Stability of different mesoporous silica particles during an in vitro digestion

Document downloaded from: This paper must be cited as: The final publication is available at Copyright Additional Information http://dx.doi.org/10.1016/j.micromeso.2016.05.004 http://hdl.handle.net/10251/78699 Elsevier Pérez-Esteve, É.; Ruiz Rico, M.; Torre, CDL.; Llorca Martínez, ME.; Sancenón Galarza, F.; Marcos Martínez, MD.; Amoros Del Toro, PJ.... (2016). Stability of different mesoporous silica particles during an in vitro digestion. Microporous and Mesoporous Materials. 230:196207. doi:10.1016/j.micromeso.2016.05.004. Pérez-Esteve et al. 2016. Micropor Mesopor Mater 1 Stability of different mesoporous silica particles during an in vitro 1 digestion 2 3 Édgar Pérez-Estevea*, María Ruiz-Ricoa, Cristina de la Torreb,c, Empar Llorcad, Félix Sancenónb,c, 4 María D. Marcosb,c, Pedro Amoróse, Carmen Guilleme, Ramón Martínez-Máñezb,c,, José Manuel 5 Barata 6 a Grupo de Investigación e Innovación Alimentaria, Universitat Politècnica de València. Camino 7 de Vera s/n, 46022, Spain 8 b Instituto Interuniversitario de Investigación de Reconocimiento Molecular y Desarrollo 9 Tecnológico (IDM), Unidad Mixta Universitat Politècnica de València – Universidad de Valencia. 10 Departamento de Química Universitat Politècnica de València, Camino de Vera s/n, 46022, 11 Valencia, Spain 12 c CIBER de Bioingeniería, Biomateriales y Nanomedicina (CIBER-BBN) 13 d Grupo de Microestructura y Química de Alimentos. Departamento de Tecnología de 14 Alimentos, Universitat Politècnica de València. Camino de Vera s/n, 46022, Valencia, Spain 15 e Institut de Ciència dels Materials (ICMUV), Universitat de València, P.O. Box 2085, 46071, 16 Valencia, Spain 17 18 *Corresponding author: [email protected] 19 20 Pérez-Esteve et al. 2016. Micropor Mesopor Mater 2 Mesoporous silica materials have the ability to entrap drugs, nutrients and functional 21 biomolecules and can be able to act as smart delivery systems capable to control and target 22 the release of their cargo in a particular part of the gastrointestinal tract when administrated 23 orally. However, the aptness of these encapsulation supports in in vivo oral controlled release 24 relies on their chemical stability through the digestive tube. In this context, we have evaluated 25 the stability of four different mesoporous silica particles, frequently used as encapsulating 26 supports, during an in vitro digestion process comprising buccal, stomach and intestinal 27 phases. Results showed that after 4 h of digestion, the textural properties of silica supports in 28 the form of nanoparticles (MCM-41 and UVM-7 nanoparticles) were lost in varying degrees, 29 whereas silica microparticles supports (MCM-41 and SBA-15 microparticles) endures better the 30 digestion process. Moreover, the functionalization of the surface with N1-(331 trimethoxysilylpropyl)diethylenetriamine, an organic moiety commonly used in the 32 preparation of pH-responsive mesoporous silica particles, resulted in an improvement of the 33 stability of the supports. 34 35 36 Keywords: mesoporous silica particles, in vitro digestion, stability, amine-functionalization 37 Pérez-Esteve et al. 2016. Micropor Mesopor Mater 3 1. Introduction 38 Mesoporous Silica Particles (MSPs) are receiving great attention in the field of oral 39 controlled release due to their capability to improve drug solubility and stability in the 40 gastrointestinal tract (GIT), [1-2] as well as to dosage the cargo along time (sustained 41 controlled release) in specific GIT places (targeted controlled release) [3-5]. These reported 42 features, that convert MSPs in unique smart delivery systems, are due to their large loading 43 capacity [6], low toxicity [7] and the fact that their surface can be functionalized with 44 molecular/supramolecular ensembles. This last feature allows the development of gated-MSPs 45 showing “zero delivery” and capable to release their cargo on-command in response to 46 specifically designated external stimuli [8-10]. Drug delivery/formulation technologies that can 47 improve bioavailability, drug stability and subsequently increase drug effectiveness are much 48 desired in the pharmaceutical sciences [11-12]. In food technology, encapsulation of bioactive 49 molecules (e.g. vitamins, antioxidants, phytochemicals, etc.) may improve their biological 50 stability, facilitate components handling, mask unpleasant sensorial properties and modulate 51 the bioaccessibility of the molecule of interest along the GIT [13]. 52 Besides a high loading capacity, controlled release and biocompatibility, the suitability of 53 MSPs in oral controlled release in in vivo applications depend on the chemical stability of the 54 supports though the whole digestive tube. However, it is known that due to the metastability 55 of MSPs, silica can be biodegraded into silicic acids, including monomeric silicic acid and 56 various polysilicic acids with different polymerization degrees under harsh environments 57 provoking a collapse of the porous structures [14]. In this line, Cauda, Schlossbauer & Bein 58 studied the biodegradation of colloidal mesoporous silica nanoparticles (50 nm) in simulated 59 body fluid of bare, globally functionalized, and surface poly(ethylene glycol)-coated colloidal 60 mesoporous silica nanoparticles in simulated body fluid (pH 7.4) for a period of 1 month at 37 61 °C [15]. After this period of time, the textural properties of the mesoporous system were lost 62 and pores were blocked because the precipitation of inorganic components from the 63 Pérez-Esteve et al. 2016. Micropor Mesopor Mater 4 simulated body solution. The stability of the particles increased by surface functionalization 64 with poly(ethylene glycol). The degradation behaviour of surfactant-extracted mesoporous 65 silica in simulated body fluid was also evaluated by He and co-workers proposing a three-stage 66 degradation process comprising a fast bulk degradation on hour-scale, a silicon concentration 67 decrease stage due to a deposition of a calcium/magnesium silicate layer, and a later 68 continuous sustained diffusion beyond days [16]. The same year, Lin, Abadeer & Haynes, 69 evaluated the stability of small mesoporous silica nanoparticles (<50 nm) functionalised with 70 poly(ethylene glycol) in H2O, phosphate buffer solution (PBS) (pH 7.5), and Dulbecco’s 71 modified Eagle’s medium (DMEM) with 10% fetal bovine serum (FBS) (pH ca. 7.5) [17]. These 72 particles exhibited long term stability in all these media at both, room and physiological 73 temperature. In a different attempt, El Mourabit et al. studied the stability of mesoporous 74 silica under acidic conditions and a loss of textural properties of the supports was observed 75 [18]. The authors also found that the degradation rate was dependent on the nature of the 76 acidic media (phosphoric acid have stronger impact than hydrochloric or sulphuric acids) and 77 the kind of mesoporous silica used in the study. More recently, Choi et al. studied the 78 biodegradation of SBA-15 in both, simulated body fluid and in vivo [19]. These authors have 79 shown that the degradation rate of SBA-15 was affected by the presence of surface functional 80 groups and synthesis methodologies. Furthermore, in vivo experiment showed that SBA-15 81 degrade in the animal and pore structure deformation occurs as a function of time. 82 Most of these studies evaluated the stability of mesoporous silica nanoparticles. However, 83 MSPs can be fabricated with a controlled size from 50 nm to a few microns. When preparing 84 smart delivery systems based on MSPs, particle size is very important since it conditions the 85 distribution and behavior of particles in living systems. In general, small MSPs can cross 86 epitheliums, can be distributed in the body and be non-specifically internalized by certain cells 87 [20]. In contrary, oversized particles (microparticles) cannot easily cross physical membranes in 88 the body, and thus large particle sizes are preferred for developing orally administrated 89 Pérez-Esteve et al. 2016. Micropor Mesopor Mater 5 controlled release devices [4]. Having in mind the importance of particle size in oral 90 administration, it may be of importance to study the stability of mesoporous silica with micro91 sized particles. However, stability of large MSPs has been barely studied. Moreover, as far as 92 we know, there are not studies about the effect that the consecutive presence of saliva (pH 93 7.5), gastric (pH 1.2-2) and intestinal fluids (pH 7.8-8) have on the stability of the small and 94 large MSPs. Thus, notwithstanding the works detailed above, a lack of information about the 95 degradability/stability of MSPs with different sizes during a whole digestion is still unavailable. 96 The aim of this study was to evaluate the stability of different bare and functionalised 97 mesoporous silica particles differing in particle size, particle shape and pore structure (pore 98 size and wall thickness) during a simulated in vitro digestion. With this purpose a deep 99 evaluation of the stability of textural properties of MSPs during the in vitro digestion was 100 performed. Descriptive studies were completed with the assessment of potential cytotoxicity 101 of digested particles or their degradation products. 102 2. Materials and methods 103 2.1 Chemicals 104 Tetraethylorthosilicate (TEOS), N-cetyltrimethylammonium bromide (CTABr), Pluronic 105 P123 (P123), triethanolamine (TEAH3), sodium hydroxide (NaOH), hydrochloric acid (HCl), N1106 (3-trimethoxysilylpropyl)diethylenetriamine (N3), and all chemicals for the preparation of the 107 simulated digestive fluids were provided by Sigma-Aldrich (Poole, Dorset, UK). HPLC grade 108 acetonitrile was provided by Scharlau (Barcelona, Spain). Rhodamine B was acquired from 109 Fluka (Missouri, USA). 110 For cell culture experiments, trypan blue solution (0.4%) cell culture grade and dimethyl 111 sulfoxide (DMSO), phosphate buffered saline (PBS) and Dulbecco's Modified Eagle's medium 112 (DMEM) with glucose, L-glutamine and pyruvate for cell culture were provided by Sigma113 Pérez-Esteve et al. 2016. Micropor Mesopor Mater 6 Aldrich (Poole, Dorset, UK). Mc Coy’s 5a Medium and Keratinocyte Serum Free Medium, Fetal 114 Bovine Serum (FBS) and trypsin were purchased from Gibco (Life Technologies, Madrid, Spain). 115 Cell proliferation reagent WST-1 was purchased from Roche Applied Science (Barcelona, 116 Spain). 117 2.2 Mesoporous silica particles synthesis 118 Synthesis of the four different silica particles was carried out following the procedures 119 described previously [4]. 120 MCM-41 (M) was synthesized following the so-called “atrane route”, using CTABr as the 121 structure-directing agent and a molar ratio fixed to 122 7TEAH3:2TEOS:0.52CTABr:0.5NaOH:180H2O. The procedure consisted in adding CTABr to a 123 solution of TEAH3 and NaOH containing TEOS at 118 °C. After dissolving CTABr in the liquor, 124 water was slowly added with vigorous stirring at 70 °C to form a white suspension. This 125 mixture was aged at room temperature overnight. 126 Nanoparticulated MCM-41 (N) was synthesized using the following procedure: NaOH was 127 added to the CTABr solution, followed by adjusting the solution temperature to 95 °C. TEOS 128 was then added dropwise to the surfactant solution. The mixture was allowed to stir for 3 h to 129 give a white precipitate. 130 UVM-7 (U) was synthesised using, once again, the “atrane route”. The molar ratio of the 131 reagents in the mother liquor was fixed at 7TEAH3:2TEOS:0.52CTABr:180H2O. The TEOS/TEAH3 132 mixture was heated to 120 °C until no elimination of ethanol was observed. The mixture was 133 cooled to 90 °C and the CTABr was added gradually in small portions, followed by water. The 134 mixture was aged for 24 h. 135 The SBA-15 (S) sample was synthesized using P123 as the structure-directing agent with 136 the reactant molar ratios: 0.017P123:1.0TEOS:6HCl:196H2O. The preparation was carried 137 mixing an aqueous solution of P123 with HCl solution, and stirring for 2 h, after which the silica 138 source, TEOS, was added. This final mixture was stirred for a further 20 h. 139 Pérez-Esteve et al. 2016. Micropor Mesopor Mater 7 After the synthesis, the different solids were recovered, washed with deionised water, 140 and air-dried at room temperature. The as-synthesized solids were calcined at 550 °C using an 141 oxidant atmosphere for 5 h in order to remove the template phase. 142 The particles were also functionalised with N1-(3-trimethoxysilylpropyl)diethylenetriamine 143 (N3). In particular, 1 g of the different MSPs were suspended in 40 mL of acetonitrile and an 144 excess of N3 (4.3 mL, 15.0 moll g-1) was then added. Final mixtures were stirred for 5.5 h at 145 room temperature. Finally, the solids were filtered off, washed with 30 mL of deionised water, 146 and dried at room temperature. 147 2.3 Simulated digestion procedure 148 An in vitro digestion model consisting of mouth, gastric and intestinal phases described by 149 Versantvoort et al. was used to simulate the typical chemical composition, pH and residence 150 time periods of each of the three main compartments of the GIT [21]. A schematic 151 representation of the in vitro digestion model is presented in Figure 1. The pH values of the 152 digestive juices were checked and, if necessary, adjusted to the appropriate interval with 153 NaOH (1 M) or HCl (37% w/w). 154 155 Pérez-Esteve et al. 2016. Micropor Mesopor Mater 8 156 157 Figure 1. Schematic representation of the in vitro digestion process. The in vitro digestion model describes a three158 step procedure simulating the digestive processes in mouth, stomach and small intestine. In each compartment, the 159 matrix is incubated at 37 °C for a time relevant for the compartment. The digestion is initiated by addition of 160 artificial saliva to the material. Subsequently, gastric juices and intestinal fluids are added to simulate the digestive 161 processes in stomach and small intestine, respectively. After each of these steps, samples were taken to conduct 162 characterization procedures. Characterization typically involves microscopy, PXRD, size distribution, zeta potential, 163 N2 adsorption-desorption isotherms, 29Si RMN, silicon analysis and biocompatibility using WST-1 test. 164 165 2.4 Characterization of the mesoporous silica particles 166 All materials, as synthetized and after a simulated digestion process, were characterized 167 by standard procedures: i.e. X-ray diffraction (XRD), N2 adsorption-desorption isotherms, 168 transmission electron microscopy (TEM), field emission scanning electron microscopy (FESEM), 169 Pérez-Esteve et al. 2016. Micropor Mesopor Mater 15 the order N (wall thickness = 1.91), closely followed by M (wall thickness = 2.05) and U (wall 294 thickness = 2.22). The differences in wall thickness among N (less than 2 nm) and M and U 295 (more than 2 nm) might be due to the synthesis route. N and U follow the atrane route of 296 synthesis, a procedure that provides wide framework walls. 297 For its part, S with a wall thickness of 2.56 nm seemed to be the particle that better conserve 298 the hexagonal structure of the particle. In fact the diffractogram of digested S was very similar 299 than that of the undigested particle meaning that mesoporous arrangement is preserved 300 during the in vitro digestion procedure. In short, the pore wall thickness seems to be the key 301 factor to preserve the mesostructure integrity, being the silica stability highly favoured when 302 the pore wall thickness increases. 303 Nitrogen sorption data (Fig 3) show a reduction of the adsorbed N2 volume in all solids, 304 which suggested that the digestion procedure provokes a loss of specific surface area and pore 305 volume. This reduction is more marked in the case of N (177 m2 g-1; 0.29 cm3 g-1) and U (372 m2 306 g-1; 0.78 cm3 g-1) that in M (321 m2 g-1; 0.46 cm3 g-1) and S (368 m2 g-1; 0.62 cm3 g-1) in consonance 307 with XRD patterns. This loss of surface area and pore volume as a consequence of the contact 308 with biological media has also been observed by other authors [15,18]. In these works, the 309 loss of textural properties of silica supports was associated with a progressive elimination of 310 the porosity by dissolution of the silica or by pore blockage due to precipitation of inorganic 311 compounds onto the surface of the porous silica. The combined result of silica redisolution and 312 salt precipitation has a marked effect on the form of the isotherms. In fact, only SBA-15 shows 313 after digestion a N2 adsorption-desorption isotherms qualitatively similar to the original one, 314 with a well-defined adsorption step at relative pressure values in the 0.6-0.8 range. In the case 315 of the remaining silica supports with thicker walls (M and U), the loss of surface area and 316 volume seems to be more pronounced than the mesostructural disorder evidenced through 317 XRD. At this point, probably the existence of small mesopores for samples M, N and U (ca. 2.5 318 nm) leads to an easier pore blocking through salt reprecipitation with the subsequent 319 Pérez-Esteve et al. 2016. Micropor Mesopor Mater 16 significant loss of surface and pore volume. In contrast, the larger values of the voids and the 320 walls in the S support favours the preservation of the mesostructure and hinder pore blocking. 321 322 323 Figure 3. Nitrogen adsorption-desorption isotherms for microparticulated MCM-41 (M) (a), nanoparticulated MCM324 41 (N) (b), SBA-15 (S) (c) and UVM-7 (U) (d) before (i) and after (ii) the in vitro digestion procedure. 325 326 To further understand which of these mechanisms are involved in the digestion-induced 327 silica degradation, in parallel to XRD and N2 absorption-desorption isotherms experiments, 328 TEM and FESEM observations of the four silica supports were carried out. Figure 4 shows 329 FESEM and TEM pictures of different bare particles before and after the in vitro digestion 330 process. This figure allows observing not only the particle size and shape of the single particle, 331 but also particle porosity. 332 After the digestion, two particles did not modify their appearance (FESEM) and pore 333 integrity (TEM). These particles are M and S which are particle with size in the microscale. This 334 implies that for these particles the loss of order observed in XRD was not provoked by a 335 collapse of the mesostructure, but probably by the formation of a small volume fraction of 336 Pérez-Esteve et al. 2016. Micropor Mesopor Mater 17 “gel” consisting mostly of (SiOH2)n18 or the adsorption of calcium and phosphate ions present 337 in the digestion fluids on the silica surface forming a hydroxyapatite phase [15, 23-25]. 338 In contrast, supports based on nanoparticulated materials, N and U biodegraded with a 339 modification of the appearance of both, surface and pore structure as a consequence of the 340 digestion procedure. Concretely, the most affected support was N. After the whole digestion 341 process, N and U nanoparticles lost the uniformity of pore structure (see TEM images) and 342 despite keeping its particle size and shape, there is a clear alteration of the surface (FESEM). In 343 those cases, besides the pore closure provoked by the apparition of new phases, the decrease 344 of the mesostructure observed by XRD and N2 adsorption-desorption isotherms can also be 345 originated, at least partially, by pore collapse. 346 347 Pérez-Esteve et al. 2016. Micropor Mesopor Mater 18 348 Figure 4. Characterization of particle size, particle shape and pore system of bare MSPs before and after the in vitro 349 digestion procedure (IDP). MCM-41 (M), MCM-41 nano (N) SBA-15 (S) and UVM-7 (U). 350 351 Silica degradation necessarily implies the break of siloxane bonds with the subsequent 352 generation of silanol groups and this was corroborated through the evolution of the 29Si NMR 353 spectra before and after digestion. For this study we selected two samples (S and N) that can 354 be considered as representative of the two observed behaviours with low (solid S) and high 355 (solid N) biodegradation according to TEM images. The 29Si NMR spectra are shown in Figure 5. 356 While in the case of sample S, the digestion does not affect the proportion of Q4:Q3:Q2 357 (68:30:2) sites, a decrease of the Q4 sites from 65% to 60% is observed in the case of the N 358 sample (Q4:Q3:Q2 from 65:31:4 to 60:33:7). 359 Pérez-Esteve et al. 2016. Micropor Mesopor Mater 19 360 361 Figure 5. 29Si NMR spectra for SBA-15 (a) and nanoparticulate MCM-41 (b), before (i) and after (ii) the in vitro 362 digestion procedure. 363 364 Having in mind these results, it is apparent that all studied MSPs are altered as a 365 consequence of the in vitro digestion process. However, the degradation degree depends on 366 the type and size of the particles. In this line, El Mourabit et al. studied the structure alteration 367 of several porous silica supports differing in particle size, particle shape, pore-size distribution, 368 specific surface area, pore volume and average of pore diameter caused by immersion in acid 369 solutions and found that the degradation of the supports was not obviously influenced by 370 textural properties of the particles [18]. Nevertheless, in our study, it seems to be clear that 371 particle size and wall thickness seem to be essential parameters that condition degradation. 372 In order to correlate the impact of each of the phases of the digestion with the particle’s 373 degradation, a further experiment was done. For this MCM-41 nanoparticles (solid N) were 374 selected given that this was the most affected support by the whole digestion process. For this 375 purpose, N was put in contact with water for 4 h. In parallel, a typical in vitro digestion process 376 (4 h) was performed. After each of these steps, samples were washed and observed by TEM. 377 Pérez-Esteve et al. 2016. Micropor Mesopor Mater 20 Figure 6 shows TEM micrographs of N after 4 h in contact with water (a) and after each of the 378 phases of the in vitro digestion process: buccal (b), gastric (c) and intestinal (d). As observed, 379 the particle size (ca. 100 nm) did not vary along the digestion suggesting that particle structure 380 remains unaltered after the whole digestion process. Moreover, surface and porosity of MCM381 41 remained unchanged after 4 h in water, meaning that particles do not collapse easily in 382 water solution. Particles are also intact after the 5 min of contact with simulated saliva. 383 However, particles change dramatically after the 2 h of gastric phase. In particular, after this 384 digestion step, particles loss clearly their spherical shape and ordered porous conformation 385 and become irregular shaped spheres with disordered porosity. Little differences among 386 particles observed after gastric and after both, gastric and intestinal phases were observed 387 suggesting that once the digestive solution is neutralized by the addition of intestinal juices, 388 the degradation process stopped. 389 390 391 Figure 6. TEM images showing particle size, particle shape and pore system of bare MCM-41 nanopaticles 392 (solid N) after 4 h in water, and buccal, gastric and intestinal phases of the in vitro digestion procedure. 393 394 These findings confirm that gastric phase (pH 2) is the responsive of particle’s degradation. The 395 role of acids in porous silica degradation has previously been described [18]. These authors 396 realised that protons play a role in the acidic alteration process of silica. Moreover, they 397 pointed out that in acidic conditions, anions (i.e. SO42-, Cl-, PO43-) present in the media can act 398 as nucleophilic catalysts accelerating the degradation reactions. Having in mind the 399 Pérez-Esteve et al. 2016. Micropor Mesopor Mater 21 composition of the two fluids comprising gastric phase of the digestion (i.e. saliva and gastric 400 juice) it can be say that all these degradative species (i.e. NaSO4, NaH2PO4, HCl, NaCl, KCl, CaCl2 401 and NH4Cl) are present in our studies. 402 403 3.3 Effect of in vitro digestion in microstructure of amine-functionalised particles 404 To investigate if the functionalization of the surface of the particles with certain organic 405 molecules have any influence in the preservation of the structure during the in vitro digestion, 406 the four particles object of study were functionalised with N1-(3407 trimethoxysilylpropyl)diethylenetriamine. This polyamine is one common organic molecule 408 used to prepare capped mesoporous silica particles able to modulate payload release in 409 response to pH changes [26]. After functionalization, particles were subjected to the digestion 410 process described in Figure 1. Digested particles were washed with water, dried and 411 characterized by XRD, TEM and FESEM. 412 Figure 2 shows XRD patterns of amine-functionalized particles before (ii) and after (iv) the 413 in vitro digestion process. In contrast to bare particles, functionalised particles showed the 414 same diffraction peaks before and after the digestion, which indicate a preservation of the 415 porous structure after the digestion procedure. The structure preservation was confirmed by 416 microscopic analysis. As observed in Figure 7, morphology, particle size and porous structure 417 of the different amine-functionalized supports is very similar before and after the in vitro 418 digestion process. These studies point out the role played by amines in the protection of 419 porous silica, especially in the nanoparticles (solids N and U), against the attack by acids and 420 chemical species present in the digestive juices. The prevention of silica degradation after 421 organic functionalization has also been observed by other authors. Lin et al. found that 422 degraded Si amounts from 42 nm diameter silica nanoparticles were greater than that from 423 the equivalent pegylated nanoparticles after both 10 days in deionized water and PBS at room 424 temperature and 37 °C [17]. Cauda et al. also observed that the attachment of a poly(ethylene 425 Pérez-Esteve et al. 2016. Micropor Mesopor Mater 22 glycol)-layer on the outer surface of colloidal mesoporous silica stabilized the particles by 426 reducing the rate of degradation in simulated body fluid at 37 °C for 1 month [15]. The 427 preventive effect of functional groups attached to the surface of the silica supports on the 428 degradation of porous silica could be attributed to the capability of these molecules to inhibit 429 the attack caused by acids and catalytic anions as well as to prevent the adsorption of 430 calcium/phosphate cations and the rearrangement of silicon species on a new “gel” fraction on 431 the walls of the particles. 432 433 434 Figure 7. Characterization of particle size, particle shape and pore system of amine-functionalised MSPs before and 435 after the in vitro digestion procedure. MCM-41 (M), MCM-41 nano (N) SBA-15 (S) and UVM-7 (U). 436 437 Pérez-Esteve et al. 2016. Micropor Mesopor Mater 23 In our study, according to CNH elemental analysis data, the post-functionalization degree 438 of our samples was comprised in the 0.14-0.15 mol of N1-(3-trimethoxysilylpropyl) groups per 439 100 g of sample. This value suppose a density of functional groups of ca. 1-1.5 groups/nm2 440 assuming that their incorporation is effective along the whole surface (external and internal). 441 As it is well known, the length of this organic group constitutes a serious drawback to achieve a 442 good diffusion and dispersion along the mesopores. Then, a certain accumulation and 443 condensation of the organic groups on the external surface and in the entrance of the 444 mesopores is expected. This relatively thick shell seems to be the responsible for the stability 445 of the silica supports after digestion. Moreover, in our case, amine groups attached to 446 particle’s surface can also able to locally neutralize the acidic environment created by HCl. 447 448 3.4 Effect of in vitro digestion in macroscopic structure 449 To determine the changes of MSPs structures and aggregation state along the whole in vitro 450 digestion procedure, confocal laser scanning microscopy (CLSM) micrographs were taken upon 451 staining M, N, S and U supports at different stage of the digestion with rhodamine B (Fig 8). 452 This study revealed that all the particles tended to form aggregates in water. This trend to 453 form large aggregates, especially observed in N and U, is in accordance with results previously 454 reported [27,28]. Figure 8 also allows observing that for all particles the gastric phase trend to 455 provoke an enlargement of particles aggregates, and that in none of the cases digestion 456 triggered the loss of macroscopic particle structure. 457 458 Pérez-Esteve et al. 2016. Micropor Mesopor Mater 24 459 Figure 8. Characterization of particle size and particle shape of bare MSPs before and after the in vitro digestion 460 procedure. MCM-41 micro (M), MCM-41 nano (N), SBA-15 (S) and UVM-7 (U). 461 462 Aggregation tendency observed in gastric phase for all particles was confirmed by particle 463 size distribution measurements using light diffraction. As shown in Figure 9, according to 464 particles’ grain size in different digestive media, S and M microparticles, exhibited a size 465 distribution in the range 0.5-2 m, while size distribution of N and U ranged from 5 up to 40 466 m. The inset graphs of the same figure shows size distribution of nano-sized silica present in 467 the digestion fluids. As presumable, only two samples (N and U) also shown particles in the 468 100-200 nm range, corresponding to those particles not participating in the particle’s clusters. 469 470