The Effect of Viscosity on the Diffusion and Termination Reaction of Organic Radical Pairs

Article abstract of DOI:10.1002/chem.201902074

The effect of viscosity on the diffusion efficiency (F_dif) of an organic radical pair in a solvent cage and the termination mechanism, that is, the selectivity of disproportionation (Disp) and combination (Comb) of the geminated caged radical pair and the diffused radicals encountered, were investigated quantitatively by following the photolysis of dimethyl 2,2′-azobis(2-methylpropionate) (V-601) in the absence and presence of PhSD. F_dif and Disp/Comb selectivity outside the cage [Disp_(dif)/Comb_(dif)] are highly sensitive to the viscosity. In contrast, the Disp/Comb selectivity inside the cage [Disp_(cage)/Comb_(cage)] is rather insensitive. The difference in viscosity dependence between Disp_(cage)/Comb_(cage) and Disp_(dif)/Comb_(dif) is explained by the spin state of the radical pair inside and outside the cage and the spin state dependent configurational changes of the radical pair upon their collision. Given that the configurational change of the radicals associates the displacement and reorganization of solvents around the radicals, the termination outside the cage, which requires larger change than that inside the cage, is highly viscosity dependent. Furthermore, while the bulk viscosity of each solvent shows good correlation with F_dif and Disp/Comb selectivity, microviscosity is the better parameter predicting F_dif and Disp_(dif)/Comb_(dif) selectivity regardless of the solvents.

Full text of DOI:10.1002/chem.201902074

A Journal of
Accepted Article
Title: The Effect of Viscosity on the Diffusion and Termination Reaction
of Organic Radical Pairs
Authors: Xiaopei Li, Tasuku Ogihara, Manabu Abe, Yasuyuki
Nakamura, and Shigeru Yamago
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201902074
Link to VoR: http://dx.doi.org/10.1002/chem.201902074
Supported by

10.1002/chem.201902074
Chemistry - A European Journal
COMMUNICATION
The Effect of Viscosity on the Diffusion and Termination Reaction
of Organic Radical Pairs
Xiaopei Li, Tasuku Ogihara, Manabu Abe, Yasuyuki Nakamura,* and Shigeru Yamago*
Abstract: The effect of viscosity on the diffusion efficiency (Fdif) of an
organic radical pair in a solvent cage and the termination mechanism,
i.e., the selectivity of disproportionation (Disp) and combination
(Comb) of the geminated caged radical pair and the diffused radicals
encountered, were investigated quantitatively by the photolysis of
dimethyl 2,2'-azobis(2-methylpropionate) (V-601) in the absence and
presence of PhSD. Fdif and Disp/Comb selectivity outside the cage
(Disp(dif)/Comb(dif)) are highly sensitive to the viscosity. In contrast, the
Disp/Comb selectivity inside the cage, Disp(cage)/Comb(cage), is rather
insensitive. The difference in viscosity dependence between
Disp(cage)/Comb(cage) and Disp(dif)/Comb(dif) is explained by the spin state
of the radical pair inside and outside the cage and the spin state
dependent configurational changes of the radical pair upon their
collision. As the configurational change of the radicals associates the
displacement and reorganization of solvents around the radicals, the
termination outside the cage which requires larger change than that
inside the cage is highly viscosity dependent. Furthermore, while the
bulk viscosity of each solvent shows good correlation with Fdif and
Disp/Comb selectivity, microviscosity is the better parameter
predicting Fdif and Disp(dif)/Comb(dif) selectivity regardless of the
solvents.
Fdif tends to decrease as viscosity increases and temperature
decreases.[7,8] However, despite the widespread use of azo
initiators in radical polymerization, no quantitative data have been
reported.
CN
R
R
R
ν
R
R
h
N
N
+
+
+
N
- N2
H
R
R
R
1
3(cage) 4(cage)
5(cage)
6(cage)
R = CN
2(cage)
a
b
: R = CN (AIBN)
: R = CO
Caged radical pair
2Me (V-601)
Diffusion
CN
N
R
Monomer
no trapping
agent
R
R
Polymer
R
+
+
+
2
H
R
R
3(dif)
4(dif)
5(dif)
6(dif)
R = CN
PhSD
2(dif)
Diffused radical
D
3-D
Scheme 1. Generation of radicals from azo initiators and the fate of the
generated radicals in the presence or absence of the monomer and radical
trapping agent PhSD
Azo initiators are also used as model compounds for
clarifying the termination reaction of radical polymerization.[9–12]
The termination of the propagating polymer-end radical
undergoes disproportionation (Disp) and combination (Comb) to
yield dead polymers, which correspond to 3 and 4 for Disp, 5 for
Comb, and 6 for the C-N combination (Comb(C-N)) when EWG =
CN in the model system. As the molecular weight and end-group
structure affect the physical properties of polymers, an accurate
understanding of the termination mechanism; i.e., Disp/Comb
selectivity is a significant importance. However, the method would
have inherent problems because while both caged and diffused
radicals generate the same termination products, 3, 4, 5, and 6,
the selectivity among Disp, Comb, and Comb(C-N) would be
different.
The cage effect and the termination mechanism of radical pairs
are important issues for a long time in various fields from physical
chemistry, organic and polymer synthesis, and further to
biochemisty.[1] While the effect of several factors, in particular
temperature and viscosity, are qualitatively known to affect the
cage effect and the termination mechanism, their quantification
has been a significant challenge.
For example, azo initiators 1, such as AIBN (1a) and V-601
(1b), which generate a pair of carbon-centered radicals 2 with the
elimination of nitrogen under thermal and photochemical
conditions, are widely used for radical polymerization, which is the
most important industrial polymerization technique.[2,3] However,
the initiation efficiency is far below the theoretical yield because a
certain amount of the initially generated caged radical 2(cage)
terminates inside the solvent cage before diffusing to the bulk
solution as the diffused radical 2(dif). The diffusion efficiency (Fdif)
of 1a is approximately 60% under conventional conditions,[3–6] and
Tanner has already reported the significantly different
viscosity effect on the Disp/Comb selectivity of the caged and
diffused t-butyl radicals generated from the corresponding azo
compound. Disp increases as the viscosity increases. However,
the
Disp/Comb
selectivity
of
the
caged
radical
(Disp(cage)/Comb(cage)) reached a plateau by increasing the
viscosity, while that of the diffused radical (Disp(dif)/C(dif)) did not,
and the difference becomes particularly significant in highly
viscous media.[13] However, the impact of the Disp/Comb
selectivity of caged and diffused radicals having polar functional
groups, i.e., 2, which mimics the polymethacrylate end radical,
has not been reported so far. Very recently, Matyjaszewski and
Poli studied the viscosity effect on the termination reaction of
radical 2b generated from an azo compound, V-601 (1b). While
they observed marginal viscosity effects on the Disp/Comb
selectivity (Scheme 1, R = CO2Me),[14] no cage effect was
considered.
[*] X. Li, T. Ogihara, Prof. Dr. S.Yamago
Institute for Chemical Research, Kyoto University
Uji, Kyoto 611-0011 (Japan)
E-mail: yamago@scl.kyoto-u.ac.jp
Prof. Dr. M. Abe
Department of Chemistry, Graduate School of Science, Hiroshima
University
1-3-1 Kagamiyama, Higashi-Hiroshima, 739-8526 (Japan)
Dr. Y. Nakamura
Research and Services Division of Materials Data and Integrated
System, National Institute for Material Science
1-2-1 Sengen, Tsukuba Ibaraki 305-0047 (Japan)
E-mail: NAKAMURA.Yasuyuki@nims.go.jp
Supporting information for this article is given via a link at the end of the
document
This article is protected by copyright. All rights reserved.

10.1002/chem.201902074
Chemistry - A European Journal
COMMUNICATION
We have recently studied the termination mechanism of
radical polymerization based on product analysis.[15–17] Polymer-
end radicals, such as 7c, with controlled molecular weight are
generated from the corresponding “living” polymer having the
organotellurium-end group 8c.[18,19] They reacted with each other,
yielding the Disp products 9c and 10c and/or the Comb product
at different temperatures according to the work by Tanner
(Scheme 1).[13] As only the diffused radical 2b(dif) is trapped by
PhSD, the diffusion efficiency Fdif (%) was directly obtained by
the amount of deuterated product 3b-D. As the termination only
takes place inside the cage, the Disp/Comb selectivity inside the
cage (Disp(cage)/Comb(cage)) was determined from the amount of
3b(cage), 4b(cage), and 5b(cage). The Disp/Comb selectivity of the
diffused radical 2b(dif) (Disp(dif)/Comb(dif)) was then estimated from
11c (Scheme 2, path
A illustrating the case of methyl
methacrylate [MMA] termination. Note that the structures of 9d,
10d, and 11d are identical to those of 3b, 4b, and 5b,
respectively). Two significant conclusions have been drawn from
the analyses of molecular weights and end-group structures: i) the
termination of a polyacrylate-end radical predominantly proceeds
by Disp under ambient temperature, while major textbooks
describe the termination of acrylate polymerization as Comb.[16] ii)
The termination of methacrylate and styrene polymerizations are
significantly changed by the viscosity of the reaction media, and
selective Disp was observed in highly viscous media in both
cases.[17]
the
Disp/Comb
ratio
carried
out
without
PhSD
(Disp(sum)/Comb(sum)), Disp(cage)/Comb(cage), and Fdif.
Recently, Tyler demonstrated that microviscosity, which is
inversely proportional to the translational diffusion coefficient of a
particle in Brownian motion, is a better parameter than bulk
viscosity for correlation with the recombination efficiency of a
radical pair in a solvent cage,[20,21] because the solvent cage effect
is a localized phenomenon. While Cp’Mo(CO3) (Cp’ = η5-
CH3C5H4) radical generated from [Cp’Mo(CO3)]2 by direct Mo-Mo
bond photolysis was used throughout the study, microviscosity
should also be important for a radical pair generated from an azo
initiator. Therefore, the effect of microviscosity of the solvents on
Fdif, Disp(cage)/Comb(cage), and Disp(dif)/Comb(dif) was also examined.
This is the first time that the effect of microviscosity on the reaction
selectivity has been examined. Note that while the relative kinetic
rate constant between kd and kc undergoing Disp and Comb
respectively should be considered for discussing the termination
selectivity, we use the Disp/Comb ratio here because the
termination takes place irreversibly so that the Disp/Comb ratio is
the same as the kd/kc ratio. This study clearly reveals the
importance of viscosity, in particular microviscosity, in
determining the diffusion efficiency Fdif and the Disp/Comb
termination selectivity of the diffused radicals, i.e.,
Disp(dif)/Comb(dif) selectivity.
CO2Me
CO2Me
path A
+
R
R
H
Te
D
C
CO2Me
CO2Me
CO2
Me
ν
-
R'
h
9
10
R
R
R
CO2Me
Te
R'
Te
R'
Diffusion
H
7(cage)
R
R
8
7(dif)
c
: R = PMMA
: R = H
CO2Me
11
(diffused radical )
d
path B
or
7
8
+ R'TeH
10
9
Te
- "
R'"
Scheme 2. Cage effects in the termination reaction of polymethacrylate and its
model radicals starting from an organotellurium compound. Diffusion of the
radical gave termination products (path A) and the tellanyl radical mediated the
abstraction proposed in reference 13 (path B).
The Disp(sum)/Comb(sum) selectivity was obtained by
photolysis of V-601 (0.050 mol L-1 solution) using a 500 W Hg
lamp in C6D6, DMSO-d6, or PEG400 at 25-80 °C in the absence
of PhSD. After complete consumption of V-601, the yields of
In contrast, Matyjaszewski and Poli proposed that the tellanyl
radical-mediated chain transfer pathway is responsible for the
formation of Disp products in our study from the marginal
Disp/Comb selectivity on the viscosity using V-601.[14] They
proposed the tellanyl radical generated from 8c abstracts the β-
hydrogen of 7c in a solvent cage, resulting in 9c and a tellurol,
which further reacts with either 7c or 8c giving 10c (Scheme 2,
path B). They also criticized all our conclusions that the results
were also affected by this tellanyl-radical mediated chain transfer
pathway.
We considered that the apparent contradictions came from
the lack of accurate data for the cage effect of the radicals
generated from azo initiators, i.e., the diffusion efficiency Fdif and
the Disp/Comb selectivity of the termination reaction inside and
outside the cage. Here, we report the detailed analysis of the
effect of viscosity and temperature on these factors. V-601 was
used as a radical precursor, not only because this was used in the
study of Matyjaszewski and Poli but also because it does not
generate the Comb(C-N) product 6, which regenerates radical
2a(cage) inside the solvent cage and further to the diffused radical
2a(dif) under thermal conditions by C-N bond homolysis. When this
reaction occurs, the amount of diffused radical generated
primarily from the azo initiator is overestimated. To clarify the
Disp/Comb selectivity inside and outside the cage, we utilized the
trapping of the diffused radical by PhSD in different solvents and
1
products 3b, 4b and 5b were determined by H NMR analyses
(see Supporting Information, Table S1 and Scheme S1). While
Disp should give a 1:1 mixture of 3b and 4b, the amount of 4b
was always lower than that of 3b, because 4b was further
oligomerized under the experimental conditions. Therefore, the
Disp(sum)/Comb(sum) selectivity was determined from the 3b/5b
ratio (Table 1). The bulk viscosity ηbulk of the solution at each
temperature was independently determined by a viscometer and
the microviscosity ηmicro values were obtained as the viscosity in
the Einstein-Stokes equation by measuring the diffusion constant
by 1H NMR DOSY experiments using methyl isobutylate as a
model compound for radical 2b (Table S2). As all reactions were
carried out at low concentrations of V-601 and they produced low-
mass products, no apparent viscosity change was observed
during the reaction. While a significant decrease in ηbulk was
observed with an increase in temperature, the effect of
temperature on ηmicro was limited, which indicates that the change
in temperature has only a small effect in the local environment of
the radical pair (Figure 1a).
The Disp(sum)/Comb(sum)s were essentially the same as those
reported by Matyjaszewski and Poli,[14] and the small but apparent
effects of viscosity and temperature were observed; Disp(sum)
increases as temperature decreases and viscosity increases in all
This article is protected by copyright. All rights reserved.

10.1002/chem.201902074
Chemistry - A European Journal
COMMUNICATION
Table 1. The effect of bulk and microviscosity on diffusion efficiency Fdif and Disp/Comb termination selectivity for caged and diffused radicals generated from
V-601 (1b)
[a]
[b]
[c]
Run
1
Solvent/Temp.
C6D6/80
ηbulk
0.35
0.41
0.60
1.1
0.82
1.2
1.7
2.3
11
ηmicro
0.081
0.096
0.13
0.17
0.38
0.44
0.52
0.59
0.63
0.78
0.83
0.88
Disp(sum)/Comb(sum)
Fdif (%)
55
Disp(cage)/Comb(cage)
43/57
Disp(dif)/Comb(dif)
45/55
Disp/Comb
46/54
44/56
2
C6D6/60
45/55
53
44/56
46/54
47/53[d]
51/49[d]
59/41[d]
67/33
3
C6D6/40
48/52
48
46/54
50/50
4
C6D6/25
51/49
46
47/53
56/44
5
DMSO-D6/80
DMSO-D6/60
DMSO-D6/40
DMSO-D6/25
PEG400/80
PEG400/60
PEG400/40
PEG400/25
55/45
33
52/48
65/35
6
58/42
30
53/47
67/33
71/29[d]
73/27[d]
76/24[d]
71/29
7
59/41
26
55/45
70/30
8
60/40
20
56/44
73/27
9
55/45
19
49/51
70/30
10
11
12
21
58/42
16
51/49
77/23
78/22[d]
90/10[d]
94/6[d]
44
62/38
11
53/47
86/14
84
66/34
7
57/43
91/9
[a] The Disp/Comb selectivity obtained in the absence of PhSD. [b] The Disp/Comb selectivity of the caged radical obtained by the addition of PhSD. [c] The
Disp/Comb selectivity of the diffused radical estimated from Disp(sum)/Comb(sum), Disp(cage)/Comb(cage), and Fdif. [d] The Disp/Comb selectivity obtained using the
organotellurium compound 8d. Data taken from ref. 16.
solvents (Table 1). The Disp(sum)/Comb(sum) in C6D6 at 80 °C (ηbulk
= 0.35 mPa·s), which is the least viscous solution in this study,
was 44/56 and that in PEG400 at 25 °C (ηbulk = 84 mPa·s), which
was the most viscous condition in this study, was 66/34.
the magnitude was significantly different. For example, the
Disp(cage)/Comb(cage) and Disp(dif)/Comb(dif) in C6D6 at 80 °C are
very similar (43/57 and 45/55, respectively, Table 1, run 1), but
those in C6D6 at 25 °C became significantly different (47/53 and
56/44, respectively, run 4). The difference between
Disp(cage)/Comb(cage) and Disp(dif)/Comb(dif) becomes more
significant in DMSO-d6 and PEG400 (runs 5-12). The
Disp(cage)/Comb(cage) and Disp(dif)/Comb(dif) in PEG400 at 80 °C are
49/51 and 70/30, respectively (run 9), and those at 25 °C are
57/43 and 91/9, respectively (run 12). The results clearly show
that the limited viscosity effect of the Disp(sum)/Comb(sum) selectivity
observed in the previous[14] and this studies is due to the marginal
viscosity effect on the Disp(cage)/Comb(cage) selectivity and high
viscosity dependence of Fdif. The results also demonstrate the
irrelevance of the use of azo compounds for clarifying the
termination mechanism of radical polymerization, as all polymer-
end radicals are diffused radicals.
Next, the same experiments were carried out in the
presence of PhSD. The kinetic isotope effect for the H/D
abstraction (4.8) and the necessary amount of PhSD (15 equiv.)
for quenching all diffused radicals were determined by the
separated experiments (Table S3-S5
and Figure S4).[22]
Furthermore, the consumed PhSD was recovered as (PhS)2
nearly quantitatively (96%) suggesting that no side reactions
mediated by the thiyl radical took place.
After complete consumption of V-601, the amounts of 3b,
1
4b, 5b, and 3b-D were determined by H NMR (Table S7). The
addition of excess PhSD slightly decreased the bulk viscosity,
especially in PEG400, but the difference is negligible in all cases
(Table S8). The diffusion efficiency Fdif in C6D6 at 80 °C was 55%
indicating that nearly half of radical 2b generated terminates
inside the cage before diffusion, and the result agrees well with
the previously reported value.[8,23] The Fdif value decreased with
the increase in viscosity, and only 7% was diffused in PEG400 at
25 °C.
The good correlations of Fdif with temperature and bulk
viscosity in each solvent clearly show the important role of bulk
viscosity (Figures 1a and 1b), but the magnitude of Fdif values
varied by solvents even having the same bulk viscosity. In
contrast, when Fdif was plotted against microviscosity, an
excellent correlation was observed across the different solvents
(Figures 1c). The results clearly demonstrate that microviscosity
is a better parameter than bulk viscosity for the prediction of the
diffusion of radicals from a solvent cage regardless of the radical
precursors.
The Disp/Comb selectivity of the diffused radical,
Disp(dif)/Comb(dif), were calculated from Fdif, Disp(sum)/Comb(sum)
,
and Disp(cage)/Comb(cage) (Table 1). Both Disp(cage)/Comb(cage) and
Disp(dif)/Comb(dif) increased with the increase in bulk viscosity, but
This article is protected by copyright. All rights reserved.

10.1002/chem.201902074
Chemistry - A European Journal
COMMUNICATION
(a)
(b)
(a)
(b)
(c)
Figure 1 Correlations between diffusion efficiency (Fdif) and temperature (a),
bulk viscosity (the inset shows the wider viscosity range) (b), and microviscosity
(c) for radical pair 2b(cage) generated from 1b. The lines and curves are for
guidance.
Figure 2. Correlation of a) bulk and b) microviscosity with Disp/Comb selectivity
in the termination reaction generated from V-601 (1b) and organotellurium
compound 8d. The lines and curves are for guidance.
The log-log plot of both Disp(cage)/Comb(cage) and
Disp(dif)/Comb(dif) vs. bulk viscosity shows excellent linear
correlation, but the dependence also varies according to the
solvents (Figure 2a). In contrast, microviscosity showed
significantly better correlation than bulk viscosity for both
Disp(cage)/Comb(cage) and Disp(dif)/Comb(dif), particularly for the
Disp(dif)/Comb(dif) (Figure 2b). These results demonstrate the
importance of microviscosity for determining not only the diffusion
of radicals from a solvent cage but also the termination
mechanism of radical pairs. The correlation between
Disp(cage)/Comb(cage) and microviscosity was lower than that of
Disp(dif)/Comb(dif), and this observation derives from the difference
in the termination pathways discussed below.
The observed cage effect on the Disp/Comb selectivity can
be explained by considering the spin state of the radical pair,
namely, the spin correlation effect,[24] and the collision model
originally proposed by Fischer and modified by us (Figure 3).[17,25]
The spin state of the caged radical pair should be predominantly
singlet while that of the encountered diffused pair is triplet. For the
diffused radicals, the triplet pair must undergo intersystem
crossing to the singlet through a conical point, in which the radical
pair must take an orthogonal orientation. The radical pair has to
rearrange to the collinear orientation before undergoing the
termination, and this change of radical configuration associates
the displacement and reorganization of the surrounding solvents.
As the transition state undergoing Comb requires greater motion
of the radicals than that of Disp, the former become less favorable
over the latter in viscous solvents.
The Disp(dif)/Comb(dif) values are almost the same as the
Disp/Comb selectivity obtained using organotellurium compound
8d. The results strongly suggest the quantitative generation of
diffused radical 2b from 8d.
In contrast, the singlet radical pair inside the solvent cage
can directly terminate from the initially generated configuration.
Thus, configurational change undergoing both Disp and Comb
should be smaller than that starting from the encountered diffused
radical pair, and this pathway is viscosity insensitive.
This article is protected by copyright. All rights reserved.

10.1002/chem.201902074
Chemistry - A European Journal
COMMUNICATION
CO2Me
Acknowledgements
CO2Me
Diffusion
V-601
- N
MeO2C
1b
(
)
MeO2C
2
This work was supported by the Japan Society for the Promotion
of Science KAKENHI Grant No. 16K05795 (Y.N.) and 16H06352
(S.Y.), and the Collaborative Research Program of the Institute
for Chemical Research, Kyoto University (grant 2018-26). The
calculations in this study were performed on the Numerical
Materials Simulator at NIMS.
Diffused radical pair (triplet)
[random orientation]
Caged radical pair (singlet)
Direct termination
from singlet state!
"Stepwise" termination
from triplet state!
Spin inversion
‡
CO2Me
CO2Me
CO2Me
CO2Me
CO Me
2
or
H
H
H
CO2Me
Keywords: Radical, Termination, Reaction mechanism, Cage
Radical pair at the conical point
(triplet/singlet)
[restricted conformation]
effect, Viscosity effect
Disp
Comb
[1]
[2]
[3]
C. Reichardt, Solvent and Solvent Effects in Organic Chemistry,
WILEY-VCH, Weinheim, 2003.
K. Matyjaszewski, T. P. Davis, Handbook of Radical Polymerization,
Wiley-Interscience, New York, 2002.
G. Moad, D. H. Solomon, The Chemistry of Radical Polymerization,
Elsevier, Amsterdam, 2006.
Scheme 3. The colliding model of the termination reaction of caged and diffused
radicals generated from V-601 (1b)
[4]
[5]
G. Ayrey, Chem. Rev. 1963, 63, 645–667.
This view is further supported by considering the time scale
of radical diffusion. The diffusion of a radical pair in a solvent cage
takes place on an ultrafast time scale (<102 ps), as studied for
iodine,[26] phenylthio,[27] and Cp’Mo(CO)3 radicals.[28] In contrast,
though a radical-radical coupling reaction is recognized as a very
fast process (~108-109 s-1), i.e., with a diffusion-controlled rate, its
time scale is far slower than that of the radical recombination
inside the cage (>1010 s-1).[29,30] Furthermore, Tyler reported that
cage diffusion of the Cp’Mo(CO)3 radical occurs in ~5 ps before
the radical starts to rotate.[28] While radical 2b is structurally less
bulky than the Cp’Co(CO)3 radical, it is not surprising that the
termination of the caged 2b(cage) also takes place before or at the
same time scale of the radical rotation. Therefore, the
Disp(cage)/Comb(cage) is less sensitive to viscosity than the
G. Moad, E. Rizzardo, D. H. Solomon, S. R. Johns, R. I. Willing,
Makromol. Chem., Rapid Commun. 1984, 5, 793–798.
J. K. Fink, J. Polym. Sci., Polym. Chem. Ed. 1983, 21, 1445–1455.
E. Niki, Y. Kamiya, N. Ohta, Bul. Chem. Soc. Jpn. 1969, 42, 3220–
3223.
T. Sato, S. Inui, H. Tanaka, T. Ota, M. Kamachi, K. Tanaka, J. Polym.
Sci. A: Polym. Chem. 1987, 25, 637–652.
S. Bizilj, D. P. Kelly, A. K. Serelis, D. H. Solomon, K. E. White, Aust. J.
Chem. 1985, 38, 1657–1673.
D. J. Trecker, R. S. Foote, J. Org. Chem. 1968, 33, 3527–3534.
V. A. Schreck, A. K. Serelis, D. H. Solomon, Aust. J. Chem. 1989, 42,
375–393.
G. Gleixner, O. F. Olaj, J. W. Breitenbach, Makrmol. Chem. 1979,
180, 2581–2590.
a) D. D. Tanner, P. M. Rahimi, J. Am. Chem. Soc. 1982, 104, 225–
229. b) M. Newcomb, Tetrahedron 1993, 49, 1151-1176.
T. G. Ribelli, S. M. W. Rahaman, K. Matyjaszewski, R. Poli, Chem.
Eur. J. 2017, 23, 13879–13882.
Y. Nakamura, S. Yamago, Macromolecules 2015, 48, 6450–6456.
Y. Nakamura, R. Lee, M. L. Coote, S. Yamago, Macromol. Rapid
Commun. 2016, 37, 506–513.
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
Disp(dif)/Comb(dif)
.
Engel reported the significant increase in the diffusion
efficiency Fdif by generating a triplet radical pair under triplet-
sensitized conditions,[31] and the results are consistent with the
proposed mechanism. However, attempts to generate a triplet
radical pair from V-601 were unsuccessful due to the extremely
low quantum yield of 1b for the sensitized reaction.[32] Further
study is clearly needed for a full understanding of the cage
diffusion and the termination selectivity.
Our results clearly demonstrate that viscosity, particularly
microviscosity is the important parameter for the estimation of
both the diffusion of radicals from a solvent cage and termination
selectivity. These results are reasonable because both the
diffusion and the termination involve transportation and
translocation of molecules in a medium and these events are local
phenomena. We believe that extension of this work would
contribute to clarify important unsolved issues in radical chemistry
and radical polymerization, such as the nature of solvent cages
and the termination mechanism of radical polymerization.
[17]
Y. Nakamura, T. Ogihara, S. Hatano, M. Abe, S. Yamago, Chem. Eur.
J. 2017, 23, 1299–1305.
S. Yamago, Chem. Rev. 2009, 109, 5051–5068.
S. Yamago, K. Iida, J. Yoshida, J. Am. Chem. Soc. 2002, 124, 13666–
13667.
J. T. Barry, D. J. Berg, D. R. Tyler, J. Am. Chem. Soc. 2016, 138,
9389–9392.
J. T. Barry, D. J. Berg, D. R. Tyler, J. Am. Chem. Soc. 2017, 139,
14399–14405.
The validity of the PhSD trapping experiments was also indicated by
the same experiments using TEMPO as a trapping agent under the
thermolysis of V-601 at 80 °C in each solvent (See supporting
infomation, Scheme S5 and Table S6).
M. Stickler, Makromol. Chem. 1986, 187, 1765–1775.
T. Koenig, H. Fischer, in Free Radicals (Ed.: J.K. Kochi), Wiley, New
York, 1973.
H.-H. Schuh, H. Fischer, Helv. Chim. Acta 1978, 61, 2463–2481.
Q. Liu, J.-K. Wang, A. H. Zewail, Nature 1996, 364, 427–430.
T. Bultmann, N. P. Ernsting, J. Phys. Chem. 1996, 100, 19417–19424.
A. B. Oelkers, L. F. Scatena, D. R. Tyler, J. Phys.Chem. A 2007, 111,
5353–5360.
M. J. Gibian, R. C. Corley, Chem. Rev. 1973, 73, 441–464.
D. Griller, in Radical Reaction Rates in Liquids (Ed.: H. Fischer),
Springer-Verlag, Berlin, 1984.
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
P. S. Engel, D. J. Bishop, M. A. Page, J. Am. Chem. Soc. 1978, 100,
7009–7017.
J. R. Fox, G. S. Hammond, J. Am. Chem. Soc. 1964, 86, 4031–4035.
This article is protected by copyright. All rights reserved.

10.1002/chem.201902074
Chemistry - A European Journal
COMMUNICATION
Entry for the Table of Contents (Please choose one layout)
Layout 2:
COMMUNICATION
‡
Xiaopei Li, Tasuku Ogihara, Manabu
Abe, Yasuyuki Nakamura*, Shigeru
Yamago*
R
R
R
hv
R
N
H
N
R
H
N2
H
43/57 - 56/44
R
Marginal
viscosity effect
Diffusion
Disproportionation
= Solvent molecule
,
PEG400)
55 - 7%
Page No. – Page No.
Termination
Strong viscosity effect
D
6, DMSO-d6
(C6
‡
R
The Effect of Viscosity on the
Diffusion and Termination Reaction
of Organic Radical Pairs
R = CO2Me
R
R
45/55 - 90/10
Strong
viscosity effect
R
Combination
Destiny of a radical pair in condensed media: The diffusion efficiency of a
radical pair in a solvent cage and the termination mechanism of a radical pair
outside the cage are significant affected by viscosity, in particular microviscosity.
On the contrary, the termination mechanism of a radical pair inside the cage shows
marginal viscosity effect.
This article is protected by copyright. All rights reserved.