The Scien i ic Wo ld Jou nal
Volume 2012, A icle ID 605610, 15 pages
doi:10.1100/2012/605610
The cien i icWo ldJOURNA
L
Resea ch A icle
Robus Op imiza ion o Algina e-Ca bopol
940 Bead Fo mula ions
J. M. L ´
opez-Cacho,1Ped o L. Gonz´
alez-R,2B. Tale o,1
A. M. Rabasco,1andM.L.Gonz´
alez-Rod ´
ıguez1
1Depa men o Pha maceu ical Technology, Facul y o Pha macy, Uni e si y o Se ille, C/ P o esso Ga c´
ıa Gonz´
alez 2,
41012 Se ille, Spain
2Depa men o Indus ial Managemen , School o Enginee ing, Uni e si y o Se ille, C/ Camino de los Descub imien os s/n,
41092 Se ille, Spain
Co espondence should be add essed o M. L. Gonz´
alez-Rod ´
ıguez, [email p o ec ed]
Recei ed 31 Oc obe 2011; Accep ed 8 Decembe 2011
Academic Edi o : Ali Nokhodchi
Copy igh © 2012 J. M. L´
opez-Cacho e al. This is an open access a icle dis ibu ed unde he C ea i e Commons A ibu ion
License, which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly
ci ed.
Fo mula ion p ocess is a e y complex ac i i y which some imes implica es aking decisions abou pa ame e s o a iables o
ob ain he bes esul s in a high a iabili y o unce ain y con ex . The e o e, obus op imiza ion ools can be e y use ul o
ob aining high quali y o mula ions. This pape p oposes he op imiza ion o diffe en esponses h ough he obus Taguchi
me hod. Each esponse was e alua ed like a noise a iable, allowing he applica ion o Taguchi echniques o ob ain a esponse
unde he poin o iew o he signal o noise a io. A L18 Taguchio hogonala aydesignwasemployed oin es iga e he
effec o eigh independen a iables in ol ed in he o mula ion o algina e-Ca bopol beads. Responses e alua ed we e ela ed
o d ug elease p o ile om beads ( 50% and AUC), swelling pe o mance, encapsula ion efficiency, shape and size pa ame e s.
Con i ma ion es s o e i y he p edic ion model we e ca ied ou and he ob ained esul s we e e y simila o hose p edic ed in
e e y p o ile. Resul s e eal ha he obus op imiza ion is a e y use ul app oach ha allows g ea e p ecision and accu acy o he
desi ed alue.
1. In oduc ion
The use o na u al polyme s o he design o d ug deli e y
sys ems has long been he subjec o g ea in e es du ing
he pas decades. Sodium algina e is a sodium sal o alginic
acid, a na u ally occu ing non oxic polysaccha ide ound in
ma ine b own algae. Algina e has been widely used as ood
and pha maceu ical addi i es, such as able disin eg an ,
hickening, and suspending agen . I con ains wo u onic
acids, α-L-gulu onic (G) and β-D-mannu onic acids (M),
andiscomposedo homopolyme icblocksandblocks
wi h an al e na ing sequence [1]. This polyme can o m a
e icula ed s uc u e when i con ac s wi h Ca2+ and hus
i has been used o p oduce sus ained elease pa icula e
sys ems o a ious d ugs, p o eins, and e en cells [2–4].
Gela ion occu s by an ionic in e ac ion be ween he calcium
ions and he ca boxyla e anions o G-G blocks as calcium
ions diffuse om he ex e nal sou ce in o he d ople o ming
a polyanionic mic ocapsule.
The addi ion o a polyca ion (poly-L-lysine o chi osan)
o he gela ion medium induces he o ma ion o polyan-
ionic-polyca ionic complexes, which s abilizes he ionic gel
ne wo k and educes he algina e pe meabili y [5,6].
The main ad an age o using algina e o encapsula e
d ugs is ha he algina e gela ion p ocess occu s unde e y
mild condi ions wi hou using high empe a u es o chemi-
cal c osslinking agen s. Ano he ad an age o using algina e
is ha he algina e gel can also be con e ed o sol by adding
chela ing agen s, such as Na+and EDTA. Howe e , he d ug
eleasing p ope ies o Ca-sodium algina e ma ices suffe
om some se ious p oblems. Fi s ly, he d ugs could be
leaked du ing he gel o ma ion due o he long imme sion
ime, which dec eased he encapsula ion efficiency. Secondly,
he bu s elease o he d ugs om pu e Ca-sodium algina e
beads is se e e due o he quick b eakdown o beads in he in
i o elease p ocess. Cu en ly, much effo has been made
o imp o ing he pe o mance o Ca-sodium algina e beads
as d ug deli e y ca ie s.
2The Scien i ic Wo ld Jou nal
Table 1: Some o he ac o s affec ing he encapsula ion p ocess o d ugs in o algina e beads.
Fac o D ug Effec Re e ence
Rela ed wi h he o mula ion
Gliclazide
Algina e concen a ion
+EE
+ diame e
−po osi y [7]
− elease a e
+ swelling
BSA + EE [5,8]
Glucose
Volume o he algina e
solu ion Gliclazide −diame e [7]
Di alen ca ion
concen a ion
BSA
−EE [5,8,9]
VEGF
Glucose
C osslinke /algina e a io
Neu al (xan han
gums, e hylcellulose,
PVA)
Diclo enac sodium +EE [10,11]
+ swelling
Ca ionic (chi osan, poli-L-lysine) P o eins
Oligonucleo ides +EE [3,5,12]
Waxy compounds (Comp i ol 888 ATO, P eci ol ATO
5, Geluci es, glyce yl palmi os ea a e, magnesium
aluminum silica e)
Tamsulosin,
diclo enac sodium
+EE
− elease a e
[13–16]
Anionic (CMC,
Eud agi L100,
dex an sulpha e) and
ca ionic
Insulin
+EE
+ diame e
− elease a e
Su ac an p esence Gliclazide −diame e [7]
+EE
Rela ed wi h he p ocess
Needle size BSA −diame e [5]
Coaxial ai olume BSA −diame e [5]
D ying p ocess BSA [5]
S i ing a e Gliclazide −EE [7]
−diame e
The e o e, many ac o s a e in ol ed in he o mula ion
o algina e mic osphe es. Some o hem a e summa ized in
Table 1.
D ug elease om calcium-algina e beads depends on
he swelling o he beads and he diffusion o he d ug
in he gel ma ix [17]. Al hough algina e beads do no
swell app eciably in acidic luid [18], he beads swell and
e ode/disin eg a e apidly in he in es inal luid, leading o
a quick elease o he loaded d ug wi hin a ew minu es
[6,19] and hence calcium algina e ma ix alone does no
seem sui able as an o al con olled elease sys em [20].
Polyme ic ma e ials ha e been widely used in o de o
con enien ly modi y and modula e he d ug elease om
con olled- elease mic opa icles. Howe e , a la ge numbe
o ac o s, including he chemical-physical p ope ies o he
aw ma e ials (bo h d ug and excipien s), he composi ion
and he ela i e amoun s o he componen s in he o mula-
ions, as well as he manu ac u ing p ocess pa ame e s, can
in luence he d ug elease beha io om he inal p oduc s
[6,7,21,22].
In he p esen s udy, we a emp ed o ein o ce calcium-
algina e beads con aining me hylene blue as a model d ug
by inco po a ing Ca bopol 940 as hyd ophilic polyme . This
is a poly ac ylic acid in anionic o m ha con ains many
ee hyd oxyl g oups. Since i can o m a gel, i has been
used in combina ion wi h o he polyme s o con ol he d ug
elease.
The e a e p e ious epo s abou he effec o ca bome s
on cha ac e is ics o calcium-algina e beads [23,24]bu no
o Ca bopol 940. The e o e, we in ended o in es iga e he
The Scien i ic Wo ld Jou nal 3
effec o his hyd ophilic and swelling polyme on elease
beha io o calcium-algina e beads.
The p epa a ion o calcium-algina e beads o d ug en-
apmen is a s epwise p ocedu e. The indus ial expe imen-
a ion includes a g ea numbe o pa ame e s and a ull
ac o ial design ha yield la ge numbe o expe imen s. Con-
en ional op imiza ion p ocedu es in ol e al e ing one inde-
penden ac o a a ime keeping all o he s emain cons an ,
which enables o assess he impac o hose pa icula pa am-
e e s on he p ocess pe o mance. These p ocedu es a e ime
consuming and cumbe some because equi e mo e expe -
imen s and canno p o ide in o ma ion abou he mu ual
in e ac ions o he pa ame e s [25]. To educe hese numbe s
o expe imen s, a small se o ials wi h all possible combi-
na ions is selec ed by using a ious s a is ical ools [26].
S a is ical expe imen al design me hods, such as he
ac ional ac o ial design [8], he Placke -Bu man me hod
[27], esponse su ace me hodology [28], he Box-Behnken
design [29,30], and he Taguchi me hod [31], p o ide a
sys ema ic and efficien plan o expe imen a ion o achie e
ce ain goals so ha many con ol ac o s can be simul ane-
ously s udied. In con as o he adi ional “one- ac o a
a ime” app oach, hese me hods can be used o examine
and op imize he ope a ional a iables while conside ing he
in e ac i e effec s among he con ol ac o s.
The Taguchi me hod ocuses on p oduc obus ness
agains uncon ollable in luences, o noise. The me hod is
designed o educe a iabili y, and op imize unc ion o pe -
o mance wi h he lowes cos componen s. Taguchi me hod
o o hogonal a ay design o expe imen (DOE) in ol es he
s udy o any gi en sys em by a se o independen a iables
( ac o s) o e a speci ic egion o in e es (le els). This
app oach also acili a es o iden i y he in luence o indi id-
ual ac o s, and es ablish he ela ionship be ween a iables
and ope a ional condi ions. In his me hodology, he desi ed
design is sough by selec ing he bes pe o mance unde
condi ions ha p oduce consis en pe o mance [32] and he
esul s om small-scale expe imen s a e alid o e he scale
up schedule [33].
Robus ness is essen ial in he p oduc ion o nea ly all
p oduc s. The a ia ion in he quali y o he p oduc may
ollow om en i onmen al ac o s and/o manu ac u ing
a iables ha canno be easily con olled (noise ac o s). In
addi ion o emphasizing ha he a iabili y in he quali y o
he p oduc and ac o s in e ac ions should be educed [34],
he Taguchi me hod uses he signal- o-noise a io (S/N ),
which is di ec ly ans o med om he quad a ic quali y
loss unc ion, as a measu e o de e mine he obus ness o a
p ocess. Thus, op imizing p ocess pa ame e s by he Taguchi
me hod is an a emp no only o b ing he a e age quali y
nea o he a ge alue, bu also o simul aneously minimize
he a ia ion in quali y [35]. The e o e, his me hod has been
ex ensi ely applied in indus y and i s applica ion caused
signi ican changes in se e al indus ies, in manu ac u ing
p ocesses and o al quali y con ol [36–38].
This echnique is somehow no e y well known o
op imiza ion o pha maceu ical p ocesses. This pape offe s a
apid and efficien me hodology o s udy and op imize pha -
maceu ical o mula ions, based on he Taguchi o hogonal
a ay design. To do his, as an example, we used algina e-
Ca bopol beads o mula ed by iono opic gela ion wi h
me hylene blue (MB) as hyd ophilic d ug model.
A e he selec ion o noise and con ol ac o s, we se-
lec ed he p o iles ( esponses) o op imize wi h he aim
o demons a e he applicabili y o his me hodology in
he op imiza ion o diffe en p o iles such as d ug elease,
swelling a e, o he mo phology o he beads. A e wa d,
he expe imen al design and da a analysis p ocedu e, we
selec ed he bes p o iles o Taguchi op imiza ion and hen
we compa ed he esul s wi h p edic ed ones by lineal
eg ession. Finally we compa ed he accu acy o he op i-
miza ion.
2. Ma e ials and Me hods
2.1. Ma e ials. Sodium algina e ( iscosi y 2% w/ : 250 cP)
and ie hanolamine (TEA) we e pu chased om Sigma
Ald ich. Calcium chlo ide (Pan eac, Spain) was selec ed as
ca ion dono sal . MB was used as hyd ophilic d ug model.
Ca bopol 940 (Aco a ma, Spain) was added o con ol he
d ug elease om he algina e beads. Sodium ipolyphos-
pha e and T is(hyd oxyme hyl) aminome hane (TRIS) we e
ob ained om Sigma (Spain).
2.2. P epa a ion o Beads. The beads we e p epa ed by ol-
lowing he ex usion/p ecipi a ion me hod [39,40]. B ie ly,
a sodium algina e aqueous solu ion con aining 0.01% (w/ )
MB and 2% (w/ ) o sodium algina e was p epa ed by a
simple mixing s ep. Ca bopol 940 was dispe sed in pu i ied
wa e and hen was added o he abo e solu ion. As a
unc ion o he expe imen , TEA was added o his solu ion.
The beads we e p epa ed by d opping he algina e-Ca bopol
solu ion (50 mL) con aining MB om a sy inge using a
pe usion pump (Pe uso m, B aun), h ough a 0.9 mm
diame e needle in o a gen ly s i ed 0.25 M calcium chlo ide
aqueous solu ion (IKA Eu os a ). Diffe en concen a ions o
Ca bopol, a diffe en d opping a e and s i ing a es we e
used, as was epo ed in Table 2. The ob ained hyd ogels we e
main ained in o he calcium chlo ide solu ion o diffe en
ime pe iods o comple e he chemical eac ion. The beads
we e collec ed by decan ing he calcium chlo ide solu ion,
washed wi h deionized wa e and d ied o a cons an weigh .
A his s age, diffe en d ying me hods we e used: oom
empe a u e (20◦C), o a y e apo a o ( educed p essu e,
acuum and 80◦C), o o en (100◦C).
2.3. Cha ac e iza ion o Beads
2.3.1. En apmen Efficiency. The MB con en in he d y sam-
ples was de e mined by dissol ing he beads, p e iously
weighed, in a ipolyphospha e solu ion (0.03 M; 100 mL).
The esul ing solu ion was il e ed and he d ug concen-
a ion was analyzed by UV spec opho ome y, a 664 nm,
using a UV- isible spec opho ome e (Hi achi U-2000).
En apmen efficiency was calcula ed as he pe cen age (w/w)
o he heo e ical d ug con en . Resul s we e based on ipli-
ca e de e mina ions.
4The Scien i ic Wo ld Jou nal
Table 2: Fac o s and hei co esponding le els implemen ed o he cons uc ion o L18 o hogonal a ay.
Fac o Le els used, ac ual (coded)
Low (−1) Medium (0) High (+1)
Independen a iables
A: addi ion o TEA No Yes
B: concen a ion o Ca bopol (% w/ ) 1 2 4
C: lux a e o polyme ic solu ion (mL/min) 1 2 3
D: s i ing a e o s i e ( pm) 100 300 500
E: eac ion ime o beads in o CaCl2solu ion (min) 5 30 60
F: d ying me hod Room Ro a y e apo a o O en
G: NaCl (% w/ ) 0 0.01 0.1
H: pH 7.2 7.6 8.0
2.3.2. In Vi o D ug Release S udies. The d ug elease p o iles
o beads we e de e mined using he USP 28 paddle me hod.
The dissolu ion medium was 900 mL (sink condi ions) o
T is-buffe solu ion adjus ed a diffe en pH and NaCl
concen a ion alues as a unc ion o he expe imen . The
s i ing speed was 50 pm and he empe a u e was main-
ained a 37 ±0.5◦C. Accu a ely weighed samples o 700 mg
beads we e used o he es . A selec ed ime in e als o
a pe iod o 420 min, aliquo s each o 3 mL we e wi hd awn
om de dissolu ion medium h ough a 0.45 μmmemb ane
il e and eplaced wi h an equi alen amoun o he esh
dissolu ion medium. Concen a ions o MB we e hen spec-
opho ome ically de e mined a 664 nm. Each expe imen
was ca ied ou in iplica e and he esul s we e a e aged.
T is-buffe solu ion was selec ed as he dissolu ion
medium because in he o ma ion o algina e gel, algina e
eac ion wi h calcium ions u ns e e sible wi h sodium ions,
and a disagg ega ion o he beads occu s [6]. We selec ed
his buffe o con ol swelling p ocess by modi ying NaCl
concen a ion.
2.3.3. Swelling P ocess. The amoun o abso bed wa e was
de e mined ollowing desc ibed me hods by many au ho s
[7,41,42]. In a i s s age, a ce ain numbe o beads,
accu a ely weighed (W0), we e placed in glass beake s
con aining 80 mL o is-buffe and NaCl concen a ion se
by DOE. Nex , he glass beake s we e placed in a he mos a ic
ba h a 25◦C and 150 pm. A ime in e als o 1 h, beads we e
il e ed and weighed (We).
Swelling pe cen age (Esw) was calcula ed using he ol-
lowing exp ession [43]:
ESW =We−W0
W0×100, (1)
whe e Weis he we weigh o beads and W0is he ini ial
weigh o beads.
2.3.4. Mo phological P ope ies o Beads. Mo phological
cha ac e is ics o beads we e obse ed by Scanning Elec on
Mic oscopy, SEM (Philips XL30). The samples we e spu e -
coa ed wi h Au/Pd using a acuum e apo a o (Edwa ds)
and examined a 20 kV accele a ing ol age.
The size and shape o he beads we e de e mined using
image analyze so wa e (Image P o Plus). The leng h and
b ead h we e measu ed om he digi al images o each
bead and i s size calcula ed om he a e age o hese wo
dimensions [44].
Fo each o mula ion, 30 beads we e andomly selec ed
o measu emen ; he esul s we e hen a e aged.
2.4. Op imiza ion Me hodology. Op imiza ion me hodology
adop ed in his s udy was di ided in o ou phases (wi h a -
ious s eps): planning, pe o mance, analysis, and alida ion.
The schema ic ep esen a ion o he designed me hodology
was depic ed in Figu e 1 (modi ied om [34]).
2.4.1. Phase I. Planning
(1) De ining he P oblem and he Goal. Fi s s ep in Phase I is
o de e mine he ac o s o be op imized in he p oduc ion
and cha ac e iza ion o algina e-Ca bopol beads ha ha e
c i ical effec on he d ug elease, encapsula ion efficiency,
and in he mo phological p ope ies o he beads. The goal
is o e alua e he effec o diffe en ac o s on he elabo a ion
o he beads and o ob ain he pe ec combina ion be ween
he le els o he ac o s o op imize hei s p ope ies.
(2) Iden i ying he Fac o s o Be Op imized. Six p ocess
pa ame e s ( lux a e, s i ing a e, incuba ion ime, d ying
mechanism, NaCl concen a ion, and pH o he dissolu ion
medium) and wo o mula ion pa ame e s (p esence o TEA
and Ca bopol concen a ion) we e e alua ed. The no mal
p ac ice is o expe imen wi h he easible ange, so ha he
a ia ion inhe en in he p ocess does no mask he ac o s
effec s. The con ols ac o and hei s le els a e exposed in
Table 2.
(3) Iden i ying he Tes Condi ions o Op imize. The espons-
es o be op imized om he cha ac e iza ion o algina e-
Ca bopol beads a e hose ha ha e c i ical effec on he
d ug elease, encapsula ion efficiency, and mo phological
p ope ies o he beads. Selec ed esponses a e exposed in
Table 3.
The Scien i ic Wo ld Jou nal 5
Valida ion expe imen s
Phase I. Planning
Indi idual
ac o
con ibu ion
Rela i e
ac o
in e ac ion
De e mina ion o
op imum le els o
ac o s
Es ablishmen o bes
op ional condi ions
o p ocess op imiza ion
Pe o mance
unde op ional
condi ions
Analyze he da a using DOE pack 2000 so wa e
P edic he pe o mance a hese le els
Iden i y he con ol ac o s and hei le els
Design he expe imen al ma ix
De ine he da a analysis p ocedu e
Pe o m he designed expe imen s
Iden i y he ac o s o be op imized
Iden i y es condi ions
De ining he p oblem and he objec i e
Phase II. Pe o mance
Phase IV. Valida ion
Phase III. Analysis
Figu e 1: Schema ic diag am o he s eps in ol ed in he Taguchi me hodology.
Table 3: Tes condi ions o op imize.
P o ile analyzed Selec ed a iable
D ug elease
Time o elease 50 pe cen o en ap-
men d ug ( 50%)
A ea unde cu e (AUC)
En apmen efficiency En apmen efficiency (EE)
Swelling pe o mance Swelling a e (SR)
Shape o he pa icles Shape ac o (SF)
Size o he pa icles Mean diame e (Dm)
(4) Expe imen al Design and Da a Analysis P ocedu e. The
undamen al p inciple o he Taguchi me hod is o imp o e
he quali y o a p oduc by minimizing he effec o he causes
o a ia ion wi hou elimina ing he causes [45]. Th ee majo
ools used in he Taguchi me hod a e he o hogonal a ays,
analysis o a iance (ANOVA), and he signal o noise a io
(S/N).
The Taguchi app oach p o ides an oppo uni y o selec
a sui able o hogonal a ay depending on he numbe o con-
ol ac o s and hei s le els [46]. The o hogonal a ays gi e
equal c edence o all he pa ame e s being in es iga ed while
ANOVA de e mines he con ibu ion o each pa ame e in
he design o he expe imen . This is a ac ional ac o ial
app oach and educes he numbe o expe imen s.
O hogonal a ay is a ma ix o numbe s a anged in
ows and columns. Each ow ep esen s he le el o ac o s
in each un and each column ep esen s a speci ic le el o a
ac o ha can be changed o each un.
S/N a io is an indica i e o quali y, and he pu pose
o he Taguchi expe imen is o ind he bes le el o each
ope a ing pa ame e o maximize he a io [47]. Analysis
o he expe imen al da a using he ANOVA and effec s o
he ac o s gi es he ou pu ha is s a is ically signi ican o
inding he op imum le els.
In he Taguchi me hod, quali y is measu ed by he
de ia ion o a cha ac e is ic om i s a ge alue, and a loss
unc ion [L(y)] is es ima ed o he de ia ion as L(y)=
k×(y−m)2,whe ekdeno es he p opo ionali y cons an ,
m ep esen s he a ge alue and yis he expe imen al alue
ob ained o each ail. In case o bigge and be e quali y
cha ac e is ics, he loss unc ion can be w i en as L(y)=
k×(1/y2) and he expec ed loss unc ion can be ep esen ed
by:
ELy=k·E1
y2,(2)
whe e E(1/y2)canbees ima ed omnnumbe o samples
as:
n
i=11/y2
i
n.(3)

6The Scien i ic Wo ld Jou nal
In o de o measu e he quad a ic loss unc ion, he
Taguchi in oduces he signal- o-noise (S/N ) a io o mea-
su ing he quali y h ough o hogonal a ay based expe i-
men s. S/N a ios ake in o accoun bo h he a iabili y in
he esponse da a and he closeness o he a e age esponse
o a a ge alue. The S/N a io cha ac e is ics can be di ided
in o h ee ca ego ies when he a iable is con inuous [48]:
(i) nominal he bes :
S
N=−10 logy2
i
νyn,(4)
(ii) smalle he be e :
S
N=−10 logiy2
i
n,(5)
(iii) bigge he be e :
S
N=−10 log1/iy2
i
n,(6)
whe e yis he a e age o obse ed da a, nis he numbe o
obse a ions, and yi he obse ed da a o he i h un. In he
nominal bes s udy, yand na e nominal alues hose a e o
be achie ed.
Fo each ype o cha ac e is ics, wi h he abo e S/N a io
ans o ma ion, he highe he S/N a io, he be e is he
esul . The uncon ollable ac o s which cause he unc ional
cha ac e is ics o a p oduc o de ia e om hei a ge alues
a e called noise ac o s. The o e all aim o quali y is o make
p oduc s ha a e obus wi h espec o all noise ac o s [49].
The design was pe o med by an L18 a ay. In he p esen
s udy, h ee le els o se en ac o s and wo le els o one ac o
(Table 2) we e conside ed and he size o expe imen a ion
was ep esen ed by symbolic a ay o L18 (which indica ed
18 expe imen al ials) (Table 2). The h ee le els o all he
ac o s ha e been assigned (excep TEA (21)), wi h a layou
o L18 (21×35).
2.4.2. Phase II. Pe o mance o Expe imen s. A e phase I,
in which he ac o s o be op imized, hei s le els and he
expe imen al ma ix we e selec ed, he expe imen s we e
ca ied ou by iplica e (Table 4) and he selec ed a iables
we e de e mined.
2.4.3. Phase III. Analysis o he Da a and Op imiza ion
(1) Analysis o Da a. A his s age, a s a is ical analysis o
a iance was pe o med o de e mine which es pa ame e s
we e s a is ically signi ican o e e y a iable. Mo eo e , he
in luence o each ac o on he diffe en esponse a iables
selec ed was de e mined.
(2) Op imiza ion. A e he analysis o he da a, we p o-
ceeded o de e mine he op imal se pa ame e con igu a-
ion. Op imiza ion p ocess was made applying he ma ginal
means me hodology (also called a e age a e o main effec s
analysis).
In o de o do ha , we ob ained he a e age alue o S/N
a io o e e y a iable in e e y le el o esponse and hen
he esul s we e plo ed in a g aphic o ma ginal means. In
he Taguchi me hod, we always ied o maximize he S/N
a io, so we selec he le els wi h he highe alues o S/N
a io in e e y ac o ha gi e us he maximum S/N inal a io.
The diffe ence in e e y esponse is he me hod ollowed o
he consecu ion o he S/N a io, as i was explained be o e.
S/N a ios o e e y a iable we e ob ained as Table 3 ex-
poses.
2.4.4. Phase IV. Valida ion Expe imen s. Wi h S/N a io and
ANOVA analysis, he op imal combina ion o he es ing
pa ame e s could be p edic ed o a 95% con idence le el.
Also we can p edic he u u e esul s ha e e y combina ion
is going o ob ain by eg ession o he main in luencing
ac o s acco ding o he ANOVA.
In o de o alida e his me hodology, alida ion expe -
imen s we e u he pe o med using he op imized condi-
ions ob ained o each esponse and hen he esul s will
be compa ed o he p edic ed by eg ession. I inal esul s
a e simila o he p edic ed, we could alida e he p oposed
me hodology.
3. Resul s and Discussion
3.1. ANOVA. The applica ion o he expe imen al design
me hodology has he objec i e o iden i ying he mos
signi ican ac o s in luencing he o mula ion o algina e-
Ca bopol beads con aining a wa e -soluble d ug, and es ab-
lishing he bes le els o op imizing diffe en esponse
a iables.
Resul s ob ained o hese pa ame e s, exp essed as mean
alues, a e p esen ed in Table 4. ANOVA could be used as
a ool o hede ec iono maineffec s. Da a epo ed
in Table 5 e ealed ha he elease pa ame e s (i.e., 50%
and AUC o elease p o iles) we e signi ican ly affec ed
by Ca bopol concen a ion and addi ion o TEA in o he
polyme ic solu ion. Also, he p esence o TEA signi ican ly
in luenced on he en apmen efficiency and he pa icle size,
whe eas he swelling pe cen age was also affec ed by he
NaCl concen a ion in o he dissolu ion medium. Finally, he
mo phological pa ame e s, exp essed as shape ac o , we e
mainly affec ed by he d ying p ocedu e and he s i ing a e
applied om he s i e appa a us.
Con ibu ion deg ees o each ac o o e e y esponse
ha e been g aphically showed in Figu e 2. These indings will
be discussed in nex sec ions.
3.1.1. Release P o iles: Time o Release 50%o En apmen
D ug ( 50%) and A ea unde he Cu e (AUC). As shown
in Table 5, all ac o s excep he NaCl concen a ion exe
signi ican in luence on 50%. In he case o AUC, NaCl con-
cen a ion nei he pH had in luence on his esponse. F om
Figu es 2(a) and 2(b), we can conclude ha he Ca bopol
concen a ion ( ac o B) had he mos signi ican effec on
The Scien i ic Wo ld Jou nal 7
Table 4: Expe imen al layou using a L18 o hogonal a ay and esul s ob ained in e e y un.
Run 50% AUC EE SR SF Dm
Mean S/N Mean S/N Mean S/N Mean S/N Mean S/N Mean S/N
1A−1B−1C−1D−173.00 28.82 268.68 49.11 48.43 37.04 2863.78 69.14 0.52 −5.74 1.34 −2.71
E−1F−1G−1H−1
2A−1B−1CoDo50.00 27.96 421.78 49.29 54.88 38.26 1937.37 65.74 0.20 −13.88 1.47 −3.43
EoFoGoH0
3A−1B−1C1D169.33 35.57 117.23 48.60 53.63 38.08 1320.67 62.42 0.25 −12.06 1.53 −3.70
E1F1G1H1
4A−1BoC−1D−191.33 31.94 461.89 48.23 54.88 38.17 1681.62 64.51 0.58 −4.77 1.72 −2.96
EoFoG1H1
5A−1BoCoDo102.33 32.18 405.74 48.66 59.00 38.92 2177.52 66.76 0.27 −11.32 1.64 −4.50
E1F1G−1H−1
6A−1BoC1D1198.33 25.60 348.62 47.04 53.51 38.08 1119.52 60.98 0.65 −3.83 1.69 −4.61
E−1F−1GoH0
7A−1B1C−1Do118.67 19.17 354.48 48.22 54.34 38.21 1610.31 64.14 0.30 −10.47 1.51 −3.65
E−1F1GoH1
8A−1B1CoD1293.33 28.10 257.57 46.46 54.46 38.24 1082.83 60.69 0.25 −12.01 1.54 −3.73
EoF−1G1H−1
9A−1B1C1D−1106.00 18.80 427.85 47.90 57.14 38.65 1459.06 63.28 0.55 −5.26 1.66 −4.42
E1FoG−1H0
10 A1B−1C−1D160.00 24.77 448.50 48.89 59.73 39.04 2786.18 68.90 0.19 −14.68 2.07 −6.33
E1FoGoH−1
11 A1B−1CoD−182.67 22.18 634.32 48.71 59.33 38.97 1151.14 61.22 0.26 −11.95 1.83 −5.27
E−1F1G1H0
12 A1B−1C1Do119.33 40.29 401.70 47.93 56.07 38.49 3685.23 71.32 0.55 −5.20 1.57 −3.97
EoF−1G−1H1
13 A1BoC−1Do111.67 23.30 1326.03 47.92 58.83 38.91 1868.79 65.39 0.40 −9.30 1.73 −5.12
E1F−1G1H0
14 A1BoCoD1148.67 36.17 225.15 47.45 59.88 39.06 3003.21 69.55 0.20 −13.88 1.72 −4.70
E−1FoG−1H1
15 A1BoC1D−1209.33 36.72 469.47 47.14 59.49 39.00 2017.31 66.10 0.38 −8.50 1.56 −4.01
EoF1GoH−1
16 A1B1C−1D1238.67 40.29 1052.68 46.04 53.97 38.15 3094.41 69.81 0.35 −9.26 1.67 −4.54
EoF1G−1H0
17 A1B1CoD−1142.00 32.25 536.70 48.08 56.39 38.52 1554.57 63.83 0.34 −9.40 1.41 −3.04
E1F−1GoH1
18 A1B1C1Do138.33 33.61 841.84 47.79 60.02 39.08 1420.16 63.05 0.26 −11.71 2.81 −8.99
E−1FoG1H−1
elease cha ac e is ics. Howe e , his effec is posi i e in case
o 50% (97.11) and nega i e in case o AUC (−6316).
The in luence o he Ca bopol pe cen age on d ug elease
p o ile is due o he effec o his polyme on he s uc u e o
he beads, as was discussed by o he au ho s [50]. The densi y
o he beads inc eases as he concen a ion o his polyme
inc eases, sugges ing ha he beads o med a highe con-
cen a ions a e mo e compac and less po ous han hose
p epa ed a low polyme concen a ion [7].Thep esenceo
a highe ne wo k densi y in he s uc u e has been demon-
s a ed by o he au ho s in he p epa a ion o Ca bopol and
PVP mic osphe es [16,50]. Mo eo e , aking in o accoun
he swelling p ope ies o his hyd ophilic polyme , i is
con i med ha he p esence o he gel laye hickness ac s as
a ba ie o he pene a ion medium he eby e a ding he
diffusion o MB [7,11].
All hese con ibu e o p oduce a slowing in MB elease
a e (inc easing he 50%), making difficul he ou pu o he
d ug o he dissolu ion medium (diminishing he AUC).
3.1.2. En apmen Efficiency. Acco ding o ANOVA analysis,
he ac o wi h highe con ibu ion and s a is ical signi i-
cance on he en apmen capaci y o he d ug was he inco -
po a ion o TEA in o o mula ions.
TEA is a weak base ha ionizes ca boxylic g oups in he
chains o ca bome s. As a esul , he s uc u e o he gel
8The Scien i ic Wo ld Jou nal
Table 5: ANOVA esul s: effec o he a iables on 50%, AUC, % en apmen (EE), swelling pe cen age (SR), shape ac o (SF), and pa icle
size (DM).
Fac o s 50% AUC EE
F- a io PI-ha F- a io PI-ha F- a io PI-ha
A 89.52 <0.001 16.48 144.1 <0.001 −1892 11.2 0.004 5.57
B 2072 <0.001 97.11 1070 <0.001 −6316 0.27 0.607 1.07
C 132.5 <0.001 24.56 136 <0.001 −2252 1.41 0.25 2.42
D 564 <0.001 50.67 251.7 <0.001 −3063 0.003 0.955 −0.12
E 172.9 <0.001 −28.06 52.87 <0.001 1404 1.27 0.274 2.30
F 83.07 <0.001 −19.44 21.36 <0.001 892.2 2.18 0.157 3.01
G 0.01 0.918 −0.22 0.01 0.921 −19.4 0.67 0.425 1.66
H 213.4 <0.001 −31.17 3.20 0.082 345.2 0.53 0.474 −1.49
Fac o s SR SF DM
F- a io PI-ha F- a io PI-ha F- a io PI-ha
A 1842 <0.001 592 34.79 <0.001 −0.072 12.78 <0.001 0.255
B 1208 <0.001 −587.2 0.78 0.383 0.013 2.19 0.148 0.128
C 808.8 <0.001 −480.5 11.9 <0.001 0.051 2.19 0.148 0.130
D 274.4 <0.001 279.9 69.35 <0.001 −0.124 1.69 0.201 0.115
E 0.00 0.99 −0.22 4.49 0.041 −0.032 2.73 0.108 −0.142
F 62.8 <0.001 −133.9 100.4 <0.001 −0.149 0.80 0.377 0.077
G 5856 <0.001 −1293 23.4 <0.001 −0.072 8.78 0.005 0.261
H 25.09 <0.001 84.64 15.63 <0.001 0.059 8.15 0.007 −0.251
changes i s con o ma ion, going om a elaxed s a e in spi al
o m o a igid o m, which p o ides iscosi y and consis-
ency o he beads. This s uc u al con o ma ion inc eases
he en apmen efficiency. The same esul s ha e been ob-
ained by o he au ho s when hey added chi osan o he
algina e ma ix o p epa e beads [5].
Once he base was added, he inc ease o en apmen
efficiency is as consequence o he c ea ion o a highe
ne wo k densi y and he elec os a ic in e ac ion be ween he
d ug and he polyme s [16]. A he end o he p ocess, he
esul ing beads become ha de and hicke , a o ing en ap-
men efficiency inside he beads bu making difficul i s la e
elease.
3.1.3. Swelling Pe cen age. A e he ANOVA analysis, we can
de e mine ha p ac ically all ac o s ha e s a is ical in luence
on his dependen a iable. The ac o p esen ing he highe
in luence was NaCl concen a ion in he dissolu ion medium
ollowed by TEA addi ion and Ca bopol concen a ion.
Acco ding o ANOM es o he ac o NaCl concen-
a ion (da a no showed), he p esence o sodium ions
ac s hinde ing he swelling p ocess. Swelling pe cen age will
inc easeasNaClconcen a iondec eases(I-ha =−1293).
This esul can be explained on he basis o he swelling
mechanism o hese polyme ic s uc u es algina e-Ca bopol.
I is well accep ed ha he algina e beads in an aqueous
medium wi hou sodium ions do no swell well, bu when
hese ions a e p esen in he medium, he p e o med complex
be ween he algina e and calcium u ns uns able [51,52].
This could be explained by he iono opic phenomenon ha
occu s be ween he Ca2+ ions o he algina e beads and he
Na+ions p esen in elease medium ha inally p oduces a
cap u e o he Ca2+ by he chlo ine ions. This ac p oduces
a disagg ega ion o he bead s uc u e leading o e osion
and dissolu ion o he swollen beads and inc easing he d ug
elease a e [5,9,51]. This can be app ecia ed in Figu e 3,
whe e beads co esponding o ba ch 11 wi h highe NaCl
concen a ion ha e lowe swelling pe cen age a he same
ime han o he s.
In addi ion, TEA inco po a ion p oduces a igid s uc-
u e inside he beads and inc ease ca boxylic g oups wi h
nega i e cha ge. These g oups ha e a high affini y o wa e
molecules so mo e quan i y o wa e can ge inside he bead,
inc easing he swelling o he beads in ime and in ensi y.
These esul s we e isualized by pho og aphing he samples
du ing he swelling p ocess es s. We can no ice a highe
swelling in ba ches wi h TEA addi ion (Figu e 4) han in
o he lo s wi hou TEA a he same ime.
3.1.4. Shape Fac o . Acco ding o ANOVA analysis, he s i -
ing a e and he d ying me hod o beads a e he ac o s ha
ha e he highe in luence on his mo phological pa ame e .
S i ing a e has an impo an s a is ical signi icance con-
ce ning he shape and size pa ame e s [7]. When algina e-
Ca bopol d ops all in o CaCl2solu ion, s i ing a e p oduce
a o ce ha molds he inal mo phological aspec o he
beads, in ela ion o hei shape and su ace. As he s i ing
a e inc eases, beads will be less sphe ical and mo e i egula
in i s su ace, which p oduces an inc ease o shape ac o
and a dec ease in i s ci cula i y (in e se o SF), as we can
demons a e in Figu e 5.
F om he e alua ion o he d ying p ocess, we mus
keep in mind he cha ac e is ics o e e y ea men , because
hey can affec many p ope ies o he beads [40]. D ying
The Scien i ic Wo ld Jou nal 9
B∗
D∗
H∗E∗C∗A∗F∗G
×103
−5
5
15
25
35
45
55
65
75
85
95
SS o 50%
(a)
B∗
D∗
A∗C∗
E∗F∗H∗G
−2.5E + 07
2.5E + 07
7.5E + 07
1.25E + 08
1.75E + 08
2.25E + 08
2.75E + 08
3.25E + 08
3.75E + 08
4.25E + 08
SS o AUC
(b)
A∗
FCEGHBD
−100
0
100
200
300
400
500
SS o EE
(c)
G∗
A∗
B∗
C∗
D∗
F∗H∗E
(d)
F∗
D∗
A∗
G∗
H∗C∗E∗B
−0.1
−0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
SS o SF
(e)
A∗
G∗
H∗
EBCDF
−0.4
−0.2
0
0.2
0.4
0.6
0.8
1
1.2
SS o Dm
( )
Figu e 2: Sc ee plo ob ained o he a iables using he esul s o he ANOVA analysis. (a) Time o elease 50% o en apmen d ug, (b) a ea
unde he cu e, (c) en apmen efficiency, (d) swelling a e, (e) shape ac o , and ( ) Diame e medium. ∗Values wi h s a is ical signi icance
(α<0, 01). A: TEA; B: Ca bopol (% w/ ); C: lux a e (mL/min); D: s i ing a e ( pm); E: e en ion ime (min.); F: d ying me hod; G: NaCl
(% w/ ); H: pH.
(a) (b) (c)
Figu e 3: Digi al images co esponding o samples du ing swelling es s (a e 180 minu es). (a) ba ch 11 (0.1 M NaCl), (b) ba ch 15 (0.01 M
NaCl) and (c) ba ch 16 (wi hou NaCl).
Submi you manusc ip s a
h p://www.hindawi.com
Pain
Resea ch and T ea men
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
The Scien i ic
Wo ld Jou nal
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
Hindawi Publishing Co po a ion
h p://www.hindawi.com
Volume 2014
Toxins
Jou nal o
Vaccines
Jou nal o
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
An ibio ics
In e na ional Jou nal o
Toxicology
Jou nal o
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
S oke
Resea ch and T ea men
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
D ug Deli e y
Jou nal o
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
Ad ances in
Pha macological
Sciences
T opical Medicine
Jou nal o
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
Medicinal Chemis y
In e na ional Jou nal o
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
Addic ion
Jou nal o
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
BioMed
Resea ch In e na ional
Eme gency Medicine
In e na ional
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
Au oimmune
Diseases
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
Anes hesiology
Resea ch and P ac ice
Scien i ica
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
Jou nal o
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
Pha maceu ics
Hindawi Publishing Co po a ion
h p://www.hindawi.com Volume 2014
MEDIATORS
INFLAMMATION
o