View Article Online

View Journal

PCCP
Physical Chemistry Chemical Physics
Accepted Manuscript

This article can be cited before page numbers have been issued, to do this please use: L. Wang, X. Zhu,
W. Cai and X. Shao, Phys. Chem. Chem. Phys., 2019, DOI: 10.1039/C8CP07153E.

Volume 18 Number 1 7 January 2016 Pages 1–636 This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
PCCP
Physical Chemistry Chemical Physics Accepted Manuscripts are published online shortly after
www.rsc.org/pccp

acceptance, before technical editing, formatting and proof reading.

Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.

You can find more information about Accepted Manuscripts in the

author guidelines.

Please note that technical editing may introduce minor changes

to the text and/or graphics, which may alter content. The journal’s
ISSN 1463-9076 standard Terms & Conditions and the ethical guidelines, outlined
PERSPECTIVE
in our author and reviewer resource centre, still apply. In no
Darya Radziuk and Helmuth Möhwald
Ultrasonically treated liquid interfaces for progress in cleaning and
separation processes
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.

rsc.li/pccp
 Page 1 of 11 PleaseChemistry
Physical do not adjust margins
Chemical Physics

View Article Online

Journal Name DOI: 10.1039/C8CP07153E

ARTICLE

Physical Chemistry Chemical Physics Accepted Manuscript

Understanding the role of water during the aggregation of poly (N,
N-dimethylaminoethyl methacrylate) in aqueous solution using
Received 00th January 20xx,
Accepted 00th January 20xx
temperature-dependent near-infrared spectroscopy
Published on 15 February 2019. Downloaded by Tulane University on 2/15/2019 5:43:26 PM.

DOI: 10.1039/x0xx00000x Li Wang,a Xuewei Zhu,a Wensheng Caia and Xueguang Shaoa,b,c,d,e*
www.rsc.org/ For understanding the role of water in the aggregation of polymers, the variation of water structures with the structural
change of polymer in the process of aggregation was studied by temperature-dependent near-infrared (NIR) spectroscopy.
The NIR spectra of the aqueous poly (N, N-dimethylaminoethyl methacrylate) (PDMAEMA) solutions of different
concentrations were measured at different temperatures. The spectral changes of the polymer and water with
temperature were analyzed by N-way principal component analysis (NPCA). It was found that, at low concentration, the
chains of the polymer tend to form a loose hydrophobic structure below 36 oC and then aggregate into the micelle around
39 oC of the lower critical solution temperature (LCST). In the process of the aggregation, the water specie with two
hydrogen bonds (S2) increases gradually before 36 oC and then a sudden decrease occurs after the temperature. The
results clearly indicate that water specie S2 plays an important role in the formation of the intermediate, i.e., the loose
hydrophobic structure of the polymer chains linked by the two hydrogen bonds of S2 water. When temperature increases,
the dissociation of the hydrogen bonds makes the intermediate destroyed to form the micelle structure. For the high
concentration solution, however, the spectral information of S2 was not found in the aggregation, suggesting a direct
formation of the micelle from the dehydrated chains.

polymers have been studied. Wu et al.10 investigated the

1 Introduction hydration of oligo (ethylene glycol) methacrylate-based
polymers in water. They suggested that the dehydration of
Aqueous solutions of temperature-sensitive polymers exhibit
side chains plays a key role in the self-aggregation process of
phase separation above lower critical solution temperature
the polymer. They also studied the hydration behavior of poly
(LCST).1 When temperature is below LCST, the polymers are
(N-vinylcaprolactam) in water,11 indicating that a hydrogen
soluble in water to form aqueous solutions. Once the
bonding transformation predominates in the first stage below
temperature of the solutions exceeds LCST, however, the
LCST while hydrophobic interactions predominate in the
polymers become insoluble in water. Because of their unique
second stage above LCST. Cho et al.12 studied the changes of
properties, LCST polymers have gained diverse applications in
the bonded water in the phase transition of poly (N-
drug delivery,2,3 tissue engineering4 and molecular actuation.5
isopropylacrylamide) aqueous solution. They found that the
Generally, the soluble state of the polymers is considered to
bonded water is released to bulk water resulting in the self-
be hydrated with water through hydrogen bonds. As the
aggregation of the polymer chains. Asakura et al.13
temperature increases above LCST, the hydrophobic moieties
demonstrated that fast exchange occurs among free water and
of the polymers release the hydrogen bonded water and
water molecules bonded on the surface of polymer below
aggregate. The interaction of water and the polymer is
LCST. Wu et al.14 observed that the polymer chains are more
supposed to be the key factor driving the aggregation.6−9
collapsed in water/methanol clusters. The reducing water
The hydration behavior in the aggregation process of
activity hinders the hydration of polymer chains. Therefore,
water may have an important role in the aggregation of
polymers. Further work is needed to understand the influence
a. Research Center for Analytical Sciences, College of Chemistry, Nankai University,
Tianjin 300071, P. R. China. of water in the phase transition.
b. Tianjin Key Laboratory of Biosensing and Molecular Recognition, Tianjin 300071, Near-infrared (NIR) spectroscopy has been used extensively
P. R. China.
c. State Key Laboratory of Medicinal Chemical Biology, Tianjin 300071, P. R. China.
to study the structure of water.15,16 Due to the strong NIR
d. Collaborative Innovation Center of Chemical Science and Engineering (Tianjin), spectral responses of water, the water clusters with different
Tianjin 300071, P. R. China. number of hydrogen bonds can be distinguished from the NIR
e. Xinjiang Laboratory of Native Medicinal and Edible Plant Resources Chemistry,

College of Chemistry and Environmental Science, Kashgar University, Kashgar spectra of water.17 Aquaphotomics is proposed that provides a
844006, P. R. China. framework for understanding the characteristic water
*E-mail: xshao@nankai.edu.cn.

This journal is © The Royal Society of Chemistry 20xx J. Name., 2013, 00, 1-3 | 1

Please do not adjust margins

 PleaseChemistry
Physical do not adjust margins
Chemical Physics
ARTICLE
absorbance patterns caused by various perturbations such as
temperature.18,19 NIR aquaphotomics provided an accurate
description of the dynamics of water associated with
amyloidogenic and the ionic hydration, e.g., the ionic liquids or
salts.20-23 The changes in the hydration and secondary
structure of human serum albumin or ovalbumin in water at
different temperatures were detected by NIR spectroscopy
with the aid of chemometric methods.24,25 In our previous
studies, the quantitative relationship between NIR spectra and
temperature was established26−28 and then the variation of
water is used to detect the content of glucose29 and estimate
the composition of n-alkane30 and water-ethanol mixtures.31
The hydrogen bonding of aliphatic amines32 and hydration of
oligopeptide,33 proteins34 and human serum samples35 were
Published on 15 February 2019. Downloaded by Tulane University on 2/15/2019 5:43:26 PM.
also analyzed by temperature-dependent NIR spectroscopy. In
addition, variation of water structure induced by glucose36 and
the function of water during gelation of globular protein37
were analyzed for understanding the biological processes.
Water was proved to be a good probe for NIR spectroscopic
analysis to explore the information of the solutes in aqueous
solutions.
Poly (N, N-dimethylaminoethyl methacrylate) (PDMAEMA) is
a well-known temperature-sensitive polymer.10,38 In this work,
based on the phase transition mechanism of the polymer in
water, temperature-dependent NIR spectroscopy combined
with chemometrics was employed to study the role of water in
the aggregation of PDMAEMA. A high order algorithm, N-way
principal component analysis (NPCA), was used to analyze the
four-dimensional data array obtained by measuring the NIR
spectra at different temperatures and time for the samples
with different concentrations. The variations of water species
in the aggregation of the polymer were observed for
understanding the role of water in the process.
2 Experimental section
2.1 Materials
N, N-dimethylaminoethyl methacrylate (DMAEMA) was
purchased from Alfa Aesar. PDMAEMA (Chart 1) was
synthesized via reversible addition-fragmentation chain
transfer (RAFT) polymerization method following the
procedure reported in the literature.38−40 The structure of the
synthesized product is confirmed by 1H NMR measurements
(in CDCl3). According to the NMR spectrum, the degree of the
polymerization was determined to be 85. The number- and
weight-average molecular weight, Mn and Mw, of synthesized
product were determined by gel permeation chromatography
(GPC) measurements with monodisperse polystyrene as
standard and THF as eluent phase at 35 oC. The two
parameters are 1.15×104 and 1.25×104, respectively, and the
poly-dispersity index Mw/Mn is 1.08, suggesting a uniform
distribution. Solutions containing 0, 10, 20, 40, 60, 100, 150,
2 | J. Name., 2012, 00, 1-3
Please do not adjust margins
Page 2 of 11
Journal Name
CN S View Article Online
DOI: 10.1039/C8CP07153E
nS
O
O
O
Physical Chemistry Chemical Physics Accepted Manuscript
N
Chart 1 Chemical structure of PDMAEMA (the structures of initiator
and transfer agent are included.)
200, 250 mg mL-1 of the synthesized product were prepared
using redistilled water as the solvent.
2.2 Methods
All NIR spectra were measured using a Vertex 70 spectrometer
(Bruker Optics Inc., Ettlingen, Germany) equipped with a
tungsten-halogen light source and InGaAs detector. The
cuvette having a length of 1 mm was used. The spectra were
measured at 4 cm−1 resolution from 12000 to 4000 cm-1 and
the spectrum of empty cell was used as reference. To increase
the signal to noise ratio, 32 scans were accumulated. The
temperature in the experiment was controlled using a model
2216e temperature controller (Bruker Optics Inc., Ettlingen,
Germany). The precision of the equipment for temperature
control is 0.1 oC. For each sample, the spectra at temperatures
from 30 to 39 oC with a step of 1 oC were measured. For each
temperature, the spectrum at every one minute within one
hour after the controlled temperature changes was measured.
Because the temperature before the first 15 minutes may not
be stable, the spectra measured after 20 minutes were
adopted in the analysis.
2.3 Calculation
Continuous wavelet transform (CWT) was used to improve the
spectral resolution and remove the variant background in the
spectra.41−44 As generally used in our previous works,29−37
Symmlet with vanishing moment 2 (Sym2) was adopted as the
wavelet filter. The scale parameter was set to 30 in this work
for enhancing the smoothing effect.
A four dimensional NPCA model was used to analyze the
data array of the NIR spectra measured at different
temperatures and time for the samples of different
concentrations. For the four-way data array X with dimensions
of I × J × K ×L, where I is the number of temperature, J is the
number of the wavelengths in the spectra, K is the number of
samples with different concentrations, and L is the number of
the spectra measured at different time, a loading, BN×J, and
three scores, AN×I, CN×K, and DN×L, can be obtained, if number of
the components is set to N. The four matrices represent the
features contained in the spectra and the weights for these
features at each temperature, concentration and time,
respectively. If ain, bjn, ckn, and dln are used to represent the
elements of the nth component at ith temperature, jth
This journal is © The Royal Society of Chemistry 20xx
 Page 3 of 11 PleaseChemistry
Physical do not adjust margins
Journal Name
wavelength (wavenumber), kth concentration and lth time, the
element of the data array xijkl can be reconstructed by,
N
=x ijkl ∑a b c d + eijkl
in jn kn ln
n =1
(i = 1, …, I; j = 1,…, J; k = 1,…, K; l = 1,…, L)
where eijkl is the residual that is not fitted by the model, and n
(=1, …, N) is the index of the component.
NPCA is a development of the ordinary PCA,45 which
explains the variance as much as possible. The loading is not
the spectra of the components in the analyzing system but a
linear combination of them. Even though, the spectral features
Published on 15 February 2019. Downloaded by Tulane University on 2/15/2019 5:43:26 PM.
contained in the measured spectra can be known from the
loadings. Therefore, the variation of the spectral features can
be observed from the corresponding scores.
3 Results and Discussion
3.1 Spectral analysis
Fig. 1 shows the NIR spectra of water, 40 mg mL-1 and 150 mg
mL-1 aqueous solutions of PDMAEMA measured at different
temperatures. Only the spectra in the region from 7400 to
5500 cm-1, which will be analyzed below, were plotted in the
figure. Obviously, only a broad peak around 6900 cm-1 can be
found in Fig. 1, which is assigned to the O−H stretching
combination of ν1 (symmetric stretching mode) and ν3
(antisymmetric stretching mode).17 The peak is generally
considered to be an overlap of the absorptions arising from
the O−H groups in water species with different hydrogen
bonding, i.e., the O-H without hydrogen bonds and the
hydrogen bonded O-H in water clusters of different
sizes.17,19,23,36,37 The free O-H in water molecules without
hydrogen bonding was named as S0, and S1, S2, S3, S4
represents the O-H in water clusters with one, two, three, four
hydrogen bonds, respectively. As the temperature increases,
Fig.1 NIR spectra of water (a), 40 mg mL-1 PDMAEMA solution (b) and 150 mg mL-1 PDMAEMA solution (c) measured at the temperature
from 30 to 39 oC in the region of 7400−5500 cm-1.
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins
Chemical Physics
ARTICLE
the position of the peak slightly shifts to higher wavenumber.
View Article Online
It can be explained as the change in DOI: contents of different
10.1039/C8CP07153E
water clusters or the different hydrogen bonding states.46-48
However, in Fig. 1 (b) and (c), a significant baseline drift of the
spectrum occurs when temperature reaches to 39 oC or 38 oC
Physical Chemistry Chemical Physics Accepted Manuscript
for 40 mg mL-1 and 150 mg mL-1 polymer solutions,
(1) respectively. This may be caused by the turbidity of solution
due to the aggregation of the polymer. When the temperature
further increases above 39 oC or 38 oC, the micelle of the
polymer forms and phase separation occurs. It was noticed
that the LCST of the low concentration solution is about 39 oC,
which is different from that of the high concentration solution
at 38 oC. This may suggest a different mechanism for the
aggregation in the solutions of different concentrations.
Due to the baseline drift of the spectra and the overlap of
the broad spectral peak, as shown in Fig. 1, it is difficult to find
further information of the polymer. Therefore, baseline
correction and resolution enhancement of the spectra are
conducted using CWT. Fig. 2 shows the transformed spectra
from that in Fig. 1. CWT with the filter “Sym2” is an
approximate equivalence of the second derivative,44 thus the
negative peaks correspond to the positive ones in the raw
spectra. The changes in absolute value of the negative peaks
were used to analyze the variation. Obviously, two peaks at
7070 and 6824 cm-1 can be found, which are ascribed to the
vibrational mode of free and hydrogen bonded water
molecules, respectively.15,49 As temperature rises, an increase
around 7070 cm-1 and a decrease around 6824 cm-1 were
observed in the transformed spectra. The result indicates that
the content of free water molecules increases and the
hydrogen bonded water molecules decreases. This can be
explained by the effect of temperature on the hydrogen
bonding, i.e., the rising temperature weakens the hydrogen
bonding and results in dissociation of the water clusters.17,36 In
addition, in Fig. 2 (b) and (c), the two peaks related to the
polymer can be found in the transformed spectra. In the inset,
the two peaks at 5946 and 5811 cm-1 can be assigned to the
J. Name., 2013, 00, 1-3 | 3
 PleaseChemistry
Physical do not adjust margins
Chemical Physics Page 4 of 11

ARTICLE Journal Name

View Article Online

DOI: 10.1039/C8CP07153E

Physical Chemistry Chemical Physics Accepted Manuscript

Published on 15 February 2019. Downloaded by Tulane University on 2/15/2019 5:43:26 PM.

Fig. 2 Transformed spectra of water (a), 40 mg mL-1 PDMAEMA solution (b) and 150 mg mL-1 PDMAEMA solution (c) measured at
temperatures from 30 to 39 oC in the spectral range of 7400−5500 cm-1. The inset shows the enlarged spectra in the range of 6050−5650
cm-1.

first overtones of the antisymmetric stretching (2ν3) of the
C−H in CH3 groups and the combination of ν1 and ν3 of the C−H
in CH2 groups in the side chain of the polymer,
respectively.8,21,28,50,51 On the other hand, the significant
change can also be found for the spectra at 39 oC or 38 oC from
the transformed spectra, indicating the formation of the
polymer micelle.
In order to investigate the spectral change of the polymer in
aqueous solution during the phase transition, the variation of
the spectral intensity at 5946 cm-1 with temperature for the
samples with different concentrations was plotted in Fig. 3. At
low temperature, only a very slight change (decrease in
Fig. 3 Variation of the spectral intensity at 5946 cm-1 with
temperature for the solutions with different concentrations.
absolute) of the spectral intensity can be seen. When
temperature rises up to 39 or 38 oC for the low (40 and 60 mg
mL-1) and high (100 and 150 mg mL-1) concentration solution,
respectively, a sharp decrease occurs, indicating the formation
of the micelle. The difference between the solutions of
different concentrations may be brought about by different
mechanism of the phase transition. It has been reported in
previous works that the LCST of PDMAEMA is dependent on
concentration.10,52,53 When the concentration of the polymer is
high, the polymer can aggregate quickly, leading to a decrease
of the LCST.52−54
3.2 Structural change of the polymer during the aggregation
For understanding the mechanism of the phase transition, the
structural changes of the polymer before the phase transition
was investigated in this study. NPCA was used to extract the
information from the spectra measured at different
temperatures and time for the solutions of different
concentrations. Due to the different LCST, the spectra for the
samples of low (10, 20, 40, 60 mg mL-1) and high (100, 150, 200,
250 mg mL-1) concentrations are treated separately, named as
dataset A and dataset B, respectively. Dataset A includes the
transformed spectra measured at temperature range 30−38 oC
and the temperature range for dataset B is 30−37 oC.
Fig. 4 shows the results for dataset A, including the loadings
of the first two principal components (PCs) and corresponding
scores in the dimension of time, concentration and
temperature, respectively. More than 99.9 % of the spectral
variances are accounted for by the two PCs. Because the
structure of the polymer is studied, only the spectral range of
6200−5600 cm-1, which belongs to the spectral response of
C−H, was used in the calculation.

4 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 20xx

Please do not adjust margins

 Page 5 of 11 PleaseChemistry
Physical do not adjust margins
Chemical Physics

Journal Name ARTICLE

View Article Online

DOI: 10.1039/C8CP07153E

Physical Chemistry Chemical Physics Accepted Manuscript

Published on 15 February 2019. Downloaded by Tulane University on 2/15/2019 5:43:26 PM.

Fig. 4 The first two PCs extracted by NPCA from dataset A in the region of 6200−5600 cm-1: the loading (a1−a2), time scores (b1−b2),
temperature scores (c1−c2) and concentration scores (d1−d2). The lines in (c1) and (c2) were obtained by piecewise fitting with a linear and
a two-degree polynomial function, respectively.
As shown in Fig. 4 (a1) and (a2), two negative peaks at 5946,
5813 cm-1 and 5944, 5809 cm-1 were obtained in the first and
second loadings, respectively. According to the
literatures,8,21,30,50,51 the two peaks correspond to the
absorption of 2ν3 and ν2+ν3 of the C−H in CH3 and CH2,
respectively. This result indicates that the CHn groups in the
system are in two kinds of atmospheres, i.e., the hydrated
ones for the peaks at high wavenumber (in PC1) and un-
hydrated ones for the peak at lower wavenumber (in PC2),
respectively.12,55
Fig. 4 (b) shows the variation of the two scores with elapsed
time since the temperature is changed. Clearly, the results
indicate that the spectral features in the two PCs do not
change with time within the examined one hour, suggesting
that the structure of the studied polymer is stable at a
constant temperature in the studied temperatures. Fig. 4 (c)
shows the variation of the two scores with temperature. It is
clear that the first score in Fig. 4 (c1) decreases gradually with
temperature before 36 oC and a sharp decrease happens after
the temperature. The result should be a clear indication that,
when temperature is low, most of the polymer chains are
stably distributed in the solution in the hydrated form. As the
temperature increases, the number of hydrated C−H groups
gradually decrease because of the dehydration and then a fast
dehydration process occurs at the higher temperature before
the phase transition. Fig. 4 (c2) shows the variation of un-
hydrated polymer chains with temperature. There is a clear
increase with the rise of temperature before 36 oC and due to
the less content in the solution, the increase seems faster than
the decrease of the hydrated ones. Such an increase can be
explained by the effect of temperature on the interaction
between water and the polymer chain.8 When the
temperature exceeds 36 oC, the increase stops and even a
slight decrease can be observed. Therefore, the polymer may
undergo a two-stage conformation change in the phase
transition, i.e., the hydrated chains of the polymer change into
an intermediate state of the loose hydrophobic structure
before 36 oC with the progress of the dehydration and the
structure is destroyed at 37 and 38 oC before the aggregation
into the micelle at LCST, 39 oC.10,56 Fig. 4 (d) displays the
variation of the two scores with concentration. Almost a linear
relationship is obtained in the concentration range of 10 − 60
mg mL-1. The result demonstrates that concentration does not
affect the structure in the studied concentration range.
Fig. 5 shows the loadings and the scores of the first two PCs
obtained by NPCA for the high concentration solution (dataset
B), which explain more than 99.9% variance of the data.
Although the location of the peaks is slightly different from
that in Fig. 4 (a), the main spectral features in the two loadings
can be explained similarly, i.e., the spectral information of the
hydrated and un-hydrated C−H groups in the side chain of the
polymer, respectively. Fig. 5 (b) shows the variation of the two
time scores. Similar result with that from Fig. 4 (b) was
obtained, indicating that the structure is stable at a constant
temperature. The variation of the temperature scores in Fig. 5
(c), however, is much different from that in Fig. 4 (c). Without
the inflection at 36 oC, the scores go downward and upward
with the increase of temperature. This clearly indicates that
the phase transition in high concentration solutions takes a
different mechanism, i.e., the aggregation happens gradually
with the dehydration of the polymer, without the intermediate
state of the loose hydrophobic structure. The reason may be
that, when the concentration is high, there is no space to form
the loose hydrophobic structure, thus the polymer chains
aggregate directly from the dehydrated form into the micelle.

This journal is © The Royal Society of Chemistry 20xx J. Name., 2013, 00, 1-3 | 5

Please do not adjust margins

 PleaseChemistry
Physical do not adjust margins
Chemical Physics Page 6 of 11

ARTICLE Journal Name

View Article Online

DOI: 10.1039/C8CP07153E

Physical Chemistry Chemical Physics Accepted Manuscript

Published on 15 February 2019. Downloaded by Tulane University on 2/15/2019 5:43:26 PM.

Fig. 5 The first two PCs extracted by NPCA from dataset B in the region of 6200−5600 cm-1: the loading (a1−a2), time scores (b1−b2),
temperature scores (c1−c2) and concentration scores (d1−d2). The lines in (c1) and (c2) were obtained by fitting with a two-degree
polynomial.
As for the variation of the two concentration scores shown in
Fig. 5 (d), similar conclusion was also obtained with that from
Fig. 4 (d), although a slight difference can be seen for the
linearity of the curves in the two figures.
3.3 Structural changes of water during the aggregation
For understanding the role of water in the aggregation process
of the polymer, the variation of water species in the process
was studied by performing NPCA on the transformed spectra
in the region 7150−6500 cm-1, which belongs to the spectral
response of water. Fig. 6 and Fig. 7 shows the first three PCs
extracted by three- and four-way NPCA, respectively, from the
spectra measured at different temperatures and time for pure
water and low concentration solutions (dataset A). The
explained spectral variance by each PC was labelled in the
figure.
The first loading in Fig. 6 (a1) explains 99.84 % of the
spectral variance, which indicates that the most of the
information is contained in this PC. Two peaks at 7069 and
6824 cm-1 can be found in the PC, which are assigned as S0 and
S2, respectively.17,19,36 The time score in Fig. 6 (b1) indicates
that the spectral component does not change at a constant
temperature. The temperature score in Fig. 6 (c1), however,
shows that the spectral information of the two water species
increase linearly with temperature. The result can be explained
by the structural changes of water, i.e., the rising temperature
weakens the hydrogen bonding and results in the dissociation
of large water clusters into small ones.17,36 Fig. 6 (a2) explains
0.15 % of the spectral variance, in which two peaks at 6907
and 6802 cm-1 can be found, which are the spectral feature of
water with different hydrogen bonding.17,19,36 These water
species also do not change at constant temperature, as shown
in Fig. 6 (b2), but a decreasing trend with temperature is
shown in Fig. 6 (c2). The result is a further indication of the
dissociation of water clusters when temperature rises.17,36 As
for the third loading shown in Fig. 6 (a3), two peaks at 6976
and 6891 cm-1 can be clearly seen, which should be the
spectral information of differently hydrogen-bonded water
clusters. However, the explained variance of the PC is only
3.3×10-4 %, which can also be seen from the values in Fig. 6
(b3). Fig. 6 (c3) further shows that the information of the PC
dose not related with temperature. The result clearly suggests
that this PC explains nothing but the tiny variation in the
measurement of the spectra.
Fig. 7 shows the results for the spectra of the polymer
solutions. Compared with Fig. 6, almost the same result was
obtained for the loading, time and temperature score of the
first PC, and the concentration score shows a decreasing trend.
This clearly indicates that the polymer dose not significantly
affect the structure of the bulk water in the solution except for

6 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 20xx

Please do not adjust margins

 Page 7 of 11 PleaseChemistry
Physical do not adjust margins
Chemical Physics

Journal Name ARTICLE

View Article Online

DOI: 10.1039/C8CP07153E

Physical Chemistry Chemical Physics Accepted Manuscript

Published on 15 February 2019. Downloaded by Tulane University on 2/15/2019 5:43:26 PM.

Fig. 6 The first three PCs extracted by NPCA from the spectra of water measured at different temperatures and time in the region of
7150−6500 cm-1: the loading (a1−a3), time scores (b1−b3) and temperature scores (c1−c3). The lines in (c) were obtained by piecewise
linear fitting of the data.

the slight decrease of the content with the increase of the
concentration. Compared with the second loading in Fig. 6 (a2),
Fig. 7 (a2) suggests the similar result, but a shift to the low
wavenumber can be found for the two peaks at 6895 and 6797
cm-1. This may indicate that the spectral information is arisen
from the hydration water around the polymer chains,12 either
in the form of so-called water clathrates around hydrophobic
group or the water bounded with the hydrophilic groups of the
polymer. In the third loading shown in Fig. 7 (a3), there is a
small peak at 6845 cm-1 besides the noise. However,
considering the explained variance (3.8×10-3 %) by the PC,
which is ten times of that in Fig. 6 (a3), the small peak should
be the spectral feature of the water with two hydrogen bonds
(S2) according to the location of the peak.17,19,36 The result may
be an evidence for the existence of the water molecules
serving as a bridge between polymer side chains to form the
loose hydrophobic structure.56
Fig. 7 (b) show the variation of the time scores. The results
are similar with that for pure water in Figs. 6 (b), indicating
that no structural change occurs at a constant temperature.
Comparing the temperature scores in Fig. 7 (c) and Fig. 6 (c),
however, it can be seen that similar results were obtained for
the first two scores, but there is a significant difference
between the third score in the two figures. A sudden down can
be found in the curve at 36 oC. The results for the former two
PCs suggest that dehydration occurs with the rising of
temperature to release more bonded water into bulk
water,6,13,57 but the result of the third score indicates the
change of S2 go up and then down with an inflection at 36 oC.
Such a result provides a clear proof for the water molecules (S2)
bridging the side chains of the polymer to form the
intermediate. When temperature exceeds 36 oC, the loose
hydrophobic structure is destroyed to release the bonded
water (S2) and starts to form the micelle till the LCST at 39 oC.
Therefore, the two-stage variation of the phase transition can
be observed from the changes of the water in the system.
Fig. 7 (d) displays the variation of three scores with
concentration. Clearly, the first score representing the spectral
information of the bulk water changes downward with the
concentration. This can be explained by the decrease of the

This journal is © The Royal Society of Chemistry 20xx J. Name., 2013, 00, 1-3 | 7

Please do not adjust margins

 PleaseChemistry
Physical do not adjust margins
Chemical Physics Page 8 of 11

ARTICLE Journal Name

View Article Online

DOI: 10.1039/C8CP07153E

Physical Chemistry Chemical Physics Accepted Manuscript

Published on 15 February 2019. Downloaded by Tulane University on 2/15/2019 5:43:26 PM.

Fig. 7 The first three PCs extracted by NPCA from dataset A in the region of 7150−6500 cm-1: the loading (a1−a3), time scores (b1−b3),
temperature scores (c1−c3) and concentration scores (d1−d3). The lines in (c) were obtained by piecewise linear fitting of the data.
water content in the system due to the increase of polymer
concentration. No changing rule, however, can be found for
the other two scores in Fig. 7 (d2) and (d3), suggesting that
there is no relationship between the spectral features in the
loadings with the concentration.
Fig. 8 shows the results of the first three PCs obtained by
NPCA for the spectra of dataset B, the samples in the
concentration of 100 − 250 mg mL-1. It can be seen that the
first two loadings in Fig. 8 (a1) and (a2) are similar with that in
Fig. 7. The third loading in Fig. 8 (a3), however, contains the
spectral feature of S1 and S3,17,19,36 rather than that of S2. Fig. 8
(b) shows the variation of the time scores, suggesting the same
conclusion as that from Fig. 7 (b). The variation of the
temperature scores in Fig. 8 (c), however, is much different
from that in Fig. 7. The inflections no longer exist in all the
three scores. The result is obviously consistent with the
assumption that the aggregation of the polymer at high
concentration happens directly from the dehydrated chains to
the micelle without the intermediate state of the loose
hydrophobic structure. Comparing the concentration scores
shown in Fig. 8 (d) and Fig. 7 (d), it can be found that the first
scores in the two figures are similar to each other, but an
increasing trend was obtained in Fig. 8 (d2) and (d3). It can be
explained by the increasing content of the hydrated water due
to the high polymer concentration.
4 Conclusions
The role of water in the aggregation of the LCST polymers was
studied using temperature-dependent NIR spectra. The NIR
spectra of the PDMAEMA aqueous solutions with different
concentrations were measured at different temperatures. By
correcting the baseline and enhancing the resolution of the
spectra with CWT, the transformed spectra were analyzed by a
high order chemometric algorithm, NPCA. At first, the
difference of the LCST between low and high concentration
solutions was found and the difference is caused by the
different mechanism of the phase transition. In high
concentration solution, the aggregation of the polymer
happens directly from the dehydrated chains to the micelle, in
the low concentration solution, however, a two-stage
conformation change was observed in the phase transition, i.e.,
the hydrated chains of the polymer change into an
intermediate state of the loose hydrophobic structure and
then the structure is destroyed before the aggregation into the

8 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 20xx

Please do not adjust margins

 Page 9 of 11 PleaseChemistry
Physical do not adjust margins
Chemical Physics

Journal Name ARTICLE

View Article Online

DOI: 10.1039/C8CP07153E

Physical Chemistry Chemical Physics Accepted Manuscript

Published on 15 February 2019. Downloaded by Tulane University on 2/15/2019 5:43:26 PM.

Fig.8 The first three PCs extracted by NPCA from dataset B in the region of 7150−6500 cm-1: the loading (a1−a3), time scores (b1−b3),
temperature scores (c1−c3) and concentration scores (d1−d3). The lines in (c1) and (c2) were obtained by piecewise linear fitting and the
line in (c3) was fitted with a two-degree polynomial.

micelle at LCST. Furthermore, S2 was found to play an

References
important role in forming the intermediate. S2 acts as a bridge
to connect the polymer chains in the loose hydrophobic 1 F. Liu and M. W. Urban, Prog. Polym. Sci., 2010, 35, 3−23.
2 Y. Qiu and K. Park, Adv. Drug Deliv. Rev., 2001, 53, 321−339.
structure, and the dissociation of the S2 bridge at high
3 J. R. McDaniel, D. J. Callahan and A. Chilkoti, Adv. Drug Deliv.
temperature is the reason for the formation of the micelle. Rev., 2010, 62, 1456−1467.
Temperature-dependent NIR spectroscopy combined with high 4 L. Yu and J. D. Ding, Chem. Soc. Rev., 2008, 37, 1473−1481.
order chemometric methods may be a powerful way to 5 B. Kim and A. Chilkoti, J. Am. Chem. Soc., 2008, 130,
analyze the high-dimensional spectra data to extract the 17867−17873.
6 K. Shiraga, H. Naito, T. Suzuki, N. Kondo and Y. Ogawa, J.
hidden information in the spectra of aqueous systems, and the
Phys. Chem. B, 2015, 119, 5576−5587.
results may be helpful to understand the role of water in the 7 M. H. Futscher, M. Philipp, P. Müller-Buschbaum and A.
aggregation of polymers. Furthermore, water may be an Schulte, Sci. Rep., 2017, 7, 17012.
effective probe for studying the aggregation of the polymers in 8 Y. Maeda, T. Kubota, H. Yamauchi, T. Nakaji and H. Kitano,
aqueous solutions. Langmuir, 2007, 23, 11259−11265.
9 L. Tavagnacco, E. Zaccarelli and E. Chiessi, Phys. Chem. Chem.
Phys., 2018, 20, 9997-10010.
10 S. T. Sun and P. Y. Wu, Macromolecules, 2013, 46, 236−246.
Conflicts of interest 11 S. T. Sun and P. Y. Wu, J. Phys. Chem. B, 2011, 115,
There are no conflicts to declare 11609−11618.
12 E. C. Cho, J. Lee and K. Cho, Macromolecules, 2003, 36.
9929−9934.
13 S. Kametani, S. Sekine, T. Ohkubo, T. Hirano, K. Ute, H. N.
Acknowledgements Cheng and T. Asakura, Polymer, 2017, 109, 287−296.
14 S. T. Sun and P. Y. Wu, Macromolecules, 2010, 43,
This study was supported by National Natural Science Foundation of 9501−9510.
China (Nos. 21775076), and the fundamental research funds for 15 V. H. Segtnan, Š. Šašić, T. Isaksson and Y. Ozaki, Anal. Chem.,
central universities (China). 2001, 73, 3153−3161.

This journal is © The Royal Society of Chemistry 20xx J. Name., 2013, 00, 1-3 | 9

Please do not adjust margins

 PleaseChemistry
Physical do not adjust margins
Chemical Physics Page 10 of 11

ARTICLE Journal Name

16 M. A. Czarnecki, Y. Morisawa, Y. Futami and Y. Ozaki, Chem.
Rev., 2015, 115, 9707−9744.
17 H. Maeda and Y. Ozaki, J. Near Infrared Spectrosc., 1995, 3,
191−201.
18 R. Tsenkova, NIR News, 2008, 19, 12−13.
19 R. Tsenkova, J. Near Infrared Spectrosc., 2009, 17, 303−314.
50 B. J. Sun, Y. N. Lin and P. Y. Wu, Appl. Spectrosc., 2007, 61,
View Article Online
765−771. DOI: 10.1039/C8CP07153E
51 Y. Maeda, H. Yamauchi and T. Kubota, Langmuir, 2009, 25,
479−482.
52 F. A. Plamper, M. Ruppel, A. Schmalz, O. Borisov, M. Ballauff
and A. H. E. Müller, Macromolecules, 2007, 40, 8361−8366.

Physical Chemistry Chemical Physics Accepted Manuscript

20 E. Chatani, Y. Tsuchisaka, Y. Masuda and R. Tsenvoka, PLoS
One, 2014, 9, e101997.
21 D. H. Dagade and S. S. Barge, ChemistrySelect, 2017, 2,
11703-11712.
22 D. Kojić, R. Tsenkova, K. Tomobe, K. Yasuoka and M. Yasui,
ChemPhysChem, 2014, 15, 4077−4086.
23 D. H. Dagade and S. S. Barge, ChemPhysChem, 2016, 17, 902-
912.
24 Y. Q. Wu, B. Czarnik-Matusewicz, K. Murayama and Y. Ozaki,
J. Phys. Chem. B, 2000, 104, 5840−5847.
25 Y. Wang, K. Murayama, Y. Myojo, R. Tsenkova, N. Hayashi
Published on 15 February 2019. Downloaded by Tulane University on 2/15/2019 5:43:26 PM.
53 K. Bebis, M. W. Jones, D. M. Haddleton and M. I. Gibson,
Polym. Chem., 2011, 2, 975−982.
54 J. Spěváček, J. Dybal, L. Starovoytova, A. Zhigunov and Z.
Sedláková, Soft Matter, 2012, 8, 6110−6119.
55 P. Schmidt, J. Dybal and M. Trchová, Vib. Spectrosc., 2006,
42, 278−283.
56 B. Zhang, H. Tang and P. Y. Wu, Macromolecules, 2014, 47,
4728−4737.
57 Y. Ono and T. Shikata, J. Am. Chem. Soc., 2006, 128,
10030−10031.

and Y. Ozaki, J. Phys. Chem. B, 1998, 102, 6655−6662.

26 X. G. Shao, J. Kang and W. S. Cai, Talanta, 2010, 82,
1017−1021.
27 J. Kang, W. S. Cai and X. G. Shao, Talanta, 2011, 85, 420−424.
28 R. F. Shan, Y. Zhao, M. L. Fan, X. W. Liu, W. S. Cai and X. G.
Shao, Talanta, 2015, 131, 170−174.
29 X. Y. Cui, X. W. Liu, X. M. Yu, W. S. Cai and X. G. Shao, Anal.
Chim. Acta, 2017, 957, 47−54.
30 L. H. Qi, W. S. Cai and X. G. Shao, Acta Chim. Sin., 2016, 74,
172−178.
31 X. Y. Cui, J. Zhang, W. S. Cai and X. G. Shao, Chemom. Intell.
Lab. Syst., 2017, 170, 109−117.
32 X. W. Zhu, X. Y. Cui, W. S. Cai and X. G. Shao, Acta. Chim. Sin.,
2018, 76, 298−302.
33 D. Cheng, W. S. Cai and X. G. Shao, Appl. Spectrosc., 2018, 72,
1354−1361.
34 M. L. Fan, W. S. Cai and X. G. Shao, Appl. Spectrosc., 2017,
71, 472−479.
35 X. W. Liu, X. Y. Cui, X. M. Yu, W. S. Cai and X. G. Shao, Chin.
Chem. Lett., 2017, 28, 1447−1452.
36 X. Y. Cui, W. S. Cai and X. G. Shao, RSC Adv., 2016, 6,
105729−105736.
37 L. Ma, X. Y. Cui, W. S. Cai and X. G. Shao,
Phys.Chem.Chem.Phys., 2018, 20, 20132−20140.
38 E. Karjalainen, V. Aseyev and H. Tenhu, Macromolecules,
2014, 47, 2103−2111.
39 J. Chiefari, Y. K. Chong, F. Ercole, J. Krstina, J. Jeffery, T. P. T.
Le, R. T. A. Mayadunne, G. F. Meijs, C. L. Moad, G. Moad, E.
Rizzardo and S. H. Thang, Macromolecules, 1998, 31,
5559−5562.
40 G. Moad, E. Rizzardo and S. H. Thang, Polymer, 2008, 49,
1079−1131.
41 X. G. Shao, A. K. M. Leung and F. T. Chau, Acc. Chem. Res.,
2003, 36, 276−283.
42 X. G. Shao and W. S. Cai, Rev. Anal. Chem., 1998, 17,
235−285.
43 X. G. Shao, C. Y. Pang and Q. D. Su, Fresenius J. Anal. Chem.,
2000, 367, 525−529.
44 X. G. Shao and C. X. Ma, Chemom. Intell. Lab. Syst., 2003, 69,
157−165.
45 S. Wold, P. Geladi, K. Esbensen and J. Ӧhman, J. Chemom.,
1987, 1, 41−56.
46 G. W. Robinson, C. H. Cho and J. Urquidi, J. Chem. Phys.,
1999, 111, 698−702.
47 C. William, C. MaCabe and F. F. Harvey, J. Phys. Chem., 1970,
74, 2990-2998.
48 S. L. Vaughan, J. M. Allan and I. Q. Terence, J. Phys. Chem. A,
2001, 105, 8916-8921.
49 Š. Šašić, V. H. Segtnan and Y. Ozaki, J. Phys. Chem. A, 2002,
106, 760−766.

10 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 20xx

Please do not adjust margins

 Page 11 of 11 Physical Chemistry Chemical Physics

View Article Online

DOI: 10.1039/C8CP07153E

Graphical abstracts

Physical Chemistry Chemical Physics Accepted Manuscript

Published on 15 February 2019. Downloaded by Tulane University on 2/15/2019 5:43:26 PM.

Water with two hydrogen bonds plays an important role during the phase transition of LCST
polymer.