Probing the influence of polymer architecture on liquid-liquid phase transitions of aqueous poly(N,N-dimethylacrylamide) copolymer solutions

Article abstract of DOI:10.1021/ma048853p

Thermosensitive poly(N,N-dimethylacrylamide-co-N-phenylacrylamide) (DMA-co-PhAm) copolymers were prepared by atom transfer radical polymerization (ATRP) in methanol/water mixtures at room temperature with methyl 2-chloropropionate as the initiator and CuCl/Me₆TREN as the catalyst. The resultant DMA-co-PhAm copolymers had tailored compositions and controlled molecular weights, and their aqueous solutions underwent liquid-liquid phase separation upon heating. These phase transition temperatures, measured by the cloud point method, were dependent on polymer concentrations, compositions, and molecular weights. The efficiency of the thermally induced liquid-liquid phase transition, i.e., the yield of phase-separated polymer, increased with increasing solution temperature, suggesting this thermally induced liquid-liquid phase transition to be a continuous equilibrium process. The efficiency of phase separation could be enhanced by adding NaCl to the solution.

Full text of DOI:10.1021/ma048853p

Macromolecules 2005, 38, 2109-2115
2109
Probing the Influence of Polymer Architecture on Liquid-Liquid Phase
Transitions of Aqueous Poly(N,N-dimethylacrylamide) Copolymer
Solutions
Xiangchun Yin and Harald D. H. Sto1ver*
Department of Chemistry, McMaster University, Hamilton, Ontario, Canada L8S 4M1
Received June 10, 2004; Revised Manuscript Received January 10, 2005
ABSTRACT: Thermosensitive poly(N,N-dimethylacrylamide-co-N-phenylacrylamide) (DMA-co-PhAm)
copolymers were prepared by atom transfer radical polymerization (ATRP) in methanol/water mixtures
at room temperature with methyl 2-chloropropionate as the initiator and CuCl/Me6TREN as the catalyst.
The resultant DMA-co-PhAm copolymers had tailored compositions and controlled molecular weights,
and their aqueous solutions underwent liquid-liquid phase separation upon heating. These phase
transition temperatures, measured by the cloud point method, were dependent on polymer concentrations,
compositions, and molecular weights. The efficiency of the thermally induced liquid-liquid phase
transition, i.e., the yield of phase-separated polymer, increased with increasing solution temperature,
suggesting this thermally induced liquid-liquid phase transition to be a continuous equilibrium process.
The efficiency of phase separation could be enhanced by adding NaCl to the solution.
Introduction
extensive studies on thermally responsive polymers,
there appears to be little information available on the
efficiency of such liquid-liquid phase separations. This
efficiency should depend on the average composition and
the compositional distribution as well as on the molec-
ular weight and the molecular weight distribution of the
copolymers. Most thermally responsive polymers stud-
ied to date were prepared by conventional radical
polymerization, with correspondingly little control over
molecular weight and molecular weight distribution.
Controlled/living radical polymerization, in particular
atom transfer radical polymerization (ATRP), has been
extensively used to prepare well-defined polymers with
comb, star, and dendritic morphologies as well as
random, gradient, and blocky comonomer sequences.18
In the present work, we report the synthesis of well-
defined random copolymers of N,N-dimethylacrylamide
(DMA) with N-phenylacrylamide (PhAm) and their
aqueous solution properties. The PhAm is added to
increase the hydrophobicity of the copolymer and permit
thermally induced phase separations. The primary
objective of this research is to probe the mechanism of
the thermally induced liquid-liquid phase transition
using copolymers with well-defined structures.
The liquid-liquid phase transitions of aqueous poly-
mer solutions have been widely used in biochemistry
and biotechnology for purifications of amino acids,
proteins, nucleic acids, and cells.1-3 Because of the fairly
high water content even in the polymer-rich phase, such
liquid-liquid phase transitions permit the separation
of biomaterials without denaturation. Classical ex-
amples include the dextran/poly(ethylene glycol) and
poly(ethylene glycol)/salts systems.4-6 Recently, the
thermally induced liquid-liquid phase transition of
aqueous polymer solutions has gained increasing at-
tention for similar reasons.7,8 These aqueous polymer
solutions are characterized by a lower critical solution
temperature (LCST), above which the polymer becomes
insoluble.
Thermally induced liquid-liquid phase transitions
are commonly seen in biomacromolecules such as elas-
tin-like polypeptides and have been used to purify
recombinant proteins from the cell lysate.9-11 The most
studied thermoresponsive polymer is poly(N-isopropy-
lacrylamide) (pNIPAM), which undergoes a liquid-solid
phase transition near its LCST of around 32 °C.12 In
contrast, aqueous solutions of the more polar poly(N,N-
dimethylacrylamide) (pDMA) do not show LCST behav-
ior. However, aqueous solutions of DMA copolymers
containing hydrophobic comonomers that enhance poly-
mer-polymer interactions are again thermally respon-
sive. Unlike pNIPAM, however, these DMA copolymers
tend to undergo liquid-liquid phase transitions, with
the phase transition temperature depending on the
amount of hydrophobic comonomer.13-15
Experimental Section
Materials. N,N-Dimethylacrylamide (DMA, 99%), aniline
(99%), acryloyl chloride (96%), triethylamine (99%), tris(2-
aminoethyl)amine (96%), methyl 2-chloropropionate (MCP)
(97%), copper(I) chloride (>99%), and silica gel (100-200 mesh,
column chromatography grade) were purchased from Aldrich.
Methanol (HPLC grade), distilled water (HPLC grade), tet-
rahydrofuran (THF), dimethylformamide (DMF), methylene
chloride, and anhydrous diethyl ether were obtained from
Caledon Laboratories. DMA was purified by vacuum distilla-
tion, while other materials were used as received. Tris(2-
(dimethylamino)ethyl)amine (Me6TREN) was synthesized ac-
cording to a literature procedure.19
Synthesis of N-Phenylacrylamide (PhAm). Aniline (36.9
mL, 0.4 mol) and triethylamine (84.5 mL, 0.6 mol) were
dissolved in 200 mL of methylene chloride. Acryloyl chloride
(25.4 mL, 0.3 mol) dissolved in 15 mL of methylene chloride
was added dropwise to this well-stirred solution at 0 °C. The
We have studied the thermally induced liquid-liquid
phase transitions of aqueous solutions of pDMA copoly-
mers and used their liquid-liquid phase transition to
prepare cross-linked hydrogel microspheres.16,17 We
found that the yield of phase-separated polymers was
typically less than 10% at the onset of phase separation
and increased with increasing temperatures. Despite
* Corresponding author. E-mail stoverh@mcmaster.ca.
10.1021/ma048853p CCC: $30.25 © 2005 American Chemical Society
Published on Web 02/19/2005

2110 Yin and Sto¨ver
Macromolecules, Vol. 38, No. 6, 2005
mixture was then warmed to room temperature and stirred
for another 4 h. The reaction solution was filtered, washed
three times with water, and dried over anhydrous magnesium
sulfate. The solvent was removed on a rotary evaporator to
yield a deep yellow solid that was recrystallized from acetone
and dried under vacuum to afford 22.0 g of PhAm (yield, 50%).
1H NMR (methylene chloride-d2) at 300 MHz: 7.0-7.7 ppm
(m, 5H, arom), 6.3-6.4 ppm (m, 2H, CH2 vinyl), 5.7-5.8 ppm
(dd, 1H, CH vinyl).
ATRP of DMA and PhAm. DMA, methanol, and water
were purged with nitrogen for 30 min before use. In a typical
procedure, a 50 mL flask was first flushed with nitrogen for
about 1 h. A solution comprised of DMA (7.83 g, 0.079 mol),
PhAm (3.38 g, 0.023 mol), methanol (8.97 g), water (2.24 g),
and copper(I) chloride (25.3 mg, 2.55 × 10-4 mol) was trans-
ferred to the flask under nitrogen. Me6TREN (68 µL, 2.55 ×
10-4 mol) was added with a syringe, and the flask was placed
into a water bath. After stirring the mixture for 10 min, MCP
(30.7 µL, 2.55 × 10-4 mol) was added. The polymerization was
carried out at room temperature. During the polymerization,
1.0 mL aliquots were withdrawn at different times with a
degassed syringe for molecular weight and conversion mea-
surements. The molecular weight of the polymers was mea-
sured after diluting 0.2 mL of the sample with THF and
passing this mixture through a short silica column to remove
the catalyst. Conversion was determined gravimetrically by
precipitating 0.8 mL of the aliquot into 20 mL of diethyl ether.
The polymer was isolated by centrifugation and dried under
vacuum at 65 °C. Only residual monomer was left after the
precipitation as confirmed by 1H NMR. Precipitation was
quantitative as verified by repeat precipitations. Samples for
the phase transition study similarly diluted with THF and
passed through a silica column to remove the catalyst. The
polymer was precipitated into diethyl ether, redissolved in
acetone, reprecipitated into diethyl ether, and then dried under
vacuum at 65 °C.
Figure 1. Conversion vs time for the ATRP of N,N-dimethy-
lacrylamide in methanol at room temperature under three
different monomer/initiator (M/I) ratios. Monomer/solvent )
1/2 (w/w), [CuCl]/[Me6TREN]/[2-chloropropionate] ) 1/1/1. The
solid lines connecting the data points are only to guide the
eye.
Results and Discussion
ATRP of DMA in Methanol Solution. The con-
trolled polymerization of DMA by ATRP is difficult,
plausibly due to three reasons: (1) the deactivation of
the ATRP catalyst by binding to monomeric or polymeric
amide groups, (2) the substitution of halide from the
propagating chain ends by amide, and (3) the low values
of the ATRP equilibrium constant.20,21 Such challenges
in the ATRP of DMA were addressed by using stronger
ATRP coordinating ligands and more stable alkyl chlo-
rides as initiators.22,23 The best system reported to date
uses the powerful ATRP catalyst CuCl/Me6TREN and
methyl 2-chloropropionate (MCP) as initiator; however,
high monomer conversions (up to 70%) were only
obtained with high catalyst-to-initiator ratios (2/1 or
3/1).23
Our initial studies concentrated on further improving
the ATRP of DMA by exploring the effects of solvents
on the polymerization. With CuCl/Me6TREN as the
catalyst and MCP as the initiator in a 1:1:1 ratio, the
polymerization proceeded initially very rapidly in ethyl
acetate, THF, and DMF but stopped within 5-10 min
at room temperature at less than 20% monomer conver-
sions. The polymerization proceeded more slowly in
toluene to reach about 40% monomer conversion after
20 h. In ethyl acetate, THF, and toluene, blue precipi-
tates were formed during the polymerization. In all
cases, the molecular weight distributions were below
1.1, indicating the polymerizations were well controlled.
In addition, the polymerizations continued upon adding
fresh catalyst. It hence appears that catalyst deactiva-
tion is still a major problem in ATRP of DMA, even with
CuCl/Me6TREN as the catalyst. Assuming that the
amide groups of DMA or pDMA displace chloride from
the catalyst, and then using a hydrogen-bonding solvent
such as an alcohol to bind to the amide groups, might
prevent this deactivation. Similar approaches have been
used in the ATRP of the strongly coordinating monomer,
vinylpyridine.24
Characterization of Copolymers. Molecular weights of
copolymers were determined using a gel permeation chro-
matograph consisting of a Waters 515 HPLC pump, three
UltraStyragel columns (500-20K, 500-30K, 5K-600K Da),
and a Waters 2414 refractive index detector, using THF as
solvent at a flow rate of 1 mL min-1, and narrow-disperse
1
polystyrene as calibration standards. H NMR spectra of the
copolymers were recorded on Bruker AC 200 and Bruker AC
300 NMR spectrometers, using methylene-d2 chloride as the
solvent.
Measurement of Phase Transition Temperatures. Phase
transition temperatures of aqueous solutions of DMA-co-PhAm
copolymers were measured using the cloud point method. A
Cary 100 Bio UV-vis spectrophotometer, coupled with a
temperature controller, was used to trace the phase transition
by monitoring the transmittance at 500 nm. The phase
transition temperature was defined as the inflection point of
the transmittance vs temperature curve, as determined by the
maximum in the first derivative. The polymer concentration
and the heating rate were 1 wt % and 1.0 °C min-1, respec-
tively.
Liquid-liquid Phase Transition of DMA-co-PhAm
Copolymer Solutions. Optical images of polymer solutions
at temperatures above or below the phase transition temper-
ature were recorded using an Olympus BH-2 microscope
equipped with a Kodak DC 120 Zoom digital camera.
To measure the efficiency of the phase separation, 20 g of a
2.0 wt % aqueous solution of DMA-co-PhAm22, containing 22
mol % PhAm, was prepared in a 24 mL glass vial. The solution
was incubated in a temperature-controlled water bath at 25
°C for 30 min. The resulting two-phase mixture was then
separated by centrifugation at 3500 rpm for 1 min into a
viscous coacervate bottom phase and a transparent superna-
tant, which were then isolated by decanting the supernatant
into a pan. Both phases were dried to constant weight at 65
°C.
Figures 1 and 2 depict typical curves for monomer
conversion vs time and molecular weight vs conversion,
for the ATRP of DMA in methanol in a water bath at
room temperature, with CuCl/Me6TREN as the catalyst

Macromolecules, Vol. 38, No. 6, 2005
Poly(N,N-dimethylacrylamide) Copolymer Solutions 2111
Table 1. Synthesis of Poly(DMA-co-PhAm) by ATRP at Room Temperature in Different Solventsa
mol % PhAm
theo
d
d
entry
solvent
monomer/solventb
in feed
in polymerc
conv (%)
Mn
Mn
Mw/Mn
I
MeOH
2/1
2/1
1/1
1/1
1/1
1/1
1/1
9.8
9.8
9.9
9.8
57
69
82
60
80
76
70
23600
28600
34000
25200
33600
32500
30700
6800
8200
1.07
1.07
1.07
1.07
1.09
1.07
1.07
II
MeOH/H2O (9/1)b
MeOH/H2O (9/1)
MeOH/H2O (9/1)
MeOH/H2O (4/1)
MeOH/H2O (4/1)
MeOH/H2O (4/1)
III
IV
V
VI
VII
9.8
9.6
9700
12.7
12.7
16.7
22.5
12.2
12.0
15.9
21.7
7900
10100
10200
8600
a [Monomer]/[2-chloropropionate]/[CuCl]/[Me6TREN] ) 400/1/1/1; reaction time, 4 h. b The ratios are by weight. c Measured by 1H NMR.
d Measured by GPC.
Figure 2. Plots of molecular weight, Mn (GPC), and molecular
weight distribution, Mw/Mn, vs monomer conversion for the
polymerization of N,N-dimethylacrylamide in methanol at
room temperature under three different monomer/initiator (M/
I) ratios. Monomer/solvent ) 1/2 (w/w), [CuCl]/[Me6TREN]/[2-
chloropropionate] ) 1/1/1. The solid lines are first-order
trendlines.
Figure 3. Conversion vs time curves for the ATRP of N,N-
dimethylacrylamide and phenylacrylamide in methanol/water
mixture at room temperature. Reaction conditions are those
in Table 1. The solid lines connecting the data points are only
to guide the eye.
The polymer yield increases to 69% in 90/10 (w/w)
methanol/water mixtures (Table 1-II and Figure 3-II)
and further to 82% in by increasing the monomer/
solvent ratio to 1:1 (w/w) (Table 1-III and Figure 3-III).
Under the same conditions, the yield drops to 60% when
increasing PhAm to 12.7 mol % (Table 1-IV and Figure
3-IV) but recovers back to 80% upon changing the
polymerization medium to 80/20 (w/w) methanol/water
(Table 1-V and Figure 3-V). Several other DMA-co-
PhAm copolymers with different compositions and dif-
ferent molecular weights were prepared by ATRP in 80/
20 (w/w) methanol/water mixtures (Tables 1 and 2). All
polymers have narrow molecular weight distributions,
indicating that water improves the activity of catalysts
without noticeably influencing the control of polymer-
izations. It is currently not clear that such enhancement
of catalyst activity is due to water changing the catalyst
structure26 or due to water preventing the binding of
amide to the catalyst.
A typical 1H NMR spectrum of DMA-co-PhAm copoly-
mer is presented in Figure 4. The copolymer composition
was determined by comparing the peak area of the
phenyl protons (5 H) from PhAm, with the total peak
area between 2.3 and 3.3 ppm, which includes seven
protons (a, c) from DMA and one proton (e) from PhAm.
The results are illustrated in Tables 1 and 2. On the
basis of the free radical reactivity ratios of DMA and
PhAm reported in the literature,14 it is expected that
the monomer units are randomly distributed in the
polymer chains. As shown in Tables 1 and 2, the
compositions of the copolymers are close to the monomer
feed ratio at all monomer conversions.
and MCP as the initiator (1:1:1). The polymerization is
rapid at the beginning, leading to 50% monomer conver-
sion within 30 min and 80% after 4 h, while maintaining
good control over the molecular weight distributions
(Mw/Mn < 1.1). In addition, no catalyst precipitates
during the polymerization. These results support the
hypothesis that H-bonding solvents are suitable reaction
media for the controlled ATRP of DMA.
The molecular weights of polymers determined by
GPC and shown in Figure 2 deviate from the theoretical
molecular weights, presumably due to the calibration
method involving narrow disperse polystyrene stan-
dards. The molecular weights of pDMA from GPC are
hence underestimated, though the trends are considered
reliable.
While this work was being carried out, Matyjaszews-
ki’s group disclosed results on the ATRP of DMA in
methanol solution with the addition of Lewis acids to
control the stereochemistry of the polymers.25 Under
their conditions, the monomer conversion was less than
50% in the presence of Lewis acid.
ATRP Copolymerization of DMA and PhAm in
Methanol/Water Mixtures. We next turned to the
copolymerization of DMA with PhAm in analogous
methanol solutions. Addition of 9.8 mol % PhAm led to
narrow-disperse copolymer, but in only 57% yield (Table
1, entry I, and Figure 3, curve I). It appears that the
added PhAm lowers the catalyst activity or catalyst
lifetime, by changing the solvency of the reaction
medium and/or by coordinating to the catalyst. We
hence added water as a very polar cosolvent in order to
increase the catalyst activity, in accordance with previ-
ous reports.26,27
Thermally Induced Liquid-liquid Phase Tran-
sitions of DMA-co-PhAm Copolymer Solutions. The

2112 Yin and Sto¨ver
Macromolecules, Vol. 38, No. 6, 2005
Table 2. Synthesis of Poly(DMA-co-PhAm) with Different Molecular Weights via ATRP in 80/20 (w/w) Methanol/Water
Mixtures at Room Temperaturea
mol % PhAm
theo
c
c
[monomer]/[MCP]
in feed
in polymerb
conv (%)
Mn
Mn
Mw/Mn
cloud pointd (°C)
60/1
16.7
16.7
16.7
16.7
22.5
22.5
22.5
22.5
15.9
16.2
16.3
15.9
21.8
21.7
21.7
22.0
89
84
77
76
80
73
70
61
5700
9000
2000
3500
4700
10200
3200
4600
8600
10600
1.08
1.06
1.05
1.07
1.07
1.07
1.07
1.07
39.3
40.2
39.1
43.5
16.3
17.2
20.1
20.1
100/1
200/1
400/1
100/1
200/1
400/1
600/1
16500
32500
8800
16000
30700
40200
a [CuCl]/[Me6TREN]/[ 2-chloropropionate] ) 1/1/1; monomer/solvent ) 1/1 (w/w); reaction time, 4 h. b Measured by 1H NMR. c Measured
by GPC. d 1.0 wt % polymer solution.
Table 3. Compositions and Phase Transition
Temperatures of DMA-co-PhAm Copolymers
PhAm in
cloud
pointc
(°C)
polymer
theo
b
b
polymer
(mol %)a Mn
Mn
Mw/Mn
DMA-co-PhAm10
DMA-co-PhAm12
DMA-co-PhAm16
DMA-co-PhAm22
9.6
12.0
15.9
21.7
34000
9700
1.07
1.09
1.07
1.07
33600 10100
32500 10200
70.1
43.5
20.1
30700
8600
a Estimated by NMR. b Measured by GPC. c Measured by the
cloud point method with 1.0 wt % polymer solution.
Figure 4. 1H NMR spectrum of poly(DMA-co-PhAm22) in
methylene-d2 chloride.
Figure 6. Optical microscope image of liquid droplets formed
from the phase transition of 1 wt % DMA-co-PhAm22 solution
at 25 °C. Scale bar is 250 µm.
decreases drastically (Table 3). It is thought that the
phase separation of such polymers is caused by the
presence of hydrophobic groups in the polymer
chains.28-30 At low temperature, strong hydration of the
hydrophilic groups dissolves the polymer and hence
forces entropically unfavorable structuring of water
around the hydrophobic groups. Above the phase transi-
tion temperature, the now dominant entropy gain drives
the dissociation of this water from the polymer and leads
to a negative free energy of desolvation. An increase in
hydrobicity of the polymer increases the entropic pen-
alty from this structured water, lowers the enthalpy
gained from solvating the hydrophilic units, and hence
decreases the cloud point.
The thermally induced phase separation observed in
aqueous DMA-co-PhAm copolymer solution is a liquid-
liquid transition. Figure 6 presents an optical micro-
scope image of the liquid coacervate droplets formed
from a 1.0 wt % aqueous DMA-co-PhAm22 solution. In
the absence of stirring, these droplets will coalesce into
a bulk layer after several minutes.
Figure 5. Cloud point curves for 1.0 wt % aqueous solutions
of DMA-co-PhAm copolymers with different compositions.
temperature-dependent phase separation of aqueous
solutions of DMA-co-PhAm copolymers was investigated
by the cloud point method. Figure 5 shows the trans-
mittance vs temperature curves for three DMA-co-
PhAm copolymers with different compositions. The
copolymer characteristics and their corresponding cloud
point temperatures are given in Table 3.
The cloud point temperature of thermally responsive
polymers is strongly influenced by changes in the
hydrophilic/hydrophobic nature of the polymer. Aqueous
solutions of DMA-co-PhAm copolymers with less than
10 mol % PhAm do not show a cloud point. With
increasing PhAm content, the cloud point temperature

Macromolecules, Vol. 38, No. 6, 2005
Poly(N,N-dimethylacrylamide) Copolymer Solutions 2113
Figure 8. Cloud point curves for aqueous solutions of DMA-
co-PhAm22 with different concentrations.
Figure 7. Effects of polymer molecular weights on the phase
transition temperature of 1.0 wt % DMA-co-PhAm copolymer
solutions. The PhAm content in the copolymer is about (A) 16
mol % and (B) 22 mol %. The molecular weights indicated are
measured by GPC (see Table 2).
Figure 9. Effect of polymer concentration on the phase
separation temperature of DMA-co-PhAm solutions as mea-
sured by the cloud point method.
Because of the difficulty in obtaining thermally
responsive polymers with narrow molecular weight
distributions, the dependence of LCST on the molecular
weight of polymers has not been systematically studied.
For example, Fujishige et al. reported the LCST of
pNIPAM to be independent of chain length for molecular
weights larger than 50 000,31 while Schild and Tirrell
indicated a small effect of molecular weight on the LCST
of pNIPAM.32 The pNIPAM used in both studies was
prepared by conventional radical polymerization and
had broad molecular weight distributions. Using frac-
tionated pNIPAMs with narrow molecular weight dis-
tributions, Tong et al. reported that the LCST of
pNIPAM increases with molecular weight.33
Figure 7 illustrates the effect of molecular weight on
the cloud point of two series of DMA-co-PhAm copoly-
mers, containing 16% and 22% PhAm. The phase
transition temperatures first increase slightly with
increasing molecular weight and then become constant
(Figure 7B). The polymers with lower molecular weights
show phase transitions over relatively broad tempera-
ture ranges. This may be due small compositional
variations or due to the influence of the initiating
methoxypropionate group. Both of these effects would
be more pronounced at lower molecular weights.
positions, respectively. The cloud point temperatures
initially decrease with increasing polymer concentra-
tions, leveling off at concentrations above 1.0 wt %
(Figure 9). At low polymer concentration (0.05 wt %),
the phase transition occurs over a broad temperature
range (Figure 8). These observations are in agreement
with the following considerations: the thermally in-
duced phase separation of aqueous polymer solutions
has been proposed to take place in two stages, with
individual polymer chains first forming primary ag-
gregates, which then macroscopically aggregate to cause
phase separation.34 The primary aggregates require
higher temperatures before aggregating macroscopi-
cally. Higher polymer concentrations would also shift
this equilibrium, causing the cloud point to occur at a
lower temperature. Compared to the liquid-solid phase
transition of pNIPAM solution,33,35 the effect of polymer
concentration on the phase transition temperature of
DMA-co-PhAm solution is hence more pronounced.
Similar observations were reported for the liquid-liquid
phase transition of elastin-like polypeptide solutions.36,37
It is well-known that the effect of salts on the cloud
point is mainly due to the resulting changes in the
structure of water.32,38 As shown in Figure 10, the cloud
point temperature of DMA-co-PhAm solutions decreases
with the addition of NaCl. NaCl is considered as a water
structure maker and hence shows a “salting-out” effect,
which can effectively enhance the hydrophobic interac-
tion and promote the collapse of polymer chains.
Effects of Polymer Concentration and Added
NaCl on the Phase Transition Temperatures. Fig-
ures 8 and 9 show the effects of polymer concentration
on the cloud point curves for DMA-co-PhAm22 and on
the cloud point temperatures for two copolymer com-

2114 Yin and Sto¨ver
Macromolecules, Vol. 38, No. 6, 2005
Figure 10. Effect of NaCl concentration on the phase separa-
tion temperature of 1.0 wt % DMA-co-PhAm solutions as
measured by the cloud point method.
Figure 12. Concentrations of DMA-co-PhAm22 in the two
separated phases at different temperatures.
peratures are required to drive more polymer to the
coacervate phase.
Figure 12 shows the concentration of DMA-co-
PhAm22 in the two separated phases as a function of
temperature. The temperature changes do not affect the
volume of the two separated phases. Hence, the con-
centrations of polymer in the coacervate phase increase
with increasing temperatures. The water content in the
coacervate phase is however still larger than 60% at 75
°C.
The polymers recovered from both phases were char-
acterized by GPC and NMR and were found to have
similar compositions and molecular weights.
As discussed above, the cloud point temperatures
decrease upon addition of NaCl to the solutions. Also
illustrated in Figure 11 is the effect of NaCl addition
on the efficiency of the phase separation. During the
polymer recovery, it was assumed that NaCl is present
in equal concentrations in both the polymer-poor and
the polymer-rich (coacervate) phases. With the addition
of NaCl, the volume of both separated phase does not
change, while the efficiency of phase separations is
enhanced.
Figure 11. Effect of temperature on the yield of phase-
separated polymer from 2.0 wt % DMA-co-PhAm copolymer
aqueous solutions and in 1.0 mol L-1 NaCl solutions. The
symbols indicate the equilibrium temperature, and the error
bars show the decrease of temperature during centrifugation.
Efficiency of Phase Separation. The efficiency of
the thermally induced liquid-liquid phase separations
was investigated by determining the yield of polymer
recovered from the coacervate phase at different tem-
peratures above the cloud points. Figure 11 shows the
percentage of the polymer found in the coacervate phase
to increase with increasing temperature. Compared to
our previously studied copolymers of DMA with glycidyl
methacrylate16 and allyl methacrylate,17 the yields of
the present DMA-PhAm copolymers are much higher
near the onset of phase transition. The polymers used
in the previous work were prepared by conventional free
radical polymerization and subsequently had broad
distributions of polymer molecular weights. Though
based on different hydrophobic comonomers, the present
results suggest that the efficiency of the phase transition
can be increased by using narrow-disperse polymers.
Furthermore, the temperature-dependent efficiency of
phase separation, even for narrow-disperse copolymers,
confirms that the thermally induced liquid-liquid phase
separation is an equilibrium process. In this process,
only part of the polymer chains are dehydrated upon
reaching the cloud point temperature, and higher tem-
Conclusions
pDMA homopolymer and DMA-co-PhAm copolymers
were synthesized by ATRP using MCP/CuCl/Me6TREN
as the initiator and the catalyst at room temperature,
with the hydrogen-bonding methanol and methanol/
water mixtures as the reaction media. The polymeriza-
tions were well controlled, leading to the production of
DMA-co-PhAm copolymers with well-defined structures
and compositions.
Aqueous solutions of DMA-co-PhAm copolymers
showed liquid-liquid phase separations at tempera-
tures dependent on the content of hydrophobic PhAm.
These cloud points were also found to depend on the
polymer concentrations and molecular weights. The
efficiency of the thermally induced liquid-liquid phase
separation was studied by measuring the yield of phase-
separated polymers, which increased with increasing
temperature. This temperature dependence of the phase
separation efficiency showed the thermally induced
liquid-liquid phase separation to be an equilibrium
process. The polymer chains only partially dehydrate
upon reaching the cloud point, and the solubility of

Macromolecules, Vol. 38, No. 6, 2005
Poly(N,N-dimethylacrylamide) Copolymer Solutions 2115
(17) Yin, X.; Sto¨ver, H. D. H. J. Polym. Sci., Part A: Polym. Chem.,
in press.
(18) Matyjaszewski, K.; Xia, J. Chem. Rev. 2001, 101, 2921-2990.
(19) Ciampolini, M.; Nardi, N. Inorg. Chem. 1966, 5, 41-44.
(20) Teodorescu, M.; Matyjaszewski, K. Macromolecules 1999, 32,
4826-4831.
(21) Rademacher, J. T.; Baum, M.; Pallack, M. E.; Brittain, W.
J.; Simonsick, W. J. Macromolecules 2000, 33, 284-288.
(22) Teodorescu, M.; Matyjaszewski, K. Macromol. Rapid Com-
mun. 2000, 21, 190-194.
polymers in the equilibrium phase progressively de-
creased with increasing temperature, leading to more
polymers driven from the continuous to the coacervate
phase. The addition of NaCl further enhances the
efficiency of phase separation without changing the
water content in the separated phases.
Acknowledgment. We thank Dr. Nicholas Burke
for valuable discussions, Yan Xia for assistance with the
1H NMR, Professor P. Berti for the use of the variable
temperature Cary 50 UV/vis instrument, and the Natu-
ral Sciences and Engineering Research Council of
Canada and 3M Canada Company for financial support.
(23) Neugebauer, D.; Matyjaszewski, K. Macromolecules 2003, 36,
2598-2603.
(24) Xia, J.; Zhang, X.; Matyjaszewski, K. Macromolecules 1999,
32, 3531-3533.
(25) Lutz, J.; Neugebauer, D.; Matyjaszewski, K. J. Am. Chem.
Soc. 2003, 125, 6986-6993.
(26) Wang, X. S.; Armes, S. P. Macromolecules 2000, 33, 6640-
6647.
References and Notes
(1) Albertsson, P-Å. Partition of Cell Particles and Macromol-
ecules, 3rd ed.; J. Wiley & Sons: New York, 1986.
(2) Li, M.; Zhu, Z. Q.; Mei, L. H. Biotechnol. Prog. 1997, 13, 105-
108.
(27) Robinson, K. L.; Khan, M. A.; de Paz Ba´n˜ez; M. V.; Wang, X.
S.; Armes, S. P. Macromolecules 2001, 34, 3155-3158.
(28) Takei, Y. G.; Aoki, T.; Sanui, K.; Ogata, N.; Okano, T.;
Sakurai, Y. Bioconjugate Chem. 1993, 4, 341-346.
(29) Feil, H.; Bae, Y. H.; Feijen, J.; Kim, S. W. Macromolecules
1993, 26, 2496-2500.
(3) Johansson, H.-O.; Karlstro¨m, G.; Tjerneld, F.; Haynes, C. A.
J. Chromatogr., B 1998, 711, 3-17.
(4) Abbott, N. L.; Blankschtein, D.; Hatton, T. A. Macromolecules
1991, 24, 4334-4348.
(30) Shibayama, M.; Mizutani, S.; Nomura, S. Macromolecules
1996, 29, 2019-2024.
(5) Carlsson, M.; Linse, P.; Tjerneld, F. Macromolecules 1993,
26, 1546-1554.
(31) Fujishige, S.; Kubota, K.; Ando, I. J. Phys. Chem. 1989, 93,
3311-3313.
(6) Persson, J.; Nystro¨m, L.; Ageland, H.; Tjerneld, F. J. Chro-
matogr., B 1998, 711, 97-109.
(32) Schild, H. G.; Tirrell, D. A. J. Phys. Chem. 1990, 94, 4352-
4356.
(7) Johansson, H.-O.; Karlstro¨m, G.; Tjerneld, F. Macromolecules
1993, 26, 4478-4483.
(33) Tong, Z.; Zeng, F.; Zheng, X.; Sato, T. Macromolecules 1999,
32, 4488-4490.
(34) Qiu, X.; Wu, C. Macromolecules 1997, 30, 7921-7926.
(35) Otake, K.; Inomata, H.; Konno, M.; Saito, S. Macromolecules
1990, 23, 283-289.
(36) Nagarsekar, A.; Crissman, J.; Crissman, M.; Ferrari, F.;
Cappello, J.; Ghandehari, H. Biomacromolecules 2003, 4,
602-607.
(37) Yamaoka, T.; Tamura, T.; Seto, Y.; Tada, T.; Kunugi, S.;
Tirrell, D. A. Biomacromolecules 2003, 4, 1680-1685.
(38) Park, T. G.; Hoffman, A. S. Macromolecules 1993, 26,
5045-5048.
(8) Johansson, H.-O.; Persson, J.; Tjerneld, F. Biotechnol. Bioeng.
1999, 66, 247-257.
(9) Urry, D. W. J. Phys. Chem. B 1997, 101, 11007-11028.
(10) Meyer, D. A.; Chilkoti, A. Nat. Biotechnol. 1999, 17, 1112-
1115.
(11) Zhang, Y.; Mao, H.; Cremer, P. S. J. Am. Chem. Soc. 2003,
125, 15630-15635.
(12) Schild, H. G. Prog. Polym. Sci. 1992, 17, 163-249.
(13) Mueller, K. F. Polymer 1992, 33, 3470-3476.
(14) Miyazaki, H.; Kataoka, K. Polymer 1996, 37, 681-685.
(15) Shibanuma, T.; Aoki, T.; Sanui, K.; Ogata, N.; Kikuchi, A.;
Sakurai, Y.; Okano, T. Macromolecules 2000, 33, 444-450.
(16) Yin, X.; Sto¨ver, H. D. H. Macromolecules 2003, 36, 9817-
9822.
MA048853P