Molecular tailoring of thermoreversible copolymer gels: Some new mechanistic insights

Article abstract of DOI:10.1063/1.476663

We earlier reported the role of hydrophobic and hydrogen bonding interactions on the transition temperatures of thermoreversible copolymer gels. We show here that the chemical structure of the hydrophobe and its concentration determine the transition temperatures [lower critical solution temperature (LCST)] and the heat of transition of new hydrophobically modified poly(N-isopropyl acrylamide) [PNIPAm] copolymer gels. The gels, prepared by copolymerizing NIPAm monomer with hydrophobic comonomers containing increasing lengths of alkyl side groups and a terminal carboxyl acid group, showed lower LCST and lower heat of transition when compared to pure PNIPAm gel. The experimental results were also compared with theoretical calculations based on a lattice-fluid-hydrogen-bond [LFHB] model. We show experimentally and theoretically that a linear correlation exists between the transition temperature and length of the hydrophobic alkyl side group. Also, in apparent contradiction to previous work, we found a reduction in the heat of transition with increasing hydrophobicity. We propose that the presence of the terminal carboxyl acid group on the hydrophobic side chain of the comonomer prevents the association of water molecules around the hydrophobe, thereby causing a reduction in the heat of transition. The LFHB model supports this argument.

Full text of DOI:10.1063/1.476663

Molecular tailoring of thermoreversible copolymer gels: Some new mechanistic
insights
M. V. Badiger, A. K. Lele, V. S. Bhalerao, S. Varghese, and R. A. Mashelkar
Citation: The Journal of Chemical Physics 109, 1175 (1998); doi: 10.1063/1.476663
View online: http://dx.doi.org/10.1063/1.476663
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/109/3?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Inhomogeneity in dehydrated NIPA/SA gel
AIP Conf. Proc. 519, 167 (2000); 10.1063/1.1291534
Molecular dynamics study of the intercalation of diblock copolymers into layered silicates
J. Chem. Phys. 112, 9112 (2000); 10.1063/1.481538
Designing new thermoreversible gels by molecular tailoring of hydrophilic-hydrophobic interactions
J. Chem. Phys. 112, 3063 (2000); 10.1063/1.480881
Equilibrium and non-equilibrium phase transitions in copolymer polyelectrolyte hydrogels
J. Chem. Phys. 107, 1645 (1997); 10.1063/1.474516
Predictions of thermoreversible volume phase transitions in copolymer gels by lattice-fluid-hydrogen-bond theory
J. Chem. Phys. 106, 4768 (1997); 10.1063/1.473474
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.255.116 On: Mon, 01 Dec 2014 12:06:40

JOURNAL OF CHEMICAL PHYSICS
VOLUME 109, NUMBER 3
15 JULY 1998
Molecular tailoring of thermoreversible copolymer gels: Some new
mechanistic insights
M. V. Badiger, A. K. Lele, V. S. Bhalerao, and S. Varghese
Chemical Engineering Division, National Chemical Laboratory, Pune-411 008, India
R. A. Mashelkara)
Council of Scientific and Industrial Research, Anusandhan Bhavan, New Delhi-110 001, India
͑
Received 6 February 1998; accepted 7 April 1998͒
We earlier reported the role of hydrophobic and hydrogen bonding interactions on the transition
temperatures of thermoreversible copolymer gels. We show here that the chemical structure of the
hydrophobe and its concentration determine the transition temperatures ͓lower critical solution
temperature ͑LCST͔͒ and the heat of transition of new hydrophobically modified poly͑N-isopropyl
acrylamide͒ ͓PNIPAm͔ copolymer gels. The gels, prepared by copolymerizing NIPAm monomer
with hydrophobic comonomers containing increasing lengths of alkyl side groups and a terminal
carboxyl acid group, showed lower LCST and lower heat of transition when compared to pure
PNIPAm gel. The experimental results were also compared with theoretical calculations based on a
lattice-fluid-hydrogen-bond ͓LFHB͔ model. We show experimentally and theoretically that a linear
correlation exists between the transition temperature and length of the hydrophobic alkyl side group.
Also, in apparent contradiction to previous work, we found a reduction in the heat of transition with
increasing hydrophobicity. We propose that the presence of the terminal carboxyl acid group on the
hydrophobic side chain of the comonomer prevents the association of water molecules around the
hydrophobe, thereby causing a reduction in the heat of transition. The LFHB model supports this
argument. © 1998 American Institute of Physics. ͓S0021-9606͑98͒50827-6͔
INTRODUCTION
show how the hydrophilic and hydrophobic interactions
could be used to control the transition temperatures of gels.12
The ability to control volume transition temperatures has
significance in synthesizing tailored gels for specific applica-
tions such as novel separation processes. Predictions of our
LFHB model have shown that large changes in the transition
temperatures of thermoreversible gels can be induced by a
small change in the hydrophobic nature of the monomer.
This has been experimentally validated by a ‘‘pin-point
variation’’ of methyl to ethyl groups in L-alanine ester side
Polymeric
hydrogels
such
as
poly͑N-
1
isopropylacrylamide͒ ͓PNIPAm͔ and poly͑vinyl methyl
2
ether͒ ͓PVME͔ are attracting increasing attention lately be-
cause of their characteristic temperature driven volume tran-
sition near the lower critical solution temperature ͑LCST͒.3
The thermoreversible swelling–collapse transitions of gels
have opened up a multitude of innovative applications in the
areas of controlled drug delivery,4
,5
bioseparations,6
7
8
13
robotics, biomedical, and even consumer products, as sum-
chains in hydrogels containing methacryloyl backbone. On
marized recently by Dagani.9
the other hand, we have also shown that the transition tem-
peratures can be controlled by incorporating hydrophobic
comonomers which reduce the LCST, and hydrophilic
comonomers which increase the LCST.
The LCST-type volume transitions of nonionic gels such
as PNIPAm can not be predicted by the Flory theory of poly-
meric networks since they do not consider specific energetic
interactions such as hydrogen bonding, which are predomi-
nantly present in hydrogels. Our previous work using the
lattice-fluid-hydrogen-bond ͑LFHB͒ theory clearly showed10
that the discontinuous volume transition of PNIPAm gel in
water is caused by a large increase in the interpolymer hy-
drogen bonds at the transition temperature and by the tem-
perature dependent effective hydrophobicity of the polymer.
Ours is the first theoretical prediction, which has pointed out
clearly the role of both hydrogen bonding and hydrophobic
interactions. In a series of recent papers we have used the
LFHB model to successfully predict reentrant volume tran-
sition of PNIPAm gel in ethanol–water mixture11 and to
The effect of comonomers has been investigated by vari-
ous groups. Mumick and McCormick14 reported that copoly-
merizing NIPAm with acrylamide ͑AAm͒ resulted in higher
transition temperatures than pure PNIPAm gel. Similarly,
Shibayama et al.15 found that copolymerizing NIPAm with
increasing acrylic acid ͑AAc͒ or dimethylacrylamide
͑DMAAm͒ content gives gels with increasing LCSTs. They
also found that the heat of demixing associated with the vol-
ume transition, which was experimentally measured by the
area under the endothermic peak in differential scanning
calorimetry ͑DSC͒ scans, decreased with the addition of AAc
16
and DMAAm. Feil et al. found that copolymerization of
hydrophilic comonomers such as AAm, AAc, and ͑diethy-
lamino͒ethyl methacrylate ͑DEAEMA͒ increased the LCST
of poly͑NIPAm-co-butyl methacrylate-co-X) terpolymer
a͒Author to whom correspondence should be addressed; electronic mail:
dgcsir@csirhq.ren.nic.in
0021-9606/98/109(3)/1175/10/$15.00
1175
© 1998 American Institute of Physics
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
32.174.255.116 On: Mon, 01 Dec 2014 12:06:40
1

1176
J. Chem. Phys., Vol. 109, No. 3, 15 July 1998
Badiger et al.
gels, while the copolymerization of hydrophobic butyl meth-
acrylate with the terpolymer decreased the LCST of the gels.
Moreover, the heat of the transition decreased with the addi-
tion of the hydrophilic comonomers and increased with the
addition of the hydrophobic comonomer. It has also been
recently reported that the copolymerization of hydrophilic
AAm with hydrophobic N,N-dimethylamino ethyl methacry-
late ͑DMAEMA͒ results in lower LCSTs than pure
which is crucially determined by its chemical structure. This
result provides a deeper mechanistic insight into the nature
of volume transition phenomenon in gels.
THEORY
We will briefly describe the extended LFHB model as
applicable for the binary system of a copolymer gel and wa-
ter. The total free energy of the system is the sum of free
energies of mixing polymer segments with water molecules
and the elastic energy of affine network deformation. The
free energy of mixing is given by the LFHB theory of Panay-
1
7
DMAEMA gels.
The heat of volume transition ⌬H is considered to be
associated with the breakage of hydration layers of water
molecules surrounding the hydrophobic groups on the poly-
mer network. Feil et al.16 proposed that the presence of hy-
drophilic comonomers reduces the concentration of the hy-
drophobic groups and consequently the hydrating water
layers, hence decreasing the ⌬H. The opposite is true for the
presence of hydrophobic comonomers in gels. Inomata
19
iotou and Sanchez, while the elastic free energy ⌬Gel is
given by the theory of rubber elasticity as
2
/3
3
␯e
V
1
V
⌬
G ϭ ͩ ͪ ͩ ͪ Ϫ1Ϫ lnͩ ͪ
,
͑1͒
el
ͭ
ͮ
2
V
V
3
V
0
0
0
where ␯ /V0 is the number density of elastically active
1
8
e
et al. have shown that n-propylacrylamide ͑NPAm͒ gel has
a lower LCST, a higher ⌬H, and a larger magnitude of the
transition than NIPAm and cyclopropylacrylamide ͑CPAm͒
gels. The authors propose that the larger ‘‘contact area’’ of
the n-propyl group as compared to isopropyl and cyclopro-
pyl groups enables the structuring of hydrated water mol-
ecules to a greater extent, which results in lowering of LCST
and increase of ⌬H for the NPAm gels.
chains in as-synthesized gels of volume V , and V is the
0
volume of the gel at given temperature and pressure.
The free energy of mixing is calculated from a mean-
field assumption in which the mers of the two components 1
and 2 are randomly placed on a lattice. For the mixture of n1
molecules of component 1 and n molecules of component 2
2
(
Nϭn ϩn ), the mer length (r), the characteristic mer vol-
1 2
ume (␯*), and the characteristic mean-field mer–mer inter-
action energy (␧*) are given by
It is clear from the foregoing that the balance of hydro-
phobic and hydrophilic groups in the copolymer gels plays a
vital role in determining the LCST. However, it is not yet
fully understood as to how various factors such as the chemi-
cal structure, the bulkiness, and the contact area of the hy-
drophobe affect the LCST of copolymer gels. In this paper
we will investigate the role of the chemical structure of the
hydrophobic comonomer on the volume phase transition and
heat of demixing in PNIPAm gel. We show that, while our
experimental results are in agreement with our earlier theo-
1
␾1
␾2
ϭ
ϩ
,
͑2͒
r
r1
r2
␯
␧
*ϭ␾ ␯*ϩ␾ ␯* ,
͑3͒
͑4͒
1
1
2
2
*ϭ␾ ␧*ϩ␾ ␧*ϩ2␾ ␾ ␨ ͑␧*␧*͒0.5,
2
2
2
1
1
2
1
2
12
1
2
where subscripts 1 and 2 represent the components water ͑1͒
and polymer ͑2͒, respectively. ␾ is the volume fraction, r is
i
1
2
the number of lattice sites occupied by a molecule of com-
retical predictions, they also provide new insights into the
subtle effects of hydrophobic interactions on volume transi-
tions.
ponent i, ␯* is the characteristic mer volume of component
i
i, and ␧* is the characteristic mer–mer interaction energy
i
for component i. A pure component i is thus characterized
We have synthesized three comonomers with increasing
alkyl chain lengths using 4-aminobutyric acid, 6-
aminocaproic acid, and 11-␻, aminoundecanoic acid. The
above compounds were condensed with acryloyl chloride in
the presence of alkali to get monomers, which were copoly-
merized with NIPAm to give new hydrophobically modified
copolymer gels. The structures of the copolymer gels were
by three parameters r , ␯* , and ␧* . ␨ is the binary inter-
i
i
i
12
action parameter between components 1 and 2.
The characteristic parameters of the pure copolymer are
obtained from Eqs. ͑2͒ to ͑4͒ by assuming that the copolymer
can be considered as a random mixture of the two homopoly-
mers A and B, which form the copolymer A–B, and that the
13
binary interaction parameter between the two polymers ␨AB
elucidated by using C cross polarization-magic angle
sample spinning ͑CP-MASS͒ nuclear magnetic resonance
is unity. In using Eqs. ͑2͒ to ͑4͒ for the copolymer, the sub-
scripts 1 and 2 are replaced by A and B representing the two
homopolymers. Thus the theory is essentially applicable only
for a random copolymer A–B. The volume fractions of the
mers of two homopolymers are calculated as
͑
NMR͒ spectroscopy. We observe a reduction in the LCST
as well as a decrease in the heat of demixing of the copoly-
mer gels by increasing the hydrophobic alkyl chain length in
the comonomer or by increasing the content of the hydropho-
bic comonomer in the copolymer. A linear relation was ob-
served between the length of the alkyl chain and the transi-
tion temperature. The reduction in the heat of demixing with
increasing hydrophobicity was surprising. However, experi-
mental results and theoretical calculations using the LFHB
model suggest that while the LCSTs depend on the overall
hydrophobicity of the gel, the heat of demixing depends on
the extent of hydration of water around the hydrophobes,
w /␳*
i
i
␾iϭ
,
͑5͒
͚iw /␳*
i
i
where w is the weight of the component i (iϭA,B) and ␳*
i
i
is its characteristic density.
In a mixture of the copolymer ͑component 2͒ and the
water ͑component 1͒, the binary interaction parameter ␨12 is
calculated as
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.255.116 On: Mon, 01 Dec 2014 12:06:40

J. Chem. Phys., Vol. 109, No. 3, 15 July 1998
Badiger et al.
1177
where,
X ϭ
␨12
ϭ␨1A␾ ϩ␨ ␾ ,
͑6͒
A
1B
B
1/2
where the binary interaction parameters between water and
the two homopolymers A and B are in general given as
␧*ϩ␧*Ϫ2␨ ͑␧*␧*͒
1 2 1 2
12
͑13͒
12
RT
␨1i
ϭa Tϩb ; iϭA,B.
͑7͒
i
i
and ␳ is the reduced density of the mixture which is related
˜
˜
The binary interaction parameter between the copolymer
and the water can be considered as simulating the ‘‘hydro-
phobicity’’ of the copolymer since the hydrophilic part of the
interactions is considered separately in the hydrogen bonding
terms in the mean-field framework of the LFHB theory as
summarized below.
Let there be i (iϭ1,m) types of proton donors and j
jϭ1,n) types of proton acceptors in the system of gel and
solvent. The total number of donors and acceptors is given
by
to the reduced pressure PϭP/P* and reduced temperature
˜
TϭT/T* by the equation of state
m
n
1
2
˜ ˜
˜
␳ ϩPϩT ln͑1Ϫ
˜
␳͒ϩ
˜
␳ 1Ϫ
Ϫ
␯ij
ͭ
ͫ
ͩ
͚ ͚
ͪ
ͬ
ͮ
r
i
j
1/3
␯e
V0
1
2
V0
˜
ϩ
T␯*
ͫ
ͩ ͪ
Ϫ
ͩ ͪ
ͬ
ϭ0.
͑14͒
(
V0
V
V
In the first line of Eq. ͑12͒, the first two terms are the
combinatorial entropy contribution, the third term is the en-
ergetic ͑effective hydrophobic͒ contribution, and the fourth
term in brackets represents the effect of the pure component
properties. The terms in the second line represent the hydro-
gen bonding contribution and the bracketed term in the third
line represents the elastic energy contribution.
i
k
i
j
a
k
N ϭ
d n and N ϭ
a n ,
͑8͒
d
͚
k
͚
j
k
k
k
k
i
k
j
where d and a are the number of donors of type i and
acceptors of type j in component k. If N is the total number
of hydrogen bonds formed between an i– j donor–acceptor
pair, then the number of undonated protons of type i and
unaccepted protons of type j are given by
ij
The equilibrium swelling capacity of the gel, qϭ1/␾2 ,
can be calculated from the condition that the swelling pres-
sure of the gel at equilibrium is equal to zero. Thus,
d
j
G
O
N ϭN Ϫ
Nij and N ϭN Ϫ
N .
͚i ij
͑9͒
␲ϭ␮ Ϫ␮ ϭ0,
͑15͒
io
i
͚j
0j
a
1
1
O
0
ij
0
ij
0
ij
where ␮1 is the chemical potential of pure water outside the
gel. Equation ͑15͒ can be solved by simultaneously solving
Eqs. ͑10͒, ͑12͒, and ͑14͒. These are the main equations con-
stituting the model.
The model parameters for a binary mixture include six
pure component parameters ͑three for each component͒, one
binary interaction parameter (␨ ), three hydrogen bonding
parameters (E ,S ,V ) for each i– j type of H bond, and
the crosslink density (␯ /V ). Parametric values are dis-
cussed in the Results and Discussions section.
The collapse of the gel at the transition temperature oc-
curs by demixing of polymer and water molecules, which is
accompanied by a heat of demixing. This heat of demixing
per gram of dry polymer gel, which is observed as an endot-
hermic peak in a DSC experiment, can be calculated from
the LFHB model by using the following thermodynamic
cycle:
If E , S , and V are the changes in the energy, en-
tropy, and volume due to the formation of an i– j hydrogen
bond, then the fraction of such i– j bonds formed is given by
n
in
i
j
a
0
ij
␯ijϭ ␯ Ϫ
␯ik ␯ Ϫ
␯kj exp͑ϪG /RT͒, ͑10͒
ͫ
d
͚
ͬͫ
͚
ͬ
k
k
12
0
0
0
ij
where,
ij ij
i
d
j
e 0
Nij
rN
0
N
N
a
i
d
j
a
␯ij
ϭ
;
␯ ϭ
;
␯ ϭ
;
rN
rN
0
ij
0
ij
0
ij
G ϭE ϩPV ϪTS .
͑11͒
ij
The free energy of mixing in the LFHB model can be
calculated from the mixing rules given by Eqs. ͑2͒–͑4͒ and
the hydrogen bonding fractions given by Eq. ͑10͒. The de-
tails of the free energy of mixing terms are described in Lele
et al. From the free energy of mixing and the elastic free
energy of Eq. ͑1͒, the chemical potential of water in the gel
phase can be calculated as
1
0
G
␮1
r1
2
ϭln͑␾ /␻ ͒ϩ 1Ϫ
␾ ϩr1␳␾ X
˜
2
2 12
1
1
ͩ ͪ
RT
r
2
˜
˜
␳
P
˜
␯
1
1
From the above cycle, the heats associated with steps 1 and 3
are responsible for the peak in the DSC endotherm. There-
fore,
ϩr1
Ϫ
ϩ
ϩ͑
˜
␯Ϫ1͒ln͑1Ϫ
˜
␳͒ϩ ln
˜
␳
ͭ
ͮ
˜
˜
1
T
T
r1
1
m
n
i
d
j
␯
␯
a
1
1
j
ϩr1
␯ijϪ
d ln
Ϫ͚
a ln
͚ ͚
͚
i
⌬H ϭϪ⌬H ϩ⌬H ,
͑16͒
4
1
3
i
j
␯io
␯0j
1
/3
where ⌬H1 and ⌬H3 are the enthalpies of mixing and can be
calculated from the LFHB theory on a per gram dry polymer
basis by the following equation:
␯e
V0
1
V0
ϩ r
␯*
˜
␯
Ϫ
,
͑12͒
ͭ
1
ͩ ͪ
1
ͫ
ͩ ͪ ͩ ͪ
ͬ
ͮ
V
V
2
V
0
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
1
32.174.255.116 On: Mon, 01 Dec 2014 12:06:40

1178
J. Chem. Phys., Vol. 109, No. 3, 15 July 1998
Badiger et al.
͑
0.02 M͒ in 2 ml of dichloromethane was added dropwise
r
⌬
Hmix
ϭ
Ϫ
˜
␳␧*ϩ͚
˜
␳ ␾ ␧*
over a period of 1–2 h. After addition of half of the acid
chloride, pH of the reaction mixture dropped to 7.5 and it
was maintained between 7.5 and 7.8 by the addition of so-
dium hydroxide solution. Unreacted acryloyl chloride and
dichloromethane were removed by extraction with ethyl ac-
etate. The clear aqueous layer was acidified to pH 5.0–5.5
with dilute hydrochloric acid and extracted with ethyl ac-
etate. The product in the ethyl acetate layer was dried with
anhydrous sodium sulfate and concentrated in rotavapor. On
cooling the concentrate, solid monomer was precipitated. It
was further purified by redissolving in ethyl acetate and re-
precipitating in petroleum ether ͑white powder, yield—49%,
melting point ͑m.p.͒ 90.1 °C͒.
ͭ
k
k
k
x2MW2
m
n
0
k
ϩ͚i ͚j ij
E
␯ijϪ
͚
␯
ͮ
,
͑17͒
␳ is the
ͫ
ij
ͬ
k
where MW is the molecular weight of the polymer,
reduced density of pure component k, and ␯ is the number
of hydrogen bonds between an i– j pair in pure component
k.
In the following section, we show that the extended
LFHB model can quantitatively fit the experimental swelling
measurements of the copolymer gels, and also qualitatively
predict the experimentally observed trends in the heat of de-
mixing of the copolymer gels.
˜
2
k
k
ij
IR ͑Nujol͒. 3278 ͑-NH and -OH stretching͒, 2950 and
2
1
ing͒.
957 ͑-CH stretching͒, 1701 ͑-CO stretching of -COOH͒,
Ϫ1
643 ͑-CO stretching of -CONH ), 1540 cm ͑-NH bend-
EXPERIMENT
2
Materials
1
H NMR (D O, ppm͒. ␦2.3 (-C Hគ COO), ␦1.8 (-C Hគ -),
2
2
2
Acrylic acid, 4-aminobutyric acid, 6-aminocaproic acid,
1, ␻-aminoundecanoic acid, ethylene glycol dimethacrylate
␦3.3 (-C Hគ 2-N), ␦4.8 ͑HOD͒, ␦5.8 (CH2ϭCH), ␦6.2
(C Hគ 2ϭCH).
1
͑EGDMA͒, and methylene bis-acryl- amide͑Bis-A͒ were all
obtained from Aldrich Chemical Company Inc.
N-isopropylacrylamide ͑NIPAm͒ was purchased from Poly-
Sciences Inc. ͑Warrington, PA͒ and was used without further
purification. AR grade 1,4-dioxane was obtained from J. T.
Baker Chemical Co., Phillipsburg, NJ. Thionyl chloride was
procured from S. D. Fine Chemicals, Mumbai, India and
purified by distillation. The initiator azo bisisobutyronitrile
Acryloyl-6-aminocaproic acid †X5‡
X5 was synthesized by the reaction between acryloyl
chloride and 6-aminocaproic acid using the same procedure
used for the preparation of X3 mentioned above. Stoichio-
metric amounts of reactants were used for the reaction ͑white
powder, yield—55%, mp 77.4 °C͒.
IR ͑Nujol͒. 3284 ͑-NH and -OH stretching͒, 2978 and
͑AIBN͒ was purchased from SAS Chemical Co., Mumbai,
2
1
1
852 ͑-CH stretching͒, 1697 ͑-CO stretching of -COOH͒,
India, and was purified by recrystallization in ethanol. The
deionized distilled water was prepared in the laboratory us-
ing standard procedures.
650 ͑-CO stretching of -CONH ), 1622 (-CϭC stretching͒,
2
Ϫ1
546 cm ͑-NH bending͒.
1
H NMR (CDCl3, ppm͒. ␦2.3 (C Hគ COO), ␦1.7
2
(
(
-C Hគ -) , ␦1.4 (-͑CH ͒ -C Hគ ), ␦3.3 (-N-C Hគ ), ␦5.6
2 2 2 2 2 2
Synthesis of monomers
Acryloyl chloride
-CH ϭCH), ␦6.2 (C Hគ ϭCH).
2
2
Acryloyl-11, ␻-aminoundecanoic acid †X10‡
Acryloyl chloride was synthesized by the reaction be-
tween acrylic acid and thionyl chloride. Thus, acrylic acid
X10 monomer was synthesized by the reaction between
acryloyl chloride and 11, ␻-aminoundecanoic acid, using the
same procedure with the stoichiometric amounts of reactants.
The dissolution of 11, ␻-aminoundecanoic acid in aqueous
alkali, however, was carried out at room temperature ͑28 °C͒
instead of 5–10 °C because of its precipitation at low tem-
perature ͑white powder, yield—53%, mp 237 °C͒.
͑20.5 ml, 0.3 M͒ along with dimethyl formamide ͑2 ml͒ and
hydroquinone ͑3 g͒ was taken in a round bottomed flask.
Freshly prepared thionyl chloride ͑23 ml, 0.3 M͒ was added
dropwise over a period of 1–2 h. After complete addition of
thionyl chloride, the reaction mixture was heated at 60 °C for
6
h with continuous stirring. The flask was then kept over-
night at room temperature. Another 3 g of hydroquinone was
added and pure acryloyl chloride was distilled out from the
flask at 70–71 °C under normal conditions. The hydrophobic
monomers were synthesized by the reaction between acry-
loyl chloride and corresponding aminoacids as per the reac-
tion scheme shown in Fig. 1. The hydrophobic comonomers
are coded on the basis of length of the alkyl chain (X).
IR ͑Nujol͒. 3304 ͑-NH, -OH stretching͒, 2924 and 2854
͑-CH stretching͒, 1693 (-CϭO stretching of -COOH͒, 1652
(-CϭO stretching of -CONH2), 1623 and (-CϭC stretch-
Ϫ1
ing͒, 1540 cm ͑-NH bending͒.
1
H NMR (CDCl3, ppm͒. ␦2.3 (C Hគ 2-COO), ␦1.6
(͑C Hគ 2͒2-CH2-COO), ␦1.3 (͑-C Hគ 2-͒6), ␦3.3 (NH-CH2), ␦5.7
(-C Hគ ϭCH2), ␦6.25 (C Hគ 2ϭCH-).
Synthesis of copolymer gels
Acryloyl-4-aminobutyric acid †X3‡
In a 100 ml beaker equipped with a pH electrode, 2.25 g
We synthesized the gels by copolymerizing X3, X5, and
X10 monomers with N-isopropylacrylamide as per the feed
composition given in Table I. The copolymerization was car-
ried out in 1,4-dioxane using EGDMA as a cross linker and
AIBN as a free radical initiator. The reaction scheme is
of 4-aminobutyric acid ͑0.02 M͒, 15 ml distilled water, and
0
.8 g sodium hydroxide were placed to give a clear solution.
The solution was stirred with a magnetic needle at 5–10 °C
ice-water bath͒. To this solution, 1.7 ml acryloyl chloride
͑
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.255.116 On: Mon, 01 Dec 2014 12:06:40

J. Chem. Phys., Vol. 109, No. 3, 15 July 1998
Badiger et al.
1179
FIG. 1. Reaction scheme for the synthesis of hydrophobic monomers and the copolymer gels.
TABLE I. Feed compositions of copolymer gels.
Mole ratioϭ1/0.1
Cross linker
Initiator
Solvent
Sample code
͑moles͒
͑moles͒
͑moles͒
͑ml͒
X3-1
NIPAm/X3
EGDMA
AIBN
1,4-Dioxane
͑20 ml͒
1,4-Dioxane
͑20 ml͒
1,4-Dioxane
͑20 ml͒
1,4-Dioxane
͑20 ml͒
(
0.0177/0.001 77)
NIPAm/X5
0.0177/0.001 77)
NIPAm/X10
0.0177/0.001 77)
NIPAm/X5
0.0177/0.005 31)
NIPAm/X5
0.0177/0.0885)
(1.5ϫ10Ϫ3)
EGDMA
(4.87ϫ10Ϫ4)
AIBN
X5-1
X10-1
X5-3
X5-5
(
(1.5ϫ10Ϫ3)
EGDMA
(4.87ϫ10Ϫ4)
AIBN
(
(1.5ϫ10Ϫ3)
EGDMA
(4.87ϫ10Ϫ4)
AIBN
(
(1.5ϫ10Ϫ3)
EGDMA
(4.87ϫ10Ϫ4)
AIBN
1,4-Dioxane
͑20 ml͒
(
(1.5ϫ10Ϫ3)
(4.87ϫ10Ϫ4)
NIPAm
N-isopropylacrylamide
EGDMA
Ethylene glycol dimethacrylate
AIBN
X3
X5
Azo-bis-isobutyronitrile
Acryloyl-4-aminobutyric acid
Acryloyl-6-aminocaproic acid
Acryloyl-11-␻-aminoundecanoic acid
X10
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
32.174.255.116 On: Mon, 01 Dec 2014 12:06:40
1

1180
J. Chem. Phys., Vol. 109, No. 3, 15 July 1998
Badiger et al.
shown in Fig. 1. The required amounts of monomers ͑with
defined mole ratios͒, crosslinking agent, and initiator were
taken in a beaker and dissolved in 1,4-dioxane with constant
stirring and N bubbling. The reaction mixture was poured
2
into test tubes and sealed. The polymerization was carried
out at 70 °C for 24 h. The gels obtained in a cylindrical form
were washed with water and ethanol for one day and dried in
an oven at 40 °C till the constant weight was obtained. Co-
polymer gels of NIPAm and monomer X5 were prepared
with three different compositions corresponding to a
NIPAm/X5 mol ratio of 1/0.1 (X5-1), 1/0.3 (X5Ϫ3), and
1/0.5 (X5Ϫ5).
Characterization of gels
The gels were characterized for their molecular structure
by 13C MASS spectroscopy. All C-NMR spectra were ob-
tained on a Bruker MSL-300 FT NMR spectrometer operat-
ing at a carbon frequency of 75.47 MHz. For the CP-MASS
experiments, cross polarization was established using
matched Hertmann–Han conditions at 50 kHz using an ada-
mantane as a standard. 13C CP-MASS spectra of each sample
were taken at two different speeds to identify the center band
and side bands. The spectra were recorded at a probe tem-
perature of 25 °C.
13
The thermal characterization of the gels was done using
a DSC ͑Model 2920, TA Instruments͒ at a heating scan rate
of 3 °C/min. Thoroughly sealed samples were equilibrated at
FIG. 2. Comparison of 13C NMR spectra of homopolymer and copolymer
gels.
5
°C and then heating scans were recorded from 5 to 50 °C.
Swelling measurements were carried out by immersing
ϫ5 mm cylindrical dry gel samples in excess water, which
4
was maintained at a controlled temperature within Ϯ0.5 °C.
The gels were allowed to equilibrate for 72 h during which
constant weight was reached, after which they were quickly
removed and weighed. The surface water was carefully
wiped off before weighing. Swelling capacity was recorded
gravimetrically as
distinguish between the -CO signals arising due to -CONH2
and -COOH of copolymer gels because of their very small
chemical shift differences. The quantitative estimation of the
copolymer composition is extremely difficult due to uncer-
tainties in the integration of the 13C signals of copolymer
gels. Moreover, the mole ratio of hydrophobically modified
monomer to NIPAm monomer used for copolymerization is
very low to get well resolved 13C signals from comonomer
structure.
wt. equilibrated gel
qϭ
.
wt. of dry gel
We assume that since the concentration of comonomer
in the gel is very small and the monomers are completely
soluble in the solvent used for synthesis ͑1,4-dioxane͒, the
copolymers thus formed are random copolymers.
RESULTS AND DISCUSSIONS
Copolymer structure
We show in Fig. 2 the 13C spectra of X3Ϫ1, X5Ϫ1, and
X10Ϫ1 copolymer gels and compare them with the spectra
of two homopolymer gels PNIPAm and X10. In the PNIPAm
Swelling behavior and LCST
homopolymer spectrum, the -CH peaks appear at 24.8 ppm,
carbonyl of -CONH appear at 174.7 ppm. The backbone -CH
The effect of the presence of hydrophobic comonomers
on the swelling of PNIPAm based gels is shown in Fig. 3.
PNIPAm homopolymer gel was found to undergo a nondis-
continuous but a steep LCST-type deswelling as the tem-
perature was raised between 27 and 30 °C. The nature of the
transition and the LCST of the gel was also confirmed from
the DSC results shown in Fig. 4. As seen from the DSC
heating scans, the volume transition is accompanied by a
heat of demixing which appears as an endothermic peak. The
broadness of the endotherm indicates that the volume transi-
tion is not completely discontinuous, which is in agreement
with the swelling measurements. Also, the peak temperature
of 28.5 °C corresponds closely to the midway temperature of
3
and -CH peaks resonate at the broad range of 40–50 ppm.
2
The incorporation of hydrophobic monomer in the copoly-
mer is confirmed by careful analysis of all the spectra. The
1
3
C
spectrum of homopolymer of acryloyl-11,␻-
aminoundecanoic acid shows a characteristic peak of hydro-
phobic side chain -CH group at 29.6 ppm which is totally
2
1
3
absent in the C spectrum of PNIPAm homopolymer. How-
ever, this peak is progressively distinct in copolymer gels
X3Ϫ1, X5Ϫ1, and X10Ϫ1 as the chain length of -CH2
increases. This clearly indicates the incorporation of hydro-
phobic comonomer in the copolymer gels. It is difficult to
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.255.116 On: Mon, 01 Dec 2014 12:06:40

J. Chem. Phys., Vol. 109, No. 3, 15 July 1998
Badiger et al.
1181
FIG. 3. Effect of alkyl chain length on swelling behavior of copolymer gels.
Experimental data: ͑᭺͒ PNIPAm; ͑᭡͒ X3-1; ͑᭿͒ X5-1; ͑᭹͒ X10-1. Lines
are theoretical fits.
FIG. 5. Effect of concentration of hydrophobic comonomer on swelling
behavior of copolymer gels. Experimental data: ͑᭿͒ X5-1: ͑᭡͒ X5-3: ͑᭹͒
X5-5. Lines are theoretical fits.
the range in which deswelling occurs. Therefore, we will
report here the peak temperatures as the LCSTs for all gels.
X3Ϫ1 gel was found to have a lower equilibrium swell-
ing capacity than the PNIPAm gel. The LCST of the gel as
measured from swelling measurements and DSC was
herm in DSC. Hence we report the LCST for this gel from
the swelling measurements alone.
The effect of increase in the concentration of the hydro-
phobic comonomer X5 on the swelling ratio is shown in Fig.
5
. Copolymer gels X5Ϫ1, X5Ϫ3, and X5Ϫ5, which contain
2
4.9 °C which is lower than that of PNIPAm gel. X5Ϫ1 has
a higher swelling capacity than PNIPAm, as seen from Fig.
. However, the LCST of 21.3 °C for X5Ϫ1 is lower than
increasing concentrations of X5, show a decrease in LCSTs
and reduced equilibrium swelling capacities. The corre-
sponding DSC endotherms for these gels are shown in Fig. 4.
X5Ϫ3 shows an LCST of 13.5 compared to 21.3 °C for
X5Ϫ1. The scan for X5Ϫ5 does not show any endotherm
for the same reasons as for X10Ϫ1 described above.
It can be seen from the above results that the main effect
of copolymerizing NIPAm with hydrophobic comonomers is
to lower the LCST of the gel. The higher the concentration
of the hydrophobic comonomer and the longer the hydropho-
bic alkyl chain length, the lower is the LCST of the gel.
Thus, increasing the concentration of the X5 comonomer
reduces the LCST significantly, and increasing the chain
length of the hydrophobe from X3 to X5 to X10 also reduces
the LCST of the gel. These effects can be understood more
quantitatively by fitting the experimental data to the ex-
tended LFHB theory as described below.
3
PNIPAm and X3Ϫ1 gels. The higher swelling capacity of
X5Ϫ1 could be due to a lower crosslink density of the gel.
X10Ϫ1 has the smallest equilibrium swelling capacity and
the lowest capacity LCST ͑10 °C͒ of all the copolymer gels.
It was not possible to observe the endotherm for the gel in
the DSC because of the low temperatures at which the LCST
transition occurs for this gel. The equilibration to baseline
from the initiation of the scan occurred from 5 to about
10 °C. Since the volume transition for the X10Ϫ1 gel oc-
curred in this range, it was not possible to record its endot-
The lines through the experimental data points in Figs. 3
and 5 are the calculations of the extended LFHB theory. The
various model parameters used for obtaining the theoretical
fits are listed in Tables II to IV and are discussed below.
Following our earlier successful work,1
0–12
we have assumed
that the pure component parameters of PNIPAm are the same
as those of PMMA, given by Sanchez and Lacombe.2
0,21
We
have shown that these parameters can quantitatively fit the
TABLE II. Molecular parameters.
Component
P* ͑MPa͒
T* ͑K͒
␳* (kg/cm3)
MW ͑g/mol͒
PNIPAM
503
431
699
627
1269
1125
10 000
10 000
Homopolymers
of hydrophobic
monomers
Water
475
578
853
18
FIG. 4. DSC heating scans of copolymer gels.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
32.174.255.116 On: Mon, 01 Dec 2014 12:06:40
1

1182
J. Chem. Phys., Vol. 109, No. 3, 15 July 1998
Badiger et al.
TABLE III. Hydrogen bonding parameters.
Donors→
H bonding
parameters
acceptors
uNH
uOH
uCOOH
2
1
2
a
↓
d ϭ80
d ϭ2
d ϭ8,24,40
1
2
3
0
Ϫ3.24ϫ103
Ϫ9.9
Ϫ16.0ϫ103
Ϫ25.8
Ϫ3.24ϫ103
Ϫ10.0
E
S
V
E
S
V
E
S
͑J/gmol͒
ij
0
͑J/gmolK͒
—CvO
ij
0
3
2
(cm /gmol)
a ϭ80
Ϫ0.85
Ϫ0.85
Ϫ0.85
ij
0
1
Ϫ12.5ϫ103
Ϫ17.8
Ϫ16.595ϫ103
Ϫ26.6
Ϫ12.5ϫ103
Ϫ17.8
͑J/gmol͒
ij
0
͑J/gmolK͒
uOH
ij
0
3
1
(cm /gmol)
a ϭ2
Ϫ0.85
Ϫ4.2
Ϫ0.85
ij
0
2
Ϫ12.5ϫ103
Ϫ7.8
Ϫ16.0ϫ103
Ϫ16.6
Ϫ12.5ϫ103
Ϫ7.8
͑J/gmol͒
ij
0
͑J/gmolK͒
uCOOH
a
ij
0
3
2
3
V
(cm /gmol)
a ϭ8, 24, 40
Ϫ0.85
Ϫ0.85
Ϫ0.85
ij
a
2
2
a ϭd ϭ8 for X3-1, X5-1 and X10-1.
3
3
2
2
a ϭd ϭ24 for X5-3.
3
3
2
2
a ϭd ϭ40 for X5-5.
3
3
swelling behavior of PNIPAm gel.10 For the copolymers of
NIPAm and the hydrophobic comonomers, the extended
LFHB model requires pure component properties of the ho-
mopolymer of the hydrophobic monomer. In the absence of
any data for the pure component properties of these ho-
mopolymers, we have assumed, in a manner similar to the
case of PNIPAm, the values for pure component properties
of poly(n-butyl methacrylate͒ as being applicable for these
homopolymers. These values have been reported by Sanchez
and Lacombe. The rationale for the choice of poly(n-butyl
methacrylate͒ is that it is more hydrophobic than poly͑m-
ethyl methacrylate͒ and that it has a similar hydrophobic side
chain as that for the comonomers in our case. Table II gives
the pure component properties assumed for PNIPAm and the
homopolymer of the hydrophobic comonomers.
mixture are assumed to be such that the hydrogen bonds
formed are slightly stronger than those for the amide groups.
The effective hydrophobicity of the copolymer gel is
represented by the binary interaction parameter ␨12 given by
Eqs. ͑6͒ and ͑7͒. The binary interaction parameters for ho-
mopolymer PNIPAm (␨1A) and homopolymer of the
comonomer (␨1B) which are required for fitting the experi-
mental swelling data are listed in Table IV. A lower value of
␨ indicates greater hydrophobicity. It is seen from the values
of ␨ that the effective hydrophobicity of all gels is tempera-
ture dependent and that the value of ␨ at any temperature
decreases for increasing chain length of the hydrophobic
group. Figure 6 shows a correlation between the experimen-
tal LCST ␨, and X for the copolymer gels. The almost linear
correlation between the LCST and X suggests that the LCST
2
2
The calculations of the hydrogen bonding terms in the
chemical potential require solving nine simultaneous equa-
tions represented by Eq. ͑10͒ for three pairs of donors and
acceptors, which are listed in Table III. The number of do-
nors and acceptors, and their hydrogen bonding energies,
entropies, and volume changes are also listed in Table III.
The hydrogen bonding parameters for the PNIPAm–water
system are the same as those reported in our earlier work.10
The parameters corresponding to the hydrogen bonding be-
tween the terminal carboxyl ͑-COOH͒ group on the hydro-
phobic comonomer and the other donors and acceptors in the
is reduced by about 2 °C for every additional ϪCH group
2
on the terminal side chain of the hydrophobe. The correlation
between X and ␨ at any given temperature is also almost
linear. These correlations can be useful not only in quick
‘‘designing’’ of gels, but could also provide insights into the
mechanism of volume transition of gels. It is also important
to note that in fitting the swelling behavior of X5Ϫ1, X5
Ϫ3, and X5Ϫ5 gels, we have not changed the interaction
TABLE IV. Binary interaction parameters and crosslink density.
3
Gel
Components
␨
(␯ /␯ ) (gmol/cm )
e
0
Ϫ3
Ϫ3
Ϫ3
Ϫ3
Ϫ3
Ϫ3
PNIPAM
X3-1
NIPAm
NIPAm
X3
NIPAm
X5
NIPAm
X10
NIPAm
X5
1.5–1.56ϫ10
1.5–1.56ϫ10
1.5–1.78ϫ10
1.5–1.56ϫ10
1.5–1.82ϫ10
1.5–1.56ϫ10
T
T
T
T
T
T
90
65
X5-1
X10-1
X5-3
X5-5
20
250
55
Ϫ3
1.5–2.1ϫ10
1.5–1.56ϫ10
1.5–1.82ϫ10
1.5–1.56ϫ10
1.5–1.82ϫ10
T
Ϫ3
Ϫ3
Ϫ3
Ϫ3
T
T
T
T
NIPAm
X5
135
FIG. 6. Correlation between LCSTs, ␨, and length of alkyl group (X) of the
comonomer.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
32.174.255.116 On: Mon, 01 Dec 2014 12:06:40
1

J. Chem. Phys., Vol. 109, No. 3, 15 July 1998
Badiger et al.
1183
TABLE V. Comparison between experimentally measured heat of demixing
been discussed in the Introduction section. However, our hy-
drophobically modified copolymer gels showed a lower heat
of transition when compared to PNIPAm homopolymer gels.
Table V shows a comparison between the heat of mixing
calculated from Eq. ͑17͒ for the transitions between 5 to
͑DSC͒ and theoretical predictions from Eq. ͑17͒.
Length of hydrophobe
Experimental
͑J/g dry polymer͒
͑area under DSC endotherm͒
Theoretical
͑J/g dry polymer͒
͑Eq. 17͒
͑No. of uCH
2
groups, X)
4
0 °C for all gels and the endotherms of the DSC. All the
0
3
5
59.2
33.3
30.7
¯
53.90
50.04
48.30
29.94
model parameters given in Tables II to IV have been used for
calculating the heat of mixing. The calculations thus become
completely predictive. The comparison is only qualitative,
probably because of errors in the baseline fitting of heating
scans and perhaps also because of the transient nature of the
DSC experiment. However, it is important to note that the
trends are in agreement: namely, the heat of demixing de-
creases with an increase in the hydrophobicity of the copoly-
mer gels.
The above observations suggest that the water molecules
are unable to structure around the hydrophobic groups in our
copolymer gels. This can probably be attributed to the pres-
ence of the terminal hydrophilic carboxyl group, which hin-
ders the formation of hydrophobic hydration around the
10
parameter ␨1B of the hydrophobic homopolymer. Since the
value of ␨ determines the transition temperature of the gels,
the predicted lowering of the LCSTs with increasing compo-
sition of X5 validates our assumption of a ‘‘random’’ co-
polymer in the model and proves the strength of the model.
The values of crosslink density of the gels required for
quantitatively fitting the experimental swelling data are re-
ported in Table IV. It is seen that X3Ϫ1 and X5Ϫ1 required
a lower value of crosslink density than PNIPAm gel: how-
ever, the X10Ϫ1 gel required a very high value of crosslink
density. Similarly, the X5Ϫ3 and X5Ϫ5 gels, which contain
increasing concentrations of X5 comonomer, required in-
creasing values of the crosslink densities. Our synthesis pro-
cedure for copolymer gels does not provide any control over
the crosslink density, hence it is possible that the copolymer
gels X3Ϫ1 and X5Ϫ1 have lower crosslink densities than
the PNIPAm gel. However, it is also possible that the effec-
tive crosslink density of the copolymer gels could increase as
a result of additional ‘‘physical’’ crosslinks such as hydro-
phobic interactions or chain entanglements of the long hy-
drophobic side groups of the comonomers. This effect will
increase with an increase in the chain length of the side
group and with increasing contents of the hydrophobe in the
copolymer gels. This could be the reason for the higher
crosslink density values for the X10Ϫ1, X5Ϫ3, and X5
Ϫ5 gels.
The DSC heating scans in Fig. 4 show the endothermic
heat associated with the volume transitions of the gel. It is
believed that this externally supplied heat is used to break
the hydration layers around the hydrophobic groups. The ex-
posed hydrophobes then associate, causing the gel to col-
lapse. The temperature at which the hydrophobic hydration
breaks down would decrease with an increase in the hydro-
phobicity of the gel. Indeed, in our copolymer gels, we find a
decrease in the transition temperature with increasing hydro-
phobic nature ͑increasing X) and content ͑composition of the
comonomer͒. The LFHB model also shows that a decrease in
the transition temperature is caused by an increase in hydro-
phobicity ͑lower ␨͒ of the gel.
ϪCH groups of the alkyl chain. Thus, our comonomers in-
2
crease the overall hydrophobicity of the copolymer gels
compared to PNIPAm gel and thereby reduce the LCSTs.
However, the comonomers do not allow structuring of water
molecules around them and therefore do not increase the heat
of demixing. The LFHB model also supports this rationale.
Being an essentially mean-field theory, the model does not
account specifically for structuring of water molecules
around hydrophobic groups. Therefore it can not predict an
increase in the heat of demixing caused by breaking down of
hydration layers around the hydrophobes. On the contrary, it
predicts a decrease in the ⌬H due to decreased polymer–
water interactions caused by the increased hydrophobicity of
the copolymer ͑as indicated by reduced ␨͒. The experimental
results match with the predictions, thus indicating that water
does not hydrate the hydrophobic groups in our copolymers.
This shows that the subtleties of chemical structure of
hydrophobes play an important role in determining the extent
of hydration. We are currently investigating these effects by
developing designed hydrophobic comonomers which can
allow structuring of water molecules.
CONCLUSIONS
We have synthesized hydrophobic vinyl monomers con-
taining varying lengths of alkyl groups X and copolymerized
these monomers with N-isopropyl acrylamide monomer to
give hydrophobically modified copolymer gels. These gels
show lower LCST-type transition temperatures than that of
pure PNIPAm gel. The reduction in the transition tempera-
ture was found to increase with the length of the alkyl group
in the hydrophobic comonomers and also with an increase in
the content of the comonomers. A linear correlation between
the transition temperature and the number of alkyl groups
was found to exist in these gels. Such a correlation could be
a useful tool in designing gels having a specific transition
temperature.
Besides the transition temperature, the heat associated
with the volume transition will also depend on the hydropho-
bicity of the gel. If the heat of transition is used for breaking
the hydration layers around the hydrophobic groups, then
greater is the extent of hydration, and the larger will be the
heat associated with the transition. Therefore it might appear
that increasing the hydrophobic alkyl chain length would fa-
cilitate the formation of a larger hydration layer, and there-
fore would result in an increased heat of transition. This has
We have also shown that the extended LFHB theory can
quantitatively fit the swelling data of the copolymer gels.
The temperature dependent binary interaction parameter ␨,
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.255.116 On: Mon, 01 Dec 2014 12:06:40

1184
J. Chem. Phys., Vol. 109, No. 3, 15 July 1998
Badiger et al.
3
4
5
which simulates the effective hydrophobicity of the gels, was
found to decrease with an increase in the alkyl chain length.
The predictions of the transition temperatures of NIPAm-X5
gels containing increasing concentrations of the comonomer
validate the assumption of a ‘‘random’’ copolymer in our
model.
M. Marchetti, S. Prager, and E. L. Cussler, Macromolecules 23, 3445
͑
1990͒.
Y. H. Bae, T. Okano, E. Hsu, and S. W. Kim, Makromol. Chem., Rapid
Commun. 8, 481 ͑1987͒.
M. G. Kulkarni, S. S. Patil, V. Premnath, and R. A. Mashelkar, Proc. R.
Soc. London, Ser. A 439, 397 ͑1992͒.
E. L. Cussler, M. R. Stokar, and J. E. Varberg, AIChE. J. 30, 578 ͑1984͒.
M. J. Snowden, M. J. Murray, and B. Z. Chowdary, Chem. Industry 15,
6
7
In apparent contradiction to previous work,16,17 our co-
polymer gels show that the heat associated with the volume
transition decreases with an increase in the hydrophobic
alkyl chain length. Our theoretical predictions of the heat of
demixing are also in qualitative agreement with experimental
observations. These results suggest that the hydrophobic
groups used in our copolymer gels do not allow structuring
of water around them, which is probably due to the presence
of the strongly hydrophilic carboxyl terminal group. The lack
of structured water reduces the heat of demixing: however,
the increase in the overall hydrophobicity of the comonomer
causes a reduction in LCSTs.
531 ͑1996͒.
8
B. Vernon, A. Gutowska, S. W. Kim, and Y. H. Bae, Macromol. Symp.
09, 155 ͑1996͒.
1
9
R. Dagani, Chem. Eng. News 75, 26 ͑1997͒.
A. K. Lele, M. V. Badiger, M. M. Hirve, and R. A. Mashelkar, Chem.
Eng. Sci. 50, 3535 ͑1995͒.
A. K. Lele, S. K. Karode, M. V. Badiger, and R. A. Mashelkar, J. Chem.
Phys. 107, 2142 ͑1997͒.
A. K. Lele, I. Devotta, and R. A. Mashelkar, J. Chem. Phys. 106, 4768
10
1
1
1
2
͑
1997͒.
13
R. Katakai, M. Yoshida, S. Hasegawa, and Y. Lijima, Macromolecules 29,
1068 ͑1996͒.
P. S. Mumick and C. L. McCormick, Polym. Eng. Sci. 34, 1419 ͑1994͒.
M. Shibayama, S. Mizutani, and S. Nomura, Macromolecules 29, 2019
1
1
4
5
͑
1996͒.
ACKNOWLEDGMENTS
1
1
6
7
H. Feil, Y. H. Bae, J. Feijen, and S. Wan Kim, Macromolecules 26, 2496
1993͒.
͑
The authors wish to thank the Department of Science
and Technology ͑DST͒, New Delhi, Grant No. 111.5 ͑131͒/
S. H. Cho, M. S. Jhon, S. H. Yuk, and H. B. Lee, J. Polym. Sci., Part B:
Polym. Phys. 35, 595 ͑1997͒.
95-ET for financial support. The authors also wish to thank
18
19
H. Inomata, S. Goto, and S. Saito, Macromolecules 23, 4887 ͑1990͒.
C. Panayiotou and I. C. Sanchez, J. Chem. Phys. 95, 10,090 ͑1991͒.
I. C. Sanchez and R. H. Lacombe, J. Phys. Chem. 80, 2352 ͑1976͒.
I. C. Sanchez and R. H. Lacombe, Macromolecules 11, 1145 ͑1978͒.
I. C. Sanchez and R. H. Lacombe, J. Polym. Sci., Polym. Lett. Ed. 15, 71
͑1977͒.
M/S Gharda Chemicals Ltd for DSC experiments.
2
0
1
21
22
H. G. Schild, Prog. Poly. Sci. 17, 163 ͑1992͒.
X. Huang, H. Unno, T. Akehta, and O. Hirasa, J. Chem. Eng. Jpn. 21, 655
2
͑1988͒.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
32.174.255.116 On: Mon, 01 Dec 2014 12:06:40
1