Hybrid Hierarchical Porous Silica Templated in Nanoemulsions for Drug Release

Article abstract of DOI:10.1002/ejic.201501127

A new nanocarrier for loading and releasing drugs is reported, and ketoprofen was used as a model drug. More precisely, the carrier is a hybrid material prepared by combining oil-in-water (O/W) nanoemulsions, into which the drug has been solubilized, with mesostructured silica. This organic-inorganic hybrid material shows a controlled release of the drug that is pH dependent. If the drug is impregnated into the bare hierarchical meso-/macroporous dual silica material, obtained after the removal of the organic components by extraction, only 8 wt.-% of ketoprofen is released in a phosphate buffer medium (pH 7.4), probably owing to its low solubility in the aqueous phase. The drug solubility and release are increased strongly by the addition of Pluronic micelles to the receptor phase; this suggests a micelle-promoted and -assisted release mechanism. Whatever the vehicle, the release profiles of ketoprofen always follow the Korsmeyer-Peppas model with a diffusional release exponent value lower than 0.5, characteristic of a pseudo-Fickian release mechanism. Moreover, the release of ketoprofen is better controlled from the hybrid nanocarrier than from the hierarchical bare porous silica.

Full text of DOI:10.1002/ejic.201501127

DOI: 10.1002/ejic.201501127
Full Paper
CLUSTER
ISSUE
Mesoporous Silica
Hybrid Hierarchical Porous Silica Templated in Nanoemulsions
for Drug Release
[
a]
[b]
[b,c]
[a]
Philippe Riachy, Ferran Roig, Maria-José García-Celma,*
Marie-José Stébé,
[
a]
[c,d]
[c,d]
[a]
Andrea Pasc, Jordi Esquena,
Conxita Solans,
and Jean Luc Blin*
Abstract: A new nanocarrier for loading and releasing drugs is (pH 7.4), probably owing to its low solubility in the aqueous
reported, and ketoprofen was used as a model drug. More pre- phase. The drug solubility and release are increased strongly by
cisely, the carrier is a hybrid material prepared by combining the addition of Pluronic micelles to the receptor phase; this
oil-in-water (O/W) nanoemulsions, into which the drug has suggests a micelle-promoted and -assisted release mechanism.
been solubilized, with mesostructured silica. This organic–inor- Whatever the vehicle, the release profiles of ketoprofen always
ganic hybrid material shows a controlled release of the drug follow the Korsmeyer–Peppas model with a diffusional release
that is pH dependent. If the drug is impregnated into the bare exponent value lower than 0.5, characteristic of a pseudo-
hierarchical meso-/macroporous dual silica material, obtained Fickian release mechanism. Moreover, the release of ketoprofen
after the removal of the organic components by extraction, only is better controlled from the hybrid nanocarrier than from the
8
wt.-% of ketoprofen is released in a phosphate buffer medium hierarchical bare porous silica.
Introduction
anticancer drug and investigated the in vitro antitumor activity
of the PTX-loaded mixed micelles. The results showed an in-
creased in vitro cytotoxicity compared with those for the injec-
tion of the only available commercial preparation of PTX (Taxol)
The growing demand for more-efficient medicines forces the
adaption of drug delivery systems and efficient production
techniques, in particular, because of the low solubility of many
active pharmaceutic ingredients in biocompatible solvents. For
example, the water solubility of ketoprofen, which is effective
and well-tolerated in the treatment of acute and chronic pain
of both rheumatic and traumatic origin as well as postoperative
[
5]
and free PTX in human lung adenocarcinoma cell line A-549.
Microemulsions can also serve as reservoirs for bioactive mole-
[
6–9]
cules.
For example, in a paper dealing with the comparison
of emulsions and microemulsions for the encapsulation of hy-
droxytyrosol (HT), Xenakis et al. showed that emulsions provide
a better HT release than microemulsions, whereas the opposite
[
1,2]
–1
pain,
is only 0.0213 mg mL . Different strategies have been
developed to overcome this drawback, and one of the most
popular approaches is the entrapment of drug molecules in
inert vehicles. Various systems based on micelles, microemul-
sions, emulsions, or inorganic materials have been formulated
[
7]
effect is noted for the storage ability. Among the different
carriers, nanoemulsions have emerged as a promising drug de-
[
10–15]
livery approach.
Nanoemulsions are emulsions consisting
[
16,17]
of small droplets, typically in the 20–200 nm size range.
for this purpose. In particular, Pluronic block-copolymer mi-
Although their size in the lower range may overlap with that of
microemulsion droplets and they also appear transparent or
translucent, nanoemulsions are in fact quite different from true
celles are interesting[
3–5]
as they increase not only the solubility
but also the metabolic stability and the circulation of the drugs
and they inhibit drug efflux transporters in the blood–brain bar-
rier and the small intestines. Wei et al. used P123/F127 mi-
celles to encapsulate paclitaxel (PTX), which is a poorly soluble
[
18]
microemulsions.
As emulsions, nanoemulsions can possess
[
3]
[5]
kinetic stability and they do not form spontaneously, as their
free energy of formation is positive; consequently, they have a
limited lifespan, and their preparation generally requires an en-
ergy input. The properties of nanoemulsions depend not only
on the thermodynamic conditions (i.e., composition, tempera-
ture, and pressure) but also on the preparation method and,
crucially, the order of component addition. The formation of
nanoemulsions can be achieved by low-energy methods, in
which external input energy might not be required and, in con-
trast to high-energy methods, droplet size is not controlled. The
most common low-energy methods are those based on phase
[
[
a] Université de Lorraine/CNRS, SRSMC, UMR7565, Faculté des Sciences et
Technologies,
BP 70239, 54506 Vandoeuvre-lès-Nancy cedex, France
E-mail: Jean-Luc.Blin@univ-lorraine.fr
http://www.srsmc.univ-lorraine.fr
b] Department of Pharmacy and Pharmaceutical Technology and Institute of
Nanoscience and Nanotechnology (IN2UB), University of Barcelona,
Av/ Joan XXIII s/n, 08028 Barcelona, Spain
E-mail: mjgarcia@ub.edu
http://www.ub.edu/farmacia/es/
[
[
c] CIBER of Bioengineering, Biomaterials and Nanomedicine (CIBER-BBN),
Barcelona, Spain
d] Institute of Advanced Chemistry of Catalonia (IQAC-CSIC),
c/Jordi Girona 18-26 Barcelona, Spain
[
19,20]
inversion (phase inversion temperature method, PIT,
or
[
21]
phase inversion composition method, PIC) and the so-called
[
22,23]
self-emulsification methods,
which are usually based on
Eur. J. Inorg. Chem. 2016, 1989–1997
1989
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
dilution processes, diffusion processes, or both. Their need for ery.[41–43] The main advantage of these carriers is their increased
small amounts of surfactants, stability towards dilution, and mass transport owing to their macroporosity, and they maintain
ease of formulation make nanoemulsions low-cost drug deliv- a large surface area owing to their mesoporosity. This allows a
ery systems and favors their mass production. Their mechanical better adsorption of a wide array of drug molecules, and good
properties, optical properties, and droplet size can be con- diffusion pathways for the drug are maintained owing to the
trolled; many water-insoluble drug molecules can be incorpo- presence of larger macropores.
rated into these droplets, and their small size gives them the
advantage of deep tissue penetration; thus, they fulfil the crite- hierarchical bare silica obtained after the mineralization of the
ria of many drug administration routes.
nanoemulsions and the hybrid material to be used as matrixes
Here, we have investigated the ability of nanoemulsions, of
Another promising drug delivery approach is the adsorption for the loading and releasing of a model hydrophobic drug,
of drug molecules on the surfaces of porous silica materials. ketoprofen. We have characterized the various carriers in detail
Thanks to their properties such as high specific surface areas and shed some light on the ketoprofen release mechanism for
and narrow pore-size distributions, ordered mesoporous silica each of them.
materials, mainly MCM-41 or SBA-15, and mesoporous silica na-
noparticles (MSNs) have been widely considered for this appli-
[
24–30]
cation.
In 2001, Vallet-Regí et al. first reported the use of
Results and Discussion
MCM-41 with different pore diameters as a carrier for ibuprofen.
The authors showed that the mesopore diameter does not in-
fluence the release because of the small size of ibuprofen mole-
Characterization of the Carriers
[
24]
cules.
Various pharmaceutically active components have
Nanoemulsions were prepared according to the PIT method
and visualized by nanoparticle tracking analysis (NTA, Figure 1,
A and B). They have been characterized by small-angle X-ray
scattering (SAXS) and dynamic light scattering (DLS). From q =
been entrapped within mesoporous materials through adsorp-
tion or impregnation and released through a diffusion-con-
trolled mechanism.[
25,27]
The syntheses of these materials are
[
31–33]
performed through templating methods. Either chemical
–1
–4
0
.1 nm , a q decay over one decade, characteristic of an ab-
[
34–36]
or biological
templating approaches can be used. For the
rupt change of scattering length density over a large interfacial
area, is detected in the SAXS pattern (Figure 1, C).
chemical approaches, the preparation combines sol–gel chem-
istry and assemblies of surfactant molecules as framework tem-
The droplet size of the nanoemulsions can be determined
[
31–33]
plates.
By this method, ordered mesoporous silica or MSNs
[44]
from the Porod law, see below.
are obtained after surfactant removal, either by extraction or
by calcination. Therefore, one alternative method for the devel-
opment of the carriers is the incorporation of the active ingredi-
2
Sv(Qext – Qint)2
I(q) =
q4
[
37–40]
ent during the material synthesis.
For this strategy, the use SV is the area per unit volume of dispersion, Qext and Qint are
of micelles has the advantage of playing a dual role; firstly, they the scattering length densities of the continuous phase (water)
allow the solubilization of the active ingredient; secondly, they and internal phase (decane), respectively. From S , a mean value
V
act as templates for the formation of the mesopore network. of the radius (R) of the droplets can be derived from R = 3Φ/
This approach allows the entrapment of larger amounts of hy- SV, in which Φ is the dispersed volume fraction. According to
[
39]
drophobic drugs.
For example, in a study dealing with silica this equation, the nanoemulsion radius is 62 nm. For the 50:50
capsules enclosing P123 triblock copolymer micelles for flurbi- mixture of the nanoemulsion and P123 micellar solution, the
profen storage and release, Martens et al. showed that the mo- DLS measurements gave an average nanoemulsion droplet ra-
lecular environment of flurbiprofen in P123-filled silica capsules dius of 74 nm. This value is slightly higher than that obtained
is different from that for ordered mesoporous silica materials from SAXS. Indeed, DLS gives the hydrodynamic size, which is
synthesized with P123 as a template with the drug molecules higher than the size obtained by SAXS. We noticed that the
[
37]
adsorbed on the silica surface.
In addition to mesoporous micelles were not detected in the nanoemulsion/micelle mix-
materials, macroporous and hierarchical macro-/mesoporous ture; this is likely due to the intensity masking effect that the
silica have also been used more recently for drug deliv- large droplet sizes (the nanoemulsions) have on the smaller
Figure 1. (A) NTA image, (B) particle-size distribution profile from NTA, and (C) SAXS pattern of the nanoemulsion.
Eur. J. Inorg. Chem. 2016, 1989–1997
www.eurjic.org
1990
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 2. SAXS patterns (A) and nitrogen adsorption–desorption isotherm (B) with the corresponding mesopore size distribution (inset) of the bare porous
silica.
Figure 3. (A) TEM and (B) SEM micrographs and (C) macropore size distribution of the hierarchical bare porous silica.
ones (the micelles). Indeed, the micellar solutions have an aver- 1056, 954, and 795 cm–1 (Figure 4, a). The most intense band
–
1
age micelle radius of 19.1 nm, as measured by DLS.
centered at ν˜ = 1056 and the shoulder at ν˜ = 1217 cm are
(F ) stretching vibration. The band
ence of the P123 micelles, in addition to a sharp peak at at ν˜ = 795 cm corresponds to the SiO (A) stretching vibra-
After the mineralization of the nanoemulsions in the pres- associated with the SiO
ν
3
4
2
–1
ν
1
4
–
1
1
1.2 nm, two peaks at 6.5 and 5.6 nm are also detected on the tion. The band at ν˜ = 954 cm is related to the Si–(OH) stretch-
SAXS pattern of the recovered bare porous silica (Figure 2, A). ing of terminal silanol groups. By comparison of the spectrum
The presence of the latter peaks reveals a hexagonal organiza- of the silica-loaded ketoprofen (Figure 4, b) with those of the
tion of the mesopores in the material, as the peak position ratio pure ketoprofen (Figure 4, c) and the bare silica (Figure 4, a),
is 1:√3:2. Indeed, these peaks are attributed to the (110) and the presence of adsorbed ketoprofen can be confirmed unam-
(
200) reflections of the hexagonal structure. A type IV isotherm, biguously. Indeed, in the investigated wavenumber range and
characteristic of a mesoporous material, was obtained by nitro- by comparison with the spectrum of bare silica, the supplemen-
gen adsorption–desorption analysis (Figure 2B).
The specific surface area and pore volume are 782 m g
tary vibrations detected in Figure 4 (b) can be attributed to
ketoprofen. In particular, the bands at ν˜ = 1695 and 1651 cm
2
–1
–1
3
–1
and 1.2 cm g , respectively. The pore diameter distribution
calculated by the Barrett–Joyner–Halenda (BJH) method is quite
narrow and centered at 9.3 nm (Figure 2, B, inset). As shown by
both TEM and SEM micrographs, the silica particles exhibit a
macroscopic network structure with a size of a few hundred
nanometers (Figure 3, A and B). The macropore size distribution,
determined by mercury intrusion porosimetry measurements,
reveals the presence of macropores with two different size dis-
tributions centered at 240 nm and 6 μm (Figure 3, C). The first
one can be attributed to the fingerprint of the oil droplets of
the nanoemulsions, and the second peak (6 μm) might arise
from interparticle porosity induced by the aggregation of parti-
cles.
After impregnation, the presence of the drug was deter-
–
1
mined by infrared spectroscopy. In the ν˜ = 400–2000 cm do-
Figure 4. Infrared spectra of (a) bare silica, (b) ketoprofen-loaded silica, and
(c) pure ketoprofen.
main, the bare silica material exhibits bands located at ν˜ = 1217,
Eur. J. Inorg. Chem. 2016, 1989–1997
www.eurjic.org
1991
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
are due to the symmetric carbonyl vibration and to both the silane (TMOS) hydrolysis[49] and the subsequent oligomerization
[
45]
dimeric carboxylic and ketonic group stretching vibrations.
of these species also help maintain strongly acidic conditions.
–1
The peak at ν˜ = 1444 cm can be attributed to the presence The effect of the generated silicic acid species on pH has al-
of CH–CH deformation, and the C–H peaks corresponding to ready been reported for the synthesis of ordered mesoporous
3
[
50]
the out-of-plane deformation of substituted aromatics are ob- materials.
For example, when the synthesis occurs through
–1
served between ν˜ = 640 and 860 cm .
the cooperative templating mechanism with decaoxyethylene
cetyl ether [C (EO) ] micellar solutions at 10 wt.-% surfactant
1
6
10
under acidic conditions (pH < 5) no variation of pH is observed,
but a decrease of pH was found upon the addition of the inor-
ganic precursor at an initial pH value of more than 5. After the
addition of TMOS, the pH of the solutions decreased to 5.5, 6.5,
and 7 from initial values of 7, 8.5, and 10, respectively. There-
fore, in the study reported here, we can assume that the surface
acidic groups of the hybrid material will also modify the local
pH in the vicinity of the surface, in particular for phosphate-
buffered saline (PBS) solutions at pH 7 and 7.4; and this pH
modification will affect the solubility of ketoprofen and its re-
lease rate.
In Vitro Release of Ketoprofen from the Hybrid Material
The cumulative release profiles of ketoprofen from the hybrid
material were determined at four pH values of the receptor
solution and are depicted in Figure 5 (A).
It is clear that ketoprofen release is favored under acidic con-
ditions. Indeed, after 5 h, the quantity of discharged drug varies
from 43 to 53 % as the pH decreases from 7.4 to 4.5. The lowest
value of 39 % was obtained in deionized water. This behavior
is quite unexpected, as for the pK of ketoprofen of approxi-
a
[
46]
mately 4.2–4.4, the opposite would be expected, as was ob-
served for ketoprofen loaded in pure mesoporous silica SBA-15
material or ibuprofen in MCM-41.
As the nanocarrier is synthesized and mineralization occurs
Indeed, for a carboxylic around the nanoemulsion droplets that contain the drug in the
[
26]
[47]
acid drug, it has been shown previously that increasing pH presence of Pluronic micelles, we also looked at the release of
would enhance the solubility and consequently the amount re- ketoprofen from nanoemulsions with and without P123 mi-
leased, particularly at pH > pK of the acid. As the pH increases, celles in a PBS solution for comparison (Figure 5, B). To avoid
a
the deprotonation of the carboxylic acid groups should lead to any precipitation owing to acidic conditions, the pH of the re-
an increase of the ketoprofen solubility and, as a consequence, ceptor solution was fixed to 7.4. A fast release of ketoprofen of
a higher released amount would be expected. However, our ca. 80 % after 5 h from nanoemulsions without P123 micelles
experiments show the opposite trend. This surprising phenom- was observed. Notably, in the presence of P123 micelles, a rapid
[
26]
enon has already been reported by Laniecki et al.
and burst release occurs, even compared with that of the hybrid
[
48]
Van Speybroeck et al. for SBA-15. The authors explained that material. For example, after 4 h, 82 and 68 % of ketoprofen was
ketoprofen can precipitate locally in the matrix under acidic released from the nanoemulsions without and with Pluronic mi-
conditions. As some ketoprofen molecules that have not precip- celles, respectively. As both experiments were performed under
itated migrate into the receptor solution, the ionic equilibrium the same conditions (the same loading of drug in the nano-
is shifted and the precipitated ketoprofen can dissolve. As a emulsions and the same pH of the receptor solution), this indi-
consequence, further amounts of drug are released from the cates that a fraction of the drug molecules are solubilized in
pores of the support. Thus, the enhancement of the ketoprofen the P123 micelles after they are released from the oil droplets.
release under acidic conditions results from the shifted ionic Therefore, in both cases, ketoprofen release is faster than that
equilibrium in an acidic medium and from the precipitation of from the hybrid material (Figure 5, B). For example, 71 and 61 %
acidic ketoprofen.
of the active agent are released from nanoemulsions after 3 h
Here, we can assume that such behavior also occurs for the in the absence and presence of P123, respectively, whereas only
hybrid material, as the pH used in the synthesis was very low 39 % of the drug is released from the hybrid material within
(
0.3) and the silicic acid species produced by tetramethoxy- the same period. Therefore, it can be concluded that the hybrid
Figure 5. Profiles of the cumulative release of ketoprofen from (A) the hybrid material as a function of receptor solution pH and (B) from both the nano-
emulsion and the hybrid material in PBS solution at pH 7.4. The scatters correspond to the experimental data, and the solid lines correspond to the fitted
curves obtained with the Korsmeyer–Peppas model.
Eur. J. Inorg. Chem. 2016, 1989–1997
www.eurjic.org
1992
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
material is a promising drug delivery system for a sustained
and slow release, in contrast to nanoemulsions, which could be
appropriate when a fast response is required, such as in oral or
topical administration of drugs. By contrast, the hybrid material
will be a better candidate, for example, for the treatment of
osteoporosis and osteoarthritis, for which the release rate of the
drug should be as slow as the bone reconstruction. Another
potential application concerns the use of nanocarriers in cancer
therapy with chosen mitochondria as the target sites. Indeed,
only a few reports have dealt with nanoparticle-based delivery
platforms to target and deliver therapeutic drugs to mitochon-
Table 1. Release kinetic parameters, calculated by fitting the Korsmeyer–Pep-
pas model to ketoprofen release from the hybrid materials and the nano-
emulsions.
Sample
Korsmeyer–Peppas
R
2
k [mg h–1
]
n
Nanoemulsion without P123
Nanoemulsion in the presence of P123
Hybrid release at pH 7.4
Hybrid release at pH 7
Hybrid release at pH 4.5
0.97
0.97
0.94
0.98
0.94
0.86
0.20
0.15
1.65
1.40
2.00
2.10
0.36
0.40
0.24
0.20
0.28
0.13
Hybrid release at pH 1.2
[
51]
[51]
dria.
For example Zhao et al.
have reported the use of
mesoporous nanoparticles to deliver the anticancer drug doxo-
rubicin (DOX) to mitochondria.
vides acceptable fits for the ketoprofen release from the hybrid
On the basis of the above observations, we propose the fol- material in the receptor solution at various pH values and from
lowing scheme for the release of ketoprofen from the hybrid the nanoemulsions in the presence or absence of Pluronic mi-
material. After the nanoemulsions loaded with ketoprofen and celles, as shown in Figure 5. By plotting Q versus log t, the
t
2
the P123 micellar solutions are mixed, TMOS is immediately values of n and k as well as the correlation coefficient (R ), can
added and its hydrolysis occurs. Hydrogen bonds form between be calculated. The results are gathered in Table 1.
the hydrogen atoms of hydrolyzed TMOS and both of the oxy-
gen atoms of the oxyethylene (EO) groups of the C (EO) and lower than 0.5, which indicates a pseudo-Fickian release mecha-
Whatever the vehicle or the pH of the receptor solution, n is
12
4
P123 surfactants, which leads to the formation of a hybrid silica nism. This means that the release is fast at the beginning, fol-
containing the drug, also labeled as the nanocarrier. The oil lowed by a sustained slow release. From the k values (Table 1),
droplets are the imprints of the macropores, and the micelles it can also be noticed that the release from the hybrid material
are involved in the formation of the mesopore network, which is faster than that from the nanoemulsion. For release at pH 7.4
constitutes the macropore walls. During the release experi- for the receptor solution, the k values are 1.65, 0.20, and 0.15
ments, the receptor solution penetrates into the nanocarrier, for the hybrid material and the nanoemulsions without and
and the ketoprofen, entrapped into the oil droplets, dissolves with P123 micelles, respectively. Concerning the variation of the
into the receptor medium. Consecutively, owing to the connec- pH of the receptor solution, a higher release rate from the hy-
tion between the macro- and mesopore networks, a part of the brid material is obtained with decreasing pH.
ketoprofen molecules can diffuse and solubilize in the core of
To investigate the release mechanism further, the release
the Pluronic micelles, which act as a stabilizing cage. Therefore, profiles were also fitted to the alternative models mentioned
they remain entrapped in the hybrid material. As the micelles above, namely, the Higuchi, first-order, and second-order re-
2
are embedded in the silica, the release of ketoprofen from the lease models (Table 2). From the correlation coefficients, R , ob-
micelles is more difficult. Silica behaves as a barrier for the diffu- tained from fitting the different kinetic equations, ketoprofen
sion of the drug molecules. This can explain why a lower release from the hybrid material could follow the Higuchi
amount of drug is released from the hybrid material than from model, which suggests that the drug is released by diffusion
the nanoemulsions under the same conditions.
and that the process is controlled by the vehicle. From Table 2,
To investigate the release mechanism, the experimental data it appears that the first-order release model can also be applied
can be fitted using different kinetic models. The commonly for the release of ketoprofen from the nanoemulsions; there-
used mathematical models for evaluating drug-release kinetics fore, the drug is released by a diffusion-controlled mechanism,
are the zero-order, first-order, Higuchi, and Korsmeyer–Peppas and its release rate depends greatly on the concentration of the
[
52]
models.
The equation corresponding to the latter model is drug loaded.
Q = kt , in which Q , k, and n stand for the fraction of drug The Higuchi model describes drug release as a diffusion
released at time t, the kinetic release, and the diffusional release process based on the Fick law and is dependent on the square
n
t
t
1
/2
exponent, respectively. The value of n provides information root of time; Q = k t , in which Q is the amount of drug
t
H
t
about the diffusion modes. The Korsmeyer–Peppas model pro- released in time t, and k is the Higuchi release rate constant.
H
Table 2. Different kinetic models fitted to the release profiles of ketoprofen from the hybrid material and from nanoemulsions to the receptor solutions.
Sample
Higuchi
R
Zero order
R
First order
R
2
[mg h–1/2
2
[mg h–1
2
1
k ]
[h–1
k
H
]
k
0
]
Nanoemulsion
0.94
0.96
0.99
0.98
0.96
0.98
0.17
0.14
2.35
2.13
2.70
3.50
0.94
0.94
0.92
0.90
0.88
0.92
0.11
0.084
1.58
1.06
1.56
1.54
0.97
0.98
0.86
0.90
0.85
0.80
0.068
0.054
0.058
0.027
0.021
0.071
Nanoemulsion + P123
Hybrid at pH 7.4
Hybrid in water
Hybrid at pH 4.5
Hybrid at pH 1.2
Eur. J. Inorg. Chem. 2016, 1989–1997
www.eurjic.org
1993
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
The zero-order model describes a system in which the drug matrix through the pores, and the dissolution of the ketoprofen
release is controlled by diffusion across the membrane and is in the medium allows its diffusion from the system along the
independent of the vehicle.; Q = Q + k t, in which Q is the solvent-filled pore channels. The solubility of ketoprofen in the
t
0
0
t
amount of drug released in time t, Q is the initial amount of release medium increases with the P123 content in the PBS
0
drug in the solution, and k is the zero-order release constant.
solution. Consequently, the amount of released drug increases.
0
The first-order model describes the release of a drug from a Therefore, in each case, a burst release of ketoprofen is noted,
system in which the release rate is concentration dependent; and a plateau is reached after 40 min. This is due to weak inter-
1
–k t
Q = Q e , in which Q is the amount of drug released in time actions between the drug and the silica. This rapid release is
t
0
t
t, Q is the initial amount of the drug in the solution, and k is usually observed if poorly soluble drugs are adsorbed on un-
0
1
the first-order release constant.
modified bare silica, which is a rather hydrophilic sub-
[
26,43,47]
strate.
Similarly to those for the hybrid materials and
nanoemulsions, the profiles for the cumulative release of keto-
profen from the hierarchical bare silica were fitted to the Kor-
smeyer–Peppas model (lines in Figure 6). The values of n (Ta-
ble 3) are lower than 0.5, which suggests that a pseudo-Fickian
release mechanism also operates here. The calculated k values
confirmed the faster release rates as the P123 concentration is
increased.
In Vitro Release of Ketoprofen from the Bare Hierarchical
Porous Silica
The release profile from the hybrid material was also compared
with that from the hierarchical bare silica (Figure 6) in which
the ketoprofen was impregnated. The release of ketoprofen
from the hierarchical bare silica was very limited and did not
exceed 8 %, which was reached within few minutes. This is
probably due to the low solubility of ketoprofen in the aqueous
phase. The solubilization rate of poorly soluble acidic drugs can
be improved by the presence of surfactants in the release me-
Table 3. Release characteristics of ketoprofen from the bare hierarchical po-
rous silica, obtained from the Korsmeyer–Peppas model.
Sample
Korsmeyer–Peppas
2
K [mg h–1
R
]
n
[
53,54]
[54]
dium;
for example, Choi et al.
have reported that the
Bare silica (0 % P123)
0.94
0.96
0.98
0.86
0.90
0.35
2.10
2.25
2.70
3.40
0.04
0.18
0.25
0.09
0.04
solubility of ibuprofen (pK = 4.4) in a buffer solution at pH 1.2
–5
a
Bare silica (10 % P123)
increases from 11.0 to 1415 μg mL–1 if 1 % (w/v) of polysorbate
–
3
Bare silica (10 % P123)
–
2
80 is added.
Bare silica (10 % P123)
Bare silica (1 % P123)
Conclusions
Oil-in-water nanoemulsions of 100 nm diameter were prepared
by the phase inversion temperature (PIT) method in a Remcopal
4/decane/water system. These nanoemulsions were mixed with
micellar and precursor solutions, and mineralization was per-
formed in the external phase of the emulsion to afford meso-/
macroporous silica materials by a dual templating mechanism
with micelles and nanoemulsions. The material has hexagonally
ordered mesopores of 9.3 nm and macropores of ca. 240 nm as
Figure 6. Profiles of the cumulative release of ketoprofen from the hierarchical
bare silica. The scatters correspond to the experimental data, and the solid
lines correspond to the fitted curves obtained with the Korsmeyer–Peppas
model. The receptor solution was PBS at pH 7.4, and various concentration
of P123 surfactant were added to the receptor solution.
2
–1
well as a high surface area of 782 m g and a large pore vol-
3
–1
ume of 1.2 cm g . Without the removal of the organic compo-
nents, the hybrid organic/inorganic nanocomposite can per-
form as a nanocarrier through the predissolution of ketoprofen
Here, we have investigated the effect of the addition of Plu- in the nanoemulsions. The results show that such a nanocarrier
ronic P123 to the PBS solution on the release of ketoprofen can release up to 43 wt.-% of the drug in a phosphate buffer
from the bare porous silica. The concentration of P123 in the at pH 7.4. The release reaches 53 wt.-% at pH 4.5; therefore, the
receptor solution was varied from 10–4 to 1 wt.-% to be above system is pH responsive for the release of ketoprofen. In addi-
–6
–1
the critical micelle concentration (CMC) value (4.4 × 10 mol L , tion, during the release process, a fraction of the ketoprofen
–
3
[55]
i.e., 2.5 × 10 wt.-%).
Indeed, in such a dissolution process, molecules solubilizes into the P123 micelles, which were added
it is assumed that solubilization occurs by partitioning of the to induce the mesostructure. The P123 micelles act as a stabiliz-
drug into surfactant micelles. Both the free and ionized acid ing cage, whereas the silica shell performs as a barrier against
forms of the ketoprofen molecule should be considered. From the diffusion of the drug molecules. In any case, the organic/
Figure 6, a beneficial effect of the presence of the Pluronic mi- inorganic hybrid materials allow a slow and sustained drug re-
celles in the PBS solution on the ketoprofen release can be lease.
noted. Indeed, after 10 min, the amount of the released drug
By contrast, owing to its low solubility in the aqueous phase,
increases from 8 to 77 % as the P123 concentration increases only a small fraction of ketoprofen can be released to the PBS
from 0 to 1 wt.-%. The release medium penetrates into the silica solution at pH 7.4 after the impregnation of this drug in bare
Eur. J. Inorg. Chem. 2016, 1989–1997
www.eurjic.org
1994
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
meso-/macrostructured silica. The solubility of ketoprofen is in- solution were then mixed prior the addition of TMOS. The mixture
was transferred and sealed in a Teflon autoclave, which was heated
creased by the addition of P123 micelles to the receptor solu-
at 40 °C for 24 h and afterwards at 70 °C for 24 h. After this step, a
powder material was obtained, which was designated as the hybrid
material loaded with ketoprofen (Scheme 2).
tion, and a maximum of 87 wt.-% of release was reached when
1
wt.-% of P123 was added to the PBS solution. Therefore, for
bare meso-/macrostructured silica, the release is surfactant-as-
sisted.
The release profiles from the different vehicles have been
fitted to the commonly used mathematical models to evaluate
the release kinetics of the drug. In all cases, a pseudo-Fickian
release mechanism was observed.
Experimental Section
General: Remcopal 4 (commercial name) it is a technical-grade
nonionic ethoxylated surfactant, which mainly consists of a hydro-
genated carbon chain with 12 carbon atoms and the average num-
ber of poly(oxyethylene) units is four; it is denoted as C (EO) . This
1
2
4
surfactant was provided by CECA. The triblock copolymer P123
(EO) (PO) (EO) ] (PO = propylene oxide) and decane were pur-
[
20
70
20
chased from Aldrich. Deionized water from a Milli-Q water purifica-
tion system was used to prepare the various samples.
Carrier Preparation and Ketoprofen Loading: The O/W nanoe-
mulsions were formulated from a water–C EO –decane system by
1
2
4
Scheme 2. Preparation of the hybrid material loaded with ketoprofen.
the PIT method. A mixture of Remcopal (6.6 wt.-%), deionized water
73.4 wt.-%) acidified with HCl (pH 0.3), and decane (20 wt.-%) was
(
For comparison, ketoprofen was also encapsulated in the bare silica
material by the impregnation method. In that case, the drug
prepared in a glass tube and stirred in a vortex. The mixtures were
heated to 40 °C and then cooled quickly to 5 °C to form O/W nano-
emulsions, which were stored at room temperature (25 °C).
(
50 mg) was dissolved in a hydro–alcoholic solution (1:1, 5 mL), and
then the solution was dispersed homogeneously on silica particles
2 g). The whole mixture was heated at 60 °C for 24 h. The sample
Hierarchical porous silica was synthesized according to a dual tem-
plating mechanism with previously prepared nanoemulsions as im-
prints for the macropore network and P123 micelles to induce the
mesoporosity. In a typical synthesis, equal volumes of the nano-
emulsion and aqueous Pluronic micellar solutions were mixed. In
the final mixture, the Remcopal, Pluronic P123, and decane contents
were 3.3, 2.5, and 10 wt.-%, respectively, and the pH was kept at
(
was considered dry when its weight became constant.
In each case, the quantity of the encapsulated drug was calculated
to make 2.5 wt.-% of the final product. This ratio corresponds to
the typical dosage embedded in commercial formulations.[
2]
Characterization: SAXS measurements were performed with a
0
.3. After the homogenization of the sample, TMOS (2.6 g) was SAXSess mc2 (Anton Paar) apparatus. It was attached to an ID 3003
added as the silica source. The mixture was stirred magnetically at
room temperature for 1 h and then transferred and sealed in a
Teflon autoclave, which was heated at 40 °C for 24 h and afterwards
at 70 °C for 24 h. To eliminate surfactants, oil, and methanol (gener-
ated by TMOS hydrolysis), Soxhlet extraction with ethanol for 48 h
was then performed. After this extraction and subsequent drying
at room temperature, a white silica powder was obtained, which
was designated to be the bare hierarchical silica.
laboratory X-ray generator (General Electric), equipped with a
sealed X-ray tube [PANalytical, λ(Cu-K ) = 0.154 nm, V = 40 kV, I =
α
50 mA]. A multilayer mirror and a block collimator provided a mon-
ochromatic primary beam. A translucent beam stop allowed the
alignment of an attenuated primary beam at q = 0. The porous
materials were introduced into a powder cell, whereas the nano-
emulsions were placed in Quartz capillaries of 1 mm diameter and
then placed inside an evacuated chamber equipped with a temper-
ature-controlled sample holder. The acquisition times were typically
With nanoemulsions as carriers, ketoprofen (Scheme 1) was incor-
porated into the oil and Remcopal 4 mixture and fully dissolved
before the addition of water, and the nanoemulsion was formed by
the PIT method as described above. The samples were kept in a
water bath at 25.0 °C. To use the hybrid materials as carriers, keto-
profen was first loaded in the nanoemulsions. Equal volumes of the
nanoemulsion containing the drug and aqueous Pluronic micellar
1
to 5 min for porous materials and 20 min to 1 h for the nanoemul-
sions. The scattering of the X-ray beam was recorded by a CCD
detector (Princeton Instruments, 2084 × 2084 pixels array with
2
–1
2
4 × 24 μm pixel size) in the q range 0.3 to 5 nm . The detector
was 309 mm from the sample holder. For the nanoemulsions, the
data were corrected for the background and the solvent (water)
scattering. The scattering intensities were evaluated on an absolute
scale with water as a reference. A Malvern zetasizer 3000HSa instru-
ment was used to measure the size of the nanoemulsion droplets
in water by dynamic light scattering with a 5 mW helium–neon
laser operated at 633 nm with the scattering angle fixed at 90°. The
sample was contained in a flow-through cell kept at 25 °C. All of
the measurements were performed under suitable conditions by
systematically diluting the samples three times with water to vali-
date the results. The water was first filtered with a mixed cellulose
ester (surfactant-free) filter (0.45 μm). The attenuator (diaphragm)
Scheme 1. Chemical structure of ketoprofen.
Eur. J. Inorg. Chem. 2016, 1989–1997
www.eurjic.org
1995
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
of the photomultiplier was automatically adapted for each sample
to prevent detector saturation. The data were analyzed by the CON-
TIN method, and the size was provided by an intensity distribution.
The NTA measurements were performed with a NanoSight NS300
instrument (Malvern) equipped with a sample chamber and a
[5] Z. Wei, J. Hao, S. Yuan, Y. Li, W. Juan, X. Sha, X. Fang, Int. J. Pharm. 2009,
3
76, 176–185.
[
[
6] M. J. Lawrence, G. D. Rees, Adv. Drug Delivery Rev. 2000, 45, 89–121.
7] M. D. Chatzidaki, N. Arik, J. Monteil, V. Papadimitriou, F. Leal-Calderon, A.
Xenakis, Colloids Surf. B 2016, 137, 146–151.
[
8] D. H. Oh, J. H. Kang, D. W. Kim, B.-J. Lee, J. O. Kim, C. S. Yong, H.-G. Choi,
Int. J. Pharm. 2011, 420, 412–418.
6
35 nm laser. The samples for TEM analysis were prepared by crush-
ing some material in chloroform. Afterwards a drop of this suspen-
sion was placed on a holey carbon coated copper grid. A Philips
FEI XL30 FEG microscope with an accelerating voltage of 7 kV was
used to record the SEM images. The infrared spectra were recorded
with a SHIMADZU IRAffinity-1 FTIR spectrometer equipped with a
PIKE GladiATR diamond attenuated total reflectance (ATR) accessory.
To perform the analysis, the sample was spread directly on the dia-
mond window and stabilized with a stainless-steel handle. The
measurements were made in absorbance mode at a resolution of
[
9] S. Setthacheewakul, S. Mahattanadul, N. Phadoongsombut, W. Pichayak-
orn, R. Wiwattanapatapee, Eur. J. Pharm. Biopharm. 2010, 76, 475–485.
[10] N. Sadurní, C. Solans, N. Azemar, M. J. García-Celma, Eur. J. Pharm. Sci.
2005, 26, 438–445.
[11] D. Mou, H. Chen, D. Du, C. Mao, J. Wan, H. Xu, X. Yang, Int. J. Pharm.
2008, 353, 270–276.
[
[
[
12] N. Ahmad, R. Ramsch, M. Llinàs, C. Solans, R. Hashim, H. A. Tajuddin,
Colloids Surf. B 2014, 115, 267–274.
13] C. Fornaguera, A. Dols-Perez, G. Calderó, M. J. García-Celma, J. Camarasa,
C. Solans, J. Controlled Release 2015, 211, 134–143.
14] S. Calligaris, P. Comuzzo, F. Bot, G. Lippe, R. Zironi, M. Anese, M. C. Nicoli,
LWT-Food Sci. Technol. 2015, 63, 77–84.
[15] S. M. Ðorđević, N. D. Cekić, M. M. Savić, T. M. Isailović, D. V. Ranđelović,
B. D. Marković, S. R. Savić, T. T. Stamenić, R. Daniels, S. D. Savić, Int. J.
Pharm. 2015, 493, 40–54.
16] C. Solans, P. Izquierdo, J. Nolla, N. Azemar, M. J. García-Celma, Curr. Opin.
Colloid Interface Sci. 2005, 10, 102–110.
17] T. G. Mason, J. N. Wilking, K. Meleson, C. B. Chang, S. M. Graves, J. Phys.
Condens. Matter 2006, 18, R635–R666.
18] K. Landfester, Macromol. Rapid Commun. 2001, 22, 896–936.
19] P. Izquierdo, J. Esquena, Th. F. Tadros, C. Dederen, M. J. Garcia, N. Azemar,
C. Solans, Langmuir 2002, 18, 26–30.
–1
2
cm . The N adsorption–desorption isotherms were determined
2
with a Micromeritics TRISTAR 3000 sorptometer at –196 °C. The pore
diameters, pore-size distributions, and cumulative pore volumes
were determined by the BJH method applied to the adsorption
branch of the isotherm.[ Macropores were detected by mercury
porosimetry with a Micromeritics Autopore IV 9500 instrument. The
penetrometer (stem volume: 0.412 mL, penetrometer volume: 1 mL)
was filled with the sample (0.1 g). A low-pressure analysis (0.1 to
56]
[
[
1
6
5 psia) was performed followed by a high-pressure analysis (15 to
0000 psia) to detect pores down to 5 nm in diameter, and the
[
[
equilibration time was 10 s.
[
[
[
20] S. L. Ee, X. Duan, J. Liew, Q. D. Nguyen, Chem. Eng. J. 2008, 140, 626–
In Vitro Release Experiments: The nanoemulsion, hybrid material,
or silica (0.2 g) containing ketoprofen was added to a dialysis bag
6
31.
21] A. Forgiarini, J. Esquena, C. González, C. Solans, Langmuir 2001, 17, 2076–
083.
22] L. Wang, X. Li, G. Zhang, J. Dong, J. Eastoe, J. Colloid Interface Sci. 2007,
14, 230–235.
(molecular weight cutoff 12000–14000 Da) and immersed in a
2
beaker containing a receptor solution consisting, in most cases, of
a phosphate buffer solution at pH 7.4. The solution was stirred at
3
1
50 rpm with a magnetic stirrer. The beakers contained buffer solu-
[23] L. Wang, J. Dong, J. Chen, J. Eastoe, X. Li, J. Colloid Interface Sci. 2009,
330, 443–448.
[24] M. Vallet-Regi, A. Rámila, R. P. del Real, J. Pérez-Pariente, Chem. Mater.
tion (150 mL) to ensure sink conditions, that is, that the concentra-
tion of the drug in the receptor solution was always lower than
1
formed at room temperature. To assist the release of ketoprofen
from the bare silica material, micelles of P123 were eventually
added to the PBS solution. For the hybrid material, the pH of the
receptor solution was varied from 1.2 (pH of the gastric fluid) to 7.4
2001, 13, 308–311.
0 % of its solubility in this medium. All experiments were per-
[
[
[
[
25] W. Xu, J. Riikonen, V. P. Lehto, Int. J. Pharm. 2013, 453, 181–197.
26] M. Moritz, M. Laniecki, Appl. Surf. Sci. 2012, 258, 7523–7529.
27] S. Wang, Microporous Mesoporous Mater. 2009, 117, 1–9.
28] L. Gao, J. Sun, L. Zhang, J. Wang, B. Ren, Mater. Chem. Phys. 2012, 135,
7
86–797.
29] V. Mamaeva, C. Sahlgren, M. Lindén, Adv. Drug Delivery Rev. 2013, 65,
89–702.
(considered to be the pH of blood and of intestinal fluid). This al-
[
lows the simulation of the behavior of the material after oral admin-
istration. An acidic medium (pH 1.2 hydrochloric acid buffer, pH 4.5
acetic buffer), deionized water (pH ≈ 6.5), and PBS (pH 7.4) were
considered as receptor solutions. Aliquots (4 mL) were taken at reg-
ular intervals and replaced with fresh solution (4 mL); the dilution
effect was considered in the calculations. The samples were ana-
lyzed with a LAMBDA 1050 UV/Vis/NIR Spectrophotometer at λ =
6
[30] P. Botella, A. Corma, M. Quesada, J. Mater. Chem. 2012, 22, 6394–6401.
[31] C. T. Kresge, W. J. Roth, Chem. Soc. Rev. 2013, 42, 3663–3670.
[32] Y. Wan, D. Zhao, Chem. Rev. 2007, 107, 2821–2860.
[33] J. L. Blin, M. Impéror-Clerc, Chem. Soc. Rev. 2013, 42, 4071–4082.
[
34] F. Wang, S. L. Nimmo, B. Cao, M. Chuanbin, Chem. Sci. 2012, 3, 2639–
645.
2
[
[
35] E. Kasotakis, A. Mitraki, Biopolymers 2012, 98, 501–509.
36] C. Mao, F. Wang, B. Cao, Angew. Chem. 2012, 124, 6517–6521; Angew.
Chem. Int. Ed. 2012, 51, 6411–6415.
2
60 nm. In all cases, the release experiments were conducted in
three replicates to confirm the results. The error between the three
measurements should not exceed 5 %.
[
37] S. Kerkhofs, F. Saïdi, N. Vandervoort, G. Van den Mooter, C. Martineau, F.
Taulelle, J. A. Martens, J. Mater. Chem. B 2015, 3, 3054–3061.
38] M. Manzano, M. Vallet-Regi, J. Mater. Chem. 2010, 20, 5593–5604.
39] T. Fontecave, C. Boissiere, N. Baccile, F. J. Ploun, C. Sanchez, Chem. Mater.
[
[
Keywords: Nanoemulsions · Mesoporous materials · Organic–
inorganic hybrid composites · Drug delivery
2013, 25, 4671–4678.
[
40] Y. H. Son, M. Park, Y. B. Choy, H. R. Choi, D. S. Kim, K. C. Park, J. H. Choy,
Chem. Commun. 2007, 27, 2799–2801.
[
1] P. Sarzi-Puttini, F. Atzeni, L. Lanata, M. Bagnasco, M. Colombo, F. Fischer,
M. D'Imporzano, Reumatismo 2010, 62, 172–188.
[41] L. Y. Xu, Y. Y. Huang, T. Long, Z. G. Shi, Mater. Manuf. Processes 2013, 28,
626–630.
[
[
2] S. Coaccioli, Eur. Rev. Med. Pharmacol. Sci. 2011, 15, 943–949.
3] A. V. Kabanov, E. Batrakova, V. Y. Alakhov, J. Controlled Release 2002, 82,
[42] F. Qu, H. Lin, X. Wu, X. Li, S. Qiu, G. Zhu, Solid State Sci. 2010, 12, 851–
856.
1
89–2012.
[43] E. Santamaría, A. Maestro, M. Porras, J. M. Gutiérrez, C. González, J. Solid
State Chem. 2014, 210, 242–250.
[44] G. Porod, in: Small Angle X-ray Scattering (Eds.: O. Kratky, O. Glatter), Aca-
demic Press, London, 1982, p. 17–51.
[
4] Y. Zhang, P. Lundberg, M. Diether, C. Porsch, C. Janson, N. A. Lynd, C.
Ducani, M. Malkoch, E. Malmström, C. J. Hawker, A. M. Nyström, J. Mater.
Chem. B 2015, 3, 2472–2486.
Eur. J. Inorg. Chem. 2016, 1989–1997
www.eurjic.org
1996
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
[
[
[
45] P. Sancin, O. Caputo, C. Cavallari, M. Passerini, L. Rodriguez, M. Cini, A.
Fini, Eur. J. Pharm. Sci. 1999, 7, 207–213.
46] T. X. Bui, V. H. Pham, S. T. Le, H. Choi, J. Hazard. Mater. 2013, 254–255,
[51] Q. Qu, X. Ma, Y. Zhao, Nanoscale 2015, 7, 16677–16686.
[52] P. Costa, J. M. S. Lobo, Eur. J. Pharm. Biopharm. 2001, 13, 123–133.
[53] J. J. Sheng, N. A. Kasim, R. Chandrasekharan, G. L. Amidon, Eur. J. Pharm.
Sci. 2006, 29, 306–314.
[54] S. H. Park, H. K. Choi, Int. J. Pharm. 2006, 321, 35–41.
[55] M. Y. Kozlov, N. S. Melik-Nubarov, E. V. Batrakova, A. V. Kabanov, Macro-
molecules 2000, 33, 3305–3313.
345–353.
47] C. Charnay, S. Bégu, C. Tourné-Péteilh, L. Nicole, D. A. Lerner, J. M. Devois-
selle, Eur. J. Pharm. Biopharm. 2004, 57, 533–540.
48] M. Van Speybroeck, R. Mols, R. Mellaerts, T. D. Thi, J. A. Martens, J.
Van Humbeeck, P. Annaert, G. Van den Mooter, P. Augustijns, Eur. J.
Pharm. Biopharm. 2010, 75, 354–365.
[56] E. P. Barrett, L. G. Joyner, P. P. Halenda, J. Am. Chem. Soc. 1951, 73, 373–
380.
[
[
49] C. J. Brinker, G. W. Scherer, The Physics and Chemistry of Sol–Gel Process-
ing, Academic Press, San Diego, 1990, p. 97.
50] A. Léonard, J. L. Blin, P. A. Jacobs, P. Grange, B. L. Su, Microporous Mesopo-
rous Mater. 2003, 63, 59–73.
Received: October 3, 2015
Published Online: January, 12, 2016
Eur. J. Inorg. Chem. 2016, 1989–1997
www.eurjic.org
1997
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim