Ci

AD

a

ARRAA

KCGSOCR

1

ocmstyoipcocdt

ci

0h

International Journal of Biological Macromolecules 51 (2012) 1070– 1078

Contents lists available at SciVerse ScienceDirect

International Journal of Biological Macromolecules

jo u rn al hom epa ge: www.elsev ier .com/ locate / i jb iomac

alcium alginate/gum Arabic beads containing glibenclamide: Development andn vitro characterization

mit Kumar Nayak, Biswarup Das ∗, Ruma Majiepartment of Pharmaceutics, Seemanta Institute of Pharmaceutical Sciences, Mayurbhanj 757086, Odisha, India

r t i c l e i n f o

rticle history:eceived 29 June 2012eceived in revised form 29 July 2012ccepted 19 August 2012vailable online 27 August 2012

eywords:alcium alginateum Arabicustained releaseptimization

a b s t r a c t

This work investigates the development, optimization and in vitro characterization of calciumalginate/gum Arabic beads by an ionotropic gelation method for prolonged sustained release of gliben-clamide. The effects of amount of sodium alginate and gum Arabic as independent process variables on thedrug encapsulation efficiency and drug release were optimized and analyzed based on central compositedesign and response surface methodology. Increment in drug encapsulation efficiency and decrease indrug release were found with the increase of both the amounts of sodium alginate and gum Arabic, usedas polymer-blend. These optimized beads showed high drug encapsulation efficiency (86.02 ± 2.97%),and suitable sustained drug release pattern over prolonged period (cumulative drug release after 7 hof 35.68 ± 1.38%). The average size of these formulated dried beads containing glibenclamide rangedfrom 1.15 ± 0.11 to 1.55 ± 0.19 mm. The in vitro dissolution of these beads showed prolonged sustained

2

entral composite designesponse surface methodology

release of glibenclamide over 7 h, which followed first-order model (R = 0.9886–0.9985) with anoma-lous (non-Fickian) diffusion mechanism (release exponent, n = 0.72–0.81). The swelling and degradationof the optimized beads were influenced by pH of test mediums. These beads were also characterizedby SEM and FTIR spectroscopy for surface morphology and excipients-drug interaction analysis, respec-tively. These developed calcium alginate/gum Arabic beads containing glibenclamide could possibly beadvantageous in terms of advanced patient compliance with reduced dosing interval.

. Introduction

Currently, the exploration of natural polymers in the devel-pment of various pharmaceutical dosage forms has advancedonsistently in terms of diversity and property. The natural poly-ers offer certain specific advantages over synthetic polymers,

uch as easy availability, biocompatibility, biodegradability, non-oxicity and pollution free processing [1,2]. Over the past fewears, a great deal of attention has been paid in the developmentf biopolymeric beads for use as drug delivery carriers throughonotropic gelation technique [3,4]. The ionic polymers (alginate,ectin, gellan gum, etc.) undergo ionotropic gelation and pre-ipitate to form beads due to electrostatic interaction betweenppositely charged species. This technique is very simple and theonditions used were very mild. In addition, physical cross-linkingue to ionotropic-gelation instead of chemical cross-linking avoidshe possible toxicity of reagents and other undesirable effects [3].

Among various ionic biopolymers, alginate, a polyanionicopolymer of mannuronic and guluronic acid residues has beennvestigated for its unique nature of forming hydrogel beads

∗ Corresponding author. Tel.: +91 9583131603.E-mail address: biswarup777@gmail.com (B. Das).

141-8130/$ – see front matter © 2012 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.ijbiomac.2012.08.021

© 2012 Elsevier B.V. All rights reserved.

through ionotropic gelation [3,4]. Alginate undergo ionotropic gela-tion in the presence of divalent cations like Ca2+, Ba2+, Cu2+, Pb2+,Zn2+, Cd2+, etc and trivalent cations like Al3+, etc due to ionicinteraction between carboxylic acid groups of alginate and thesecations [5,6]. Various drugs have been successfully encapsulatedin ionotropically gelled alginate beads and exhibited different drugrelease profiles [7,8]. Although, ionotropically gelled alginate beadscan be prepared by simple and mild procedures, this method has amajor limitation of drug loss during bead preparation due to leach-ing of drugs through the pores [9]. Therefore, many modificationsof ionotropically gelled alginate beads based on the use of anotherpolymer as blend with alginate were investigated for drug deliveryapplications [2,10–18].

Gum Arabic, a biocompatible and biodegradable natural gum,is mainly used in oral and topical pharmaceutical formulations[19]. The popularity of gum Arabic is due to physical propertiesincluding high solubility, pH stability, non-toxicity and gellingcharacteristics [20]. According to United States Food and DrugAdministration (USFDA), gum Arabica enjoys the ‘Generally Rec-ognized as Safe (GRAS)’ status [20]. It is highly branched and

slightly acidic polysaccharide found as a mixed calcium, magne-sium, and potassium salt of polysaccharidic acids with main chainof (1→3)-�-d-galactopyranosyl units and side chains containingl-arabinofuranosyl, l-rhamnopyranosyl, d-galactopyranosyl, and

ological Macromolecules 51 (2012) 1070– 1078 1071

dtpiiaeudp

t[mitp

mimtmoddmimcfa

2

2

(Iu

2g

ccB

Table 1Factors and levels of the central composite design used for optimization.

Normalized levels Experimental settings

Sodium alginate (mg)(A)

Gum Arabic (mg)(B)

−1.414 179.29 79.29−1 200.00 100.00

TE

A.K. Nayak et al. / International Journal of Bi

-glucopyranosyl uronic acid units [20,21]. In the current inves-igation, the combination of sodium alginate and another naturalolymer, gum Arabic was used as polymer-blend to develop

onotropically gelled calcium alginate/gum Arabic beads for usen sustained drug release application. However, any report onlginate-gum Arabic blend beads is unavailable in the previous lit-rature, which adds the work novel of its kind. Glibenclamide wassed as a model drug in the present study to evaluate the sustainedrug release potential of the calcium alginate/gum Arabic beadsrepared using ionotropic gelation.

Glibenclamide, a second generation sulfonylurea, is used in thereatment of non-insulin dependent diabetes mellitus (Type-II)22]. Its plasma half-life is 4–6 h, which makes multiple dosing to

aintain the therapeutic blood level [23]. To reduce multiple dos-ng, sustained release dosage of glibenclamide, which will be ableo deliver glibenclamide at a slow release rate over an extendederiod is essential.

Optimization by means of statistical experimental designethodologies has been widely applied by designing a set of exper-

ments that will reliably measure the response variables, fitting aathematical model to the data, conducting appropriate statistical

est to assure that the best possible model is chosen, and deter-ining the values of independent formulation variables to produce

ptimum response [24]. Among various statistical optimizationesigns, central composite design, which is a response surfaceesign, has been widely used for formulation and process opti-ization [13,16]. It is very efficient and flexible, providing much

nformation on experiment variable effects and overall experi-ental error in a minimal number of required runs [25]. In the

urrent investigation, the central composite design was employedor the formulation optimization of ionotropically gelled calciumlginate/gum Arabic beads containing glibenclamide.

. Experimental

.1. Materials

Glibenclamide (B.S. Trader Pvt. Ltd., India), sodium alginateCentral Drug House, India), gum Arabic (Central Drug House,ndia), and calcium chloride (Mark Specialties Pvt. Ltd., India) weresed. All chemicals and reagents used were of analytical grade.

.2. Preparation of calcium alginate/gum Arabic beads containinglibenclamide

The calcium alginate/gum Arabic beads containing gliben-lamide were prepared using ionotropic gelation method. Calciumhloride (CaCl2) was used as cross-linker in ionotropic gelation.riefly, sodium alginate and gum Arabic aqueous dispersions were

able 2xperimental plan and observed response values from randomized run in central compos

Experimental formulations Factors

Sodium alginate (mg)(A) G

F-1 200.00 1F-2 200.00 2F-3 300.00 1F-4 300.00 2F-5 179.29 1F-6 320.71 1F-7 250.00 2F-8 250.00

F-9 250.00 1

a Observed response values: Mean ± S.D. (n = 3).b DEE (%) = Drug encapsulation efficiency (%).c R7h (%) = drug release at 7 h.

0 250.00 150.001 300.00 200.001.414 320.71 220.71

prepared separately using distilled water. These dispersions werewell mixed with stirring for 10 min at 1000 rpm using a magneticstirrer (Remi Motors, India). Afterwards, glibenclamide was addedto the dispersion mixture. The ratio of drug to polymer wasmaintained 1:4 in all formulations and mixed thoroughly using ahomogenizer (Remi Motors, India). The final sodium alginate/gumArabic gels containing glibenclamide were ultrasonicated for 5 minfor debubbling. The resulting dispersion was then added via a21-gauge needle drop wise into100 ml of 10% (w/v) CaCl2 solution.Added droplets were retained in the CaCl2 solution for 15 minto complete the curing reaction and to produce rigid beads. Thewet beads were collected by decantation, and washed two timeswith distilled water and dried at 37 ◦C for overnight. The calciumalginate/gum Arabic beads containing glibenclamide were storedin a desiccator until used.

2.3. Experimental design

A central composite design (spherical type, single center point,and ̨ = 1.414) was employed for the formulation optimization ofcalcium alginate/gum Arabic beads containing glibenclamide. Theamount of sodium alginate (A) and the gum Arabic (B) as poly-meric blend were defined as the selected independent formulationvariables (factors); while drug encapsulation efficiency (DEE, %),and drug release at 7 h (R7 h, %) were used as dependent variables(responses). The process variables (factors) and levels with exper-imental values are reported in Table 1. The matrix of the designincluding investigated factors and responses are also shown inTable 2. Design-Expert 8.0.6.1 software (Stat-Ease Inc., USA) wasused for generation and evaluation of experimental design. Thepolynomial mathematical model generated by circumscribed cen-tral composite design is following [25]:

Y = b0 + b1A + b2B + b3AB + b4A2 + b5B2,

where Y is the response; b0 is the intercept, and b1, b2, b3, b4, b5are regression coefficients. A and B are individual effects; A2 and B2

are quadratic effects; AB is the interaction effect. One-way ANOVAwas applied to estimate the significance of the model (p < 0.05). The

ite design.

Responsesa

um Arabic (mg)(B) DEE (%)b R7 h (%)c

00.00 54.17 ± 1.54 53.23 ± 2.2200.00 67.36 ± 2.15 44.24 ± 2.0300.00 60.25 ± 2.12 49.76 ± 2.1600.00 70.28 ± 3.04 43.75 ± 1.7750.00 59.13 ± 2.22 50.81 ± 1.9750.00 65.80 ± 1.98 46.36 ± 2.0320.71 73.93 ± 3.37 40.54 ± 1.5479.29 55.38 ± 1.76 53.06 ± 1.9850.00 58.04 ± 2.05 51.27 ± 2.07

1 iologic

stga

2

wT7tuifiuiu

D

2

boo

2

bBslc

2

bEpbs

2

gc(mrwitoiwfml

072 A.K. Nayak et al. / International Journal of B

urface response plots and contour plots were analyzed to revealhe effect of independent factors (amount of sodium alginate andum Arabic ratio) on the measured responses (DEE, and R7 h) werenalyzed.

.4. Determination of DEE (%)

Accurately weighed 100 mg of prepared beads from each batchere taken separately and were crushed using pestle and mortar.

he crushed powders were placed in 500 ml of phosphate buffer, pH.4, and kept for 24 h with occasionally shaking at 37 ± 0.5 ◦C. Afterhe stipulated time, the mixture was stirred at 500 rpm for 20 minsing a magnetic stirrer. The polymer debris formed after the dis-

ntegration of beads was removed by filtering through Whatman®

lter paper (No. 40). The drug content in the filtrate was determinedsing a UV–vis spectrophotometer (Shimadzu, Japan) by measur-

ng absorbance at �Max of 228 nm. The DEE of beads was calculatedsing the following formula:

EE(%) = actual drug content in beadstheoretical drug content in beads

× 100.

.5. Determination of bead size

Particle size of 100 dried beads from each batch was measuredy optical microscopic method for average particle size using anptical microscope (Olympus). The ocular micrometer was previ-usly calibrated by stage micrometer.

.6. Surface morphology analysis

The surface morphology of the formulated beads was analyzedy scanning electron microscope (SEM) (JEOL, JSM-5800, Japan).eads were gold coated by mounted on a brass stub using double-ided adhesive tape and under vacuum in an ion sputter with a thinayer of gold (3–5 nm) for 75 s and at 15 kV to make them electricallyonductive and their morphology was examined.

.7. Fourier transform-infrared (FTIR) spectroscopy

Samples were reduced to powder and analyzed as KBr pelletsy using a Fourier transform-infrared (FTIR) spectroscope (Perkinlmer Spectrum RX I, USA). The pellet was placed in the sam-le holder. Spectral scanning was taken in the wavelength regionetween 3800 and 400 cm−1 at a resolution of 4 cm−1 with scanpeed of 1 cm/s.

.8. In vitro drug release studies

The release of the glibenclamide from various ionotropicallyelled calcium alginate/gum Arabic beads containing gliben-lamide was tested using a dissolution apparatus USP/BP/IPCampbell Electronics, India). The baskets were covered with 100-

esh nylon cloth to prevent the escape of the beads. The dissolutionates were measured at 37 ± 1 ◦C under 50 rpm speed. Accuratelyeighed quantities of calcium alginate/gum Arabic beads contain-

ng glibenclamide equivalent to 5 mg glibenclamide were addedo 900 ml of simulated gastric fluid (pH 1.2). The test was carriedut in simulated gastric fluid (pH 1.2) for 2 h, and then, continuedn simulated intestinal fluid (pH 7.4) for next 5 h. 5 ml of aliquots

as collected at regular time intervals, and the same amount ofresh dissolution medium was replaced into dissolution vessel to

aintain the sink condition throughout the experiment. The col-ected aliquots were filtered, and suitably diluted to determine the

al Macromolecules 51 (2012) 1070– 1078

absorbance using a UV–vis spectrophotometer (Shimadzu, Japan)by measuring absorbance at �Max of 228 nm.

2.9. Analysis of in vitro drug release kinetics and mechanism

In order to predict and correlate the in vitro drug release behav-ior from formulated calcium alginate/gum Arabic beads containingglibenclamide, it is necessary to fit into a suitable mathematicalmodel. The in vitro drug release data were evaluated kineticallyusing various important mathematical models like zero order,first order, Hixson–Crowell, Weibull, Baker–Lonsdale, Higuchi, andKorsmeyer–Peppas models [26].

Zero-order model: Q = kt + Q0; where Q represents the drugreleased amount in time t, Q0 is the start value of Q, and k is therate constant.First-order model: Q = Q0 ekt; where Q represents the drug releasedamount in time t, Q0 is the start value of Q, and k is the rate constant.Hixson–Crowell model: Q 1/3 = kt + Q 1/3

0 ; where Q represents thedrug-released amount in time t, Q0 is the start value of Q, and k isthe rate constant.Weibull model: m = 1 − exp [−(t)b/a]; where m represents the drugreleased amount in time t, a is the time constant, and b is the shapeparameter.Baker–Lonsdale model: 3/2 [1 − (1 − Q)2/3] − Q = kt; where Q repre-sents the drug released amount in time t, and k is the rate constant.Higuchi model: Q = kt0.5; where Q represents the drug releasedamount in time t, and k is the rate constant.Korsmeyer–Peppas model: Q = ktn; where Q represents the drugreleased amount in time t, k is the rate constant, and n is thediffusional exponent, indicative of drug release mechanism.

The accuracy and prediction ability of these models were com-pared by calculation of squared correlation coefficient (R2) usingKinetDS 3.0 Rev. 2010 software.

Again, The Korsmeyer–Peppas model was employed in the invitro drug release behavior analysis of these formulations to dis-tinguish between competing release mechanisms: Fickian release(diffusion-controlled release), non-Fickian release (anomaloustransport) and case-II transport (relaxation-controlled release).When n is ≤0.43, it is Fickian release. The n value between 0.43 and0.85 is defined as non-Fickian release. When, n ≥ 0.85, it is case-IItransport [13,15,16].

2.10. Swelling behavior measurement

Swelling measurement of optimized calcium alginate/gum Ara-bic beads containing glibenclamide was carried out in two differentaqueous media: simulated gastric fluid (pH 1.2), and simulatedintestinal fluid (pH 7.4). 100 mg beads were placed in vessels of dis-solution apparatus (Campbell Electronics, India) containing 500 mlrespective media. The experiment was carried out at 37 ± 1 ◦Cunder 50 rpm paddle speed. The swelled beads were removed atpredetermined time interval and weighed after drying the surfaceby using tissue paper. Swelling index was determined using thefollowing formula:

Swelling index

= weight of beads after swelling − dry weight of beadsdry weight of beads

× 100.

A.K. Nayak et al. / International Journal of Biological Macromolecules 51 (2012) 1070– 1078 1073

Table 3Summary of ANOVA for the response parameters.

Source Sum of squares d.f.a Mean square F-Value p-ValueProb > F

(a) For DEE (%)b

Model 379.95 5 75.99 130.06 0.0011 (S)A 42.47 1 42.47 72.69 0.0034 (S)B 305.71 1 305.71 523.25 0.0002 (S)AB 2.50 1 2.50 4.27 0.1306 (NS)A2 12.54 1 12.54 21.46 0.0189 (S)B2 29.26 1 29.26 50.08 0.0058 (S)

(b) For R7 h (%)c

Model 163.47 5 32.69 61.25 0.0032 (S)A 13.14 1 13.14 24.62 0.0157 (S)B 133.71 1 133.71 250.50 0.0005 (S)AB 2.22 1 2.22 4.16 0.1341 (S)A2 5.14 1 5.14 9.63 0.0532 (NS)B2 14.36 1 14.36 26.91 0.0139 (S)

A and B represent the main effects (factors); A2 and B2 are the quadratic effect; AB is the interaction effect.

2

8cKai

3

3g

tcactribdinapsoldcmrettgatst

a d.f. = Degree of freedom.b DEE (%) = Drug encapsulation efficiency (%).c R7h (%) = drug release at 7 h.

.11. Statistical analysis

Statistical optimization was performed using Design-Expert.0.6.1 software (Stat-Ease Inc., USA). The squared correlationoefficients (R2) of all kinetic models were determined usinginetDS 3.0 Rev. 2010 software. All measured data are expresseds mean ± standard deviation (S.D.). Each measurement was donen triplicate (n = 3).

. Results and discussion

.1. Preparation of calcium alginate/gum Arabic beads containinglibenclamide

The divalent calcium ions fit into electronegative cavities ofhe sodium alginate like eggs in an “Egg Box” model to formalcium alginate due to electrostatic ionic interaction between neg-tively charged carboxylate group of sodium alginate and positivelyharged calcium ions present in the cross-linking solution [27]. Athe cross-linking sites, polyvalent cations cause interpolysaccha-ide binding, which are called as junction zones [28]. The calciumons compete with the sodium ions of sodium alginate and thus,ring the two polymer chains together. Calcium ions are accommo-ated in the interstices of two polyuronate chains having a close

on-pair interaction with carboxylate anions of the sodium algi-ate and sufficient coordination by other electronegative oxygentoms [29]. Besides anionic nature of sodium alginate, gum Arabicossesses slightly anionic nature with a mixed calcium, magne-ium, and potassium salt of polysaccharidic acids with main chainf (1→3)-�-d-galactopyranosyl units and side chains containing-arabinofuranosyl, l-rhamnopyranosyl, d-galactopyranosyl, and-glucopyranosyl uronic acid units [20,21]. Therefore, calciumations present in the cross-linking solution might compete withagnesium and potassium cations present in the polysaccha-

idic structure of gum Arabic. When these polysaccharides werexposed to divalent calcium cations in the current investigation,here could occur an electrostatic ionic interaction between posi-ively charged calcium cations and negatively charged carboxylateroups of sodium alginate and gum Arabic. Thus, various calcium

lginate/gum Arabic beads containing glibenclamide was preparedhrough ionotropic gelation method when various dispersions ofodium alginate, gum Arabic, and glibenclamide were dropped intohe solutions containing calcium ions.

3.2. Formulation optimization by central composite design

Designing pharmaceutical formulations with the minimumnumber of trials is very crucial for pharmaceutical scientists [30].In conventional optimization process by changing one factor at atime, the optimization is usually carried out by varying a single fac-tor and keeping all other factors fixed at a specific set of conditions.This method is time consuming and incapable of effective optimiza-tion as it does not consider the interactive effects of all the primaryfactors [31]. It is therefore important to understand the influenceof formulation variables on the formulation quality with a minimalnumber of experimental trials and subsequent selection of formu-lation variables to develop optimized formulation using establishedstatistical tools such as central composite design. Central compos-ite design is a response surface design, which provides informationon individual effects, pair-wise interactions of various individualeffects and curvilinear variables effects [25]. In the current inves-tigation, a central composite design (spherical type, single centerpoint, and ̨ = 1.414) with total 9 experimental formulations of cal-cium alginate/gum Arabic beads containing glibenclamide wereproposed by Design-Expert 8.0.6.1 software (Stat-Ease Inc., USA)for two independent process variables (factors): amount of sodiumalginate (A) and amount of gum Arabic (B) in the polymer-blendused (Table 1). The effects of these independent variables on DEE(%), and R7 h (%) were investigated as optimization response param-eters. According to the trial proposal of central composite design, 9experimental formulations of calcium alginate/gum Arabic beadscontaining glibenclamide were prepared using ionotropic gela-tion method. Overview of the experimental plan and the observedresponse values are presented in Table 2. The Design-Expert 8.0.6.1software provided suitable polynomial model equations involvingindividual main factors and interaction factors for each investigatedresponses after fitting these data. These models were evaluated sta-tistically by applying one-way ANOVA (p < 0.05), which is shown inTable 3. The model p values of less than 0.05 for both the measuredresponses implied the models were significant (p < 0.05).

The model equation relating DEE (%) as response became:

DEE(%) = 96.57 − 0.32 A − 0.15 B − 3.16 × 10−4 AB + 8.31

×10−4 A2 + 1.27 × 10−4 B2[R2 = 0.9954; p < 0.05]

The model equation relating R7 h (%) as response became:

R7 h(%) = 27.89 + 0.20 A + 0.11 B + 2.98 × 10−4 AB − 5.32

×10−4 A2 − 8.89 × 10−4 B2[R2 = 0.9903; p < 0.05]

1074 A.K. Nayak et al. / International Journal of Biological Macromolecules 51 (2012) 1070– 1078

F odiumc g) and

sm

D 2

R

riritvfbrt(tOri

ig. 1. Three-dimensional response surface plots showing the effects of amounts of sorresponding contour plots showing the effects of amounts of sodium alginate (m

Model simplification was carried out by eliminating non-ignificant terms (p > 0.05) in model equations resulting from theultiple regression analysis [32], giving:

EE(%) = 96.57 − 0.32 A − 0.18 B + 8.31 × 10−4 A2 + 1.27 × 10−4 B

7 h(%) = 27.89 + 0.20 A + 0.11 B − 8.89 × 10−4 B2

Design-Expert 8.0.6.1 software generated three-dimensionalesponse surface plots and corresponding contour plots relatingnvestigated responses, DEE (%) and R7 h (%). The three-dimensionalesponse surface plot is very useful in learning about the main andnteraction effects of the independent variables (factors), whereaswo-dimensional contour graph gives a visual representation ofalues of the response [15,33]. The three-dimensional response sur-ace plots relating DEE (%) and R7 h (%) are presented in Fig. 1(a and, respectively). The two-dimensional corresponding contour plotselating DEE (%) and R6 h (%) are presented in Fig. 1(c and d, respec-ively). The three-dimensional response surface plot relating DEE%) (Fig. 1c) depicted the increase in DEE with the increase of both

he amount of sodium alginate (A), and amount of gum Arabic (B).n the other hand, the three-dimensional response surface plots

elating R7 h (%) (Fig. 1d) also indicated the decrease in R7 h with thencrease of both the amount of sodium alginate (A), and amount

alginate (mg) and gum Arabic (mg) on (a) DEE (%) and (b) R7 h (%). Two-dimensional gum Arabic (mg) on (c) DEE (%) and (d) R6 h (%).

of gum Arabic (B) in the formulated calcium alginate/gum Arabicbeads containing glibenclamide.

A numerical optimization technique using the desirabilityapproach was employed to develop new formulations with desiredresponse (desired quality). A constraint to maximizing the DEE andminimizing the R7 h was to set the goal to locate the optimum sett-ings of independent variables for the optimized formula by thecentral composite design using the the Design-Expert 8.0.3 soft-ware based on the criterion of desirability. To get the desired opti-mum responses, independable variables (factors) were restrictedto 200.00 mg ≤ A ≤ 400.00 mg, and 100.00 mg ≤ B ≤ 200.00 mg;whereas the desirable ranges of responses were restricted to85% ≤ DEE ≤ 100%, and 30% ≤ R7 h ≤ 40%. The desirability plot indi-cating desirable regression ranges for optimal process variablesettings and overlay plot indicating the region of optimal pro-cess variable settings are presented in Fig. 2(a and b, respectively).In order to evaluate the optimization capability of these modelsgenerated according to the results of central composite design,optimized calcium alginate/gum Arabic beads containing gliben-clamide were prepared using one of the optimal process variablesettings proposed by the design (R2 = 1). The selected optimal

process variable setting used for the formulation of optimizedformulation was A = 395.13 mg and B = 178.98 mg. The optimizedbeads containing glibenclamide (F-O) were evaluated for DEE(%) and R7 h (%). Table 4 lists the results of experiments with

A.K. Nayak et al. / International Journal of Biological Macromolecules 51 (2012) 1070– 1078 1075

Fig. 2. The desirability plot: (a) indicating desirable regression ranges and the overlay plot and (b) indicating the region of optimal process variable settings.

Table 4Results of experiments for confirming optimization capability.

Code Factors Responsesa

Sodium alginate (mg) Gum Arabic (mg) DEE (%)b R7 h (%)c

Predicted value Actual value Error (%)d Predicted value Actual value Error (%)d

F-O 395.13 178.98 85.24 86.02 ± 2.97 −0.92 34.66 35.68 ± 1.38 −2.86

a Observed response values: mean ± S.D. (n = 3).

e] × 10

pasemw

3

c7boTniptsvs

3

gItti

debris which was seen on the bead surface could be due to themethod of preparation (i.e., simultaneous gel bead preparation andformation of the polymer blend matrix).

Table 5Mean diameter of calcium alginate/gum Arabic beads containing glibenclamide.

Formulation codes Mean diameter (mm)a

F-1 1.18 ± 0.11F-2 1.32 ± 0.13F-3 1.39 ± 0.16F-4 1.50 ± 0.10F-5 1.15 ± 0.11F-6 1.51 ± 0.18F-7 1.32 ± 0.16

b DEE (%) = Drug encapsulation efficiency (%).c R7h (%) = drug release at 7 h.d Error (%) = [difference between predicted value and actual value/Predicted valu

redicted responses by the mathematical models and those actu-lly observed. The optimized beads containing glibenclamide (F-O)howed DEE of 86.02 ± 2.97% and R7 h of 35.68 ± 1.38% with smallrror-values (−0.92 and −2.86, respectively). This reveals thatathematical models obtained from the central composite designere well fitted.

.3. DEE

The DEE (%) of all these calcium alginate/gum Arabic beadsontaining glibenclamide was within the range, 54.17 ± 1.54 to3.93 ± 3.37% (Tables 1 and 4). It was observed that DEE (%) of theseeads containing glibenclamide was increased with the incrementf sodium alginate and gum Arabic amount in the polymer-blend.he increased DEE (%) with the increasing amount of sodium algi-ate and gum Arabic in these beads may be due to the increase

n viscosity of the polymer solution with the increasing amount ofolymer addition, so that, it might have prevented drug leachingo the cross-linking solution. In addition, the increasing amount ofodium alginate and gum Arabic in polymer-blend might have ele-ated the cross-linking by CaCl2 through availing more numbers ofites for ionic cross-linking.

.4. Bead size

The average size of these formulated dried beads containinglibenclamide ranged from 1.15 ± 0.11 to 1.55 ± 0.19 mm (Table 5).

ncreasing the bead size was found with the increasing amount ofhe polymers, sodium alginate and gum Arabic into bead formula-ions containing glibenclamide. This could be attributed due to thencrease in viscosity of the polymer blend (sodium alginate and gum

0.

Arabic) solution with incorporation of both the polymers in increas-ing ratio that in turn increased the droplet size during addition ofthe polymer blend solution to the cross-linking solution.

3.5. Bead morphology

The morphological analysis of calcium alginate/gum Arabicbeads containing glibenclamide was visualized by SEM at differentmagnifications and is presented in Fig. 3(a and b). The SEM photo-graph of these beads showed spherical shape with a rough surface.Detailed examination of the bead surface topography revealedcracks and wrinkles, which might be caused by partly collapsingthe polymeric gel network during drying [34]. However, polymeric

F-8 1.20 ± 0.12F-9 1.24 ± 0.13F-O 1.55 ± 0.19

a Mean ± S.D.

1076 A.K. Nayak et al. / International Journal of Biological Macromolecules 51 (2012) 1070– 1078

algin

3

aAata(gf1nost

atgAobmsgA

FAc

Fig. 3. Scanning electron microphotograph of optimized calcium

.6. FTIR spectroscopy

The FTIR spectra of sodium alginate, gum Arabic and calciumlginate/gum Arabic beads without drug, calcium alginate/gumrabic beads containing glibenclamide, and pure glibenclamidere shown in Fig. 4. The FTIR spectra of sodium alginate showedhe band around 3000–3600, 1614, 1417 and 1032 cm−1, whichre due to the stretching of OH, COO (asymmetric), COOsymmetric), and C O C, respectively. In the FTIR spectrum ofum Arabic showed characteristic peaks around 3000–3600 cm−1

or OH stretching, 2993 and 2918 cm−1 for CH stretching and612 cm−1 for C O stretching. The FTIR spectrum of calcium algi-ate/gum Arabic beads without drug showed characteristic peaksf both sodium alginate and gum Arabic without any significanthifting or deviations. In the FTIR spectrum of pure glibenclamide,he principal absorption peaks appeared at 3314 cm−1 due to theNH stretching, 3116 cm−1 for aromatic hydrogen absorption, and

peak at 1717 cm−1 occurred due to C O absorption peak. Inhe FTIR spectrum of calcium alginate/gum Arabic beads containinglibenclamide, various characteristic peaks of sodium alginate, gumrabic, and glibenclamide appeared without any significant shiftingr deviation of these peaks. In short, the calcium alginate/gum Ara-ic beads containing glibenclamide prepared by ionotropic gelationethod had significant characters of glibenclamide in the FTIR

pectrum, suggesting, there was no interaction between the drug,libenclamide and the polymers used (sodium alginate and gumrabic).

ig. 4. The FTIR spectra of sodium alginate (a), gum Arabic (b), calcium alginate/gumrabic beads without drug (c), calcium alginate/gum Arabic beads containing gliben-lamide (F-O) (d), and pure glibenclamide (e).

ate/gum Arabic beads containing glibenclamide (F-O) (a and b).

3.7. In vitro drug release

The in vitro glibenclamide release studies were carried outfor various calcium alginate/gum Arabic beads containing gliben-clamide in the 0.1 N HCl (pH, 1.2) for first 2 h and then, in phosphatebuffer (pH 7.4) for next 5 h. All these calcium alginate/gum Arabicbeads showed prolonged release of glibenclamide over 7 h (Fig. 5).Glibenclamide release from these beads in the acidic medium wasless than 15% after 2 h for all these beads containing glibenclamidedue to the shrinkage of pH sensitive alginate at acidic pH. The traceamount of glibenclamide release in the initial period could prob-ably be due to the surface adhered drug crystals. After that, drugrelease was observed faster in phosphate buffer (pH 7.4) compar-atively. This might be due to the higher swelling rate of alginate inphosphate buffer. The percentage drug released from calcium algi-nate/gum Arabic beads containing glibenclamide in 7 h (R7 h, %) waswithin the range of 35.68 ± 1.38–53.23 ± 2.22%, and was found tobe lower with the increasing of the polymers, sodium alginate andgum Arabic, present in the polymer-blend. In case of beads con-taining higher polymer content, the more hydrophilic property ofthe polymers could probably bind better with water to form vis-cous gel-structure, which might blockade the pores on the surfaceof these beads and sustain the drug release for prolonged period.

The in vitro drug release data from various calcium algi-nate/gum Arabic beads containing glibenclamide were evaluatedkinetically using various mathematical models like zero order,first order, Hixson–Crowell, Weibull, Baker–Lonsdale, Higuchi, andKorsmeyer–Peppas models. The R2 values of these models were

determined for evaluation of accuracy and prediction ability ofthese models using KinetDS 3.0 Rev. 2010 software. The resultof the curve fitting into various mathematical models is given in

Fig. 5. In vitro drug release from various calcium alginate/gum Arabic beads con-taining glibenclamide [mean ± S.D., n = 3].

A.K. Nayak et al. / International Journal of Biological Macromolecules 51 (2012) 1070– 1078 1077

Table 6Results of curve fitting of the in vitro glibenclamide release data from different calcium alginate/gum Arabic beads.

Models Formulation codes

F-1 F-2 F-3 F-4 F-5 F-6 F-7 F-8 F-9 F-O

Zero order R2 0.9347 0.9073 0.9218 0.8985 0.9237 0.9166 0.9011 0.9434 0.9214 0.9051First order R2 0.9985 0.9904 0.9955 0.9886 0.9953 0.9946 0.9876 0.9957 0.9895 0.9900Hixson–Crowell R2 0.9875 0.9737 0.9804 0.9672 0.9815 0.9797 0.9670 0.9878 0.9771 0.9715Weibull R2 0.8691 0.8442 0.8508 0.8271 0.8531 0.8547 0.8256 0.8707 0.8431 0.8413Baker–Lonsdale R2 0.7576 0.7594 0.7797 0.7570 0.7790 0.7680 0.7642 0.8082 0.7768 0.7662Higuchi R2 0.6539 0.5750 0.6130 0.5883 0.6196 0.5837 0.6052 0.6622 0.6250 0.5417Korsmeyer–Peppas R2 0.8970 0.8654 0.8759 0.8492 0.8787 0.8772 0.8452 0.8957 0.7679 0.8567

Tbl7gbdmaA

3

b1b7alasonsbnibm

Fin

na 0.73 0.77 0.75 0.74

a Diffusional exponent.

able 6. When the respective R2 of calcium alginate/gum Arabiceads containing glibenclamide was compared, it was found to fol-

ow the first-order model (R2 = 0.9886–0.9985) over a period of h. The value of release exponent (n) determined from in vitrolibenclamide release data of various calcium alginate/gum Arabiceads ranged from 0.72 to 0.81, indicating anomalous (non-Fickian)iffusion mechanism for drug release. The anomalous diffusionechanism of drug release demonstrates both diffusion controlled,

nd swelling controlled drug release from calcium alginate/gumrabic beads containing glibenclamide.

.8. Swelling behavior

The swelling behavior of optimized calcium alginate/gum Ara-ic beads containing glibenclamide was evaluated in 0.1 N HCl, pH.2, and phosphate buffer, pH 7.4. The swelling behaviors of theseeads in both the pH, 0.1 N HCl (pH 1.2), and phosphate buffer (pH.4) are shown in Fig. 6. The swelling index of optimized calciumlginate/gum Arabic beads containing glibenclamide was initiallyower in acidic pH (0.1 N HCl, pH 1.2) in comparison with that of inlkaline pH (phosphate buffer, pH 7.4), indicating a pH-sensitivewelling behavior. This might have occurred due to shrinkagef alginate at acidic pH. Maximum swelling of these beads wasoticed at 1–2 h in phosphate buffer, pH 7.4 and after which, ero-ion and dissolution of swollen cross-linked alginate/gum Arabiceads took place. The swelling behavior of optimized calcium algi-

ate/gum Arabic beads in alkaline pH could be explained by the

on exchange phenomenon between calcium ions, which are ininding with carboxylic groups of cross-linked alginate/gum Arabicatrix and the sodium ions present in phosphate buffer due to the

ig. 6. Swelling behavior of optimized calcium alginate/gum Arabic beads contain-ng glibenclamide in 0.1 N HCl, pH 1.2 and phosphate buffer, pH 7.4 [mean ± SD,

= 3].

0.75 0.79 0.72 0.74 0.73 0.81

influence of calcium-sequestrant phosphate ions. This argumentmay be explained by the observation of some turbidity, which wasappeared in the phosphate buffer might be due to the formationof calcium phosphate on the same analogy as mentioned by Pas-parakis and Bouropoulos [34]. Finally, the alginate beads begin todisintegrate, when calcium ions in the egg-box buckled structurediffuse out into the swelling medium. Similar ion exchange processmay occur in case of calcium coordinated carboxylic groups presentin the gum Arabic chain. These results clearly suggested that theoptimized calcium alginate/gum Arabic beads containing gliben-clamide might have the capability to swell slightly in the stomachat acidic pH as they subsequently move to the upper intestine andswell more in the intestinal alkaline pH.

4. Conclusion

Calcium alginate/gum Arabic beads containing glibenclamidewere successfully prepared by ionotropic gelation method andoptimized using central composite design. These optimized cal-cium alginate/gum Arabic beads are excellent combination of highdrug encapsulation efficiency (86.02 ± 2.97%), and suitable sus-tained drug release pattern over prolonged period (cumulative drugrelease after 7 h of 35.68 ± 1.38%), which could possibly be advan-tageous in terms of advanced patient compliance with reduceddosing interval. The technique for the preparation of calcium algi-nate/gum Arabic beads containing glibenclamide was found to besimple, reproducible, easily controllable, economical and consis-tent. Besides, the excipients such as sodium alginate, gum Arabicand calcium chloride used for the formulation of these beads werecheap and easily available.

Acknowledgements

Authors are thankful to Principal and Management of SeemantaInstitute of Pharmaceutical Sciences, Mayurbhanj-757086, India forproviding necessary facilities.

References

[1] T.R. Bharadwaj, M. Kanwar, R. Lal, A. Gupta, Drug Development and IndustrialPharmacy 26 (2000) 1025–1038.

[2] S. Hua, H. Ma, X. Li, H. Yang, A. Wang, International Journal of Biological Macro-molecules 46 (2010) 517–523.

[3] S. Racovita, S. Vasilu, M. Popa, C. Luca, Revue Roumaine de Chimie 54 (2009)709–718.

[4] J.S. Patil, M.V. Kamalapur, S.C. Marapur, D.V. Kadam, Digest Journal of Nanoma-terials and Biostructures 5 (2010) 241–248.

[5] A. Shilpa, S.S. Agarwal, A.R. Rao, Journal of Macromolecular Science PolymerReviews 43 (2003) 187–221.

[6] T.A. Davis, F. Llanes, B. Volesky, G. Diaz-pulido, L. Mccook, A. Mucci, AppliedBiochemistry and Biotechnology 110 (2003) 75–90.

[7] P. Smrdel, M. Bogataj, A. Mrhar, Scientia Pharmaceutica 76 (2008) 77–89.[8] M.M. Morshad, J. Mallick, A.K. Nath, M.Z. Uddin, M. Dutta, M.A. Hossain, M.H.

Kawsar, Bangladesh Pharmaceutical Journal 15 (2012) 53–57.[9] B. Singh, V. Sharma, D. Chauhan, Chemical Engineering Research and Design 88

(2010) 997–1012.

1 iologic

[[[

[

[[

[[

[

[

[

[

[[

[

[

[

[

[[[[

078 A.K. Nayak et al. / International Journal of B

10] A. Gursoy, F. Kalkan, I. Okar, Journal of Microencapsulation 15 (1998) 621–628.11] P. Liu, T.R. Krishnan, Journal of Pharmacy and Pharmacology 51 (1999) 141–149.12] Y.-N. Dai, P. Lin, J.-P. Zhang, A.-Q. Wang, Q. Wei, Journal of Biomedical Materials

Research Part B: Applied Biomaterials 86B (2008) 493–500.13] A.K. Nayak, S. Khatua, M.S. Hasanin, K.K. Sen, DARU Journal of Pharmaceutical

Sciences 19 (2011) 356–366.14] A.K. Nayak, M.S. Hasanin, S. Beg, M.I. Alam, ScienceAsia 36 (2010) 319–325.15] A.K. Nayak, D. Pal, International Journal of Biological Macromolecules 49 (2011)

784–793.16] D. Pal, A.K. Nayak, AAPS PharmSciTech 12 (2011) 1431–1441.17] S. Chakraborty, M. Khandai, A. Sharma, N. Khanam, C.N. Patra, S.C. Dinda, K.K.

Sen, Acta Pharmaceutica 60 (2010) 255–266.18] K.S.V. Krishna Rao, M.C.S. Subha, B.V.K. Naidu, M. Sairam, N.N. Mallikarjuna,

T.M. Aminabhavi, Journal of Applied Polymer Science 102 (2006) 5708–5718.19] M.R. Avadi, A.M.M. Sadeghi, N. Mohammadpour, S. Abedin, F. Atyabi, R.

Dinarvand, M. Rafiee-Tehrani, Nanomedicine: Nanotechnology, Biology andMedicine 6 (2010) 58–63.

20] P.S. Gils, D. Ray, P.K. Sahoo, International Journal of Biological Macromolecules46 (2010) 237–244.

21] B.H. Ali, A. Ziada, G. Blunden, Food and Chemical Toxicology 47 (2009) 1–8.

[[

[

al Macromolecules 51 (2012) 1070– 1078

22] A.A. Attama, O.J. Nwabunze, Acta Pharmaceutica 57 (2007) 161–171.23] J.R.D. Gupta, R. Irchhiaya, N. Garud, P. Tripathi, P. Dubey, J.R. Patel, International

Journal of Pharmaceutical Science and Drug Research 1 (2009) 46–50.24] M.D. Bhavsar, S.B. Tiwari, M.M. Amiji, Journal of Controlled Release 110 (2006)

422–430.25] G. Ye, S. Wang, P.W.S. Heng, L. Chen, C. Wang, International Journal of Pharma-

ceutics 337 (2007) 80–87.26] P. Costa, J.K.S. Lobo, European Journal of Pharmaceutical Sciences 13 (2001)

123–133.27] Y.S. Khotimchenko, V.V. Kovalev, O.V. Savchenko, O.A. Ziganshina, Russian Jour-

nal of Marine Biology 27 (2001) S53–S64.28] W.R. Gombotz, S.F. Wee, Advanced Drug Delivery Reviews 31 (1998) 267–285.29] D.A. Rees, Pure and Applied Chemistry 53 (1981) 1–14.30] E. Hamed, A. Sakr, Journal of Controlled Release 73 (2001) 329–338.31] M. Sarkar, P. Majumdar, Chemical Engineering Journal 175 (2011) 375–387.

32] A.K. Nayak, B. Laha, K.K. Sen, Acta Pharmaceutica 61 (2011) 25–36.33] J. Malakar, A.K. Nayak, D. Pal, International Journal of Biological Macro-

molecules 50 (2012) 138–147.34] G. Pasparakis, N. Bouropoulos, International Journal of Pharmaceutics 323

(2006) 34–42.