Poly(vinyl ketone)s by controlled boron group transfer polymerization (BGTP)

Article abstract of DOI:10.1002/anie.200907163

(Figure Presented) Boron can do it! Readily prepared catecholboron enolates act as regulators in the controlled radical polymerization of alkyl and aryl vinyl ketones. Control is achieved by an unprecedented reversible radical boron group transfer (BGT) between enoyl radicals (see scheme). Polymers with M _n<= 14000 g mol^-1 and a polydispersity index of 1.3 can be obtained by this process.

Full text of DOI:10.1002/anie.200907163

Angewandte
Chemie
DOI: 10.1002/anie.200907163
Controlled Radical Polymerization
Poly(vinyl ketone)s by Controlled Boron Group Transfer
Polymerization (BGTP)**
Kazuhiro Uehara, Christine B. Wagner, Thomas Vogler, Heinrich Luftmann, and Armido Studer*
[
8]
Controlled radical polymerizations have been intensively
studied during the past ten years, and several methods have
been developed: ATRP (atom-transfer radical polymeri-
key importance for this selectivity. As a-enoyl radicals carry
spin density at oxygen, we assumed that reversible inter-
change of the catecholboron moiety between an a-enoyl
radical and a catecholboron ketone enolate might be feasible.
To develop a boron group transfer radical polymerization
(BGTP), this degenerate transfer process has to be kinetically
competent with the addition of the a-enoyl radical to the
monomer present in solution [Eq. (2)]. In our concept, the
boron group transfer requires spin density at oxygen, while
the growth of the polymer chain relies on the spin density at
the carbon atom of the a-enoyl radical.
[1]
zation), RAFT (reversible addition–fragmentation chain-
[
2]
[3]
transfer polymerization), iodo, tellurium, antimony, and
[
4]
[5]
bismuth group transfer, cobalt-mediated processes, and
[
6]
NMP (nitroxide-mediated polymerization). Various poly-
mers with defined molecular weight and polydispersity (PDI)
below the theoretical limit (1.5) can be conveniently prepared
using these techniques. To our knowledge, only one report on
the successful controlled radical polymerization of vinyl
[
7]
ketones has appeared. Herein, we will introduce a new
method for controlled polymerization of vinyl ketones that is
[
8,9]
based on a radical boron group transfer process.
We recently showed that formal homolytic substitution at
boron in catecholboron enolates 1 with the 2,2,6,6-tetrame-
[10]
thylpiperidine-1-oxyl radical (TEMPO)
delivers reso-
nance-stabilized a-enoyl radicals 2 that undergo TEMPO
[
11]
trapping to afford a-oxygenated ketones 3 [Eq. (1)].
We tested our hypothesis with enolate 4a as a polymer-
ization regulator, which was prepared in situ by the reaction
of the corresponding a,b-unsaturated ketone with catechol-
[
11]
borane. NMR spectroscopy studies revealed that 4a was
formed within one hour at room temperature as a single
isomer, which was tentatively assigned as E isomer (over 95%
yield; see the Supporting Information). Initial polymerization
experiments were conducted with methyl vinyl ketone
(
(
MVK) and V-70 as a stable, low-temperature radical initiator
Scheme 1). Enolate 4a, MVK, and V-70 in dry THF
It is well known that organocatecholboron compounds
react only with heteroatom- and not with C-centered radicals.
Interaction of the Lewis acidic boron atom with an electron
lone pair of the heteroatom-centered radical seems to be of
(50 vol%) were heated to 708C, and the resulting polymer
[*] Dr. K. Uehara, C. B. Wagner, Dr. T. Vogler, Dr. H. Luftmann,
Prof. Dr. A. Studer
Organisch-Chemisches Institut, Westfꢀlische Wilhelms-Universitꢀt
Corrensstrasse 40, 48149 Mꢁnster (Germany)
Fax: (+49)281-833-6523
E-mail: studer@uni-muenster.de
[
**] We thank the DFG for supporting our work. Wako Chemicals is
acknowledged for the donation of V-70 and Tyler W. Wilson for
assistance in the preparation of the manuscript.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200907163.
Scheme 1. Polymerization of various a,b-unsaturated ketones.
Angew. Chem. Int. Ed. 2010, 49, 3073 –3076
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3073

Communications
precipitated upon addition of Et O/pentane. Mean molecular
upon further increasing the monomer/initiator ratio to 200/1.
For these experiments, the PDI slightly increased to around
1.3 (Table 1, entries 9 and 10).
In the optimization of the reaction, replacing THF by
toluene had a detrimental effect on the polymerization and
resulted in lower molecular weights and worse reaction
control (Table 1, entry 11; see discussion below). Varying
ratios of regulator/initiator, at constant monomer concentra-
tion, only showed an effect on yield (Table 1, entries 9, 12, 13).
2
weight (M ) and polydispersity index values were determined
n
by gel permeation chromatography (GPC) relative to linear
poly(methyl methacrylate) standards. The results are sum-
marized in Table 1.
Table 1: Polymerization of MVK with 4a (1.3 equiv) under different
conditions to give PMVK 5.
Entry MVK
V-70
t
Yield M (calcd)
M (GPC) PDI
n
n
À1
À1
Interestingly, M was not altered to a large extent in these
(
equiv) (equiv) [h] [%] [gmol
]
[gmol
]
n
experiments, and thus a ratio of 1.3:1 for 4a/V-70 was used for
the following investigations. Replacing V-70 with a,a’-azobis-
isobutyronitrile (AIBN) as a radical initiator delivered lower
polymer yields compared to the V-70 initiated reactions
(Table 1, entries 14 and 15), and lowering the amount of V-70
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
50
1.0
1.0
1.0
1.0
1.0
–
1
6
1
6
6
1
1
73
74
72
70
97
2110
2130
4020
3910
5370
–
2800
2700
5600
5400
7800
–
1.27
1.24
1.28
1.27
2.21
–
50
100
100
100
100
100
100
200
200
200
200
200
100
100
100
100
100
100
100
100
[
a]
<2
<2
to 0.2 equiv provided a lower conversion, higher M , and
n
[
b]
c]
1.0
1.0
1.0
1.0
1.0
–
–
–
larger PDI values (Table 1, entry 16). Optimization of the
concentration in the reaction was not successful: when the
reaction was conducted under more dilute conditions, a lower
yield and molecular weight was obtained (Table 1, entry 17),
whereas at higher concentration, little effect on the polymer-
[
12 39
4350
7630
7420
5850
6490
7730
3140
3680
5530
2710
4390
4390
3800
1650
9300
11900
13700
8300
12800
13000
5500
8100
7200
3300
5700
6500
5500
1600
1.44
1.33
1.32
1.45
1.32
1.32
1.30
1.44
1.49
1.27
1.35
1.42
1.37
1.26
1
6
6
1
1
6
69
67
53
59
70
57
0
1
2
3
4
5
6
7
8
9
0
1
[
d]
0.75
1.5
[
12]
ization was observed (Table 1, entry 18).
[
e]
1.0
1.0
0.2
1.0
1.0
1.0
1.0
1.0
We then studied what effect the Lewis acidity of the boron
atom in 4 had on the polymerization. With the tert-butyl-
substituted regulator 4b, PMVK with a slightly larger PDI
was formed (Table 1, entry 19). The more electrophilic fluoro-
substituted derivative 4c did not improve the result (Table 1,
entry 20), whereas the methoxy-substituted congener 4d led
to a sharp decrease in yield (Table 1, entry 21).
[
e]
12 67
1
6
6
6
6
6
20
48
79
79
68
28
[
[
[
[
[
f]
g]
h]
i]
j]
Next, we tested whether our new method supports the
controlled polymerization of aryl vinyl ketones (Table 2).
Pleasingly, with phenyl vinyl ketone, polymers with a PDI of
[
a] 4a was not added. [b] Galvinoxyl radical (5 equiv, 5 mol% with respect
to MVK) was added. [c] Conducted at 308C. [d] Toluene was used as a
solvent. [e] Conducted with AIBN. [f] MVK in THF at 10 vol%. [g] Con-
ducted in neat MVK. [h] Conducted with 4b. [i] Conducted with 4c.
1
.3 were obtained (Table 2, entry 1). Increasing the reaction
[j] Conducted with 4d.
time to 6 h did not change the result, thus indicating that
polymerization was complete under the applied conditions
within one hour (Table 2, entry 2). Polymerization in absence
of 4a afforded a large PDI, which again clearly shows the
ability of 4a to act as a regulator in these processes (Table 2,
entry 3). However, control was not perfect for a larger
We were pleased to observe that 4a indeed regulates the
polymerization of MVK and provides direct validation of our
hypothesis. A low PDI of 1.27 clearly demonstrated a
controlled process when the reaction was conducted for 1 h
with 50-fold excess of MVK over V-70 (Table 1, entry 1).
Polymerization was fast, as extending the reaction time to 6 h
delivered a similar yield and molecular weight (Table 1,
entry 2). Polymers with larger Mn values were readily
obtained by increasing the monomer/initiator ratio to 100/1.
Importantly, the polymerization remained controlled with
unchanged PDI (Table 1, entries 3 and 4). As expected, the
polymerization was not controlled in the absence of regulator
targeted M (Table 2, entry 4). Gratifyingly, p-methoxyphenyl
n
vinyl ketone and p-bromophenyl vinyl ketone could be
conveniently polymerized by BGTP (Table 2, entries 5–8).
Renaud and co-workers showed that alkyl catecholboron
derivatives undergo efficient formal homolytic substitution at
Table 2: Polymerization of various aryl vinyl ketones with 4a (1.3 equiv)
and V-70 (1 equiv) in THF (50 vol%) to give 6–8.
4
a and under otherwise identical conditions, which is clearly
Entry Ketone
R
t
Yield Mn
Mn
PDI
shown in the significant increase of the PDI above the
theoretical limit of 1.5 (Table 1, entry 5). Moreover,
attempted polymerization in the absence of V-70 did not
work, thus supporting the radical nature of the process
(equiv)
[h] [%] (calc)
(GPC)
[gmol ]
À1
À1
[gmol
]
1
2
3
4
5
6
7
8
50
50
50
Ph
Ph
Ph
Ph
1
6
6
1
6
6
6
6
78
78
84
73
86
72
71
79
4100
4100
4410
7560
5500
4490
5870
6520
7000
6700
10300
12800
8400
15900
6900
18300
1.33
1.33
3.15
1.53
1.39
2.92
1.43
4.22
(
Table 1, entry 6). Along this line, polymerization in the
[a]
presence of the galvinoxyl radical (5 equiv, 5 mol% relative
to MVK) was inhibited, which clearly shows that a radical
mechanism must be operative (Table 1, entry 7). Reaction at
100
50
4-MeOC H
4-MeOC H
4-BrC
4-BrC H
6
4
4
[
a]
a]
50
50
6
3
08C for 12 h provided poly(methyl vinyl ketone) (PMVK) in
6
H
4
[
50
low yield with a rather high Mn value (Table 1, entry 8).
PMVK with M values of up to 14000 gmol was isolated
6
4
À1
n
[a] 4a was not added.
3
074 www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 3073 –3076

Angewandte
Chemie
[8]
boron with a-carbonyl radicals deriving from ketones.
However, analogous reactions with a-carbonyl radicals deriv-
ing from amides and esters were reported to not work. This
result is in agreement with our observations that controlled
polymerization of n-butyl acrylate and N-isopropyl-
acrylamide by using 4a as a regulator was not observed.
The spin density at oxygen in these a-carbonyl radicals is
probably too low to effect formal homolytic substitution at
[
8]
boron with catecholboron enolates.
We showed the controlled character of the 4a-mediated
MVK polymerization by determining conversion as a func-
tion of time and analyzing molecular weight as a function of
monomer conversion (Figure 1). Both plots had the typical
behavior expected for a controlled process, further validating
the experimental results shown in Table 1.
Figure 2. Part of an ESI-MS spectrum of a MVK polymerization
regulated with 4a in benzene.
identified (see the Supporting Information). Furthermore,
1
H NMR spectroscopy allowed to estimate M of the poly-
n
mers independently to the GPC method. The NMR analysis
provided systematically lower Mn values (by about
À1
1
200 gmol ; Supporting Information, Table S1). Decreasing
the amount of V-70 to 0.2 equiv provided the polymer in
lower yield, with a larger M showing that initiation occurred
n
Figure 1. a) Monomer conversion versus time. b) Molecular weight
mostly by the V-70 derived radicals (Table 1, entry 16).
Currently, we do not fully understand why only a very
small fraction of polymers was initiated by the enoyl radical
deriving from 4a. A possible explanation might be the
probable low reactivity of the enoyl radical deriving from
4a towards MVK. As dimerization products of two of the
enoyl radicals deriving from 4a were not identified, it seems
that these enoyl radicals are able to efficiently undergo formal
homolytic substitution at boron with a polymeric B-enolate to
regenerate 4a. This process keeps the concentration of the
4a-derived enoyl radicals low and does not lead to chain
termination. In fact, after polymerization and precipitation,
we isolated from solution 42% of p-methoxyphenethyl
phenyl ketone that probably derived from hydrolysis of 4a.
Importantly, for reactions performed in benzene, a peak series
assigned to an aldol condensation product (~) was identified
and PDI versus monomer conversion (conditions: MVK (200 equiv),
4
a (1.3 equiv), V-70 (1 equiv), THF, 708C).
&
M , c linear fit, ~ PDI.
n
To understand the new polymerization concept in more
detail and to support the suggested mechanism depicted in
Eq. (2)], we performed mechanistic studies by using HR-ESI
mass spectrometry. Polymerizations with 4a (MVK/4a =
0:1) were conducted in either benzene or in THF. ESI-
[
2
mass spectrometric analysis showed a clean polymerization
process with two (in THF) or three main peak series (in
benzene) separated by one monomer mass (Figure 2). These
three series could be clearly identified, and they differ in the
radical chain initiation and termination. Only a small fraction
of polymers carry the enoyl radical, which is derived from 4a,
as the initiating moiety in their backbone (series indicated by
a * in Figure 2). Thus, most of the polymers were initiated by
the more reactive radical derived from the V-70 (~ and ^).
(Figure 2). This aldol-terminated polymer was derived from a
catecholboron enolate by intramolecular aldol condensa-
1
[13]
This fact was further supported by H NMR spectroscopic
tion. This side reaction strongly supports our suggestion
analysis of the polymers: a signal for the methoxy group of the
enoyl moiety was lacking, and only the resonance of the
methoxy substituent of the V-70-derived unit was clearly
that the growing polymer chain carries a catecholboron
enolate moiety. Interestingly, polymeric aldol condensation
products were identified only in traces for reactions con-
[
14]
Angew. Chem. Int. Ed. 2010, 49, 3073 –3076
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3075

Communications
ducted in THF (see the Supporting Information), which is
[3] G. David, C. Boyer, J. Tonnar, B. Ameduri, P. Lacroix-Desmazes,
B. Boutevin, Chem. Rev. 2006, 106, 3936.
likely to be the reason why a better polymerization control
was achieved in THF as compared to the reactions conducted
in toluene (see above).
[
4] a) S. Yamago, J. Polym. Sci. Part A 2006, 44, 1; b) S. Yamago, E.
Kayahara, M. Kotani, B. Ray, Y. Kwak, A. Goto, T. Fukuda,
Angew. Chem. 2007, 119, 1326; Angew. Chem. Int. Ed. 2007, 46,
To further support our mechanism, we investigated the
1304; c) Y. Kwak, M. Tezuka, A. Goto, T. Fukuda, S. Yamago,
1
thermal stability of the B-enolate using H NMR spectros-
Macromolecules 2007, 40, 1881; d) S. Yamago, Chem. Rev. 2009,
109, 5051.
copy. At 708C, we did not see any decomposition of the B-
enolate, which indicates that reversible O–B homolysis does
not occur and thus a mechanism in analogy to the NMP
[5] a) B. B. Wayland, G. Poszmik, S. L. Mukerjee, M. Fryd, J. Am.
Chem. Soc. 1994, 116, 7943; b) H. Kaneyoshi, K. Matyjaszewski,
Macromolecules 2006, 39, 2757 and references therein; c) R.
Poli, Angew. Chem. 2006, 118, 5180; Angew. Chem. Int. Ed. 2006,
[
6]
process, in which reversible radical generation by homolysis
controls the process, is most likely not operative. This was
further supported by the experiment lacking the V-70 initiator
45, 5058; d) A. Debuigne, R. Poli, C. Jerome, C. Detrembleur,
Prog. Polym. Sci. 2009, 34, 211; e) A. Debuigne, J.-R. Caille, R.
Jerome, Angew. Chem. 2005, 117, 1125; Angew. Chem. Int. Ed.
2005, 44, 1101.
(
see Table 1, entry 6), where polymerization did not occur.
We also checked whether the catecholboron moiety can be
transferred to the a-cyanoalkyl radical derived from V-70. To
this end, a mixture of 4a and V-70 was heated to 708C in THF
for 1 h. Product analysis revealed a 86% yield of p-methox-
yphenethyl phenyl ketone resulting from hydrolysis of 4a
along with 67% of 2,3-bis-(2-methoxy-2-methylpropyl)-2,3-
dimethylsuccinonitrile, which derived from in-cage dimeriza-
tion of the initiating radicals. The reduced V-70-derived
radical (4-methoxy-2,4-dimethylpentanenitrile), which could
be formed by either disproportionation or by boron-group
transfer followed by hydrolysis, was identified as being only
present in traces by GC-analysis. Therefore, boron-group
transfer from 4a to the V-70-derived a-cyanoalkyl radical can
be ruled out.
[
6] a) C. J. Hawker, A. W. Bosman, E. Harth, Chem. Rev. 2001, 101,
3661; b) V. Sciannamea, R. Jꢀrꢁme, C. Detrembleur, Chem. Rev.
2008, 108, 1104; c) A. Studer, Chem. Eur. J. 2001, 7, 1159; d) A.
Studer, Chem. Soc. Rev. 2004, 33, 267; e) A. Studer, T. Schulte,
Chem. Rec. 2005, 5, 27; f) M. K. Brinks, A. Studer, Macromol.
Rapid Commun. 2009, 30, 1043.
[7] ATRP is not working for this polymer class, see: a) A. Mittal, S.
Sivaram, D. Baskaran, Macromolecules 2006, 39, 5555. For a
successful RAFT polymerization, see: b) C. Cheng, G. Sun, E.
Khoshdel, K. L. Wooley, J. Am. Chem. Soc. 2007, 129, 10086.
[
8] Boron compounds in radical chemistry: a) C. Ollivier, P.
Renaud, Chem. Rev. 2001, 101, 3415; b) A.-P. Schaffner, P.
Renaud, Eur. J. Org. Chem. 2004, 2291; c) V. Darmency, P.
Renaud, Top. Curr. Chem. 2006, 263, 71.
[9] For a boron-containing alkoxide as polymerization regulator,
In conclusion, we presented a novel method for controlled
radical polymerization of alkyl and aryl vinyl ketones. This
process comprises an unprecedented boron group transfer
reaction as the key step. To date, controlled radical polymer-
ization of this substance class has been limited to the RAFT
process; compared to RAFT-mediated MVK-polymeri-
zations, BGTP is faster and also gives good control over the
reaction. Mass spectrometry studies support the suggested
mechanism and elucidate possible termination processes.
see: T. C. Chung, W. Janvikul, H. L. Lu, J. Am. Chem. Soc. 1996,
118, 705. Boron does modulate the reactivity of the alkoxyl
radical, however, does not directly interact with the radical.
Therefore, this process should not be regarded as a boron group
transfer reaction.
[
10] T. Vogler, A. Studer, Synthesis 2008, 1979.
[11] M. Pouliot, P. Renaud, K. Schenk, A. Studer, T. Vogler, Angew.
Chem. 2009, 121, 6153; Angew. Chem. Int. Ed. 2009, 48, 6037.
[
12] Calculated M values for these experiments were always below
n
the experimentally determined Mn. M (calc) = (MW
ꢂ
n
M
yield[%] ꢂ [M] /[I ] ) + 141, where I is the effective concen-
0
eff
0
eff
tration of the V-70 derived initiating radicals and 141 corre-
sponds to the molecular weight of the initiating moiety. We
assume that about 35% of the radicals that derive from V-70 do
not act as initiating species, and instead undergo in-cage
dimerization. With 1 equiv of V-70, we assume that within 1 h
at 708C around 1.3 equiv of initiating radicals are available for
the polymerization process.
Received: December 18, 2009
Revised: January 20, 2010
Published online: March 22, 2010
Keywords: boron · enolates · mass spectrometry ·
.
polymerization · poly(vinyl ketone)s
[
13] R. R. Huddleston, D. F. Cauble, M. J. Krische, J. Org. Chem.
2003, 68, 11.
[
1] a) N. V. Tsarevsky, K. Matyjaszewski, Chem. Rev. 2007, 107,
[14] Unfortunately, all attempts to oxidize the polymeric boron
enolate with TEMPO failed so far. We recently reported (see
Ref. [11]) that in B-enolate formation/TEMPO-trapping process
about 10% of reduced enone was always formed. The mecha-
nism of that reduction is currently still not understood.
2270; b) M. Ouchi, T. Terashima, M. Sawamoto, Chem. Rev.
2009, 109, 4963.
[
2] a) A. Favier, M.-T. Charreyre, Macromol. Rapid Commun. 2006,
27, 653; b) C. Barner-Kowollik, S. Perrier, J. Polym. Sci. Part A
2008, 46, 5715; c) G. Moad, E. Rizzardo, S. H. Thang, Acc. Chem.
Res. 2008, 41, 1133.
3
076 www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 3073 –3076