resulted in the formation
of a colorless product, but the polymerization was less controlled
than that in the presence of 1a in terms of both Mn and PDI
(Table 1, run 4). This result clearly indicates the importance of
using a sterically bulky mesityl group as a substituent. Polym-
(49) Power, P. P. J. Organomet. Chem. 2004, 689, 3904–3919.
(50) Ellison, J. J.; Ruhlandt-Senge, K.; Power, P. P. Angew. Chem., Int.
Ed. 1994, 33, 1178–1180.
(51) Clegg, W.; Errington, R. J.; Fisher, G. A.; Hockless, D. C. R.; Norman,
N. C.; Orpen, A. G.; Stratford, S. E. J. Chem. Soc., Dalton Trans.
1992, 1967–1974.
9
2510 J. AM. CHEM. SOC. VOL. 131, NO. 7, 2009

Development of an Arylthiobismuthine Cocatalyst
A R T I C L E S
Figure 1. Effect of 1a on the color of polystyrene crude product prepared
(a) in the presence of 1a (Table 1, run 1) and (b) in its absence (Table 1,
run 5).
Figure 2. Correlation between the degree of polymerization (DP) of styrene
and the number-average molecular weight (Mn) and polydispersity index
(PDI ) Mw/Mn) of the resulting polystyrenes, for the bulk polymerization
of styrene with 3 at 60 °C in the presence or absence of 1a.
erization under the same reaction conditions without the addition
of 1a resulted in a high monomer conversion (96%), but a black
precipitate appeared in the reaction mixture (Figure 1b).
Furthermore, the Mn of the resulting polymer was ∼20% lower
than the theoretical value, and the PDI was larger than that of
the product formed in the presence of 1a (Table 1, run 5). The
effects of tetraphenyldibismuthine, which is thermally more
stable than the tetramethyl derivative,41 were also investigated.
However, the reaction mixture became black, and no improve-
ment of the PDI control was observed (run 6). These results
clearly indicate that the addition of 1a considerably suppresses
the loss of the organobismuthanyl polymer end group and
increases the control over the Mn and PDI of the product.
The effects of 1a were further investigated by carrying out
the polymerization of 2000 equiv of styrene (Table 1, runs
7-10). Polymerization in the presence of 1a (0.2-1.0 equiv)
resulted in high monomer conversions and gave colorless
polymers in all cases (Table 1, runs 7-9). Highly controlled
polystyrenes with low PDIs and Mn’s close to the theoretical
values were obtained in all cases. The use of 0.2 equiv of 1a
was most effective in attaining a high monomer conversion, a
high degree of control of the Mn, and a low PDI. Polystyrene
with an Mn of 2.0 × 105 and a PDI of 1.15 was obtained under
optimized conditions (run 7). In contrast, polymerization carried
out in the absence of 1a (Table 1, run 10) did not reach a high
monomer conversion after 96 h, and it afforded polystyrene with
a rather high PDI of 1.34. The formation of a black precipitate
was also observed in the reaction mixture.
The Mn of the polystyrene could be controlled by the styrene/3
ratio but not the styrene/1a ratio, and thus, low (Mn ≈ 10 000)
and high molecular weight polystyrenes could be selectively
synthesized by varying the styrene/3 ratio in the presence of
0.2 equiv of 1a (Table 1, runs 11-15, and Figure 2). These
results clearly indicate that 1a does not initiate the polymeri-
zation. The effects of 1a on the Mn and PDI of the product
were marginal for the synthesis of low molecular weight
polystyrenes (Mn < 50 000), but it became more significant as
the targeted molecular weight increased (Mn > 50 000). The
Mn’s for polystyrene prepared in the presence of 1a linearly
increased with the amount of styrene used, as shown by the
filled squares in Figure 2, and were all close to the corresponding
theoretical values. This linear correlation is consistent with the
living character of the polymerization. GPC traces of the
polystyrenes were unimodal in all cases, and the peak maxima
shifted toward higher molecular weight as the styrene/3 ratio
increased (Figure 3).
Figure 3. GPC traces of polystyrenes synthesized by varying the styrene/3
ratio.
carried out by using 100 equiv of styrene. The plot of monomer
conversion versus time shown in Figure 4a indicates that pseudo-
first-order kinetics with respect to the monomer were followed,
implying that the concentration of the radical species remained
constant during the polymerization. The apparent rate constant
(kpapp) was determined to be 0.11 h-1. A linear correlation
between the Mn of the polystyrene and the monomer conversion
was observed (Figure 4b). The PDI was rather high at low
monomer conversion (PDI ) 1.23 at 24% conversion) but
decreased as the polymerization progressed (PDI ) 1.08 at 95%
conversion). Second, a chain-extension experiment was exam-
ined. Thus, polystyrene (Mn ) 8200, PDI ) 1.09) was prepared
from 3 and 100 equiv of styrene in the presence of 0.2 equiv of
1a and AIBN at 60 °C in 85% monomer conversion, and styrene
(100 equiv) was added to the crude reaction mixture. After the
mixture was heated at 60 °C for 18 h, we obtained the chain-
extended polystyrene with Mn ) 19 600 and PDI ) 1.12. GPC
traces clearly revealed that the initially formed polystyrene was
completely converted (see the Supporting Information). All of
these results are consistent with the living character of the
polymerization under the current conditions.
Polymerization of n-Butyl Acrylate by BIRP in the Pres-
ence of a Cocatalyst. We next studied the polymerization of
100 equiv of n-butyl acrylate (BA) by BIRP using chain-transfer
agent 3 and AIBN (0.2 equiv) in the presence of 1a (0.2 equiv)
at 60 °C. The polymerization was complete within 0.5 h, and
the desired poly(butyl acrylate) (PBA) was obtained in 82%
yield. The Mn of 11 600 was close to the theoretical value
The living character of the polymerization was further
confirmed by control experiments. At first, kinetic studies were
9
J. AM. CHEM. SOC. VOL. 131, NO. 7, 2009 2511

A R T I C L E S
Kayahara and Yamago
Figure 4. (a) First-order plot of monomer conversion vs time and (b)
correlation of Mn and PDI with monomer conversion for the bulk
polymerization of styrene (100 equiv) at 60 °C in the presence of 3 and 1a.
Figure 5. GPC traces of PBA samples prepared by varying the BA/3 ratio,
as measured by GPC columns with exclusion limits of (a) 2 × 106 and (b)
2 × 107.
Table 2. Organobismuthine-Mediated Living Radical
Polymerization of BA in the Presence of Cocatalyst 1aa
the exclusion limit of the first system was Mn ) 2 × 106 and
that of the second was 2 × 107. We found that the Mn increased
linearly with the BA/3 ratio, although the monomer conversion
decreased as a result of the increase in viscosity of the reaction
mixture. Highly controlled PBAs with low PDIs and Mn’s close
to their theoretical values were obtained in all cases. The GPC
traces for all of the PBAs were unimodal, and the peak maxima
shifted to higher molecular weight as the targeted molecular
weight increased (Figure 5). It is worth noting that a PBA with
an Mn of 1.4 × 106 and a PDI of 1.22 was obtained at 47%
monomer conversion when 20 000 equiv of BA was employed
(Table 2, run 7). Moreover, a PBA with an Mn of 2.8 × 106
was obtained at 41% monomer conversion when 50 000 equiv
of BA was employed (Table 2, run 8). Although the PDI of the
latter product was slightly larger (1.43) than those of the smaller
molecular weight polymers, it is still acceptable. This product
represents the highest molecular weight polyacrylate synthesized
to date in a controlled manner by LRP from a monofunctional
chain-transfer agent under ambient conditions.
c
run
equiv of BA
time (h)
conv. (%)b
Mn,theor
Mn,exp
PDIc
1
2
3
4
5
100
500
1000
4000
8000
0.5
1.5
2
4
5
82
85
82
72
63
10 500
54 400
105 000
369 000
646 000
11 600
45 400
131 000
295 000
528 000
(491 000) (1.18)
735 000 1.20
(701 000) (1.21)
1.06
1.08
1.12
1.18
1.19
6
10000
6
58
743 000
7
8
20000
50000
7
9
47
41
1 200 000 1 370 000
2 630 000 2 820 000
1.22
1.43
a A mixture of 3, 1a (0.2 equiv), AIBN (0.2 equiv), and BA
(100-50000 equiv) was heated under a nitrogen atmosphere. b Monomer
conversion was determined by 1H NMR. c Number-average molecular
weight (Mn) and polydispersity index (PDI ) Mw/Mn) were obtained by
size exclusion chromatography, which was calibrated using polyMMA
standards. Two linearly connected columns with exclusion limits of 2 ×
106 using CHCl3 as the eluent were used for runs 1-6, and a similar
setup with exclusion limits of 2 × 107 using THF as the eluent was
used for runs 5-8. The data obtained for runs 5 and 6 using the THF
system are shown in parentheses.
Mechanism of Action of the Cocatalyst. The role of cocatalyst
1 was examined. First, a 1:1 mixture of organobismuthine chain-
transfer agent 4 and thiobismuthine 1a was heated in the
presence of AIBN (0.1 equiv) in THF-d8 at 80 °C in an NMR
tube (Scheme 4), and the reaction mixture was monitored by
1H NMR. Equal amounts of 4 and 1a were slowly consumed,
and a new set of signals corresponding to organobismuthine 5
and thiobismuthine 1c appeared. The reaction reached a 4/5 (or
1a/1c) ratio of 8:2 after 3 h, and the ratio did not change further
on prolonged heating. Next, the reverse reaction, starting from
a mixture of 5 and 1c, was examined at 80 °C in the presence
of 0.1 equiv of AIBN. Although the reaction was slow and
calculated from the BA/2 ratio, and the PDI was 1.06 (Table 2,
run 1). The rate of polymerization followed first-order kinetics
with respect to the monomer, and Mn increased linearly with
monomer conversion (see the Supporting Information). These
results are again consistent with the living character of the
polymerization.
We also examined the synthesis of high and ultrahigh
molecular weight PBAs by increasing the BA/3 ratio from 500
to 50 000 in the presence of 0.2 equiv of 1a (Table 2, runs 2-8).
The range of Mn’s obtained for the resulting PBAs was so wide
that two sets of GPC columns were necessary for the analysis:
9
2512 J. AM. CHEM. SOC. VOL. 131, NO. 7, 2009

Development of an Arylthiobismuthine Cocatalyst
A R T I C L E S
Scheme 4. Bismuthanyl-Group Exchange Reaction between Organobismuthine and Thiobismuthine
Scheme 5. Radical-Mediated Exchange Reaction of Bismuthanyl
Groups
of the Bi-S bond toward carbon-centered radicals and the
presence of a sterically bulky aryl group on the sulfur atom of
the cocatalyst are responsible for our success in the synthesis
of high molecular weight polymers. We believe that this work
opens the possibility of designing new heteroatom compounds
for the precise control of radical reactions and living radical
polymerization.
Experimental Section
Preparation of Diphenyl(2,6-dimesitylphenylthio)bismuth-
ine (1a). n-Butyllithium (11.0 mL, 1.55 M solution in hexane, 17
mmol) was slowly added to a solution of 2,6-dimesitylben-
zenethiol50 (5.89 g, 17 mmol) in ether (30 mL) at -78 °C. The
resulting mixture was stirred for 1 h at this temperature and then
slowly warmed to room temperature over 1 h. Diphenylbismuthanyl
bromide51 (6.65 g, 15 mmol) was added to the reaction mixture at
0 °C, after which the mixture was stirred for 2 h at room
temperature. Water was added, and the organic phase was separated.
The organic phase was successively washed with saturated aqueous
NH4Cl solution and saturated aqueous NaCl solution, dried over
MgSO4, and then passed through a pad of Celite. Removal of the
solvent under reduced pressure and recrystallization of the residue
from chloroform/hexane gave 7.32 g of 1a as pale-yellow crystals
further addition of AIBN was required (0.1 equiv at 3, 6, and
9 h), an 8:2 mixture of 4 (or 1a) and 5 (or 1c) was eventually
reached after 11 h. These results indicate that equilibrium was
reached in both cases.
A plausible mechanism to account for this exchange reaction
is shown in Scheme 5. The carbon-centered radical 6 generated
from AIBN reacts with 1a or 1c to give 5 or 4, respectively,
and 2,6-dimesitylphenylthiyl radical (2a). This thiyl radical then
reacts with 5 or 4 to regenerate 6 and 1a or 1c, respectively.
Both the reaction of 6 with 1 and that of 2a with 4 or 5 take
place unselectively, regardless of the substituent on the bismuth
atom (R1 ) Ph or Me). Therefore, the reaction eventually
reaches equilibrium to give a statistical mixture of compounds.
These results also imply that the homolytic substitution
reaction of the polymer-end radical with 1a occurs as shown in
Scheme 2. This effective deactivation of the polymer-end radical
must be responsible for both the increased control of the PDI
and the successful synthesis of high molecular weight polymers
that was possible in the presence of 1a. The results may also
be explained by the formation of a stable (but not persistent)
hypervalent bismuthanyl radical via the reaction of the polymer-
end radical with 1a. However, this scenario is less likely,
because if such a radical formed, retardation of the polymeri-
zation rate would be observed as a result of the occurrence of
chain-breaking termination reactions, as observed in RAFT
polymerization.52-56
1
(69% yield). Mp: 140.0-142.4 °C. H NMR (400 MHz, CDCl3):
δ 2.03 (s, 12H, o-CH3 of mesityl), 2.30 (s, 6H, p-CH3 of mesityl),
6.87 (s, 4H, m-H of mesityl), 7.13 (d, J ) 7.6 Hz, 2H, m-H of
SAr), 7.20 (tt, J ) 7.6, 1.2 Hz, 2H, p-H of BiPh2), 7.26-7.32 (m,
1H, p-H of SAr, 4H, m-H of BiPh2), 7.44 (dt, J ) 6.8, 1.2 Hz, 4H,
o-H of BiPh2). 13C NMR (100 MHz, CDCl3): δ 21.00, 21.12,
127.19, 127.55, 128.28, 128.82, 130.65, 133.84, 135.79, 136.46,
136.92, 139.37, 146.63, 165.51. HRMS (EI) m/z: calcd for
C36H35SBi (M)+, 708.2262; found, 708.2263. IR (neat): 815, 1135,
1185, 1270, 1460, 1695, 2940 cm-1
.
Typical Procedure for Polymerization of Styrene. A solution
of styrene (1.15 mL, 10.0 mmol), 1a (14.2 mg, 0.020 mmol), AIBN
(3.28 mg, 0.020 mmol), and 317 (18.28 µL, 0.10 mmol) was heated
at 60 °C for 24 h with stirring under a nitrogen atmosphere in a
glovebox. A small portion of the reaction mixture was extracted
and dissolved in CDCl3. The conversion of the monomer (96%)
was determined by 1H NMR. The rest of the reaction mixture was
dissolved in a 10 mM TEMPO/CHCl3 solution (4 mL) and poured
into vigorously stirred methanol (200 mL). The product was
collected by filtration and dried under reduced pressure at 40 °C to
give 0.99 g of polystyrene. The Mn (10 900) and PDI (1.07) were
determined by GPC calibrated using polystyrene standards.
Conclusion
We have shown that highly controlled BIRP is possible by
carrying out the polymerization in the presence of a newly
designed arylthiobismuthine cocatalyst 1a. High molecular
weight polystyrenes and ultrahigh molecular weight poly(butyl
acrylate)s have been synthesized in a controlled manner in the
presence of this cocatalyst. In addition to its synthetic utility,
this work also provides insight into the fundamental reactivities
of diheteroatom compounds in radical reactions. The reactivity
Acknowledgment. This work was partly supported by a Grant-
in-Aid for Scientific Research from the Ministry of Education,
Culture and Sports and the Japan Science and Technology Agency.
The authors thank Dr. Kunihiko Sugo of Otsuka Chemical Co. for
useful discussions.
(52) Monteiro, M.; de Brouwer, H. Macromolecules 2001, 34, 349–352.
(53) Kwak, Y.; Goto, A.; Tsujii, Y.; Murata, Y.; Komatsu, K.; Fukuda, T.
Macromolecules 2002, 35, 3026–3029.
Supporting Information Available: Synthesis of 1b, 1c, and
4; typical procedure and kinetic plots for polymerization of butyl
(54) Calitz, F. M.; McLeary, J. B.; McKenzie, J. M.; Tonge, M. P.;
Klumperman, B.; Sanderson, R. D. Macromolecules 2003, 36, 9687–
9690.
1
acrylates; and H and 13C NMR spectra for all of the new
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
(55) Kwak, Y.; Goto, A.; Komatsu, M.; Sugiura, Y.; Fukuda, T. Macro-
molecules 2004, 37, 4434–4440.
(56) Feldermann, A.; Coote, M. L.; Stenzel, M. H.; Davis, T. P.; Barner-
Kowollik, C. J. Am. Chem. Soc. 2004, 126, 15915–15923.
JA8092899
9
J. AM. CHEM. SOC. VOL. 131, NO. 7, 2009 2513