This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
Journal of Membrane Science 346 (2010) 296–301
Contents lists available at ScienceDirect
Journal of Membrane Science
journal homepage: www.elsevier.com/locate/memsci
Quantitative characterization of membrane formation process of
alginate–chitosan microcapsules by GPC
Weiting Yu a , Junzhang Lin a , Xiudong Liu b,∗∗ , Hongguo Xie a , Wei Zhao a , Xiaojun Ma a,∗
a
b
Laboratory of Biomedical Materials Engineering, Dalian Institute of Chemical Physics (DICP), Chinese Academy of Sciences (CAS), 457 Zhongshan Road, Dalian 116023, PR China
College of Environment and Chemical Engineering, Dalian University, Dalian Economic Technological Development Zone, Dalian 116622, PR China
a r t i c l e
i n f o
Article history:
Received 23 June 2009
Received in revised form
23 September 2009
Accepted 23 September 2009
Available online 1 October 2009
Keywords:
Alginate–chitosan microcapsule
Membrane formation process
Quantitative characterization
Gel permeation chromatography
a b s t r a c t
The semi-permeable membrane of alginate–chitosan (AC) microcapsules has been proven important to
control the microcapsule stability and selective substance diffusion rate. Therefore, a novel and operable
methodology based on gel permeation chromatography (GPC) was established for quantitative characterization of the membrane formation process, so as to provide guidance on performance improvement
of AC microcapsules in biomedical applications. Not only the molecular weight (Mw ) and its distribution
of chitosan can be obtained by GPC, but also the area integral of molecular weight peaks can be linearly
correlated to chitosan concentration in certain range. The dynamic membrane formation process was
then obtained by quantitatively analyzing reaction amount of chitosan with time, which showed that
for chitosan molecules with wide Mw distribution, only parts of molecules bind with alginate to form
microcapsule membrane. Moreover, the contribution of chitosan molecules participating in the membrane formation process was also different. These new findings, therefore, are helpful for adjusting and
controlling the membrane formation process and properties of microcapsule membrane.
© 2009 Elsevier B.V. All rights reserved.
1. Introduction
Microencapsulation is the process of enclosing a substance
inside a semi-permeable membrane to form a microcapsule [1].
Due to the protection and selective permeation properties of the
semi-permeable membrane, microcapsules have been widely used
not only in biochemical engineering such as cell immobilization
fermentation, but also in biomedical fields such as drug delivery
and cell transplantation [2]. Although many polymers have been
attempted with various preparation technologies, alginate-based
microcapsules are still the commonly used ones. Alginate–chitosan
(AC) microcapsule, formed by two naturally occurred polysaccharides with opposite charges, has been attractive owing to the
inherent properties such as biodegradability, nontoxicity, and biocompatibility [3]. Numerous papers have been published reporting
studies on membrane permeability [4], mechanical strength [5]
of AC microcapsules [6], as well as applications in drug delivery
systems [7,8] enzyme immobilization [9], cell immobilization fermentation [10], and cell transplantation [11].
Whatever in short-term or long-term in vivo applications, AC
microcapsules have to keep structural stability and exert function
while facing, enduring, or interacting with complicated envi-
∗ Corresponding author. Tel.: +86 411 84379139; fax: +86 411 84379096.
∗∗ Corresponding author. Tel.: +86 411 87402448; fax: +86 411 84379096.
E-mail addresses: [email protected] (X. Liu), [email protected] (X. Ma).
0376-7388/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.memsci.2009.09.049
ronments. Besides alginate gel core, the complex microcapsule
membrane, formed by electrostatic interaction between alginate
and chitosan, has been demonstrated to play an important role
in controlling the microcapsule stability and substance diffusion
rate [12]. The membrane formation process includes at least chitosan molecules diffusion into three-dimensional (3D) alginate gel
network and simultaneous binding between protonated amino
groups of chitosan and carboxyl groups of alginate, which is
affected by complicated factors such as molecular weight (Mw )
and wide Mw distribution, chain flexibility, charge density of both
polysaccharides [13]. Therefore, a better quantitative understanding on microcapsule membrane formation process, especially on
chitosan amount and fraction binding with alginate will be helpful to improve the functional properties of membrane mechanical
strength and selective substance permeability, and then the performance of AC microcapsules in biomedical applications. However,
it is usually difficult to realize effective quantitative analysis of
polysaccharides. Curotto and Aros introduced the use of ninhydrin
reaction for quantitative determination of chitosan [14]. Prochazkova et al. demonstrated that the reaction of chitosans with
ninhydrin was sensitive and reproducible only when a reliable
calibration against a reference of similar composition was available [15]. Moreover, proteins hard to be removed from chitosan
often affect the analysis accuracy. Gaserod et al. reported radioactive label method detecting the interaction between alginate and
chitosan [16]. Although they gave quantitative reaction amount
of radioactive labeled chitosan on alginate gel surface, the special
Author's personal copy
W. Yu et al. / Journal of Membrane Science 346 (2010) 296–301
experimental condition and safety concern make it not acceptable
for routine analysis means.
In this paper, a novel methodology for quantitative characterization of alginate–chitosan membrane formation process was
established based on gel permeation chromatography (GPC), which
has been generally used as a credible and rapid method to characterize the molecular weight and its distribution of polymers [17,18].
AC microcapsules were made by two-stage procedure, that is, calcium alginate gel (CAG) beads were produced firstly, and then the
beads were immersed in chitosan solution to form the membrane
on the surface of the beads. After establishing calibration curves
with standard polymers, chitosan molecular weight and its distribution can be measured by GPC. And then the amount of chitosan
molecules binding with alginate beads can be calculated by the
integral of molecular weight peak area during AC microcapsule
membrane formation process. Therefore, the membrane formation
process can be characterized and further understood by quantitatively analyzing reaction amount of chitosan with time.
2. Materials and methods
297
2.2. Preparation of AC microcapsules
AC microcapsules were prepared according to the method
developed in our lab [19]. Sodium alginate was dissolved in 0.9%
(w/v) NaCl solution to form concentration of 1.5% (w/v). After
being filtered through a 0.22 ␮m membrane filter, the solution was
stored overnight before use to facilitate deaeration. Sodium alginate solution as above was extruded through a 0.4-mm needle
into calcium gelling solution using electrostatic droplet generator
(YD-06, Dalian Institute of Chemical Physics, Chinese Academy of
Sciences, China) to form CAG beads. The beads were rinsed with distilled water and immersed in 0.5% (w/v) chitosan solution to form
alginate–chitosan membrane around the surface.
2.3. Characterization of morphology and size distribution of AC
microcapsules
The morphology of AC microcapsules was observed under convert optical microscope (Olympus CK-40, Olympus Corp. Japan).
The size distribution of AC microcapsules was determined with
laser diffraction particle analyzer (LS100 Q, Beckman-Coulter Corp.,
USA).
2.1. Materials
2.4. Gel permeation chromatography (GPC)
Chitosan samples were chemically modified from raw material
(Yuhuan Ocean Biomaterials Corporation, China) by our laboratory.
The degree of deacetylation (DD) was 96–98%. The average degree
of polymerization (DP) was 130, 400, and 560, respectively, which
was calculated from intrinsic viscosity. Therefore, DP130, DP400
and DP560 were used to denote the different batch of chitosan
samples in the study. Sodium alginate was purchased from the
Chemical Reagent Corp (Shanghai, China), whose viscosity was over
0.02 Pa s when dissolved to form a 1.0% (w/v) aqueous solution at
20 ◦ C. The powder size was less than 200 mesh. Nanoparticulate calcium carbonate was purchased from Taihua Corporation (Zhejiang,
China). Pullulan standards (Shodex Standard P-82) were obtained
from Shoko Co., Ltd., Japan. All other reagents were analytical grade
and used as received.
Pullulan standards (Shodex Standard P-82, Showa Denko
K.K.) were firstly injected into a TSK G4000PWxl column
(7.8 mm × 300 mm, 10 ␮m particle diameter, Tosoh Corporation,
Tokyo, Japan) of GPC equipped with a refractive index detector (Waters, Model 2414, Milford, MA, USA) and an HPLC pump
(Waters, Model 515), which would give a universal calibration
curve showing the relationship between weight-average molecular
weight (Mw ) and the elution time. Then, chitosan samples (DP130,
DP400 and DP560) dissolved in buffer of 0.2 M CH3 COOH/0.1 M
CH3 COONa were eluted through a TSK G4000PWxl column with a
20 ␮L injection volume at a flow rate of 0.6 mL/min with the same
buffer used as eluent. The column and the detector were both set
at the temperature of 30 ◦ C. All data provided by the GPC system
Fig. 1. Photographs of CAG beads (A) and AC microcapsules with chitosan of DP130 (B), DP400 (C), and DP560 (D).
Author's personal copy
298
W. Yu et al. / Journal of Membrane Science 346 (2010) 296–301
were collected and analyzed by the Empower Workstation software package (Waters Corp., MA, USA). Finally, Mw and molecular
weight distribution (Mw /Mn ) of each chitosan were measured by
referring to the universal calibration curve [20,21].
2.5. Quantitative analysis of chitosan during the membrane
formation process by GPC
Firstly, six aliquots of the same chitosan sample were respectively injected into column to determine the stability and
reproducibility of GPC method. Then, chitosan solutions of each
sample (DP130, DP400, and DP560) with the concentration of 1, 2,
3, 4, and 5 mg/mL, respectively, were analyzed by GPC. By correlating the area integral of all peaks of each chromatogram (elution
curve) with the concentration of chitosan solution, a concentrationarea integral curve was obtained to use as the standard curve for
quantitative analysis of each chitosan sample. When CAG beads
were immersed in chitosan solution to form the membrane on the
surface, 20 ␮L supernatant at certain time interval was collected
and analyzed by GPC. Then the concentration of chitosan solution
during the membrane formation process could be calculated by the
concentration-area integral standard curve, from which the reaction amount of chitosan onto alginate beads could be calculated.
3. Results and discussion
Fig. 2. Size distributions of AC microcapsules.
clear outer layer demonstrating the membrane formation. The
membrane thickness is usually 10–100 ␮m and can be adjusted
by control reaction conditions, which would affect the properties of AC microcapsules. Fig. 2 demonstrates the size distribution
of AC microcapsules was narrow, and the average diameter of
monodispersed AC microcapsules was about 300 ␮m, counted and
calculated automatically by the software of the particle analyzer.
3.1. Morphology and size distribution of AC microcapsules
3.2. Standard curves for quantitative analysis of chitosan by GPC
AC microcapsules were made by two-stage procedure. Firstly,
alginate solution was dropped into calcium chloride solution to
produce spherical CAG beads with smooth surface (Fig. 1A). And
then the beads were immersed in chitosan solution (DP130, DP400,
and DP560) to form AC microcapsules (Fig. 1B–D), which showed
Six aliquots of the same chitosan sample were respectively
injected into TSK G4000PWxl column, the six elution curves almost
overlapped with the same peak positions. The relative standard
deviation (RSD) of six measurements was calculated to be 1.7% that
Fig. 3. A representative GPC chromatograms with different concentrations of chitosan (A) and linear concentration-area integral standard curves of chitosan DP130 (B),
DP400 (C), and DP560 (D).
Author's personal copy
W. Yu et al. / Journal of Membrane Science 346 (2010) 296–301
299
Table 1
The correlation of chitosan fractions with molecular weight interval.
Molecular weight interval
Fraction 1
Fraction 2
Fraction 3
Fraction 4
>100 kDa
40–100 kDa
10–40 kDa
<10 kDa
demonstrated GPC method was stable and reproducible. Therefore,
GPC method was feasible to be used for the following quantitative
analysis of chitosan samples.
Each chitosan sample (DP130, DP400, and DP560) was dissolved
and prepared with concentrations in the range of 1–5 mg/mL,
and 20 ␮L solution of each concentration was injected into TSK
G4000PWxl column and analyzed by GPC. The different shape and
peak position of three chromatograms meant the three chitosan
samples had different molecule composition, which suggested different Mw as well as different molecular weight distribution (Fig. 3).
The area integral of all molecular weight peaks of every run was
calculated and correlated to the corresponding concentration, thus
three linear concentration-area integral curves were obtained and
used as the standard curves for each chitosan sample. The standard
curve equation and the reliability of regressions (R2 ) were shown
in Fig. 3.
3.3. Quantitative analysis of chitosan during the membrane
formation process by GPC
As mentioned in Section 2.2, CAG beads were immersed in chitosan solution to form the membrane on the surface. During the
membrane formation process, 20 ␮L supernatant at time interval
0, 10, 20, 30, and 60 min was collected and analyzed by GPC. Fig. 4
shows the decrease of peak height with time, which intuitionistically reflected the dynamic reaction process between chitosan and
alginate molecules. To quantitatively characterize the membrane
formation process, the area integral of peaks at each time interval was carried out, the concentration of chitosan unreacted could
subsequently be obtained according to the standard concentrationarea integral curves. Therefore, the amount of chitosan binding with
alginate beads could be calculated during the membrane formation
process (Fig. 5). It can be found that the total binding amount of chitosan on square centimeter of alginate bead surface increased with
the time of membrane formation process. Moreover, it showed the
lower DP of chitosan sample, the higher binding amount of chitosan
molecules on the surface of alginate beads, which suggested molecular weight of chitosan was an important factor for the membrane
formation as a whole.
To get insight of elution curves in Fig. 4, an interesting phenomenon was noticed that the change of shape and decrease of
peak height were unsymmetrical. For example, the shape and
height of curve section at elution time interval 11–13 min almost
kept unchanged, while the shape and height of curve section at
elution time interval 13–18 min changed and decreased with time.
According to the calibration curve, elution time interval 11–13 min
corresponded to chitosan molecules with Mw over 100 kDa (defined
as Fraction 1), and elution time interval 13–18 min corresponded
to chitosan molecules with Mw under 100 kDa (Fraction 2, Fraction
3 and Fraction 4) (Table 1). The phenomenon in Fig. 4, therefore, meant that for chitosan with wide distribution of molecular
weight, not all chitosan molecules took part in the membrane formation process by binding alginate gel beads. The total binding
amount of chitosan on square centimeter of alginate bead surface
seemed mainly resulting from fractions with smaller molecules,
which were easier to diffuse into 3D gel network and binding with
alginate molecules to form membrane.
The amount of each chitosan fraction binding with alginate
beads during the membrane formation process was recalculated to
further investigate the real contribution of each fraction on mem-
Fig. 4. GPC chromatograms of chitosan samples during the membrane formation
process: (A) DP130, (B) DP400, and (C) DP560.
brane formation. Fig. 6 shows that Fraction 1 was hardly reactive,
a few Fraction 2 was reactive, while Fraction 3 and Fraction 4 were
dominant parts participating in the membrane formation. Moreover, it was also noticed that the bound amount of Fraction 3 was
higher than Fraction 4 for DP400 and DP560 chitosan samples,
which seemed against the above conclusion.
The reason for the contradictive results could be found in Fig. 7.
At the moment of adding CAG beads into chitosan solution, that
is, reaction time 0 min, the initial concentration of each fraction of
DP130, DP400 and DP560 chitosan was quite different. It showed
that the initial concentration of Fraction 3 of three chitosan sam-
Author's personal copy
300
W. Yu et al. / Journal of Membrane Science 346 (2010) 296–301
Fig. 5. The amount of chitosan binding with alginate beads during the membrane
formation process.
ples were higher than that of Fraction 4. For each chitosan sample,
the proportions of each fraction were calculated and shown in
Fig. 8. The proportions of Fraction 3 and Fraction 4 were 50.75% and
38.53% in DP130, 49.23% and 18.95% in DP400, 35.75% and 13.07% in
DP560, respectively. It can be further calculated the initial concentration (or proportion) of Fraction 3 was 2.6-fold higher than that
of Fraction 4 in DP400, and 2.7-fold higher than that of Fraction 4
in DP560. It could result in much higher driving force for Fraction
3 to diffuse into alginate network than Fraction 4, which led the
bound amount of Fraction 3 was higher than Fraction 4 for DP400
and DP560 chitosan. Therefore, the contribution of chitosan fractions participating in the membrane formation process was also
Fig. 7. GPC chromatograms of chitosan samples (DP130, DP400 and DP560) at the
initial state (reaction time is 0 min).
Fig. 8. Initial proportions of four fractions in chitosan DP130 (A), DP400 (B), and
DP560 (C).
different, and the bound amount of chitosan fractions depended on
their initial concentration.
4. Conclusion
Fig. 6. The amount of chitosan fraction binding with alginate beads during the
membrane formation process: (A) DP130, (B) DP400, and (C) DP560.
Alginate–chitosan microcapsules are prepared by two naturally
occurred polysaccharides with high Mw and wide Mw distribution, which makes the membrane formation process complicated.
By use of GPC analysis, quantitative characterization of microcapsule membrane formation process was realized. Firstly, chitosan
Mw and its distribution can be measured by calibration curves
with standard polymers. And then the area integral of molecular
weight peaks can be linearly correlated to chitosan concentration in certain range, which was used to calculate the amount
Author's personal copy
W. Yu et al. / Journal of Membrane Science 346 (2010) 296–301
of chitosan molecules binding with alginate beads during microcapsule membrane formation process. The results showed not all
chitosan molecules participating in membrane formation process,
but only fractions with smaller molecules (in this case, Mw under
100 kDa) mainly contributed to the reaction, which suggested
chitosan Mw was an important factor for the membrane formation. Moreover, the contribution of chitosan fractions participating
in the membrane formation process was also different, and the
bound amount of chitosan fractions also depended on their initial
concentration.
Acknowledgements
The authors thank the National Natural Science Foundation
(People’s Republic of China) for the financial support (Nos.:
20806080, 20876018, and 20736006), the grants of National Basic
Research Development Program of China (973 Program) (No:
2007CB714305) and National Key Technology R&D Program in the
11th Five-year Plan of China (No: 2006BAD27B04) from the Ministry of Science & Technology, China.
References
[1] T.M. Chang, Semipermeable microcapsules, Science 146 (1964) 524–525.
[2] W. Wang, X.D. Liu, Y.B. Xie, H.A. Zhang, W.T. Yu, Y. Xiong, W.Y. Xie, X.J. Ma,
Microencapsulation using natural polysaccharides for drug delivery and cell
implantation, J. Mater. Chem. 16 (2006) 3252–3267.
[3] M. Rinaudo, Chitin and chitosan: properties and applications, Prog. Polym. Sci.
31 (2006) 603–632.
[4] A. Bartkowiak, D. Hunkeler, Alginate–oligochitosan microcapsules. II. Control
of mechanical resistance and permeability of the membrane, Chem. Mater. 12
(2000) 206–212.
[5] A. Bartkowiak, D. Hunkeler, Alginate–oligochitosan microcapsules: a mechanistic study relating membrane and capsule properties to reaction conditions,
Chem. Mater. 11 (1999) 2486–2492.
[6] O. Gaserod, A. Sannes, G. Skjak-Braek, Microcapsules of alginate–chitosan II. A
study of capsule stability and permeability, Biomaterials 20 (1999) 773–783.
301
[7] M. George, T.E. Abraham, Polyionic hydrocolloids for the intestinal delivery of
protein drugs: alginate and chitosan—a review, Control. Rel. 114 (2006) 1–14.
[8] M.S. Crcarevska, M.G. Dodov, K. Goracinova, Chitosan coated Ca–alginate
microparticles loaded with budesonide for delivery to the inflamed colonic
mucosa, Eur. J. Pharm. Biopharm. 68 (2008) 565–578.
[9] E. Taqieddin, M. Amiji, Enzyme immobilization in novel alginate–chitosan
core–shell microcapsules, Biomaterials 25 (2004) 1937–1945.
[10] S. Graff, S. Hussain, J.C. Chaumeil, C. Charrueau, Increased intestinal delivery of
viable Saccharomyces boulardii by encapsulation in microspheres, Pharm. Res.
25 (2008) 1290–1296.
[11] T. Haque, H. Chen, W. Ouyang, C. Martoni, B. Lawuyi, A.M. Urbanska, S. Prakash,
In vitro study of alginate–chitosan microcapsules: an alternative to liver
cell transplants for the treatment of liver failure, Biotechnol. Lett. 27 (2005)
317–322.
[12] H.V. Saether, H.K. Holme, G. Maurstald, O. Smidsrod, B.T. Stokke, Polyelectrolyte
complex formation using alginate and chitosan, Carbohydr. Polym. 74 (2008)
813–821.
[13] J.Z. Knaul, M.R. Kasaai, V.T. Bui, K.A.M. Creber, Characterization of deacetylated chitosan and chitosan molecular weight review, Can. J. Chem. 76 (1998)
1699–1706.
[14] E. Curotto, F. Aros, Quantitative determination of chitosan and the percentage
of free amino groups, Anal. Biochem. 211 (1993) 240–241.
[15] S. Prochazkova, K.M. Varum, K. Ostgaard, Quantitative determination of chitosans by ninhydrin, Carbohydr. Polym. 38 (1999) 115–122.
[16] O. Gaserod, O. Smidsrod, G. Skjak-Braek, Microcapsules of alginate–chitosan I.
A quantitative study of the interaction between alginate and chitosan, Biomaterials 19 (1998) 1815–1825.
[17] M.L. Tsaih, R.H. Chen, Molecular weight determination of 83% degree of
deacetylation chitosan with non-Gaussian and wide range distribution by highperformance size exclusion chromatography and capillary viscometry, Appl.
Polym. Sci. 71 (1999) 1905–1913.
[18] J. Brugnerotto, J. Desbrieres, G. Roberts, M. Rinaudo, Characterization of chitosan by steric exclusion chromatography, Polymer 42 (2001) 9921–9927.
[19] X.D. Liu, W.Y. Yu, Y. Zhang, W.M. Xue, W.T. Yu, Y. Xiong, X.J. Ma, Y. Chen, Q. Yuan,
Characterization of structure and diffusion behaviour of Ca–alginate beads prepared with external or internal calcium sources, J. Microencapsul. 18 (2002)
775–782.
[20] D.W. Ren, H.F. Yi, H.A. Zhang, W.Y. Xie, W. Wang, X.J. Ma, A preliminary study
on fabrication of nanoscale fibrous chitosan membranes in situ by biospecific
degradation, J. Membr. Sci. 280 (2006) 99–107.
[21] M.R. Kasaai, Calculation of Mark–Houwink–Sakurada (MHS) equation viscometric constants for chitosan in any solvent-temperature system using
experimental reported viscometric constants data, Carbohydr. Polym. 68
(2007) 477–488.