Engineering a sharp physiological transition state for poly(n- isopropylacrylamide) through structural control

Article abstract of DOI:10.1002/pola.25854

Poly(N-isopropylacrylamide) (pNIPAAm), a well-studied, biologically inert polymer that undergoes a sharp aqueous thermal transition at 32 °C, has been a subject of widespread interest for possible biological applications. A major hindrance to its successful application is due to the difficulty of maintaining a sharp transition when the polymer is modified for a physiological transition temperature, especially in isotonic solutions. Current copolymer blends raise the transition temperature but also make the transition significantly broader. We have combined the use of reversible addition-fragmentation chain transfer (RAFT) polymerization with tacticity control to synthesize well-defined pNIPAAm that demonstrates sharp transitions under physiological conditions. By selecting a RAFT agent with appropriate end groups, controlling molecular weight, and increasing the racemo diad content, we were able to increase the thermal transition temperature of pure pNIPAAm to a sharp transition at 37.6 °C under isotonic conditions.

Full text of DOI:10.1002/pola.25854

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
Engineering a Sharp Physiological Transition State for
Poly(n-isopropylacrylamide) Through Structural Control
Kai Chang, Zachary T. Dicke, Lakeshia J. Taite
School of Chemical & Biomolecular Engineering, Georgia Institute of Technology, 311 Ferst Drive NW, Atlanta,
Georgia 30332-0100
Correspondence to: L. J. Taite (E-mail: lakeshia.taite@chbe.gatech.edu)
Received 5 August 2011; accepted 11 November 2011; published online 29 November 2011
DOI: 10.1002/pola.25854
ABSTRACT: Poly(N-isopropylacrylamide) (pNIPAAm),
a
well-
pNIPAAm that demonstrates sharp transitions under physiologi-
cal conditions. By selecting a RAFT agent with appropriate end
groups, controlling molecular weight, and increasing the racemo
diad content, we were able to increase the thermal transition tem-
studied, biologically inert polymer that undergoes a sharp aque-
ꢀ
ous thermal transition at 32 C, has been a subject of widespread
interest for possible biological applications. A major hindrance
to its successful application is due to the difficulty of maintaining
a sharp transition when the polymer is modified for a physiologi-
cal transition temperature, especially in isotonic solutions. Cur-
rent copolymer blends raise the transition temperature but also
make the transition significantly broader. We have combined the
use of reversible addition-fragmentation chain transfer (RAFT)
polymerization with tacticity control to synthesize well-defined
ꢀ
perature of pure pNIPAAm to a sharp transition at 37.6 C under
isotonic conditions. VC 2011 Wiley Periodicals, Inc. J Polym Sci
Part A: Polym Chem 50: 976–985, 2012
KEYWORDS: reversible addition fragmentation chain transfer
(RAFT); stimuli-sensitive polymers; structure-property relations;
syndiotactic
INTRODUCTION New developments in biomedical diagnos-
tics, theranostics, and sensing applications increasingly rely
upon ‘‘smart’’ materials, materials which have properties that
can be triggered to change upon exposure to an external
stimulus. One of the most well-studied polymers in this class
is poly(N-isopropylacrylamide) (pNIPAAm), a biologically
pNIPAAm-based thermoresponsive polymers lies in the prep-
aration of amphiphilic random, block, graft, or star-shaped
copolymers for biomedical applications, through a combina-
tion of simple LCST modification techniques, we have synthe-
sized a pure pNIPAAm that can be manipulated to an LCST
at physiological temperature under isotonic conditions.
These polymers, and the ease with which the transition tem-
perature can be modified using these synthesis techniques
we describe, present a novel strategy for the formation of
polymer systems with highly homogeneous properties that
can be used in applications in biotechnology.
1
inert polymer that exhibits an aqueous lower critical solu-
ꢀ
2
tion temperature (LCST) of 32 C. Because of this well-
defined transition, pNIPAAm has been extensively studied by
many groups for applications in biotechnology, ranging from
3
4,5
6
protein purification to drug delivery to biosensing. One
of the many requirements for successful application in bio-
logical systems is an LCST at physiological or higher temper-
The LCST of pNIPAAm is influenced by a variety of factors,
and there are several methods used to modify it. The most
common is to copolymerize with a small amount of hydro-
philic comonomer. Although several copolymer blends exist
3
,5,7,8
atures.
For instance, although current technology ena-
bles the use of pNIPAAm systems in protein purification
3
from acellular systems, increasing the LCST to physiological
9
,10
temperatures would make it possible to use pNIPAAm to
extract hydrophilic proteins in the context of a continuing
cell culture without extreme stresses on the cells. In addi-
tion, pNIPAAm-based controlled drug delivery systems also
require LCSTs at least as high as physiological temperatures
to use these properties as drug release triggers. Because
these applications must operate within very narrow temper-
ature ranges, the ability to manipulate the LCST to higher
temperatures without sacrificing the sharpness of the transi-
tion is essential. Although the current trend in research on
to raise the LCST with sharp transitions,
some of the
most popular blends for biological applications such as
acrylic acid (AAc) have a widening effect on the LCST.
Copolymerization with AAc has the effect of widening the
ꢀ
LCST from a transition that occurs over <0.5 C to a transi-
ꢀ
tion that takes place over a range of 5–10 C or even larger
depending on the desired LCST.
2
,7,11
This increase in transi-
tion range is due to the inhibition of water exclusion and is
less than ideal for any application requiring a sharply
1
2
defined response.
Nevertheless, copolymerizing as a
Additional Supporting Information may be found in the online version of this article.
2011 Wiley Periodicals, Inc.
V
C
976
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 976–985

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
method of raising LCST is widely used, and prior research
has examined the use of comonomers precisely designed to
cal applications. Furthermore, the incorporation of this living
process allows for additional chain extension polymerization,
which can be used in the synthesis of hydrogels or for the
addition of functionalities through subsequent polymeriza-
tions and conjugations.
1
3
change the LCST under certain conditions. This concept
has recently extended into attaching specific end groups to
the polymer, which can significantly affect lower molecular
weight (MW) pNIPAAm although its utility is inversely pro-
1
4
EXPERIMENTAL
portional to the MW.
N-Isopropylacrylamide was purchased from TCI America and
recrystallized in a 9:1 ratio of hexanes:benzene. Carbon di-
sulfide, tetrabutylammonium hydrogen sulfate, mineral spi-
rits, 1,4 dioxane, Aliquat 336, and 3-methyl-3-pentanol
Another method of modifying the LCST is through changing
the polymer architecture. Branching, for example, has been
shown to lower the LCST of pNIPAAm, whereas stereospe-
cific polymerization and control over the tacticity of the
polymer have been shown to increase or decrease the LCST,
depending on whether the polymer is syndiotactic or isotac-
(
3Me3PenOH) were purchased from Sigma Aldrich and used
without further purification. Chloroform and acetone were
purchased from BDH Chemicals and used without further
purification.
1
5–22
tic.
For the purposes of raising the LCST, selectively pol-
1
9
ymerizing in the racemo conformation can be a useful tool.
This method modifies the rotational energy required to ori-
ent the polymer such that it undergoes the cooperative dehy-
dration that is observed macroscopically as the LCST.
0
0
00
S,S -Bis(a,a -dimethyl-a -acetic acid)trithiocarbonate (1)
Synthesis
2
3,24
Synthesis of 1 was done similarly to the procedure set forth
pNIPAAm with a majority of meso diads exhibits a lower
LCST, whereas pNIPAAm with a majority of racemo diads
3
0
by Lai et al. A total of 6.62 mL (0.1 mol) acetone was
reacted with 7.26 mL (0.1 mol) chloroform, 2.16 mL (0.04
mol) carbon disulfide, and 0.241 g (0.7 mmol) tetrabutylam-
monium hydrogen sulfate in 12 mL of mineral spirits. The
reaction mixture was purged with nitrogen for 5 min and run
in a water bath at room temperature. Ten milliliters of 50%
NaOH was added dropwise over 90 min and the reaction was
left to run overnight. Ninety milliliters of water was then
added, followed by 42 mL of 6 N HCl. The reaction mixture
was then purged under nitrogen for half an hour and filtered.
The resulting product was recrystallized in acetone to yield 4
g of product. Synthesis of 1 was confirmed by electrospray
mass spectrometry (see Supporting Information).
1
9,21,22
exhibits a higher LCST.
The difference in LCST seen
ꢀ
using this method is usually on the order of 3–5 C, a signifi-
cant but limited enhancement.
In addition to the intrinsic properties of the polymer, solvent
properties play a large role in the observed transition tem-
perature. Certain cosolvents such as methanol can signifi-
2
5
cantly reduce the LCST, whereas pH can have a mild
1
1
26
effect and salts can have a large effect on the LCST. This
is especially important in biological applications because bio-
logical systems require certain osmolarity and salt concen-
trations, without which cells will undergo apoptosis. The cu-
mulative result of these LCST-depressing effects renders the
polymer all but useless for biological applications in its
native form.
Polymerization
Polymerization of NIPAAm was carried out under six differ-
ent conditions. High transition temperature pNIPAAm was
synthesized using a ‘‘temperature shock’’ treatment in which
Secondary to the issue of LCST manipulation is the need for
well-defined polymers, a problem that is easily rectified
using a living radical polymerization scheme, in this case, re-
versible addition-fragmentation chain transfer (RAFT) poly-
merization. RAFT polymerization controls the polydispersity
ꢀ
the reaction was thermally initiated at 65 C for 1 h and im-
mediately placed into a room temperature bath to react at
room temperature for the rest of the polymerization time,
typically 95 h. The purpose of this method was to slow the
reaction kinetics to allow for better tacticity control. It also
served as a way to control for MW. Typically, a 3.2 g mixture
of 100:1:0.5 ratio of NIPAAm:1:azobisisobutyronitrile (AIBN)
was placed in a sealed 25-mL round-bottom flask equipped
with a magnetic stir bar. The mixture was purged with nitro-
gen for 10 min and 20 mL of nitrogen-purged 1,4 dioxane
index (PDI, defined as the weight-average MW, M , divided
w
by the number-average MW, M ) of a polymer by introducing
n
7,28
2
a chain transfer agent (CTA).
The CTA is reversibly reac-
tive toward growing polymer chains and forms a dynamic
equilibrium between the actively growing chains and the
dormant chains. This reduces the number of actively poly-
merizing chains and in effect increases the polymerization
ꢀ
2
9
0
0
was added. The solution was reacted at 65 C for 1 h and at
room temperature for 95 h.
time. The CTA chosen for this study was S,S -bis(a,a -di-
0
0
methyl-a -acetic acid)trithiocarbonate (1),
a
well-docu-
that has been previously shown
to be very versatile and produce good results with pNI-
3
0–35
mented symmetric CTA
To test the effects of majority of racemo diads, 3Me3PenOH
was added to the reaction mixture. Accordingly, a 3.2 g mix-
ture of 100:1:0.5 ratio of NIPAAm:1:AIBN was placed in a
sealed 50-mL round-bottom flask equipped with a magnetic
stir bar. A total of 6.7 mL of 3Me3PenOH was added to the
reaction mixture. The mixture was purged with nitrogen for
10 min and 20 mL of nitrogen-purged 1,4 dioxane was
3
1,34,35
PAAm.
By incorporating RAFT polymerization along
with the LCST manipulation principles outlined above, we
have successfully synthesized well-defined pNIPAAm with a
ꢀ
sharp LCST of 37.6 C in phosphate-buffered saline (PBS)
without the use of copolymerization, thereby introducing
another method to optimize pNIPAAm synthesis for biologi-
ꢀ
added. The solution was reacted at 65 C for 1 or 1.5 h to
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 976–985
977

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
initiate polymerization and at room temperature for up to
5 h thereafter.
9
Control polymers were synthesized using typical RAFT poly-
merization techniques with 1. Briefly, a 3.2 g mixture of
100:1:0.5 ratio of NIPAAm:1:AIBN was placed in a sealed 25-
mL round-bottom flask equipped with a magnetic stir bar. A
total of 6.7 mL of 3Me3PenOH was added to the reaction
mixture for a control polymer with majority of racemo diads,
whereas this step was omitted for the atactic polymer con-
trol. The mixture was purged with nitrogen for 10 min and
2
0 mL of nitrogen-purged 1,4 dioxane was added. The solu-
ꢀ
tion was reacted at 65 C for 48 h.
For radical-polymerized pNIPAAm controls, a 3.2 g mixture
of 100:1 ratio of NIPAAm:AIBN was placed in a sealed 25-
mL round-bottom flask equipped with a magnetic stir bar.
The mixture was purged with nitrogen for 10 min and 20
mL of nitrogen-purged 1,4 dioxane was added. The solution
1
FIGURE 1 H NMR spectra of pNIPAAm synthesized using 1 in
ꢀ
chloroform-d. This polymer was synthesized at 65 C for 48 h
ꢀ
was reacted at 65 C for 48 h.
(
pNIPAAm-1-HT; GPC M
n
¼ 7541, PDI ¼ 1.23). Peaks b and c
correspond to residual solvent peaks of 1,4 dioxane and water,
respectively. The peaks shown in d correspond to polymer
backbone peaks in various configurations.
A ‘‘temperature shock’’ radical polymerization control experi-
ment was also conducted in which a 3.2 g mixture of 100:1
ratio of NIPAAm:AIBN was placed in a sealed 25-mL round-
bottom flask equipped with a magnetic stir bar. The mixture
was purged with nitrogen for 10 min and 20 mL of nitro-
gen-purged 1,4 dioxane was added. The solution was reacted
Varian Mercury Vx 400 spectrometer using DMSO-d
solvent at 90 C. The high temperature was used to resolve
the methine backbone peaks.
as a
6
ꢀ
1
8,36,37
ꢀ
at 65 C for 1 h as a thermal initiation or temperature shock
Mass spectrometry was
and then removed from the heat to react at room tempera-
ture for 95 h.
run on an Applied Biosystems 4700 Proteomics Analyzer
with a 200 Hz Nd:YAG laser using CHCA matrix and reflect-
ing detector. UV–vis spectrometry was conducted using a
Cary 50 UV–Vis Spectrophotometer (Agilent) with the single
cell Peltier thermostatted cell holder and accessory for tem-
perature control. Temperature was ramped at a rate of 0.5
A copolymer with 4% AAc, pNIPAAm-co-AAc, was also syn-
thesized to compare with the results. This was done by
reacting 1.5 g of NIPAAm with 37.9 lL of AAc and 2.18 g of
AIBN in a sealed 25-mL round-bottom flask equipped with a
magnetic stir bar. The mixture was purged with nitrogen for
ꢀ
ꢀ
C/min and data points were taken every 0.1 C.
1
0 min and 20 mL of nitrogen-purged 1,4 dioxane was
RESULTS AND DISCUSSION
added. The solution was reacted for various lengths of time
at 65 C. An 8% AAc copolymer was also synthesized using
ꢀ
MW Control
pNIPAAm was synthesized under various reaction times at
room temperature. The initial reaction temperature of all
polymers was 65 C to thermally initiate the reactions. This
the same method and incorporated 75.8 lL of AAc.
Upon completion of reactions, all pNIPAAm samples were
precipitated in anhydrous diethyl ether and collected via fil-
tration. The samples were then dissolved in nanopure water
and dialyzed with a 2000 MWCO membrane. The water was
changed at 1 h, 3 h, and overnight. The samples were then
frozen and lyophilized. pNIPAAm and majority of syndiotac-
tic pNIPAAm polymerized by this method are hereafter
denoted as pNIPAAm-1 and pNIPAAm-1s, respectively.
ꢀ
was maintained for 1 h, and the reactions were then placed
in room temperature baths to slowly polymerize over the
course of 7 days. This ‘‘temperature shock’’ treatment was
used to form well-controlled low-molecular-weight pNIPAAm.
The primary goal of using this method, rather than control
using feed concentrations, is to allow for better tacticity con-
trol because it has been previously shown that reducing po-
lymerization temperature increases the efficacy of bulky
alcohols and Lewis bases as syndiotacticity-inducing
Characterization
Polymers were characterized using gel permeation chroma-
tography (GPC), NMR, matrix-assisted laser desorption ioni-
zation mass spectrometry (MALDI), and ultraviolet (UV)–visi-
ble spectrometry. GPC was conducted on a PL-GPC 50 with
UV, refractive index (RI), and evaporative light scattering
1
8,38
1
agents.
pNIPAAm synthesis was confirmed using
H
NMR as shown in Figure 1.
GPC was conducted on pNIPAAm-1 at various reaction times
using a RI detector; traces are shown in Figure 2. Clearly,
the MW continues to increase with time, indicating contin-
ued polymerization after reaching room temperature. The
MWs and PDIs for the polymers are shown in Table 1. As
expected for RAFT polymerization, the PDIs exhibited by the
polymers are low, on the order of 1.1. This indicates good
(
ELS) detectors (Agilent) equipped with two Plgel 3-lm
MIXED-E columns. Filtered stabilized tetrahydrofuran was
used as the polymer solvent and eluent at a flow rate of 1
mL/min. Chromatograms were compared with those of poly-
1
styrene standards (Agilent). H NMR was conducted on a
978
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 976–985

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
FIGURE 2 pNIPAAm-1 polymerized for various lengths of time using a 100:1:0.5 ratio of NIPAAm:1:AIBN. The time indicated is the
ꢀ
ꢀ
total reaction time with 1 h signifying a reaction for 1 h at 65 C, 3 h signifying reaction for 1 h at 65 C, 2 h at room temperature,
and so forth. As seen in the figure, molecular weight increases with increasing time with the dotted blue line representing the
peak retention time for 1 h of polymerization.
TABLE 1 The Molecular Weights and PDIs for the Polymers
Total
Feed Molar Ratio
NIPAAm:1:AIBN)
Polymerization
Time (h)
Conversion
(%)
a
b
(
M
n
(NMR)
M
w
(GPC)
M
n
(GPC)
PDI
pNIPAAm-1-1h
pNIPAAm-1-3h
pNIPAAm-1-6h
pNIPAAm-1-12h
pNIPAAm-1-1d
pNIPAAm-1-2d
pNIPAAm-1-3d
pNIPAAm-1-4d
pNIPAAm-1-7d
pNIPAAm-r-1h
pNIPAAm-r-3h
pNIPAAm-r-6h
pNIPAAm-r-12h
pNIPAAm-r-1d
pNIPAAm-r-2d
pNIPAAm-r-3d
pNIPAAm-r-7d
pNIPAAm-1-HT
100:1:0.5
100:1:0.5
100:1:0.5
100:1:0.5
100:1:0.5
100:1:0.5
100:1:0.5
100:1:0.5
100:1:0.5
100:0:1
1
3
3,194
3,316
3,597
3,815
3,974
3,981
4,122
4,127
4,150
10,699
10,732
10,350
10,330
10,668
10,466
9,958
10,946
5,762
3,825
4,005
4,014
4,128
4,157
4,219
4,251
4,288
4,328
11,333
14,533
15,277
16,333
16,555
19,840
19,752
7,402
9,338
3,483
3,665
3,655
3,686
3,776
3,838
3,866
3,885
3,961
6,902
9,615
9,231
9,242
9,372
13,231
13,188
4,743
7,541
1.10
1.09
1.10
1.12
1.10
1.10
1.10
1.10
1.09
1.64
1.51
1.66
1.77
1.77
1.50
1.50
1.56
1.23
55.2
57.3
62.1
65.9
68.6
68.8
71.2
71.3
71.7
92.4
92.7
89.4
89.2
92.1
90.4
86.0
94.5
97.0
6
12
24
48
72
96
168
1
100:0:1
3
100:0:1
6
100:0:1
12
24
48
72
168
48
100:0:1
100:0:1
100:0:1
100:0:1
100:1:0.5
pNIPAAm was polymerized for varying periods of time at room temperature using 0.9 M NIPAAm solution in 1,4 dioxane. Thermal initiation was
ꢀ
conducted by polymerizing in a preheated 65 C oil bath for 1 h and subsequent polymerization was conducted at room temperature. A high temper-
ꢀ
ature control polymer (pNIPAAm-1-HT) was polymerized at 65 C for 48 h.
a
Theoretical molecular weight calculated by multiplying conversion and theoretical maximum molecular weight based on feed ratios.
b
PDI ¼ M
w n
(GPC)/M (GPC).
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 976–985
979

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
FIGURE 3 Molecular weight versus conversion of (A) pNIPAAm-1 series and (B) pNIPAAm-1s series. The solid line represents the-
2
n
oretical values. Linear correlations between conversion and M (R values of 0.86 and 0.96 for pNIPAAm-1 series and pNIPAAm-1s
series, respectively) confirm ‘‘living’’ controlled radical polymerization. Nearly constant low PDI indicates that conversion is inde-
pendent of PDI for these levels of conversion.
control despite the low conversion and overall MW. Free rad-
ical polymerization, however, imparts a typically higher PDI
ity of syndiotactic polymer. The results of this experiment
are seen in Table 2 and Figure 3(B). As expected, the linear
relationship between conversion and Mn continues to be
observed. The PDIs of the system (on the order of 1.15) are
also within the range of RAFT polymerization although they
are slightly higher than those polymerized without the pres-
ence of 3Me3PenOH.
(>1.5). The table also indicates relatively consistent MWs
regardless of polymerization times for free radical polymer-
ization, indicating that within an hour of polymerization the
reaction has already approached completion and further po-
lymerization time at room temperature did not affect conver-
sion or final MW.
Although both polymerization methods exhibited CRP, a
direct comparison could not be made because of a longer
thermal initiation time for the pNIPAAm-1s-a to pNIPAAm-
1s-g series. Subsequently, a new polymer (pNIPAAm-1s-4d,
When the MW of RAFT-polymerized pNIPAAm (the pNI-
PAAm-1 series) was plotted against the conversion, the
approximation of a straight line, as seen in Figure 3, further
confirms RAFT polymerization. Notably, although pNIPAAm
M
n
¼ 4100, PDI ¼ 1.15) was synthesized to directly compare
ꢀ
polymerized for varying reaction times at a constant 65
C
pNIPAAm-1-4d with a majority of racemo diad version of the
same polymer (pNIPAAm-1s-4d). Both polymers were poly-
merized under the same conditions for 4 days, and the GPC
traces are shown in Figure 4. An ELS detector was used
because of its higher sensitivity to low concentrations of
polymer. As seen in Figure 4, pNIPAAm-1s-4d is slightly
larger than the pNIPAAm-1-4d. This is as expected because
the bulky alcohol acts as an accelerator during the polymer-
ization process when used in conjunction with free radical
reached 97% conversion within 4 h, the level of conversion
only reached 71.1% after 7 days of polymerization at room
temperature. This is to be expected because of the much
higher glass transition temperature (T ) of pNIPAAm (135
g
ꢀ
39
C) when compared with the reaction temperature. At
moderate to high conversions, polymerization slows down
considerably because of the vitrification effect.
4
0–43
This
effect is more pronounced when polymerized at room tem-
ꢀ
19
perature than when polymerized at 65 C because of the
polymerization. Therefore, it is not surprising that it has a
larger temperature difference between T and the reaction
similar effect in RAFT polymerization. This acceleration may
have also contributed to the slightly higher PDI of pNIPAAm-
1s. In addition to having a larger polymer overall, pNIPAAm-
1s shows a small peak at 12.5 min into the elution. This
peak corresponds to higher MW polymers and/or aggregates
(15,200 Da) that may have formed as a result of termination
reactions and represent 2.4% of the total polymers. The
traces indicate that this sample of pNIPAAm-1-4d has a MW
of 3700 Da with a PDI of 1.13, whereas pNIPAAm-1s-4d has
a MW of 4100 Da with a PDI of 1.15.
g
temperature. Nearly complete conversion such as that seen
in the high-temperature polymerization is therefore not
expected. Nevertheless, the values shown in Figure 3(A) indi-
cate that reducing the reaction temperature, while slowing
the reaction kinetics, did not change the characteristic linear
increase of M as a function of conversion. It should also be
n
noted that the PDI remained almost completely constant.
Therefore, we can conclude that this method of polymeriza-
tion does not affect the ‘‘living’’ controlled radical polymer-
ization (CRP) while providing finer control over MW.
MW and End Group Control Influence on LCST
In a separate experiment, the racemo diad promoting agent
The measured cloud point temperature (T ), indicative of
cp
3
Me3PenOH was included in the polymerization process to
the LCST, is taken in this article to be the temperature at
which normalized transmittance drops to 50%. As expected,
confirm continued CRP behavior when synthesizing a major-
980
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 976–985

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
TABLE 2 pNIPAAm Polymerization Results in the Presence of 3Me3PenOH
Polymerization
Time at Room
Feed Molar Ratio
a
b
Temperature (h)
(NIPAAm:1:AIBN)
M
n
(NMR)
M
n
(GPC)
PDI
Conversion (%)
pNIPAAm-1s-a
pNIPAAm-1s-b
pNIPAAm-1s-c
pNIPAAm-1s-d
pNIPAAm-1s-e
pNIPAAm-1s-f
pNIPAAm-1s-g
0
100:1:0.5
100:1:0.5
100:1:0.5
100:1:0.5
100:1:0.5
100:1:0.5
100:1:0.5
6,269
6,388
6,695
6,736
6,835
7,324
7,760
6,381
6,728
6,919
6,777
6,888
7,571
7,957
1.17
1.16
1.15
1.16
1.16
1.15
1.15
54.1
55.2
57.8
58.2
59.0
63.2
67.0
ꢀ
2
5
23
47
71
95
0
.9 M NIPAAm was reacted for varying periods of time in a 2:1 ratio of 3Me3PenOH to NIPAAm. Thermal initiation was conducted at 65 C for 1.5 h.
Theoretical molecular weight calculated by multiplying conversion and theoretical maximum molecular weight based on feed ratios.
a
b
PDI ¼ M
w n
(GPC)/M (GPC).
pNIPAAm polymerized through free radical polymerization
shows a sharp Tcp at 32 C in deionized water [Fig. 5(A)].
polymer. First, these acid groups act in a similar manner to
hydrophilic comonomers, especially at low MWs. This is by
ꢀ
14
This is independent of the polymerization conditions and
can be seen in long-term low-temperature polymerization as
well as short-term high-temperature polymerization. This
far the more noticeable of the two effects. Second, these end
groups can act as syndiotacticity-inducing agents, hydrogen
bonding to the acrylamide group in much the same way that
bulky alcohols are reported to do.
ꢀ
temperature is shifted to 28 C when measured in PBS [Fig.
5
(A)], the commonly accepted ion concentration and pH for
To further analyze the first effect, especially as it related to
the MW of the polymer, we compared our sample of pNI-
PAAm-1-4d to pNIPAAm-co-AAc copolymers with a target
degree of polymerization of 50 and an AAc content of 4%.
Such polymers should theoretically have approximately two
AAc groups per polymer, similar to the two acid groups from
the RAFT polymerization, and comparable MWs. Because of
the difficulty in achieving a specific target MW using free
radical polymerization, two different samples with number-
averaged MWs near that of pNIPAAm-1-4d are presented.
The Tcp of the polymers is shown in Figure 6.
physiological systems. The decrease in temperature is due to
the destabilizing effects of the salt ions in aqueous solu-
tion.
known to decrease the LCST and even split it into two or
more transitions depending on concentration.
physiological concentrations of NaCl and NaH PO are insuf-
ficient to induce the splitting of the LCST, the reduction in
LCST is still significant.
2
6
Both sodium chloride and sodium phosphate are
2
6
Although
2
4
With the addition of 1 as the RAFT agent, the Tcp is signifi-
cantly increased [Fig. 5(B)]. The acetic acid end groups
(
from the RAFT agent) have two significant effects on the
Compared with the copolymers of comparable MWs, pNI-
PAAm-1-4d shows a higher Tcp than both the 4900 MW co-
polymer and the 6100 MW copolymer, with the higher MW
copolymer exhibiting a closer Tcp curve to pNIPAAm-1-4d
than the lower MW samples. It is interesting to note that
sensitivity of the Tcp to MW is not eliminated in the copoly-
mers and is contrary to the MW sensitivity of the RAFT poly-
mers. In addition, the characteristic widening of the Tcp
curve in these copolymers is small at these low MWs.
Because these copolymers were formed through free radical
polymerization and have high PDIs, it is likely that many
chains contain fewer than the anticipated number of hydro-
philic groups. Because of the low number of expected AAc
groups per polymer (approximately two), the variation can
lead to lower thermal transition temperatures and the still
somewhat sharp transitions exhibited.
When polymerized to higher MW (M ¼ 11,800, PDI ¼ 1.7)
n
to reduce this effect, as seen in Figure 6(B), the start of the
thermal transition is almost identical to that of pNIPAAm-1-
4d; however, the range of transition for pNIPAAm-1-4d was
FIGURE 4 GPC traces using ELS detector of pNIPAAm-1-4d
and pNIPAAm-1s-4d (M
¼ 3700 and 4100, respectively).
ꢀ
n
2.3 C, whereas the range for high MW pNIPAAm-co-AAc was
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 976–985
981

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
FIGURE 5 Changes in the Tcp of polymers assessed in deionized water (18 mX) and PBS. (A) pNIPAAm-r-4d and (B) pNIPAAm-1-
ꢀ
ꢀ
4
d. The Tcp changes from 32 to 28 C and 40.5 to 35.5 C for pNIPAAm-r-3d and pNIPAAm-1-4d, respectively.
ꢀ
greater than 10 C. Furthermore, pNIPAAm-co-AAc did not
comonomer. The manipulation required to synthesize copoly-
mers having thermal transition characteristics comparable to
our RAFT homopolymer, a more or less iterative process
involving multiple variables, confirms the superiority of our
method of LCST modification over the traditional copolymer-
ization method.
ꢀ
reach a stable transmittance until 26 above the start of the
transition. The Tcp for the two polymers is also quite differ-
ent: 44.9 and 40.5 C for pNIPAAm-co-AAc and pNIPAAm-1,
ꢀ
respectively. The similar starting temperatures for the transi-
tions suggest that comparable ratios of hydrophilic groups
were incorporated into both polymers, and the large differ-
ence in transition temperature ranges can again be attrib-
uted to the less well-defined nature of the copolymer. On the
other hand, the exclusion or inclusion of one or two AAc
groups per polymer chain will not change the AAc content
as drastically as in the case of the lower MW copolymers. In
an effort to reproduce both the MW and the thermal transi-
Tacticity Control Over LCST
A secondary effect of using 1 in the polymerization scheme
is the increase of racemo diads in the overall pNIPAAm poly-
mer. Although the number of acetic acid groups is limited to
the small amount of RAFT agent available during the poly-
merization and is inconsequential compared with the con-
centration of a solvent additive like 3Me3PenOH, the effect is
still pronounced, as seen in Figure 7(A,B).
tion of pNIPAAm-1-4d, an 8% AAc copolymer with an M of
000 was also synthesized [Fig. 6(B)]. As expected, this poly-
n
5
mer showed a much higher thermal transition than the com-
parable 4% AAc copolymers but also a broader transition,
indicative of more widespread incorporation of the AAc
As seen from the methine backbone peaks, there is a slight
increase in the percentage of racemo diads when using 1.
The racemo content increases from 54.6% in free radical
FIGURE 6 Normalized transmittance of 1 wt % aqueous solutions of pNIPAAm-1-4d and pNIPAAm-co-AAc. (A) pNIPAAm-co-AAc
with 4% AAc content and comparable molecular weights to pNIPAAm-1-4d (M of 4900 and 6100 with PDIs of 1.9 and 2.2, respec-
¼ 11,800, PDI ¼ 1.7) as well as lower MW 8% AAc polymer (M ¼ 5000,
n
tively) and (B) higher MW 4% AAc content polymer (M
PDI ¼ 1.9).
n
n
982
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 976–985

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
FIGURE 7 Methine backbone peaks of (A) free radical-polymerized pNIPAAm showing 54.6% racemo diads, (B) pNIPAAm-1-4d
showing 58.6% racemo diads, and (C) pNIPAAm-1s-4d showing 61.1% racemo diads. The peaks at 1.67 and 1.27 ppm correspond
to meso diads, whereas the peak at 1.46 ppm corresponds to racemo diads.
ꢀ
FIGURE 8 Normalized transmittance of 1 wt % aqueous solution of (A) pNIPAAm-1-4d and pNIPAAm-1s-4d (Tcp ¼ 40.5 C and
3.3 C respectively), (B) pNIPAAm-1-7d and pNIPAAm-1s-7d (Tcp ¼ 39.7 ꢀC and 42.4 C respectively), (C) pNIPAAm-1-HT polymer-
ꢀ
ꢀ
4
ꢀ
ꢀ
ꢀ
ized to completion at 65 C (Tcp ¼ 33.7 C and 35.0 C for pNIPAAm-1-HT and pNIPAAm-1s-HT respectively).
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 976–985
983

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
with an M of 3456 (PDI ¼ 1.13) for chains polymerized for
n
4
days. As expected, the higher MW slightly inhibits the
effect of the acetic acid end groups.
Because of the further increase in LCST by inducing a larger
percentage of racemo diads, and the relative stability of that
ꢀ
2
.7 C increase, we are, therefore, able to combine tacticity
control with the properties of 1 to produce a NIPAAm poly-
mer that undergoes a sharp thermal transition temperature
ꢀ
at 37.6 C, exactly within physiological temperature range, in
a solution of PBS as seen in Figure 9. This transition takes
ꢀ
place 2.1 above the transition temperature of the pNIPAAm
synthesized in the presence of 3Me3PenOH and occurs
ꢀ
within a span of 2 C. Such a polymer with its low PDI of
1.15 and lack of comonomers can greatly improve the sensi-
tivity to temperature that pNIPAAm copolymer systems for
biological applications currently lack.
FIGURE 9 Normalized transmittance of 1 wt % solution of pNI-
PAAm-1-4d polymerized with additional 3Me3PenOH in PBS.
CONCLUSIONS
Successful applications of pNIPAAm for biological purposes
have been limited in part due to the trade-off between hav-
ing an LCST that is high enough and having a very sharp
transition. We have shown in this article that this problem
can be overcome by implementing various polymerization
methods and tools. By polymerizing slowly with 1 over the
course of 4 days while inducing racemo diad formation, we
were able to synthesize well-defined pNIPAAm with a sharp
polymerization to 58.4% in polymerization with 1. When
polymerized with 3Me3PenOH, however, the racemo content
further increases to 61.1% [Fig. 7(C)]. Previous studies on
the relationship between diad tacticity and racemo diad pro-
moting agent properties have shown an inverse relationship
4
4
between pK and racemo diad content. Therefore, it is not
a
surprising that the RAFT agent has an outsized effect on the
tacticity despite its low concentration, because the acetic
acid groups on the end of the RAFT agent have a much
smaller pKa than 3Me3PenOH. These tacticity changes,
though small, have a significant effect on the observed LCSTs,
which is consistent with previous work in which pNIPAAm
having a higher percentage of racemo diads displays higher
ꢀ
LCST of 37.6 C in a solution of PBS. This reaction scheme
combines tacticity control with RAFT polymerization, MW
control, and end group control. Such polymers can be used
for more accurate transitions for drug delivery, diagnostics,
BioMEMs, and other applications in biotechnology.
1
9
Funding for this project was provided by the Georgia Institute
of Technology, a United States Department of Education Gradu-
ate Assistance in Areas of National Need (GAANN) fellowship
through the Georgia Institute of Technology Center for Drug
Design, Development, and Delivery (CD4), and Air Products
and Chemicals, Inc. The authors thank Mark Fisher with Varian
Inc. for UV–Vis Spectrometer programming assistance and
Leslie Gelbaum for assistance with NMR. They also thank Aswin
Natarajan, Lindsey A. Bergman, and Nathan C. Rubright for
assistance.
LCSTs.
Figure 8 shows the difference that a small change in tacticity
can make in the LCST of pNIPAAm. When polymerized for 4
days, we see a transition temperature increase from 40.5 to
ꢀ
4
3.3 C. When polymerized for 7 days at room temperature,
ꢀ
we see a transition temperature increase from 39.7 to 42.4 C.
When polymerized normally at 65 C for 48 h to ensure com-
pletion, pNIPAAm polymerized with 1 has an LCST at 33.7 C,
ꢀ
ꢀ
slightly higher than that of free radical-polymerized pNIPAAm,
but not high enough for applications in biotechnology. With
the inclusion of 3Me3PenOH to induce stereospecific polymer-
ꢀ
ization, the transition temperature is increased to 35 C.
REFERENCES AND NOTES
From this, we deduce that by polymerizing slowly for a long
period of time the bulky alcohol has more opportunity to
induce racemo diads. The LCST shift stays constant at ꢁ2.7
1
Vihola, H.; Laukkanen, A.; Valtola, L.; Tenhu, H.; Hirvonen, J.
Biomaterials 2005, 26, 3055–3064.
Boutris, C.; Chatzi, E. G.; Kiparissides, C. Polymer 1997, 38,
2567–2570.
2
ꢀ
C for both the 4-day and 7-day polymerizations, whereas it
ꢀ
decreases to a 1.3 C difference with faster, 48 h reactions at
3
Carter, S.; Rimmer, S.; Rutkaite, R.; Swanson, L.; Fairclough,
ꢀ
65
C. This is further confirmed using NMR, in which the
J. P. A.; Sturdy, A.; Webb, M. Biomacromolecules 2006, 7,
1
124–1130.
methine backbone of the higher reaction temperature pNI-
PAAm with 3Me3PenOH shows a lower racemo diad content
of 60% (see Supporting Information).
4
Hendrickson, G. R.; Smith, M. H.; South, A. B.; Lyon, L. A.
Adv. Funct. Mater. 2010, 20, 1697–1712.
Bikram, M.; Gobin, A. M.; Whitmire, R. E.; West, J. L. J. Con-
trolled Release 2007, 123, 219–227.
Su, S.; Ali, M.; Filipe, C. D. M.; Li, Y.; Pelton, R. Biomacromo-
lecules 2008, 9, 935–941.
5
The slight decrease in LCST between the 4-day and 7-day
polymerizations can be attributed to the larger size of the
6
7
-day polymer, having a M of 3997 (PDI ¼ 1.07) compared
n
984
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 976–985

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
7
1
Ankareddi, I.; Brazel, C. S. J. Appl. Polym. Sci. 2011, 120,
597–1696.
27 Chiefari, J.; Chong, Y. K.; Ercole, F.; Krstina, J.; Jeffery, J.;
Le, T. P. T.; Mayadunne, R. T. A.; Meijs, G. F. Macromolecules
1
998, 31, 5559–5562.
8
Ta, T.; Convertine, A. J.; Reyes, C. R.; Stayton, P. S.; Porter,
T. M. Biomacromolecules 2010, 11, 1915–1920.
Shen, Z.; Terao, K.; Maki, Y.; Dobashi, T.; Ma, G.; Yama-
moto, T. Colloid Polym. Sci. 2006, 284, 1001–1007.
0 Sanclimens, G.; Crespo, L.; Giralt, E.; Royo, M.; Albericio, F.
Biopolymers 2004, 76, 283–297.
Pei, Y.; Chen, J.; Yang, L.; Shi, L.; Tao, Q.; Hui, B.; Li, J. J.
Biomater. Sci. Polym. Ed. 2004, 15, 585–594.
Ito, S.; Ogawa, K.; Suzuki, H.; Wang, B.; Yoshida, R.; Koku-
futa, E. Langmuir 1999, 15, 4289–4294.
Ebara, M.; Yamato, M.; Nagai, S.; Aoyagi, T.; Kikuchi, A.;
Sakai, K.; Okano, T. Surf. Sci. 2004, 570, 134–141.
Li, Z.; Kim, Y.-H.; Min, H. S.; Han, C.-K.; Huh, K. M. Macro-
mol. Res. 2010, 18, 618–621.
Carter, S.; Rimmer, S.; Sturdy, A.; Webb, M. Macromol.
Biosci. 2005, 5, 373–378.
28 Perrier, S.; Takolpuckdee, P. J. Polym. Sci. Part A: Polym.
Chem. 2005, 43, 5347–5393.
9
29 Fukunda, T.; Goto, A.; Ohno, K. Macromol. Rapid Commun.
2
000, 21, 151–165.
30 Lai, J. T.; Filla, D.; Shea, R. Macromolecules 2002, 35,
754–6756.
1
6
11
31 Narain, R.; Gonzales, M.; Hoffman, A. S.; Stayton, P. S.;
Krishnan, K. M. Langmuir 2007, 23, 6299–6304.
1
2
32 Rieger, J.; Zhang, W.; Stoffelbach, F.; Charleux, B. Macro-
molecules 2010, 43, 6302–6310.
13
33 Thomas, D. B.; Convertine, A. J.; Myrick, L. J.; Scales, C.
W.; Smith, A. E.; Lowe, A. B.; Vasilieva, Y. A.; Ayres, N.; McCor-
mick, C. L. Macromolecules 2004, 37, 8941–8950.
14
3
4
Convertine, A. J.; Lokitz, B. S.; Vasileva, Y.; Myrick, L. J.;
15
Scales, C. W.; Lowe, A. B.; McCormick, C. L. Macromolecules
2
3
006, 39, 1724–1730.
Nuopponen, M.; Kalliomaki, K.; Laukkanen, A.; Hietala, S.;
16
Vogt, A. P.; Gondi, S. R.; Sumerlin, B. S. Aust. J. Chem.
007, 60, 396–399.
7 Ray, B.; Isobe, Y.; Matsumoto, K.; Habaue, S.; Okamoto, Y.;
5
2
Tenhu, H. J. Polym. Sci. Part A: Polym. Chem. 2008, 46, 38–46.
1
3
6
Isobe, Y.; Fujioka, D.; Habaue, S.; Okamoto, Y. J. Am.
Kamigaito, M.; Sawamoto, M. Macromolecules 2004, 37,
1702–1710.
Chem. Soc. 2001, 123, 7180–7181.
1
8
Hirano, T.; Ishizu, H.; Seno, M.; Sato, T. Polymer 2005, 46,
37 Ray, B.; Isobe, Y.; Morioka, K.; Habaue, S.; Okamoto, Y.;
Kamigaito, M.; Sawamoto, M. Macromolecules 2003, 36,
10607–10610.
5
43–545.
38 Hirano, T.; Miki, H.; Seno, M.; Sato, T. Polymer 2005, 46,
693–3699.
1
9 Hirano, T.; Okumura, Y.; Kitajima, H.; Seno, M.; Sato, T. J.
Polym. Sci. Part A: Polym. Chem. 2006, 44, 4450–4460.
3
20
Hirano, T.; Miki, H.; Seno, M.; Sato, T. Polymer 2005, 46,
5
501–5505.
39 Silva, M. E. S. R. e.; Dutra, E. R.; Mano, V.; Machado, J. C.
Polym. Degrad. Stab. 2000, 67, 491–495.
21
Katsumoto, Y.; Kubosaki, N. Macromolecules 2008, 41,
5
955–5956.
40 Odian, G. G. Principles of Polymerization, Fourth Edition;
Wiley: New Jersey, 2004; Chapter 3, pp 285–286.
22 Ray, B.; Okamoto, Y.; Kamigaito, M.; Sawamoto, M.; Seno,
K.-i.; Kanaoka, S.; Aoshima, S. Polym. J. 2005, 37, 234–237.
Koyama, M.; Hirano, T.; Ohno, K.; Katsumoto, Y. J. Phys.
Chem. B 2008, 112, 10854–10860.
41 Mita, I.; Horie, K. J. Macromol. Sci. Rev. Macromol. Chem.
Phys. 1987, C27, 91–169.
23
4
2
Friis, N.; Hamielec, A. E. In Emulsion Polymerization;
2
1
4
Ono, Y.; Shikata, T. J. Am. Chem. Soc. 2006, 128,
0030–10031.
Winnik, F. M.; Ringsdorf, H.; Venzmer, J. Macromolecules
Piirma, I., Garden J.L., Eds.; ACS Symp. Ser. 24; American
Chemical Society: Washington DC, 1976, Chapter 5, pp 82–91.
25
43 Sundberg, D. C.; James, D. R. J. Polym. Sci. Polym. Chem.
Ed. 1978, 16, 523–529.
1990, 23, 2415–2416.
2
6 Zhang, Y.; Furyk, S.; Bergbreiter, D. E.; Cremer, P. S. J. Am.
44 Yamada, K.; Nakano, T.; Okamoto, Y. Macromolecules
1998, 31, 7598–7605.
Chem. Soc. 2005, 127, 14505–14510.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 976–985
985