06influence of mesoporous structure type on the controlled delivery of drugs

J Sol-Gel Sci Technol (2009) 50:421–429 DOI 10.1007/s10971-009-1932-3

ORIGINAL PAPER

Influence of mesoporous structure type on the controlled delivery of drugs: release of ibuprofen from MCM-48, SBA-15 and functionalized SBA-15 Isabel Izquierdo-Barba Æ Edesia Sousa Æ Juan Carlos Doadrio Æ Antonio Luis Doadrio Æ Joaquı´n Pe´rez Pariente Æ Africa Martı´nez Æ Florence Babonneau Æ Marı´a Vallet-Regı´

Received: 20 October 2008 / Accepted: 18 February 2009 / Published online: 9 March 2009 Ó Springer Science+Business Media, LLC 2009

Abstract Ordered mesoporous materials exhibit potential features to be used as controlled drug delivery systems, including their wide range of chemical compositions and their outstanding textural and structural properties. Therefore, it is possible to control the drug release kinetics by tailoring such parameters. In this paper, mesoporous materials such as MCM-48 and SBA-15, which present different pore sizes (3.7 and 8.8 nm) and structural characteristics (3D-bicontinuous cubic and 2D-hexagonal, respectively) have been synthesized to evaluate their application as drug delivery system and to determine their influence on release kinetic of ibuprofen. Moreover, a chemical modification of the SBA-15 mesoporous material with octadecyltrimethoxysilane has also been performed to study its influence on the release rate of ibuprofen. The structural characteristics (3D cubic and 2D hexagonal pore system) do not affect the release kinetic profiles of ibuprofen. On the contrary, the pore size affects highly to the release kinetic profiles from first-order kinetic to zero-order kinetic for MCM-48 and SBA-15, respectively. Moreover, the importance of surface functionalization was

I. Izquierdo-Barba
E. Sousa
J. C. Doadrio
A. L. Doadrio
J. P. Pariente
A. Martı´nez
M. Vallet-Regı´ (&) Departamento de Quı´mica Inorga´nica y Bioinorga´nica, Facultad de Farmacia, Universidad Complutense, 28040 Madrid, Spain e-mail: [email protected] I. Izquierdo-Barba
M. Vallet-Regı´ Centro de Investigacio´n Biome´dica en Red, Bioingenierı´a, Biomateriales y Nanomedicina, CIBER-BBN, Madrid, Spain F. Babonneau Chimie de la Matie`re Condense´e de Paris, UPMC-Paris6 and CNRS, Paris, France e-mail: [email protected]

demonstrate through the very fast delivery of ibuprofen from SBA-15 mesoporous materials functionalized with octadecyl chains. Keywords Ordered mesoporous materials
Structural characteristics
2D-hexagonal
3D-bicontinuous cubic structure and functionalization

1 Introduction Supramolecular chemistry to obtain silica-based ordered mesoporous materials offers a large range of possibilities to tailor at the nanometric scale, the chemical composition, pore size and structural characteristics of the final material depending on the targetted properties for a given application [1, 2]. Polymerization of an inorganic precursor and further removal of the occluded surfactant results in a rigid silica shell that delimits the structural shape of the mesoporous. Mesoporous materials have high potential for applications, and possible areas of use include heterogeneous catalysis [3], sensors [4], filtration membranes [5] and more recently matrices for the controlled delivery of drugs [6–11]. The administration of drugs by a drug delivery system provides advantages over conventional drug therapies [12] increasing the oral bioavailability of the poorly water soluble drugs [13]. The entire drug dose needed for a desired period of time is administered at one time and released in a controlled manner. A number of different systems have been studied for controlled drug delivery, such as biodegradable polymers [14], and cements [15]. Ordered mesoporous materials possess a high pore volume that is usually close to 1 cm3/g, very homogeneous in size, which offers the possibility of embedding a large

123

422

variety of organic molecules having therapeutic activity [16, 17]. The feasibility of MCM-41, a member of the M41S mesoporous family of materials that contains a hexagonal array of pores of 3–4 nm of diameter, as a system for the controlled release of ibuprofen has recently been proven [6]. It has also been shown that the pore size of MCM-41 matrices influence the delivery rate [18] and it could be expected that the pore architecture of the mesoporous silica material can shape the release pattern of guest molecules as well. Moreover, it has been shown that the delivery process is also influenced by an adequate modification of the surface chemistry of the silica matrix [19–21]. For example, covering the MCM-41 surface by aminopropyl groups affects the delivery of the model drug from the matrix to a simulated body fluid media under in vitro conditions [22–24]. Taken all these considerations into account, we have explored in this work the influence of the mesoporous structure type on the release pattern of drugs by using ibuprofen as a model molecule. Two different materials have been used, SBA-15 and MCM-48. Highly ordered hexagonal mesoporous silica structure SBA-15, have been synthesized using commercially available block-copolymer surfactants in strong acid media [25]. It possesses an ordered hexagonal arrangement of unidirectional mesoporous channels having 6.0–7.0 nm of diameter, much larger therefore than the *3.0 nm channels currently present in MCM-41. Indeed, the presence of micropores that connect the main mesopore channels has also been evidenced [26]. Moreover, the material is characterized by a wall thicker than that of MCM41, owing to its much larger unit cell. Therefore, SBA-15 would be suitable for studying the influence of pore size on drug delivery in a wide range of pore size. In contrast with the unidirectional channels present in both MCM-41 and SBA-15, MCM-48 structure (cubic space group Ia-3d) contains two independent intertwined networks of channels [27, 28]. According to this tridimensional channel topology, MCM-48 would also be attractive for developing systems aiming to release drugs in a controlled manner. One of the characteristics of the mesoporous materials is the presence of a high concentration of silanol groups in the mesopores, which can be functionalized to control the pore size and surface properties. The control of inner surface properties induced by such surface modification is required for the design of adsorption/desorption media for specific purposes. The effect of the surface modification of MCM-41 with aminopropyl groups on the release of ibuprofen has been described [22, 24]. Other possible chemical change is the modification of the surface by silanes with long hydrocarbon chain [21]. This is a standard procedure to decrease the pore size and hydrophilic character of silica surfaces, widely used for example in gas and liquid chromatography for efficient separation [29].

123

J Sol-Gel Sci Technol (2009) 50:421–429

The objective of this paper is to study the release of ibuprofen from different matrices based on mesoporous materials such as SBA-15 and MCM-48, which present different pore sizes and structures, hexagonal and cubic, respectively. On the other hand, the influence of the chemical interaction of the ibuprofen with the mesoporous matrix will also be studied. For this purpose, the surface of SBA-15 materials will be modified with octadecyltrimethoxysilane. These studies of ibuprofen release have been carried out in static and stirred conditions in simulated body fluid (SBF).

2 Experimental 2.1 Synthesis of mesoporous materials SBA-15 mesoporous material has been synthetized by using the triblock polyethylene oxide–propylene oxide block copolymer (P123) as structure directing agent in acidic condition and tetraethylorthosilicate as silica source, according to the method reported by Zhao [30]. MCM-48 materials has been synthetized according to the method reported by Kim [31] by using hexadecyltrimethylamonium bromide (HTABr) as structure directing agent and colloidal silica Ludox AS40 (40 wt% of SiO2) as silica source. In both cases, the surfactants have been removed by calcination method at 550 °C under an initial nitrogen atmosphere followed by an air atmosphere. The pore wall functionalization has been performed by the reaction between the calcined mesoporous materials and octadecyltrimethoxysilane (C21H46O3Si; Fluka) following the method described in a previous publication [21]. 200 mg of cacined mesoporous materials (SBA-15 and MCM-48) were kept under Ar and refluxed with 4 meq of the alkoxide in 30 mL of methanol at 65 °C, during 24 h. The final samples, were filtered and washed with methanol, and dried at 60 °C overnight. In the MCM-48 mesoporous materials, after the functionalization process, a collapse of the mesoporous structure is observed by XRD and N2 adsorption characterization. 2.2 Drug loading The ibuprofen loading process has been perfomed on the calcined powders (SBA-15 and MCM-48) and the modified mesoporous material (SBA-15-C18), which were previously compacted in disks (13 9 3 mm) by uniaxial (2.75 MPa) and isostatic pressure (3 MPa). The disks (100 mg) were soaked in a solution of ibuprofen in hexane (10 mg/mL) for 4 days at room temperature. The weight ratio between mesoporous matrix and drug were changed

J Sol-Gel Sci Technol (2009) 50:421–429

from 1:1, 1:5 and 1:7 in order to obtain larger amount of drug inside mesoporous matrices.

423

3 Results and discussion 3.1 Characterization of the encapsulated samples

2.3 Drug release studies The in vitro study of ibuprofen delivery from the mesoporous matrices was performed by soaking the disks in 40 mL of a simulated body fluid, denoted as SBF, with similar ionic composition to the human plasma [32] at 37 °C. The procedures to evaluate the delivery were carried out with stirring (200 rpm) and without stirring of the SBF solution. UV spectrometry (UV-500 UNICAM) at 273 nm was the analytical method used for monitoring the amount of ibuprofen delivered as a function of time.

Figure 1 shows the XRD patterns of SBA-15 and MCM-48 mesoporous materials before and after adsorption of ibuprofen. In the case of the SBA-15 material, the pattern shows the (10) (11) and (20) reflections corresponding to a 2Dhexagonal mesoporous structure (p6 mm plain group) with ˚ [30]. The pattern corresponding to the d10 spacing of 87 A sample MCM-48 shows reflections characteristics of Ia-3d symmetry of a 3D-biocontinuous cubic structure, in good agreement with reported patterns of pure siliceous MCM-48 ˚ . The XRD materials [27, 31], with a d211 spacing of 38 A pattern corresponding to SBA-15-C18 (not shown) shows that the functionalization process of the pore wall does not

2.4 Characterization of the samples All the samples were characterized by X-ray diffraction (XRD), thermogravimetric analyses (TGA), elemental analysis, N2 adsorption and solid state nuclear magnetic resonance (NMR). The XRD patterns were obtained using a Philips X‘Pert MDP (Cu Ka radiation) diffractometer with multipurpose sample holder. The TGA was carried out between 30 and 900 °C in air (flow rate 100 mL/min with a heating rate of 10 °C/min) using a Perkin Elmer instrument. The surface area and pore size of the materials were determined by N2 adsorption using a Micromeritics ASAP 2010 porosimeter. Previously, all samples were degassed at 100oC for 24 h under vacuum (10-2 Torr), except the matrices with ibuprofen inside, that were degassed at 60 °C to avoid the sublimation of organic molecules. Pore size distribution was determined by assuming cylindrical pores, which are characteristics of all the studied samples (SBA-15, MCM41 and MCM-48). The method reported by Kruk-Jaroniec [33, 34], and based on a simple geometric relation between the specific pore volumes, obtained by N2 adsorption (t-Plot Method), and the lattice spacing d, obtained from XRD data, was used. The solid state NMR spectra were recorded on a Bruker AVANCE 300 spectrometer at 75.47 MHz for 13C, 59.63 MHz for 29Si and 300.13 MHz with a Bruker CPMAS probe using 7 mm ZrO2 rotors with spinning rate ranging from 4 to 6 kHz. 29Si MAS NMR spectra were recorded with a pulse width of 2 ls (ca. 30°) and a recycle delay of 150 s to overcome long T1 relaxation times, reported in the literature for MCM-41 type silicas. For the 13 C MAS NMR spectra, pulse width and recycle delay were, respectively 5 ls and 20 s. 13C CP MAS NMR spectra were recorded with a contact time of 1 ms and a recycle delay of 5 s. All 13C NMR experiments were conducted with proton decoupling.

Fig. 1 XRD patterns of SBA-15 and MCM-48 before and after the adsorption of ibuprofen

123

424

J Sol-Gel Sci Technol (2009) 50:421–429

Table 1 Amount (wt%) of ibuprofen adsorbed in different matrices as a function of the weight ratio of mesoporous matrix:ibuprofen Samples

Matrix : ibuprofen ratio

Drug adsorption (wt%)

MCM-48

1:1

29

MCM-48

1:5

27

SBA-15

1:1

21

SBA-15

1:7

32

SBA-15-C18

1:7

26

affect the structural characteristics, observing a decreasing ˚ which could be attributed to of d10 spacing from 87 to 80 A the decreasing of pore size due to the funtionalization precess. Moreover, XRD patterns corresponding of mesoporous matrices after ibuprofen adsorption show no loss of structural ordering in both matrices (see Fig. 1). The content of functional groups present in SBA-15-C18, was evaluated by TG and elemental analysis assuming no presence of methoxy groups, which could have been introduced during the grafting procedure (see NMR results). The amount of carbon present in SBA-15 is 9.7 g per 100 g of SiO2, according to elemental analysis. This value corresponds to a functional group content of 1.01 mmol/g SiO2. The effective uptake of ibuprofen by the mesoporous materials when soaked into the hexane solution of drug can be determined by TGA and elemental analysis. Table 1 shows the ibuprofen content for different materials in different adsorption conditions. It can be observed that by keeping the weight ratio—drug/matrix—constant, MCM48 has a higher capability of loading (29 wt%) than SBA15 (21 wt%) due to its larger surface area [24], the loading Fig. 2 13C SP (bottom) and CP (top) MAS NMR spectra of samples containing ibuprofen. (a) (b) from MCM-48 mesoporous matrix (c) (d) from SBA-15 mesoporous matrix; (e) (f) from SBA-15-C18 mesoporous matrix

123

in both cases correspond to a similar amount of ibuprofen per surface area unit (ca. 2.5 g/m2). Moreover, the amount of ibuprofen adsorbed in MCM-48 material as function of weight ratio (drug: matrix) is practically the same, indicating that the greatest ibuprofen adsorption has been achieved with the lower weight ratio. On the contrary, in the case of SBA-15 material, the amount of ibuprofen adsorbed per unit of surface area increased from 2.7 to 4.0 g/m2 for 1:1 and 1:7 (drug: matrix) weight ratios, respectively. Such differences could be attributed to higher drug accessibility in SBA-15 material compared to MCM48 materials due to larger pore size. From these results, the ratio (1:7) was chosen to adsorb ibuprofen on the modified SBA-15-C18 sample. The samples containing ibuprofen have been characterized by multinuclear solid state NMR. For the case of MCM-48 with 1:1 ratio, the 29Si MAS NMR spectrum (not shown) evidences that this sample consists of highly condensed framework with respective Q4, Q3 and Q2 compositions (Qn = Si(OSi)n(OH)4-n, n = 2–4) of 70, 24 and 6%. Assuming that the detected SiOH are surface sites, one can calculate a value of 3 OH/nm2, which indeed represents the upper limit value for accessible surface sites. The 13C MASNMR spectrum recorded with a single pulse (SP) experiment (Fig. 2a) shows the characteristic peaks of ibuprofen due to the sp3 CH3 and CH2 carbons (20–45 ppm), the sp2 carbons in the aromatic ring (125–145 ppm) and the carboxylic group at 180.6 ppm. The assignments are reported in Table 2. All the peaks are very sharp (1–2 ppm of LWHM (Linewidth at Half Maximum)) and no spinning side bands are observed despite the relatively low spinning rate (5 kHz). This behaviour is

J Sol-Gel Sci Technol (2009) 50:421–429 Table 2 Assignments of Samples

425

13

C NMR signals of ibuprofen. The numbering of the C atoms is presented in Fig. 2

Chemical shift (ppm) C1

C2

C3

C4-7

C5-6

C8

C9

C10

MCM-48

180.6

17.9

45.9

138.4–142.2

128.1–130.3

45.8

30.9

21.7

SBA-15

181.5

18.7

45.9

138.3–141.6

128.4–130.2

45.9

31.0

22.4

SBA-15-C18

180.6

17.8

45.3

140.4

129.2

45.3

30.1

21.7

different from that of crystalline ibuprofen that shows sharp peaks as well, but with spinning side bands for the sp2 carbons (aromatic ring and carboxylic group), due to a strong chemical shift anisotropy. This spectrum thus strongly suggests a large mobility of the ibuprofen molecules in the MCM-48 network. This is indeed confirmed by the poor 13C signal obtained using a 1H-13C cross-polarisation (CP) sequence (Fig. 2b), whatever the contact time. The low efficiency of the CP transfer is due to a reduction of the 13C-1H dipolar coupling by motional processes. This large mobility is also responsible for the excellent resolution of the 1H MAS–NMR spectrum recorded with a spinning rate of 5 kHz (Fig. 3a). One can recognize the peaks due to the ibuprofen molecules at 0.7, 1.3, 1.7, 2.3, 3.6 and 7.0 ppm, and an additional broader peak at 5.5 ppm that could be assigned to OH groups from SiOH and H2O molecules engaged in H-bonding [35]. This unexpected behaviour for such encapsulated molecules is similar to what was already reported in mesoporous MCM-41 sample, [36–38] which presents a monodirectional hexagonal structure with pore size of 2.5 nm [6]. This high mobility does not seem compatible with the existence of strong interactions between the carboxylic groups of ibuprofen and the Si-OH surface groups. For the case of SBA-15 with 1:7 ratio, the 29Si MAS– NMR spectrum (not shown) evidences that the network is slightly more condensed than for MCM-48 with respective Q4, Q3 and Q2 compositions of 79, 19 and 2%. Now if one considers the lower surface area value with respect to MCM48, the same value of 3OH/nm2 is found as the upper limit value for the accessible surface sites. The 13C SP MAS– NMR spectrum (Fig. 2c) shows peaks whose assignments are reported in Table 2, which are slightly broader (LWHM = 2–3 ppm) than for MCM-48, indicating a lower mobility. This is also in agreement with the 1H MAS–NMR spectrum, which shows much broader peaks (Fig. 3b). Interestingly, a 13C CP MAS–NMR spectrum with a good signal/noise ratio was obtained (Fig. 2d), which presents differences from the SP spectrum. The signals due to the C = O (181.5 ppm) and aromatic rings (129.2, 138.5 ppm) are much broader (LWHM = 3–4 ppm) than those of the SP spectrum. This suggests the existence of two families of ibuprofen molecules, with different mobilities. The majority of those detected by the SP sequence, which does not interact with the surface, and those detected by CP, which could interact with the Si-OH present at the pore wall, causing a

Fig. 3 1H MAS NMR spectra of samples containing ibuprofen

broadening of the C = O signal. However restricted mobility of a portion of the ibuprofen molecules could also be due to their incorporation into the micropores of the host (see discussion below). The SBA-15-C18 sample has been characterized before and after adsorption of ibuprofen. The 29Si CP-MAS–NMR spectrum (not shown) shows peaks at -45 and -55 ppm due to the T2 and T3 silane units bonded to the C18 chains, with one and two Si–O–Si bonds, respectively. The relative intensity of these peaks is weak (the molar ratio between T and Q units in the CP spectrum is around 5%), in agreement with the value of functional groups obtained from TG and elemental analysis. The presence of C18 chains is confirmed by the 13C CP-MAS NMR spectrum, that presents a main peak at 30.2 ppm (CH2 in the chain), peaks at 22.9 ppm (CH3–CH2– and Si–CH2–CH2–), and at

123

426

J Sol-Gel Sci Technol (2009) 50:421–429

Fig. 4 Nitrogen adsorption/ desorption isotherms of calcined SBA-15, calcined MCM-48, and SBA-15-C18 before and after adsorption of ibuprofen

13.2 ppm (CH3–CH2– and Si–CH2–). A small peak at 50.8 ppm reveals the presence of OCH3 groups, which have been introduced during the grafting procedure done in methanol. These results are in agreement with the presence of 1H signals at 1–1.4 ppm due to the C18 chains and at 3.6 ppm due to OCH3 groups. For the sample containing ibuprofen, the 13C SP and CP MAS–NMR spectra are very similar (Fig. 2e–f). The peaks are rather broad, with line widthes similar to what has been observed in the CP spectrum of SBA-15, indicating restricted mobility. This is also confirmed by the 1H MAS– NMR spectrum (Fig. 3c) with broader peaks. In SBA-15, this restriction in mobility for part of the ibuprofen molecules was assigned to interactions of these molecules with the surface sites. In SBA-15-C18, this may also arise from the presence of the long C18 chains. Nitrogen adsorption desorption isotherms corresponding to MCM-48, SBA-15, and SBA-15-C18 are shown in Fig. 4. Table 3 summarizes these results, which shows the different pore size (8.8 and 3.7 nm) of SBA-15 and MCM48 materials. It should be noticed the existence of micropores in the SBA-15 material which are interconnecting the mesoporous as other authors have demonstrated [26]. As expected, the introduction of the organic fragment leads to a decrease in pore diameter, surface area and pore volume. However, the small decrease in pore size of the functionalized sample as compared with calcined SBA-15 indicates a rather low content of organic moieties, in agreement with the NMR results.

123

Table 3 N2 adsorption results obtained for the samples before and after ibuprofen adsorption Samples

Dadsorption (nm)

SBET (m2/g)

Vlp (cm3/g)

V (cm3/g)

MCM-48

3.7

1,166

–

1.048

MCM-48-Ibu

2.6

784

–

0.516

SBA-15 SBA-15-Ibu

8.8 7.1

787 254

0.061 –

1.056 0.374

SBA-15-C18

7.9

678

0.005

0.898

SBA-15-C18-Ibu

6.9

316

–

0.486

Moreover, it can be observed in Table 3 that the adsorption of ibuprofen in the different materials leads to a decrease in pore diameter, surface area, and pore volume whereas no microporosity is found in the SBA-15 containing ibuprofen. These facts evidence the presence of the drug into the pores after the adsorption of ibuprofen. The different amount of ibuprofen adsorbed on MCM-48 and SBA-15 mesoporous materials could be explained considering the differences in sample surface area and affinity between matrix and drug [9]. The adsorption of drugs into mesoporous materias is a surface phenomenon that is governed by the adsorption properties. Therefore, the surface area is expected to be the main factor that determines the amount of adsorbed drug molecules. Moreover, considering the greatest amount of loaded drug per surface area unit, SBA-15 material with 4.0 g/m2 performs better than MCM-48 material with 2.5 g/m2, which could be

J Sol-Gel Sci Technol (2009) 50:421–429

427

Fig. 5 Model of interaction between the ibuprofen molecules and the pore surface of different mesoporous matrices (SBA-15 and MCM-48 and SBA-15-C18). (ref = effective pore radii)

attributed to a higher accesibility of the drug in SBA-15 material due to its larger pore size. At the same time, as can be seen by analyzing the adsorption isotherms (Fig. 4) and BET results, the ibuprofen molecules do not fully occupy the available intrachannel space, which can be attributed to a limited capacity of adsorption of the mesopores. Indeed, the reduction in pore diameter produced by the adsorption of ibuprofen inside the matrices varies depending on the host material, from 1.1 nm for MCM-48 to 1.7 nm for SBA-15. Taken into account the molecular dimensions of ibuprofen,*1.0 9 0.6 nm, the observed differences in pore size reduction could be accounted for by assuming that the ibuprofen molecule adsorbs onto the pore surface with its major axis either parallel (for MCM-48) or nearly perpendicular (for SBA-15) to the surface which could also be explained by the larger amount of ibuprofen adsorption per surface area unit (Fig. 5). Following this model, the molecule-to-molecule interaction that would take place in the nearly perpendicular orientation could also contribute to the lower mobility evidenced by NMR for ibuprofen adsorbed on calcined SBA-15. The restricted mobility detected in SBA-15-C18, which shows ibuprofen orientation similar to that of MCM-48, could be attributed to the presence of the long hydrocarbon chains. 3.2 Study of release kinetic profile of ibuprofen from mesoporous matrices

Fig. 6 Release profiles of ibuprofen from SBA-15 mesoporous matrices in static and stirring conditions Table 4 Parameters of controlled release of ibuprofen from SBA-15 material in stirring and static conditions. kb and ki constitute the release zero-order kinetic constant corresponding to the initial burst step and subsequently second slow step, respectively [(Qt/ 2nd step System 1st step (kb 9 102 h-1) Q0)max]burst (ki 9 102 h-1)

2nd step (Qt/Q0)max

Static

305

0.74

10

0.85

Stirring 427

0.60

93

0.99

3.2.2 Effect to the pore arrangement and pore size 3.2.1 Effect to the stirring of the in vitro system In order to study the influence of the stirring on the in vitro release system, two independent experiments in static and stirring (200 rpm) conditions on disks of SBA-15 materials containing ibuprofen were perfomed. Such results are collected in Fig. 6 and Table 4. Both systems present a similar drug release profiles with two steps; an initial burst release effect followed by a very slow release pattern. The initial burst is attributed to the fast dissolution and release of the portion of ibuprofen located on or near the surface of the disks while the slow step could be attributed to the release of ibuprofen confined inside the microporus. In the case of static conditions the second step goes very slowly which could be due to the poor fluid diffusion in the poor system. Similar results (not shown) were found for MCM-48 structure.

Figure 7a and b show the release kinetic profiles of ibuprofen from MCM-48 and SBA-15, respectively. The release profiles show obviously differences in both matrices. In the case of MCM-48, the release profile has been adjusted following an exponential equation based on the Noyes–Whitney equation and applying the Fick’s first law [39]: Qt ¼ 1  ek1 t ; Q0 where Qt and Q0 are respectively the amount of drug in the porous matrix at time t and the initial amount at t = 0, and k1 is the first-order release constant independent of the drug concentration and that contain information about the solvent accessibility to substrate and the diffusion coefficient through mesoporous channels.

123

428

J Sol-Gel Sci Technol (2009) 50:421–429

centered at *3.5 nm, a value similar to that of MCM-48, but a 2D hexagonal pore arrangement. By comparing both kinetic constants, an increase in release rate of ibuprofen is observed for the 3D-pore cubic system, which has been demonstrated to provide easier diffusion process and mass transport than 2D-pore hexagonal system [28]. 3.2.3 Effect of surface modification

Fig. 7 Release profiles of ibuprofen from mesoporous matrices (MCM-48, SBA-15 and MCM-41) and C18-modified matrix (SBA15-C18) under stirring condition

On the contrary, in the case of SBA-15, the release kinetic profile exhibits two steps, an initial burst release effect and very slow release pattern which could be adjusted following zero-order kinetic profile characterized by the equation: Qt ¼ k0 t; Q0 where k0 is the zero-order release constant independent of the adsorbate concentration as well as of the solvent accessible area. Similar results have been shown in the case of alendronate release from SBA-15 matrices [24]. Both materials contain different pore size (3.6 and 8.8 nm, respectively) and structural characteristics (3D pore system (Ia-3d symmetry) and 2D pore system (p6 mm symmetry)). In that sense, mesoporous matrices with similar pore size but different structural characteristics should be compared. For this purpose, the observed in vitro release profile for MCM-48 has been compared with that of MCM41, a mesoporous material previously studied for ibuprofen delivery [40]. The release kinetic profile of the ibuprofen from MCM-41 appeared to be similar to that of MCM-48 following a first-order kinetic profile (see Fig. 7c and Table 5). MCM-41 presents a pore size distribution Table 5 Parameters of controlled release of ibuprofen from MCM48, MCM-41, SBA-15 and SBA-15C18 mesoporous materials (ki = k1 for MCM-48 and MCM-41 and ki = k0 for SBA-15 matrices, respectively Sample

(Qt/Q0)max

(ki 9 102 h-1)

r

MCM-48

0.93

11.25

0.9911

MCM-41

0.99

9.7

0.9960

SBA-15

0.98

93

0.9963

SBA-15-C18

0.99

1,023

0.9934

123

The mesoporous silica with the largest pore size used in this study, SBA-15, has been modified with a silane containing a long C18 hydrocarbon chain; it is thus expected that the pore size decreases and the surface becomes more hydrophobic. Figure 7d shows the effect of the surface modification with C18 on ibuprofen release from SBA-15. The experiments demonstrate a faster release of ibuprofen from the functionalized system, showing a total release in one step with a release kinetic profile of zero-order. As the k0 values are compared (Table 4), a more favourable ibuprofen release kinetic for the functionalized system is evidenced. It can thus be concluded that the chemical nature of the pore wall surface seems to play a key role for shaping the drug delivery pattern. In regard to this, the interaction between ibuprofen with the silanol groups presents in SBA15 must be taken into account. The fact that ibuprofen release from SBA-15-C18 seems to be faster can be ascribed to the decrease of the silanol groups presents in the original mesoporous materials or at least a decrease in their accessibility. The unmodified sample contains a high concentration of OH groups before functionalization, but the amount of OH groups present in the pore wall decrease when the hydrocarbons chains are anchored on the pore surface. Furthermore, the long chains may cover the surface and prevent the accessibility of ibuprofen to the Si– OH groups. Indeed, the presence of the hydrocarbon chain confers the pore surface a rather hydrophobic character, and hence a low affinity for polar groups. The inner surface of modified SBA-15 became less polar after modification, decreasing the ibuprofen interaction with the functionalized surface, which results in a very fast delivery of ibuprofen from this system. It is also noteworthy that the differences in the ibuprofen/matrix interactions reflected in the mobility of the ibuprofen detected by NMR do not result in significant modifications of the release rate. This observation is in agreement with the kinetic model, used to fit the release pattern, where the rate-controlling step is the diffusion of the drug through the matrix porous network.

4 Conclusions The applicability of mesoporous silica SBA-15 (hexagonal) and MCM-48 (cubic) materials as matrices for the

J Sol-Gel Sci Technol (2009) 50:421–429

controlled delivery of drugs was studied, in order to establish the influence of the pore architecture and size of the host on the ibuprofen release. The experiments have demonstrated no significant effect of the structure type on the release kinetic profiles of ibuprofen. However, the pore size affects such profiles from first-order kinetic for MCM48 to zero-order kinetic profiles for SBA-15. The polarity of the surface of the SBA-15 material has been modified by anchoring hydrophobic long hydrocarbon chain (C18) on the surface. This treatment decreases the interaction of ibuprofen with the modified surface, which results in a very fast delivery of ibuprofen from this system. 13 C and 1H NMR evidence differences in the mobility of ibuprofen molecules adsorbed into the matrices but they are not reflected in the overall release pattern, which obeys a diffusion model. Acknowledgments Financial support of CICYT Spain, through research projects MAT2005-01486 and CAM S-0505/MAT/000324 is acknowledged.

References 1. Ying JY, Mehner CP, Wong MS (1999) Angew Chem Int Ed 38:56. doi:10.1002/(SICI)1521-3773(19990115)38:1/2\56::AIDANIE56[3.0.CO;2-E 2. Soler-Illia GJAA, Sanchez C, Lebeau B, Patarin J (2002) Chem Rev 102:4093. doi:10.1021/cr0200062 3. Lapkin A, Bozkaya B, Mays T, Borello L, Edler K, Crittenden B (2003) Catal Today 81:611. doi:10.1016/S0920-5861(03)00159-7 4. Lan EH, Dave BC, Fukutu JM, Dunn BS, Zink JI, Valentino JS (1999) J Mater Chem 9:45. doi:10.1039/a805541f 5. MacGrath KM, Dabbs DM, Yao N, Elder KJ, Akasay IA, Gruner SM (2000) Langmuir 16:398. doi:10.1021/la990098z 6. Vallet-Regı´ M, Ramila A, Del Real RP, Pe´rez-Pariente J (2001) Chem Mater 13:311. doi:10.1021/cm0011559 7. Vallet-Regı´ M, Balas F, Colilla M, Manzano M (2007) Solid State Sci 9:768. doi:10.1016/j.solidstatesciences.2007.03.026 8. Vallet-Regı´ M, Balas F, Colilla M, Manzano M (2007) Drug Metab Lett 1:37. doi:10.2174/187231207779814382 9. Vallet-Regı´ M, Balas F, Arcos D (2007) Angew Chem Int Ed 47:7448 10. Vallet-Regı´ M (2006) Chem Eur J 12:5934. doi:10.1002/chem. 200600226 11. Salonen J, Laitinen L, Kaukonen AM, Tuura J, Bjo¨rkqvist M, Heikkila¨ T, Va¨ha¨-Heikkila¨ K, Hirvonen J, Lehto V-P (2005) J Control Release 108:362. doi:10.1016/j.jconrel.2005.08.017 12. Langer R (1999) J Control Release 62:7. doi:10.1016/S01683659(99)00057-7 13. Mellaerts R, Mols R, Jammaer JAG, Aerts CA, Annaert P, Humbeeck JV, Van den Mooter G, Augustijns P, Martens JA (2008) Eur J Pharm Biopharm 1:223. doi:10.1016/j.ejpb.2007. 11.006 14. Shin Y, Chang JH, Liu J, Williford R, Shin YK, Exarhos GJJ (2001) Contr Release 73:1. doi:10.1016/S0168-3659(01)00247-4

429 15. Doadrio JC, Arcos D, Caban˜as MV, Vallet-Regı´ M (2004) Biomaterials 25:2629. doi:10.1016/j.biomaterials.2003.09.037 16. Yanagisawa T, Shimizu T, Kuroda K, Kato C (1990) Bull Chem Soc Jpn 63:988. doi:10.1246/bcsj.63.988 17. Kresge CT, Leonowicz ME, Roth WJ, Vartuli JC, Beck JS (1992) Nature 359:710. doi:10.1038/359710a0 18. Horcajada P, Ra´mila A, Pe´rez-Pariente J, Vallet-Regı´ M (2004) Microporous Mesoporous Mater 68:105. doi:10.1016/j.micromeso. 2003.12.012 19. Nieto A, Balas F, Colilla M, Manzano M and Vallet-Regı´ M (2008) Micropor Mesopor Mater doi: 10.1016/j.micromeso.2008. 03.025 20. Izquierdo-Barba I, Martı´nez A, Doadrio AL, Pe´rez-Pariente J, Vallet-Regı´ M (2005) Eur J Pharm Sci 26:365. doi:10.1016/ j.ejps.2005.06.009 21. Doadrio JC, Sousa EMB, Izquierdo-Barba I, Doadrio AL, Pe´rezPariente J, Vallet-Regı´ M (2006) J Mater Chem 16:462. doi: 10.1039/b510101h 22. Mun˜oz B, Ramila A, Perez-Pariente J, Diaz I, Vallet-Regı´ M (2003) Chem Mater 15:500. doi:10.1021/cm021217q 23. Song S-W, Hidajat K, Kawi S (2005) Langmuir 21:9568. doi: 10.1021/la051167e 24. Balas F, Manzano M, Horcajada P, Vallet-Regı´ M (2006) J Am Chem Soc 128:8116. doi:10.1021/ja062286z 25. Zhao DY, Feng JL, Huo QS, Melosh N, Fredrickson GH, Chmelka BF, Stucky GD (1998) Science 279:548. doi:10.1126/ science.279.5350.548 26. Che S, Lund K, Tatsumi T, Iijima S, Joo SH, Ryoo R, Terasaki O (2003) Angew Chem Int Ed 42:2182. doi:10.1002/anie. 200250726 27. Kaneda M, Tsubakiyama T, Carlsson A, Sakamoto Y, Oshuna T, Terasaki O, Joo H, Ryoo R (2002) J Phys Chem B 106:1256. doi: 10.1021/jp0131875 28. Sakamoto Y, Kim TW, Ryoo R, Terasaki O (2004) Angew Chem Int Ed 43:5231. doi:10.1002/anie.200460449 29. Gru¨n M, Kurganov AA, Schacht S, Schu¨th F, Unger KK (1996) J Chromatogr A 740:1. doi:10.1016/0021-9673(96)00205-1 30. Zhao D, Huo Q, Fena J, Chmelka BF, Stucky GD (1998) J Am Chem Soc 120:6024. doi:10.1021/ja974025i 31. Kim JM, Kim SK, Ryoo R (1998) Chem Commun (Camb) 259. doi:10.1039/a707677k 32. Kokubo T, Kushitani H, Sakka S, Kisugi T, Yamamuro T (1990) J Biomed Mater Res 24:721. doi:10.1002/jbm.820240607 33. Kruk M, Jaroniec M, Sayari A (1997) J Phys Chem B 101:583. doi:10.1021/jp962000k 34. Kruk M, Jaroniec M, Sarayi A (1997) Langmuir 13:6267. doi: 10.1021/la970776m 35. Bacile N, Laurent G, Bonhomme C, Innocenza P, Babonneau F (2007) Chem Mater 19:1343. doi:10.1021/cm062545j 36. Babonneau F, Camus L, Steunou N, Ramila A, Vallet-Regi M (2003) Mater Res Soc Symp Proc 775:77 37. Azaı¨s T, Tourne´-Pe´teilh C, Aussenac F, Baccile N, Coelho C, Devoisselle J-M, Babonneau F (2006) Chem Mater 18:6382. doi: 10.1021/cm061551c 38. Babonneau F, Yeung L, Steunou N, Gervais C, Mun˜oz B, Ramila A, Vallet-Regi M (2004) J Sol-Gel Sci Techn 31:219 39. Costa P, Sousa-Lobo JM (2001) Eur J Pharm Sci 13:123. doi: 10.1016/S0928-0987(01)00095-1 40. Ramila A, Mun˜oz B, Perez-Pariente J, Vallet-Regı´ M (2003) J Sol-gel Sci Tech 26:1199

123