J . h.1 ICI<OI;NCA 1’s UI A’I’I0 N .

’Taylor6rFrancis
[Link]-Dcceniber, 2003, VOL. 20, NO. 6, 791 -797 ‘0 healthsciences

Chitosan microparticle preparation for controlled drug

release by response surface methodology

J . A. KO?, H. J. PARK?$”, 1’. S. PARK$, S. J . H\YANGl\ and

J . 13. PARK11
?Graduate School of Riotechnology and $Department of Statistics,
Korea University, Seoul 136-701, Korea
Journal of Microencapsulation Downloaded from [Link] by University of Guelph on 06/14/13

:Department of Packaging Science, Clemson University, Clenison,

SC 29634-0370, USA
:!College of Pharmacy, Chungnam National University, Taejon 305-764, Korea
IISam-A Pharmaccutical Co. Ltd, Seoul 135-101, Korea

([Link] I 7 Feb~ircru~~
200.3; revised I 7 June 200.3; accepted 2.3 Jiiire 200.3 )

I tie objectives \yere to investigate the effects of formulation variables on the

7 7

release of drug and to optimize the formulation of chitosan microparticles

loaded \\it11 drug for controlled release using response surface methodology.
Chitosan micropar-ticles were prepared by dropping a chitosan solution into
sodium tripolyphosphate (TP P ) through ionic cross-linking. T h e release
behaviocir of felodipine as a model drug was affected by preparation variables.
A central composite design w a s used to evaluate and optimize the effect of
For personal use only.

preparation variables, chitosan concentration (A’,), the pH of the T P P solution

( X 2 ) and cross-linking time (
‘1’3) on the cumulative per cent drug release ( v ) in
24 h. Chitosan concentration and cross-linking time affected negatively the
release uf fclodipine, while the p H of the T P P did so positivclp and was the
highest inHuenti;il factor. T h e optimum rate of drug release, 100’%1 in 24 h, was
achieved at 1.8‘%, chitosan concentration, a pH 8.7 for the ‘I’PP solution and
9.7 niin cross-linking time.

k’ejwortls: Chitrisan, microparticle, drug-delivery system, response surface

methodology .

Introduction

Chitosan is a natural cationic polysaccharide derived from chitin, which is a

copolymer of glucosamine and N-acetyl glucosamine units (Mi et a / . 1999b, Ravi
I<umar 2001). I t has been estensively studied in the pharmaceutical industry as a
carrier for drugs owing to its biocompatibility and biodegradability (Bayomi et a / .
1998, Genta et a/. 1998, Mi et al. 1999a). It can form a gel matrix with counter-ions
such as sodium tripolyphosphate (TPP) (Rodmeier et at. 1989, Shiraishi et al.
1993, Calvo et a / . 1997). 1‘PP is a non-toxic and multivalent anion. I t can form
either inter- or intramolecular links between positively charged amino groups of
chitosan and negatively charged counter-ion of TPP (Aral and Akbuka 1998,
Mi et a / . 1999c, Shu and Z h u 2000).

*To \vtioni correspondence should be addressed at: Graduate School of Riotechnology,

Korea University, 1, S-ICa, Anani-Dong, Sungbuk-Ku, Seoul 136-701, Korea. e-mail:
hjpark@ kore;r.:[Link]

Joiirttol I S S N 02/15-2048 p r i n t / l S S N 1465-5240 cmlinc

of ~l,lirrn~~trrnpsri/ntbrr 0 2003 T a y l o r & I:ranc!s I,td
litt~~://\~~~\~.tnndf.~~,.uk/j~,nmals
1101: [Link]/1~~~~s~o~r~~11~oo1~~~os1~
 792 J . A. KOet al

D r u g release from chitosan microparticles can be controlled by the matrix

density, which is affected by various factors, such as the molecular weight and
concentration of chitosan, and the degree of cross-linking (concentration of cross-
linking agent and cross-linking time). R e m u d h - L o p e z and Bodmeier ( I 997)
reported that t h e release of d r u g from chitosan films decreased as the concentration
of the T P P solution increased. I n addition, they showed that the swelling and
permeability characteristics of chitosan films were affected by pH and the concen-
tration of the cross-linking agent. Puttipipatkhachorn et a/. (2001) reported that
the higher the molecular weight and degree of deacetylation of chitosan, the lower
Journal of Microencapsulation Downloaded from [Link] by University of Guelph on 06/14/13

the release rate of chitosan film.

However, quantitative aspects of the effects and relationships among these
various factors have not been studied yet. I n a previous study, we showed that the
effect of the preparation process on the release behaviour of chitosan microparticles,
which w a s prepared under various conditions such as concentration and molecular
weight of chitosan, p H and concentration of TPP solution, and curing time.
1 herefore, the present study attempts to determine the relationships among some
r7

factors that affect the release profile using response surface methodology (RSM).
R S M is a useful statistical technique for formulation and process optimization,
which is accomplished by studying the mutual interactions among the variables over
a range of values in a statistically valid manner (Hwang et a / . 2002).
T h e objectives were to investigate the effects of formulation variables on the
For personal use only.

release of d r u g from chitosan microparticles and to optimize the preparation of

chitosan microparticles loaded with d r u g for controlled release using RSM.
Felodipine (C18H19Cl2N04, M W 384.3), the treatment of hypertension, was
used a s a model drug.

Materials and methods

Materials
Chitosan ( M W 3.5 x lo", degree of deacetylation 86.4%) was obtained from
Biotech. Co. (Mokpo, Korea). Felodipine w a s obtained from Nivedita Chemicals
(Mumbai, India). Sodium tripolyphosphate ( T P P ) and all other reagents were
of analytical grade.

Response surface methodology

A centi-al composite design was employed to tit a second-order model for three
variables at five levels. Chitosan concentration (XI), pH of the T P P solution (&)
and cross-linking time ( X , ) were three independent variables (factors) considered
in the preparation of chitosan microparticles, while the release of d r u g in 24 h was a
dependent variable (response). T h e actual and corresponding coded values of three
factors ( X I ,X 2 , X,)are given in table 1. T h e complete design consisted of 16
experimental points that included eight vertex points, six axial points and two
replications at the centre point. T h e following second-order model in XI, X 2 and
X 3 w a s fitted using the data in table 1:
 [Link] nricvopavticle preparntiorr by response surface methodology 793

Table 1. Ccntral composite design.

Coded values
s, Indepcndcnt variables - 1.682 -1 0 1 1.682

A‘, chitosan concentration (I%) 1.359 1 .s 2 2.5 2.841

x 2 ‘I’PP p H 3.636 5 7 9 10.364
X3 cross-linking time (niin) 3.18 10 20 30 36.82
Journal of Microencapsulation Downloaded from [Link] by University of Guelph on 06/14/13

I able 2. Experimental runs and rcsponscs.

7 7

Factors
Response
I<llns s, x2 s 1 Y

-1 -1 -1 75.353
-I -1 1 60.39 7
-1 1 -1 96.103
-I 1 I 84.539
I -I -1 53.289
1 -1 1 45.240
1 93.134
For personal use only.

I -1
I I I 73.683
0 0 0 76.777
J 0 0 0 78.562
Ii - 1.082 0 0 89.352
I, 1.682 0 0 49.876
h‘l 0 -1.682 0 43.902
N 0 1.682 0 89.173
0 0 0 -1.682 99.930
P 0 0 1.682 47.133

where thc S , ’ s are coded independent variables. T h e S A S program w a s used for

analysis of this second-order model (SAS 1996) and the Mathernatica program
w a s iised for a four-dimensional response surface plot (Gray 1‘994).

I’repnrcr t iotr of TPP-chitosari W I icvo p a r t i d e s

r l
I he preparation of the microparticles followed the method described by I<o
et nl. (2002). Briefly, 16 types of chitosan microparticles were prepared according to
the central composite design (table 2). Chitosan solutions of varying concentrations
were prepared h y dissolving them in dilute acetic acid (1% v/v). rhTeen 80.2% ( v / v )
\\‘as added into the solution as a surfactant. T h e core material, felodipine, dissolved
in CI-I;?CI,?(2: lo), w a s mixed with the i q u e o u s phase (chitosan solution) in a
homogenizer (HR’I%-20DN, Hana Co., Shihung, Korea) at 5000 rpm for 20 min.
r 7
I he volume ratio of CHzClz:aqueuus phase was 1 :lo. The emulsion was cross-
linked b y dropping through a spray gun into the TPP solution (10%). After cross-
linking was allowed for w r y i n g time, microparticles were washed with distilled
water repeatedly and viicc~unidried for 1 2 11.
 794 J . A. KOet al.

In vitro release stitdies

TPP-chitosan microparticles (1 0 mg) were suspended in 500 ml phosphate-
buffered saline (PBS) ( p H 7.4) at 37"C, 50rpni. Tween 80 (O.S'%, V/V) was added
into the medium as an emulsifier. Samples (1 ml) were withdrawn periodically and
the volume of each sample was replaced by the same volume of fresh medium. T h e
amount of released felodipine was assayed spectrophotometrically at 362 nm
(Abrahamsson et al. 1994). T h e in vitro release study was performed in triplicate
for each sample.
Journal of Microencapsulation Downloaded from [Link] by University of Guelph on 06/14/13

Results and discussion

Dissolution profiles and the results of the 16 formulations prepared using the
experimental design are shown in figure 1 and table 2, respectably. T h e fitted
model was:

Y = 77.35 - 8.61X'i + 13.87X2 - 10.45X3 + 2.91XlX2 - 0.13X,X3

- 0.99X2X3 - 2.08Xf - 3.17X; - 0.69X:,

where Y is cumulative per cent drug released after 24 h , XI is the chitosan solution
concentration, X 2 is the p H of the TPP solution and XDis cross-linking time. Lack
For personal use only.

of fit was not significant ( p = 0.1036) and regression was strongly significant
( p = 0.0090, R2 = 0.9258), so it was concluded that the second-order model
adequately approximated the true surface. T h e estimated second-order RSM
(figures 2 and 3) indicated that chitosan concentration and cross-linking time
were affected negatively, while the pH of the TPP was affected positively. One
can also see that X2 was the most significant factor.

+A
O B
1-C
- v D
t E
U F
-+G
u n
A I
4 - J
+ K
U L
+ M
U N
t o
v p

0 10 20 30
Time (hr)
Figure 1 . Release profiles of drug from the 16 formulations of chitosan microparticles
prepared using the experimental clesign.
 Cliitosnn iiiicvopnrticle preparatioii ~ J response
J sirrfnce methodology 795

CI
4
N
109.81
cl
.“

w
Journal of Microencapsulation Downloaded from [Link] by University of Guelph on 06/14/13

55.96
-8
z

’ 29.04
36

Figure 2. Response surface plot (3D) showing the effect of pH of the ‘I’PP solution (A’,)
and cross-linking time (X,) o n the cumulative per cent released in 2411 ( Y )
(concentration o f chitosan solution = 2%).

DH of TPP
For personal use only.

-1.3 -
Chitosan Conc. (“h)
Figure 3. Rcsponw surface plot (4D) showing t h e effect of chitosan concentration (A’,),
o n the cumulativc pcr cent
the p H of ‘I’PP solution ( X , ) and cross-linking titnc (S,)
relc‘lscd I l l 24 I1 (v).

Figiirc 2 i s the response surface plot, which shows the efTect of the pl-I of the ‘I‘PP
(A’,) and cross-linking time (X,) on the d r u g release response, Y . As expected from
the estimated RSNI, the 1’ increased as X2 increased. T h i s can be attributed to the
ionization of TPP, which is controlled by the pH of the solution. ‘TPP w a s
dissociated into OH- and tripolyphosphate ions (HP301(,’-and P3OlO5-),where
only P3Olo5-occurred at low p H . Moreover, as the pH of the solution increased,
the ionization of the amine group of chitosan decreased, resulting in weak ionic
 796 J . A . K Oet al.

cross-linking, which led to a formation of less dense chitosan niicroparticles. On the

other hand, chitosan microparticles prepared in acidic T P P solution formed a dense
matrix through complete ionic cross-linking, enabling controlled release (Mi et al.
1999b, Shu and Zhu 2000, Lee et al. 2001, KO et al. 2002).
T h e negative coeflicient of X
I and X3 refers to the decreased amount (X,) of drug
released as chitosan concentration and cross-linking time increased (figure 3).
Figure 3 is four-dimensional surface plot of all formulation variables for the
cumulative per cent drug release from chitosan microparticles. An increase in
drug release was observed at low chitosan concentration and a high pH of the
Journal of Microencapsulation Downloaded from [Link] by University of Guelph on 06/14/13

r 7
1 PP solution. A high chitosan concentration resulted in a more dense matrix, which
causes reduced swelling ability of microparticles. Consequently, slowed diffusion
takes place, and it leads to decreased release of drug. Aqikgoz et al. (1996) and
Bayomi et a/. (1998) also showed that a significant increase in drug release was a
result of the formation of a more porous structure from chitosan of lo\v concentra-
tion. When chitosan solution was dropped into the T P P solution, an ionic interac-
tion occurred between tripolyphosphoric ions (PRO 10s-) and the amino group
(NH,+) of chitosan, resulting in cross-linking (Mi et al. 1999a). Therefore,
increased curing time would increase the interaction time, leading to the formation
of a dense chitosan matrix, which is thought to be associated with a decrease in the
release of drug.
T h e optimum values of preparation variables for the chitosan microparticles and
For personal use only.

response Y (100% drug release in 24 h) were determined by the estimated ridge

of maximum response. Results of the estimate ridge of maximum response test
revealed that optimum chitosan concentration (XI), the p H of the TPP solution
(X,) and curing time (X,) were 1.8'%,pH 8.7 and 9.7 min, respectively, for loo'%,
drug release in 24 h.

Conclusions

T h e response surface methodology w a s used to investigate the effects of

formulation variables and to optimize the formulation of chitosan microparticles
loaded with drug for controlled release. T h e release of drug w a s highly dependent
on the formulation variables, with significant interactions among these variables.
Chitosan concentration and cross-linking time had negative impacts and the pH of
the T P P had a positive impact on the release behaviour of felodipine. T h e p H of the
T P P solution was the most influential factor that affected release of drug from
chitosan microparticles.

Acknowledgement

T h e study was supported by a grant from the Korea Health 21 RSLD Project,
Ministry of Health PL Welfare, Republic of Korea (HMP-00-PT-21700-0017).

References

B.,
AI~[Link], JOIIANSSOS, D., TOIWI~I;NSSOS, A. and \ V I N G S T R ~ \ N I I ,K., 1994,
Evaluation of stabilizers in the drug relcase testing of hydrophilic matrix cstcndcd-
release tablets of fclodipine. Pharwracer~ticalreserrrclr, 11, 1093-1 097.
 Clritosnrr irricroparticle pveparntmi bji resporise sitrfnce nretlrodology 797

.A~:ii;r;Oz, WI., I<A$, H . S., ORMAN,NI. and H ~ S C A LA., A , , 1996, Chitosan microspheres
of diclofenac sodium: I . Application of factorial design and evaluation o f re1
kinetics. Jorrrirol .f [Link], 13, 141-1 60.
ARAI,, C . and AI<IJ(J~A, J ., 1998, Alternative approach to the preparation of chitosan I x i d s .
of Jorrrrrnl of Plrormcrcerctics, 168, 9-1 5.
lrrterrr~rtiorr~rl
HAYOAII,PI.A , , AI.-SU\VAYKII,S. A , , EI,.HI:I,W, A. M. and N~I;SNAII, A. F., 1998, Preparation
of casein-cliitosan microspheres containing diltiazcni hydrochloride by an aqueous
coiiccrvation technique. Plrnrrrrncerrticn Actcr Helvetine, 73, 187-1 92.
I~OIIAII;II:I<, R., 011,I<. H. and I’I<ANAI<, Y., 1989, Preparation and evaluiition of d r u g -
containing chitosan beads. Drug Developmerrt nrrd /rrclrrstrial Plrnr?rrncy, 15,
Journal of Microencapsulation Downloaded from [Link] by University of Guelph on 06/14/13

C., VII,A-J>YIX),J . L. atid [Link], ’&I. J., 1997, N~\nel

-LOI~I<X,
tosnn-polyctliyleiie oxide nanoparticlcs :is protein carriers. Joitrrrnl of
Applied Polyrrrer Scierrce, 63, 125-1 32.
GISNTA,I . , C~STATINI, M., .[Link], A , , COXTI, B. and MONTASAI<I, Id., 1998, Influence of
glut:iraldeliydc on d r u g releasc and mucoadhesive properties o f chitosan
ni i cros ph cres . Crr rbolry drat e Poly viers, 36, 8 1-8 8.
G R A YJ ,. 1-1 ., 1994, Mnstevirrg A4rrtlrerrrrrticn Progrnwrtrrirrg Metlrotls ~ i r dAppliccrtiorrs (Boston:
Acadcni ic Press/A P Professional).
IH\VM, I<, ‘ I - , , J U S C ; , S. T.. idl;1?, G . D., cIIINNAN, n4. s., PMK, 17, s. an d PAI<K,H. J . ,
2002, Controlling molecular \\,eight and degree of dcacetylation of chitosan by
response surface methodology. Jorrrnnl of Agrictrltrircrl crrrrl Food Clremistry, 50,
1876-1 882.
KO,J . A , , PAI<K,H. J . , ~-I\vAsc;, S. J., [Link]<K, J . B. and LI:I<,J . S . , 2002, Preparation and
For personal use only.

character i za t i ( ) n o f cli i tnsan mi cropart i cles in tend cd for control 1ed d r u g delivery .
Ilrterrr~rtiot/nlJo/r,t/crl of l’l~~~~~tlncerrtic.s, 249, 16s-I 74.
Idix, S. T., M I , F. Id., SIIKS, Y. J . and S i i Y u , S . S.,2001, Equilibrium and kinetic studies of
copper( 1 I ) ion uptake b y chitosan-tripolyphosphate chelating resin. Polynrer, 42,
1879-1892.
R ~ I , 17. I.,., SIIYLJ, S . S., CIII:~, C . 7‘. and S C I I ~ U NJ .GY., , 1999c, Porous ctiitosan
microsphere for controlling the antigen release of Nc\vcastlc disease vaccine:
preparation of antigen-adsorbed microsphere and in vitro release. f3ionratericrls, 20,
1603-1 61 2.
wil, 12. id., [Link](J, s. s., i<UAN, C . Y.,idEk:, S . T., I,U, I<. T. and J,.\NC;, s. F., 1 9 ~ ~ 1 ,
Cliitosan-1)olyelectrnlytc coinplexation for the preparation of gel beads iind
controlled release of anticancer d r u g . I . Enzymatic hydrolysis of polymer. Jorrrtrnl of
Applied Po1~:IlrerSciellce, 74, 1868-1 879.
R41, F. I.,., SIIY~J,S. S.. LIX, S. ‘1’. and Wosc;, 1‘. I3., 199913, Kinetic study o f chitosan-
tripol!pliosptiate complex reaction iind acid-resistive properties of the chitosan-
tripol!~pliospliatc gel beads prepared by in-liquid curing method. Jortrrirrl of I’oly~~rer
Scierrce: P o r t 13: Polyrrrer Plrysics, 37, 1551-1 564.
I1’1 I’ATK IIA C 1.1 ( ) I<S, S ., NuS’I‘I-I:\N I11, J ., 1’A M :\M( 1, K . and PIXII, C . E., 2001, I3rug
physical state and drug-polymer interaction on d r u g release from chitosan matrix
tilnis. Jorrrrrrrl of Coufrolled Relense, 75, 143-1 5 3 .
RAVI I<IJ~\.IAH. R4. N. \I., 2001, A revie\v of chitin and cliitosan applications. l i e t r c t i ? ~arid
Frirrctiorr~rlPolvrrrers, 46, 1-27.
I< l:h I 1J A A N - Idc’)lllsz, C . arid DO IM I;II:I<,R ., 1 99 7, R4 ech an i cal , \\’a ter u p ta kc a t i d perm cab i 1i ty
prnpcrtics of crosslinked chitosan glutamate and alginatc films. Jortrrrnl of Cowtrolled
Rclecrse, 44, 2 15-225.
SAS, 1996, S A S Sj~.stetrr,for Regression, 3rd edn (Cary: SAS Institute, Inc.).
SIIII<:\ISIII, S., IAIAI, ’1‘. and O.I,AC;II<I,WI., 1993, Controlled release of indonicthacin b y
chi t o s a ti-po I ye1cc t 1-0y1te con1p I ex : opt im iza t inn and irr uivo/irr vit ro eva I i t it t i o n .
Jortrrrnl (d Corrtrolled Releirse, 25, 21 7-225.
SIIU, S . Z. and ZIIU, I<. J . , 2000, A novel approach to prepare tripolyphosphatc/chitos~tn
complex beads for controlled d r u g delivery. frrterrrotional of Jorrrrrnl of
Pl~orlrrrrcerrtirs,201, 5 1-58,