Macromolecular
Chemistry and Physics
Full Paper
Well-defined N-Isopropylacrylamide DualSensitive Copolymers with LCST ≈38 °C in
Different Architectures: Linear, Block and
Star Polymers
Lorenzo A. Picos-Corrales, Angel Licea-Claverie,* Jose M. Cornejo-Bravo,
Simona Schwarz, Karl-Friedrich Arndt
Reversible addition-fragmentation chain transfer (RAFT) polymerization is used to prepare
temperature- and pH-sensitive statistical copolymers with lower critical solution temperature (LCST) close to 38 °C at pH 7.4 based on N-isopropylacrylamide and methacrylic acid
derivative comonomers with a pKa close to 6. Statistical copolymers are re-activated to prepare amphiphilic block copolymers and star polymers with
cross-linked core. The LCST is maintained by varying the
architecture; however, the LCST originated behaviour changes
due to self-aggregation. Statistical copolymers and short
block copolymers show complex aggregation, whereas midsize block copolymers and star polymers show shrinkage of
aggregate dimensions. The pH of the medium has a profound
impact on the self-assembling behaviour of the different
polymer architectures.
1. Introduction
Responsive water-soluble polymers have been extensively
investigated for the development of “smart” macromolecules
that respond to external stimuli. Such stimuli-responsive
L. A. Picos-Corrales, Prof. A. Licea-Claverie
Instituto Tecnológico de Tijuana, Centro de Graduados e
Investigación, A.P. 1166, 22000 Tijuana, B.C., Mexico
E-mail: aliceac@tectijuana.mx
Prof. J. M. Cornejo-Bravo
Facultad de Ciencias Químicas e Ingeniería, Universidad
Autónoma de Baja California, Calzada Universidad No. 14418,
22300 Tijuana, B.C., Mexico
Dr. S. Schwarz
Department of Polylelectrolytes and Dispersions, Leibniz
Institute of Polymer Research Dresden, Hohe Strasse 6, 01069
Dresden, Germany
Prof. K.-F. Arndt
Physical Chemistry of Polymers, Dresden University of
Technology, 01062 Dresden, Germany
Macromol. Chem. Phys. 2012, 213, 301−314
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
polymers exhibit a transition in physical/chemical properties under a change in pH or temperature.[1–4] Therefore,
they have been proposed and partially tested as drugdelivery carriers to improve the efficacy and reduce the
toxicity of pharmacological therapies. Polymer-drug conjugates with prolonged circulation times target tumours by
the “enhanced permeation retention” (EPR) effect. Once in
the tumour interstitium, the polymer-drug can only enter
cells by the endocytic route. After endocytosis, the drug
could be released in early or secondary endosomes by pHcontrolled hydrolysis (pH drops from physiological 7.4 to
5 ≈ 6 in endosomes or to 4 ≈ 5 in lysosomes) or specifically
by enzymolysis in lysosomes.[5,6]
For drug delivery applications, the thermal responsiveness, for example, “lower critical solution temperature”
(LCST) of the polymers, needs to be tuned to human body
conditions.[7–9] Poly(N-isopropylacrylamide) (PNIPAAm)
has been the most studied thermoresponsive (also called
thermosensitive) polymer. In water, it undergoes a sharp
coil to globule transition turning the chain insoluble, followed by hydrophobic aggregation and phase separation
wileyonlinelibrary.com
DOI: 10.1002/macp.201100468
301
Macromolecular
Chemistry and Physics
L. A. Picos-Corrales et al.
www.mcp-journal.de
above 32 °C.[10,11] Calorimetric studies have shown that
the process of phase separation is preceded by disruption
of hydrogen bonding to water molecules.[12] The temperature range of the whole process may be narrow (discontinuous) or broad (continuous) arising from experimental
conditions (non-equilibrium) or by specific polymer
chain interactions with water. Furthermore, the process
is reversible, which means that precipitated polymer redissolves upon cooling. This is a distinctive aspect that
makes PNIPAAm different to natural thermosensitive
polymers such as proteins that usually do not “renaturate”
by cooling. And, finally, the phase transition temperature
may show a hysteresis in cyclic heating–cooling experiments.[13] If the PNIPAAm chains are chemically attached
(cross-linked) forming a network structure, the coil to
globule transition of the single chains leads to a cascadelike contraction of the macroscopic network. The LCST of
PNIPAAm can be reversibly adjusted by a change in the
conditions of the aqueous environment: presence of salts,
surfactants and so on.[14–16] However, for drug delivery
applications the main method used for LCST adjustment
is the non-reversible change of LCST by chemical modification of the polymer chain. The number of chemically
introduced units into the polymer chain can be one (end
functionalization or initiator functionalization),[17,18] or
more (copolymerization).[3,10,17,19−31] Aiming to increase
the LCST of PNIPAAm from 32 °C to above body temperature, copolymerization with hydrophilic monomers, containing carboxylic acids, amines or hydroxyls as substituents, were abundantly studied.[3,10,17,19–31] Some of those
reported acid monomers are good candidates for adjusting
the LCST to above body temperature at pH 7.4.[21,22,24,26]
In comparison, the use of tertiary amine containing
comonomers to increase the LCST of NIPAAm has being
less successful because at those pH values (7–8) amines
are usually not ionized.[17,27,31]
The effect of polymer architecture on LCST is a subject
that has not been targeted in deep in the literature. In
the case of PNIPAAm chains, there are reports that show
that a star polymer changes its LCST with increasing
number of arms compared to linear PNIPAAm.[32,33] This
change in LCST can be ascribed to the hydrophobic effect
of end groups in the star arms of relatively low molecular weight. In the case of grafted PNIPAAm chains onto
a hydrophilic or pH-sensitive polymer backbone, some
authors report a change of LCST with the number of
grafts,[31,34] whereas others do not find any change.[35,36]
There are also reports on grafting of polyoxazoline side
chains into a PNIPAAm backbone.[37] In this case, the
LCST increased with increasing content on polyoxazoline
grafts, but only when carboxylic acid groups were introduced by hydrolysis of methyloxazoline units. The finding
is confusing for block copolymers. From seven references
in which one block is PNIPAAm and the block copolymer
302
do not self-assemble below its LCST,[38–44] three reported
no change in LCST compared to pure PNIPAAm (31 or
32 °C),[38,39,44] two slight changes (33 or 34 °C)[40,41] and two
changes up to 44 °C.[42,43] In the case of block copolymers
that self-assemble below its LCST to micelles or vesicles,
all this materials showed an effect of the second or even
third block into their temperature sensitivity.[45–48] In the
case of self-assembled nanostructures and also in the case
of nano/microgels,[49–53] the LCST behaviour of PNIPAAm
do not result necessarily in aggregation and precipitation, in most cases a shrinkage of the nanostructure or a
change of a simple micelle into more complex aggregated
structures were observed. Occurrence of precipitation
depends on the properties of the other components of the
nanostructure and additional conditions such as pH and
the presence of salts. In the case of nano/microgels, they
change its “LCST” behaviour as a function of comonomer,
similarly to random linear copolymers.[49–53]
Starting from these literature reports, we try to make
some preliminary statements that would be the starting
point of our own studies:
1
2
3
Random copolymers, terpolymers and nano/microgels
show a change in LCST behaviour as a result of a hydrophilic (increase) or hydrophobic (decrease) comonomer. Usually, a low content on comonomer (up to
15 mol%) is enough to make the change.
Graft copolymers, diblock and triblock copolymers (not
assembling below LCST): some report a change, others report a minimal change and, finally, some report
no change in LCST. The rationale that we find is that
a change is observed only when the molar content of
PNIPAAm is below 50%. Increasing the PNIPAAm content results in diminishing the effect until it disappear
(between 50 and 65 mol% of PNIPAAm).
Self-assembled block copolymers (below LCST), form
nanostructures that change its size or type as a function of temperature. This temperature (we call it LCST
for simplicity) is shifted as effect of the second (third)
polymer chain. In case that the molar ratio of PNIPAAm
is well above 50 mol%, still a change in LCST can be
found. As for all block copolymers, the change in LCST
diminishes with increasing PNIPAAm content.
We must emphasize that, for these preliminary statements, different kind of chemistries in the copolymers
are compared, which may be an oversimplification.
In our investigation, the versatility of the RAFT polymerization technique[54,55] was exploited to prepare different topologies with the LCST close to the human body
temperature: linear (statistical) copolymers, linear block
copolymers and star-copolymers. Statistical copolymers
of NIPAAm with hydrophobic weak-acid comonomers
were prepared with the goal of effectively adjusting the
Macromol. Chem. Phys. 2012, 213, 301−314
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.MaterialsViews.com
Macromolecular
Chemistry and Physics
Well-defined N-Isopropylacrylamide . . .
www.mcp-journal.de
temperature of phase transition to 38 °C at pH 7.4. On
the other hand, as the weak-acid groups contained in
the comonomers ionize depending on pH, the resulting
LCST of the copolymers may be influenced by the pH
which opens the possibility of dual temperature and pHtriggered drug delivery. Our goal polymeric materials
are soluble at physiological conditions and capable of
responding by both a slight temperature increase or by a
pH decrease at the physiologically relevant pHs of 7.4 and
5.0, respectively. Furthermore, the question of the effect
of the copolymer architecture on LCST behaviour can be
answered more easily in our model system since only one
type of comonomer is used for all architectures. Finally,
the reversibility of the LCST behaviour in the different
architectures was also studied.
2. Experimental Section
2.1. Materials
The monomers, 2-methacryloyloxybenzoic acid (2MBA), 4-methacryloyloxybenzoic acid (4MBA) and 5-methacryloyloxypentanoic acid (5MPA) (Figure 1A) were synthesized according to
literature reports.[56,57] The RAFT agents, 4-cyanopentanoic acid
dithiobenzoate (CPADB) and 2-hydroxyethyl 2-phenylacetate
dithiobenzoate (HPDB) were synthesized following previously
reported procedures.[58,59] 4,4′-Azobis(4-cyanopentanol) (ACP) was
also prepared as described in literature.[60] All other chemicals
were purchased. N-Isopropylacrylamide (NIPAAm, 97%, Aldrich)
was purified by recrystallization from n-hexane. 4,4′-Azobis(4cyanopentanoic acid) (ACPA, 98%, Fluka) was recrystallized from
methanol. Styrene (St, 99%, Spectrum), hexyl acrylate (HA, 98%,
Aldrich) and divinylbenzene (DVB, 80%, Aldrich) were purified
by passing through an inhibitor remover column (Aldrich). pDioxane (ACS grade, Fermont), diethyl ether (ACS grade, Fermont)
and tetrahydrofuran (HPLC grade, Aldrich) were used as received.
2.2. Polymerizations
2.2.1. Preparation of PNIPAAm and Statistical Copolymers
Macro-CTAs were synthesized with prescribed molecular weights
via RAFT polymerization. PNIPAAm-S(C⫽S)Ph was prepared with
molar ratios monomer/CTA/initiator 283:1.0:0.2 and 50:1.0:0.2.
Random copolymers were synthesized with various molar per
cents of comonomers in order to optimize the LCST to 38–40 °C.
Poly(NIPAAm-co-4MBAX)-S(C⫽S)Ph (X = 2, 5%) was synthesized
using molar ratios of 283:1.0:0.2 and 50:1.0:0.2. Poly(NIPAAm-co2MBAY)-S(C=S)Ph (Y=5%, 7%, 10%, 15%) was prepared with molar
ratio 283:1.0:0.2. Poly(NIPAAm-co-5MPAZ)-S(C⫽S)Ph (Z = 5%, 7%)
was synthesized with molar ratios 283:1.0:0.4 and 50:1.0:0.4.
Monomer, CPADB and ACPA were dissolved in p-dioxane using
monomer concentration of 1.412 mol L−1 to
prepare macro-CTAs with Mn > 20 000 g mol−1.
Monomer, HPDB and ACP were dissolved
in p-dioxane using monomer concentration of 3.0 mol L−1 to prepare macro-CTAs
with Mn < 20 000 g mol−1. The reagents
were added to an ampoule, solutions were
degassed by three freeze–evacuate–thaw
cycles and, afterwards, the ampoules were
sealed under vacuum. Reactions were carried
out with stirring at 70 °C (CPADB) and 80 °C
(HPDB) for prescribed polymerization times.
The reactions were stopped by freezing,
ampoules were then opened. The p-dioxane
was removed by rotary evaporation and the
residual polymer was re-dissolved in acetone.
Polymer solutions were precipitated into
excess of cold diethyl ether for removing
monomers and non-polymeric residues. The
residual polymer was dissolved in acetone
and reprecipitated in excess of cold diethyl
ether. This procedure was repeated three
times. The polymer was finally dried under
vacuum at room temperature overnight.
2.2.2. Preparation of Block Copolymers
Figure 1. Chemical structures of (A) monomers and (B) synthetic strategy used.
www.MaterialsViews.com
Block copolymers were prepared using styrene and hexyl acrylate for the second block.
Molar ratio monomer/macro-CTA/initiator
22:1.0:0.2 and macro-CTA concentration of
20.0 mmol L−1 were used to prepare block
Macromol. Chem. Phys. 2012, 213, 301−314
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
303
Macromolecular
Chemistry and Physics
L. A. Picos-Corrales et al.
www.mcp-journal.de
copolymers of Mn < 10 000 g mol−1. Molar ratio 50:1.0:0.2 (HA),
100:1.0:0.2 (St) and macro-CTA concentration of 7.0 mmol L−1
were used to prepare Block copolymers of Mn > 15 000 g mol−1.
Macro-CTAs were dissolved in p-dioxane before adding monomer
and initiator. Solutions were added to an ampoule, then degassed
by three freeze–pump–thaw cycles, sealed under vacuum and
placed in a thermostated oil bath (70 °C) for 24 h. Polymer solutions were precipitated into excess of cold diethyl ether without
removing the p-dioxane. In this way, residual monomers and
also homopolymers were removed. The residual block copolymer
was re-dissolved in acetone and reprecipitated in excess of cold
diethyl ether. This procedure was repeated three times. The block
copolymer was finally dried under vacuum at room temperature
overnight.
2.2.3. Preparation of Star Polymers
Star polymers have been prepared by “arm-first” RAFT method
with cross-linking adapting a procedure described previously.[61,62] Different molar ratios of DVB:macro-CTA were studied
for determining optimum conditions for star preparation. For
preparing star polymers with PNIPAAm arms, PNIPAAm-S(C ⫽S)
Ph with Mn,GPC = 28 100 g mol−1 (molar ratios 10:1, 15:1, 20:1) and
Mn,GPC = 25 900 g mol−1 (molar ratios 25:1, 30:1, 35:1) were tested.
On the other hand, for preparing star polymers with copolymeric arms a molar ratio of DVB:macro-CTA = 35:1 was used with
poly(NIPAAm-co-5MPA5%) ( Mn,GPC = 26 700 g mol−1) and with
poly(NIPAAm-co-4MBA10%) ( Mn,GPC = 21 000 g mol−1). ACPA was
used as initiator and a molar ratio of macro-CTA:initiator was
5:1 in all cases. The reagents were dissolved in p-dioxane using
macro-CTA concentration of 4.8 mmol L−1. Solutions were added
to an ampoule and degassed by three freeze-evacuate-thaw
cycles. Ampoules were sealed under vacuum and immersed in an
oil bath at 70 °C for prescribed polymerization times (24, 36 or
48 h). The reactions were stopped by freezing, ampoules were
then opened. Polymer solutions were precipitated with diethyl
ether for removing DVB; the precipitate was collected and dried
under vacuum at room temperature overnight.
HPLC grade) for NIPAAm polymers and THF/CH3COOH (50:1)v
for random copolymers were used as eluent at a flow rate of
0.5 mL min−1 at room temperature. Monodisperse polystyrene
was used as calibration standard. GPC was equipped with a
refractive index detector (Varian RI-4) and a tri-angle light scattering detector (MINI-DAWN, Wyatt).
Dynamic light scattering (DLS) was used to determine the
hydrodynamic diameters (Dh). Measurements were carried out
at room temperature (25 °C) using a Zeta-sizer “nano-ZS” from
Malvern Instruments (ZEN3500) equipped with a green laser
operating at λ = 532 nm. The angle of measurement is 173o
(backscattering) and the size analysis was performed by CONTIN.
The reported hydrodynamic diameters (Dh) were calculated
using the Stokes–Einstein equation for spheres, as usual.[63] The
scattering intensity as a function of temperature was also used
for determining the LCST of the prepared polymeric materials.
The hysteresis of the LCST behaviour at different pH values was
also studied for selected polymers. For these studies a Zeta-sizer
“Nano-S” from Malvern Instruments (ZEN1600) equipped with a
red laser operating at λ = 632.8 nm and an angle of measurement
173o was used. The size analysis was performed at different
temperatures using 3 min equilibrating time at each temperature
in the heating cycle followed by 10 min equilibrating time
at each temperature in the cooling cycle. The reported Dh
values are the maxima in size distribution by intensity from
CONTIN analysis and where calculated as described above. The
shrinking temperature (Ttr) is reported as the temperature for
the definitive size drop, which corresponds within experimental
error with the LCST of the system. The polymer concentration
was 1 mg mL−1 for determination of LCST and for hysteresis
studies on statistical copolymers, whereas it was 0.4 mg mL−1 for
hysteresis studies on block copolymers and star polymers. The
solutions were prepared, shaken for 12 h and stored overnight in
refrigerator. Before measurement, they were filtered off using a
0.45 micron syringe filter for eliminating dust. 1H NMR spectra
were recorded on a Varian, Mercury-200 MHz nuclear magnetic
resonance instrument with CDCl3 or CD3OD as solvents and
tetramethylsilane (TMS) as internal reference.
2.2.4. Fractionation of Star Polymers and Free Arms
Separation of star polymers from linear polymers not attached to
the star polymer (free arms) was carried out by polymer fractionation. The used procedure was as following: into a beaker containing a magnetic stir bar, the precipitated star polymer product
(1 g) was dissolved in acetone (15 mL). Star polymer fraction was
precipitated by dropwise addition of cold diethyl ether (15 mL)
into the solution. The precipitate was collected and dried under
vacuum at room temperature overnight.
2.3. Characterization
Polymer yields were determined gravimetrically. The molecular
weights, polydispersity index (PDI = Mw/ Mn) and radius of gyration (Rg) were determined by gel permeation chromatography
(GPC) using a Varian 9002 chromatograph with a Phenogel 10
guard column (50 mm × 7.8 mm), a Phenogel 5 linear mixed bead
column (300 mm × 4.60 mm) and a Phenogel 10 linear (2) mixed
bed column (300 mm × 7.80 mm) in series. Tetrahydrofuran (THF,
304
3. Results and Discussion
3.1. Statistical Copolymers with LCST ≈38 °C
RAFT polymerization of NIPAAm is well documented.[54,55,61]
The kinetics of NIPAAm polymerization using 4-cyanopentanoic acid dithiobenzoate (CPADB) as RAFT agent
in p-dioxane as solvent was reported previously.[61]
By the addition of an acid comonomer with less than
15 mol% in the monomer mixture, we do not expect a strong
influence on the kinetics. CPADB is a type of RAFT agent
that also works very well for methacrylate-type monomers.[55] Table 1 summarizes experiments at different
polymerization times adding to NIPAAm each of the three
acid comonomers used in this investigation. As expected,
the molecular weight increases with polymerization time
while low polydispersities (PDI = Mw/Mn, also simply
Macromol. Chem. Phys. 2012, 213, 301−314
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.MaterialsViews.com
Macromolecular
Chemistry and Physics
Well-defined N-Isopropylacrylamide . . .
www.mcp-journal.de
Table 1. Preparation of random copolymers by RAFT with different acid comonomers.
Random copolymer
Poly(NIPAAm-co-5MPA5%)
Molar ratios
M/CTA/Ia)
Reaction time
[h]
Conv.b)
[wt%]
Mn(calc.)c)
[g mol−1]
Mn(GPC)
[g mol−1]
Mw/Mn
(PDI)
283:1.0:0.2
14
32
10 857
12 000
1.027
16
40
13 502
16 000
1.042
283:1.0:0.4
Poly(NIPAAm-co-4MBA5%)
Poly(NIPAAm-co-2MBA15%)
283:1.0:0.2
283:1.0:0.2
13
50
16 808
24 700
1.156
15
72
24 080
29 900
1.176
17
73
24 410
30 100
1.152
19
71
23 750
31 800
1.230
12
44
14 950
22 600
1.213
14
63
21 280
28 300
1.084
16
74
24 950
34 800
1.088
12
35
12 870
20 500
1.039
14
51
18 625
24 800
1.039
16
72
26 180
30 300
1.022
a)Polymerizations at 70 °C and using a monomer concentration of 1.412 mol L−1; b)Conversions determined gravimetrically; c)Calculated
molecular weights based on Equation (1)
called dispersity) are mantained. The measured molecular
weights by GPC correspond well with the calculated ones
using the ideal RAFT equation:
[M]0
Mn(calc) =
· MW,i · Rw + MW,CT A
fi ·
(1)
[CT A]0
where fi is mole fraction of monomer i, Mw is the molecular
weight of monomer i and Rw is the fraction of conversion.
The typical RAFT behaviour for copolymerization experiments with 4MBA and 2MBA was observed, showing a
linear increase of molecular weight with time maintaining
a low polydispersity (see Supporting Information). The
composition of the statistical copolymers prepared was
determined by 1H-NMR in deuterated methanol a good solvent for all type of monomers. In the spectra, the signals of
the acid comonomer can be fairly good seen and the composition was calculated by integration of selected protons
as described in the Supporting Information.
Table 2 shows a series of synthesized copolymers produced with the aim to get different compositions and
molecular weights. In some cases, the acid comonomer
content was higher as expected from the comonomer
Table 2. Characteristics of random copolymers with acid groups.
Random copolymer
Mn(GPC) Mw/Mn
[g mol−1] (PDI)
NIPAAm
content
[mol%]a)
LCST by DLS [°C]
pH = 7.8 7.4
Dhb) [nm]
6.0
5.0
4.4 pH = 7.8
7.4
6.0
4.4
Poly(NIPAAm-co-5MPA5%)
31 800
1.230
95.0
–
40
34
28
23
–
12.5
7.9
9.6
Poly(NIPAAm-co-5MPA5%)
6600
1.009
90.0
–
38
34
–
–
–
3.1
–
–
Poly(NIPAAm-co-5MPA7%)
30 700
1.249
93.3
40
–
–
–
28
17.6
–
–
19.7
Poly(NIPAAm-co-4MBA2%)
26 800
1.214
–
–
34
–
–
–
–
8.7
–
–
Poly(NIPAAm-co-4MBA5%)
35 400
1.224
88.2
–
38
32
–
–
–
8.6
7.8
–
–
–
Poly(NIPAAm-co-4MBA5%)
c)
17 200
1.012
90.0
– 40
32
–
c)
–
–
7.0
Poly(NIPAAm-co-4MBA5%)
5600
1.007
88.0
43 38
36
–
–
3.5
2.8
–
–
Poly(NIPAAm-co-2MBA5%)
38 700
1.538
–
–
32
–
–
–
–
–
–
–
Poly(NIPAAm-co-2MBA7%)
35 300
1.655
–
–
34
–
–
–
–
–
–
–
Poly(NIPAAm-co-2MBA10%)
45 400
1.293
–
–
36
–
–
–
–
10.7
–
–
Poly(NIPAAm-co-2MBA15%)
38 900
1.038
84.0
–
40
–
–
–
–
7.8
–
–
a)Content
was determined by NMR; b)Dh data for copolymers as determined by DLS at 26 °C; c)Data was determined at pH 7.0.
www.MaterialsViews.com
Macromol. Chem. Phys. 2012, 213, 301−314
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
305
Macromolecular
Chemistry and Physics
L. A. Picos-Corrales et al.
www.mcp-journal.de
Figure 2. LCST of random copolymers: (A) LCST of poly(NIPAAm95%-co-5MPA5%) Mn = 31 800 g mol−1 at different pH values; (B) LCST as a
function of 2MBA content in poly(NIPAAm-co-2MBA) at pH 7.4.
mixture. This is an indication that the incorporation
of the acid comonomer do not follow the rules for ideal
copolymerization; however, determination of copolymerization parameters was out of the scope of this investigation. The penultimate column in Table 2 summarizes also
the LCST values measured at different pH for the statistical copolymers prepared. For each type of copolymer, an
LCST close to 38 °C at pH 7.4 was obtained at certain acid
comonomer content. This LCST value was reproduced also
for short block copolymers ( Mnbelow 7000 g mol−1). Furthermore, the LCST value decreases with decreasing pH
value. This trend is clearly seen in Figure 2A. This means
that by adding an acid commoner to PNIPAAm the LCST
is pH dependent. This statement is not always valid and
deserves clarification. The selection of the acid comonomers
for this purpose was very important. An acid comonomer
with a pKa value higher than that of acrylic acid (AA)
or methacrylic acid (MA) was needed given the fact that
the LCST of copolymers of NIPAAm with AA or MA are to
high at pH 7–8.[21] We have chosen three acid comonomers
with pKa values between 5 and 6, namely, 5MPA with
pKa = 6.14,[57] 4MBA with pKa = 5.3[56] and 2MBA with
pKa = 5.8.[56] Furthermore, all three acid comonomers contain hydrophobic moieties as spacer groups (see Figure 1).
This results in the fact that they are water insoluble at pH
values lower than 5. These comonomers introduce into
the PNIPAAm chain both, hydrophobic interacting units
as well as ionizable acid units. At pH value of 7.4, more
than 90% of acid units are expected to be ionized; however, at pH 6 the degree of ionization decreases differently
for each acid comonomer. This can be calculated using
the Henderson–Hasselbach equation[64]
306
1 −α
(2)
α
where α is the degree of ionization of the acid units. Therefore, at pH 6, 42% of 5MPA units are ionized, while 83% of
4MBA units are ionized. The LCST change from 38 °C (pH
7.4) to 34 °C (pH 6) for the short block poly(NIPAAm-co5MPA), while it change from 38 °C (pH 7.4) to 36 °C (pH 6),
for the similar in size and composition poly(NIPAAm-co4MBA). This change is remarkable in view of a possible
biomedical application of these copolymers. It means that
these copolymers are expected to change from expanded
state to collapsed state in the endosome, if they go through
endocytosis into a cell. The change in ionization degree is
more dramatic at pH 5, only 6.8% of acid units in 5MPA are
ionized and the LCST drops below that of PNIPAAm. The
hydrophobic effect from the spacer chains is overwhelming
at this pH. For the comonomer 2MBA, which has a pKa
value in between the other ones discussed, we study the
effect of comonomer content on the LCST. In other words,
how much of this comonomer is needed to attain an LCST
value ≈38 °C at a pH 7.4. Figure 2B shows the results in a
graph. Between 10 and 15 mol% of 2MBA is needed to
attain an LCST of ≈38 °C at pH 7.4. For comparison, only
5 mol% of the 5MPA acid monomer is needed to attain
the desired LCST. It implicates that the acidity constant is
the most important factor affecting the LCST at given pH
value; however, other aspects such as hydrophobic interactions from other groups in the monomer are needed
to be taken into account to have a real prediction. In any
case, it has been shown that all three comonomers chosen
can shift the LCST of PNIPAAm at pH 7.4 to values close to
38 °C at a certain concentration in the statistical copolymer
pKa = pH + log
Macromol. Chem. Phys. 2012, 213, 301−314
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.MaterialsViews.com
Macromolecular
Chemistry and Physics
Well-defined N-Isopropylacrylamide . . .
www.mcp-journal.de
between 5 and 15 mol%. In addition, the LCST decreases to
lower values when the pH decreases. This can be expected
for acid comonomers, as their ionization degree falls and
their hydrophobic interaction contributions overwhelm
the hydrophilic ones.
This implies that these copolymers are soluble at physiological conditions and show a coil to globule transition
with precipitation either by increasing the temperature
or by a change to lower pH values. These materials are
therefore double-sensitive statistical copolymers.
3.2. Block Copolymers with LCST ≈38 °C
Temperature-sensitive amphiphilic PNIPAAm copolymers
were studied shortly after the LCST behaviour on PNIPAAm was discovered.[65] The main interest of introducing
a second hydrophobic block into a PNIPAAm chain is to be
able to attain nanosized aggregates in aqueous systems
capable of transporting in their interior biologically active
molecules.[66,67] In this investigation, we chose a frequently
used hydrophobic block: polystyrene, which is a glassy
polymer; and a non-reported one for this purposes: poly(nhexyl acrylate), which is rubbery. Both monomers have
been proved to successfully grow in RAFT polymerization
conditions to form block copolymers with PNIPAAm.[68,69]
It was our goal to get block copolymers with micelle
forming capacity in water at pH 7.4. By chain extending
the statistical copolymers described in the previous section with 5 to maximal 15 mol% of hydrophobic units,
we expect to get micelles as expected theoretically in the
literature for amphiphilic block copolymers.[70] Table 3
shows the resulting block copolymers with very low polydispersity as expected for a controlled radical copolymerization method, and with hydrophobic comonomer
Figure 3. LCST behaviour of “short” block copolymers at different
pH values: poly(NIPAAm83%-co-5MPA7%)-b-PSt10%, Mn = 8800 g mol−1
(, pH 7.4; , pH 6.0) and poly(NIPAAm78%-co-5MPA10%)-b-PHA12%,
Mn = 9100 g mol−1 (, pH 7.4; Δ, pH 6.0).
content within the chosen range. Short blocks ( Mn <
10 000 g mol−1) and mid size blocks (up to 35 000 g mol−1)
were prepared successfully. More important, LCST values
close to 38 °C were still achieved. However, the influence of pH on LCST is different and need to be discussed
more in detail. Figure 3 shows the LCST determination
by turbidity for two (short) block copolymers differing
only in the hydrophobic block. As can be seen, the LCST
values for both block copolymers are the same, and this
is also true at two pH values: 7.4 and 6.0. Nevertheless
the difference in LCST at different pH values is, compared
with the statistical copolymer used in its preparation,
Table 3. Characteristics of block copolymers.
Block copolymer Mn(macro-CTA)a) Mn(block)a) Mw/Mn
[g mol−1]
[g mol−1] [PDI]
PNIPAAm-b-PSt
Composition
Dh [nm]
1st/2nd block
pH = 7.4 7.0
6.0
[mol%]b)
– 38.8c)
LCST by DLS [°C]
4.4 pH = 7.4 7.0
6.0 4.4
–
–
– 30c)
–
–
–
–
–
44
–
30
–
19.3
–
–
–
44
–
30
–
(86-6.2)/7.8
119
–
–
87
37
–
–
24
1.044
(82-12)/6
–
94.0 27.6d)
–
– C.S.e) 36d)
–
1.042
(81-9)/10
–
42.6 34.0d)
–
– C.S.e) 38d)
–
18 000
22 100
1.024
84/16
poly(NIPAAm-co5MPA10%)-b-PSt
6500
8800
1.097
(83-7)/10
14.9
Poly(NIPAAm-co5MPA10%)-b-PHA
6500
9100
1.067
(78-10)/12
Poly(NIPAAm-co5MPA7%)-b-PHA
30 700
35 000
1.320
Poly(NIPAAm-co4MBA12%)-b-PSt
18 000
21 300
Poly(NIPAAm-co4MBA12%)-b-PHA
17 200
20 100
a)
Molecular weights determined by GPC; b)Composition determined by NMR; c)Data were determined in pure water; d)Data were determined at pH 5.8; e)Continuous shrinkage.
www.MaterialsViews.com
Macromol. Chem. Phys. 2012, 213, 301−314
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
307
Macromolecular
Chemistry and Physics
L. A. Picos-Corrales et al.
www.mcp-journal.de
bigger. Furthermore, the LCST for block copolymers at pH
7.4 is higher and at pH 6 lower, than for the statistical
copolymers. The difference can be caused by a different
aggregation behaviour. A statistical copolymer behaves
similarly as a linear PNIPAAm chain does, it shows a coil
to globule transition at LCST followed by self-aggregation
and precipitation. Amphiphilic block copolymers already
show aggregation below the LCST. This aggregation
behaviour depends on block copolymer composition and
on environmental parameters such as concentration, pH,
presence of salts and also temperature.
Table 4 shows the measured hydrodynamic radii in
water together with estimates of contour lengths (Lc) of
block copolymers and end to end distances of similarsize polymer coils in good solvent (h). On the one side, it
can be considered that an spherical micelle would have a
diameter around two times h (or Lc) of the block copolymer
chains (radius would be therefore close to h or Lc of the
block copolymer). On the other side, a vesicle (also named
polymersome) would be constructed with two double
layers including space in the core for water; resulting in
more than five times h or Lc of the block copolymer (radius
2.5 times h or Lc). However, because the extended chains
calculated from Lc are unrealistic for amphiphillic blockcopolymers in water, we took h as the best estimate. Therefore, we postulate the aggregates described in the last
column of Table 4. Both short blocks studied form micelles
below the LCST; when the temperature is increased above
the LCST a coil to globule transition of the PNIPAAm
copolymeric blocks would result in a change in the
self-aggregation behaviour of the micelles. If the micelles
are not very stable, they may disrupt to form bigger, more
stable aggregates such as worms and vesicles; and they
may finally precipitate; but if they are stable they can also
shrink in its size without forming bigger structures. In the
case of the short blocks, it is evident that their micelles are
not very stable because the turbidity increases rapidly and
they finally precipitate. However, the additional energy
required to disrupt the micelles has the effect that the
LCST is shifted to higher temperatures at pH value of 7.4
where the acid comonomer groups are almost completely
ionized (≈95%). As less than half of the acid units are ionized (42%) at pH 6, the number of hydrophobic units from
the non-ionized acid groups added to the hydrophobic
block contribution leading to an LCST below that of the
non-block copolymer. Figure 4 show the results of LCST
for mid-size block copolymers. In this case, the size as Dh
of block copolymers is plotted versus temperature to get a
picture of the size of aggregates formed. Below the LCST,
this block copolymers form micelles or slightly elongated
micelles (see Table 4). A block copolymer PNIPAAm-b-PSt
is used for comparison purposes. We can see that the size
for the block copolymer with no acid comonomer doubles
above the LCST. The block copolymers with acid 4MBA
comonomer also increase it size above the LCST at pH 5.8,
whereas at pH 7.4 and pH 7.0 they do not show a clear LCST
up to 50 °C; however, its aggregates shrink continuously by
heating (see Supporting Information). At pH 5.8, the LCST
was measured to be between 36 and 37 °C. By comparison
with statistical copolymers containing 4MBA units, we
Table 4. Characteristics of aggregates formed by block copolymers in water
Mn (block)a)
[g mol−1]
Contentb)
[mol% 1°/2° block]
Rhc)
[nm]
Lcd)
[nm]
he)
[nm]
Aggregate typef)
22 100
84/16
19.4g)
49.9
15.1
Micelleg)
poly(NIPAAm-co5MPA10%)-b-PSt
8800
(83-7)/10
7.45h)
18.9
7.1
Micelleh)
Poly(NIPAAm-co5MPA10%)-b-PHA
9100
(78-10)/12
9.65h)
17.4
7.1
Micelleh)
Poly(NIPAAm-co5MPA7%)-b-PHA
35 000
(86-6.2)/7.8
59.5h) 43.5i)
72.0
15.4
Vesicleh) Vesiclei)
Poly(NIPAAm-co4MBA12%)-b-PSt
21 300
(82-12)/6
47.0j) 13.8k)
43.8
15.1
Vesiclej) Micellek)
Poly(NIPAAm-co4MBA12%)-b-PHA
20 100
(81-9)/10
21.3j) 17.0k)
39.8
13.2
Micellej) Micellek)
Block copolymer
PNIPAAm-b-PSt
a)
Molecular weights data were determined by GPC, THF was used as eluent; b)Content was determined from NMR; c)Rh data were determined by DLS at 25 °C; d)Lc (contour length), equal to the product of overall degree of polymerization times the contribution of one
monomer repeat unit to the contour length, which is 0.254 nm[71]; e)h, the end to end distance of a coiled polymer chain in a good
solvent.[63] Rg = 1.78 Rh was used.[72] Comparative molecular weights PNIPAAm ( Mn = 37 000 g mol−1, Dh(THF) = 6.5 nm[62]; Mn = 26 720 g
mol−1, Dh(THF) = 6.4 nm[61]; Mn = 21 020 g mol−1, Dh(THF) = 5.6 nm[61]; Mn = 12 000 g mol−1, Dh(THF) = 3.0 nm)[62] was used; f) Micelle radius ≈ h,
ellipsoidal micelle between 1.5 and 2 times h and vesicle ≥3 times h; g)Data were determined in water; h)Data were determined at pH 7.4;
i)
Data were determined at pH 4.4; j)Data were determined at pH 7.0; k)Data were determined at pH 5.8.
308
Macromol. Chem. Phys. 2012, 213, 301−314
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.MaterialsViews.com
Macromolecular
Chemistry and Physics
Well-defined N-Isopropylacrylamide . . .
www.mcp-journal.de
In general terms, although the LCST value from the statistical copolymer is kept close to human body conditions
by amphiphilic block formation, we find specific effects
from the ionization degree of the acid comonomer (major)
in addition to the hydrophobic block contribution (minor),
resulting in slightly higher (high ionization) or lower (low
ionization) LCST values. Furthermore, because amphiphilic
block copolymers form aggregates in water, the LCST
behaviour is not equal to that of linear chains. Aggregate
type and its stability may shift the transition temperature
value (Ttr better describes what is happening than LCST).
3.3. Star Polymers with LCST ≈38 °C
Figure 4. LCST behaviour of mid-size block copolymers at different pH values: PNIPAAm84%-b-PSt16%, Mn = 25 000 g mol−1
in pure water (), poly(NIPAAm82%-co-4MBA12%)-b-PSt6%, Mn =
24 500 g mol−1 in buffer pH 5.8 () and poly(NIPAAm81%-co-4MBA9%)b-PHA10%, Mn = 19 200 g mol−1 in buffer pH 5.8 ().
find here an increased LCST at pH 5.8. For this acid comonomer, the ionization degree is 76% at pH 5.8, whereas
at pH 7.4 a 99% ionization is attained. At both pH values,
we have an overall hydrophilic effect of the comonomer.
The hydrophobic PSt block seems to favour aggregation at
slightly lower temperatures than the PHA block. The midsize block in Table 4 containing 5MPA as acid comonomer
shows at pH 7.4 LCST values below that expected from its
acid comonomer content, 7% of 5MPA should result in an
LCST higher than for 5% 5MPA (40 °C) but is not. This result
suggests that the aggregates formed (vesicles, Table 4) are
not stable at pH 7.4 and easily rearrange to other structures. The behaviour of this block copolymer is discussed
in detail in the last section of this paper.
The use of RAFT copolymerization for the preparation of
branched polymer structures relies on the competition
between chain growth by block copolymerization of a
divinyl monomer into the re-activated macro-CTAs (copolymeric NIPAAm arms) and the coupling of growing chains
containing a number of divinyl units into branched structures. If the amount of divinyl monomer related to that
of the macro-CTA is high, then an undesired macroscopic
polymer network would be obtained. On the other extreme,
if the amount of divinyl monomer is too low, diblock
copolymers with some branches of poly(NIPAAm-co-acid
comonomer) are obtained. A star polymer with crosslinked core is only obtained at conditions where the divinyl
monomer content is enough for branching but not enough
for a wall to wall macroscopic cross-linking. In previous
studies, we found that the concentration of macro-CTA and
divinyl monomer in the recipe plays an important role in
this reaction.[61,62] Therefore, we fixed a macro-CTA concentration and swapped a range of DVB cross-linker to macroCTA ratio to find conditions for star polymer synthesis.
The synthetic scheme is described in Figure 1, whereas in
Figure 5A we show preliminary experiments performed
Figure 5. GPC traces of crude products for the synthesis of PNIPAAm-PDVB stars after 24 h polymerization time: (A) Molar ratio DVB:macroCTA for (a) 30:1; (b) 25:1 and (c) 20:1. (B) Molar ratio DVB:macro-CTA (30:1) and star product after fractionation.
www.MaterialsViews.com
Macromol. Chem. Phys. 2012, 213, 301−314
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
309
Macromolecular
Chemistry and Physics
L. A. Picos-Corrales et al.
www.mcp-journal.de
using PNIPAAm macro-CTA as a model system to fix the
cross-linker to macro-CTA ratio. As we can see from the
GPC traces, with increasing ratio of cross-linker to macroCTA the content on star polymer increases. Nevertheless, a
large amount of non-coupled macro-CTA (arms) are found
in the crude product. Further increase of the ratio to 35:1
increased more the GPC peak for the star product, whereas
a 40:1 ratio resulted in a macro gel formation. In Figure 5B,
GPC traces of the star product (30:1) before and after fractionation are shown and compared to the macro-CTA used
for its synthesis. The star product is enriched; however, a
shoulder towards lower molecular weights is still observed,
suggesting that the purification process was not complete.
Further fractionation did not improved the GPC chromatogram; however, when we studied the product by DLS, we
obtain a narrow distribution of sizes. We can see this pictured in Figure 6 for the star product (after fractionation)
poly(NIPAAm-co-5MPA)arms-PDVBcore. This measurement
was performed in THF, a good solvent for both arms and
core of the star. The average size is 17 nm. Furthermore,
we can see in Figure 6 that the size distribution in water
at pH 7.4 is broad and much bigger, an indication of selfaggregation of the star polymer. The average size in water
is 45 nm.
Table 5 show the results of characterization experiments on star polymer products prepared with three different type of arms: PNIPAAm and two PNIPAAm copolymers with acid comonomers having an LCST close to
38 °C. As we can see, the polymer products are polydisperse
(PDI = 1.6 to 2) and the molecular weight increases, as
compared with that of the macro-CTA, from 4 to 8 times,
suggesting that 4–8 arms are incorporated in the star products; in fact, the number of arms should be higher because
the determination of molecular weight of star products
by GPC is not very reliable. Hydrodynamic volume of star
polymers is contracted compared to linear analogs of the
Figure 6. Size distribution by volume for poly(NIPAAm-co5MPA5%)arms-PDVBcore star polymer (35:1) by DLS: (a) macro-CTA in
THF; (b) star in THF; (c) star in water at pH 7.4.
same molecular weight, so the value of molecular weight
obtained by GPC may be underestimated. The composition of the star polymers as estimated by NMR (spectra in
Supporting Information) shows that DVB is incorporated
in 11 mol% or less in the star polymers. Apparently, the
content on DVB is lower in the case of the star products
with copolymeric arms; however, the NMR determination
for these polymer products was performed in deuterated
methanol, a non-solvent for DVB, whereas in the case of
PNIPAAm arms deuterated chloroform was used, which is
a good solvent for DVB. This procedure was necessary for a
better solvatation of the polymeric arms to determine the
acid comonomer content in the star polymers. In any case,
we have star polymer products, which can be confirmed
by further determination of architecture parameters. For
Table 5. Characteristics of star polymers in THF.
Mn,star/Mn,arm Dh,star/Dh,arm Composition arm/core
[mol%]e)
Star polymer
[Cros.:macroCTA]
Mn(GPC)d)
[g mol−1]
Mw/Mn
[PDI]
Rgd)
[nm]
Rhd)
[nm]
PNIPAAmarms-PDVBcorea)
[30:1]
136 800
1.947
10.7
8.7
5.276
3.164
89/11
Poly(NIPAAm-co5MPA5%)arms-PDVBcoreb)
[35:1]
108 100
2.005
8.6
8.3
4.046
2.594
(89-7)/4
Poly(NIPAAm-co4MBA10%)arms-PDVBcorec)
[35:1]
167 900
1.613
9.6
6.5
7.988
2.321
(85-10)/5
Prepared using PNIPAAm with Mn = 25 930 g mol−1, PDI = 1.093 and Dh(THF) = 5.5 nm; b)Prepared using P(NIPAAm-co-5MPA5%) with
Mn = 26 720 g mol−1, PDI = 1.125 and Dh(THF) = 6.4 nm; c)Prepared using P(NIPAAm-co-4MBA10%) with Mn = 21 020 g mol−1, PDI = 1.010 and
Dh(THF) = 5.6 nm; d)Data were obtained in THF at 25 °C for star with PNIPAAm arms and in THF:CH3COOH (50:1)v for star with random
copolymer arms; e)Determined by 1H-NMR.
a)
310
Macromol. Chem. Phys. 2012, 213, 301−314
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.MaterialsViews.com
Macromolecular
Chemistry and Physics
Well-defined N-Isopropylacrylamide . . .
www.mcp-journal.de
Table 6. Architecture parameters for star polymers and its aggregates.
ga,b)
ρc)
No. of armsd)
Dh,e,f)
[nm]
Dh,(pH 7.4)f)
[nm]
Xaggg)
LCSTh)
[°C]
PNIPAAmarms-PDVBcore
[30:1]
0.667
1.229
≈4
17.4
91.9
5
32
Poly(NIPAAm-co-5MPA5%)arms-PDVBcore
[35:1]
0.367
1.036
10–12
16.6
66.8
4
40
Poly(NIPAAm-co-4MBA10%)arms-PDVBcore
[35:1]
0.457
Star polymer
[Cros.:macroCTA]
38i)
1.477
6–8
13.0
59.0
4
>40
34j)
factors calculated related to linear polymers of PNIPAAm; first entry a polymer with following characteristics: ( Mn(GPC) =
60 500 g mol − 1 and Rg = 13.1 nm), second and third entry, following characteristics ( Mn(GPC) = 52 300 g mol − 1, Rg = 14.2 nm); b)Contraction
factor calculated by the equation g = Rg2,star/ Rg2,lin; c)Calculated by the ratio: ρ = Rg,star/Rh,star.d)According to W. Burchard;[72] e)Data were
obtained in THF for star polymer with PNIPAAm arms and in THF:CH3COOH (50:1)v for star polymer with random copolymeric arms; f)Dh
was determined at 25 °C.g)Number of aggregates calculated from the formula Xagg = Dh(pH 7.4)/Dh(THF); h)LCST was determined by turbidity
at pH 7.4; i)By DLS; j)By DLS at pH 7.0.
a)Contraction
example, it is well known that branched macromolecules
show a lower viscosity than linear polymers of the same
molecular weight. Furthermore, the contraction factors g
(based on R g) and g′ (based on viscosity) can be related to
the number of arms of a star polymer.[72] Viscosity results
show that the star polymers have a lower intrinsic viscosity
than linear counterparts (see Supporting Information). In
Table 6, a series of parameters for the star polymers are
shown. Results can be related to the number of arms in
the star polymer. Taking the g factor and the ρ parameter
in THF as described for star polymers in a good solvent.[72]
We postulate that the number of arms goes from 4 in the
star with PNIPAAm arms to between 6 and 12 for the star
products with copolymeric arms. As expected, the hydrodynamic diameter (Dh) in THF is lower for the star polymers
with more arms (copolymeric arms). The measurements of
Dh in water at pH 7.4 show that star to star aggregation
is occurring because the diameter is several times that of
the star polymer in THF. The number of stars per aggregate is calculated to be 4 to 5 (Table 6). The LCST studied by
turbidity shows that, in general, the star polymers kept the
LCST of the polymeric arms, although for the star polymer
with poly(NIPAAm-co-4MBA) arms the turbidity decreases
showing only a local maxima around 36 °C (Table 6, last
row and Figure 7). We must take into consideration that
the star polymers self-assemble in water below the LCST
of their arms. This aggregation process results from the
highly hydrophobic nature of DVB units in their cores. The
aggregates formed are relatively stable and do not precipitate further by increasing the temperature; therefore, the
turbidity is not very much affected by temperature rise.
For a more detailed discussion on this behaviour, we made
size investigations by DLS in heating and cooling cycles,
as is going to be discussed below. Preliminary conclusions
are that the LCST of the polymeric arms are kept when star
polymers are formed, however, the typical LCST behaviour—coil to globule transition followed by self-aggregation and precipitation—is not occurring. The core of the
stars plays a role in changing the LCST behaviour but not
the LCST value.
Figure 7. LCST by turbidity for star polymers: (A) PNIPAAmarms-PDVBcore star polymer; (B) poly(NIPAAm-co-5MPA5%)arms-PDVBcore star polymer
and (C) poly(NIPAAm-co-4MBA10%)arms-PDVBcore star polymer.
www.MaterialsViews.com
Macromol. Chem. Phys. 2012, 213, 301−314
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
311
Macromolecular
Chemistry and Physics
L. A. Picos-Corrales et al.
www.mcp-journal.de
Figure 8. Hysteresis of LCST for poly(NIPAAm90%-co-5MPA10%), Mn =
6600 g mol−1 (, heating; Δ, cooling) and for poly(NIPAAm88%-co
-4MBA12%), Mn = 5600 g mol−1 (, heating; , cooling).
3.4. Effect of Polymer Architecture in Reversibility
of LCST Behaviour
The next question we wanted to answer was if the LCST
behaviour is reversible for all type of structures we prepared with PNIPAAm and copolymeric PNIPAAm units. In
the literature, it is reported that linear PNIPAAm shows a
hysteresis in its LCST behaviour of less than 2 °C,[13] and the
authors demonstrate that the LCST cannot be fully reversible because of other transition states involved and strong
interactions with water molecules. We take this report as
starting point to compare the LCST behaviour by heating
and by cooling of the materials prepared. The purpose was
to see if the behaviour is similar to that of pure PNIPAAm,
to see how is the effect of comonomer and also to see the
effect of architecture, if any.
First of all, the LCST behaviour of PNIPAAm at a concentration of 0.4 mg mL−1 at pH 7.0 shows a higher hysteresis than the one reported in the literature. At this
polymer concentration, aggregation and precipitation is
observed, while in the literature report they prevent this
by studying a very low concentration of a high-molecularweight PNIPAAm.[13] In our case, the cooling down cycle
shows that the precipitate formed, re-dissolves only after
long time and after solvating of the PNIPAAm chain is
good enough for solution without stirring (see Supporting
Information). In Figure 8, we can se the comparison of
heating and cooling curves for random copolymers of
PNIPAAm with short chains. As we can see, the behaviour
is similar to that of pure PNIPAAm: large hysteresis with
formation of big size aggregates at temperatures above
the LCST value. A similar trend was observed for other
statistical copolymers. In the case of block copolymers,
the situation is very different. First of all, we find that
312
Figure 9. Hysteresis of LCST for poly(NIPAAm83%-co-5MPA7%)b-PSt10%, Mn = 8800 g mol−1 at pH 6.03 (, heating; Δ, cooling)
and at pH 5.1 (, heating; , cooling).
hysteresis is small to negligible (less than 2 °C). Furthermore, it depends strongly on the aggregation behaviour
of the block copolymers and aggregate type below its
LCST. For example, in Figure 9 we can see the behaviour
of a short block copolymer at two different pH values. In
both cases, we see that the polymeric micelles formed at
temperatures below the LCST aggregate into more complex and big structures above the LCST, and that these
structures returned to the micellar structure by cooling
down with a small temperature hysteresis. However,
in Figure 10 we can see that the vesicles formed by the
mid-size block copolymer do not further aggregate above
Figure 10. Hysteresis of LCST for poly(NIPAAm86%-co-5MPA6.2%)-bPHAt7.8%, Mn = 35 000 g mol−1 at pH 4.4 (, heating; Δ, cooling) and
at pH 7.6 (, heating; , cooling).
Macromol. Chem. Phys. 2012, 213, 301−314
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.MaterialsViews.com
Macromolecular
Chemistry and Physics
Well-defined N-Isopropylacrylamide . . .
www.mcp-journal.de
Figure 11. Hysteresis of LCST for star polymers. Poly(NIPAAmco-5MPA5%)arms-PDVBcore at pH 7.5 (, heating; , cooling) and
poly(NIPAAm-co-4MBA10%)arms-PDVBcore at pH 7.0 (, heating; Δ,
cooling).
its LCST; they shrink in an almost continuously fashion
at pH 7.6, while at pH 4.4 they shrink and most likely
change structure from vesicle to a micelle. In both cases,
the behaviour seems to be fully reversible with a negligible hysteresis. Also in the case of a block copolymer
showing no clear LCST, poly(NIPAAm-co-4MBA12%)-b-PHA
at pH 7.0, the shrinkage of the aggregate was followed
up to 70 °C and by cooling the aggregate size was recovered with no hysteresis (see Supporting Information). In
the case of star polymers, we also find a shrinkage of the
aggregated star ensemble by going through the LCST value
of the arms (Figure 11); however, in this case the reversibility is not granted because a hysteresis of up to 2 °C is
observed together with a high scatter in size change. In
fact, the aggregated star ensemble is not recovered totally,
showing clearly non-reversibility.
4. Conclusion
RAFT polymerization is a methodology that allows for the
preparation of statistical, block and star polymers with
LCST close to 38 °C at pH 7 to 7.4. This can be achieved
with different comonomers and also in different molecular weights, in our case using acid comonomers. Once a
statistical copolymer is prepared with the wanted LCST
by RAFT, the fact that it is a “living” polymer chain, in this
case called a macro-CTA; allows that this chain can be reactivated to be used for the preparation of more complex
polymer structures such as block copolymers and star polymers. The so prepared complex polymer structures maintain in general terms the set LCST; however, the fact that
the complex polymer structures may self-aggregate under
www.MaterialsViews.com
specific conditions of concentration, pH and temperature,
provoke important behavioural differences for complex
polymer systems containing the same copolymeric chains.
For example, the same block copolymer may form at a specific pH value, micelles that further aggregate and finally
precipitate by heating above the LCST of the copolymeric
temperature sensitive block; however, at another pH value
it may form vesicles that shrink by heating above the LCST
value of the copolymeric block changing its structure to
a more stable micelle. Even if the transition temperature
(Ttr) is not very different to the LCST from the copolymeric
blocks, it is very important to study the overall effect for
the application of the complex polymer structures. Star
polymers show that Ttr close to the LCST of their copolymeric arms, however if they aggregate and precipitate
or only shrink by heating, depends strongly on the core
forming polymer. In our case, a highly hydrophobic core
resulted in star self-aggregation even below the LCST of
the copolymeric arms. Reversibility of the temperaturesensitive effect cannot be truly achieved: the behaviour of
statistical copolymers is similar to a pure PNIPAAm chain,
while for block copolymers surprisingly we found several
cases where, judging on overall size only, reversibility was
better than for a PNIPAAm chain. Self-aggregating star
polymers prepared showed non-reversibility of the temperature transition behaviour. The main advantage of using
acid comonomers with pKa values close to 6 for tailoring
the LCST of PNIPAAm is that a change in pH value of the
environment from 7 to 6 already results in a clear change
in LCST, making all the polymeric materials prepared good
candidates for further investigations aiming biomedical
application.
Supporting Information
Supporting Information is available from the Wiley Online
Library or from the author.
Acknowledgements: L.A.P.-C. thank DAAD for financing his
PhD sandwich program. We thank J. Alvarez-Sanchez for the
synthesis of one sample. Technical support by I.A. Rivero and
A. Ochoa (IT-Tijuana) and M. Mende (IPF Dresden) is gratefully
acknowledged. This investigation was supported by CONACYT
(Mexico) SEP2006-60792.
Received: August 8, 2011; Revised: October 18, 2011; Published
online: January 3, 2012; DOI: 10.1002/macp.201100468
Keywords: block copolymers; LCST, PNIPAM; RAFT; star polymers
[1] E. S. Gil, S. M. Hudson, Prog. Polym. Sci. 2004, 29, 1173.
[2] I. Dimitrov, B. Trzebicka, A. H. E. Müller, A. Dworak,
C. B. Tsvetanov, Prog. Polym. Sci. 2007, 32, 1275.
[3] T. Chen, R. Ferris, J. Zhang, R. Ducker, S. Zauscher, Prog.
Polym. Sci. 2010, 35, 94.
Macromol. Chem. Phys. 2012, 213, 301−314
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
313
Macromolecular
Chemistry and Physics
L. A. Picos-Corrales et al.
www.mcp-journal.de
[4] D. Roy, J. N. Cambre, B. S. Sumerlin, Prog. Polym. Sci. 2010, 35,
278.
[5] R. Satchi-Fainaro, R. Duncan, C. M. Barnes, Adv. Polym. Sci.
2006, 193, 1.
[6] K. Ulbrich, V. Subr, Adv. Drug Delivery Rev. 2004, 56, 1023.
[7] E. S. Lee, Z. Gao, Y. H. Bae, J. Control. Rel. 2008, 132, 164.
[8] A. K. Bajpai, S. K. Shukla, S. Bhanu, S. Kankane, Prog. Polym.
Sci. 2008, 33, 1088.
[9] P. Bawa, V. Pillay, Y. E. Choonara, L. C. du Toit, Biomed. Mater.
2009, 4, 022001.
[10] A. Saeed, D. M. R. Georget, A. G. Mayes, React. Funct. Polym.
2010, 70, 230.
[11] A. E. Smith, X. Xu, C. L. McCormick, Prog. Polym. Sci. 2010, 35, 45.
[12] H. G. Schild, D. A. Tirrell, J. Phys. Chem. 1990, 94, 4352.
[13] X. Wang, X. Qiu, C. Wu, Macromolecules 1998, 31, 2972.
[14] F. Garret-Flaudy, R. Freitag, Langmuir 2001, 17, 4711.
[15] R. Freitag, F. Garret-Flaudy, Langmuir 2002, 18, 3434.
[16] Y. Zhang, S. Furyk, D. E. Bergbreiter, P. Cremer, J. Amer. Chem.
Soc. 2005, 127, 14505.
[17] J. E. Chung, M. Yokoyama, T. Aoyagi, Y. Sakurai, T. Okano,
J. Controlled Release 1998, 53, 119.
[18] Z. Li, S. O. Kyeremateng, K. Fuchise, R. Kakuchi, R. Sakai,
T. Kakuchi, J. Kressler, Macromol. Chem. Phys. 2009, 210,
2138.
[19] C. Tsitsilianis, G. Gotzamanis, Z. Iatridi, Eur. Polym. J. 2011,
47, 497.
[20] C. Diehl, H. Schlaad, Macromol. Biosci. 2009, 9, 157.
[21] X. Yin, A. S. Hoffman, P. S. Stayton, Biomacromol. 2006, 7,
1381.
[22] D. Kuckling, H. J. Adler, K. F. Arndt, L. Ling, W. D. Habicher,
Macromol. Chem. Phys. 2000, 201, 273.
[23] C. Poncet-Legrand, F. M. Winnik, Polym. J. 2001, 33, 277.
[24] R. Salgado-Rodriguez, A. Licea-Claverie, K. F. Arndt, Eur.
Polym. J. 2004, 40, 1931.
[25] A. Licea-Claverie, E. Rogel-Hernandez, R. Salgado-Rodriguez,
J. A. Lopez-Sanchez, L. A. Castillo, J. M. Cornejo-Bravo,
K. F. Arndt, Macromol. Symp. 2004, 207, 193.
[26] C. Erbil, Y. Yildiz, N. Uyanik, Polym. Adv. Technol. 2009, 20,
926.
[27] Y. Deng, R. Pelton, Macromolecules 1995, 28, 4617.
[28] A. Tuncel, D. Demirgöz, S. Patir, E. Piskin, J. Appl. Polym. Sci.
2002, 84, 2060.
[29] M. Hahn. E. Görnitz, H. Dautzenberg, Macromolecules 1998,
31, 5616.
[30] T. Nonaka, Y. Hanada, T. Watanabe, T. Ogata, S. Kurihara,
J. Appl. Polym. Sci. 2004, 92, 116.
[31] A. Durand, D. Hourdet, Macromol. Chem. Phys., 2000, 201, 858.
[32] R. Plummer, D. J. T. Hill, A. K. Whittaker, Macromolecules
2006, 39, 8379.
[33] A. P. Vogt, B. S. Summerlin, Macromolecules 2008, 41, 7368.
[34] E. Hasan, K. Jankova, V. Samichkov, Y. Ivanov,
Ch. B. Tsvetanov, Macromol. Symp. 2002, 177, 125.
[35] C. Gao, H. Möhwald, J. Shen, Polymer 2005, 46, 4088.
[36] S. Wohlrab, D. Kuckling, J. Polym. Sci., Part A: Polym. Chem.
2001, 39, 3797.
[37] J. C. Rueda, S. Zschoche, H. Komber, F. Krahl, K. F. Arndt,
B. Voit, Macromol. Chem. Phys. 2010, 211, 706.
[38] C. Zhao, X. Zhuang, C. He, X. Chen, X. Jing, Macromol. Rappid.
Commun. 2008, 29, 1810.
[39] D. Roy, J. N. Cambre, B. S. Sumerlin, Chem. Commun. 2009, 2106.
[40] C. M. Schilli, M. Zhang, E. Rizzardo, S. H. Thang, Y. K. Chong,
K. Edwards, G. Karlsson, A. H. E. Müller, Macromolecules
2004, 37, 7861.
314
[41] J. Virtanen, S. Holappa, H. Lemmetyinen, H. Tenhu, Macromolecules 2002, 35, 4763.
[42] S. I. Yusa, Y. Shimada, Y. Mitsukami, T. Yamamoto,
Y. Morishima, Macromolecules 2004, 37, 7507.
[43] A. J. Convertine, B. S. Lokitz, Y. Vasileva, L. J. Myrick,
C. W. Scales, A. B. Lowe, C. L. McCormick, Macromolecules
2006, 39, 1724.
[44] Y. You, C. Hong, W. Wang, W. Lu, C. Pan, Macromolecules
2004, 37, 9761.
[45] A. E. Smith, X. Xu, S. E. Kirkland-York, D. A. Savin,
C. L. McCormick, Macromolecules 2010, 37, 7507.
[46] Supporting Information from Y. Li, A. E. Smith, B. S. Lokitz,
C. L. McCormick, Macromolecules 2007, 40, 8524.
[47] Y. Li, B. S. Lokitz, C. L. McCormick, Macromolecules 2006, 39,
81.
[48] B. S. Lokitz, A. J. Convertine, R. G. Ezell, A. Heidenreich, Y. Li,
C. L. McCormick, Macromolecules 2006, 39, 8594.
[49] M. Karg, I. Pastoriza-Santos, B. Rodriguez-Gonzalez,
R. von Klitzing, S. Wekert, T. Hellweg, Langmuir 2008, 24, 6300.
[50] X. Ma, Y. Xing, Polym. Bull. 2006, 57, 207.
[51] V. G. Spizzirri, F. Lemma, F. Puoci, F. Xue, W. Gao, G. Cirillo,
M. Curcio, D. I. Parisi, N. Picci, Polym. Adv. Technol.
2011, 22, 1705.
[52] M. Keerl, W. Richtering, Colloid Polym. Sci. 2007, 285, 471.
[53] J. D. Debord, L. A. Lyon, Langmuir 2003, 19, 7662.
[54] A. B. Lowe, C. L. McCormick, Prog. Polym. Sci. 2007, 32,
283.
[55] G. Moad, E. Rizzardo, S. H. Thang, Acc. Chem. Res. 2008, 41,
1133.
[56] A. Licea-Claveríe, E. Rogel-Hernández, J. A. López-Sánchez,
L. A Castillo-Arámbula, J. M. Cornejo-Bravo, K. F. Arndt, Des.
Monom. Polym. 2003, 6, 67.
[57] J.
M.
Cornejo-Bravo,
J.
E.
Vazquez-Altamirano,
E. Rogel-Hernandez, A. Licea-Claverie, M. A. Ramos-Ibarra,
Des. Monom. Polym. 2007, 10, 263.
[58] Y. Mitsukami, M. S. Donovan, A. B. Lowe, C. L. McCormick,
Macromolecules 2001, 34, 2248.
[59] N. A. Cortez-Lemus, R. Salgado-Rodríguez, A. Licea-Claveríe,
J. Polym. Sci., Part A: Polym. Chem. 2010, 48, 3033.
[60] G. Clouet, M. Knipper, J. Brossas, Polym. Bull. 1984,
11, 171.
[61] A. Licea-Claverie, J. Alvarez-Sanchez, L. A. Picos-Corrales,
C. Obeso-Vera, M. C. Flores, J. M. Cornejo-Bravo, C. J. Hawker,
C.W. Frank, Macromol. Symp. 2009, 283–284, 56.
[62] J. Alvarez-Sanchez, A. Licea-Claverie, J. M. Cornejo-Bravo,
C. W. Frank, React. Funct. Polym. 2011, 71, 1077–1088.
[63] K. F. Arndt, G. Müller, in Polymercharakterisierung, Carl
Hanser Verlag, München, Germany 1996.
[64] G. Conio, E. Patrone, S. Russo, V. Trefiletti, Makromol. Chem.
1976, 177, 49.
[65] F. M. Winnik, A. R. Davidson, G. K. Haner, H. Kitano, Macromolecules 1992, 25, 1876.
[66] M. L. Adams, A. Lavasanifar, G. S. Kwon, J. Pharm. Sci. 2003,
92, 1343.
[67] G. Gaucher, M. H. Dufresne, V. P. Sant, N. Kang, D. Maysinger,
J. C. Leroux, J. Controlled Release 2005, 109, 169.
[68] M. Nuopponen, J. Ojala, H. Tenhu, Polymer 2004, 45, 3643.
[69] A. Vega-Rios, A. Licea-Claverie, J. Mex. Chem. Soc. 2011, 55,
21.
[70] M. Motornov, Y. Roiter, I. Tokarev, S. Minko, Prog. Polym. Sci.
2010, 35, 174.
[71] D. Kafouris, C. S. Patrickios, Eur. Polym. J. 2009, 45, 10.
[72] W. Burchard, Adv. Polym. Sci. 1999, 143, 113.
Macromol. Chem. Phys. 2012, 213, 301−314
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.MaterialsViews.com