RAFT polymerization of N-vinylcaprolactam and effects of the end group on the thermal response of poly(N-vinylcaprolactam)

Article abstract of DOI:10.1016/j.reactfunctpolym.2012.04.002

The reversible addition-fragmentation chain transfer (RAFT) radical polymerization of N-vinylcaprolactam (NVCL) was performed using either S-benzyl-S-(benzyl propionate) trithiocarbonate (CTA 1) or N, N-diethyl-S-(α,α′- dimethyl-α″-acetic acid) dithiocarbamate (CTA 2) as a chain transfer agent (CTA). The polymerizations were controlled processes that yielded polymers with high conversion (>60%), controlled molecular weights that were close to the theoretical values and a narrow molecular weight distribution (minimal value: 1.13). The cloud point temperatures of poly(N-vinylcaprolactam) (PNVCL) were measured by turbidimetry and shifted to lower temperature and concentrations as the hydrophobicity of the end groups and the molecular weights of the polymers increased.

Full text of DOI:10.1016/j.reactfunctpolym.2012.04.002

Reactive & Functional Polymers 72 (2012) 407–413
Contents lists available at SciVerse ScienceDirect
Reactive & Functional Polymers
journal homepage: www.elsevier.com/locate/react
RAFT polymerization of N-vinylcaprolactam and effects of the end group
on the thermal response of poly(N-vinylcaprolactam)
⇑
Lidong Shao, Minqi Hu, Lin Chen, Li Xu, Yunmei Bi
College of Chemistry and Chemical Engineering, Yunnan Normal University, Kunming 650092, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
The reversible addition–fragmentation chain transfer (RAFT) radical polymerization of N-vinylcaprolactam
(NVCL) was performed using either S-benzyl-S-(benzyl propionate) trithiocarbonate (CTA 1) or N,
Received 26 January 2012
Received in revised form 2 April 2012
Accepted 3 April 2012
N-diethyl-S-(a,
a0- dimethyl-a00-acetic acid) dithiocarbamate (CTA 2) as a chain transfer agent (CTA). The
polymerizations were controlled processes that yielded polymers with high conversion (>60%), controlled
molecular weights that were close to the theoretical values and a narrow molecular weight distribution
(minimal value: 1.13). The cloud point temperatures of poly(N-vinylcaprolactam) (PNVCL) were measured
by turbidimetry and shifted to lower temperature and concentrations as the hydrophobicity of the end
groups and the molecular weights of the polymers increased.
Available online 11 April 2012
Keywords:
RAFT polymerization
Poly(N-vinylcaprolactam)
End groups
Ó 2012 Elsevier Ltd. All rights reserved.
Thermoresponsive polymer
1. Introduction
the cytotoxicity of PNVCLs having different molecular weights
has shown that the lower molecular weight polymers were
During the past decade, controlled radical polymerization
(CRP) techniques have been widely used for the preparation of var-
ious polymers that have controlled molecular weight, narrow
polydispersity, and well-defined architecture. The most useful CRP
techniques include nitroxide-mediated polymerization (NMP) [1],
atom transfer radical polymerization (ATRP) [2], reversible addi-
tion-fragmentation chain transfer polymerization (RAFT) [3–7],
and macromolecular design via the interchange of xanthates
(MADIX) [8]. CRP techniques require less strict polymerization con-
ditions than living anionic and cationic polymerization. Therefore,
these techniques have been used to synthesize a wide range of con-
jugated monomers. A variety of thermosensitive water-soluble
polymers have been synthesized using CRP [9–14].
Poly(N-vinylcaprolactam) (PNVCL) is one of several well-known
synthetic thermoresponsive polymers, and it exhibits a dissolu-
tion/precipitation transition in aqueous solution at temperatures
close to physiological temperatures. PNVCL is non-toxic [15,16]
and biocompatible [17,18], and the hydrolysis of the amide group
does not produce small amine compounds. These properties make
PNVCL an important polymer for use in biomedical applications
[19,20]. It has been reported that the lower critical solution
temperature (LCST) of PNVCL and its copolymers was sensitive to
changes in their molecular weights [21–23]. Research examining
well tolerated [15]. Therefore, the ability to control the molecule
weights of these polymers is desirable. Additionally, controlling
the polymerization will contribute to the development of more
sophisticated PNVCL-based materials. However, N-vinylcaprolac-
tam (NVCL) is representative of the unconjugated monomers that
can only be polymerized into high molecular weight polymers
through a radical mechanism similar to that of N-vinylpyrrolidone
(NVP) because the vinyl
P-electron is not conjugated with the C@O
functional group. It has been known that the CRP of unconjugated
monomers is not an easy task because of the highly reactive radical
species that are derived from the monomers and because of the
lack of suitable reagents or catalysts that can induce fast intercon-
version between the highly reactive radical and the dormant
species. The CRP of NVCL has only recently been reported. Devasia
et al. [24] reported the first RAFT polymerization of NVCL using
methyl 2-(ethoxycarbonothioylthio)propanoate as a chain transfer
agent (CTA), but no details were included. Later, Wan et al. [25]
described the RAFT polymerization of NVCL, which was mediated
by 2-diphenylthiocarbamoylsulfanyl-2-methylpropionic acid, ((O-
ethylxanthyl)methyl) benzene, and (1-(O-ethylxanthyl) ethyl)
benzene. However, the polymerization of NVCL did not proceed
in a strictly living manner. Tebaldi et al. [26] reported the synthesis
of a triblock copolymer that consisted of poly(t-butylacrylate) and
poly(N-vinylcaprolactam) by sequential RAFT polymerization, but
the CTA (dibenzyltrithiocarbonate) used during the synthesis
was not the most suitable for polymerizing NVCL. Negru et al.
⇑
Corresponding author. Tel./fax: +86 871 5516061.
E-mail address: yunmeibi@hotmail.com (Y. Bi).
1381-5148/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.reactfunctpolym.2012.04.002

408
L. Shao et al. / Reactive & Functional Polymers 72 (2012) 407–413
o
2
o
Scheme 1. Synthesis of CTA 1.
[27] synthesized PNVCL-b-PEG-b-PNVCL triblock copolymers by
the ATRP of NVCL using poly(ethylene glycol) that contained a
chloride end group (PEG-Cl) as the initiator and CuCl/5, 7, 7, 12,
14, 14-hexamethyl-1, 4, 8, 11-tetraazacyclotetradecane (Me6Cy-
clam) as the catalyzed system. However, these reported polymer-
izations of NVCL did not proceed in a strictly living manner or at
least in a lower control degree of the process. Thus, the living
and controlled radical polymerization of NVCL has not yet been
completely established.
The use of a CTA is crucial for successful RAFT polymerization. It
has been reported that the use of N, N-dialkyl dithiocarbamate was
effective for controlling the polymerization of unconjugated
monomers [28–31]. Additionally, trithiocarbonate belongs to a
family of CTAs (Scheme 2), which contain an S atom connected
to a Z group. Compared to other CTAs (dithiobenzoates, dithiocar-
bamates and xanthates), trithiocarbonates can be employed for the
RAFT polymerization of more monomers because of their moderate
Ctr, which results in less retardation of the polymerization and im-
parts hydrolytic stability [32]. Moreover, a CTA that has a large Z
group can cause retardation of the polymerization process,
whereas a CTA with a small or flexible Z group may be helpful
for controlling the polymerization. Based on these observations,
we have designed S-benzyl-S-(benzyl propionate) trithiocarbonate
(CTA 1), which is a new nonsymmetrical trithiocarbonate contain-
2.2. Characterizations
1H and 13C NMR spectra were obtained using a Bruker DRX-500
spectrometer with either CDCl3 or D2O as the solvent. The chemical
shifts were calculated relative to tetramethylsilane. Infrared spec-
tra were collected using potassium bromide pellets on a Nicolet
AVATAR 360 spectrometer. ESI–MS analyses were performed using
a Finnigan LCQ-Advantage Mass Spectrometer. The molecular
weight and the polydispersity of the polymers were determined
using gel permeation chromatography with multi-angle laser light
scattering (GPC/MALLS). The GPC/MALLS system consisted of a
Waters 2690D separations module, a Waters 2414 refractive index
detector (RI), and a Wyatt DAWN EOS MALLS detector. Styragel
HR1 THF and HR2 THF columns (Waters) were used at 40 °C with
polystyrene as standards and DMF as a mobile phase at a flow rate
of 0.3 ml/min.
2.3. Synthesis of S-benzyl-S-(benzyl propionate) trithiocarbonate (CTA 1)
The synthesis of S-benzyl-S-(benzyl propionate) trithiocarbonate
(CTA 1) was described in Scheme 1. 3-mercaptopropionic acid (4 ml,
46 mmol) was added to a solution of KOH (5.2 g, 93 mmol) in water
(50 ml) and cooled in an ice-water bath. Carbon disulfide (6 ml,
100 mmol) was added to the solution dropwise, and then the
resulting orange solution was stirred for 5 h. BnCl (5.5 ml, 46 mmol)
was added to the mixture, which was then heated for 16 h at 80 °C.
The mixture was cooled and CHCl3 (60 ml) was added, the reaction
mixture was acidified with hydrochloric acid until the organic layer
became yellow. The water phase was extracted using CHCl3 (2 ꢀ
60 ml). The combined organic layers were washed with 10% Na2CO3
(aq) (2 ꢀ 100 ml) and then dried over anhydrous MgSO4. After the
solvent had evaporated, the remaining product was purified using
flash gel column chromatography withpetroleum ether as an eluent,
which yielded a yellow oil (14.5 g, 87%). 1H NMR (500 MHz, CDCl3, d,
ppm): 7.28–7.41 (m, 10H), 5.17 (s, 2H), 4.63 (s, 2H), 3.68 (t, J = 7 Hz,
2H), 2.84 (t, J = 7 Hz, 2H). 13C NMR (125 MHz, CDCl3, d, ppm): 223.32,
171.62, 136.01, 135.29, 129.66, 129.13 ꢀ 2, 129.02 ꢀ 2, 128.78 ꢀ 2,
128.69 ꢀ 2, 128.22, 67.16, 41.93, 33.61, 31.76. IR (cmꢁ1): 3029
(ArAH), 2919 (CH2CH2), 1733 (C@O), 1499, 1453 (Ar), 1064 (C@S),
840 (CAS). ESI–MS(m/z): calculated for C18H18O2S3, 362.53; found,
385.37 [M + Na]+.
ing a flexible Z group of SCH2CH2COOCH2Ph and N, N-diethyl-S-(a,
a0-dimethyl-a00-actetic acid) dithiocarbamate (CTA 2), which be-
longs to a low Ctr family of dithiocarbamates and has a smaller Z
group of N(Et)2, to be used as CTAs for mediating the RAFT poly-
merization of NVCL (Scheme 1). Kinetic studies were used to com-
pare the effectiveness of CTA 1 and CTA 2. The phase separation
behavior of thermoresponsive polymer solutions has been reported
to be dependent on the physical properties and structure of the
polymer, such as the concentration, molecular weight (MW) and
end groups [33–38]. It has been demonstrated that the end groups
have a significant effect on the cloud point of low-MW poly(N-iso-
propylacrylamide) (PIPAAm) [39]. However, the effect of the end
groups on the phase separation behavior of PNVCL is unknown, lar-
gely due to the lack of control over both the MW and the end group
chemistry for PNVCL. In this study, the effect of the end groups,
MW and concentration on the cloud point temperatures (TCP) of
PNVCL were also investigated.
2. Experimental
2.1. Materials
N-vinylcaprolactam (98%, Sigma Aldrich, St. Louis, Missouri,
USA) was distilled under reduced pressure to remove the inhibitor
o
and then stored at 4 °C.
a, a-azobisisobutyronitrile (AIBN) (98%,
Shanghai Chemical Reagents Co., Shanghai, China) was purified
by recrystallizing from methanol two times. 3-mercaptopropionic
acid (99%, Alfa Aesar, Ward Hill, Massachusetts, USA), benzyl
chloride (BnCl) (98%) and carbon disulfide (99%, The Shanghai
Chemical Reagents Co., Shanghai, China) were used as received.
2
2
2
2
N, N-diethyl-S-(a,
a0-dimethyl-a00-actetic acid)dithiocarbamate
2
3 2
(CTA 2) was prepared according to the method described in the
literature [40].
Scheme 2. RAFT polymerization of NVCL mediated by CTAs.

L. Shao et al. / Reactive & Functional Polymers 72 (2012) 407–413
409
Table 1
RAFT polymerization of NVCL in presence of CTA 1 and CTA 2.
d
Polymers
CTA
Time (h)
Conv. (%)
Yield (%)
Mnth (g/mol)
Mn (g/mol)
PDId
PNVCL1-1a
PNVCL1-2a
PNVCL1-3a
PNVCL1-4a
PNVCL1-5a
PNVCL2-1b
PNVCL2-2b
PNVCL2-3c
PNVCL2-4c
PNVCL2-5c
PNVCL
CTA1
CTA1
CTA1
CTA1
CTA1
CTA 2
CTA 2
CTA 2
CTA 2
CTA 2
None
8
16
24
40
60
12
42
6
18.0
24.1
45.3
52.0
64.4
43.9
51.3
21.3
33.0
47.1
51.8
16.1
21.3
44.2
49.7
63.9
43.4
50.8
18.6
31.7
46.4
50.2
5373
7071
12946
14834
18290
3290
3805
3199
4828
6791
–
6800
7900
13600
15700
20600
3840
4030
3720
5710
7200
7357
1.29
1.13
1.42
1.28
1.33
1.21
1.32
1.28
1.16
1.15
2.75
21
25
4
a
b
c
The theoretic molecular weight was calculated by the formula: Mnth = 200 ꢀ 139.2 ꢀ conv. + molecular weight of CTA 1. which [monomer]:[CTA]:[AIBN] = 200:1:0.25.
The theoretic molecular weight was calculated by the formula: Mnth = 50 ꢀ 139.2 ꢀ conv. + molecular weight of CTA 2. which [monomer]:[CTA]:[AIBN] = 50:1:0.25.
The theoretic molecular weight was calculated by the formula: Mnth = 100 ꢀ 139.2 ꢀ conv. + molecular weight of CTA 2. which [monomer]:[CTA]:[AIBN] = 100:1:0.25.
Mn and PDI were determined by GPC/MALLS.
d
2.4. RAFT polymerization of NVCL
A
round-bottom flask was charged with CTA
1 (0.016 g,
0.044 mmol) or CTA (0.016 g, 0.0688 mmol), AIBN (3 mg,
2
0.018 mmol), NVCL (1.22 g, 8.8 mmol) or (0.971 g, 7.00 mmol);
then, the flask was degassed using three freeze-evacuate-thaw
cycles, followed by sealing under vacuum. The polymerization
was carried out at 70 °C for a predetermined time (Scheme 2). The
polymer was purified by precipitating from THF into petroleum
ether (b.p. 30–60 °C) three times to remove unreacted NVCL,
filtered, and then dried under vacuum at 40 °C for 24 h. The conver-
sion was determined using the gravimetric method. The results are
listed in Table 1.
2.5. Transmittance measurements
The PNVCL polymer (Table 1, PNVCL1-1, Mn = 6800 g/mol,
PNVCL1-2, Mn = 20,600 g/mol, PNVCL2-1, Mn = 3840 g/mol or
PNVCL2-5, Mn = 7200 g/mol) was dissolved in deionized water at
different concentrations (0.01 mg/ml-6.00 mg/ml). The solutions
were maintained at room temperature for 12 h to reach equilib-
rium. The optical transmittance at 500 nm of each polymer
solution was monitored as a function of temperature using a UV–
vis spectrometer (CARY UV-50, VARIAN) equipped with a water-
circulation heating stage. The heating rate was 1 °C/5 min. The
TCP was defined as the temperature that corresponded to the initial
break points in the resulting transmittance versus temperature
curve.
Fig. 1. IR spectra of CTA 1, CTA 2 and PNVCL prepared via RAFT polymerization
mediated by CTA 1 (a) and CTA 2 (b).
7.4 ppm), benzyl methylene protons (at 5.2 ppm), methylene
protons (at 3.7 ppm) and benzyl methylene protons (at 2.4 ppm)
appeared, indicating that the trithiocarbonate CTA 1 participated
in the polymerization and yielded the corresponding PNVCL that
contained a trithiocarbonate end group.
Similarly, we examined the bulk radical polymerization of NVCL
by using CTA 2 at two different concentrations ([NVCL]0:[CTA
2]0:[AIBN]0 = 100:1:0.25 or [NVCL]0:[CTA 2]0:[AIBN]0 = 50:1:0.25).
The presence of the R-group of the trithiocarbonate moiety at the
ends of the polymer chain (Scheme 2) could also be observed in
the IR and 1H NMR spectra of PNVCL. Fig. 2b shows a 1H NMR spec-
trum of PNVCL obtained from the CTA 2 system, from which chem-
ical shifts at 4.40 ppm (NCH), 3.22 ppm (NCH2), 2.33–2.50 ppm
(COCH2), 1.45–1.75 ppm (NCCH2CH2CH2 of caprolactam ring and
CH2 of backbone) due to the PNVCL chain were observed. Specifi-
cally, the quartet signal at 3.71 ppm (CH2NCH2) and the triplet sig-
nal at 1.22 ppm (CH3) were assigned to the CTA 2 moiety. In
addition, from the IR spectrum of PNVCL (Fig. 1b), the bands at
2925 cmꢁ1 (ACH2CH2CH2A of the caprolactam ring and ACH2A
of the backbone) and 1640 cmꢁ1 were assigned to the carbonyl
band of PNVCL. The characteristic bands of the CTA 2 moiety
(C@S, 1084 cmꢁ1; CAS, 842 cmꢁ1) appeared, indicating that the
synthesized PNVCL contained a dithiocarbamate end group.
The molecular weights (MWs) and polydispersities (PDIs) of the
obtained polymers were measured using GPC/MALLS and are
summarized in Table 1. Although PNVCL has been widely studied,
3. Results and discussion
3.1. RAFT polymerization of NVCL mediated by CTA 1 and CTA 2
The RAFT polymerization of NVCL was performed using CTA 1
and the AIBN initiator by maintaining the molar ratio [NVCL]0:[CTA
1]0:[AIBN]0 = 200:1:0.4 in the bulk solution at 70 °C. The presence of
the R-group in the trithiocarbonate moiety at the ends of the
polymer chain (Scheme 2) could be observed in the IR and 1H
NMR spectra of PNVCL (Fig. 1a). The strong, broad peak at
2926 cmꢁ1 corresponds to the absorption of the PNVCL backbone
and caprolactam ring because of the polymerization. The intense
absorption peak at 1634 cmꢁ1 was assigned to the C@O band of
PNVCL. The characteristic bands of the CTA 1 moiety (C@S,
1083 cmꢁ1; CAS, 842 cmꢁ1) appeared, which demonstrates that
the resulting PNVCL has a trithiocarbonate end group. The 1H
NMR spectrum of PNVCL is shown in Fig. 2a. Compared with the
1H NMR of CTA 1, the peaks assigned to the aromatic protons (at

410
L. Shao et al. / Reactive & Functional Polymers 72 (2012) 407–413
(a)
DDOO
22
g
h
O
S
d
f
a
c
e
O
S
S
n
b
O
N
i
j
m
k
l
PNVCL11--55
(b)
HHOO
2 2
a
b
b
j
j
S
d
c
N
COOH
S
n
a
O
N
f
g
e
h
i
PNVCL22--55
CDCl3
Fig. 2. 1H NMR spectra of PNVCL prepared via RAFT polymerization mediated by CTA 1 (a) and CTA 2 (b).
the reports on its GPC data are very limited [22–24]. In our study,
we examined the following eluents: THF, THF+n-BuNH4Br, H2O,
H2O + NaCl, and DMF. We found that in all of the solvents, the mea-
sured molecular weights were much lower (1 ꢀ 102 g/mol with
THF system) or much higher (1 ꢀ 106 g/mol with H2O system) than
the theoretical values, with the exception of DMF, which resulted
in a molecular weight that was closer to the theoretical values than
the other solvents. Therefore, DMF was selected as the GPC eluent
in our study. As shown in Table 1, in the presence of any of the
CTAs, polymers with relatively narrow polydispersities (with
PDI 6 1.42) and conversion that was dependent on MW were
obtained, whereas in the absence of CTA, the corresponding PDI
was wide. This observation indicated that the polymerization
was under control.
Kinetic studies of the bulk polymerization of NVCL were per-
formed at 70 °C in the presence of CTA 1 with the molar ratio
[NVCL]0:[CTA1]0:[AIBN]0 = 200:1:0.4 or CTA 2 with [NVCL]0:[CT
A2]0:[AIBN]0 = 100:1:0.25 or 50:1:0.25 to evaluate the effectiveness
of CTA 1 and CTA 2. For the polymerization with CTA 1, induction
periods of 1 h could be observed in the time-conversion plot
(Fig. 3a). The ln([M0]/[M]) increased with the reaction time in a
nearly linear fashion, suggesting that the radical concentration re-
mained essentially constant during the course of the polymerization
following the 1 h induction period. Additionally, the GPC/MALLS
analysis (Fig. 3a) showed that the MW increased nearly linearly
with conversion and that the corresponding PDI varied in a small
domain from 1.13 to 1.42 with increasing conversion. Fig. 4 shows
the corresponding GPC traces of the polymers from the above-men-
tioned kinetic study. It is easily observed that the GPC traces are
symmetric and unimodal and that no shoulder peak appears at a
high MW position, which would be attributed to linear–linear
polymer coupling. All of the results indicated that the polymeriza-
tion of NVCL mediated by CTA 1 has controlled polymerization
characteristics.
As shown in Fig. 5, the pseudo-first-order kinetics indicate a
constant concentration of radicals, and the experimental MW
values that are close to the theoretical values indicate that the
polymerizations mediated by CTA 2 also proceeded in a controlled
manner. Although a small shoulder or bump at low retention
volumes was present in most of the GPC traces from the polymers,
which has been attributed to bimolecular coupling in some studies
of RAFT polymerization [12], the PDIs of all polymers throughout
the polymerization are constant, in the range of 1.15–1.32.
Additionally, Fig. 5 showed that the PDI of the polymers prepared
using the ratio of [NVCL]0:[CTA2]0:[AIBN]0 = 100:1:0.25 were
lower than that of the polymers prepared using the ratio of
[NVCL]0:[CTA2]0:[AIBN]0 = 50:1:0.25, and the curve obtained from
the former ratio was closer to the origin. Such results imply that
the polymerization could be better controlled by changing the ratio
of [NVCL]0:[CTA2]0:[AIBN]0. These results indicate that CTA 1 and

L. Shao et al. / Reactive & Functional Polymers 72 (2012) 407–413
411
3.0
2.8
2.6
2.4
2.2
2.0
1.8
1.6
1.4
1.2
1.0
70
60
50
40
30
20
10
0
3.0
Mn(th)
20000
(a)
(b)
Conv. (%)
Mn(GPC)
2.5
ln ([M0]/[M])
PDI
16000
2.0
12000
8000
4000
1.5
1.0
0.5
0.0
0
10
20
30
40
50
60
10
20
30
40
50
60
70
Time (h)
conversion (%)
Fig. 3. Trace of time versus monomer conversion and ln[M0]/[M] (where [M0] = concentration of the monomer at time t = 0 h and [M] = concentration of the monomer at the
corresponding time) (a); and dependence of MW and PDI on conversion for the RAFT polymerization of NVCL in the presence of CTA 1 (b).
vs temperature plots (cloud point curves) for the aqueous solutions
of PNVCL1-1 and PNVCL2-5 at different concentrations. PNVCL1-1
exhibited a sharper transition at a lower temperature above a poly-
mer concentration of 0.10 mg/ml than PNVCL2-5 did. This finding
is attributed to PNVCL1-1 having more hydrophobic end groups.
The hydrophobic end groups may associate with each other at
temperatures that are close to the transition point and change
the hydrophobic–hydrophilic balance, resulting in a lower cloud
point [11]. It is known that the phase transition temperature of
PNVCL is sensitive to the molecular weight of the polymer [21].
To examine how the molecular weight affected the cloud point of
NVCL polymers obtained from the CTA 1 and CTA 2 systems, the
phase transitions of PNVCL1, which had a different MW, and
PNVCL2-1 (Mn = 3840 g/mol) were also studied using TCP measure-
ments. Fig. 7 shows that the TCP of PNVCL1 solutions with identical
concentrations (1 mg/ml and 0.5 mg/ml) decreases with increasing
polymer MW. Additionally, the TCP of PNVCL2-1 and PNVCL2-5 had
an inverse dependence on their MW. This behavior is similar to
that observed for solutions of the PNVCL obtained using conven-
tional radical polymerization, where the TCP shifts to a lower
temperature and concentration as the molecular weight of the
polymer increases [21]. This phenomenon is attributed to the fact
that the solution behavior of the PNVCL/water system corresponds
Fig. 4. Gradual traces of GPC in the kinetic study of the bulk polymerization of
N-vinylcaprolactam mediated by CTA 1 with the molar ratio of [NVCL]0:[CTA1]0:
[AIBN]0 = 200:1:0.4.
CTA 2 are effective chain transfer agents for controlling the radical
polymerization of NVCL and that CTA 1 is more suitable than CTA
2, suggesting that the former shows higher chain transfer ability
than the latter.
3.2. Thermoresponsive properties
It has been reported that the end groups of polymer molecules
may affect the thermal phase transition [38,39]. The NVCL polymer
molecules from the CTA 1 system and the CTA 2 system have dif-
ferent end groups. To investigate the effects of the end groups on
the phase transition, the cloud point temperatures (TCP) of aqueous
solutions of PNVCL1-1 (Mn = 6800 g/mol, Table 1) and PNVCL2-5
(Mn = 7200 g/mol), which have similar MW, were determined from
turbidimetry measurements. Fig. 6a and b shows the transmittance
to a typical Flory–Huggins (Type I) de-mixing behavior with TCP
.
As shown in Fig. 6a and c, the water solubility of the PNVCLs
obtained from the CTA 1 system decreased with increasing MW.
For example, at
a polymer concentration of 6.0 mg/ml, the
transmittance of the PNVCL1-1 solution at room temperature is
approximately 80%, whereas the transmittance of the PNVCL1-5
solution is only 42%. From visual inspection, the PNVCL1-5
(b)
(a)
3.0
Mn(th)
Mn(GPC)
PDI
2.8
2.6
2.4
2.2
2.0
1.8
1.6
1.4
1.2
1.0
[NVCL]:[Z2n]:[AIBN]=100:1:0.25
[NVCL]:[Z2n]:[AIBN]=50:1:0.25
8000
6000
4000
2000
10
20
30
40
50
60
conversion (%)
Fig. 5. Dependence of MW and PDI on conversion for the RAFT polymerization of NVCL in the presence of CTA 2 (a), and GPC traces of PNVCL prepared via RAFT
polymerization mediated by CTA 2 (b).

412
L. Shao et al. / Reactive & Functional Polymers 72 (2012) 407–413
100
80
60
40
20
0
(b)
(a)
100
80
0.005mg/ml
0.01mg/ml
0.02mg/ml
0.10mg/ml
0.50mg/ml
1.00mg/ml
2.60mg/ml
4.00mg/ml
6.00mg/ml
0.05mg/ml
0.10mg/ml
60
0.20mg/ml
0.50mg/ml
1.00mg/ml
2.00mg/ml
4.00mg/ml
5.00mg/ml
6.00mg/ml
40
20
0
30
35
40
45
30
35
40
45
o
o
Temperature ( C)
Temperature ( C)
(c)
(d)
100
0.005mg/ml
100
0.01mg/ml
0.02mg/ml
0.10mg/ml
0.50mg/ml
1.00mg/ml
2.60mg/ml
4.00mg/ml
6.00mg/ml
80
80
60
40
0.05mg/ml
0.10mg/ml
0.20mg/ml
0.50mg/ml
1.00mg/ml
2.00mg/ml
4.00mg/ml
5.00mg/ml
6.00mg/ml
60
40
20
0
20
0
30
35
40
45
30
35
40
45
50
o
o
Temperature ( C)
Temperature ( C)
Fig. 6. Transmittance measurements as a function of temperature for different concentrations of PNVCL1-1 (a), PNVCL2-5 (b), PNVCL1-5 (c) and PNVCL2-1 (d).
(b) 100
(a) 100
PNVCL1-1
PNVCL1-2
PNVCL1-3
PNVCL1-4
PNVCL1-5
80
60
40
20
0
80
60
40
20
0
PNVCL1-1
PNVCL1-2
PNVCL1-3
PNVCL1-4
PNVCL1-5
30
35
40
45
30
35
40
o
45
o
Temperature ( C)
Temperature ( C)
Fig. 7. Transmittance measurements as a function of temperature for 1 mg/ml (a) and 0.5 mg/ml (b) aqueous solutions of PNVCL1-1, PNVCL1-2, PNVCL1-3, PNVCL1-4 and
PNVCL1-5.
solution underwent gelation above a polymer concentration of
2.6 mg/ml. The same phenomenon has been observed for telechelic
poly(N-isopropylacrylamides) (PNIPAMs) that have an n-octadecyl
group at each chain end [41,42]. In dilute aqueous solutions below
the LCST, the telechelic HM-PNIPAMs form flower-like associates
consisting of loops of hydrated polymer chains that have both
end groups entrapped in the micellar core. At higher concentra-
tions (c > 20 g/l), the HM-PNIPAMs chains form bridges between
the rosettes, which trigger significant enhancement of the viscos-
ity, and eventually, result in the formation of a gel phase [41]. In
our case, the PNVCL1 that has a hydrophobic benzyl group at each
chain end is similar to the telechelic PNIPAM. The gelling behavior
of PNVCL1-5 with the higher MW implies that there is hydropho-
bic association of the chain ends.
medium, which decreased with increasing solution concentration,
but the TCP was almost independent when the concentration in-
creased to nearly 4.0 mg/ml for PNVCL1-1 and PNVCL1-5 and to
5.0 mg/ml for PNVCL2-1 and PNVCL2-5. We can also observe from
Fig. 7 that the PNVCL1 exhibits a sharper decrease in transmittance
at 1.0 mg/ml than at 0.5 mg/ml. This finding is consistent with the
generally accepted LCST principle for dilute solutions, which states
that higher water content enhances the hydrogen-bonding interac-
tions between water and the polymer chain, which requires more
thermal energy to break the water structure, thereby resulting in
an increase of the LCST [36].
4. Conclusion
The concentration-dependence in Fig. 6 showed that the solution
concentration also affected the phase transition temperature of the
PNVCL obtained from the CTA 1 and CTA 2 systems in aqueous
The radical polymerization of NVCL using AIBN as an initiator
and one of two new compounds, S-(benzyl propionate) trithiocar-
bonate (CTA 1) or N, N-diethyl-S- (a,
a0-dimethyl-a00-actetic acid)

L. Shao et al. / Reactive & Functional Polymers 72 (2012) 407–413
413
[11] F. Hua, X. Jiang, D. Li, B. Zhao, J. Polym. Sci., Part A: Polym. Chem. 44 (2006)
2454–2467.
[12] Y. Li, S.B. Lokitz, L.C. McCormick, Macromolecules 39 (2006) 81–89.
[13] X.J. Lu, L.F. Zhang, L.Z. Meng, Y.H. Liu, Polym. Bull. 59 (2007) 195–206.
[14] L.S. Wan, H. Lei, Y. Ding, L. Fu, J. Li, Z.K. Xu, J. Polym. Sci., Part A: Polym. Chem.
47 (2009) 92–102.
[15] H. Vihola, A. Laukkanen, L. Valtola, H. Tenhu, J. Hirvonen, Biomaterials 26
(2005) 3055–3064.
[16] I. Dimitrov, B. Trzebicka, A.E. Müller, A. Dworak, C.B. Tsvetanov, Prog. Polym.
Sci. 32 (2007) 1275–1343.
[17] A. Pich, A. Tessier, V. Boyko, Macromolecules 39 (2006) 7701–7707.
[18] N. Hntzschel, F.B. Zhang, F. Eckert, Langmuir 23 (2007) 10793–10800.
[19] E.A. Markvicheva, N.E. Tkachuk, S.V. Kuptsova, T.N. Dugina, S.M. Strukova, Y.E.
Kirsh, V.P. Zubov, L.D. Rumsh, Appl. Biochem. Biotechnol. 61 (1996) 75–84.
[20] H. Vihola, A. Laukkanen, J. Hirvonen, H. Tenhu, Eur. J. Pharm. Sci. 16 (2002) 69–
74.
[21] F. Meeussen, E. Nies, H. Berghmans, Polymer (2000) 8597–8602.
[22] N.I. Shtanko, W. Lequieu, E.J. Goethals, F.E. DuPrez, Polym. Int. 52 (2003) 1605–
1610.
[23] F.S. Medeiros, C.S.J. Barboza, I.M. Ré, R. Giudici, M.A. Santos, J. Appl. Polym. Sci.
118 (2010) 229–240.
[24] R. Devasia, R. Borsali, S. Lecommandoux, R.L. Bindu, N. Mougin, Y. Gnanou,
Polym. Prepr. (ACS, Div. Polym. Chem.) 46 (2005) 448–449.
[25] D.C. Wan, Q. Zhou, H.T. Pu, J. Polym. Sci., Part A: Polym. Chem. 46 (2008) 3756–
3765.
dithiocarbamate (CTA 2), as a chain transfer agent exhibited
controlled polymerization characteristics, which was indicated by
a well-controlled molecular weight, narrow molecular weight
distribution, and linear relationship between the molecular weight
and monomer conversion. CTA 1 was a good mediator for the RAFT
polymerization of NVCL, whereas the polymerization of NVCL that
was mediated by CTA 2 was not an ideal RAFT process. However,
the polymerization could be better controlled by changing the
ratio of [NVCL]0:[CTA2]0:[AIBN]0. With the use of the above men-
tioned trithiocarbonate and dithiocarbamate, new thiocarbonyl-
thio compounds were applied during the RAFT polymerization of
NVCL. The influence of the end groups on the TCP of PNVCL was
demonstrated for the first time. The TCP of PNVCL with the more
hydrophobic CTA 1 end groups was lower than that of the PNVCL
obtained from the CTA 2 system in the range of MW studied. Final-
ly, the transition temperatures decreased with increasing MW and
concentration for all of the samples studied.
Acknowledgements
[26] M.L.D.S. Tebaldi, T.C. Chaparro, A.M. Santos, Mater. Sci. Forum. 636–637 (2010)
76–81.
[27] I. Negru, M. Teodrescu, P.O. Stanescu, C. Draghici, A. Lungu, A. Sarbu, Mater.
Plast. 47 (2010) 35–41.
[28] M. Benaglia, J. Chiefari, Y.K. Chong, G. Moad, E. Rizzardo, S.H. Thang, J. Am.
Chem. Soc. 131 (2009) 6914–6915.
[29] E. Rizzardo, J. Chiefari, M. Benaglia, S.H. Thang, G. Moad, RAFT polymerization,
PCT Int WO 2010/083569 A1, 2010.
The authors gratefully acknowledge the support for this work
from the National Natural Foundation of China (No. 20864003),
Program for Changjiang Scholars and Innovative Research Team
in University, PCSIRT, and the Natural Science Foundation of
Yunnan Province (2009ZC064M), PR China.
[30] S. Perrier, P. Takolpuckdee, J. Polym. Sci., Part A: Polym. Chem. 43 (2005) 5347–
5393.
[31] A. Favier, M.-T. Charreyre, Macromol. Rapid. Commun. 27 (2006) 653–692.
[32] G. Moad, E. Rizzardo, S.H. Thang, Aust. J. Chem. 58 (2005) 379–410.
[33] S. Aoshima, H. Oda, E. Kobayashi, J. Polym. Sci., Part A: Polym. Chem. 30 (1992)
2407–2413.
[34] A. Saeed, D.M.R. Georget, A.G. Mayes, React. Funct. Polym. 72 (2012) 77–82.
[35] Z. Tong, F. Zeng, X. Zheng, T. Sato, Macromolecules 32 (1999) 4488–4490.
[36] S. Verbrugghe, K. Bernaerts, F.E. DuPrez, Macromol. Chem. Phys. 204 (2003)
1217–1225.
[37] J.M. Rathfon, G.N. Tew, Polymer 49 (2008) 1761–1769.
[38] S. Furyk, Y. Zhang, D. Ortiz-Acosta, P.S. Cremer, D.E. Bergbreiter, J. Polym. Sci.,
Part A: Polym. Chem. 44 (2006) 1492–1501.
[39] Y. Xia, N.A.D. Burke, H.D.H. Stöver, Macromolecules 39 (2006) 2275–2283.
[40] J.T. Lai, R. Shea, J. Polym. Sci., Part A: Polym. Chem. 44 (2006) 4298–4316.
[41] P. Kujawa, H. Watanabe, F. Tanaka, F.M. Winnik, Eur. Phys. J. E 17 (2005) 129–
137.
[42] P. Kujawa, F. Segui, S. Shaban, C. Diab, Y. Okada, F. Tanaka, F.M. Winnik,
Macromolecules 39 (2006) 341–348.
References
[1] M.K. Georges, R.P. Veregin, P.M. Kazmaier, Macromolecules 26 (1993) 2987–
2988.
[2] K. Matyjaszewski, J. Xia, Chem. Rev. 101 (2001) 2921–2990.
[3] T.P. Le, G. Moad, E. Rizzardo, Polymerizations with living characteristics, PCT
Int WO 98/01478, 1998.
[4] G. Moad, E. Rizzardo, H.S. Thang, Polymer 49 (2008) 1079–1131.
[5] D. Zhou, X.l. Zhu, J. Zhu, Z.P. Cheng, React. Funct. Polym. 69 (2009) 55–61.
[6] X.Q. Xue, J. Zhu, Z.B. Zhang, N.C. Zhou, X.L. Zhu, React. Funct. Polym. 70 (2010)
456–462.
[7] J.Y. Yang, K. Luo, H.Z. Pan, P. Kopecková, J. Kopecek, React. Funct. Polym. 71
(2011) 294–302.
[8] J. Chiefari, R.T.A. Mayadunne, G. Moad, Polymerizations with living
characteristics and polymers made therefrom, PCT Int WO 99/31144, 1999.
[9] J. Huang, K. Matyjaszewski, Macromolecules 38 (2005) 3577–3583.
[10] Y. Xia, X. Yin, N.A.D. Burke, H.D.H. Stöver, Macromolecules 38 (2005) 5937–
5943.
ˇ
ˇ