Water dispersible pH-responsive chitosan nanogels modified with biocompatible crosslinking-agents

Article abstract of DOI:10.1016/j.polymer.2012.05.027

Chitosan nanoparticles were obtained by crosslinking with two biocompatible dicarboxylic acids: polyethylene glycol dicarboxylic acid and tartaric acid. The water-in-oil (W/O) microemulsion method yielded particle sizes around 10-15 nm in the dried state (TEM) and 200-700 nm in the swollen state (QELS) from commercial chitosan. All the synthesized nanogels showed improved water solubility and most of them were stable at physiological pH. QELS studies revealed the influence of the hydrophilic character and flexibility of the crosslinker on the swelling behaviour of the nanogels. The nanoparticles showed a pH-sensitive volume transition that was consistent with the pKa of chitosan. The collected zeta potential data corroborated the electrostatic repulsion mechanism responsible for the pH-responsive behaviour.

Full text of DOI:10.1016/j.polymer.2012.05.027

Polymer 53 (2012) 3107e3116
Contents lists available at SciVerse ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
“Water dispersible pH-responsive chitosan nanogels modified with biocompatible
crosslinking-agents”
*
Maite Arteche Pujana, Leyre Pérez-Álvarez, L. Carlos Cesteros Iturbe , Issa Katime
Grupo de Nuevos Materiales y Espectroscopia Supramolecular, Departamento de Química Física, Facultad de Ciencia y Tecnología, Universidad del País Vasco (UPV/EHU),
644 Bilbao, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
Chitosan nanoparticles were obtained by crosslinking with two biocompatible dicarboxylic acids:
polyethylene glycol dicarboxylic acid and tartaric acid. The water-in-oil (W/O) microemulsion method
yielded particle sizes around 10e15 nm in the dried state (TEM) and 200e700 nm in the swollen state
(QELS) from commercial chitosan. All the synthesized nanogels showed improved water solubility and
most of them were stable at physiological pH. QELS studies revealed the influence of the hydrophilic
character and flexibility of the crosslinker on the swelling behaviour of the nanogels. The nanoparticles
showed a pH-sensitive volume transition that was consistent with the pKa of chitosan. The collected zeta
potential data corroborated the electrostatic repulsion mechanism responsible for the pH-responsive
behaviour.
Received 7 December 2011
Received in revised form
24 April 2012
Accepted 16 May 2012
Available online 23 May 2012
Keywords:
Chitosan
Nanogel
Ó 2012 Elsevier Ltd. All rights reserved.
pH-responsive
1. Introduction
formed by electrostatic interaction between positively charged
chitosan and negatively charged mucosal surfaces. Thus, it can be
The well known biological and physicochemical properties of
chitosan lead to the identification of this biopolymer as one of the
used to enhance drug penetration across the mucosa. Conse-
quently, in recent years chitosan has also been extensively inves-
tigated as a carrier for mucosal drug delivery [9], as an absorption
enhancer [10], as suitable material in ophthalmology and as carrier
for gene delivery [11].
most promising biomaterials. Chitosan, poly[b-(1-4)-linked-2-
amino-2-deoxy- -glucose], is a natural, linear and cationic poly-
D
aminosaccharide obtained by alkaline partial deacetylation of
chitin. This polymer can be degraded by general lysozymes in the
body and subsequently excreted as non-toxic, nonimmunogenic,
and noncarcinogenic degradation products [1,2].
However, despite all its favorable features, the use of chitosan in
biomedical field is limited by its poor solubility in physiological
media [2]. The development of water soluble chitosan is a prereq-
uisite for a successful biomedical use of chitosan and its derivatives.
The water solubility of chitosan has been improved by different
methods such as depolymerization [12], quaternization of the
amino group [13], N-carboxymethylation [14], and PEGylation [15].
The presence of reactive hydroxyl and amino groups enhances
the easy modification of chitosan to create new biofunctional
materials. Reactions involving the eNH2 groups at the C-2 position
are the most frequently employed [16]. In particular, the possibility
to modify chitosan by crosslinking has attracted attention as a new
and exciting way to develop sophisticated biocompatible and
biodegradable gels [17]. Chitosan gels have been prepared as
macroscopic hydrogels [18], films [19], microspheres [20] or
nanoparticles [21].
In addition to biodegradability, chitosan exhibits many valuable
characteristics such as low toxicity and biocompatibility. For these
exceptional features, as well as for being an inexpensive material,
chitosan has attracted interest in various fields including the food
industry, cosmetics, water treatment, agriculture and more recently
in biomedicine. Researches in the latter field have shown that
chitosan exhibits many interesting properties such as, immune
stimulating properties [3] suppressing tumor growth [4],
promoting resistance to infections by microorganisms [5], and
enhancing both humoral and cell-mediated immune responses [6].
Moreover, in contrast to many other natural polymers, chitosan
has
a unique cationic nature [7] and is mucoadhesive [8].
Mucoadhesive properties are due to molecular attractive forces
Chitosan nanoparticles are typically obtained by physical or
covalent crosslinking. Physical crosslinking is based on ionic gela-
tion method. Sodium tripolyphosphate [22] is the most employed
ionic crosslinker. However, the produced particles are not water
* Corresponding author. Tel.: þ34 946015966.
E-mail address: c.cesteros@ehu.es (L.C. Cesteros Iturbe).
0032-3861/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.polymer.2012.05.027

3108
M.A. Pujana et al. / Polymer 53 (2012) 3107e3116
dispersible at physiological pH. Since these particles are formed by
electrostatic interactions, changes in the pH of the medium could
disrupt its stability [23,24]. Consequently, covalent crosslinking is
more interesting than the ionic gelation method. In covalent
crosslinking reaction, crosslinkers are molecules with at least two
reactive functional groups that allow the formation of bridges
between chitosan chains. The most common crosslinkers of chito-
san are aldehydes [25], epoxides [26], and other agents [27,28].
Some of these crosslinkers, such as glutaraldehyde, are toxic [29].
Thus, using non-toxic alternative crosslinker is desirable to prepare
chitosan crosslinked networks for biomedical uses.
crosslinking ratio and crosslinker nature on several important
physicochemical properties of the nanogels, such as, particle size,
solubility, swelling properties, surface charge and pH-sensitivity
are evaluated. This study revealed significant dependences that
clarify and provide helpful support in the knowledge of the
swelling behaviour of this promising biomaterial for biomedical
applications.
2. Experimental
2.1. Materials
Poly(ethylene glycol) (PEG) is one of the most popular polymers
for the chemical modification of biomaterials. PEG shows excep-
tional physicochemical and biological properties, such as high
solubility in water, biocompatibility, and ease of chemical modifi-
cation. Thus, PEG has been extensively used for surface modification
of chitosan improving the solubility of the modified polymers [30].
Several studies have been published on chitosan covalently
crosslinked with PEG [31,32]. Bodnar et al. [33] prepared cross-
linked chitosan nanoparticles by the direct amidation reaction of
the activated carboxylic groups of PEGdiacid and amino groups of
chitosan. These authors also employed natural di- and tricarboxylic
acids with short chains, including tartaric acid, for the intra-
molecular crosslinking of chitosan [34]. These systems were stable
in aqueous medium at pH 6.5. However, no studies were performed
at physiological pH. In these investigations the particle size
measured by TEM was ranged from 60 to 280 nm [34]. The particle
size was reduced to 4e24 nm by previous degradation of the
precursor chitosan [33].
Aldrich low molecular weight chitosan was purified by a previ-
ously described method [42]. The deacetylation degree of chitosan
measured by NMR was 79% which is in good agreement with the
value reported by the supplier (75e85%). The viscosity average
molecular weight of chitosan measured by an Ubbelohde capillary
viscometer (HAc 0.3 M/NaAc 0.2 M, 25 ꢀC) [43] was 66,000 g/mol.
Poly(ethylene glycol)bis(carboxymethyl)ether, Mn ¼ 600, (PEGdia-
cid), tartaric acid, triton X-100, N-hydroxysuccinimide (NHS), 1-
ethyl-3-(3-dimethylaminopropyl carbodiimide hydrochloride)
(EDC), Folic acid and hexanol (for synthesis, 98%) were purchased
from SigmaeAldrich. Cyclohexane (for synthesis, 98%), acetic acid
(for analysis, 99.8%) and ethanol (for analysis, 96%) were supplied
from Panreac.
2.2. Synthesis of nanoparticles
The crosslinked nanoparticles were obtained by mixing sepa-
rately prepared chitosan and crosslinker agent microemulsions.
First, a chitosan solution was prepared by dissolving 1.0 g of chi-
tosan powder in 100 ml of 1% acetic acid. Cyclohexane, n-hexanol
and chitosan solution were mixed in a flask in a fixed ratio of
2.75:1:1 (v/v). The chitosan microemulsion was formed by adding
Triton X-100 drop by drop into the mixture under vigorous stirring
until the mixture became transparent. The W/O microemulsion of
crosslinker (PEGdiacid or tartaric acid) was prepared following the
same procedure but the diacid was previously activated with N-(3-
Dimethylaminopropyl)-Nʹ-ethylcarbodiimide hydrochloride (EDC)
and N-hydroxysuccinimide (NHS). In order to ensure the efficient
activation, the crosslinker and NHS were first mixed in water for
15 min and then EDC was gradually added. The ratio between EDC,
NHS and COOH was 5:2:1 (v/v) and the pH adjusted to 5.4 by the
addition of 2 M NaOH solution. The mixture was stirred at room
temperature during 4 h.
On the other hand, reverse (water-in-oil, W/O) microemulsion
has emerged as an effective way to prepare chitosan covalently
crosslinked nanoparticles. Microemulsion is
a
transparent,
isotropic and thermodynamically stable medium [35]. Colloidal
aqueous droplets of chitosan acidic solution plus crosslinker are
dispersed in a continuous oil phase and stabilized by surfactant
molecules at the water/oil interface [36]. These colloidal sized
droplets act as nanoreactors where the shape and size distribution
of particles are controlled. Several works have focused on cova-
lently crosslinked chitosan prepared by this method. However,
most of these researches have employed glutaraldehyde as cross-
linker agent [37,38]. Tallury et al. [39] reported a preliminary study
of the synthesis of chitosan nanoparticles crosslinked with tartaric
acid by water-in-oil (W/O) microemulsion instead of the direct
crosslinking reaction in water. This method led to highly mono-
dispersed nanogels with a particle size of 26 nm in the dried state.
Tartaric acidechitosan nanogels prepared by microemulsion
showed high water solubility. Nevertheless, the effect of the
external pH was not considered in the investigation. Instead of
microemulsion medium more complex systems based on block
copolymers have been developed as nanoreactors for chitosan
crosslinking in order to avoid organic solvents [40].
Another exploited property of chitosan is its weak base nature.
Its pKa is 6.3e7 [41]; depending on its molecular weight and
deacetylation degree. The glucosamine units are protonated at
pHw6.5 and chitosan behaves as a soluble cationic polyelectrolyte
in dilute acidic solutions. This fact is the origin of the pH-responsive
swelling behaviour of covalently crosslinked chitosan. At acidic pH
the accumulation of opposite electrostatic charges leads to the
electrostatic repulsive force that causes the swelling of the
network.
The crosslinking reaction took place during 24 h at room
temperature after the addition of the crosslinker microemulsion
into the chitosan microemulsion. The nanoparticles were isolated
and washed by dispersion in ethanol followed by centrifugation.
Finally, the obtained nanogels were dispersed in acetic acid solu-
tion, ultrafiltered and dried.
2.3. Methods
2.3.1. 1H NMR
The crosslinked chitosan samples were analyzed by NMR
spectroscopy. 1H NMR spectra were recorded on a Bruker Avance
500 MHz instrument. The samples were dispersed in 2% w/w
CD3COOD/D2O.
The purpose of this paper is to gain deeper insight into the
swelling behaviour of chitosan nanoparticles prepared via reverse
microemulsion method by crosslinking with PEGdiacid and tartaric
acid. The nanogels were prepared by a slight modification of the
procedure reported by Tallury et al. [39] The effect of the
2.3.2. FTIR
Infrared spectra of the samples were recorded on a Thermo 520
Fourier transform infrared spectrophotometer equipped with
a Smart Orbit diamond ATR (attenuated total reflection) accessory.

M.A. Pujana et al. / Polymer 53 (2012) 3107e3116
3109
Spectra were taken with a resolution of 4 cmꢁ1 and were averaged
over 64 scans.
Fig. 2 shows FTIR spectra of pure chitosan and its (A)
PEGechitosan nanogels and (B) tartaric acidechitosan nanogels
with different percentages of crosslinking agent. The spectrum of
pure chitosan exhibits an absorption band around 1647 cmꢁ1 cor-
responding to the amide I mode (C]O stretching) of N-acetylglu-
cosamine units. The peak at 1593 cmꢁ1 corresponds to the eNH
stretching of deacetylated N-glucosamine groups. The intensity of
amide I band (1647 cmꢁ1) and amide II band (1555 cmꢁ1) increased
with crosslinking. This is due to the absorption of the new carbonyl
moieties formed by the reaction of chitosan eNH2 groups with the
activated carboxylic groups of the crosslinkers.
2.3.3. QELS
To determine the size distribution of the nanogels, a QELS
spectrometer was used at an angle of 90ꢀ. Intensity correlation
function measurements were carried out using a Brookhaven BI-
9000AT 522-chanel digital correlator equipped with a water-
cooled Argon-ion laser operating at 514.5 nm as a light source.
The dried powder samples were dispersed in doubly distilled water
(1 mg/L) at different pH values. The obtained homogeneous
dispersions were diluted until the concentration was 40 mg/L. All
measurements were made at room temperature.
The average hydrodynamic diameters were calculated by the
NNLS (non-negative least squares) method from a total of 20
measurements of samples from the same batch. The uncertainties
represented the standard deviation of the mean of the replicate
runs.
For samples with high content of crosslinker, the NH2 peak at
1593 cmꢁ1 disappeared, indicating the complete reaction of amino
groups in chitosan, and a new shoulder at 1730 cmꢁ1 appeared. This
new peak could be attributed to ester bond formation between
chitosan and unreacted N-hydroxysuccinimide units. As Fig. 2
shows, the ester band is less intense for nanogels prepared with
PEGdiacid suggesting that this crosslinker is more efficient for
crosslinking with chitosan than tartaric acid. The shorter length of
tartaric acid molecules may restrict the number of chitosan amino
groups available for the crosslinking reaction.
2.3.4. Zeta potential
Electrophoretic mobility measurements were performed with
a Zeta-Sizer IV (Malvern Instruments) equipment. Chitosan parti-
cles were first dispersed in acid water solution (pH ¼ 4), then
samples with different pH values were prepared by the addition of
2 M NaOH solution. The final chitosan concentration was 40 mg/L.
The average values and uncertainties were calculated as has been
explained for QELS measurements.
Although pure chitosan shows bands at 1309 cmꢁ1 and
1430 cmꢁ1 (CeH bending), a remarkable increase in the intensity of
these bands was observed for PEG-modified systems which agrees
with the larger amount of eCH2 units of this crosslinker.
3.1.2. Determination of the compositions of the nanogels
The degree of crosslinking was evaluated by 1H NMR. Fig. 3C
shows a typical 1H NMR spectrum of chitosan. The peak at
2.0e2.1 ppm is due to the three protons of N-acetylglucosamine
(GlcNAc) units, and the peak at 3.1e3.2 ppm corresponds to H-2
proton of glucosamine (GlcN) residues which represent the free
2.3.5. TEM
The morphological characteristics and particle size distribution
of dried nanoparticles were studied with a Philips CM120 trans-
mission electron microscope operating at 120 kW. A drop of
nanoparticles aqueous dispersion was settled on a carbon-coated
TEM grid and it was dried and glow-discharged in a high vacuum
chamber.
amino group content of chitosan.
A clear decrease in the
3.1e3.2 ppm peak intensity was observed in all the crosslinked
samples due to the formation of amide bonds between the amine
groups of chitosan and the activated carboxylic groups of the
crosslinking agents. Moreover, as Fig. 3A shows, the resonance
signal of new formed H-2 proton of amide moiety was observed at
4.0e4.1 ppm in the case of PEGechitosan samples.
2.3.6. UVevis
UVevis spectroscopy was employed for the study of water
solubility of the obtained nanogels. Solubility of synthesized
crosslinked chitosan nanoparticles was measured at different pH
values. Briefly, nanogels were dissolved in 1% HAc solution (1 mg/
mL) and the pH of the solution was adjusted by the addition of 2 N
NaOH solution. The transmittance of the solution at 750 nm was
recorded on a UVevis spectrophotometer (Cintra 303 UVevis
Spectrophotometer).
The non-anomeric protons of chitosan, which are connected to
ring-skeleton in a glycosyl residue, have similar electron densities
and similar chemical shifts. In the spectrum of the linear chitosan,
the signals of the non-anomeric protons overlap and produce
a broad signal which is observed between 3.5 and 4 ppm. In the
case of PEGechitosan samples crosslinking was also observed by
the typical broad signal of alkyl protons of eOeCH2eCH2eO frag-
ments of the incorporated PEG. As can be observed in Fig. 3A, this
contribution overlaps with that of the non-anomeric protons of
chitosan.
3. Results and discussion
3.1. Preparation and characterization of nanoparticles
The 1H NMR spectra of tartaric acidechitosan nanogels exhibi-
ted an isolated peak at 4.5 ppm corresponding to the resonance of
Hc protons of tartaric acid units, HOOeCOeCHcOHeCHcOHeCOOH
(Fig. 3B). The appearance of this resonance indicates the successful
reaction of linear chitosan with tartaric acid.
Chitosan was crosslinked by reverse microemulsion method
with two biocompatible dicarboxylic acids with different chain
length: tartaric acid and PEGdiacid. The carboxylic groups of the
crosslinkers were previously activated with EDC and NHS. A w/o
microemulsion was prepared with the activated diacid without any
purification as has been above described. Separately, chitosan
microemulsion was formed and the amidation reaction leading to
the formation of crosslinked nanogels took place as is schematically
displayed in Fig. 1. Different chitosan/crosslinker ratios were used
to prepare nanosized chitosan particles according to the data
summarized in Table 1.
The assignments and chemical shifts of the 1H NMR signals are
given as follows: chitosan 1H-RMN (D2O/CD3COOD, 500 MHz,
20 ꢀC):
d
¼ 4.85 (1-H of GlcN), 4.75 (1-H of GlcNAc), 3.45e4 (3-H, 4-
H, 5-H, 6-H, 2-H of GlcNAc), 3.21 (2-H of GlcN), 2.08 (HNeCOCH3).
PEGechitosan 1H-RMN (D2O/CD3COOD):
d
¼ 4.85 (1-H of GlcN),
4.75 (1-H of GlcNAc), 3.45e4 (3-H, 4-H, 5-H, 6-H), 3.21 (2-H of
GlcN), 2.08 (HNeCOCH3), 3.5e3.75 (a-CH2, b-CH2 and c-CH2 of
PEGDC), 4.05 (2-H of crosslinked GlcNH). Tartaric acidechitosan
3.1.1. FTIR studies of crosslinked nanoparticles
The structure of the synthesized nanogels was analyzed by FTIR
spectroscopy.
1H-RMN (D2O/CD3COOD):
GlcNAc), 3.45e4 (3-H, 4-H, 5-H, 6-H), 3.21 (2-H of GlcN), 2.08
(HNeCOCH3), 2.8 (NHS), 4.45e4.55 (a’-CH2 of tartaric acid).
d
¼ 4.85 (1-H of GlcN), 4.75 (1-H of

3110
M.A. Pujana et al. / Polymer 53 (2012) 3107e3116
1
O
O
O
C
O
C
O
C
O
C
EDC, NHS
pH=5.4, 4h
HO
R
OH
N
O
R
O
N
O
O
R = -CH2-(OCH2CH2)-O-CH2 for PEGdiacid
and -CHOH-CHOH- for tartaric acid
2 A
2 B
6
HOH C
H N
H N
2
2
2
4
O
HO
5
O
HO
2
O
O
3
O
HO
O
1
NH
2
HOH C
HOH C
2
2
n
+
O
O
O
C
O
CHa2-CH2 -O-CH2-CH2 -O-CHa2-C
8
N
O
O
N
O
O
6
HOH C
2
H N
H N
2
2
4
O
HO
5
O
HO
2
O
O
3
1
O
HO
O
NH
CO
HOH C
HOH C
2
2
CH
O
2
a
b
n
8
CH
2
2
c
CH
O
CH
CH
2
2
c
b
O
CH
CO
2
a
CH OH
2
CH OH
2
HN
O
OH
O
O
O
OH
O
O
OH
NH
NH
2
2
CH OH
2
Fig. 1. Activation of the diacids (1) and crosslinking reaction of linear chitosan with PEGdiacid (2A), and tartaric acid (2B).
Among various bands of the 1H NMR spectrum of chitosan, the
one corresponding to the methyl protons at 2.0e2.1 ppm, possess
the highest resolution [44] and was used as reference for the
quantitative determination of the composition of the nanogels. This
was done by analyzing the decrease of the peak located at 3.2 ppm
in the 1H NMR spectra assigned to H-2 protons of glucosamine
moieties. The compositions of the nanogels determined by 1H NMR
relating to the glucosamine residues of pure chitosan are shown in
Table 2. The overall degree of crosslinking of chitosan was esti-
mated assuming that no intramolecular crosslinkings are formed
([(Total modified eNH2) ꢁ (Activated eNHS pendant chains)]/2).
The differences between the stoichiometric crosslinker feed and
the modification rate of the chitosan reveals the typical incomplete
activation of the carboxylic acid moieties [45].
Table 1
Feed compositions in the preparation of crosslinked chitosan nanoparticles.
Chitosan (mg) PEG (mg) Tart (mg) EDC (mg) NHS (mg)
Chi-PEG15.0
Chi-PEG 25.0
Chi-PEG 30.0
Chi-PEG 37.5
Chi-PEG 45.0
Chi-PEG 75.0
Chi-Htart 15.0 200
Chi-Htart 25.0 200
Chi-Htart 30.0 200
Chi-Htart 37.5 200
Chi-Htart 45.0 200
Chi-Htart 75.0 200
200
200
200
200
200
200
57.5
96.0
115.0
144.0
172.5
287.5
0.0
0.0
0.0
0.0
0.0
0.0
150
250
300
375
450
750
150
250
300
375
450
750
44.0
73.5
88.0
110.0
132.0
220.0
44.0
0.0
14.4
24.0
28.8
36.0
43.0
72.0
0.0
0.0
0.0
0.0
73.5
88.0
110.0
132.0
220.0
Samples with incomplete crosslinking showed a signal at
2.8 ppm corresponding to the unreacted eNHS groups of the
0.0

M.A. Pujana et al. / Polymer 53 (2012) 3107e3116
3111
0,7
0,6
0,5
0,4
0,3
0,2
0,1
0
A 0,6
B
0,5
75%
75%
37.5%
0,4
0,3
0,2
0,1
0
37.5%
30%
15%
0%
0%
1800 1750 1700 1650 1600 1550 1500 1450 1400
Wavenumber (cm-1)
1800 1750 1700 1650 1600 1550 1500 1450 1400
Wavenumber (cm-1)
Fig. 2. FTIR spectra of chitosan nanogels modified with PEG (A) and tartaric acid (B) for different contents of crosslinker in the feed.
activated crosslinker. A typical spectrum of this incompletely
crosslinked chitosan is shown in Fig. 3B. The free amine content and
the overall degree of modification of the nanogels can be deter-
mined by integrating this signal. The presence of eNHS residues
was observed in all the samples crosslinked with tartaric acid and
only for highly crosslinked PEG nanogels, as is shown in Table 2.
This fact is consistent with the FTIR analysis and again reveals the
influence of the chain length of the crosslinker on the crosslinking
efficiency.
in the swelling of the nanogels. Finally, the third factor controlling
the swelling degree of nanogels is the negative effect of the elastic
contribution of crosslinked polymer chains which reduces the
swelling capacity of nanogels as crosslinking increases [48].
By increasing the degree of crosslinking, the number of amine
group decreases and this fact adversely affects the ionic contribu-
tion to the swelling. Additionally, the network becomes more
compact and the swelling is limited. However, simultaneously the
hydrophilicity of the system increases as a result of the increase of
PEG or tartaric acid content. The introduction of pendant groups on
amine primaries of chitosan leads to a decrease in the intermo-
lecular interactions attributed to the destruction of the rigid crys-
talline structure of chitosan enhancing the hydrophilicity [49].
On the other hand, there are some additional factors that could
affect the swelling behaviour of the nanogel but are difficult to
quantify. For instance, a possible increase of chitosan content per
nanoparticle for high crosslinker feeds, the contributions of intra-
molecular crosslinks and the presence of mono tethered crosslinker
molecules.
3.2. Size and morphology of nanoparticles
The chitosan nanoparticles characterized by TEM showed
a more or less spherical geometry, favorable solubility and no
severe agglomeration of the particles. As can be seen in Fig. 4, the
nanoparticles showed mean diameters ranging from 10 to 50 nm
and a narrow size distribution.
The particle size of swollen nanogels was determined by qua-
sielastic light scattering (QELS) measurements, showing a typical
polydispersity index of 0.15. Fig. 5 shows a representative size
distribution profile. Since chitosan is an adhesive polymer in
aqueous solution, these nanoparticles tend to form aggregates. So,
for interacting particles, the average particle size is always found
higher than the actual size of the particles [46]. The progressive
dilution reduces the inter-particle interaction and therefore, the
formation of large aggregated particles. In order to reduce the
presence of aggregates and, as have been proposed by other authors
[47], we measured the sizes of the dispersed nanogels at highly
diluted conditions.
The measured hydrodynamic diameters of the nanoparticles
reveal a large swelling capacity of the gels that depends on the
degree of crosslinking. The nanoparticles crosslinked with PEG-
diacid showed a hydrodynamic diameter 5e35 times higher than
the observed by TEM for the dried samples. In the case of the gels
prepared with tartaric acid this factor varied between 9 and 20.
The swelling properties of the nanogels depend on three factors:
the osmotic pressure, the electrostatic contribution and the elastic
force of the network. On the one hand, the swelling is dependent on
the osmotic contribution due to polymeresolvent interaction of the
network. The modification with PEGdiacid or tartaric acid results in
a stronger interaction between network and water increasing the
osmotic factor. The swelling of polymeric networks also depends on
the electrostatic factor. The free amino groups of chitosan are
protonated at acidic pH values (pKa z 6e6.5). The charge repul-
sions between neighbouring protonated groups cause an increase
As shown in Fig. 6 an increase in the degree of crosslinking
results in a higher particle size for both systems. This trend is
remarkably more pronounced for PEG-crosslinked nanogels than
for the nanoparticles crosslinked with tartaric acid. In case of PEG
nanogels the remarkable increase in the hydrodynamic diameters
for the nanoparticles with higher crosslinker contents can be
attributed to the flexibility and high hydrophilicity of this polymer
(
c
¼ 0.41), probably the presence of mono tethered PEG molecules
(detected in NMR spectra) also plays an important role. The lower
swelling capacity of tartaric acidechitosan nanogels may be the
result of the shorter length and higher rigidity of the crosslinker,
which restrict the previously mentioned contributions.
Previous studies of these systems for samples which had been
directly crosslinked in aqueous medium did not find a strong
dependence between swelling and stoichiometric ratio of tartaric
acid or PEG [33,34]. The microemulsion method employed in this
work leads to highly monodisperse nanoparticles and more
homogeneous compositions that could explain the good correlation
between the crosslinker content and the hydrodynamic diameters
of the nanogels.
3.3. Water solubility of crosslinked chitosan
Chitosan is only soluble in aqueous acidic solutions, signifi-
cantly restricting its application in biological fields. Therefore, the
development of chitosan nanoparticulated systems with

3112
M.A. Pujana et al. / Polymer 53 (2012) 3107e3116
Fig. 3. 1H NMR spectra of (A) PEGechitosan nanogels (25% mol. PEG), (B) tartaric acidechitosan nanogels (25% mol. tartaric acid) and (C) unmodified chitosan.

M.A. Pujana et al. / Polymer 53 (2012) 3107e3116
3113
Table 2
of aqueous dispersion and so, the solubility of nanogels was
improved.
The pH value when the transmittance at 750 nm reached 50%,
pH50, was employed as a parameter to express the water solubility
of the synthesized chitosan nanogels. The pH50 values obtained for
different chitosan nanogels are shown in Fig. 8.
Composition and overall degree of crosslinking of the chitosan nanogels determined
by 1H NMR.
Stoichiometric
crosslinker feed
(mol. %)
eNH2
modified
(mol. %)
eNHS
(mol. %)
Overall degree
of crosslinking
(mol. %)
PEG
15.0
25.0
30.0
37.5
45.0
75.0
15.0
25.0
30.0
37.5
45.0
75.0
10
18
18
32
14
47
15
21
26
30
39
45
0
0
0
0
0
6
4
4
5
5
9
9
16
7
21
6
Solubility of the chitosan nanoparticles is related to the hydro-
philic character of the crosslinking agent and the amount of free
amino groups in the chitosan chains. As Figs. 7 and 8 shows, the
increase in the solubility was more remarkable for PEGechitosan
nanogels than for tartaric acid ones. The hydrophilic nature of
PEG enhances chitosan solubility at physiological pH and leads to
fully dispersible nanogels even at basic pH values. The solubility of
the nanogels was greater as the crosslinker content increased,
except for highly crosslinked tartaric acidechitosan nanogels. In the
case of tartaric acid nanogels with 75% degree of crosslinking, the
increase in the crosslinker content enhanced the opalescence of the
solutions. As has been previously pointed out in the study of
hydrodynamic diameters of the nanogels, this result may be related
to the more compact structure of the networks prepared with
tartaric acid. In fact, chitosan nanoparticles crosslinked with tar-
taric acid at a stoichiometric ratio of 75% precipitated in aqueous
media, despite its hydrophilic character.
TART
9
11
11
15
17
8
9
12
enhanced water solubility at physiological pH is an important
goal in this research field. The water solubility of the obtained
PEGechitosan and tartaric acidechitosan crosslinked nano-
particles was assayed as a function of pH and compared with
pure chitosan.
Solutions changed from clear to opaque by increasing the pH of
the aqueous medium and were colloidally stable at room temper-
ature. Fig. 7 shows the transmittance of aqueous solutions of pure
chitosan and chitosan-nanogels crosslinked with PEG and tartaric
acid as a function of pH. Decreasing the pH, as the protonation of
free amino groups of chitosan chains takes place, the transmittance
3.4. pH-responsive swelling properties of nanogels
The swelling behaviour of chitosan nanogel particles was
studied by QELS as function of the external pH. Fig. 9 shows the
Fig. 4. TEM micrographs and particle size distribution of A) PEG 37.5% and B) tartaric acid 37.5% chitosan nanogels.

3114
M.A. Pujana et al. / Polymer 53 (2012) 3107e3116
120
100
80
60
40
20
0
A
B 120
100
80
60
40
20
0
345 349 353 357 360 364 368 372
D (nm)
p
340 349 353 357 360 364 368 375
D (nm)
p
Fig. 5. QELS particle size distribution of (A) PEGechitosan nanogels and (B) tartaric acidechitosan nanogels with a crosslinker ratio of 37.5% at pH ¼ 7.0.
hydrodynamic diameters of PEGechitosan (A) and tartaric
acidechitosan (B) nanogels with the same degree of crosslinking
(37.5%) measured at different pHs. It can be seen that the average
size increases for both systems when the pH decreases. The
primary amine groups are protonated in acidic media, and hydro-
dynamic diameters increase because of the repulsive interaction.
Chitosan crosslinked nanogels showed a pH-responsive volume
transition which was consistent with the pKa value of linear chi-
tosan chains [50]. PEGechitosan nanogels exhibited a large volume
transition (w90%) when compared to tartaric acid-modified
particles (w15%), due to the more rigid network originated by
this last crosslinker.
(Fig. 10). Although the z-potential characterizes the surface
charge of the particles, it could be assumed that the amino
groups inside the nanogel behave in a similar way to the surface
ones [22].
The influence of the chemical structure of the crosslinkers was
also observed in the study of the zeta potential. As Fig.11 shows, the
plot of zeta potential versus the crosslinker content showed
opposite behaviours for PEG and tartaric acid.
In both systems the evolution of the zeta potential with the
crosslinker content is in consonance with the observed colloidal
stability of high-crosslinked nanoparticles. Thus the progressive
decrease in the charge of the protonated network for tartaric acid
nanogels results in the observed loss of solubility for high-
crosslinked nanoparticles. The opposite behaviour in the zeta
potential leads to the observed increase in the solubility of high-
crosslinked PEG nanoparticles.
The progressive protonation of amino groups at low pH values
was confirmed by measuring the electrokinetic potential (
z-
potential) of the nanogels. The value of the -potential increased
z
from (ꢁ8) e (e10) to 30e45 mV in the range of pH from 10 to 4
The different behaviour of both systems is difficult to explain if
we consider that the mechanism and the degrees of crosslinking
900
800
700
600
500
400
300
200
100
110
100
90
80
70
60
50
40
30
pH50
Chitosan 7.00
Chi-TART 7.35
Chi-PEG 7.50
0
20
40
60
80
100
2
4
6
8
10
12
14
Crosslinker (mol.%)
pH
Fig. 6. Effect of crosslinker content on hydrodynamic diameter determined by QELS for
(C) PEGechitosan and (,) tartaric acidechitosan nanogels dispersed in water at
pH ¼ 4.0.
Fig. 7. Transmittance variation with external pH for pure chitosan (B), PEGeChitosan
(:) and tartaric acideChitosan (-) nanogels with a modification degree of 25.0 mol%.

M.A. Pujana et al. / Polymer 53 (2012) 3107e3116
3115
10
9,5
9
are similar. Experimental data showed evidence that hydrolysis
reaction of mono tethered crosslinker molecules had not occurred.
Taking all above into consideration, the main difference between
both chitosan nanoparticles is the swelling capacity (Fig. 6).
Apparently, for the nanogels prepared with PEG, the progressive
decrease in the charge of the protonated network resulting from
the increase in the degree of crosslinking can be offset by the higher
flexibility and swelling capacity of these nanogels. Thus, both
factors can provide a higher net charge distribution at the particle
surface. For the tartaric acid nanogels the progressive increase of
the rigidity introduced by the crosslinker may acts in the opposite
sense.
Chitosan
Tartaric-Chit
PEG-Chit
8,5
8
The pKa values of synthesized nanogels were calculated from
data shown in Fig. 10 using the following equation [22],
7,5
7
0:4343Fzplateau
pKa ¼ pHz
ꢁ
¼
zplateau=2
2RT
6,5
6
where, pHz plateau/2 is the pH where the maximum value of the
¼
z
zeta potential “zplateau” is reduced by half. T is the temperature and
0
15
25
30
37,5
45
75
F, R, are the Faraday and the gas constant respectively. The pKa
values thus obtained were 5.9 for PEG and 6.2 for tartaric acid,
respectively.
Crosslinking agent (mol. %)
Fig. 8. pH50 values for pure chitosan, and dispersions of tartaric acid and PEGechi-
tosan nanogels.
B 440
A
500
420
400
380
360
340
320
450
400
350
300
250
200
0
2
4
6
8
10
12
14
4
5
6
7
8
9
10
11
12
pH
pH
Fig. 9. Average diameter as function of external pH for (A) PEG and (B) tartaric acidechitosan nanogels (37.5 mol% crosslinker stoichiometric ratio).
50
A
B
30
20
10
0
40
30
20
10
0
-10
-10
3
4
5
6
7
pH
8
9
10
11
0
2
4
6
8
10
12
14
pH
Fig. 10. Zeta potential as function of pH of the medium for (A) PEG and (B) tartaric acid-crosslinked nanogels (37.5 mol%).

3116
M.A. Pujana et al. / Polymer 53 (2012) 3107e3116
[2] Chandy T, Sharma CP. Artif Cells Artif Organs 1990;18(1):1e24.
45
40
35
30
25
[3] Nishimura K, Nishimura S, Nishi N, Numata F, Tone Y, Tokura S, et al. Vaccine
1985;3(5):379e84.
[4] Qi L, Xu Z, Chen M. Eur J Cancer 2007;43(1):184e9.
[5] Van der Lubben IM, Verhoef JC, Borchard G, Junginger HE. Adv Drug Deliv Rev
2001;52(2):139e44.
[6] Zaharoff DA, Rogers CJ, Hance KW, Schlom J, Greiner JW. Vaccine 2007;25(11):
2085e94.
[7] Rinaudo M. Prog Polym Sci 2006;31(7):603e32.
[8] Lehr CM, Bouwstra JA, Schacht EH, Junginger HE. Int J Pharm 1992;78(1e2):
43e8.
[9] Illum L, Farraj NF, Davis SS. Pharm Res 1994;11(8):1186e9.
[10] Alonso MJ, Salve P, Remunan-Lopez C, Vila-Jato JL. Eur Patent; 1998. 0860-
166-A1.
[11] Alonso MJ, Sánchez AJ. Pharm Pharmacol 2003;55(11):1451e63.
[12] Nishimura SI, Kohgo O, Kurita K, Kuzuhara H. Macromolecules 1991;24(17):
4745e8.
[13] Fischer D, Bieber T, Li Y, Elsaesser HP, Kissel T. Pharm Res 1999;16(8):1273e9.
[14] Muzzarelli RAA, Tanfani F, Emanuelli M. Carbohydr Res 1984;126(2):225e31.
[15] Saito H, Wu X, Harris J, Hoffman A. Macromol Rapid Commun 1997;18(7):
547e50.
[16] Enescu D, Olteanu CE. Chem Eng Commun 2008;195(10):1269e91.
[17] Berger J, Reist M, Mayer JM, Felt O, Peppas NA, Gurny R. Eur J Pharm Biopharm
2004;57(1):19e34.
10
20
30
40
50
60
70
80
[18] Kristl J, Smid-Korbar J, Strue E, Schara M, Rupprecht H. Int J Pharm 1993;
99(1):13e9.
Cross-linker (mol. %)
[19] Park HJ, Jung ST, Song JJ, Kang SG, Vergano PJ, Testin RF. Chitin Chitosan Res
1999;5(1):19e26.
[20] Wang LY, Ma GH, Su ZG. J Control Rel 2005;106(1e2):62e75.
[21] Agnihotri SA, Aminabhavi TM. Drug Dev Ind Pharm 2007;33(11):1254e62.
[22] Calvo P, Remuñán-López C, Vila-Jato JL, Alonso MJ. J Appl Polym Sci 1997;
63(1):25e132.
Fig. 11. Zeta potential for (,) tartaric acidechitosan and (C) PEGechitosan nanogels
at pH ¼ 4.0 as function of crosslinker content.
4. Conclusions
[23] López-León T, Carvalho ELS, Seijo B, Ortega-Vinuesa JL, Bastos-González D.
J Colloid Interface Sci 2005;283(2):344e51.
Chitosan was successfully crosslinked by amidation reaction
using biocompatible dicarboxylic acids (PEGdiacid and tartaric
acid) by w/o microemulsion to form ultrafine nanoparticles
(10e15 nm TEM). PEGdiacid showed to be more efficient for chi-
tosan crosslinking than tartaric acid. Water solubility of the
samples was improved in comparison with linear chitosan and
furthermore, in most of cases the new systems were stable at
physiological pH.
The synthesized nanogels showed a pH-responsive swelling
behaviour that corresponds to the pKa of the non-crosslinked
chitosan. The particle size of colloid dispersion varied from 250 to
450 nm by decreasing the pH from 8 to 4. The pKa values of
resulting nanoparticles estimated by Zeta potential measurements
were around 5.9e6.2. The swelling capacity and surface charge
clearly depends on the hydrophilic character and flexibility of the
crosslinker.
[24] Zhang H, Oh M, Allen C, Kumacheva E. Biomacromolecules 2004;5(6):2461e8.
[25] Khalid MN, Ho L, Agnely JL, Grossiord JL, Couarraze G. STP Pharm Sci 1999;
9(4):359e64.
[26] Yang ZK, Yuan Y. J Appl Polym Sci 2001;82(8):1838e43.
[27] Crescenzi V, Paradossi G, Desideri P, Dentini M, Cavalieri F, Amici E, et al.
Polym Gels Netw 1997;5(3):225e39.
[28] Mi FL, Tan YC, Liang HF, Sung HW. Biomaterials 2002;23(1):181e91.
[29] Beauchamp O, St Clair MB, Fennell TR, Clarke DO, Morgan KT. Crit Rev Toxicol
1992;22(3e4):143e74.
[30] Du J, Hsieh JL. Cellulose 2007;14(6):543e52.
[31] Kulkarni AR, Hukkeri VI, Sung HW, Liang HF. Macromol Biosci 2005;5(10):
925e8.
[32] Kiuchi H, Kai W, Inoue YJ. Appl Polym Sci 2008;107(6):3823e30.
[33] Bodnar M, Hartmann JF, Borbely J. Biomacromol 2006;7(11):3030e6.
[34] Bodnar M, Hartmann JF, Borbely J. Biomacromol 2005;6(5):2521e7.
[35] Chow PY, Gang LM. Adv Polym Sci 2005;175:257e98.
[36] Leong YS, Candau F. J Phys Chem 1982;86(13):2269e71.
[37] Liu G, Shao L, Gea F, Chena J. China Particuol 2007;5(6):384e90.
[38] Banerjee T, Mitra S, Singh AK, Sharma RK, Maitra A. Int J Pharm 2002;
243(1e2):93e105.
[39] Tallury P, Kar S, Bamrungsap S, Huang YF, Tand W, Santra S. Chem Commun
2009;17:2347e9.
[40] Maggi F, Ciccarelli S, Diociaiuti M, Casciardi S, Masci G. Biomacromol 2011;
12(10):3499e507.
The water solubility at physiological pH and the pH-responsive
swelling of the synthesized nanogels make them attractive candi-
dates as biocarriers for a large variety of biomedical applications.
[41] Schulz PC, Rodriguez MS, Del Blanco LF, Pistonesi M, Agullo E. Colloid Polym
Sci 1998;276(12):1159e65.
Acknowledgements
[42] Kasaai MR. Carbohydr Polym 2010;79(4):801e10.
[43] Rinaudo M, Milas M, Dung PL. Int J Biol Macromol 1993;15(5):281e5.
[44] Sánchez A, Sibaja M, Vega-Baudrit J, Madrigal S. Rev Iber Polim 2007;8(5):4.
[45] Olde Damink LHH, Dijkstra PJ, van Luyn MJA, van Wachem PB, Nieuwenhuis P,
Feijen J. Biomaterials 1996;17(8):765e73.
[46] Wu C, Zhou S, Wang W. Biopolymers 1995;35(3):385e92.
[47] Sorlier P, Rochas C, Morfin I, Viton C, Domard A. Biomacromol 2003;4(4):
1034e40.
Financial support for this work was provided by the Basque
Government and the University of the Basque Country (UPV/EHU),
for awarding a fellowship.
References
[48] Flory PJ, Rehner J. J Chem Phys 1943;11:521.
[49] Chan P, Kurisawa M, Chung JE, Yang YY. Biomaterials 2007;28(3):540e9.
[50] Schatz C, Pichot C, Delair T, Viton C, Domard A. Langmuir 2003;19(23):
9896e903.
[1] Kockisch S, Rees GD, Young SA, Tsibouklis J, Smart JD. J Pharm Sci 2003;92(8):
1614e23.