Applied Catalysis A: General 264 (2004) 175–181
Synthesis and catalytic properties of nanostructured aluminas obtained by sol–gel method J. Sanchez-Valente a,∗ , X. Bokhimi b,1 , J.A. Toledo a a
Instituto Mexicano del Petróleo, Area de Investigacion en FCC, Edif. 24, Cub. 213, Eje Central L. Cárdenas 152, A.P. 14805, 07730 Mexico, D.F., Mexico b Institute of Physics, The National University of Mexico (UNAM), A.P. 20364, 01000 Mexico, D.F., Mexico Received 1 October 2003; received in revised form 23 December 2003; accepted 23 December 2003
Abstract Alumina was synthesized by using the sol–gel method. In order to study the aging effect gels, they were aged for 1, 14, 30, and 50 days. Samples were characterized with X-ray powder diffraction (XRD), Fourier transform infrared spectroscopy (FT-IR) with pyridine adsorption, nitrogen physisorption, and by their catalytic activity for cyclohexene isomerization. XRD of boehmite samples did not show the (0 2 0) reflection, indicating that the double layers characteristic of its structure did not correlate between each other, and the unit cell was not built. The diffraction pattern showed two new atom ordering at low angle regions corresponding to distances of 1.3 and 3 nm. When samples were transformed into ␥-alumina the atom ordering of 3 nm was maintained in all samples. In the one aged for 30 days, two additional diffraction peaks were observed, which correspond to distances between 5 and 6 nm. The specific surface area of the samples aged for 1, 14, and 30 days remained around 350 m2 /g. This area, however, was 307 m2 /g for the sample aged for 50 days. The highest catalytic activity for cyclohexene isomerization was obtained with the sample aged for 30 days, whereas the lowest activity was obtained with the sample aged for 1-day. The intrinsic activity correlated linearly with the Lewis acid site density, which at the same time correlated with the pore volume and size and the largest atom ordering distance, suggesting that the catalytic cyclohexene isomerization reaction on aluminas depends on the Lewis acid site density, which is favored with the aging time treatment. © 2004 Elsevier B.V. All rights reserved. Keywords: Aging treatment; Low angle XRD measurements; Textural properties; ␥-Alumina acidity; Cyclohexene reaction test
1. Introduction Aluminas have been extensively used as adsorbents, catalysts or supports in several petrochemical and petroleum refining processes, mainly due to their low cost, good thermal stability, high specific surface area, surface acidity, and the important interaction that they exhibit with deposited transition metals [1–3]. Aluminas are generally produced industrially by precipitation, drying, and calcination of aluminum oxy-hydroxides. The catalytic properties of aluminas largely depend on their crystalline structure and texture. Therefore, a great effort has been devoted to master these physicochemical properties [3–9]. Differences on surface properties have been ∗ Corresponding author. Tel.: +52-55-30038531; fax: +52-55-30038541. E-mail address:
[email protected] (J. Sanchez-Valente). 1 Advisor at the Instituto Mexicano del Petroleo, Mexico
0926-860X/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.apcata.2003.12.041
reported for aluminas synthesized by different methods, although they had the same crystalline structure [10], which can be partially explained by the fact that the synthesis method produces different hydration of the solid. Of course, a possible effect of impurities into alumina crystalline structure is not discarded. Recently, several research groups at the worldwide have driven their efforts to synthesize nanostructured aluminas, to control their textural properties. Hence, many strategies to obtain such characteristics have been proposed; for instance, the use of hypercritical drying conditions of xerogels [11,12] and the employment of surfactant assemblies [13–15]. In the present work, the sol–gel method was used to prepare aluminas, because it provides an attractive and convenient route; to control the structure, pore volume, and specific surface area of the solids [12,16,17]. In order to deepen on this knowledge, the present work was focused to study the structural, textural, and catalytic properties of alumina synthesized by the sol–gel method and
176
J. Sanchez-Valente et al. / Applied Catalysis A: General 264 (2004) 175–181
aged for several days at room temperature. Specific surface area, pore size, and shape were measured in the samples by nitrogen adsorption. Large angle X-ray powder diffraction (XRD), refining of the structure and quantitative analysis in the small angle region were used to characterize the samples. Since acid properties are important for the performance of catalysts they were characterized by Fourier transform infrared spectroscopy (FT-IR) of pyridine adsorption. The Catalytic properties of the aluminas were evaluated in cyclohexene isomerization reaction.
2. Experimental 2.1. Synthesis Aluminum tri-sec-butoxide (ATB) (Aldrich 97%) was dissolved and refluxed in absolute anhydrous ethyl alcohol (EtOH) (J.T. Baker) for 1 h. Hydrochloric acid, the hydrolysis catalyst, was added drop-wise into the solution while stirring and refluxing at 75 ◦ C for 3 h. Eventually, the system was cooled down to room temperature, at this point water was added letting the hydrolysis to complete and give a transparent gel. The molar ratios of reactants were EtOH:ATB = 60:1, HCl:ATB = 0.03:1, and H2 O:ATB = 1:1. The gel was placed in a glass vessel and aged for several days. After this treatment, the product was dried overnight at 100 ◦ C. The samples were labeled as Al1, Al14, Al30 or Al50; according to the number of aging days. The dried solids were calcined in air from room temperature up to 400 ◦ C at 2 ◦ C/min, and then from this temperature to 700 ◦ C at 4 ◦ C/min, where they remained for 4 h. These calcined solids were identified by adding the letter (C) to the nomenclature used for fresh samples. 2.2. Textural properties The specific surface area, pore volume, and porosity distribution of fresh and calcined samples were obtained from nitrogen adsorption and desorption isotherms, determined at liquid nitrogen temperature with an ASAP 2000 apparatus; BET equation was used to compare the surface areas. 2.3. X-ray diffraction X-ray diffraction patterns of the samples packed in an aluminum glass holder were recorded at room temperature with Cu K␣ radiation in a Bruker Advance D-8 diffractometer having theta–theta configuration and a graphite secondary-beam monochromator. The diffraction intensity of boehmite samples was measured in the 2θ range between 2 and 75◦ , with a 2θ step of 0.02◦ during 10 s per point; ␥-alumina samples were measured between 10 and 127◦ with a step of 0.1◦ for 37 s per point. Low angle measure-
ments were performed from 1 to 15◦ , with a 2θ step of 0.02 and 4 s per point, in the geometry with variable divergence (V20) and antiscattering (V20) slits, which facilitated the selection of the optimum aperture angle and radiation suppression. After each measurement of the sample a low angle, the corresponding intensity produced by the empty sample holder was measured under the same conditions, and subtracted from the one get with the sample. ␥-Alumina crystalline structure was refined with the Rietveld technique by using FULLPROF-98 code2 ; peak profiles modeled with a pseudo-Voigt function [18] contained average crystallite size as one of its characteristic parameters [19]. The standard deviations given in text and in tables, which shows the last figure variation of a number, are given in parentheses. When they correspond to Rietveld refined parameters, their values are not estimates of the probable error in the analysis as a whole, but only of the minimum possible probable errors based on their normal distribution [20]. 2.4. FT-IR pyridine adsorption Samples acidity was measured by using pyridine adsorption. The calcined samples were evacuated to 0.01 Pa; and then they were maintained under a flow of nitrogen saturated with pyridine during 15 min. Thereafter, the samples were outgassed at room temperature for 1 h, and eventually they were heated up to the desired temperature to take the FT-IR spectrum. 2.5. Cyclohexene isomerization reaction Catalytic activity of ␥-aluminas was evaluated in cyclohexene isomerization reaction. The ability of cyclohexene to play a dual role as donor or acceptor of hydride ions, provides an excellent test to evaluate the catalytic performance of several solids [22]; for instance, the selectivity and activity of the zeolites and aluminas properties on FCC catalysts has been studied via this method [21–23]. The reaction was carried out at 400 ◦ C using a conventional continuous flow system operated at atmospheric pressure in a fixed bed reactor containing 0.1 g of the calcined sample. The feeding gas supplied 0.011 mol/h of cyclohexene and 1.95 l/h of nitrogen, used as diluent. Reaction products were analyzed on line in a gas chromatograph (GC Varian Star 3400 CX) equipped with a PONA capillary column with dimension 50 m × 0.5 mm × 5 m film thickness, and a flame ionization detector (FID). The cyclohexene conversion was defined as the percentage of reactant converted to products, which were basically the three isomers of methylcyclopentenes (MCyC5 2− ).
2 J. Rodr´ıguez-Carbajal, Laboratoire Leon Brillouin (CEA-CNRS), France. Tel.:+33-1-6908-3343, fax: +33-1-6908-8261. E-mail address:
[email protected].
J. Sanchez-Valente et al. / Applied Catalysis A: General 264 (2004) 175–181
177
3. Results and discussion 500 Al30C
Specific surface area, pore diameter, and pore volume of the fresh and calcined samples are reported in Table 1. Fresh samples have higher surfaces areas than the calcined ones; this observation is in agreement with those observed in previous work, where upon calcination, the area decreased considerably [24]. This fact was attributed to the high mobility of the quasi-isolated boehmite sheets with a thickness of only half the unit cell [25]. The adsorption–desorption isotherms of the fresh and calcined samples are presented in Figs. 1 and 2. All isotherms are of type IV which characterizes the mesoporous solids. The hysteresis loop for fresh samples is of H2 type [26] which is in agreement with the randomly folded boehmite sheets of solids [24]. When the samples were calcined, an important change was observed on the hysteresis loop, the H2 loop type become into H1. This change, already reported by us, was attributed to a pore’s shape and ordering change during the calcination [24]. Table 1 Specific surface area, average pore diameter, and total pore volume of fresh and calcined samplesa Sample
BET surface area (m2 /g)
Average pore diameter (nm)
Total pore volume (cm3 /g)
Al1F Al14F Al30F Al1C Al14C Al30C Al50C
528 515 665 345 346 358 307
3.4 4.0 4.3 6.1 6.3 8.1 6.0
0.69 0.79 1.12 0.82 0.95 1.00 0.69
a
Samples were calcined at 700 ◦ C for 4 h.
700 Al30F
630 Vol. adsorbed (cm3/g STP)
560 490
Al14F
420 350 Al1F
280
Vol. Adsorbed (cm3/g STP)
3.1. Textural properties 400 Al114C Al50C
300
200
Al1C
100
0 0.0
0.1
0.2
0.3
0.4 0.5 0.6 0.7 0.8 Relative pressure (P/Po)
0.9
1.0
Fig. 2. Adsorption–desorption isotherms of calcined aged sol–gel boehmite (␥-alumina): (䉱) Al1C, (䊐) Al114C, (䊏) Al30C, and (䊊) Al50C.
From Table 1 it is noticed that the values of textural parameters of all samples were improved as the aging time increase, reaching the optimum values on sample aged for 30 days, Al30. This behavior agrees with the hysteresis evolution observed as a function of aging time. Hence, an ordering of boehmite sheets during aging has taken place, which points out to an unfinished stacking process of boehmite phase. Furthermore, the volume of the hysteresis loop rises and become narrow when the aging time increases, Figs. 1 and 2, which is reflected on the total pore volume (Vp ) and average pore diameter. When the aging time surpasses the 30 days, the textural properties values decrease significantly; for instance, the values reported for Al50C in Table 1 are lower than those determined for Al1C. One suitable explanation to this could be that the ordering process reaches a maximum, after which a collapse on the structure occurs. It is worth to remark that unexpected values of total pore volume were reached by this way, the Al30F presents a Vp = 1.12 cm3 /g (Table 1); these high values were preserved even after calcination at 700 ◦ C (see Vp for Al1, Al14, and Al30 fresh and calcined). It was observed that the textural parameters, which are very important in catalysis, could be regulated easily by choosing the adequate conditions of synthesis and post-treatments of the solids.
210
3.2. Crystalline structure
140 70 0 0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Relative pressure (P/Po)
Fig. 1. Adsorption–desorption isotherms of fresh sol–gel boehmite after aging the gels for several days: (䊏) Al1F (1-day), (䉱) Al14F (14 days), and (䊉) Al30F (30 days).
The fresh sol–gel boehmite corresponding to the sample aged for 1-day, had a diffraction pattern similar to the one of crystalline boehmite (Fig. 3), but it did not had the (0 2 0) reflection. This indicates that the double layers of boehmite crystalline structure [27] were not correlated between each other and could not build boehmite’s unit cell, this could be produced by the presence of OR groups on
178
J. Sanchez-Valente et al. / Applied Catalysis A: General 264 (2004) 175–181
Sol-Gel Boehmite
Sol-Gel Boehmite
Intensity (a. u.)
50 days
50 days
Intensity (a. u.)
30 days
14 days
14 days
(200)
(132)
(002)
(150)
(130)
(020)
(021)
1 day
30 days
1 day
10
20
30
40
50
60
70
Two Theta (degree)
the boehmite double layers, which is caused by the reaction stoichiometry of alkoxide/water, giving a boehmite phase partly alkoxylated. The presence of the other reflections of boehmite in the diffraction pattern, however, indicates that the single boehmite layer built of coupled octahedra made with oxygen, aluminum, and hydroxyls [27] were yet formed. In all samples, the (0 2 1) diffraction peak, produced by the atom planes coupled via the hydrogen bonding [25], shifted to lower diffraction angles in comparison with crystalline boehmite (Fig. 3); this means that the hydrogen bonding in sol–gel boehmite was weaker than in the crystalline one. The aging time effect on this bonding was to homogenize it along the sample, producing a more symmetric (0 2 1) peak as this time was increased (Fig. 3). Fresh and aged samples had an atom ordering at long distances, not reported previously, for boehmite (Figs. 3 and 4). This ordering gave rise to two diffraction peaks at 2θ angles lower than 10◦ . In the sample aged for 1-day the stronger peak corresponded to a distance of 1.35 nm, which shifted to 1.12 nm and had a lower intensity when the aging time was increased. The second peak was very broad and not well-defined; in average, it corresponded to a distances centered at 3 nm. The integrated intensity of both peaks was similar, which means that the corresponding atom distances were equally frequently. Because the transformation of boehmite into ␥-alumina is pseudomorphic, the alumina porosity is determined by the corresponding one of boehmite. Therefore, we measured in detail the low diffraction angles of the ␥-alumina samples obtained after calcining the boehmite at 700 ◦ C (Fig. 5). For doing that, it was necessary to operate the diffractometer with variable divergence and antiscattering slits, and to mea-
1
2
3
4
5
6
7
d(nm) Fig. 4. Low angle X-ray powder diffraction pattern of the samples aged for several days. Because of its physical interpretation, intensity is plotted as a function of atom distance and not as a function of 2θ. Crosses correspond to experimental data, and the continuous line to a smoothing of the data with a Fast Fourier Transform filter.
sure the background produced by the empty sample holder to be subtracted from the diffraction pattern of the sample. The diffraction pattern at low angles provides information about the correlation at long distances. This pattern for the samples aged for 1, 14, and 50 days, had a peak with
Sol-Gel γ-Alumina 50 days
Intensity (a. u.)
Fig. 3. High angle X-ray powder diffraction pattern of the samples aged for several days. As reference, the pattern of a crystalline boehmite is included.
30 days
14 days
1 day
1
2
3
4
5
6
7
d(nm) Fig. 5. Low angle X-ray powder diffraction pattern of the calcined boehmite aged for several days. Intensity is plotted as a function of distance between atoms and not as a function of 2θ. Crosses correspond to experimental data, and the continuous to a smoothing of the data with a Fast Fourier Transform filter.
J. Sanchez-Valente et al. / Applied Catalysis A: General 264 (2004) 175–181
179
Sol-Gel γ-Alumina
Intensity (a. u.)
50 days
30 days
14 days
1 day
20
40
60
80
100
120
Two Theta (degree) Fig. 6. High-angle X-ray powder diffraction pattern of the calcined boehmite aged for several days. Tick marks represent the reflections associated to the monoclinic crystalline structure.
its maximum at 3 nm and an intensity that decreased notoriously at higher distances (Fig. 5). In contrast to that, the diffraction pattern of the sample aged for 30 days had also the above peak, but overlapped to other peaks of similar intensity located at distances between 5 and 6 nm, which give rise to an apparent plateau in the diffraction pattern at these distances. At high angles the diffraction pattern for all the calcined samples was essentially equal (Fig. 6). In order to obtain information about the crystallite size of these aluminas, the crystalline structure was modeled with a monoclinic unit cell reported for -alumina [28], which is a good approximation because the crystallite size was smaller than 2 nm (Table 2) and consequently, the diffraction peaks were very broad. 3.3. FT-IR pyridine adsorption The infrared spectra of adsorbed pyridine, performed on all samples, showed only characteristic absorption bands of pyridine adsorbed on Lewis acid sites [29] which is in agreement with those results found in previous work [24]. Fig. 7 summarizes the results of all samples, where the pyridine’s desorption temperature was increased from 100 to 500 ◦ C; Table 2 ␥-Alumina average crystallite size as a function of aging time Aging time (days)
d (nm)
3 14 30 50
1.46(4) 1.32(2) 1.39(4) 1.43(3)
Fig. 7. Lewis acidity density determined from FT-IR pyridine adsorption on ␥-aluminas at several desorption temperatures.
only the acid sites that retained pyridine above 100 ◦ C were considered because we assumed that they were chemisorbed. The amount of pyridine adsorbed decreases as the temperature rises (Fig. 7). It is well-known that the strength of the acid sites is related to the maximum temperature at which the pyridine is retained. Thus, we observe that the aging treatment produced stronger acid sites. Since Al30C retain pyridine adsorbed even at 400 ◦ C it is the more acidic solid of this study (Fig. 7). This result agrees with those observed in the textural properties section, where the Al30 sample shows the best values on textural parameters of the group. Table 4 reports the total intrinsic acidity (TIA) which is the total acidity referred to the specific surface area of a solid. The largest amounts of acid sites were obtained on Al1C and Al30C samples. However, even if the TIA values are very close for these solids, around 3 mol pyridine/m2 , their strength distribution are not the same, Al30C shows larger amounts of stronger acid sites than the Al1C (Fig. 7). Hence, a clear enhancement on Lewis acid sites is observed when the aging treatment is done. A maximum is reached in the sample aged for 30 days. The best acidic properties determined on Al30C sample could not be only related to its pore volume and pore diameter (Table 1), because the porosity measured on all calcined solids allows the accessibility of the pyridine molecule to the acid sites. Another possible explanation may be a generation of new type of acid sites during the aging process. 3.4. Catalytic test In Table 3 are reported the conversion and selectivity of ␥-aluminas in cyclohexene isomerization reaction. Although small differences in BET surface areas were observed in aluminas aged for 1, 14, and 30 days (Table 1), quite differences were observed in cyclohexene conversion. The conversion increases as the aging time increases, being the most active
180
J. Sanchez-Valente et al. / Applied Catalysis A: General 264 (2004) 175–181
Table 3 Results of catalytic activity of alumina in cyclohexene isomerization reaction Sample
Al1C Al14C Al30C Al50C
Conversion (mol %)
Selectivity (mol %) MCyC5 2−
MCyC5
CyC6
2.0 4.2 28.5 7.8
100 100 99.3 98.2
0.0 0.0 0.7 1.0
0.0 0.0 0.0 0.8
Table 4 Intrinsic activity and total acidity on ␥-aluminas Sample
Intrinsic activitya (mol/m2 h) ×105
Total intrinsic aciditya (mol Pyridine/m2 )
Al1C Al14C Al30C Al50C
0.64 1.32 8.76 2.81
3.046 1.728 3.045 2.100
a
Total intrinsic acidity considering values from 50 ◦ C.
sample the alumina for 30 days, while the lowest activity was showed for that aged 1-day, even thought both samples had similar specific surface areas (Tables 1 and 3). A best way to compare the catalytic activity of different solids is by using its intrinsic activity, which is the activity by surface area instead of the weight of a given sample. In Table 4 are presented the intrinsic activities of all aluminas. It is worth to notice that the alumina aged for 30 days showed a higher activity than those aged for 1, 14, and 50 days. Furthermore, the Al50C, with a surface area of 307 m2 /g, showed a higher activity than the one determined on Al1C and Al14C samples, with specific surface areas of 345 and 346 m2 /g, respectively, indicating that the catalytic activity not only depends on the surface exposed to the reactants. As a matter of fact, the intrinsic activity correlated linearly with the Lewis acid site density (Fig. 8), suggesting that Lewis acid sites in the samples are responsible of the activity
2
Intrinsic Activity (mol/m *s)x10
3
10 9 Al30C
8 7 6 5 4 3 2
Al50C
Al14C
1 0 1.6
Al1C 1.8
2.0
2.2
2.4
2.6
2.8
3.0
3.2
2
Lewis Acidity (µmol Py/m )
Fig. 8. Linear correlation between intrinsic activity and Lewis acid site density on ␥-aluminas, obtained from boehmite precursors.
observed on cyclohexene reaction, which agrees with those findings in others systems, where the activity was related to the aluminum amount [22,23]. The reaction products obtained indicate that mainly isomerization of cyclohexene to methylcyclopentenes occurs during the reaction (Table 3). However, aluminas aged for 30 and 50 days produced a small amount of the saturated products, methylcyclopentane (MCyC5 ) and cyclohexane (CyC6 ), which is explained through a hydrogen transfer mechanism [22,30]. It has been demonstrated that cyclohexene isomerization to methylcyclopentenes occurs basically over Lewis acid sites [21,22,30], this result agrees with the Lewis acidity measured in each of samples and could explain the linearity observed in Fig. 8, between the intrinsic activity and the total Lewis acidity. Therefore, it can be concluded that the aging treatment given to the alumina precursor could control the type of acid sites. The amount and the type of acid sites will depend on the demand of a particular application.
4. Conclusions Aging treatment of sol–gel boehmite affects greatly its textural properties. An ordering of boehmite sheets seems to have place during the treatment, pointing out that the stacking process continues, which is confirmed by the changes observed on the hysteresis loop as a function of aging time. Hence, important textural parameters, as pore size distribution and total pore volume, can be easily tuned. High surface area and pore volume values were obtained even after calcining the sample at 700 ◦ C; the optimum was reached for the sample aged for 30 days with a surface area of 358 m2 /g and a Vp=1 cm3 /g. Similar findings were obtained from XRD analysis, where the aging time effect is reflected on the hydrogen bonding homogenization along the sample, and the better ordering was observed on the sample aged for 30 days. Only Lewis acid sites were observed by FT-IR of pyridine adsorption on ␥-aluminas. The strength and the amount of acid sites depended on aging time. The amount of pyridine adsorbed decreases as the temperature rises. The sample aged for 30 days, Al30C, retain pyridine adsorbed even at 400 ◦ C; therefore, it is the more acidic solid of the group. However, its best acidic properties could not only be related to its better textural parameters, but also to the generation of a new type of acid sites during the aging process. In fact, even if in the cyclohexene test the intrinsic activity correlated linearly with the Lewis acid site density of calcined solids, the selectivity is not the same for the samples aged for 30 and 50 days, since between the products appear, in small amounts, saturated compounds. These compounds are explained through a hydrogen transfer mechanism, which may be due to a different acid site. Thus, aging treatment could also control the type of acid sites.
J. Sanchez-Valente et al. / Applied Catalysis A: General 264 (2004) 175–181
In conclusion, important insights on how control textural and acidic properties of aluminas, are disclosed from this study.
Acknowledgements This work was financially supported by IMP-Maya Crude Oil Research Program, Project D.00187. We would like to thank Eng. Manuel Aguilar for his technical assistance on XRD measurements.
References [1] T. Kotanigawa, M. Yamamoto, M. Utiyama, H. Hattori, K. Tanabe, Appl. Catal. 1 (1981) 185. [2] K. Hellgardt, D. Chadwick, Ind. Eng. Chem. Res. 37 (1998) 405. [3] H. Knözinger, P. Ratnasamy, Catal. Rev. Sci. Eng. 17 (1978) 31. [4] B. C. Lippens, Thesis Structure and Texture of Aluminas, Uitgeverij Waltman -Delft, 1961. [5] T. Ono, Y. Ohguchi, O. Togari, in: G. Poncelet, P. Grange, P. A. Jacobs (Eds.), Preparation of Catalysts III, Elsevier, 1983, p. 631. [6] D.L. Trimm, A. Stanislaus, Appl. Catal. 21 (1986) 215. [7] D. Mishra, S. Anand, R.K. Panda, R.P. Das, Mater. Lett. 42 (2000) 38. [8] T. Tsuchida, J. Eur. Ceram. Soc. 20 (2000) 1759.
181
[9] F. Vaudry, S. Khodabandeh, M.E. Davis, Chem. Mater. 8 (7) (1996) 1451. [10] K. Jiratova, L. Beranek, Appl. Catal. 2 (1982) 125. [11] G.M. Pajonk, Catal. Today 35 (1997) 319. [12] M. Schneider, A. Baiker, Catal. Rev. Sci. Eng. 37 (4) (1995) 515. [13] M. Yada, H. Hiyoshi, K. Ohe, M. Machida, T. Kijima, Inorg. Chem. 36 (1997) 5565. [14] P. Yang, D. Zhao, D.I. Margolese, B.F. Chmelka, G.D. Stucky, Chem. Mater. 11 (1999) 2813. [15] Z. Zhang, T.J. Pinnavaia, J. Am. Chem. Soc. 124 (2002) 12294. [16] J. Livage, Catal. Today 41 (1998) 3. [17] B.E. Yoldas, J. Mater. Sci. 10 (1975) 1856. [18] P. Thompson, D.E. Cox, J.B. Hasting, J. Appl. Crystallogr. 20 (1987) 79. [19] R.A. Young, P. Desai, Arch. Nauki Mater. 10 (1989) 71. [20] E.J. Prince, Appl. Crystallogr. 14 (1981) 157. [21] U. Sedran, Catal. Rev. Sci. Eng. 36 (3) (1994) 405. [22] G. de la Puente, U. Sedran, Chem. Eng. Sci. 55 (2000) 759. [23] G. de la Puente, E.F. Sousa-Aguiar, A. Figueiredo-Costa, U. Sedran, Appl. Catal. A. 242 (2003) 381. [24] J. Sánchez-Valente, X. Bokhimi, F. Hernández, Langmuir 19 (2003) 3583. [25] X. Bokhimi, J. Sánchez-Valente, J. Solid State Chem. 166 (2002) 182. [26] K.S.W. Sing, et al., Pure Appl. Chem. 57 (1985) 603. [27] X. Bokhimi, J.A. Toledo-Antonio, M.L. Guzmán, B. Mar, F. Hernández, J. Navarrete, J. Solid State Chem. 161 (2001) 319. [28] E. Husson, Y. Repelin, J. Solid State Inorg. Chem. 33 (1996) 1223. [29] E.P. Parry, J. Catal. 2 (1963) 371. [30] A. Corma, J. Mart´ınez-Triguero, J. Catal. 165 (1997) 102.
