Polymer phase-transition behavior driven by a charge-transfer interaction

Article abstract of DOI:10.1002/anie.201210261

Cool down and get it together! The lower critical solution temperature (LCST) behavior of a polymer containing pyrene units was controlled by a charge-transfer (CT) interaction with an electron-accepting effector (see picture). The appearance of CT bands enabled quantitative evaluation of the thermodynamics of the association. Copyright

Full text of DOI:10.1002/anie.201210261

.
Angewandte
Communications
DOI: 10.1002/anie.201210261
Thermoresponsive Polymers
Polymer Phase-Transition Behavior Driven by a Charge-Transfer
Interaction**
Shogo Amemori, Kenta Kokado, and Kazuki Sada*
Charge-transfer (CT) interactions are intermolecular inter-
actions between p-electron-rich (donor molecules) and p-
electron-deficient species (acceptor molecules). The high
specificity of CT interactions enables the arrangement
between a donor and an acceptor to be controlled, and the
resulting CT complex shows a characteristic absorption band
in the visible region; this absorption band provides informa-
also potential applications in smart materials. As an out-
standing example, Ritter and co-workers reported that
a polymer bearing adamantyl moieties showed LCST behav-
ior in the presence of methylated b-cyclodextrins (b-CD) in
water owing to the dissociation of b-CD from the adamantyl
group in the polymer chain upon heating and the subsequent
association of adamantyl groups through a hydrophobic
[1–4]
[8]
tion on the association of supramolecular complexes.
interaction.
We also demonstrated readily adjustable
Therefore, CT interactions are among the most powerful
tools for the design of supramolecular complexes, such as
organic crystalline materials with superconductivity or con-
LCST behavior based on hydrogen bonding between a poly-
mer bearing urea units and an alcohol as an “effector” for
[
9]
polymer solubility. In both cases, the dissociation of
supramolecular complexes at an increased temperature
triggered a drastic change in solubility of the polymers.
These findings inspired us to explore the applicability of other
weak intermolecular interactions between polymers and
effectors to clarify the molecular design of LCST polymer
chains and external effectors. Herein, we demonstrate the
first example of LCST behavior based on a CT interaction as
the driving force (Figure 1). As the result of the appearance of
a CT band in the UV/Vis absorption spectrum, a quantitative
evaluation of the relationship between LCST behavior and
the formation of supramolecular complexes was possible.
Previously, the distinction between polymer–polymer inter-
actions and polymer–effector (solvent) interactions was
considered problematic in the common case of LCST
behavior driven by hydrogen bonding.
[
1]
[2]
ductivity, low-molecular-weight organic gelators, preor-
[
3]
ganized building blocks for rotaxane, and supramolecular
[
4]
polymers. In many cases, the supramolecular CT complexes
can be collapsed readily by heating, since a CT interaction
provides a bond of modest strength unless a cavity, such as
[4c]
that of cucurbit[8]uril, is used to stabilize the CT complex.
Herein, we demonstrate the precise and facile control of
the lower critical solution temperature (LCST) of a polymer
solution, that is, its phase-transition behavior, by the use of
a CT interaction. The well-known phase-transition behavior
associated with the LCST of a polymer solution occurs as
a result of primitive molecular recognition between the
polymer chains and solvent molecules (e.g., H O). It has
2
attracted broad interest with respect to the development of
stimuli-sensitive materials because the solubility of the
[
5]
polymer chains changes drastically upon heating. However,
one of the major drawbacks of this phenomenon is the
restriction imposed by the required conditions (temperature,
solvent, or pressure); as result, stimuli-sensitive materials are
limited to those based on intrinsic LCST polymers, such as
[6]
poly(N-isopropylamide) (PNIPAM)
glycol) (PEG).
and poly(ethylene
[7]
Sophisticated supramolecular control of the LCST has
been paid much attention, because such studies can uncover
not only a fundamental perspective of the phenomenon but
[
*] S. Amemori, Dr. K. Kokado, Prof. Dr. K. Sada
Graduate School of Chemical Sciences and Engineering
Hokkaido University
Kita 10 Nishi 8, Kita-ku, Sapporo, Hokkaido 060-0810 (Japan)
E-mail: sadatcm@mail.sci.hokudai.ac.jp
Dr. K. Kokado, Prof. Dr. K. Sada
Department of Chemistry, Graduate School of Science
Hokkaido University (Japan)
[
**] We thank the Creative Research Institution of Hokkaido University
for elemental analysis and electron spray ionization mass spec-
troscopy. This work was supported by Grant-in-Aid for Scientific
Research (23350048).
Figure 1. Concept of the LCST behavior of supramolecular complexes
as caused by a CT interaction, and structure of the effectors studied
and the polymer PPMA, which contains pyrene groups as donor
moieties.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201210261.
4
174
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 4174 –4178

Angewandte
Chemie
There are two requirements for LCST behavior: a high
aggregation ability of the polymer and an interaction that can
be cleaved readily upon heating, such as a charge-transfer
interaction. Our approach to the design of such a system
rested on poly((1-pyrene)methyl acrylate) (PPMA), in which
the pyrene units on the polymer side chain induce aggregation
through a p–p interaction that results in reduced solubility of
the polymer, and the electron-accepting molecules 1–5 as
effectors to form CT complexes with PPMA. The long alkyl
chains in 1–5 were expected to change the solubility of PPMA
in nonpolar media, and the relatively low binding constant
between pyrene and the effectors were expected to give rise
to the dissociation of CT complexes upon heating (Figure 1).
PPMA was obtained by the controlled radical polymeri-
zation of monomer 6, which was synthesized by a condensa-
tion reaction of acryloyl chloride with 1-pyrenemethanol
(
Scheme 1). The number-average and weight-average molec-
Scheme 1. Synthetic route to PPMA. AIBN=azobisisobutyronitrile,
DMAc=N,N-dimethylacetamide.
Figure 2. a) UV/Vis spectra of 1 in 1,2-dichloroethane with PPMA at
4
ular weights of PPMA were determined to be M = 2.2 ꢀ 10
various temperatures (solid lines) and without PPMA (dashed line).
b) Transmittance (measured at 800 nm) as a function of temperature
n
4
and M = 4.4 ꢀ 10 , respectively, by size-exclusion chromatog-
w
À1
for PPMA with 1 in 1,2-dichloroethane (scan rate: 18Cmin ). For (a)
raphy with polystyrene standards. The effectors 1–5 with long
alkyl chains were synthesized by the corresponding dehydra-
tion–condensation reaction (see Scheme S1 in the Supporting
Information).
À1
and (b), [1]=0.14m, [PPMA]=10 gL (0.035m with respect to the
pyrene units in the polymer).
À1
First, the solubility of PPMA (10 gL ) in various organic
1 and the pyrene unit in PPMA caused the dissociation of
stacked pyrene groups and the consequent dissolution of
PPMA in 1,2-dichloroethane.
solvents at room temperature was investigated (see Table S1
in the Supporting Information). In some organic solvents,
such as acetonitrile, acetone, ethyl acetate, 1,2-dichloro-
ethane, toluene, and hexane, PPMA was practically insoluble,
mainly because of the strong p–p interaction between the
large aromatic rings on the polymer chain. Thus, a certain
amount of an effector seemed to be necessary to solubilize
PPMA by interruption of the p–p interaction between the
pyrene moieties on the polymer chain.
Figure 2b shows the change in the transmittance of
PPMA in the presence of 1 in 1,2-dichloroethane at various
temperatures. A drastic change in transmittance was observed
at 438C upon heating. The cloud point was determined as the
temperature at which 90% transmittance was observed.
Upon subsequent cooling, transmittance was recovered to
100%, and PPMA dissolved completely again. Repeated
heating–cooling cycles resulted in a repeated change in
transmittance; thus, reversible LCST behavior was observed.
The absorption of the CT band in the UV/Vis spectrum
decreased upon heating to 358C (Figure 2a). These results
indicated that the dissociation of CT complexes between
1 and PPMA at increasing temperatures led to a decrease in
the solubility of PPMA and promoted its aggregation, which
resulted in precipitation. The addition of effectors 2–5
induced similar LCST behavior of PPMA in 1,2-dichloro-
ethane (see Figures S1–S4 in the Supporting Information).
Ethyl acetate and toluene also acted as poor solvents for
PPMA and caused LCST behavior under similar conditions
(see Figures S5 and S6). On the basis of these findings, we
As a poor solvent for PPMA, 1,2-dichloroethane was
selected for the establishment of LCST behavior. When
a solution of 1 (0.14m) in 1,2-dichloroethane was added to
À1
a suspension of PPMA (10 gL , [pyrene unit in PPMA] =
0
.035m) in 1,2-dichloroethane at room temperature, the
suspension dissolved, and the white turbid mixture became
a clear yellow solution. The change in turbidity indicated that
PPMA became soluble in 1,2-dichloroethane upon the
addition of 1, and the yellow color was attributed to the
charge-transfer absorption band observed at 430 nm in the
UV/Vis spectra of PPMA in 1,2-dichloroethane in the
presence of 1 (Figure 2a). These results provided evidence
that the formation of CT complexes between the effector
Angew. Chem. Int. Ed. 2013, 52, 4174 –4178
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4175

.
Angewandte
Communications
suggest that CT interactions can function as efficient inter-
molecular interactions to trigger LCST behavior.
We next investigated the dependence of the cloud point
on the concentration of the effector. When the effector
concentration was increased, the cloud point gradually
increased (Figure 3). Higher temperatures were necessary to
Table 1: Association constant of complexes between PPMA and effectors
1–5 in 1,2-dichloroethane.
À1 [a]
Effector Ka [m
]
Effector
concentration [m] point [8C]
Cloud
Degree of
association [%]
[
b]
[c]
1
2
3
4
5
0.74
0.50
4.23
3.99
4.68
0.119
0.155
0.036
0.032
0.024
25
26
30
28
24
8.1
9.1
10.3
9.1
9.1
[
a] The association constant (K ) was determined by the Benesi–
a
Hildebrand method at 258C. [b] The cloud point was determined by
a transmittance experiment versus temperature at the effector concen-
tration given. [c] The degree of association was defined as the ratio [CT
complex]/[PPMA] at the LCST.
the formation of a CT complex involving 1 and 8.1% of the
pyrene groups in PPMA. Similarly, the other effectors 2–5
formed CT complexes around the cloud points with 6–12% of
the pyrene moieties in the polymer chain (Table 1; see also
Table S2 and Figure S20).
The use of effectors 1 and 2, for which low association
constants were found, required a larger effector concentra-
tion; thus, a critical degree of association appears to exist for
LCST behavior to occur (Table 1). These results showed that
the solubility of PPMA and the cloud points depended on the
degree of association between PPMA and the effector
molecules. The reversible thermoresponsivity can be attrib-
uted to the relatively low association constant of pyrene
groups and effector molecules in the CT interaction. This
study provides the first quantitative description of the
relationship between thermoresponsivity and the formation
of a supramolecular complex; it had been assumed previously
that the hydration number of poly(N-isopropylacrylamide)
Figure 3. Dependence of the cloud point of PPMA in 1,2-dichloro-
ethane on the concentration of the effector for effectors 1–5.
À1
[
PPMA]=10 gL (0.035m with respect to the pyrene units in the
polymer).
rupture the CT complexes at high effector concentrations.
Furthermore, a decline in the cloud point was observed as the
concentration of PPMA was increased (see Figure S7). These
observations revealed that the equilibrium between the CT
complex (between the pyrene groups of PPMA and the
effector) and its dissociated state clearly dominated the
thermoresponsivity and LCST behavior of the polymer.
To elucidate the relationship between the CT interaction
and the thermoresponsivity of PPMA in the presence of the
effectors, we evaluated the association constants (Ka)
between PPMA and effectors 1–5 and the thermodynamic
parameters for the association (DH and DS) by the con-
struction of Benesi–Hildebrand and vanꢁt Hoff plots, respec-
tively (see Figures S8–S17). The linearity of the Benesi–
Hildebrand plots implies the formation of a CT complex with
a 1:1 donor–acceptor composition. All calculated association
[
10]
(PNIPAM) governed its thermoresponsivity.
To further control the LCST behavior of the polymer, we
added 1,2-dimethoxybenzene as a competitive donor and
a good solvent molecule to the ternary mixture of PPMA and
5 in 1,2-dichloroethane. This additive indeed functioned as
a competitive donor: the charge–transfer absorption band of
monomer 6 and effector 5 in 1,2-dichloroethane decreased
upon the addition of 1,2-dimethoxybenzene (see Figure S21).
The addition of a small amount of 1,2-dimethoxybenzene
induced a drastic decrease in the cloud point, although only
1,2-dimethoxybenzene is able to act as a good solvent for
PPMA. This phenomenon can be interpreted as a cononsol-
vency effect (Figure 4a); such an effect was also reported for
À1
constants were relatively low (< 10m ), and they decreased
with an increase in temperature (Table 1; see also Table S2).
The low binding ability of the effectors to the pyrene moieties
on the PPMA chain probably led to the dissociation of the CT
complexes upon heating up to 808C. To gain deeper insight
into the thermoresponsivity, we evaluated the degree of
association (concentration of CT complexes/concentration of
the pyrene unit in PPMA) required for the formation of CT
complexes under our experimental conditions on the basis of
the calculated thermodynamic parameters. For example,
under the conditions of a concentration of 0.119m in 1 and
a ternary system consisting of PNIPAM, H O, and other
2
[
11]
solvents. This result indicated that the molar fraction of 5
accessible to PPMA was diminished owing to a competitive
CT interaction between 1,2-dimethoxybenzene and 5 (Fig-
ure 4c). For example, the lowest effective concentration of 5
decreased from 0.025 to 0.017m upon the addition of 1,2-
dimethoxybenzene (0.70m), and the cloud point decreased
from 26 to 18C. (The effective concentration was calculated
on the basis of the association constant of the complex
between 5 and 1,2-dimethoxybenzene; see Table S4 and
Figures S18 and S19.) The calculated effective concentration
agreed well with the observation that a mixture of 5 (0.017m)
and PPMA showed a cloud point of 28C (without 1,2-
0
2
.035m (for pyrene unit) in PPMA, for which a cloud point of
58C was found, 8.1% of the pyrene units in the polymer
chain were found to form a CT complex with 1. In other
words, PPMA in 1,2-dichloroethane became soluble through
4
176
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 4174 –4178

Angewandte
Chemie
behavior of PPMA can be con-
trolled readily by changing the effec-
tor concentration or structure and
by the further addition of a compet-
itive donor. To our best knowledge,
the design of polymers with LCST
behavior on the basis of a p–p
interaction and a charge-transfer
interaction has not been described
previously. We conclude that proper
selection of the intermolecular inter-
action between the pendant group of
the polymer chain and the effector
by employing knowledge based on
supramolecular
chemistry
can
govern the solubility and thermo-
sensitivity of the polymer solution.
Many other intermolecular interac-
tions should provide a wide range of
LCST behavior in various media and
at various temperatures. We are now
exploring the use of different inter-
actions for the design of smart
thermosensitive systems.
Received: December 24, 2012
Published online: March 7, 2013
Keywords: charge transfer ·
.
phase transitions · polymers ·
supramolecular complexes ·
thermodynamics
Figure 4. a,b) Cloud point (^ =LCST) of PPMA in 1,2-dichloroethane in the presence of 5 versus the
À1
concentration of 1,2-dimethoxybenzene (a) or 1,1,2,2-tetrachloroethane (b). [PPMA]=10 gL
(
[
1] a) F. Wudl, H. Yamochi, T. Suzuki,
H. Isotalo, C. Fite, H. Kasmai, K.
Liou, G. Srdanov, J. Am. Chem. Soc.
À1
À1
0.035m with respect to the pyrene units in the polymer), [5]=0.025m , scan rate=18Cmin . The
dashed lines show the assumed boundaries between types of thermoresponsivity. c) Presumed
mechanism of the competitive effect.
1990, 112, 2461 – 2462; b) G. Saito,
H. Sasaki, T. Aoki, Y. Yoshida, A.
Otsuka, H. Yamochi, O. O. Droz-
dova, K. Yakushi, H. Kitagawa, T.
Mitani, J. Mater. Chem. 2002, 12, 1640 – 1649; c) D. Jꢂrome,
Chem. Rev. 2004, 104, 5565 – 5591.
dimethoxybenzene; Figure S22). The variation in solubility
was caused by the competitive association of 5 with pyrene
groups in the PPMA or with 1,2-dimethoxybenzene. In other
words, the addition of 1,2-dimethoxybenzene diminished the
association between PPMA and 5. Whereas an increase in the
amount of 1,2-dimethoxybenzene resulted in a decrease in the
cloud point, an elevation of the cloud point was observed
upon the addition of 1,1,2,2-tetrachloroethane, which is
known to act as a good solvent for PPMA but not to associate
with 5 (Figure 4b). From these results, we conclude that the
control of the degree of association between the pendant
functional groups in the polymer chain and effectors enables
the solubility of the polymer to be changed as desired.
In summary, we have demonstrated the controllable
LCST behavior of pyrene-containing PPMA in organic
solvents with effectors 1–5 and clearly showed a relationship
between thermosensitivity and a CT interaction. Since the
association between pyrene units and effectors plays a crucial
role in determining the solubility of PPMA, the LCST
[
2] a) U. Maitra, P. V. Kumar, N. Chandra, L. J. DꢁSouza, M. D.
Prasanna, A. R. Raju, Chem. Commun. 1999, 595 – 596; b) A.
Friggeri, O. Gronwald, K. J. C. van Bommel, S. Shinkai, D. N.
Reinhoudt, J. Am. Chem. Soc. 2002, 124, 10754 – 10758; c) P.
Babu, N. M. Sangeetha, P. Vijaykumar, U. Maitra, K. Rissanen,
A. R. Raju, Chem. Eur. J. 2003, 9, 1922 – 1932; d) D. Rizkov, J.
Gun, O. Lev, R. Sicsic, A. Melman, Langmuir 2005, 21, 12130 –
1
2138; e) P. Mukhopadhyay, Y. Iwashita, M. Shirakawa, S.
Kawano, N. Fujita, S. Shinkai, Angew. Chem. 2006, 118, 1622 –
625; Angew. Chem. Int. Ed. 2006, 45, 1592 – 1595; f) K. V. Rao,
1
K. Jayaramulu, T. K. Maji, S. J. George, Angew. Chem. 2010, 122,
4314 – 4318; Angew. Chem. Int. Ed. 2010, 49, 4218 – 4222.
[3] a) L. M. Klivansky, G. Koshkakaryan, D. Cao, Y. Liu, Angew.
Chem. 2009, 121, 4249 – 4253; Angew. Chem. Int. Ed. 2009, 48,
4185 – 4189; b) T. Ikeda, M. Higuchi, D. G. Kurth, J. Am. Chem.
Soc. 2009, 131, 9158 – 9159; c) J.-C. Olsen, A. C. Fahrenbach, A.
Trabolsi, D. C. Friedman, S. K. Dey, C. M. Gothard, A. K.
Shveyd, T. B. Gasa, J. M. Spruell, M. A. Olson, C. Wang, H.-P.
Jacquot de Rouville, Y. Y. Botros, J. F. Stoddart, Org. Biomol.
Chem. 2011, 9, 7126 – 7133.
Angew. Chem. Int. Ed. 2013, 52, 4174 –4178
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4177

.
Angewandte
Communications
[
4] a) J. S. Park, K. Y. Yoon, D. S. Kim, V. M. Lynch, C. W.
Bielawski, K. P. Johnston, J. L. Sessler, Proc. Natl. Acad. Sci.
USA 2011, 108, 20913 – 20917; b) G. Fernꢃndez, E. M. Perez, L.
Sanchez, N. Martin, Angew. Chem. 2008, 120, 1110 – 1113;
Angew. Chem. Int. Ed. 2008, 47, 1094 – 1097; c) Y. Liu, Y. Yu,
J. Gao, Z. Wang, X. Zhang, Angew. Chem. 2010, 122, 6726 – 6729;
Angew. Chem. Int. Ed. 2010, 49, 6576 – 6579.
Fitzpatrick, D. Fournier, J. Lyskawa, M. Nutley, F. Stoffelbach, P.
Woisel, Chem. Commun. 2009, 5266 – 5268; c) U. Rauwald, J.
del Barrio, X. J. Lou, O. A. Scherman, Chem. Commun. 2011, 47,
6000 – 6002.
[7] Z.-X. Zhang, X. Liu, F. J. Xu, X. J. Loh, E.-T. Kang, K.-G. Neoh,
J. Li, Macromolecules 2008, 41, 5967 – 5970.
[8] a) S. Schmitz, H. Ritter, Angew. Chem. 2005, 117, 5803 – 5806;
Angew. Chem. Int. Ed. 2005, 44, 5658 – 5661; b) O. Kretschmann,
C. Steffens, H. Ritter, Angew. Chem. 2007, 119, 2764 – 2767;
Angew. Chem. Int. Ed. 2007, 46, 2708 – 2711.
[
5] a) E. S. Gil, S. M. Hudson, Prog. Polym. Sci. 2004, 29, 1173 –
1222; b) D. Schmaljohann, Adv. Drug Delivery Rev. 2006, 58,
1655 – 1670; c) H. Wei, S.-X. Cheng, X.-Z. Zhang, R.-X. Zhuo,
Prog. Polym. Sci. 2009, 34, 893 – 910; d) C. Weber, R. Hoogen-
boom, U. S. Schubert, Prog. Polym. Sci. 2012, 37, 686.
[9] S. Amemori, K. Kokado, K. Sada, J. Am. Chem. Soc. 2012, 134,
8344 – 8347.
[
6] a) O. Kretschmann, S. W. Choi, M. Miyauchi, I. Tomatsu, A.
Harada, H. Ritter, Angew. Chem. 2006, 118, 4468 – 4472; Angew.
Chem. Int. Ed. 2006, 45, 4361 – 4365; b) J. Bigot, M. Bria, S. T.
Caldwell, F. Cazaux, A. Cooper, B. Charleux, G. Cooke, B.
[10] Y. Ono, T, Shikata, J. Am. Chem. Soc. 2006, 128, 10030 – 10031.
[11] a) H. G. Schild, M. Muthukumar, D. A. Tirrell, Macromolecules
1991, 24, 948 – 952; b) H. Yamauchi, Y. Maeda, J. Phys. Chem. B
2007, 111, 12964 – 12968.
4
178
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 4174 –4178