G Model

COLSUB-6552; No. of Pages 10 ARTICLE IN PRESS

Colloids and Surfaces B: Biointerfaces xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces

journal homepage: www.elsevier.com/locate/colsurfb

Promising dissolution enhancement effect of soluplus on crystallized

celecoxib obtained through antisolvent precipitation and high
pressure homogenization techniques
Alireza Homayouni a,b , Fatemeh Sadeghi a,b , Jaleh Varshosaz c ,
Hadi Afrasiabi Garekani b,d,∗∗ , Ali Nokhodchi e,f,∗
a
Targeted Drug Delivery Research Center, School of Pharmacy, Mashhad University of Medical Sciences, Mashhad, Iran
b
Department of Pharmaceutics, School of Pharmacy, Mashhad University of Medical Sciences, Mashhad, Iran
c
School of Pharmacy, Esfahan University of Medical Sciences, Esfahan, Iran
d
Pharmaceutical Research Center, School of Pharmacy, Mashhad University of Medical Sciences, Mashhad, Iran
e
Medway School of Pharmacy, University of Kent, Chatham Maritime, ME4 4TB Kent, UK
f
Drug Applied Research Center and Faculty of Pharmacy, Tabriz University of Medical Sciences, Tabriz, Iran

a r t i c l e i n f o a b s t r a c t

Article history: Poor solubility and dissolution of hydrophobic drugs have become a major challenge in pharmaceuti-
Received 24 April 2014 cal development. Drug nanoparticles have been widely accepted to overcome this problem. The aim of
Received in revised form 21 July 2014 this study was to manufacture celecoxib nanoparticles using antisolvent precipitation and high pres-
Accepted 22 July 2014
sure homogenization techniques in the presence of varying concentrations of soluplus® as a hydrophilic
Available online xxx
stabilizer. Antisolvent crystallization followed by freeze drying (CRS-FD) and antisolvent crystallization
followed by high pressure homogenization and freeze drying (HPH-FD) were used to obtain celecoxib
Keywords:
nanoparticles. The obtained nanoparticles were analyzed in terms of particle size, saturation solubility,
Celecoxib nanoparticles
Antisolvent crystallization
morphology (optical and scanning electron microscopy), solid state (DSC, XRPD and FT-IR) and dissolution
High pressure homogenization behavior. The results showed that celecoxib nanoparticle can be obtained when soluplus was added to the
Solid state analysis crystallization medium. In addition, the results showed that the concentration of soluplus and the method
Dissolution used to prepare nanoparticles can control the size and dissolution of celecoxib. Samples obtained in the
presence of 5% soluplus through HPH technique showed an excellent dissolution (90%) within 4 min. It is
interesting to note that celecoxib samples with high crystallinity showed better dissolution than those
celecoxib samples with high amorphous content, although they had the same concentration of soluplus.
DSC and XRPD proved that samples obtained via HPH technique are more crystalline than the samples
obtained through only antisolvent crystallization technique.
© 2014 Elsevier B.V. All rights reserved.

1. Introduction
It has been reported that approximately 40% of marketed drugs
and 70% of new drug candidates suffered from poor water solubil-
ity [1,2]. The oral bioavailability of these drugs could be improved
by enhancing the solubility and dissolution rate. Several strategies
have been developed to improve dissolution rate of poorly water
soluble drugs [3]. Nanoparticle engineering is one of the recent
strategies that has been used to improve the oral bioavailability of
∗ Corresponding author at: Medway School of Pharmacy, University of Kent,
Chatham Maritime, ME4 4TB Kent, UK. Tel.: +44 1634 202947.
∗∗ Corresponding author.
E-mail addresses: hadiafrasiabi@yahoo.com (H. Afrasiabi Garekani),
a.nokhodchi@kent.ac.uk (A. Nokhodchi).
poorly water soluble drugs [4]. A reduction in particle size increases
the effective surface area of particles and this in turn increases
the dissolution rate according to the Noyes–Whitney equation
[3]. Moreover, the reduction in the particle size of drugs down
to nanometer range enhances the saturation solubility due to the
Ostwald–Freundlich equation which is normally not observed for
micronized particles [5]. The preparation of nanoparticles can be
basically classified into two major methods, top-down and bottom-
up methods. In top-down methods nanoparticles are produced by
breaking down the large particles to smaller particle size using high
energy processes such as wet milling and high pressure homog-
enization [2,6]. In contrast, preparation of nanoparticles through
bottom-up methods starts from molecular level. In this method,
nanoparticles can be produced by precipitation/crystallization [7].
In precipitation technique simply the drug is dissolved in an organic
solvent and then the drug is precipitated by the addition of an

http://dx.doi.org/10.1016/j.colsurfb.2014.07.037
0927-7765/© 2014 Elsevier B.V. All rights reserved.

Please cite this article in press as: A. Homayouni, et al., Promising dissolution enhancement effect of soluplus on crystallized cele-
coxib obtained through antisolvent precipitation and high pressure homogenization techniques, Colloids Surf. B: Biointerfaces (2014),
http://dx.doi.org/10.1016/j.colsurfb.2014.07.037
G Model
COLSUB-6552; No. of Pages 10 ARTICLE IN PRESS
2 A. Homayouni et al. / Colloids and Surfaces B: Biointerfaces xxx (2014) xxx–xxx
antisolvent. One of the main problems of nanoparticles is the
crystal growth after the precipitation [8]. The presence of stabi-
lizer could be more useful to inhibit the crystal growth and also
aggregation of nanoparticles. Moreover, a hydrophilic stabilizer
could increase the wettability of hydrophobic drugs hence enhanc-
ing the water solubility and oral bioavailability [7]. In order to
control the particle size and prevent aggregation of nanocrystals
in suspension, combination of both techniques (bottom-up and
top-down) has recently been used. The Baxter company invented a
combination technology called “NANOEDGE” in which the precip-
itated nanocrystals was immediately subjected to high pressure
homogenization (HPH) to inhibit the crystal growth and break
the large agglomerates [2,9]. Polyvinyl caprolactam–polyvinyl
acetate–polyethylene glycol (soluplus® ) is a new graft copolymer
with amphiphilic properties that has been introduced by BASF.
This polymer could act as stabilizer and also as solubilizing agent
in formulation of poorly water soluble drugs. Soluplus has been
shown excellent solubilizing properties for poorly water soluble
drugs [10–12]. Furthermore, it has been reported that this polymer
was capable of increasing the oral bioavailability of danazol,
fenofibrate and itraconazol [13]. Similar results were reported for
bicalutamide nanocrystals prepared by antisolvent crystallization
using soluplus [14].
Celecoxib is a selective cyclooxygenase 2 inhibitor used for the
treatment of osteoarthritis, rheumatoid arthritis, and acute pain.
Furthermore, it is used as a combined therapy with chemotropic
agents in cancer treatments [15]. This drug is classified as a BCS
Class II drug (poor solubility, high permeability) with an aqueous
solubility of 1–3 ␮g/ml [16,17]. The poor solubility of celecoxib
(CLX) in gastrointestinal tract leads to an incomplete absorp-
tion and hence poor bioavailability. It has been reported that
the oral bioavailability of CLX capsule is 30% when administrated
to dogs [15]. Thus, increasing the solubility and dissolution rate
could enhance the oral bioavailability of CLX. Many formulation
approaches have been used to increase the solubility and dissolu-
tion rate of CLX in last few years [16–28]. These include solid dis-
persion with PVP, meglumine, HPMC, PVA, PEG, Kollicoat [18–21],
polymer induced crystallization using PVP, HPMC, Cremophor, SDS
[23,24,29], cocrystallization using nicotinamide [16], salt formation
in combination with crystallization inhibitor [25], complexation
with ␤-cyclodextrin [26] and nanoparticle formation [15,27,28]. To
the best of our knowledge there is no systematic study on the poten-
tial use of soluplus (SOL) on the solubility/dissolution enhancement
of CLX through making micro/nanocrystals. In other words, the aim
of the present study was to enhance the dissolution rate of CLX by
making micro/nanocrystals of CLX in the presence of soluplus as a
hydrophilic stabilizer using high pressure homogenization (HPH)
technique. In addition, a full comprehensive solid state analysis of
the resultant micro/nanoparticles was performed to elaborate the
results obtained in the present study.
2. Materials and methods
2.1. Materials
Celecoxib (CLX) was purchased from Arastoo chemical com-
pany, soluplus was donated from BASF (Germany) and sodium
dodecyl sulfate was obtained from Merck (Germany). All other sol-
vents and chemicals were of analytical grade.
2.2. Methods
2.2.1. Preparation of CLX:SOL freeze dried microparticles
precipitated by antisolvent crystallization method (CRS-FD)
Celecoxib and soluplus were dissolved in methanol and distilled
water respectively. Antisolvent crystallization was performed by
Please cite this article in press as: A. Homayouni, et al., Promising dissolution enhancement effect of soluplus on crystallized cele-
coxib obtained through antisolvent precipitation and high pressure homogenization techniques, Colloids Surf. B: Biointerfaces (2014),
http://dx.doi.org/10.1016/j.colsurfb.2014.07.037
adding 10 ml of the methanolic solution of CLX using a syringe at
a rate of 10 ml/min into 200 ml of water containing soluplus under
stirring condition (800 rpm at room temperature). The amount of
SOL was varied from 0.025 to 0.25% (w/v). The concentration of
methanolic solution was varied from 5 to 9.5% (w/v). Different sam-
ples of CLX:SOL were prepared to achieve final composition ratios of
95:5, 90:10, 75:25 and 50:50. After the addition of methanolic solu-
tion, CLX particles start precipitating immediately (the total time
for stirring after the addition of methanolic solution was 1 min).
After 1 min, the obtained suspension was immediately placed in
−80 ◦ C and freeze dried with the Heto freeze dryer (Denmark) for
48 h. Each crystallization experiment was repeated three times.
2.2.2. Preparation of CLX:SOL freeze dried nanoparticles
precipitated by antisolvent crystallization method followed by
high pressure homogenizer (HPH-FD)
The suspensions obtained from the antisolvent crystalliza-
tion (CRS) process were further homogenized by a high pressure
homogenizer (Avestin, Emulsiflex C3) at 500 and 1000 bars for 3 and
5 cycles respectively. The obtained suspension was immediately
freeze dried at −80 ◦ C for 48 h using the freeze dryer as described
above.
2.2.3. Preparation of physical mixtures of drug carrier
For better comparison between various samples, the physical
mixtures of CLX and SOL with the same ratios used in other methods
mentioned above were also prepared. To this end, the sieved frac-
tions (smaller than 250 ␮m) of the both CLX and SOL were mixed
using mortar and pestle for 20 min to obtain a uniform mixture. The
formulations were kept in glass vials until further studies.
2.2.4. Preparation of amorphous CLX
Amorphous CLX was prepared by melting the pure CLX at 170 ◦ C
under nitrogen atmosphere and then rapidly cooling it in an ice bath
as described by Andrews et al. [20]. The prepared amorphous CLX
was used in DSC analysis.
2.2.5. Morphological analysis
A scanning electron microscope (LEO 1450 VP, Germany) at an
acceleration voltage of 20 kV was used to investigate the morpholo-
gies of pure CLX and CLX:SOL samples. All freeze-dried samples
were coated with a thin gold-palladium layer by sputter coater
(SC 7620, England). Optical microscope (Olympus BX-60, Japan)
was also used to investigate the morphology of nanosuspensions
obtained from CRS and HPH methods before freeze drying.
2.2.6. Particle size measurement
To measure the size of crystals obtained through the anti-
solvent crystallization (CRS) and high pressure homogenization
(HPH) techniques, a nano-zetasizer (Malvern, UK) was used. The
nanosuspension was diluted with deionized water to approxi-
mately 1 mg/ml for each measurement. The measurements were
run three times to determine the mean and standard deviation of
Z-average size and zeta potential of the particles.
2.2.7. Solubility measurement
The saturation solubilities of pure CLX, physical mixtures
of CLX:SOL and treated CLX-SOL samples obtained via differ-
ent methods were determined. To this end, an excess amount
of samples were added to 15-ml screw-capped glass vials con-
taining 10 ml double-distilled water. The vials were shaken at
200 rpm in an air bath (25 ◦ C) for 48 h. Preliminary results showed
that 48 h was enough to reach the equilibrium condition. The
resulting suspensions were then filtered through a 0.22 ␮m filter
(Biofil MCE membrane). The concentration of CLX was determined
spectrophotometrically at 253 nm (Cecile 900, USA). The solubility
G Model
COLSUB-6552; No. of Pages 10 ARTICLE IN PRESS
A. Homayouni et al. / Colloids and Surfaces B: Biointerfaces xxx (2014) xxx–xxx 3

Table 1
Mean particle size, zeta-potential, glass transition temperature (Tg ), melting point and enthalpy of fusion (H) of pure CLX, samples prepared by antisolvent crystallization
(CRS) and samples prepared by antisolvent crystallization followed by high pressure homogenization (HPH) in the presence of various soluplus (SOL) ratios (data are the
mean and standard deviations of three determinations).

SampleCLX:SOL Z-average (nm) Zeta-potential 1st peak (◦ C)Tg /H (J/g) 2rd peak (◦ C)/H (J/g)

CLX 5528 ± 265 −23 – 163.39/90.26

CLX CRS – – – 162.91/95.73
95:5 CRS 1248 ± 451 −24 – 160.67/52.07
90:10 CRS 1474 ± 368 −21 – 161.59/35.83
75:25 CRS 1306 ± 344 −16 66.62/0.35 153.50/9.11
50:50 CRS 293 ± 221 −13 68.97/0.42 –
CLX HPH 514 ± 40 −26 – 163.75/94.59
95:5 HPH 872 ± 51 −22 – 162.99/55.78
90:10 HPH 577 ± 62 −17 – 160.67/42.86
75:25 HPH 677 ± 86 −15 – 155.42/11.70
50:50 HPH 439 ± 35 −14 – 161.75/3.47

results are the mean and standard deviation of three determina-
tions.
2.2.8. Differential scanning calorimetry (DSC)
DSC analysis was conducted to study the thermal behaviors of
the samples. To this end, DSC 822e (Mettler Toledo, Switzerland)
equipped with a refrigerated cooling system was used and the
instrument was calibrated using indium standard. Samples of
pure celecoxib and CLX:SOL micro/nanoparticle samples (2–3 mg)
were placed in aluminum pans sealed with a lid. The crimped
aluminum pans were placed inside the DSC and were scanned
from 20 to 200 ◦ C at a scanning rate of 10 ◦ C/min under nitro-
gen gas at a flow rate of 80 ml/min. To compare the thermal
behavior of various samples their onset temperatures, melting
points and endothermic and exothermic enthalpies were calcu-
lated using the software provided (STARe Ver. 10.00 Mettler Toledo,
Switzerland).
2.2.9. X-ray powder diffraction studies (XRPD)
X-ray powder diffraction patterns of the selected samples were
collected on a Bruker, D8 Advance, Germany diffractometer, with
Cu K␣ radiation (1.54056 Å). Each sample was scanned in the range
of 5–40◦ (2) with a step size of 0.05.
2.2.10. Fourier transform infrared spectroscopy (FT-IR)
FT-IR was used to investigate any changes in crystallized CLX
on the molecular level during the preparation of CLX:SOL particles.
Each sample was mixed with KBr at a ratio of 1:5 and was com-
pressed at a pressure of 7 tonnes for 2 min in a hydraulic press to
make disks for FT-IR. The spectrum was obtained using a Perkin
Elmer spectrum II (PerkinElmer, Waltham, MA). The obtained disk
(sample-KBr) was scanned against pure KBr disk at wavenumbers
ranging from 450 to 4000 cm−1 with a resolution of 1.0 cm−1 .
2.2.11. Dissolution studies
A USP dissolution apparatus 2 (paddle method) (Pharmatest,
Germany) was used to monitor the dissolution profiles of CLX
formulations. Each dissolution vessel was filled with 1000 ml dis-
tilled water containing 0.25% (w/v) sodium dodecyl sulfate (SDS)
equilibrated to 37 ◦ C and the paddles rotated at 50 rpm. An accu-
rate weight of samples equivalent to 40 mg CLX was placed at
the top of the dissolution medium and immediately the parti-
cles were sank to the bottom of the vessels. Samples were taken
from the vessels at selected time intervals through sintered fil-
ter by a peristaltic pump (Alitea, Sweden), and the concentration
of CLX in the samples assayed at 253 nm by a multi-cell changer
spectrophotometer (Shimadzu, Japan) based on a calibration curve
obtained for CLX at this wavelength. The dissolution of each sam-
ple was determined in triplicate. Soluplus showed no absorption
at 253 nm.
2.2.12. Dissolution parameters
In most cases it is more sophisticated and desirable to consider
the whole dissolution profile, since the use of a single point such as
t50% or t90% neglects all other data points. Therefore, the dissolution
efficiency (DE) and mean dissolution time (MDT) were calculated
to reflect the overall dissolution performance of various celecoxib
formulations. The DE is the area under the dissolution curve (y)

Table 2
Saturation solubility, dissolution efficiency (DE%) and mean dissolution time (MDT) of CLX, it physical mixtures with soluplus (PM), samples prepared by antisolvent crystal-
lization (CRS) and samples prepared by antisolvent crystallization followed by high pressure homogenization (HPH) in the presence of various ratios of celecoxib:soluplus
(data are the mean and standard deviations of three determinations).

SampleCLX:SOL Saturation solubility (␮g/ml) DE%90 min MDT (min)

CLX 2.1 ± 0.7 42.2 ± 0.4 17.4 ± 0.3

Amorphous CLX 2.2 ± 0.5 15.8 ± 0.3 31.8 ± 0.2
95:5 PM 2.2 ± 0.1 64.3 ± 1.4 19.0 ± 0.4
90:10 PM 2.5 ± 0.2 60.0 ± 0.9 23.1 ± 1.8
75:25 PM 2.7 ± 0.2 53.3 ± 0.8 25.9 ± 1.7
50:50 PM 3.2 ± 0.3 53.1 ± 1.8 31.2 ± 2.0
CLX CRS 2.2 ± 0.5 47.1 ± 0.5 17.2 ± 0.3
95:5 CRS 5.9 ± 0.3 91.4 ± 0.6 4.0 ± 0.5
90:10 CRS 6.5 ± 0.2 85.2 ± 0.7 8.7 ± 0.6
75:25 CRS 6.7 ± 0.1 79.6 ± 1.6 14.3 ± 0.4
50:50 CRS 8.2 ± 0.2 88.8 ± 1.8 6.2 ± 0.4
CLX HPH 3.9 ± 0.4 62.7 ± 0.4 11.6 ± 0.3
95:5 HPH 7.9 ± 0.1 95.1 ± 1.8 2.6 ± 1.2
90:10 HPH 8.0 ± 0.2 92.3 ± 3.4 4.7 ± 1.5
75:25 HPH 8.1 ± 0.2 88.8 ± 0.4 4.8 ± 0.2
50:50 HPH 8.1 ± 0.1 86.8 ± 1.3 6.2 ± 0.4

Please cite this article in press as: A. Homayouni, et al., Promising dissolution enhancement effect of soluplus on crystallized cele-
coxib obtained through antisolvent precipitation and high pressure homogenization techniques, Colloids Surf. B: Biointerfaces (2014),
http://dx.doi.org/10.1016/j.colsurfb.2014.07.037
G Model
COLSUB-6552; No. of Pages 10 ARTICLE IN PRESS
4 A. Homayouni et al. / Colloids and Surfaces B: Biointerfaces xxx (2014) xxx–xxx

Fig. 1. DSC traces of pure CLX, freeze dried samples prepared by antisolvent crystallization (CLX-CRS) and freeze dried samples prepared by antisolvent crystallization
followed by high pressure homogenization (CLX-HPH) obtained in the presence of different concentrations of soluplus.

up to a certain time t, expressed as a percentage of the area of the 3. Results and discussions
rectangle described by 100% dissolution in the same time [30].
t 3.1. Morphological analysis
0
y · dt
DE% = × 100
y100 · t Morphological analysis (optical microscopy and SEM) was per-
formed in order to investigate if any changes occurred in crystal
Another parameter that describes the rate of dissolution is the mean
habit of CLX particles during antisolvent crystallization or high
dissolution time (MDT); MDT is the statistical analyses that reflect
pressure homogenization (HPH) process. As it was shown in Figure
to the mean time of dissolution and has the advantage of being
S1-A (refer to supplementary materials), freshly crystallized CLX
applicable to all types of dissolution profiles. MDT is the most likely
without any stabilizer exhibited thin needle shaped crystals aggre-
time for a molecule to be dissolved from a solid dosage form or solid
gated to form star-shaped crystals. When this sample was subjected
particle and can be calculated using the following equations.
to the high pressure homogenization smaller particles with dif-

t̄ · Mi ti + ti+1 ferent morphologies were obtained (Figure S1-B, supplementary
MDT = i t̄ = Mi = (Mi+1 − Mi ) materials). CLX crystals obtained in the presence of 5% soluplus
Mi 2
showed different morphology (Figure S1-C, supplementary mate-
where t̄ is the midpoint of the time period during which the fraction rials). When this sample was subjected to homogenization through
M of the drug has been released from sample [30]. HPH, smaller and separate particles were obtained (Figure S1-D,

Please cite this article in press as: A. Homayouni, et al., Promising dissolution enhancement effect of soluplus on crystallized cele-
coxib obtained through antisolvent precipitation and high pressure homogenization techniques, Colloids Surf. B: Biointerfaces (2014),
http://dx.doi.org/10.1016/j.colsurfb.2014.07.037
G Model
COLSUB-6552; No. of Pages 10 ARTICLE IN PRESS
A. Homayouni et al. / Colloids and Surfaces B: Biointerfaces xxx (2014) xxx–xxx
Fig. 2. XRPD pattern of raw CLX, physical mixture and CRS and HPH samples pre-
pared by antisolvent crystallization (CRS) and samples prepared by antisolvent
crystallization followed by high pressure homogenization (HPH) in the presence
of various concentrations of soluplus.
supplementary materials). For better visualization (three dimen-
sional), SEM was performed after freeze drying which is discussed
below.
SEM image of pure CLX showed rod-shaped crystals (Figure
S2-A, supplementary materials) while samples obtained by antisol-
vent crystallization followed by high pressure homogenization and
freeze drying (HPH-FD) showed smaller particles with platy mor-
phology (Figure S2-B, supplementary materials). Freeze-dried CLX
samples obtained in the presence of 5% (Figure S2-C, supplementary
materials) or 10% (Figure S2-D, supplementary materials) soluplus
(CLX:SOL CRS-FD) exhibited needle shaped crystals. It appears that
the presence of SOL inhibited the crystal growth and the crys-
tal grew faster only in one dimension during the crystallization.
The results also showed that samples with higher ratio of soluplus
(above 25%) produced a big lump of aggregated particles (figures
not shown). It is interesting to note that when the above sam-
ples containing soluplus (5 and 10%) were subjected to HPH the
elongated particles became smaller (Figure S2-E, supplementary
materials) or in the case of 10% soluplus these elongated particles
disappeared and different particles with a different morphology
(irregular shape) was obtained (Figures S2-E and S2-F, supple-
mentary materials). This shows that the HPH method can break
down the elongated particles (micrometer size) to smaller particles
(nanometer size) to produce platy and rod-shaped nanoparticles
after freeze drying.
3.2. Particle size measurement
Particle size analysis was carried out to investigate the effect
of processing method (antisolvent crystallization or high pressure
homogenization) and soluplus concentration on the zeta potential
and size of the particles. The results in Table 1 showed that HPH
process generated smaller particles compared to simple antisol-
vent crystallization (CRS) technique in the presence of soluplus.
The same trend was obtained for the samples containing the
other concentrations of SOL. One exception in these samples was
CLX:SOL 50:50 CRS samples which showed smaller particles com-
pared to the samples subjected to HPH technique (Table 1). This
sample produced nanoparticles with the Z-average of 293 nm. This
value is remarkably lower than other CRS samples that could be
due to the inhibitory effect of SOL as a stabilizer during anti-
solvent crystallization. Soluplus is a graft copolymer containing
Please cite this article in press as: A. Homayouni, et al., Promising dissolution enhancement effect of soluplus on crystallized cele-
coxib obtained through antisolvent precipitation and high pressure homogenization techniques, Colloids Surf. B: Biointerfaces (2014),
http://dx.doi.org/10.1016/j.colsurfb.2014.07.037
5
polyvinyl caprolactam–polyvinyl acetate as a hydrophobic portion
and polyethylene glycol as a hydrophilic part. This amphiphilic
polymer could be adsorbed on the precipitated hydrophobic drug
surface, while the hydrophilic part provides stabilization in the
aqueous medium hence inhibiting the crystal growth [10]. Gener-
ally increasing the SOL concentration decreased the mean particle
size of the samples obtained via CRS and HPH methods. One should
be noted that due to the formation of gel like suspension the instru-
ment was not able to provide any reading for CLX crystallized
sample in the absence of soluplus.
3.3. Solubility measurement
Solubility measurement was carried out for pure CLX, physical
mixtures (PM) and all freeze dried (FD) samples. As it is shown in
Table 2 pure CLX showed very low solubility in water (2.1 ␮g/ml)
compared to the treated samples in the presence of SOL. In physi-
cal mixture samples (CLX:SOL), increasing the concentration of SOL
enhanced the saturation solubility of CLX that could be attributed
to the solubilizing effect of SOL [10]. Comparing the solubility data
of PM and crystallized samples (both CRS and HPH techniques)
demonstrated that antisolvent crystallization process in the pres-
ence of SOL has been more effective in enhancing the solubility
of CLX than its physical mixtures with CLX (Table 2). It seems the
mean particle size of samples plays a critical role in the solubil-
ity enhancement of CLX in the absence of SOL (for example the
saturation solubility of CLX HPH samples showed twofold higher
than untreated CLX). According to Freundlich–Ostwald equation
decreasing the particle size to nanometer range could increase the
saturation solubility [5]. As it was shown in Table 2, generally HPH
process could be more effective than simple antisolvent crystalliza-
tion technique to increase the saturation solubility of CLX except
when high concentration of SOL (50%) was used which there was no
significant difference in the solubility of these two samples (ANOVA
test; p > 0.05).
3.4. Differential scanning calorimetry (DSC)
DSC analysis was performed in order to investigate any interac-
tion between CLX and SOL or any changes in polymorphic form of
CLX during the crystallization process. Fig. 1 shows the DSC traces
of CLX obtained via CRS-FD and HPH-FD samples and their melting
points and enthalpy of fusion in Table 1. Pure CRS and HPH sam-
ples exhibited an endothermic peak around 163 ◦ C that is similar to
untreated CLX, and this is in agreement with previously published
data [31]. The results reported in Table 1 show that the presence of
SOL in freeze dried samples decreased the melting point of CLX and
the enthalpy of fusion which could be due to the partial dissolution
of CLX in polymer during heating process. On the other hand in
antisolvent crystallization process, SOL could attach to CLX particle
surfaces hence following freeze drying the SOL remained among
CLX particles. This occurrence caused smaller particle size of CLX
as well as fine dispersion of SOL among CLX particles. All of these
reasons could cause a reduction in melting point and enthalpy of
fusion of CLX. Similar results were obtained for antisolvent crys-
tallization of many API in presence of stabilizers [14,24,29,32]. As
it was shown in Table 1 CRS samples with 25 and 50% SOL exhib-
ited a small endothermic peak at 66.62 and 68.97 ◦ C that could be
attributed to the formation of amorphous CLX (these small peaks
are visible at larger smaller scale in Fig. 1). In previous studies it has
been demonstrated that amorphous CLX showed Tg around 59 ◦ C
[20,31]. In these samples deviation of endothermic peak (corre-
sponding to Tg of CLX) to higher temperature could be attributed
to the dispersion of CLX in SOL. SOL showed the Tg around 74 ◦ C
(data not shown). In HPH samples this small endothermic peak
did not appear. Moreover as it was shown in Table 1 and Fig. 1 in
G Model
COLSUB-6552; No. of Pages 10 ARTICLE IN PRESS
6 A. Homayouni et al. / Colloids and Surfaces B: Biointerfaces xxx (2014) xxx–xxx

Fig. 3. FT-IR spectra of various CLX samples obtained through antisolvent crystallization techniques (CRS) in the presence of different concentrations of soluplus (SOL).

most HPH samples the melting point and fusion enthalpy of CLX are
higher than similar CRS samples. These results demonstrated that
the crystalinity of CLX in HPH samples was higher than similar CRS
samples. In other words semi-crystalline and unstable fragile form
converted to more stable and crystalline form. This phenomenon
could be due to the high pressure homogenization process that has
been patented as Nanoedge® by Baxter Inc. [8,9,33]. However in
50% SOL CRS and 50% SOL HPH samples it can be concluded that
most of CLX converted to amorphous form.
3.5. X-ray powder diffraction studies (XRPD)
As it is shown in Fig. 2 X-ray diffraction pattern of untreated
CLX showed sharp peaks at 2 values of 5◦ , 10.5◦ , 16◦ and 21.5◦
which are diagnostic peaks for crystalline nature of CLX [31] while
SOL did not show any peak and exhibited amorphous state. CLX-
CRS and CLX-HPH samples showed similar pattern diffraction to
raw CLX but with less intensity which is an indication of less
crystallinity for crystallized CLX. Reduction in intensity of char-
acteristics peaks of these samples could also be attributed to the
smaller mean particle size [34]. Moreover, by increasing the ratio
of SOL in both CRS and HPH samples the intensity of peaks become
smaller so that in the crystallized samples (CRS and HPH) con-
taining 50% SOL the peaks of CLX completely diminished. This
indicates that the amorphous nature of the crystallized samples
in the presence of high concentration of SOL increases. This was
not the case for physical mixtures of CLX:SOL (50:50) as it shows
high crystallinity. The authors believe that in the crystallized sam-
ples (CRS or HPH) followed by freeze drying, CLX may disperse as
very fine particles (nm size) within the amorphous structure of
SOL leading to the observed amorphous structure. These results
were in good agreement with DSC studies. Careful comparison
between XRPD of samples obtained through CRS and HPH showed
that the intensity of peaks in HPH samples is slightly greater
than the intensity of CRS samples (Fig. 2). For example in 5% SOL
HPH sample the intensity of peaks at 2 of 10.5◦ , 16◦ and 21.5◦
increased approximately 7–13% compared to 5% SOL CRS sam-
ple. These increases in intensity could be due to the effect of
high pressure homogenization on crystallinity as described before
(Table 1).

Please cite this article in press as: A. Homayouni, et al., Promising dissolution enhancement effect of soluplus on crystallized cele-
coxib obtained through antisolvent precipitation and high pressure homogenization techniques, Colloids Surf. B: Biointerfaces (2014),
http://dx.doi.org/10.1016/j.colsurfb.2014.07.037
G Model
COLSUB-6552; No. of Pages 10 ARTICLE IN PRESS
A. Homayouni et al. / Colloids and Surfaces B: Biointerfaces xxx (2014) xxx–xxx 7

Fig. 4. FT-IR spectra of various CLX samples obtained through antisolvent crystallization techniques followed by high pressure homogenization (HPH) in the presence of
different concentrations of soluplus (SOL).

3.6. Fourier transform infrared spectroscopy (FT-IR)
The FT-IR studies were performed to investigate if any interac-
tion between CLX and SOL has been occurred in molecular level
during the process of crystallization. The FT-IR spectra of pure CLX,
SOL and all crystallized samples are shown in Fig. 3. The FT-IR spec-
tra of CLX showed characteristic peaks at 3341 and 3234 cm−1 that
attributed to N H stretching vibration of SO2 NH2 group, 1348 and
1165 cm−1 for the S O asymmetric and symmetric stretching and
1230 for C F stretching. These results were in good agreement with
previous data [31]. It has been previously reported that in amor-
phous CLX the doublet peak corresponded to NH2 group shifted
to 3388 and 3264 cm−1 . In addition the wavenumber of asymmet-
ric group of S O shifted from 1348 to 1340 cm−1 . These shifts are
attributed to the intermolecular hydrogen bonding in sulfonamide
group that does not occur in crystalline state of CLX [20,31]. The
spectra of SOL showed ester carbonyl stretching at 1739 cm−1 and
C O stretching for tertiary amide at 1643 cm−1 . On the basis of this
information Fig. 3 proves that all CRS samples prepared in the pres-
ence of SOL showed both amorphous and crystalline nature, while
in HPH samples (according to wavenumbers of NH2 groups) there
is no shift in the wavenumber of all HPH samples compared to crys-
talline CLX (Fig. 4). These results demonstrated that amorphous CLX
particles converted to crystalline state when CRS samples were sub-
jected to high pressure homogenization process. All these findings
were supported by DSC and XRPD data (Figs. 1 and 2).
3.7. Dissolution studies
The dissolution test was performed on all treated, untreated
and physical mixtures of CLX:SOL and their dissolution profiles are
shown in Fig. 5. The figure shows that in all cases the presence of
soluplus improved the dissolution of CLX. This dissolution enhance-
ment was more pronounced when 5% SOL was used regardless of

Please cite this article in press as: A. Homayouni, et al., Promising dissolution enhancement effect of soluplus on crystallized cele-
coxib obtained through antisolvent precipitation and high pressure homogenization techniques, Colloids Surf. B: Biointerfaces (2014),
http://dx.doi.org/10.1016/j.colsurfb.2014.07.037
G Model
COLSUB-6552; No. of Pages 10 ARTICLE IN PRESS
8 A. Homayouni et al. / Colloids and Surfaces B: Biointerfaces xxx (2014) xxx–xxx

Fig. 5. Dissolution profiles for different samples: (A) physical mixtures, mixtures of CLX and soluplus, (B) freeze dried samples prepared by antisolvent crystallization (CRS-
FD), and (C) freeze dried samples prepared by antisolvent crystallization followed by high pressure homogenization (HPH-FD) samples (n = 3, and error bars are standard
deviation).

Please cite this article in press as: A. Homayouni, et al., Promising dissolution enhancement effect of soluplus on crystallized cele-
coxib obtained through antisolvent precipitation and high pressure homogenization techniques, Colloids Surf. B: Biointerfaces (2014),
http://dx.doi.org/10.1016/j.colsurfb.2014.07.037
G Model
COLSUB-6552; No. of Pages 10 ARTICLE IN PRESS
A. Homayouni et al. / Colloids and Surfaces B: Biointerfaces xxx (2014) xxx–xxx
the technique used to prepare the samples. This could be due to the
hydrophilic nature of SOL resulting in a better wettability of the
samples. The results showed that the presence of SOL more than
5% caused a reduction in the dissolution of the drug from physical
mixtures of CLX:SOL, but its dissolution still is faster than the dis-
solution of pure CLX (Fig. 5A). This may be due to the formation of
a viscose gel around drug particles when the concentration of SOL
is above the optimum level. All these data indicated that although
the hydrophilicity of SOL should increase the dissolution rate of
CLX, the increase in the viscosity of the microenvironment around
drug particles produced by high concentration of SOL should not
be ignored.
The dissolution of amorphous CLX was also investigated and
the results showed that the dissolution of amorphous CLX is slower
compared to crystalline CLX (Fig. 5B). The slower dissolution rate of
amorphous CLX could be attributed to the devitrification of amor-
phous CLX when in contact to aqueous environment and hence
recrystallization to more hydrophobic structure that was reported
previously by Puri et al. [35]. Similar results have been reported
for capecitabine [36], felodipine [37] and diazepam [38]. In the
case of CRS-FD samples, all exhibited faster dissolution rate com-
pared to untreated CLX which could be due to the presence of
smaller particles for all treated samples (see Fig. 5B and Table 1). In
CRS-FD samples the 5 and 50% SOL samples showed higher disso-
lution compared to the samples obtained in the presence of 10 and
25% SOL. Although at higher concentration of SOL it was expected
slower dissolution compared to the samples containing 10 and
25% SOL (as described before in the case of physical mixtures), the
increase in the dissolution of this sample could be attributed to the
small size of the sample (293 nm) which could overcome the nega-
tive effect of the viscosity on the dissolution of CLX. In case of 10 and
25% SOL as they showed similar range of particle size then the effect
of particle size on their dissolution should be the same, therefore,
the negative effect of viscosity on dissolution would be more dom-
inant. In addition to the viscosity effect, the effect of amorphous
content of CLX in the samples should not be neglected as it has
been shown that the dissolution of amorphous CLX is poorer than
the dissolution of crystalline CLX (Fig. 5B). XRPD results showed
that the crystallinity of CRS samples containing 5% SOL is higher
than CLX samples containing 10 and 25%. Therefore, the dissolu-
tion rate and dissolution efficiency (DE) is higher than other CRS
samples. In the case of 50% SOL CRS sample the hydrophilicity of
SOL could overcome the negative effect of amorphous nature.
The increase in dissolution rate of CRS samples with 50% SOL
could also be due to very small size of CLX particles (Table 1;
293 nm) which could overcome the negative effect of viscosity of
high concentration of SOL in the dissolution of CLX (Fig. 5B). The
figure also shows that CLX-CRS sample in the absence of SOL did
not show any remarkable increase in the dissolution compared to
the untreated CLX. This indicated that the presence of SOL during
the crystallization is essential to improve the dissolution rate of
CLX.
Fig. 5C shows that HPH-FD samples obtained in absence or
the presence of SOL showed higher dissolution rate compared to
untreated CLX that could be due to the smaller particle size of these
samples compared to raw CLX (Table 1). This trend was backed up
by the percentage dissolution efficiency (DE) and mean dissolution
time (MDT) reported in Table 2. Comparing DE values of samples
obtained via CRS and HPH showed the better performance of parti-
cles obtained through HPH method (Table 2) particularly when the
concentration of SOL was 0, 5, 10 and 25%. Several studies have been
demonstrated the high efficiency of HPH process for enhancing
the dissolution rate of several drugs [15,27]. The high dissolution
rates obtained for HPH-FD samples compared to CRS-FD samples
could be attributed to the smaller particle size of HPH-FD sam-
ples (Table 1). Furthermore, the conversion of amorphous CLX to
Please cite this article in press as: A. Homayouni, et al., Promising dissolution enhancement effect of soluplus on crystallized cele-
coxib obtained through antisolvent precipitation and high pressure homogenization techniques, Colloids Surf. B: Biointerfaces (2014),
http://dx.doi.org/10.1016/j.colsurfb.2014.07.037
9
crystalline state after high pressure homogenization (XRPD, Fig. 2)
could be another reason for the higher dissolution rate of HPH-FD
samples compared to their counterpart samples obtained through
CRS. The results of one-way analysis of variance (ANOVA) showed
significant difference between DE% and MDT of CRS-FD and HPH-FD
samples (p < 0.05) with the exception of 50% SOL samples. It can be
concluded that CLX samples obtained through HPH technique show
better dissolution performance compared to the samples obtained
via antisolvent precipitation technique.
4. Conclusion
It was shown that soluplus is a potential stabilizer to be used
in the crystallization of celecoxib by antisolvent crystallization
and high pressure homogenization to produce nanoparticles with
improved dissolution rate. The concentration of soluplus during
the crystallization can modulate the drug dissolution and also the
particle size of celecoxib. Solid state analysis of celecoxib samples
showed that the method used to prepare celecoxib particles had
a big impact on the crystallinity of the samples and the particles
obtained through HPH showed higher crystallinity compared to the
samples obtained via a simple antisolvent crystallization technique.
The results showed that the dissolution of celecoxib nanoparticels
is controlled by particle size, viscosity of the microenvironment
around the drug particles and solid state of celecoxib. By controlling
these three parameters the dissolution rate of celecoxib particles
can be optimized to achieve the highest dissolution rate.
Acknowledgments
Financial support of Vice Chancellor for Research of Mashhad
University of Medical Sciences, Iran (Grant No. 900213) is greatly
appreciated. The authors are also thankful to Pasaddak Company,
Iran, for providing gift sample of Soluplus® .
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.colsurfb.
2014.07.037.
References
[1] E. Merisko-Liversidge, G.G. Liversidge, E.R. Cooper, Eur. J. Pharm. Sci. 18 (2003)
113–120.
[2] R. Shegokar, R.H. Müller, Int. J. Pharm. 399 (2010) 129–139.
[3] A. Singh, Z.A. Worku, G. Van den Mooter, Expert Opin. Drug Deliv. 8 (2011)
1361–1378.
[4] J. Hu, K.P. Johnston, R.O. Williams 3rd, Drug Dev. Ind. Pharm. 30 (2004) 233–245.
[5] F. Kesisoglou, S. Panmai, Y. Wu, Adv. Drug Deliv. Rev. 59 (2007) 631–644.
[6] L. Gao, D. Zhang, M. Chen, J. Nanopart. Res. 10 (2008) 845–862.
[7] A.A. Thorat, S.V. Dalvi, Chem. Eng. J. 181/182 (2012) 1–34.
[8] B. Sinha, R.H. Müller, J.P. Möschwitzer, Int. J. Pharm. 453 (2013) 126–141.
[9] C.M. Keck, R.H. Muller, Eur. J. Pharm. Biopharm. 62 (2006) 3–16.
[10] H. Hardung, D. Djuric, S. Ali, Drug Deliv. Technol. 10 (2010) 20–27.
[11] Z.K. Nagy, A. Balogh, B. Vajna, A. Farkas, G. Patyi, A. Kramarics, G. Marosi, J.
Pharm. Sci. 101 (2012) 322–332.
[12] N.K. Thakral, A.R. Ray, D. Bar-Shalom, A.H. Eriksson, D.K. Majumdar, AAPS
PharmSciTech 13 (2012) 59–66.
[13] M. Linn, E.-M. Collnot, D. Djuric, K. Hempel, E. Fabian, K. Kolter, C.-M. Lehr, Eur.
J. Pharm. Sci. 45 (2012) 336–343.
[14] V.B. Pokharkar, T. Malhi, L. Mandpe, Pharm. Dev. Technol. 18 (2013) 660–666.
[15] A. Dolenc, J. Kristl, S. Baumgartner, O. Planinsek, Int. J. Pharm. 376 (2009)
204–212.
[16] J.F. Remenar, M.L. Peterson, P.W. Stephens, Z. Zhang, Y. Zimenkov, M.B. Hickey,
Mol. Pharm. 4 (2007) 386–400.
[17] O.A. Abu-Diak, D.S. Jones, G.P. Andrews, Mol. Pharm. 8 (2011) 1362–1371.
[18] E.A. Fouad, M. El-Badry, G.M. Mahrous, F.K. Alanazi, S.H. Neau, I.A. Alsarra, Drug
Dev. Ind. Pharm. 37 (2011) 1463–1472.
[19] P. Gupta, V.K. Kakumanu, A.K. Bansal, Pharm. Res. 21 (2004) 1762–1769.
[20] G.P. Andrews, O. Abu-Diak, F. Kusmanto, P. Hornsby, Z. Hui, D.S. Jones, J. Pharm.
Pharmacol. 62 (2010) 1580–1590.
G Model
COLSUB-6552; No. of Pages 10 ARTICLE IN PRESS
10 A. Homayouni et al. / Colloids and Surfaces B: Biointerfaces xxx (2014) xxx–xxx

[21] J. Lee, M. Kim, H. Yoon, C. Shim, H. Ko, S. Cho, D. Lee, G. Khang, J. Pharm. Invest.
43 (2013) 205–213.
[22] T. Lee, C.W. Zhang, Pharm. Res. 25 (2008) 1563–1571.
[23] V.R. Gupta, S. Mutalik, M.M. Patel, G.K. Jani, Acta Pharm. 57 (2007) 173–184.
[24] M. Nasr, AAPS PharmSciTech 14 (2013) 719–726.
[25] H.R. Guzmán, M. Tawa, Z. Zhang, P. Ratanabanangkoon, P. Shaw, C.R. Gard-
ner, H. Chen, J.-P. Moreau, Ö. Almarsson, J.F. Remenar, J. Pharm. Sci. 96 (2007)
2686–2702.
[26] M.N. Reddy, T. Rehana, S. Ramakrishna, K.P. Chowdary, P.V. Diwan, AAPS J. 6
(2004) 68–76.
[27] S. Wang, Y. Liu, C. Sun, Y. Hao, T. Jiang, L. Zheng, J. Pharm. Pharm. Sci. 13 (2010)
589–606.
[28] A. Tan, S. Simovic, A.K. Davey, T. Rades, C.A. Prestidge, J. Control. Release 134
(2009) 62–70.
[29] H. Lee, J. Lee, J. Cryst. Growth 374 (2013) 37–42.
[30] P. Costa, J.M. Sousa Lobo, Eur. J. Pharm. Sci. 13 (2001) 123–133.
[31] G. Chawla, P. Gupta, R. Thilagavathi, A.K. Chakraborti, A.K. Bansal, Eur. J. Pharm.
Sci. 20 (2003) 305–317.
[32] J. Varshosaz, R. Talari, S.A. Mostafavi, A. Nokhodchi, Powder Technol. 187 (2008)
222–230.
[33] B.E. Rabinow, Nat. Rev. Drug Discov. 3 (2004) 785–796.
[34] H. Al-Hamidi, A.A. Edwards, M.A. Mohammad, A. Nokhodchi, Colloids Surf. B:
Biointerfaces 81 (2010) 96–109.
[35] V. Puri, A.K. Dantuluri, M. Kumar, N. Karar, A.K. Bansal, Eur. J. Pharm. Sci. 40
(2010) 84–93.
[36] J. Meulenaar, J.H. Beijnen, J.H. Schellens, B. Nuijen, Int. J. Pharm. 441 (2013)
213–217.
[37] Z.A. Langham, J. Booth, L.P. Hughes, G.K. Reynolds, S.A.C. Wren, J. Pharm. Sci.
101 (2012) 2798–2810.
[38] D.J. van Drooge, W.L. Hinrichs, H.W. Frijlink, J. Control. Release 97 (2004)
441–452.

Please cite this article in press as: A. Homayouni, et al., Promising dissolution enhancement effect of soluplus on crystallized cele-
coxib obtained through antisolvent precipitation and high pressure homogenization techniques, Colloids Surf. B: Biointerfaces (2014),
http://dx.doi.org/10.1016/j.colsurfb.2014.07.037