Optimal Surface Amino-Functionalization Following Thermo-Alkaline Treatment of Nanostructured Silica Adsorbents for Enhanced CO2 Adsorption

Special preparation of Santa Barbara Amorphous (SBA)-15, mesoporous silica with highly hexagonal ordered, these materials have been carried out for creating adsorbents exhibiting an enhanced and partially selective adsorption toward CO2. This creation starts from an adequate conditioning of the silica surface, via a thermo-alkaline treatment to increase the population of silanol species on it. CO2 adsorption is only reasonably achieved when the SiO2 surface becomes aminated after put in contact with a solution of an amino alkoxide compound in the right solvent. Unfunctionalized and amine-functionalized substrates were characterized through X-ray diffraction, N2 sorption, Raman spectroscopy, electron microscopy, 29Si solid-state Nuclear Magnetic Resonance (NMR), and NH3 thermal programmed desorption. These analyses proved that the thermo-alkaline procedure desilicates the substrate and eliminates the micropores (without affecting the SBA-15 capillaries), present in the original solid. NMR analysis confirms that the hydroxylated solid anchors more amino functionalizing molecules than the unhydroxylated material. The SBA-15 sample subjected to hydroxylation and amino-functionalization displays a high enthalpy of interaction, a reason why this solid is suitable for a strong deposition of CO2 but with the possibility of observing a low-pressure hysteresis phenomenon. Contrastingly, CH4 adsorption on amino-functionalized, hydroxylated SBA-15 substrates becomes almost five times lower than the CO2 one, thus giving proof of their selectivity toward CO2. Although the amount of retained CO2 is not yet similar to or higher than those determined in other investigations, the methodology herein described is still susceptible to optimization.


The Problem
The effective and abundant capture of CO 2 on the surface of nanoparticulate substrates is a fundamental experimental aim, since this gas plays a leading role in the upsurge of the greenhouse effect; this last phenomenon being nowadays one of the most important ecological problems to deal with. Several strategies for achieving satisfactory CO 2 levels of capture already exist; one implementation of particular and novel interest consists in the trapping and eventual in situ transformation of this gas into useful and environmentally friendly compounds, via chemical procedures. Hence, with the objective in mind of reaching an efficient carbon capture and storage The CO 2 molecule is a Lewis acid since this molecule can accept a pair of electrons from a base in order to establish a covalent coordinated bond. Therefore, the crucial steps for designing a proper CO 2 adsorbent include the selection of a substrate whose surface can be habilitated to possess a good number of basic sites; i.e., the surface has to be endowed with chemical species that can be combined with other matching molecules for attaining the desired characteristics.
Nanoporous silica supports [16][17][18] are provided, in general, with a high surface area and a considerable total pore volume. These characteristics make these substrates appropriate for adsorbing an assortment of chemical species. The employment of amphiphilic triblock copolymers as pore templating agents has allowed the creation of well-ordered mesoporous networks, which have been labeled as SBA-n (Santa Barbara) materials [27]. Among them, the SBA-15 [28] substrate results to have an ideal surface for performing an efficient and enhanced CO 2 adsorption. On the one hand, this material is formed by a collection of cylindrical capillaries, all practically having the same diameters, in a hexagonal arrangement and lying parallel to each other along micrometric lengths. On the other hand, the capillaries open at both ends that are forming these mesoporous structures are endowed of such sizes as to allow the free transit of different molecules along the nanocylinder axis. This is a very important property since the adsorptive molecules can enter, while the adsorbate molecules can exit the structure easily.
Besides, the functionalization of the surface via, for instance, amino-alkoxides (i.e., the type of compounds that we will be employing for the amination of the SiO 2 surface) can also benefit from this size characteristic, given that the deposition of molecules usually occurs after an adsorption from solution process. Another important property of amorphous silica is the presence of silanol (Si-OH) species throughout the surface. These species are determinant when the surface is to be functionalized with a given molecule. In the case of CO 2 , this molecule is in itself a weak Lewis acid, while the Si-OH groups are likewise. Then, in order for the silica surface to capture CO 2 , it has to be functionalized with pertinent species, e.g., amino-substituted alkoxides. The alkoxide groups can be attached to the surface after reacting with the silanol species; in turn, the CO 2 molecules become coordinated to the NH 2 groups of the amino-functionalizing compound.

Habilitation of the SiO 2 Surface to Attain an Efficient and Sufficient Amine-Functionalization Process
Silica by itself manifests no significant affinity toward CO 2 adsorption, for this reason, it is necessary to perform the chemical modification of this surface. This transformation can rely on the silanol groups, since the character and concentration of these species are crucial for attaining a widespread amino-functionalization of the SiO 2 surface. Importantly, because of the thermal treatment required to eliminate remnant pore templating species, a relatively small population of silanol groups (-SiOH) subsists in the final network. Therefore, a hydroxylation process via treatment of the SiO 2 substrate with a lowly concentrated NaOH solution, set at the right temperature, has been found to be a suitable operation for achieving an efficient hydroxylation of the surface. The reason of the low NaOH concentration to be used in this process resides in preventing an excessive desilication (i.e., dissolution and leaching) of the SiO 2 matrix when put in contact with an alkaline solution [29].
Nonetheless, the previous chemical procedure can still result useful, not only for attaining more SiOH species on the surface, but to eliminate the micropores present in the parent silica material.

Supplementary Aspects in Relation to the Choice of Santa Barbara Amorphous Silica No. 15 (SBA-15) for Enhanced CO 2 Trapping
Other aspects related to the convenience of choosing SiO 2 as substrate for CO 2 adsorption are the possibility of readily regenerating the adsorbent surface by means of a mild thermal treatment in an oven, thus reducing the power requirements necessary to perform the surface regeneration process. Another striking characteristic is the selectivity that the silica surface shows toward the adsorption of CO 2 instead of adsorptives such as CH 4 .
Furthermore, the deposition of a functionalizing molecule on the silica surface can be made from either an organic or inorganic solution of the species in question. In the case of amine-functionalization, it is better to perform this process employing an appropriate organic liquid for dissolving the amino alkoxide compound chosen for functionalizing the surface, e.g., toluene instead of ethanol. The reason of this is the chemical character of the silanol groups, which can interact synergistically or not with the solvent molecules with respect to an effective deposition of the functionalizing compound [30].
The objective of this work consists in showing evidence of a more abundant and efficient CO 2 capture on the basis of some appropriate textural changes that can be inflicted on model SBA-15 silica adsorbents via thermal and chemical surface functionalization procedures. A concomitant aim is to increase the number of silanol groups (i.e., -Si-OH) on the silica surface, since this is a key aspect for achieving an efficient functionalization of the surface via chemical anchorage methods. The ultimate goal is to accomplish an abundant capture of miscellaneous adsorptives (other than CO 2 ) for their storage and eventual transformation into other molecules of diverse practical interest.
This work is organized as follows. First, the experimental section detailing the synthesis of the silica substrates is presented. Next, the thermo-alkaline habilitation of the surface of these solids is described. The following step consists in carrying out the functionalization of the previous materials with an especially selected amino-alkoxide compound. Subsequently, N 2 sorption isotherms at 76 K are shown together with a list of the most important structural parameters of the porous solids involved in this investigation. CO 2 adsorption isotherms, the center of this research, in the range of 263-313 K on the diverse SBA-15 materials are presented. Additionally, Fourier Transform Infrared (FTIR), X-ray Diffraction (XRD), Transmission Electron Microscopy (TEM), Thermogravimetric Analysis (TGA), solid-state Nuclear Magnetic Resonance (NMR), and Temperature Programed Desorption (TPD) characterization experiments are depicted. Finally, CH 4 adsorption isotherms at 298 K on some of the SBA-15 adsorbents are presented, in order to ascertain the selectivity of these substrates between this last adsorptive and CO 2 .

Synthesis of Precursor SBA-15 Solids
The following molar ratio sequence was employed to prepare SBA-15 silica:TEOS:197H 2 O:6.2HCl:0.017P123, the synthesis was performed according to the method devised by Beck et al. [24] and Zhao et al. [31]. The ensuing material was washed, filtered, put in

Soft Thermo-Alkaline Treatment Performed over SBA-15 Precursor Materials
The mesoporous silica adsorbents of the SBA-15 (S15 denoted) kind were subjected to a thermo-(313 K) alkaline treatment through the employment of diluted (<1 M) NaOH solutions. Specifically, a 250 mg sample of each S15 substrate was subjected to a series of treatments under the previous conditions, each of these operations taking place every two hours. Subsequently, a desilication (i.e., hydroxylation) process was performed by adding a 0.025 M NaOH solution at 313 K (2 h). The resultant solid was washed with deionized water, filtered, and dried at 373 K for 2 h. The ensuing substrate was labeled as S15H.

Chemical Anchoring of 3-Aminopropyltriethoxysilane (APTES) on the Surface of SiO 2 Mesoporous Solids
The chemical bonding of APTES on the surfaces of the S15 and S15H silica substrates was achieved once these solids were calcined in air at 723 K for 2 h. The advantage of this route is to guarantee the total removal of the surfactant molecules after the substrate has been annealed at 723 K.
Before APTES surface deposition, the precursor SiO 2 solids were treated at 473 K under vacuum for 2 h in order to remove organic impurities and also to promote the activation of the silanol (SiOH) groups already residing on the surface. The molar ratio involved in this functionalization reaction corresponded to 3SiO 2 :2APTES and this procedure was carried out at 353 K under a N 2 atmosphere and continuous magnetic stirring of a reflux system during 20 h, while employing toluene as the APTES dispersion medium. The final solids were separated by filtration and dried at 373 K for 2 h. These materials were labeled as S15NH 2 and S15HNH 2 , given that these aminated materials proceed from the respective S15 and S15H samples.

CO 2 Sorption Experiments at Different Temperatures
Prior to CO 2 sorption experiments, each sample was outgassed at 373 K under high vacuum for 6 h. The adsorption temperature range utilized for these experiments was chosen between 273 and 298 K in step intervals of 10 K. Finally, the adsorption capacity of the amine-functionalized SiO 2 substrates was determined experimentally, in an automatic adsorption instrument (Quantachrome Autosorb 1LC, Quantachrome Instruments, Boynton Beach, FL, USA). Additional textural and thermal (isosteric heat of adsorption) properties were, calculated from the corresponding CO 2 adsorption isotherms evaluated at the different temperatures. CH 4 adsorption at 273 K, on some of the silica samples, was also carried out with the aim of visualizing the selectivity of the modified SBA-15 adsorbents towards CO 2 .

Middle-Infrared Spectroscopy Studies
The FTIR spectra in Figure 1 evidence vibrational changes of the molecules; therein, it is possible to note a prominent signal widening corresponding to physisorbed water in the S15H samples subjected to the hydroxylation treatment, if compared to the spectrum of the S15 precursor specimen. This difference surges even if all samples were previously annealed under vacuum at 473 K for 2 h. On the other hand, the intramolecular Si-OH (963 cm −1 ) and Si-O-Si flexion (802 cm −1 ) signals appear overlapped forming a broad band in the S15H spectrum. However, for hydroxylated samples, the band assigned to free silanol groups (Si-OH) appearing at 3769 cm −1 [32,33] results masked by the widening of the vibrational band of physisorbed water. Nevertheless, a larger amount of free SiOH groups sites interact with water through hydrogen bonds and promotes a rapidly hydration, which can be reflected in a larger amount of physisorbed water, again with respect to the same quantity in the S15 precursor sample. The samples under scrutiny were now functionalized through a chemical bonding method, i.e., a process employing APTES and that was already described in the experimental section. In this case, a covalent bond is established between the surface of the substrate and the bridging amino species (i.e., APTES) ( Figure 2). In this FTIR spectrum, it is possible to note silica vibrational bands as well as typical primary amine bands (-NH2), as in the case of APTES [34,35]. Notice that the bands corresponding to S15H, substrates previously subjected to a treatment with 0.025 M NaOH are the most intense of all, especially if these signals are compared to those obtained in the parent S15 specimen that is subjected to no alkaline exposure.

Raman Spectra of Precursor and Thermo Alkaline Treated Silica Samples
Raman spectra are useful in order to provide additional structural characterization, besides of that corresponding to the substrate itself, in particular to evidence the presence of free or vicinal silanol groups that are lying on the silica surface. The samples under scrutiny were now functionalized through a chemical bonding method, i.e., a process employing APTES and that was already described in the experimental section. In this case, a covalent bond is established between the surface of the substrate and the bridging amino species (i.e., APTES) ( Figure 2). In this FTIR spectrum, it is possible to note silica vibrational bands as well as typical primary amine bands (-NH 2 ), as in the case of APTES [34,35]. Notice that the bands corresponding to S15H, substrates previously subjected to a treatment with 0.025 M NaOH are the most intense of all, especially if these signals are compared to those obtained in the parent S15 specimen that is subjected to no alkaline exposure. The samples under scrutiny were now functionalized through a chemical bonding method, i.e., a process employing APTES and that was already described in the experimental section. In this case, a covalent bond is established between the surface of the substrate and the bridging amino species (i.e., APTES) ( Figure 2). In this FTIR spectrum, it is possible to note silica vibrational bands as well as typical primary amine bands (-NH2), as in the case of APTES [34,35]. Notice that the bands corresponding to S15H, substrates previously subjected to a treatment with 0.025 M NaOH are the most intense of all, especially if these signals are compared to those obtained in the parent S15 specimen that is subjected to no alkaline exposure.

Raman Spectra of Precursor and Thermo Alkaline Treated Silica Samples
Raman spectra are useful in order to provide additional structural characterization, besides of that corresponding to the substrate itself, in particular to evidence the presence of free or vicinal silanol groups that are lying on the silica surface.

Raman Spectra of Precursor and Thermo Alkaline Treated Silica Samples
Raman spectra are useful in order to provide additional structural characterization, besides of that corresponding to the substrate itself, in particular to evidence the presence of free or vicinal silanol groups that are lying on the silica surface.
The traditional Raman spectra of vitreous SiO 2 are characterized by bands appearing at around 410, 800, 1065, and 1200 cm −1 and are associated to vibrations of the amorphous silica network [36]. In our case, however, the bands that exist at around 410, 500, and 610 cm −1 can be linked to the vibrations of siloxane groups (Si-O-Si), which are assigned to rings of diverse sizes [37]. The signal at around 410 cm −1 is associated to the bending mode vibration of oxygen atoms inside a ring made of more than four silicon atoms (n > 4); this peak is usually known as the R band. The signal at around 490 cm −1 is associated to the breathing mode vibration of rings depicting four silicon atoms; this signal usually being known as the D 1 band; the peak appearing at around 605 cm −1 is associated to the breathing mode vibration of rings formed by three silicon atoms, and branded as D 2 bands [38]. Additionally, the peak at around 980 cm −1 is associated with the vibrational mode of (OH)-groups with respect to Si.
In the Raman spectra of the pristine S15 sample, it can be observed shallow R, D 1 , and D 2 bands which are associated to rings of diverse sizes existing in the silica network ( Figure 3). Additionally, the two bands appearing at around 850 and 1060 cm −1 can be linked [39] to the emissions located at around 850, 900, 950-1000, and 1050-1100 cm −1 and that can be due to tetrahedral silicon atoms having four, three, two or only one bridging oxygen atoms, respectively [40]. Non-bridging oxygen atoms could correspond to hydroxyl groups or to different metals, such as polysilicate alkaline cations [39]. In the Raman spectrum of the alkaline treated sample (S15H), an intense D 1 band appears at around 490 cm −1 . Besides, an intense and broad band appears from 600 to 1100 cm −1 . This signal pathway suggests the existence of a slightly high population of silica rings formed by four silicon atoms. Furthermore, the intense band located at higher frequencies suggests the presence of a higher amount of tetrahedral silicon atoms having no bridging oxygens as consequence of the alkaline treatment to which the samples were subjected. The traditional Raman spectra of vitreous SiO2 are characterized by bands appearing at around 410, 800, 1065, and 1200 cm −1 and are associated to vibrations of the amorphous silica network [36]. In our case, however, the bands that exist at around 410, 500, and 610 cm −1 can be linked to the vibrations of siloxane groups (Si-O-Si), which are assigned to rings of diverse sizes [37]. The signal at around 410 cm −1 is associated to the bending mode vibration of oxygen atoms inside a ring made of more than four silicon atoms (n > 4); this peak is usually known as the R band. The signal at around 490 cm −1 is associated to the breathing mode vibration of rings depicting four silicon atoms; this signal usually being known as the D1 band; the peak appearing at around 605 cm −1 is associated to the breathing mode vibration of rings formed by three silicon atoms, and branded as D2 bands [38]. Additionally, the peak at around 980 cm −1 is associated with the vibrational mode of (OH)-groups with respect to Si.
In the Raman spectra of the pristine S15 sample, it can be observed shallow R, D1, and D2 bands which are associated to rings of diverse sizes existing in the silica network ( Figure 3). Additionally, the two bands appearing at around 850 and 1060 cm −1 can be linked [39] to the emissions located at around 850, 900, 950-1000, and 1050-1100 cm −1 and that can be due to tetrahedral silicon atoms having four, three, two or only one bridging oxygen atoms, respectively [40]. Non-bridging oxygen atoms could correspond to hydroxyl groups or to different metals, such as polysilicate alkaline cations [39]. In the Raman spectrum of the alkaline treated sample (S15H), an intense D1 band appears at around 490 cm −1 . Besides, an intense and broad band appears from 600 to 1100 cm −1 . This signal pathway suggests the existence of a slightly high population of silica rings formed by four silicon atoms. Furthermore, the intense band located at higher frequencies suggests the presence of a higher amount of tetrahedral silicon atoms having no bridging oxygens as consequence of the alkaline treatment to which the samples were subjected. (a) S15 (b) S15H Figure 3. Raman spectra of SBA-15 materials: (a) precursor SBA-15 substrate (S15); and (b) silica after being treated with NaOH 0.025 M (S15H).

Thermogravimetric Analysis (TGA) Analysis of Precursor and Functionalized Silica Samples
The S15 precursor simple was subjected to TGA analysis ( Figure 4). The weight loss of 6.0 wt % observed at around 373 K could be due to water desorption. Afterwards, a slightly decreasing slope is developed as the temperature is raised and in such a way that when the analysis is fulfilled at 823 K an additional mass loss of 2.0 wt % is detected. A slight mass loss of about 2.0 wt % is possibly due to the partial dehydroxylation of the SiO2 matrix, or in other words, the Si-OH groups condense to generate water molecules and Si-O-Si siloxane species, consequently the total mass loss is about 8.0 wt %. Nonetheless, for the S15H sample subjected to hydroxylation, a mass loss of 11.0 wt % arises thus confirming that this sample was more hydrated than the precursor solid. This, in principle, may

Thermogravimetric Analysis (TGA) Analysis of Precursor and Functionalized Silica Samples
The S15 precursor simple was subjected to TGA analysis ( Figure 4). The weight loss of 6.0 wt % observed at around 373 K could be due to water desorption. Afterwards, a slightly decreasing slope is developed as the temperature is raised and in such a way that when the analysis is fulfilled at 823 K an additional mass loss of 2.0 wt % is detected. A slight mass loss of about 2.0 wt % is possibly due to the partial dehydroxylation of the SiO 2 matrix, or in other words, the Si-OH groups condense to generate water molecules and Si-O-Si siloxane species, consequently the total mass loss is about 8.0 wt %. Nonetheless, for the S15H sample subjected to hydroxylation, a mass loss of 11.0 wt % arises thus confirming that this sample was more hydrated than the precursor solid. This, in principle, may be due to the fact that the hydroxylation treatment causes the rupture of siloxane groups to generate vicinal Si-OH groups, which can be easily rehydrated and, as the temperature is raised during the TGA analysis, become dehydrated promoting, at the same time, the condensation of them, since, as is well known, silica hydration-condensation processes are reversible phenomena [41]. TGA analysis, become dehydrated promoting, at the same time, the condensation of them, since, as is well known, silica hydration-condensation processes are reversible phenomena [41]. Temperature (K) S15 S15H Figure 4. Thermogravimetric analysis under a N2 atmosphere of the S15 parent sample and of the S15H solid proceeding from the corresponding alkaline hydroxylation of the previous substrate.
The TGA study corresponding to the samples functionalized with APTES are shown in Figure 5; namely, for the S15NH2 and S15HNH2 substrates. The S15HNH2 hydroxylated sample suffers the highest mass loss of all. The first TGA region is linked to desorption of the physisorbed water at around 373 K; i.e., 3.0 wt % for the S15NH2 specimen, and 5.0 wt % for the S15HNH2 solid. Nevertheless, the significant mass loss depicted from 543 K onwards, is caused by APTES thermal degradation. Specifically, the S15NH2 material presents a mass loss of 14.1 wt % while the S15HNH2 solid corresponds to a mass loss of 17.5 wt %. This implies a difference between these specimens of 19.4 wt %, which evidences that a major weight loss occurs for the hydroxylated and subsequently APTES functionalized simple. Temperature (K) S15NH2 S15HNH2 Figure 5. Thermogravimetric analysis, under N2 flow, of the S15NH2 and S15HNH2 APTES-functionalized silica specimens.

X-ray Diffraction (XRD) Analysis at Low Angle
The XRD diffractograms obtained for both precursor and functionalized silica materials are shown in Figure 6. Therein, it is observed for all cases, the structural parameters of a pore network Figure 4. Thermogravimetric analysis under a N 2 atmosphere of the S15 parent sample and of the S15H solid proceeding from the corresponding alkaline hydroxylation of the previous substrate.
The TGA study corresponding to the samples functionalized with APTES are shown in Figure 5; namely, for the S15NH 2 and S15HNH 2 substrates. The S15HNH 2 hydroxylated sample suffers the highest mass loss of all. The first TGA region is linked to desorption of the physisorbed water at around 373 K; i.e., 3.0 wt % for the S15NH 2 specimen, and 5.0 wt % for the S15HNH 2 solid. Nevertheless, the significant mass loss depicted from 543 K onwards, is caused by APTES thermal degradation. Specifically, the S15NH 2 material presents a mass loss of 14.1 wt % while the S15HNH 2 solid corresponds to a mass loss of 17.5 wt %. This implies a difference between these specimens of 19.4 wt %, which evidences that a major weight loss occurs for the hydroxylated and subsequently APTES functionalized simple. TGA analysis, become dehydrated promoting, at the same time, the condensation of them, since, as is well known, silica hydration-condensation processes are reversible phenomena [41]. Temperature (K) S15 S15H Figure 4. Thermogravimetric analysis under a N2 atmosphere of the S15 parent sample and of the S15H solid proceeding from the corresponding alkaline hydroxylation of the previous substrate.
The TGA study corresponding to the samples functionalized with APTES are shown in Figure 5; namely, for the S15NH2 and S15HNH2 substrates. The S15HNH2 hydroxylated sample suffers the highest mass loss of all. The first TGA region is linked to desorption of the physisorbed water at around 373 K; i.e., 3.0 wt % for the S15NH2 specimen, and 5.0 wt % for the S15HNH2 solid. Nevertheless, the significant mass loss depicted from 543 K onwards, is caused by APTES thermal degradation. Specifically, the S15NH2 material presents a mass loss of 14.1 wt % while the S15HNH2 solid corresponds to a mass loss of 17.5 wt %. This implies a difference between these specimens of 19.4 wt %, which evidences that a major weight loss occurs for the hydroxylated and subsequently APTES functionalized simple. Temperature (K) S15NH2 S15HNH2 Figure 5. Thermogravimetric analysis, under N2 flow, of the S15NH2 and S15HNH2 APTES-functionalized silica specimens.

X-ray Diffraction (XRD) Analysis at Low Angle
The XRD diffractograms obtained for both precursor and functionalized silica materials are shown in Figure 6. Therein, it is observed for all cases, the structural parameters of a pore network Figure 5. Thermogravimetric analysis, under N 2 flow, of the S15NH 2 and S15HNH 2 APTES-functionalized silica specimens.

X-ray Diffraction (XRD) Analysis at Low Angle
The XRD diffractograms obtained for both precursor and functionalized silica materials are shown in Figure 6. Therein, it is observed for all cases, the structural parameters of a pore network consisting of cylindrical pore in a hexagonal packing with relatively sharp peaks appearing at (100), (110), and (200) planes, which are proof of this assertion.  Figure 6. Low angle X-ray Diffraction (XRD) diffractograms of S15, S15H, S15NH2, and S15HNH2 solids.
As can be seen in Figure 6, there is no change in the interplanar 100 distances depicted by both S15 precursor and S15NH2 functionalized samples. However, when comparing the Wd pore sizes obtained from Equation (1) [42] for the S15 and S15NH2 specimens, it is readily realized an evident pore size diminution provoked by the APTES presence inside the pores of the SBA-15 substrate.
where Vp and Vmic are the mesopore and micropore volumes, respectively, δ is the silica substrate density, and d100 is the 100 interplanar distance obtained from the XRD pattern.
In the case of the samples treated with alkaline solution, it can be noted a slight increase in the interplanar distance; this effect confirms that the hydroxylation treatment increases the pores size of the precursor solid without significantly altering the array and shape of cylindrical pores that constitute void volume of the substrate. Contrastingly, those samples functionalized with APTES suffered partial structural disarray, thus concluding that the S15HNH2 sample accept a highest amount of chemisorbed APTES in comparison to the S15NH2 sample, even if the two samples are put in contact with the same 3SiO2:2APTES molar ratio. This suggests that the hydroxylation process favored a large generation of free Si-OH sites. It is thus adequate to stress that the abundance and distribution of OH groups on the surface of silica adsorbents determine the occurrence of subsequent surface reactions.

Transmission Electron Microscopy (TEM)
The TEM micrographs shown in Figure 7 corroborate that the S15 precursor specimen consists of an ordered array of cylindrical capillaries disposed in a hexagonal packing, something that is evident in these images. Low angle X-ray Diffraction (XRD) diffractograms of S15, S15H, S15NH 2 , and S15HNH 2 solids.
As can be seen in Figure 6, there is no change in the interplanar 100 distances depicted by both S15 precursor and S15NH 2 functionalized samples. However, when comparing the W d pore sizes obtained from Equation (1) [42] for the S15 and S15NH 2 specimens, it is readily realized an evident pore size diminution provoked by the APTES presence inside the pores of the SBA-15 substrate.
where V p and V mic are the mesopore and micropore volumes, respectively, δ is the silica substrate density, and d 100 is the 100 interplanar distance obtained from the XRD pattern.
In the case of the samples treated with alkaline solution, it can be noted a slight increase in the interplanar distance; this effect confirms that the hydroxylation treatment increases the pores size of the precursor solid without significantly altering the array and shape of cylindrical pores that constitute void volume of the substrate. Contrastingly, those samples functionalized with APTES suffered partial structural disarray, thus concluding that the S15HNH 2 sample accept a highest amount of chemisorbed APTES in comparison to the S15NH 2 sample, even if the two samples are put in contact with the same 3SiO 2 :2APTES molar ratio. This suggests that the hydroxylation process favored a large generation of free Si-OH sites. It is thus adequate to stress that the abundance and distribution of OH groups on the surface of silica adsorbents determine the occurrence of subsequent surface reactions.

Transmission Electron Microscopy (TEM)
The TEM micrographs shown in Figure 7 corroborate that the S15 precursor specimen consists of an ordered array of cylindrical capillaries disposed in a hexagonal packing, something that is evident in these images. On the other hand, when the hydroxylation treatment is performed (S15H), there exist no obvious changes of the ordered silica pore arrangement with respect to the precursor S15 solid (Figure 8). This indicates, in a way similar to that suggested by the X-ray Diffraction analysis, that the sample subjected to the mild hydroxylation process, previously described, preserves the arrangement of its cylindrical voids and that only a slight lixiviation phenomenon occurs in a relatively uniform and soft manner. The TEM micrographs of the S15NH2 and S15HNH2 samples were not included in this manuscript because there did not exist obvious changes of the ordered silica pore arrangement with respect to the precursor S15 solid and the hydroxylated S15H samples (Figures 7 and 8). Furthermore, it is important to remember that the APTES deposition was performed at low coverages.

N2 Adsorption at 76 K
In Figure 9, N2 sorption isotherms at 76 K on precursor, S15, and hydroxylated, S15H, specimens and on their respective APTES functionalized solids (samples S15NH2 and S15HNH2) are displayed. The isotherms correspond to an IUPAC Type IV shape with an associated H1 hysteresis loop [43,44]. The figure inset in Figure 9, shows the pore-size distributions (PSD) of the same samples obtained by means of the Non Local Density Functional Theory (NLDFT) treatment for N2 desorption taking place along the boundary descending curve and according to a cylindrical pore model. Even if the dihydroxylation process is somewhat intense, because of lixiviation, the concentration of the caustic compound that was used to accomplish this end was sufficiently low as to keep the pore structure mostly unaffected, with practically no significant deformation of the pore shape. This can be also assessed by the close appearance of the XRD peak analysis. It is also pertinent to mention that the alkaline treatment, in the case of the S15 substrate, caused a pore width increment of 0.29 nm (see Table 1). On the other hand, when the hydroxylation treatment is performed (S15H), there exist no obvious changes of the ordered silica pore arrangement with respect to the precursor S15 solid (Figure 8). This indicates, in a way similar to that suggested by the X-ray Diffraction analysis, that the sample subjected to the mild hydroxylation process, previously described, preserves the arrangement of its cylindrical voids and that only a slight lixiviation phenomenon occurs in a relatively uniform and soft manner. The TEM micrographs of the S15NH 2 and S15HNH 2 samples were not included in this manuscript because there did not exist obvious changes of the ordered silica pore arrangement with respect to the precursor S15 solid and the hydroxylated S15H samples (Figures 7 and 8). Furthermore, it is important to remember that the APTES deposition was performed at low coverages. On the other hand, when the hydroxylation treatment is performed (S15H), there exist no obvious changes of the ordered silica pore arrangement with respect to the precursor S15 solid (Figure 8). This indicates, in a way similar to that suggested by the X-ray Diffraction analysis, that the sample subjected to the mild hydroxylation process, previously described, preserves the arrangement of its cylindrical voids and that only a slight lixiviation phenomenon occurs in a relatively uniform and soft manner. The TEM micrographs of the S15NH2 and S15HNH2 samples were not included in this manuscript because there did not exist obvious changes of the ordered silica pore arrangement with respect to the precursor S15 solid and the hydroxylated S15H samples (Figures 7 and 8). Furthermore, it is important to remember that the APTES deposition was performed at low coverages.

N2 Adsorption at 76 K
In Figure 9, N2 sorption isotherms at 76 K on precursor, S15, and hydroxylated, S15H, specimens and on their respective APTES functionalized solids (samples S15NH2 and S15HNH2) are displayed. The isotherms correspond to an IUPAC Type IV shape with an associated H1 hysteresis loop [43,44]. The figure inset in Figure 9, shows the pore-size distributions (PSD) of the same samples obtained by means of the Non Local Density Functional Theory (NLDFT) treatment for N2 desorption taking place along the boundary descending curve and according to a cylindrical pore model. Even if the dihydroxylation process is somewhat intense, because of lixiviation, the concentration of the caustic compound that was used to accomplish this end was sufficiently low as to keep the pore structure mostly unaffected, with practically no significant deformation of the pore shape. This can be also assessed by the close appearance of the XRD peak analysis. It is also pertinent to mention that the alkaline treatment, in the case of the S15 substrate, caused a pore width increment of 0.29 nm (see Table 1). Figure 8. TEM images of the SBA-15 original sample, previously to the NaOH mild hydroxylation treatment, of the S15H hydroxylated specimen. 5.6. N 2 Adsorption at 76 K In Figure 9, N 2 sorption isotherms at 76 K on precursor, S15, and hydroxylated, S15H, specimens and on their respective APTES functionalized solids (samples S15NH 2 and S15HNH 2 ) are displayed. The isotherms correspond to an IUPAC Type IV shape with an associated H1 hysteresis loop [43,44]. The figure inset in Figure 9, shows the pore-size distributions (PSD) of the same samples obtained by means of the Non Local Density Functional Theory (NLDFT) treatment for N 2 desorption taking place along the boundary descending curve and according to a cylindrical pore model. Even if the dihydroxylation process is somewhat intense, because of lixiviation, the concentration of the caustic compound that was used to accomplish this end was sufficiently low as to keep the pore structure mostly unaffected, with practically no significant deformation of the pore shape. This can be also assessed by the close appearance of the XRD peak analysis. It is also pertinent to mention that the alkaline treatment, in the case of the S15 substrate, caused a pore width increment of 0.29 nm (see Table 1). Figure 9. N2 sorption isotherms at 76 K and pore size distribution curves (figure inset) of SBA-15 substrates: S15, S15H, S15NH2, and S15HNH2. P/P 0 is the relative pressure, while P 0 represents the saturation pressure, finally STP means Standard Temperature and Pressure.
The main textural parameters of the different substrates were also calculated through the same N2 sorption analysis. The changes of the structural properties underwent by the S15 precursor sample can be due to both hydroxylation with NaOH and subsequent functionalization with APTES; these parameters are listed in Table 1. The temperature of 76 K corresponds to the approximate boiling point of N2 at Mexico City's altitude conditions. ABET is the Brunauer-Emmett-Tell (BET) surface area, Vt is the total adsorbed volume, Aext is the external surface area, Amic is the virtual micropore surface area and Dmode is the modal pore diameter arising from the Non Local Density Functional Theory (NLDFT) approach by employing a kernel that utilizes the boundary desorption isotherm relative to N2 at 76 K in cylindrical pores.
The above results confirm the beneficial effect of subjecting the surface of the substrates to a previous hydroxylation treatment that produces the elimination of micropores (i.e., the loss of Amic), and concomitantly, an increase of the mean pore size favored by the deposition of APTES on the surface of mesopores, while involving no pore mouth blocking. This happens since when SiO2 substrates are exposed to a hydroxylation process, some silica lixiviation occurs, the intensity of which depends on the concentration of the caustic solution; this process being known as alkaline desilication. It is also important to remember that the hydroxylation operation was performed at 313 K. Now, when comparing the textural parameters listed in Table 1 and associated to the pioneering materials and their corresponding solids obtained after APTES functionalization of the surface, the next important effects are as follows: (1) Compared to the precursor unhydroxylated substrate (S15), the presence of APTES has caused a decrease of the mean pore size in the S15HNH2 sample by 10.1%. Likewise, a decrease of 40.6% of the total pore volume. In turn, the S15H adsorbent depicted a pore size diminution of 13.2% and a Figure 9. N 2 sorption isotherms at 76 K and pore size distribution curves (figure inset) of SBA-15 substrates: S15, S15H, S15NH 2 , and S15HNH 2 . P/P 0 is the relative pressure, while P 0 represents the saturation pressure, finally STP means Standard Temperature and Pressure. A BET is the Brunauer-Emmett-Tell (BET) surface area, V t is the total adsorbed volume, A ext is the external surface area, A mic is the virtual micropore surface area and D mode is the modal pore diameter arising from the Non Local Density Functional Theory (NLDFT) approach by employing a kernel that utilizes the boundary desorption isotherm relative to N 2 at 76 K in cylindrical pores.
The main textural parameters of the different substrates were also calculated through the same N 2 sorption analysis. The changes of the structural properties underwent by the S15 precursor sample can be due to both hydroxylation with NaOH and subsequent functionalization with APTES; these parameters are listed in Table 1. The temperature of 76 K corresponds to the approximate boiling point of N 2 at Mexico City's altitude conditions.
The above results confirm the beneficial effect of subjecting the surface of the substrates to a previous hydroxylation treatment that produces the elimination of micropores (i.e., the loss of A mic ), and concomitantly, an increase of the mean pore size favored by the deposition of APTES on the surface of mesopores, while involving no pore mouth blocking. This happens since when SiO 2 substrates are exposed to a hydroxylation process, some silica lixiviation occurs, the intensity of which depends on the concentration of the caustic solution; this process being known as alkaline desilication. It is also important to remember that the hydroxylation operation was performed at 313 K. Now, when comparing the textural parameters listed in Table 1 and associated to the pioneering materials and their corresponding solids obtained after APTES functionalization of the surface, the next important effects are as follows: (1) Compared to the precursor unhydroxylated substrate (S15), the presence of APTES has caused a decrease of the mean pore size in the S15HNH 2 sample by 10.1%. Likewise, a decrease of 40.6% of the total pore volume. In turn, the S15H adsorbent depicted a pore size diminution of 13.2% and a reduction of 44.1% in the total pore volume. These facts suggest that more APTES molecules have been anchored on the hydroxylated specimen with respect to the unhydroxylated specimen. (2) Additionally, the surface area decreased by 10.4% with respect to its original magnitude (S15).
As the textural analysis shows, the S15H material depicts no micropores, thus suggesting that practically all the APTES molecules reside in the interior of mesopores and that they are uniformly dispersed on the surface. However, the disappearance of micropores from the S15 solid indicates that the access to these voids is somewhat precluded by the APTES species, perhaps anchored at pore entrances.
From the above statements, it is evident that the S15H adsorbent becomes the most efficient toward APTES capture if compared to the S15 parent specimen. The origin of this phenomenon may reside on the generation of higher number of free silanol groups (Si-OH) on the silica surface by action of the NaOH alkaline treatment.

NH 3 Thermal Programmed Desorption from Silica Specimens (TPD-NH 3 )
The amount of acid sites on the surface of the different adsorbents was determined by the amount of NH 3 desorbed from the surface in the temperature range of 298-923 K. The results are schematically shown in Figure 10. In this figure, the silica solids depict two types of sites: (i) weak sites are evident by the band appearing from 443 to 673 K; and (ii) sites considered strong appear between 743 and 923 K. This implies the presence of Si-OH centers on the surface of the silica adsorbents, which are responsible of NH 3 adsorption. The tendency found with respect to the abundance of acid centers, in our samples, is as follows: S15H >> S15. This result confirms that the hydroxylation treatment employed induces the formation of a greater number of acid sites, a fact that is reflected by the area displayed under each of the TPD peaks. reduction of 44.1% in the total pore volume. These facts suggest that more APTES molecules have been anchored on the hydroxylated specimen with respect to the unhydroxylated specimen. (2) Additionally, the surface area decreased by 10.4% with respect to its original magnitude (S15).
As the textural analysis shows, the S15H material depicts no micropores, thus suggesting that practically all the APTES molecules reside in the interior of mesopores and that they are uniformly dispersed on the surface. However, the disappearance of micropores from the S15 solid indicates that the access to these voids is somewhat precluded by the APTES species, perhaps anchored at pore entrances.
From the above statements, it is evident that the S15H adsorbent becomes the most efficient toward APTES capture if compared to the S15 parent specimen. The origin of this phenomenon may reside on the generation of higher number of free silanol groups (Si-OH) on the silica surface by action of the NaOH alkaline treatment.

NH3 Thermal Programmed Desorption from Silica Specimens (TPD-NH3)
The amount of acid sites on the surface of the different adsorbents was determined by the amount of NH3 desorbed from the surface in the temperature range of 298-923 K. The results are schematically shown in Figure 10. In this figure, the silica solids depict two types of sites: (i) weak sites are evident by the band appearing from 443 to 673 K; and (ii) sites considered strong appear between 743 and 923 K. This implies the presence of Si-OH centers on the surface of the silica adsorbents, which are responsible of NH3 adsorption. The tendency found with respect to the abundance of acid centers, in our samples, is as follows: S15H >> S15. This result confirms that the hydroxylation treatment employed induces the formation of a greater number of acid sites, a fact that is reflected by the area displayed under each of the TPD peaks. Intensity (u.a) Temperature (K) S15 S15H Figure 10. NH3 thermal programmed desorption from silica specimens (TPD-NH3) analysis of the S15 precursor solid as well as for the S15H hydroxylated material following thermo-alkaline treatment. Figure 11 shows an assessment of the High Power Decoupling (HPDEC) NMR spectra corresponding to the precursor, still unfunctionalized SBA-15 silica materials. On the other hand, Figure 12 depicts a comparison between the NMR spectra of the S15H hydroxylated substrate before and after being chemically functionalized with APTES. Finally, the NMR spectra displayed in Figure 13 shows a comparison of the signals emitted by the S15 (unhydroxylated) and S15H (hydroxylated) solids after being modified on their surface by the anchoring of APTES. All these NMR spectra exhibit signals appearing at chemical displacements of −112, −103, and −93 ppm and are due to the silica structure, and denoted as Q 4 , Q 3 , and Q 2 , respectively. The Q 2 , Q 3 , and Q 4 bands are referred to silicon Figure 10. NH 3 thermal programmed desorption from silica specimens (TPD-NH 3 ) analysis of the S15 precursor solid as well as for the S15H hydroxylated material following thermo-alkaline treatment. Figure 11 shows an assessment of the High Power Decoupling (HPDEC) NMR spectra corresponding to the precursor, still unfunctionalized SBA-15 silica materials. On the other hand, Figure 12 depicts a comparison between the NMR spectra of the S15H hydroxylated substrate before and after being chemically functionalized with APTES. Finally, the NMR spectra displayed in Figure 13 shows a comparison of the signals emitted by the S15 (unhydroxylated) and S15H (hydroxylated) solids after being modified on their surface by the anchoring of APTES. All these NMR spectra exhibit signals appearing at chemical displacements of −112, −103, and −93 ppm and are due to the silica structure, and denoted as Q 4 , Q 3 , and Q 2 , respectively. The Q 2 , Q 3 , and Q 4 bands are referred to silicon atoms bonded to two, three or four siloxane chains (-O-Si-) and the respective silanol groups (Si-OH) in order to complete four unions, respectively [45,46].

Nuclear Magnetic Resonance (NMR) Studies
The 29 Si NMR spectra (HPDEC) of the S15 and S15H samples, Figure 11 exposes similar population of Q 4 and Q 2 type silicon sites and an increased one of the Q 3 type for the hydroxylated sample. This situation suggests that the pristine SBA-15 network maintains most of its silicon atoms totally integrated (condensed) into the network (Q 4 ) while still possessing silicon atoms having one or two hydroxyl groups, but in lower amounts (Q 3 and Q 2 ); Furthermore, the above evidence suggests that hydroxylation promotes the formation of silicon atoms depicting a solitary hydroxyl group (Q 3 ), as reveals the spectrum of sample S15H of Figure 11. atoms bonded to two, three or four siloxane chains (-O-Si-) and the respective silanol groups (Si-OH) in order to complete four unions, respectively [45,46].   Figure 11. 29 Si Nuclear Magnetic Resonance (NMR) spectra (High Power Decoupling) for the S15 and S15H samples. The figure inset exemplifies a peak analysis performed after deconvolution of the set of different signals obtained through 29 Si NMR analysis on the S15 solid. Intensity (u.a) 29 Si (ppm) S15H S15HNH2 Figure 12. 29 Si NMR spectra (HPDEC) for the S15H precursor and APTES (S15HNH2) modified samples. Figure 11. 29 Si Nuclear Magnetic Resonance (NMR) spectra (High Power Decoupling) for the S15 and S15H samples. The figure inset exemplifies a peak analysis performed after deconvolution of the set of different signals obtained through 29 Si NMR analysis on the S15 solid.  Figure 11. 29 Si Nuclear Magnetic Resonance (NMR) spectra (High Power Decoupling) for the S15 and S15H samples. The figure inset exemplifies a peak analysis performed after deconvolution of the set of different signals obtained through 29 Si NMR analysis on the S15 solid. Intensity (u.a) 29 Si (ppm) S15H S15HNH2 Figure 12. 29 Si NMR spectra (HPDEC) for the S15H precursor and APTES (S15HNH2) modified samples. Figure 12. 29 Si NMR spectra (HPDEC) for the S15H precursor and APTES (S15HNH 2 ) modified samples.
The signals centered at around −69 and −59 ppm observed in the spectra of Figures 12 and 13 are associated to silicon atoms bonded to organic chains (Si-C) and labeled as T 3 and T 2 signals, which are equally distributed [45,46] and merge as consequence of the chemical bond established between the free silanol groups and the APTES functionalizing molecule. These bands can be attributed to the existence of silicon atoms bonded to carbon atoms and either two (T 2 ) or three (T 3 ) siloxane chains (Si-O-Si); additional surface hydroxyl groups that are required for completing the four unions of each silicon atom [45,46]. Therefore, therein it is evident the presence of T 3 and T 2 signals for the S15HNH 2 solid, thus confirming the efficient anchoring of APTES on the surface of the precursor S15H material. Intensity (u.a) 29 Si (ppm) S15NH2 S15HNH2 Q 4 Q 3 Q 2 T 3 T 2 Figure 13. 29 Si NMR (HPDEC) spectra of solids modified on the surface via APTES: S15NH2 and S15HNH2 substrates.
The 29 Si NMR spectra (HPDEC) of the S15 and S15H samples, Figure 11 exposes similar population of Q 4 and Q 2 type silicon sites and an increased one of the Q 3 type for the hydroxylated sample. This situation suggests that the pristine SBA-15 network maintains most of its silicon atoms totally integrated (condensed) into the network (Q 4 ) while still possessing silicon atoms having one or two hydroxyl groups, but in lower amounts (Q 3 and Q 2 ); Furthermore, the above evidence suggests that hydroxylation promotes the formation of silicon atoms depicting a solitary hydroxyl group (Q 3 ), as reveals the spectrum of sample S15H of Figure 11.
The signals centered at around −69 and −59 ppm observed in the spectra of Figures 12 and 13 are associated to silicon atoms bonded to organic chains (Si-C) and labeled as T 3 and T 2 signals, which are equally distributed [45,46] and merge as consequence of the chemical bond established between the free silanol groups and the APTES functionalizing molecule. These bands can be attributed to the existence of silicon atoms bonded to carbon atoms and either two (T 2 ) or three (T 3 ) siloxane chains (Si-O-Si); additional surface hydroxyl groups that are required for completing the four unions of each silicon atom [45,46]. Therefore, therein it is evident the presence of T 3 and T 2 signals for the S15HNH2 solid, thus confirming the efficient anchoring of APTES on the surface of the precursor S15H material.
However, in the case of the 29 Si NMR spectra in Figure 12 there can be also observed a slight chemical shift of the Q 4 and Q 3 bands for the case of the APTES functionalized sample (S15HNH2), with respect to the signals of the S15H hydroxylated solid (S15H). This could be due to a kind of screening effect because of the presence of APTES in the functionalized material, something that modifies the chemical ambient around the silicon atoms causing some shrinkage of the starting silica matrix. The emergence of a signal between the Q 2 and Q 3 bands in the spectrum of the S15HNH2 can be interpreted as consequence of the surface attaching of APTES, preferably trough one or two unions established over the silicon atoms having one hydroxyl group (Q 3 ). This assumption is confirmed by the apparition the T 2 and T 3 groups. In other words, the attachment occurs by the substitution of one or two of the ethoxy groups (C2H5O-) of the APTES molecule with similar number of surface hydroxyl group (Si-OH) of the mesoporous substrate. Then, the slight increase of the Q 2 population can be due to the integration of new silicon atoms of the APTES molecule with two remnant hydroxyl groups as consequence of the hydrolysis of their ethoxy groups. In order to avoid confusion, we have to remember that it is a well-known fact that the alkoxide groups of APTES or of any silicon alkoxide are readily substituted by the hydroxyl groups arising from water. However, in the absence of this Figure 13. 29 Si NMR (HPDEC) spectra of solids modified on the surface via APTES: S15NH 2 and S15HNH 2 substrates.
However, in the case of the 29 Si NMR spectra in Figure 12 there can be also observed a slight chemical shift of the Q 4 and Q 3 bands for the case of the APTES functionalized sample (S15HNH 2 ), with respect to the signals of the S15H hydroxylated solid (S15H). This could be due to a kind of screening effect because of the presence of APTES in the functionalized material, something that modifies the chemical ambient around the silicon atoms causing some shrinkage of the starting silica matrix. The emergence of a signal between the Q 2 and Q 3 bands in the spectrum of the S15HNH 2 can be interpreted as consequence of the surface attaching of APTES, preferably trough one or two unions established over the silicon atoms having one hydroxyl group (Q 3 ). This assumption is confirmed by the apparition the T 2 and T 3 groups. In other words, the attachment occurs by the substitution of one or two of the ethoxy groups (C 2 H 5 O-) of the APTES molecule with similar number of surface hydroxyl group (Si-OH) of the mesoporous substrate. Then, the slight increase of the Q 2 population can be due to the integration of new silicon atoms of the APTES molecule with two remnant hydroxyl groups as consequence of the hydrolysis of their ethoxy groups. In order to avoid confusion, we have to remember that it is a well-known fact that the alkoxide groups of APTES or of any silicon alkoxide are readily substituted by the hydroxyl groups arising from water. However, in the absence of this substance, the only way for reaching condensation consists in the substitution of alkoxide groups by oxygen atoms proceeding from hydroxyl groups with the subsequent formation of the respective alcohol. Figure 13 also depicts the same shifting effect, i.e., for the S15HNH 2 substrate a chemical shift of about 4 ppm is evident with respect to the signals emitted by the S15NH 2 solid [45,46]. Therefore, the S15HNH 2 silica matrix provides emissions toward higher chemical displacements; this can be due to the change in the chemical environment around the silicon atoms or a contraction of the porous matrix caused by the APTES species existing on the pore walls of the previously hydroxylated specimen; this can be confirmed after N 2 adsorption characterization.
This event is not surprising given that chemical displacements are very sensible to the chemical environment, and these conditions can be different even for the same molecule on the same support, depending on the amount of these chemisorbed species. In the same fashion, it is also evident in Figure 13 that the S15HNH 2 solid possesses a slightly higher amount of T 3 groups, if compared to those present in the S15NH 2 specimen. This effect is also reflected on the relative decrease of the amount of available Q 3 groups and due to Si-OH groups that have been employed to bind the APTES molecules to the surface. However, in this S15NH 2 sample the T 2 band displays a higher intensity than the T 3 signal.
The intensity differences found between the S15NH 2 and S15HNH 2 samples confirm that a condensation reaction occurred between the hydroxyl groups (Q 2 and Q 3 ) on the respective surface of the substrate (S15 or S15H) and two or one ethoxy groups of APTES. For this reason, on the S15NH 2 sample the most intense T 2 band (having one hydroxyl group) and the increased Q 3 band can only be associated to the presence of the silicon atoms of APTES. Additionally, the higher intensity of the T 3 band in the S15HNH 2 sample suggests a higher integration or condensation of the silicon atom of APTES as consequence of higher amount of vicinal silanol groups over the surface of the S15H substrate used.
Finally, in Table 2, the areal percentages associated to the Q and T NMR peaks are listed, once an appropriate deconvolution of these signal patterns has been performed (just as it was exemplified in Figure 11 for the case of the S15 sample). Table 2. Areal percent of signals proceeding from Nuclear Magnetic Resonance (NMR) spectra obtained from the deconvolution of the respective signals of precursor and amine-functionalized substrates. The results listed in Table 2 confirm the statements induced from the aspects of Figures 11-13. In this way, the argument advanced from the NMR characteristics of the S15 and S15H materials corroborates that these materials are appropriate for the anchoring of different amounts of APTES on their surface. Even if every synthesis is carried out, under the same conditions and employing the same amount of APTES, the S15H specimen adsorbs a larger amount of this compound. The net effect of hydrolyzing (S15H) is reflected in a larger proportion of Q 3 species, i.e., free or adjacent silanol groups. There exists a difference of 15.6% in the amount of Q 3 species between the S15 and S15H substrates. Besides, it can also be observed that the hydroxylated matrix appears to be less condensed than the matrix that was not subjected to this treatment; in this respect, just check the Q 4 ratios against each other for the two materials.

Material
In the same way, when comparing the functionalized solids, the largest amounts of T 2 and T 3 signals were found in the S15HNH 2 sample rather than in the S15NH 2 solid, something that can only be associated to the attached APTES species. Once the respective areal ratios are observed and considering the molar ratios chosen for carrying out the chemical anchoring on silica, as described in the Section 3.4, it is found (in a semiquantitative way) that 1.85 and 3.3 mmol of APTES have been anchored on the surface of the S15NH 2 and S15HNH 2 samples, respectively. This difference strongly confirms that the soft alkaline hydroxylation improves as much as 43.8% the anchoring efficiency of APTES towards the SiO 2 surface (Table 2).
In brief, the hydroxylation alkaline process generates a higher efficiency toward the anchoring of APTES molecules than the unhydroxylated pioneering specimen does. This is mostly due to the larger amount of silanol groups on the pore walls of the hydroxylated samples. Under the presence of these species, the anchoring of the functionalizing amine species occurs more efficiently under molar ratios of 2:1 and 1:1, In contrast the unfunctionalized S15 precursor specimen is only able to anchor APTES molecules under a 2:1 Si-OH: APTES molar ratio.

CO 2 Adsorption Studies and Determination of the Isosteric Enthalpy of Adsorption
CO 2 adsorption isotherms on the assorted four silica substrates were determined at 263, 273, 283, 293 and 298 K or 303 K for each one of them (Figures 14-16). When the isotherms of the unfunctionalized substrates (i.e., S15 and S15H) were analyzed, it was observed that the hydroxylation treatment endowed these adsorbents with a higher CO 2 adsorption capacity ( Figure 14). Nevertheless, the values of the isosteric enthalpy of adsorption evidence a relatively weak interaction (since these data are rather lower than the enthalpy of sublimation of CO 2 ) between the adsorbent surface and the CO 2 molecules. The possible reason for this behavior can be due to the intrinsic acidic nature of the silica surface [47]; there is no a drastic difference in the isosteric enthalpies depicted by the S15 and S15H silica samples Figure 15.
In brief, the hydroxylation alkaline process generates a higher efficiency toward the anchoring of APTES molecules than the unhydroxylated pioneering specimen does. This is mostly due to the larger amount of silanol groups on the pore walls of the hydroxylated samples. Under the presence of these species, the anchoring of the functionalizing amine species occurs more efficiently under molar ratios of 2:1 and 1:1, In contrast the unfunctionalized S15 precursor specimen is only able to anchor APTES molecules under a 2:1 Si-OH: APTES molar ratio.

CO2 Adsorption Studies and Determination of the Isosteric Enthalpy of Adsorption
CO2 adsorption isotherms on the assorted four silica substrates were determined at 263, 273, 283, 293 and 298 K or 303 K for each one of them (Figures 14-16). When the isotherms of the unfunctionalized substrates (i.e., S15 and S15H) were analyzed, it was observed that the hydroxylation treatment endowed these adsorbents with a higher CO2 adsorption capacity ( Figure 14). Nevertheless, the values of the isosteric enthalpy of adsorption evidence a relatively weak interaction (since these data are rather lower than the enthalpy of sublimation of CO2) between the adsorbent surface and the CO2 molecules. The possible reason for this behavior can be due to the intrinsic acidic nature of the silica surface [47]; there is no a drastic difference in the isosteric enthalpies depicted by the S15 and S15H silica samples Figure 15. Pressure (Bar) Isotherms at 293 K S15 S15H Figure 14. CO2 adsorption at 293 K, on SBA-15 precursor (S15) samples and hydroxylated (S15H) samples. Adsorbed amount (mmol g -1 ) S15 S15H Figure 15. CO2 adsorption enthalpies of SBA-15 precursor and hydroxylated silicas (S15 and S15H). confirms that the soft alkaline hydroxylation improves as much as 43.8% the anchoring efficiency of APTES towards the SiO2 surface (Table 2). In brief, the hydroxylation alkaline process generates a higher efficiency toward the anchoring of APTES molecules than the unhydroxylated pioneering specimen does. This is mostly due to the larger amount of silanol groups on the pore walls of the hydroxylated samples. Under the presence of these species, the anchoring of the functionalizing amine species occurs more efficiently under molar ratios of 2:1 and 1:1, In contrast the unfunctionalized S15 precursor specimen is only able to anchor APTES molecules under a 2:1 Si-OH: APTES molar ratio.

CO2 Adsorption Studies and Determination of the Isosteric Enthalpy of Adsorption
CO2 adsorption isotherms on the assorted four silica substrates were determined at 263, 273, 283, 293 and 298 K or 303 K for each one of them (Figures 14-16). When the isotherms of the unfunctionalized substrates (i.e., S15 and S15H) were analyzed, it was observed that the hydroxylation treatment endowed these adsorbents with a higher CO2 adsorption capacity ( Figure 14). Nevertheless, the values of the isosteric enthalpy of adsorption evidence a relatively weak interaction (since these data are rather lower than the enthalpy of sublimation of CO2) between the adsorbent surface and the CO2 molecules. The possible reason for this behavior can be due to the intrinsic acidic nature of the silica surface [47]; there is no a drastic difference in the isosteric enthalpies depicted by the S15 and S15H silica samples Figure 15. Pressure (Bar) Isotherms at 293 K S15 S15H Figure 14. CO2 adsorption at 293 K, on SBA-15 precursor (S15) samples and hydroxylated (S15H) samples. Adsorbed amount (mmol g -1 ) S15 S15H Figure 15. CO2 adsorption enthalpies of SBA-15 precursor and hydroxylated silicas (S15 and S15H). Figure 15. CO 2 adsorption enthalpies of SBA-15 precursor and hydroxylated silicas (S15 and S15H).
On the other hand, the CO 2 adsorption isotherms corresponding to the APTES functionalized substrates confirm that the hydroxylated specimens better contribute to a more efficient and energetic CO 2 capture at different temperatures (Figures 16 and 17). In addition, the affinity toward CO 2 of amino-functionalized silica substrates (hydroxylated or not) is better than that depicted by the unfunctionalized solids.
On the other hand, the CO2 adsorption isotherms corresponding to the APTES functionalized substrates confirm that the hydroxylated specimens better contribute to a more efficient and energetic CO2 capture at different temperatures (Figures 16 and 17). In addition, the affinity toward CO2 of amino-functionalized silica substrates (hydroxylated or not) is better than that depicted by the unfunctionalized solids. Isotherms at 273.1 K S15NH2-CO2 S15HNH2-CO2 S15NH2-CH4 S15HNH2-CH4 Figure 16. CO2 and CH4 adsorption isotherms at 273 K on APTES functionalized materials SBA-15 (pristine S15) and hydroxylated silica S15H. Figure 16 displays a comparison between CO2 and CH4 adsorption at 273 K on the precursor and hydroxylated amino-functionalized silica materials. Therein, it can be appreciated that the aforementioned substrates present a strong selectivity toward CO2 rather than to CH4. This could be due to the fact that the methane molecule is non-polar and it is not establishing a meaningful attractive interaction with the amine-functionalized SiO2 surface. The adsorbed volumes shown by the silica samples toward CH4 are too low for being reflected in a significant CH4 uptake. These contrasting adsorption characteristics between CO2 and CH4 will indeed be very useful (i.e., selective) for achieving a fine separation of mixtures including these gases. Likewise, note the better affinity at 298 K towards CO2 depicted by S15HNH2 results to be 19.2% higher than the amount adsorbed by the S15NH2 matrix at 1 bar. However, even at 298 K and 0.15 bar, there is a difference between the two samples of 36.2% ( Figure 17). Even at 0.15 bar, S15HNH2 shows a better performance at 313 K than the S15NH2 specimen, thus stressing this way that the thermos-alkaline treatment, previous to APTES surface functionalization, is very convenient in order to significantly raise CO2 capture capacity with respect to the amount adsorbed by the APTES functionalized but not previously hydroxylated S15NH2 substrate. Therefore, the S15HNH2 solid is more recommendable than the conventional (i.e., not hydroxylated) samples for pursuing enhanced CO2 adsorption this effect being valid at 0.15 bar and temperature range of 318-348 K [48]. It is also important to mention that the CO2 isotherms shown in Figure 17 depict low-pressure hysteresis loops at 298 K. However, this phenomenon is more accentuated for the S15HNH2 specimen than for the S15NH2 material. This suggests that there may exist a stronger interaction between CO2 and the hydroxylated surface (before being functionalized with APTES) than with the analogous but unfunctionalized sample [49]. Pressure (bar) S15NH2-298 K S15HNH2-298 K S15HNH2-313 K Figure 17. CO2 adsorption isotherms at 298 K on APTES functionalized materials SBA-15 (pristine and previously treated SBA-15) and isotherm at 313 K for the sample SBA-15 hydroxylated and functionalized with APTES. In all cases the adsorption and desorption processes depict upper and lower pathways, respectively.
Importantly, the S15HNH2 even if substrate presents the best performance in terms of CO2 adsorption. The total adsorbed volume of this gas is not too high when compared to other SBA-15 materials reported in the literature. Nevertheless, this comparison become relative ever since the overall CO2 capture capacity depends on a number of factors, among them: (1) The synthesis conditions and the method employed to eliminate the templating block copolymer to leave behind the pore entities. This process is intrinsically linked to the amount of SiOH groups on the surface [50,51]. (2) The method of amine deposition on the silica surface (in-situ or ex-situ) [52,53].
(3) The nature of the amine molecule to be surface deposited on the pore walls [54,55]. (4) The quantity of amine compound employed for the surface functionalization [35]. (5) The functionalization method (e.g., impregnation, chemical anchorage or a combination of both).
Olea et al. [56] have taken into account most of the above factors, including the exploration of different surface functionalization approaches of SBA-15 materials. (6) The experimental conditions selected for CO2 capture (e.g., temperature, pressure, CO2 flow rate, etc.).
Nonetheless, the present work demonstrates that the alkaline pretreatment is adequate for favorably allowing the polarity of the conventional SBA-15 surface in such a way that the substrates Figure 17. CO 2 adsorption isotherms at 298 K on APTES functionalized materials SBA-15 (pristine and previously treated SBA-15) and isotherm at 313 K for the sample SBA-15 hydroxylated and functionalized with APTES. In all cases the adsorption and desorption processes depict upper and lower pathways, respectively. Figure 16 displays a comparison between CO 2 and CH 4 adsorption at 273 K on the precursor and hydroxylated amino-functionalized silica materials. Therein, it can be appreciated that the aforementioned substrates present a strong selectivity toward CO 2 rather than to CH 4 . This could be due to the fact that the methane molecule is non-polar and it is not establishing a meaningful attractive interaction with the amine-functionalized SiO 2 surface. The adsorbed volumes shown by the silica samples toward CH 4 are too low for being reflected in a significant CH 4 uptake. These contrasting adsorption characteristics between CO 2 and CH 4 will indeed be very useful (i.e., selective) for achieving a fine separation of mixtures including these gases. Likewise, note the better affinity at 298 K towards CO 2 depicted by S15HNH 2 results to be 19.2% higher than the amount adsorbed by the S15NH 2 matrix at 1 bar. However, even at 298 K and 0.15 bar, there is a difference between the two samples of 36.2% (Figure 17). Even at 0.15 bar, S15HNH 2 shows a better performance at 313 K than the S15NH 2 specimen, thus stressing this way that the thermos-alkaline treatment, previous to APTES surface functionalization, is very convenient in order to significantly raise CO 2 capture capacity with respect to the amount adsorbed by the APTES functionalized but not previously hydroxylated S15NH 2 substrate. Therefore, the S15HNH 2 solid is more recommendable than the conventional (i.e., not hydroxylated) samples for pursuing enhanced CO 2 adsorption this effect being valid at 0.15 bar and temperature range of 318-348 K [48]. It is also important to mention that the CO 2 isotherms shown in Figure 17 depict low-pressure hysteresis loops at 298 K. However, this phenomenon is more accentuated for the S15HNH 2 specimen than for the S15NH 2 material. This suggests that there may exist a stronger interaction between CO 2 and the hydroxylated surface (before being functionalized with APTES) than with the analogous but unfunctionalized sample [49]. Importantly, the S15HNH 2 even if substrate presents the best performance in terms of CO 2 adsorption. The total adsorbed volume of this gas is not too high when compared to other SBA-15 materials reported in the literature. Nevertheless, this comparison become relative ever since the overall CO 2 capture capacity depends on a number of factors, among them: (1) The synthesis conditions and the method employed to eliminate the templating block copolymer to leave behind the pore entities. This process is intrinsically linked to the amount of SiOH groups on the surface [50,51]. (2) The method of amine deposition on the silica surface (in-situ or ex-situ) [52,53].
(3) The nature of the amine molecule to be surface deposited on the pore walls [54,55]. (4) The quantity of amine compound employed for the surface functionalization [35]. (5) The functionalization method (e.g., impregnation, chemical anchorage or a combination of both).
Olea et al. [56] have taken into account most of the above factors, including the exploration of different surface functionalization approaches of SBA-15 materials. (6) The experimental conditions selected for CO 2 capture (e.g., temperature, pressure, CO 2 flow rate, etc.).
Nonetheless, the present work demonstrates that the alkaline pretreatment is adequate for favorably allowing the polarity of the conventional SBA-15 surface in such a way that the substrates becomes more appropriate for being subjected to further reactions with amine species without compromising the thermal stability of the pore substrate at temperatures lower than the calcination one.
On the other hand, in the Section 3.4 the APTES quantity that was used for functionalizing of surface of the parent solids. i.e., S15 and S15H is relatively small (3SiO 2 :2APTES, e.g., for preparing 150 mg of either S15 or S15H sit is required a volume of 0.4 mL of APTES). This action is useful with the aim of elucidating the way by which the APTES molecules interact with the surface at low surface coverage. However, eventually, analogous experiments could be performed by slightly increasing the APTES:SiO 2 molar ratio as to increase the concentration of APTES molecules on the surface of silica substrates.
Finally, in order to know the forces that dominate the adsorbate-adsorbent interactions, a study of the isosteric enthalpy of adsorption was performed in relation to CO 2 adsorption at different temperatures; this study was carried out for the solids modified on the surface with APTES. The adsorption runs were made at the following temperatures: 263, 273, 288 and 298 K; the determination of the isosteric enthalpy of adsorption was calculated from a Clausius-Clapeyron type equation (which simulated the quasi-solidification of CO 2 on the silica surface) (Figure 18). The results shown with respect to the enthalpy of adsorption of CO 2 , for the APTES functionalized substrates, emphasize that in these materials prevails an adsorption potential that strengthen the adsorbate-adsorbent forces, in comparison to the process carried out on the unfunctionalized silica solid. Then, the thermo-alkaline pretreatment and its subsequent functionalization with APTES lead to a higher capacity for CO 2 adsorption in comparison to that achieved by the amine-functionalized silica precursor specimen. The absolute −∆H values result higher than the CO 2 enthalpy of sublimation (∆H sub ) at 298 K and 1 bar (i.e., 26.2 kJ·mol −1 ). These results constitute an evidence of that a larger amount of CO 2 is being anchored on the surface of S15HNH 2 solid than on that of the S15NH 2 adsorbent. Even more, the S15HNH 2 adsorbent establishes a stronger interaction with CO 2 than the unhydroxylated specimen, which is extended up to large surface coverages. For instance, for a CO 2 surface coverage of 0.72 mmol·g −1 , an enthalpy of adsorption of −23.2 kJ·mol −1 is found for the S15NH 2 specimen in contrast to the −39.2 kJ·mol −1 evolved by the S15HNH 2 adsorbent. adsorbent forces, in comparison to the process carried out on the unfunctionalized silica solid. Then, the thermo-alkaline pretreatment and its subsequent functionalization with APTES lead to a higher capacity for CO2 adsorption in comparison to that achieved by the amine-functionalized silica precursor specimen. The absolute −ΔH values result higher than the CO2 enthalpy of sublimation (ΔHsub) at 298 K and 1 bar (i.e., 26.2 kJ·mol −1 ). These results constitute an evidence of that a larger amount of CO2 is being anchored on the surface of S15HNH2 solid than on that of the S15NH2 adsorbent. Even more, the S15HNH2 adsorbent establishes a stronger interaction with CO2 than the unhydroxylated specimen, which is extended up to large surface coverages. For instance, for a CO2 surface coverage of 0.72 mmol·g −1 , an enthalpy of adsorption of −23.2 kJ·mol −1 is found for the S15NH2 specimen in contrast to the −39.2 kJ·mol −1 evolved by the S15HNH2 adsorbent. In addition, the chemical reason for this adsorption behavior can reside on the evidence provided by the NMR spectra; the ideal case would be that three Si-OH silanol groups be required to substitute the three alkoxi groups of the APTES molecule. However, this occurs in such a way that the remaining Si-OH species still be available for establishing more bondings with more alkoxi groups proceeding from other APTES molecules. That is, the attaching of one APTES molecule render new Si-OH groups for the attachment of other APTES molecules. The above behavior can be attributed to the chemical nature of the adsorbent-adsorbate pair. There are two possible chemical explanations for these phenomena to occur: (a) The basic property of amino functionalized silica acts as a Lewis base that provides an electron pair to the CO2 molecule, i.e., a coordination link is formed. The adsorbed CO2 molecule then establishes additional coordination links with the hydrogen or oxygen atoms proceeding from adjacent APTES or with the hydroxyl groups of the silica substrate. (b) When the surface population of Si-OH species is increased via the hydroxylation treatment, the APTES molecules exist over larger domains of the pore structure, therefore a larger amount of CO2 molecules gets captured. This last process could be carried out by the zwitterionic reaction that renders ammonium carbonate after the reaction of amine and CO2 species [57]. The capture of CO2 by neighboring (i.e., more concentrated) amine species results to be more stable than if only an individual, isolated amine species reacts with CO2. A proton of one of the neighboring amines is transferred to the adjacent one; the former NH2 group attracts the carbon dioxide to it due to its refurbished electron density. This is a reflection of the assertion, since therein a In addition, the chemical reason for this adsorption behavior can reside on the evidence provided by the NMR spectra; the ideal case would be that three Si-OH silanol groups be required to substitute the three alkoxi groups of the APTES molecule. However, this occurs in such a way that the remaining Si-OH species still be available for establishing more bondings with more alkoxi groups proceeding from other APTES molecules. That is, the attaching of one APTES molecule render new Si-OH groups for the attachment of other APTES molecules. The above behavior can be attributed to the chemical nature of the adsorbent-adsorbate pair. There are two possible chemical explanations for these phenomena to occur: (a) The basic property of amino functionalized silica acts as a Lewis base that provides an electron pair to the CO 2 molecule, i.e., a coordination link is formed. The adsorbed CO 2 molecule then establishes additional coordination links with the hydrogen or oxygen atoms proceeding from adjacent APTES or with the hydroxyl groups of the silica substrate. (b) When the surface population of Si-OH species is increased via the hydroxylation treatment, the APTES molecules exist over larger domains of the pore structure, therefore a larger amount of CO 2 molecules gets captured. This last process could be carried out by the zwitterionic reaction that renders ammonium carbonate after the reaction of amine and CO 2 species [57]. The capture of CO 2 by neighboring (i.e., more concentrated) amine species results to be more stable than if only an individual, isolated amine species reacts with CO 2 . A proton of one of the neighboring amines is transferred to the adjacent one; the former NH 2 group attracts the carbon dioxide to it due to its refurbished electron density. This is a reflection of the assertion, since therein a considerable enthalpy of adsorption is attained by the amine-functionalized and pre-hydroxylated SBA-15 species (Figure 18).

Conclusions
The present investigation shows that a soft alkaline hydroxylation is suitable for raising the population of Si-OH surface groups in conventional silica substrates; something that could be due to the breaking of Si-O-Si species into groups and, at the same time, involving the loss of micropores while preserving the mesoporous structure. For example, in the case of the S15 substrate, the alkaline treatment caused a pore width increment of 2.9 Å, which resulted appropriate for our purposes. In any case, these additional Si-OH groups allow the attachment on the surface of diverse species, such as the APTES which can help to fix a myriad of active chemical species over mesoporous substrates.
The studies of the amount of acid sites on the surface of the different adsorbents determined by the TPD-NH 3 desorption and 29 Si NMR confirm that the hydroxylation treatment induces the formation of a greater number of acid sites. In this way, the argument advanced from the NMR characteristics of the S15 and S15H suggests that these materials are appropriate to anchor different amounts of APTES on their surface, even if every synthesis is carried out under the same conditions and employing the same APTES: solid molar ratio. This is likely due to the abundant presence of Q 3 species mainly, i.e., free or adjacent silanol groups on which the bonding of APTES occurs and that allows the functionalization of the surface with other important chemical species. The 29 Si NMR analysis reveals, in a semiquantitative way, that 1.85 and 3.3 mmol of APTES have been anchored on the surface of the S15NH 2 and S15HNH 2 samples, respectively. Furthermore, the fixing of APTES over these substrates preferably occurs through only one chemical union. Distinctively, for the case of non-treated SBA-15 substrates, the APTES species remain fixed to the silica network through either one or two chemical bonds. These results demonstrate the effectiveness of the alkaline treatment for increasing the population of Si-OH anchoring sites without affecting the structure of mesoporous substrate.
Finally, the adsorption enthalpy of the APTES-functionalized substrates, confirms a higher CO 2 adsorption when the silica substrates are previously treated with an alkaline solution. All the evidence shown herein suggests that the interaction of CO 2 molecules with the systems treated with alkaline solutions could be due to the existence of a larger amine group population, arising from the APTES species. However, the adsorption of CO 2 could be the result of combined interactions with the hydroxyl groups of remnant and vicinal S-OH groups existing on the substrate or coming from APTES. Although the amount of CO 2 retained in the systems herein described could result to be similar or lower than those obtained in other investigations, the possibility of optimization is still attainable.