Structure-LCST relationships for end-functionalized water-soluble polymers: An "accelerated" approach to phase behaviour studies

Article abstract of DOI:10.1039/b702067h

A novel "high throughput" technique for LCST measurement was developed which is able to identify the effect of subtle changes in end group composition on the aqueous phase behaviour of water-soluble poly(2- (dimethylamino)ethyl methacrylate). The Royal Society of Chemistry.

Full text of DOI:10.1039/b702067h

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Structure–LCST relationships for end-functionalized water-soluble
polymers: an ‘‘accelerated’’ approach to phase behaviour studies{
Satyasankar Jana,a Steven P. Rannard*ab and Andrew I. Cooper*a
Received (in Cambridge, UK) 8th February 2007, Accepted 17th April 2007
First published as an Advance Article on the web 11th May 2007
DOI: 10.1039/b702067h
a uniform, linear temperature gradient throughout the well-plate,
controlling the temperature of liquids placed in each of the wells.
Fig. 1a shows the consistency of temperatures achieved within
the columns and the temperature gradients that have been
produced within the rows of the well-plate when one end of the
plate is heated to 90 uC and the other is cooled to 30 uC. Fig. 1b
demonstrates the flexibility of the approach: many different linear
temperature gradients may be achieved by simply adjusting the
temperature of each end of the plate.
A novel ‘‘high throughput’’ technique for LCST measurement
was developed which is able to identify the effect of subtle
changes in end group composition on the aqueous phase
behaviour of water-soluble poly(2-(dimethylamino)ethyl
methacrylate).
Thermoresponsive polymers are characterized by changes in
solubility behavior with temperature as a result of coil-to-globule
transitions.1a,b The lower critical solution temperature (LCST) for
such polymers is defined as the temperature at which the polymer
phase-separates from solution (often water) due to the breakdown
in hydrogen bonding interactions. Water-soluble polymers such as
poly(N-isopropylacrylamide) (pNIPAM), poly(ethylene oxide),
poly(2-(dimethylamino)ethyl methacrylate) (pDMAEMA) and
many proteins exhibit LCST phase behavior.2 This ‘‘intelligent’’
phenomenon has been exploited in supported catalysts,3a drug
delivery,3b biomaterials,3c nanomaterials,3d microfluidics3e and
other applications.3f,g
To demonstrate the applicability of this novel HT measurement,
the technique has been validated through the determination of the
LCST transitions are affected by a number of variables such as
the polymer composition, structure and molecular weight,
concentration in solution, electrolyte levels, solvent quality and
(particularly for pDMAEMA) pH. Typical determination meth-
ods, such as cloud point measurements, can be time-consuming
and involve slow temperature ramps of around 1–2 uC min21
.
Using these manual approaches, detailed multivariate studies can
be complex and laborious, particularly in a multidimensional study
to investigate the interactions between variables. The development
of accelerated approaches4a–d therefore offers benefits for the rapid
discovery of new functional thermoresponsive materials.
We present here a novel ‘‘high throughput’’ (HT) temperature
gradient approach for the determination of LCST. Temperature
gradients have been exploited previously by Amis and coworkers
for combinatorial polymer thin film phase behavior screening,4d,5
and by Mao et al, using a capillary-based gradient device for
aqueous solution state LCST measurement.6a,b Our approach,
however, allows the HT analysis of larger numbers of samples and
enables complex multivariate, multidimensional studies through
the utilisation of a clear polystyrene 384-well microplate format
placed upon a recessed copper base. The base is simultaneously
heated and cooled by passing silicone fluid at controlled
temperature through each end. This differential heating produces
aDepartment of Chemistry, University of Liverpool, Crown Street,
Liverpool, UK L69 3BX. E-mail: aicooper@liv.ac.uk
bUnilever R & D, Port Sunlight Laboratory, Quarry Road East,
Bebington, Wirral, UK L63 3JW
{ Electronic supplementary information (ESI) available: Details of
equipment, validation experiments, synthesis procedures and expansions
of optical micrographs. See DOI: 10.1039/b702067h
Fig. 1 Generation of static temperature gradients in 384 well-plates (a)
Reproducibility of temperature and gradient across rows and through
columns, (b) Linear temperature gradients controlled by base-plate
temperature (inset shows hot and cold circulator temperatures, uC).
2962 | Chem. Commun., 2007, 2962–2964
This journal is ß The Royal Society of Chemistry 2007

View Article Online
often reported LCST for pNIPAM (1 wt%, Mw = 300 6
103 g mol21) in neutral distilled water (pH = 9.0–9.5). Importantly,
we have taken 16 static repeat measurements (Fig. 2a), allowing
for a more accurate determination than conventional techniques.
The LCST was found to be 30.6 ¡ 0.4 uC from the data presented
in Fig. 2a which is in accordance with the literature.6a,7 The phase
transition can also be analyzed in more detail by ‘‘dynamically
scaling’’ the temperature gradient to expand the LCST region
(Fig. S4{); a more precise LCST value of 30.7 ¡ 0.1 uC was
obtained in this case. Image analysis (Fig. 2b) allowed the
observation of the LCST transition even within a single well
(Fig. 2c), demonstrating that the working temperature resolution is
in fact better than the nominal R/24 uC for a given temperature
range (where R = the overall temperature range and 24 is the
number of wells per row).
Processed images (Fig. S4b{) show a distinct similarity with a
more standard plot of transmittance versus temperature from
traditional devices such as UV-visible spectrophotometers,1b
despite the fact that the data was acquired in all 16 channels
(i.e., rows) in ‘‘one shot’’. ‘‘Two-dimensional’’ measurements
investigating the effect of KCl (16 concentrations) on the LCST
transition were also accomplished (Fig. 2d). There was excellent
agreement with the previously reported linear relationship found
using conventional cloud point measurements7 (see data super-
imposed in Fig. 2d).
Scheme 1 Synthesis of end-functionalized pDMAEMA using aqueous
ATRP.
ATRP polymerization of DMAEMA, 2, under our conditions,
samples taken at specific times have predictable degrees of
polymerization (DP) and narrow molecular weight distributions
(PDI). It was thus possible to generate well-defined libraries with a
range of molecular weights by ‘‘temporal sequential sampling
ATRP’’ (see Supplementary Information{). For each end group,
narrow polydispersity polymers with DPs in the ranges a = 7–9,
b = 12–13, c = 15–16, d = 17–18, e = 19–20, f = 21–22 and g = 22–
23 monomer units were produced and fully characterized by both
1H nuclear magnetic resonance spectroscopy (NMR) and gel
permeation chromatography (GPC), Fig. 3. Thus, a library of
pDMAEMA polymers, 3, varying systematically in molecular
weight and end group was generated.
The method was developed primarily to allow the investigation
of subtle and complex structure-phase behavior relationships for
various polymer structures and topologies. The controlled
synthesis of a library of 49 pDMAEMA materials with seven
systematically varying end groups and seven selected degrees of
polymerization (DP), has been accomplished using aqueous atom
transfer radical polymerization techniques,8 (Scheme 1). End
group variation was accomplished by using a small library of seven
tertiary bromide ATRP initiators, 1. Since the number average
molecular weight (Mn) varies linearly over the timescale of the
Fig. 4a shows the variation in LCST with changing end group
for two sub-libraries; the lowest DP samples (R1a–7a, DP = 7–9)
and those with the highest DP (R1g–7g, DP = 22–23). Fig. 4b
illustrates the image processing used to extract the LCST values.
The higher DP polymers (R1g–7g) generally have higher LCST
values for a given end group. This is probably due to the effect of
the hydrophobic end group becoming less pronounced with
increased polymer chain length. This was supported by a more
detailed study of the effect of DP on LCST for the ethyl- and
propyl- end-functionalized series of samples (Supplementary
Information, Fig. S9{). Likewise, it was shown previously that a
Fig. 2 High throughput determination of poly(N-isopropylacrylamide)
phase behaviour (a) LCST in distilled water, 16 repeats; (b) image analysis
of 384 well plate (temperatures referred to are in uC); (c) phase transitions
observed within single wells; (d) static ‘‘two dimensional’’ phase transition
measurement (temperature versus KCl concentration).7
Fig. 3 Overlaid GPC chromatograms of the ethyl end-group
pDMAEMA sub-library (polymers R1a–g).
This journal is ß The Royal Society of Chemistry 2007
Chem. Commun., 2007, 2962–2964 | 2963

View Article Online
transitions were easily determined for the more hydrophobic end
groups at increased DP (Fig. 4a). In general, more hydrophobic
end groups led to lower LCST values. The ordering of the butyl-
and octyl-end-functionalized polymers (R3g and R4g) is anom-
alous in this respect and cannot currently be explained, although
analysis of the characterization data (NMR, GPC) and validation
of these results by conventional LCST measurements suggests that
this ordering is correct. In general, the trends are less regular for
the low molar mass samples (R1a–R7a) although the lowest LCST
values are observed for the most hydrophobic end-groups (benzyl
and phenethyl). This can probably be explained by the
polymerization process: while the control over molecular weight
is very good, even variations of just one DMAEMA repeat unit
might be expected to compete with single methylene changes in the
end group structure of oligomers with an average DP of just 7–9.
With these caveats, the data presented in Fig. 4 and Table S2
represent the basis of a rational ‘‘hydrophobicity scale’’ for
structure-induced changes in phase behavior for such polymers.
In conclusion, we have demonstrated a novel HT approach
capable of investigating subtle solution structure-phase behavior
relationships which may greatly aid in the design of functional
thermoresponsive materials. We have used this methodology to
show convincingly that the LCST phase behavior for low DP
pDMAEMA is affected by very subtle (single methylene group)
changes in the structure of the initiating end group.
We thank EPSRC and Insight Faraday for support (GR/
T04328/01; EP/C511794/1) and Mr Richard Dewson for his skill in
designing and building the apparatus.
Notes and references
1 (a) H. G. Schild, Prog. Polym. Sci., 1992, 17, 163–249; (b) X. Yin,
A. S. Hoffman and P. S. Stayton, Biomacromolecules, 2006, 7, 1381–1385.
2 E. S. Gil and S. M. Hudson, Prog. Polym. Sci., 2004, 29, 1173–1222.
3 (a) H. Hamamoto, Y. Suzuki, Y. M. A. Yamada, H. Tabata, H. Takashi
and S. Ikegami, Angew. Chem., Int. Ed., 2005, 44, 4536–4538; (b)
A. S. Hoffman and P. S. Stayton, Macromol. Symp., 2004, 207, 139–151;
(c) K. Haraguchi and H. J. Li, Angew. Chem., Int. Ed., 2005, 44, 2–6; (d)
P. W. Zhu, J. Mat. Sci: Mater. Med., 2004, 15, 567–573; (e) M. Ebara,
J. M. Hoffman, A. S. Hoffman and P. S. Stayton, Lab Chip, 2006, 6,
843–848; (f) S. Balamurugan, G. B. Bantchev, Y. Yang and
R. L. McCarley, Angew. Chem., Int. Ed., 2005, 44, 4872–4576; (g)
Q. Sun and Y. Deng, J. Am. Chem. Soc., 2005, 127, 8274–8275.
4 (a) C. M. Stafford, C. Harrison, K. L. Beers, A. Karim, E. J. Amis,
M. R. Vanlandingham, H. C. Kim, W. Volksen, R. D. Miller and
E. E. Simonyi, Nat. Mater., 2004, 3, 545–550; (b) M. A. R. Meier,
J. F. Gohy, C. A. Fustin and U. S. Schubert, J. Am. Chem. Soc., 2004,
126, 11517–11521.
Fig. 4 Determination of the Lower Critical Solution Temperature for
poly (2-(dimethylamino)ethyl methacrylate) (1 wt% of polymer and at
pH 9.0–9.5) (a) Static observation of the phase transitions polymer
libraries. End group variation: R1 = ethyl; R2 = n-propyl; R3 = n-butyl;
R4 = n-octyl; R5 = benzyl; R6 = phenethyl; R7 = 2-(dimethylamino)ethyl.
Degree of polymerization variation: top DP = 7–9; bottom DP = 22–23
monomer units. Temperature range = 36.4–55.1 uC. (b) Image analysis
and half-height gray scale determination of transition temperature for
polymer R2g (propyl end group, DP = 22–23 monomer units).
decrease in LCST occurs with increasing molar mass for
pDMAEMA with a benzomethoxy end group.9
5 J. C. Meredith, A. Karim and E. J. Amis, Macromolecules, 2000, 33,
5760–5762.
6 (a) H. Mao, C. Li, Y. Zhang, D. E. Bergbreiter and P. S. Cremer, J. Am.
Chem. Soc., 2003, 125, 2850–2851; (b) H. Mao, C. Li, Y. Zhang, S. Furyk,
P. S. Cremer and D. E. Bergbreiter, Macromolecules, 2004, 37,
1031–1036.
Surprisingly, we have been able to discern marked variations in
LCST resulting from very small structural differences in these
polymers. For example, three distinct LCST values for the ethyl,
propyl-, and butyl- end-functionalized pDMAEMA materials
were observed for both the low DP series (R1a–R3a; 47.8, 43.7,
47 uC, respectively) and the high DP series (R1–R3g; 51, 50.2,
47 uC), respectively, despite differences of only a single methylene
unit (Table S2).
7 H. Inomata, S. Goto, K. Otake and S. Saito, Langmuir, 1992, 8, 687–690.
8 (a) S. Coca, C. B. Jasieczek, K. L. Beers and K. Matyjaszweski, J. Polym.
Sci., Polym. Chem., 1998, 36, 1417–1424; (b) X. S. Wang and S. P. Armes,
Macromolecules, 2000, 33, 6640; (c) F. Zeng, Y. Shen, S. Zhu and
R. Petton, J. Polym. Sci., Polym. Chem., 2000, 38, 3821–3827; (d) B. Mao,
L. H. Gan, Y. Y. Gan, X. Li, P. Ravi and K. C. Tan, J. Polym. Sci., Part
A: Polym. Chem., 2004, 42, 5161–5169; (e) J. V. M. Weaver, I. Bannister,
K. L. Robinson, X. Bories-Azeau, S. P. Armes, M. Smallridge and
P. McKenna, Macromolecules, 2004, 37, 2395–2403; (f) V. Bu¨tu¨n,
S. P. Armes and N. C. Billingham, Polymer, 2001, 42, 5993–6008.
9 Q. Liu, U. Yu and P. Ni, Colloid Polym. Sci., 2004, 282, 387–393.
The least hydrophobic end group, 2-(dimethylamino)ethyl, is an
analog of the DMAEMA monomer repeat unit. Polymers initiated
from this group (R7a,g) showed only a faint LCST at 53.5 uC for
the lowest DP sample (studied from 36.4–55.1 uC). LCST
2964 | Chem. Commun., 2007, 2962–2964
This journal is ß The Royal Society of Chemistry 2007