Langmuir 2003, 19, 7621-7627
7621
Acid-Functionalized Amorphous Silica by Chemical Grafting-Quantitative Oxidation of Thiol Groups E. Cano-Serrano, G. Blanco-Brieva, J. M. Campos-Martin, and J. L. G. Fierro* Instituto de Cata´ lisis y Petroleoquı´mica, CSIC, c/Marie Curie, s/n, Cantoblanco, 28049 Madrid, Spain Received March 26, 2003. In Final Form: July 4, 2003 This work describes the synthesis and structural features of functionalized amorphous silica with sulfonic acid groups. The approach followed involved the condensation of (3-mercaptopropyl)trimethoxysilane or (3-mercaptopropyl)methyldimethoxysilane and/or phenyltrimethoxysilane on the hydroxyl groups of the silica substrate followed by oxidation with hydrogen peroxide. This oxidation reaction was effective, although some lixiviation of sulfur occurred, and the amount of S released depended on the hydrophobicity of the surface and the temperature of the treatment. The extent of oxidation of the thiol precursor, determined by photoelectron spectroscopy (XPS), was moderate at ambient temperature and much more effective at 333 K. The XPS technique proved to be extremely useful not only to discriminate between the -SH and -SO3H, whose chemical shifts differ by 5 eV, but also to quantify the proportion of these species and hence the conditions that make the oxidation process quantitative. The solid samples prepared according to this grafting-oxidation methodology exhibited strong acid sites, as revealed by their performance in the esterification reaction of acetic acid by methanol in liquid phase. Activity data indicated that the reaction is accelerated in the presence of the sulfonic acid-functionalized silica catalyst, being higher than that of the reference Nafion silica composite. Finally, this grafting procedure and the methodologies to quantify and test these acid-functionalized silica materials can be applied to other substrates.
Introduction Functionalized amorphous silica surfaces are of great interest owing to their potential applications in environmental and industrial processes. This potential derives from the fact that the covalently anchored functional groups on the silica cause its surface to have specific attributes, such as a stereochemical configuration, binding sites, charge density, and acid-base or redox properties.1-5 Recent developments have been made in the mesoporous ordered silica (MOS) family incorporating organic groups using either sol-gel methodology6-8 or postgrafting techniques.3,9 This chemistry has been successfully exploited to prepare pure Brønsted sulfonic acid-functionalized mesoporous silica, i.e., MCM-41 materials10-13 or amorphous silica.14-18 * To whom correspondence may be addressed. Fax: +34 915854760. E-mail:
[email protected]. http://www.icp.csic.es/eac/ index.htm. (1) Scherbaum, K. D. Science 1994, 265, 1413. (2) Sayari, A. Chem. Mater. 1996, 8, 1840. (3) Feng, X.; Fryxell, G. E.; Wang, L.-Q.; Kim, A. Y.; Liu, J.; Kemner, K. M. Science 1997, 276, 923. (4) Wight, A. P.; Davis, M. E. Chem. Rev. 2002, 102, 3589. (5) Dijs, I. J.; van Ochten, H. L. F.; van Walree, C. A.; Geus, J. W.; Jenneskens, L. W. J. Mol. Catal. A: Chem. 2002, 188, 200. (6) Macquarrie, D. J. J. Chem. Soc., Chem. Commun. 1996, 1961. (7) Melero, J. A.; Stucky, G. D.; van Grieken, R.; Morales, G.; J. Mater. Chem. 2002, 12, 1664. (8) Shen, J. G. C., Herman, R. G., Klier, K.,J. Phys. Chem. B 2002, 106, 9975 (9) Mercier, L., Pinnavia, T., J. Chem. Mater. 2000, 74, 188. (10) VanRhijn, M.; DeVos, D. E.; Sels, B. F.; Bossaert, W. D.; Jacobs; P. A. Chem. Commun. 1998, 317. (11) Bossaert, W. D.; DeVos, D. E.; VanRhijn, W. M.; Bullen, J.; Grobet, P. J.; Jacobs, P. A. J. Catal. 1999, 182, 156. (12) Margolese, D.; Melero, J. A.; Christiansen, S. C.; Chmelka, B. F.; Stucky, G. D. Chem. Mater. 2000, 12, 2448 (13) Diaz, I.; Marquez-Alvarez, C.; Mohino, F.; Perez-Pariente, J.; Sastre, E. J. Catal. 2000, 193, 283. (14) Badley, R. D.; Ford, W. T. J. Org. Chem. 1989, 54, 5437 (15) Harmer, M. A.; Sun, Q.; Farneth, W. E. J. Am. Chem Soc. 1996, 118, 7708 (16) Harmer, M. A.; Sun, Q.; Farneth, W. E. J. Catal. 1996, 118, 62.
Within this framework, our aim in the present work followed two directions: (i) modification of an amorphous silica surface by covalent anchoring of thiol-containing molecules followed by careful oxidation under specific reaction conditions, and (ii) testing the acid function developed on the silica surface in a liquid-phase catalyzed reaction. Concerning objective (i), it should be emphasized that techniques for synthesizing sulfonic acid-functionalized silica use a thiol route, grafting, or the sol-gel route, followed by thiol group oxidation, but no attention has been paid to the oxidation step, giving incomplete oxidation,19 or disulfide species S-S.13 In fact, the oxidation of thiol groups does not quantitatively yield sulfonic groups, which is detrimental for acid-catalyzed reactions. By using a highly surface-sensitive technique, such as photoelectron spectroscopy, here we were able to quantify the percentage of oxidation of thiol to sulfonic acid groups and, as an extension, to define the precise methodology to be used for the oxidation reaction to be quantitative. As a test reaction for the acid function, we selected an esterification reaction taking place in liquid phase. Esterification of carboxylic acids with alcohols belongs to classical chemical reactions, whose kinetics and equilibria have been investigated along the history of physical chemistry. Currently, organic esters are valuable intermediates in several branches of the chemical industry. Esterification proceeds in both the absence and the presence of an added catalyst. In the absence of a catalyst, the reaction, however, is extremely slow since its rate depends on autoprotolysis of the carboxylic acid. Therefore, esterifications are carried out in the presence of an acid catalyst, which acts as a proton donor to the carboxylic acid. Typical homogeneous acid catalysts are inorganic (17) Harmer, M. A.; Sun, Q.; Michalczyk, M. J.; Yang, Z. Chem. Commun. 1997, 1803. (18) Cano-Serrano, E.; Campos-Martin, J. M.; Fierro, J. L. G. Chem. Commun. 2003, 246. (19) Wilson, K.; Lee, A. F.; Macquarrie, D. J.; Clark, J. H. Appl. Catal., A 2002, 228, 1273
10.1021/la034520u CCC: $25.00 © 2003 American Chemical Society Published on Web 08/05/2003
7622
Langmuir, Vol. 19, No. 18, 2003
Cano-Serrano et al.
Table 1. Samples Prepared, Precursor Quantities Added on 2 g of Silica, and Synthesis Conditions sample silica G-952 E1RT E1333 E2RT E2333 E3RT E3333 E4RT E4333 E0b a
MPMDMS
silane precursor (g) MPTMS
PTMS
0.68 0.68 0.68 0.68
1.32 1.32 0.68 0.68 0.68 0.68
0.68
1.32 1.32 1.32
Tox (K)
BET area (m2/g)
pore volume (mL/g)
S (wt %)
a 333 a 333 a 333 a 333 a
224 97 98 168 199 176 164 208 211 224
1.44 0.56 0.55 1.11 1.25 0.81 0.75 1.22 1.30 1.44
4.10 3.35 4.46 2.86 2.27 2.20 3.72 1.61
Room temperature. b Treated with concentrated H2SO4.
mineral acids, such as H2SO4, HCl, HI, and ClSO3OH. The disadvantage of mineral acids is their miscibility with the reaction medium, which causes separation problems and the corrosion of equipment at higher catalyst concentrations. Accordingly, heterogenized acid catalysts provide an attractive alternative to homogeneous catalysts.20 There are quite a few publications in the literature reporting polystyrene-based ion exchange fiber as an acid catalyst.21 New heterogeneous catalysts have been developed in order to surmount the separation problem. Ionexchange resin catalysts have been used in esterification reactions for some years.22,23 Typical resin catalysts are sulfonic acids bonded to a polymer carrier, such as polystyrene cross-linked with divinylbenzene (DVB). Several catalysts of this kind are commercially available, e.g., the Amberlyst family. In the use of poly(vinylbenzene)based catalysts, the process typically becomes diffusionlimited, and the polymer is swollen and deactivated in aggressive reaction media. However, resin catalysts suffer from a lack of mechanical strength, thermal stability, and low specific areas. Due to this low specific area, acidic resin catalysts are less active for fast reactions than the corresponding homogeneous catalysts because the reaction rates are limited by the transport of reactant to the active sites on the particle surface. Finally, another incentive coming from the test reaction selected for the purpose of the present work is that esterification of carboxylic acids with alcohols is very sensitive to the presence of water in the reaction medium. Since surface functionalization by an organic route may induce some hydrophobicity, it would be of interest to examine the role of water in catalytic performance in the target reaction. Experimental Section Catalyst Preparation. Catalysts were prepared using three different silane precursors ((3-mercaptopropyl)trimethoxysilane (MPTMS), (3-mercaptopropyl)methyldimethoxysilane (MPMDMS), and/or phenyltrimethoxysilane (PTMS)). The procedure was as follows: Each precursor was added dropwise onto the amorphous silica support (2 g) (Grace Davison G952) until incipient wetness. To ensure that the reaction between the precursor and the silica substrate had occurred, the impregnate was kept at room temperature for 24 h. The anchored thiol groups of the modified silica were then oxidized with a 33% H2O2 solution (32 mL). This amount represents an excess of about 100 times higher than that required for complete oxidation of thiol groups. The suspension of the solid was kept under stirring for 1 h at room temperature or 333 K. The solid was then filtered and washed three times with deionized water (32 mL). To ensure (20) Okuhara, T. Chem. Rev. 2002, 102, 3641. (21) Streat M. Ion Exchange for Industry; Ellis Horwood: Chichester, 1988. (22) Yoshioka, T.; Shimamura, M. Bull. Chem. Soc. Jpn. 1994, 57, 334. (23) Xu, Z. P.; Chuang, K. T. Can. J. Chem. Eng. 1996, 74, 493.
that all sulfonic groups had been protonated, the solid was suspended in a 10 wt % H2SO4 solution (32 mL) for 1 h. The solid was filtered, washed three times with deionized water (32 mL), and finally dried at 333 K for 24 h. The nomenclature of the catalysts synthesized is shown in Table 1. For comparison, an E1RT sample pretreated with concentrated H2SO4 (32 mL) for 1 h (catalyst labeled E0) and a Nafion SAC-13 silica composite sample supplied by Aldrich were also employed as references. Characterization of Samples. Sulfur contents were determined by inductively coupled plasma atomic emission spectroscopy (ICP-AES) (Perkin-Elmer Optima 3300 DV). The number of exchangeable protons was evaluated. The acid capacity of the sulfonic acid-modified samples was determined by means of acidbase titration, using a solution of Na+ in propan-2-ol as an ionexchange agent. Textural properties were measured by isothermal adsorption of nitrogen at 77 K with a Micromeretics TriStar 3000. Specific area was measured using the BET method over a relative pressure (P/P0) range between 0.03 and 0.3 and a cross section of adsorbed nitrogen of 0.162 nm2. Pore distributions were calculated using the BJH model to desorption branch of the isotherms. Thermogravimetric analyses of the used catalysts were performed with a Perkin-Elmer TGS 2 instrument, working at a heating rate of 10 K min-1 under an air flow (60 mL min-1). X-ray photoelectron spectra (XPS) were acquired with a VG Escalab 200R spectrometer equipped with a hemispherical electron analyzer and an Mg KR (hν ) 1253.6 eV) nonmonochromatic X-ray source. The samples were outgassed in the pretreatment chamber at room temperature for 1 h prior to being transferred into the ultrahigh vacuum analysis chamber of the instrument. The signals of silicon, oxygen, sulfur, and carbon were scanned at a pass energy of 20 eV for a number of times in order to obtain good signal-to-noise ratios. The binding energies (BEs) were referenced to the BE of C 1s core-level spectrum at 284.9 eV. Surface atomic ratios were estimated from the areas of the peaks, normalized to silicon, and corrected using the corresponding sensitivity factors.24 Catalytic Tests. The catalytic performance of the samples was evaluated in a liquid-phase glass batch reactor working at atmospheric pressure. The reaction test selected for the purpose of the present work was the esterification of acetic acid with methanol. The procedure consisted in mixing 1.5 mol of acetic acid with 1.5 mol of methanol (molar ratio 1:1) under vigorous stirring while heating to 333 K. Once this temperature remained stable, 1 g of the catalyst was added to the reaction mixture. Since the objective of this work was to compare catalyst performance of samples differing in the population of sulfonic acid groups, these reaction conditions were maintained constant. Aliquots of the liquid phase were periodically withdrawn from the reactor, and acetic acid conversion was evaluated by acid titration with a 2 M NaOH solution. Stability tests were also carried out. Basically, the procedure was as follows: 0.3 g of sulfonic acid-modified silica was suspended in 36 g of methanol and stirred magnetically for different times at 333 K. After each treatment, the solid was filtered, washed several times with methanol, and finally air-dried overnight. The chemical analysis of sulfur was used as a direct measurement of the leaching of sulfonic groups to the solution. (24) Wagner, C. D.; Davis, L. E.; Zeller, M. V.; Taylor, J. A.; Raymond, R. M.; Gale, L. H. Surf. Interface Anal. 1981, 3, 211
Acid-Funtionalized Amorphous Silica
Langmuir, Vol. 19, No. 18, 2003 7623
Results Silica functionalization was achieved by using different thiol precursors in different amounts with the simultaneous incorporation of phenyl group-containing precursors. A summary of the thiol and phenyl group-containing precursors employed in the synthesis is offered in Table 1. Also included in Table 1 are the experimental conditions used to oxidize the thiol groups into the corresponding sulfonic acid function. Chemical analysis indicated that the amount of S incorporated depended on the type of S precursor and also on the oxidation conditions employed. Moreover, S loading was always below the theoretical amount used in preparation. This observation points to a partial solubilization of surface thiol groups along the oxidation process. Comparison of samples E1 and E2 with E3 and E4 indicated that the amounts of S incorporated were higher with the MPMDMS than with the MPTMS precursor. Surprisingly, the final S content was also modified by the addition of PTMS, in such a way that the presence of phenyl groups increased the S content; this can be seen on comparing E1 vs E2 and E3 vs E4. This finding, which is clear for the samples treated at room temperature and more evident for the samples after oxidation at 333 K, means that the introduction of phenyl-containing silane may occupy some Si-OH sites and reduced the amount of -SH initially introduced, but because of the protection of hydrophobic phenyl groups, the S-containing group would possess higher stability than that without the protection of phenyl group. On the other hand, samples oxidized at 333 K showed substantially lower sulfur contents than their counterparts oxidized at room temperature. This suggests that treatment at 333 K leads to a larger desorption of S-containing species. At this temperature, the protective effect of hydrophobic groups was clearer. Although no attempt has been made to determine the chemical nature of desorbed sulfur species at 333 K. Sulfur loss could be due to leaching of silane precursors giving soluble species of the type (OH)3SiCH2CH2CH2SH and/or (OH)3SiCH2CH2CH2SO3H. In any case, if both Sx and SO2 products could be formed, SO2 appears to be the major oxidation product because no yellow film of Sx was observed on the liquid surface at the end of oxidation and in any case S-S species was not observed by FT-Raman. Nitrogen adsorption-desorption isotherms of all the samples and a silica reference were type IV of the IUPAC classification. A hysteresis loop at high relative pressure was observed. The addition of silane precursors led to a decrease in adsorption capacity (Figure 1a), and this was even more marked in samples containing phenyl groups (E1 and E3). The reduction of the adsorption capacity of functionalized silica samples is clearly illustrated by the decrease in specific area and pore volume (Table 1).). With the exception of sample E1, pore size distributions (BJH model, desorption branch) (Figure 1b) showed that samples modified using silane precursors containing phenyl groups had a slight decrease in pore size. In sample E1 there is a strong reduction in BET area with a parallel broadening of pore size distribution. In addition, no changes due to the nature of the thiol precursor used (MPMDMS or MPTMS) were observed. Thermogravimetric (TG) analyses of oxidized samples at ambient temperature and at 333 K gave virtually identical results. However, the use of different silane precursors elicited clear differences (Figure 2). At temperatures lower than 383 K, water desorption was observed, this desorption being smaller for samples
Figure 1. (a) Adsorption isotherms of nitrogen at 77 K of the samples and the G-952 starting silica. (b) Pore distribution using the BJH model to desorption branch of the 77 K nitrogen isotherm of the samples.
containing hydrophobic (phenyl and methyl) surface groups. Thus, weight loss followed an order of E4 > E2 > E3 > E1. A second weight loss, due to S-containing groups, occurred at 600-800 K, while the samples containing phenyl groups showed an additional weight loss at 800-850 K. The TG profiles for samples using MPMDMS and MPTMS precursors showed that the decomposition of sulfur groups takes place in two steps. However, for the first precursor a larger mass loss was observed between 500 and 600 K, followed by a less evident one between 600 and 800 K, whereas solids synthesized
7624
Langmuir, Vol. 19, No. 18, 2003
Cano-Serrano et al.
Figure 2. TGA measurements in air flow of samples.
with MPMTS showed a low mass loss between 575 and 625 K and a more pronounced loss between 625 and 800 K. Thus, the chemical nature of the precursor influences the thermal stability of the S-functionalized surface. High-resolution X-ray photoelectron spectra of S 2p core level for the in situ outgassed samples revealed a characteristic S 2p3/2-S 2p1/2 spin-orbital splitting (Figure 3). The conclusion about the chemical properties of the samples should be taken from the more intense component S 2p3/2. Samples showed two types of sulfur species: one at low BE (163.7 eV), corresponding to a -SH groups, and another at higher BE (168.5 eV), associated with sulfonic -SO3H groups. This energy difference makes XPS a useful tool for evaluating the degree of oxidation of thiol groups to sulfonic groups. The samples oxidized with hydrogen peroxide at ambient temperature exhibited a very small proportion of oxidized sulfur (Table 2, Figure 3a). Only sample E4RT exhibited 57% of sulfur oxidized to -SO3H. However, oxidation with hydrogen peroxide at 333 K proved to be much more effective for all samples (Table 2, Figure 3). On comparison of samples E4RT and E2RT, or E2333 and E4333, it appears that the oxidation of thiol groups is easier with the MPTMS than with the MPMDMS precursor. In addition, the ability of -SH groups to be oxidized depends on the sulfur content. Thus the samples with S contents ranging from 3.72 to 4.46% (E1RT, E2RT, and E4RT) were oxidized to some extent (17-57%), while for the E3RT counterpart with S content of 2.27% no oxidation of -SH group was discerned. A further comparison of samples E1RT and E0 pointed to a higher proportion of -SO3H groups in the latter (Table 2). In addition, sample E0 exhibited a substantially higher proportion of -SO3H groups than the other samples. This finding should be ascribed to sample pretreatment in a solution of concentrated H2SO4, which results in an additional incorporation of sulfur on the silica surface. This observation receives further support from the observation that the larger amount of sulfur incorporated ran parallel with the increase in sulfonic acid groups on the surface. To assess the amount of exchangeable protons, the acid capacity of the sulfonic-modified materials, previously oxidized at 333 K, was measured by acid-base titration,
Figure 3. Photoelectron spectra in the S 2p core level region of samples oxidized at room temperature (A) and 333 K (B).
using Na+ in propan-2-ol as an ion-exchange agent (Table 3). The exchange capacity values of the MPTMS samples were high (ca. 80% of sulfur content) and somewhat lower (ca. 40% of sulfur) for the MPMDMS counterparts. Even lower values were found for the E2333 sample, which is in agreement with the partial oxidation of thiol groups (see XPS section). The catalytic performance of the samples oxidized at 333 K was evaluated with the esterification of acetic acid
Acid-Funtionalized Amorphous Silica
Langmuir, Vol. 19, No. 18, 2003 7625
Figure 4. Esterification of acetic acid with methanol with reference sample (Nafion SAC-13) and without catalyst at 333 K, with an initial molar ratio of 1:1 between the reagents.
Figure 5. Esterification of acetic acid with methanol with samples E1333, E2333, E3333, and E4333 at 333 K, with an initial molar ratio of 1:1 between the reagents.
Table 2. Binding Energy (eV) of S 2p3/2 Levels and Atomic Ratios Measured by XPS sample silica G-952 E1RT E1333 E2RT E2333 E3RT E3333 E4RT E4333 E0
S 2p3/2
S/Si atomic ratio
163.9 (78%) 168.4 (22%) 168.5 163.8 (83%), 168.2 (17%) 163.7 (44%), 168.2 (56%) 163.7 168.4 163.8 (43%), 168.1 (57%) 168.2 163.9 (42%) 168.8 (58%)
0.04 0.03 0.04 0.03 0.04 0.04 0.03 0.02 0.08
Table 3. Exchange Capacity of Solid Oxidized at 333 K
sample
S contents (ICP-AES) (mmol g-1)
ion-exchange capacity (mmol g-1)
E1333 E2333 E3333 E4333
1.04 0.89 0.69 0.50
0.46 0.27 0.53 0.40
with methanol. All the samples showed much higher conversion levels than the blank (no catalyst) (Figures 4 and 5), pointing to the involvement of the surface acid groups produced in silica functionalization. Interestingly, catalyst E2333 exhibited a lower conversion than catalysts E1333, E3333, and E4333. In consonance with the XPS data, sample E2333 still retained 44% of sulfur atoms as thiol groups, which necessarily implies a lower density of -SO3H groups in this sample. Samples with all their sulfur atoms as sulfonic groups exhibited higher conversion levels than the reference sample: the commercial Nafion silica composite (SAC-13) purchased from Aldrich. After reaction, the E1333 and E3333 catalysts were filtered off, dried at 333 K, and reused in the tests. In this second test, only a slight decrease in conversion was detected (26 vs 29% at 180 min).
Figure 6. Stability of the sulfonic groups against leaching in methanol at 333 K.
To evaluate the loss of sulfur from the samples during the esterification reaction, stability tests were performed in methanol at 333 K. The S/Si atomic ratios of catalysts E1333, E3333 and E4333 aged at 12, 24, and 48 h are shown in Figure 6. Since the S/Si ratiossas determined by photoelectron spectroscopysremained essentially constant in all cases, it can be inferred that the samples are stable to aging in methanol. Sample E0 displayed a high conversion, although lower than that of the samples oxidized at 333 K (Figure 7). However, the activity of the aged sample E0 was almost completely lost, and its conversion level approached that of the blank experiment. This behavior is due to the fact that the sulfur incorporated during treatment with sulfuric acid remains weakly adsorbed on the surface and is released to the liquid phase along the aging experiments. The removal of S6+ species in aged samples had already been confirmed by recording the S2p core-level spectra (Figure 8), in which sulfonic acid groups disappeared while the -SH groups were maintained. This indicates that -SO3H groups chemically bonded to the surface are a
7626
Langmuir, Vol. 19, No. 18, 2003
Cano-Serrano et al.
arranged stereochemically on accessible locations of the pore network. The development of sulfonic moieties by the oxidation of anchored mercaptopropyl groups with hydrogen peroxide under mild conditions has been successfully demonstrated for -SH groups bound to alkaline-synthesized MCM-41 or neutral-synthesized HMS systems,10,11,19 silica,5,14,18 and SBA-15 mesoporous material.7,8 The oxidation process
-SH + 3H2O2 f -SO3H + 3H2O
Figure 7. Esterification of acetic acid with methanol with samples E1333 and E0, before and after aging, with an initial molar ratio of 1:1 between the reagents.
Figure 8. Photoelectron spectra in the S2p core level region of sample E0 before and after aging.
necessary condition for the catalysts to exhibit stable activity in the target reaction. Discussion The functionalization of amorphous silica with chemically anchored sulfonic acid groups of high acid strength is of considerable interest from the standpoint of application in environmental and industrial reactions. Several factors can be invoked in favor of the reliability of the sulfonic acid-functionalized silica surface for catalytic applications: (i) owing to the relative high average specific surface, the rates of the catalytic processes do not appear to be influenced to any significant extent by mass-transport phenomena; (ii) excellent mechanical strength and high specific area facilitate the withdrawal of the silica-based material from the reaction medium when experiments are conducted in liquid phase, and (iii) the availability of a moderate concentration of strong sulfonic acid groups
(1)
appears to be safe and reliable since it is conducted at temperatures between ambient temperature and 350 K. However, prolonged oxidation treatment results in samples that progressively lose the mesoscopic order. This is particularly clear, and indeed risky, when working with ordered mesoporous materials such as that described above but less marked with an amorphous silica substrate, such as the one employed in the present work. Indeed, reaction 1 is stoichiometric but a certain number of the anchored -SH groups become detached from the surface along the oxidation processes. This is illustrated in the last column of Table 1, where it may be seen that the S content drops when reaction 1 is conducted at 333 K. The low S-leaching levels for samples E3333 (3%) and E1333 (18%) with respect to their counterparts oxidized at ambient temperature, in which the PTMS silicon precursor was deliberately incorporated during the synthesis, suggests that PTMS plays a significant role in the anchoring and stability of -SH groups. The effect of PTMS on the surface distribution of thiol groups, generated by the grafting of functional MPMDMS and MPTMS organosilanes using surface hydroxyl groups as anchor points, may be understood assuming parallel reactions. If the competition of the PTMS silicon precursor for surface -OH groups is equal to that of the MPMDMS or MPTMS thiolcontaining groups, a more homogeneous distribution should be expected, because PTMS would essentially act as a spacer of the MPMDMS or MPTMS S-precursors. As a consequence, the average distance between the -SH nearest neighbors should increase. The simplistic interpretation of the spacer effect of PTMS on the S precursor is reinforced by the quantitative S/Si atomic ratios calculated by photoelectron spectroscopy (Table 2) and also by the ion exchange capacities shown in Table 3. Thus, the S/Si b atomic ratio in sample E3 is high and does not change during the oxidation process according to eq 1 at 333 K, which is consistent with the effective anchoring of -SH groups remaining uniformly distributed on the silica substrate. Further support in favor of the more uniform distribution of the anchored -SH groups on the surface and therefore of the sulfonic acid groups generated in the oxidation process according to eq 1, is provided by the ion exchange capacity of the sulfonic acid groups. The ion exchange capacity of H+ from -SO3H by Na+ is given by eq 2
-SO3H + Na+ f -SO3Na + H+
(2)
The values shown in Table 3 indicate greater ion exchange capacities in samples E1333 and E3333. In other words, the amount of exchangeable protons, and therefore the concentration of -SO3H moieties generated on the silica surface, is higher in E3333 and E1333 than in the other samples. Comparison of samples E2 and E4, in which MPMDMS or MPTMS were incorporated in the absence of the PTMS precursor, reveals that the surface concentration of -SO3H groups and their stability are rather
Acid-Funtionalized Amorphous Silica
low. In particular, if the MPTMS precursor is used, then S leaching reaches a high level. In the absence of PTMS spacer, the hydrolysis reaction appears to take place in local areas of the pores of the substrate, resulting in a weak interaction with the surface and therefore with only minor changes in its textural properties. Scrutiny of the nitrogen adsorption-desorption isotherms clearly reveals that their hysteresis loops and shape of the post-synthesis-oxidized samples virtually coincide with those of the nongrafted SiO2 reference. Notwithstanding, the BET areas and pore volumes (Table 1) decrease in the grafted samples with MPMDMS or MPTMS precursors, and substantially more so in samples E1 and E3, in which the second PTMS precursor was incorporated. This latter observation can be interpreted as being due to a carpeting of the narrower mesopores (and micropores) of the silica substrate by the silylant agent where it reacted with surface hydroxyl groups. Indeed, silylation of amorphous silica and silica xerogels or aerogels with different Si-containing organic precursors has frequently been carried out with a view to modifying their porous structure and also the hydrophobic character of the surface, resulting in tailored performances for many clean petrochemical processes. Photoelectron spectroscopy proved to be an excellent tool for monitoring not only the chemical structure of the anchored -SH structures but also those structures of their oxidation products according to eq 1. The large difference in the chemical shift of S2- and S6+ species (ca. 5 eV) allowed us to precisely determine the proportion of -SH groups and their oxidation products -SO3H ( eq 1). On examination of the S 2p profiles of samples oxidized with H2O2 at ambient temperature (Figure 3A), it is evident that the extent of oxidation of the -SH groups, according to eq 1, was only partial in E1RT, E2RT, and E4RT and absent in the E3RT sample. However, the absence of any component in the S 2p core-level spectra at binding energies of 163.5 eV in the E1RT, E3RT, and E4RT samples oxidized at 333 K with H2O2 (Figure 3B) clearly indicate that thiol species were fully oxidized to sulfonic acid groups. We would like to stress that the methodology outlined to identify the true chemical structure and abundance of the S-remaining anchored species by XPS is one of the most important issues discussed in this work Esterification of carboxylic acids with alcohols was carried out in the presence of an acid catalyst, which acts as a proton donor to the carboxylic acid and enhances the rate of reaction. For simplicity, the esterification of acetic acid with methanol was used as a test reaction of the acidic function of -SO3H groups. Activity data, expressed as acetic acid conversion, indicated that although the
Langmuir, Vol. 19, No. 18, 2003 7627
reaction occurred in the absence of catalyst (Figure 4), it was accelerated in the presence of the sulfonic acidfunctionalized silica catalyst. Conversion data at 333 K, a reaction time of 180 min, and an initial [CH3COOH]/ [CH3OH] ) 1:1 molar ratio (Figure 7) clearly indicated that samples with all sulfur atoms present as sulfonic groups exhibited higher conversion levels than the reference Nafion silica composite catalyst. It should be emphasized that the kinetic curves of sample E2333, in which photoelectron spectroscopy detected only 56% of the S-containing species in the form of -SO3H groups (Table 2 and Figure 3B), and those of the reference Nafion silica composite virtually coincided and were below those of the E1333, E3333, and E4333 samples, with 100% of S atoms oxidized to -SO3H groups. Finally, the better performance exhibited by the organically functionalized silica, developed in the present work, with respect the commercial Nafion silica composite sample together with the good stability demonstrated by the stability tests make these solid materials excellent candidates as replacements for the environmentally noxious mineral acids (H2SO4, HI, ClSO3H) usually employed as homogeneous catalysts for the esterification reaction. An additional advantage is the relative ease with which these solid catalysts can be separated from the liquid reaction medium and then reused again. Conclusions In this work, a simple and effective procedure for incorporating sulfonic groups on silica through the controlled oxidation of a thiol-functionalized surface, which quantitatively rendered sulfonic acid groups, has been developed. Along the oxidation of thiol groups, some leaching of sulfur occurs, and the amount of S released depends on the hydrophobicity of the surface and the temperature of the treatment employed. The extent of oxidation of the thiol precursor was moderate at ambient temperature and much more effective at 333 K. The chemical structure of the functionalization agent was found to influence the kinetics and the extent of oxidation of thiol groups. Finally, the solid samples prepared according to this grafting-oxidation methodology exhibited strong acid sites, as revealed by their performance in the esterification reaction of acetic acid with methanol in liquid phase. Acknowledgment. The authors acknowledge financial support from Repsol-YPF (Spain). Three of us (E.C.S., G.B.B., and J.M.C.M.) gratefully acknowledge fellowships also granted by Repsol-YPF. LA034520U
