Shape selectivity of MWW-type aluminosilicate zeolites in the alkylation of toluene with methanol

Shape selectivity of MWW-type aluminosilicate zeolites in the alkylation of toluene with methanol

Applied Catalysis A: General 318 (2007) 22–27 www.elsevier.com/locate/apcata Shape selectivity of MWW-type aluminosilicate zeolites in the alkylation...

424KB Sizes 1 Downloads 33 Views

Applied Catalysis A: General 318 (2007) 22–27 www.elsevier.com/locate/apcata

Shape selectivity of MWW-type aluminosilicate zeolites in the alkylation of toluene with methanol Satoshi Inagaki a,*, Kohei Kamino b, Eiichi Kikuchi b, Masahiko Matsukata b,c a

Chemical Resources Laboratory, Tokyo Institute of Technology, 4259, Nagatsuta, Midori-ku, Yokohama, Kanagawa 226-8503, Japan b Department of Applied Chemistry, Waseda University, 3-4-1, Okubo, Shinjuku-ku, Tokyo 169-8555, Japan c Advanced Research Institute of Science and Engineering, Waseda University, 3-4-1, Okubo, Shinjuku-ku, Tokyo 169-8555, Japan Received 3 June 2006; received in revised form 16 October 2006; accepted 16 October 2006 Available online 21 November 2006

Abstract We investigated the alkylation of toluene with methanol at 250 8C over MCM-22 and ITQ-2 zeolites possessing an MWW structure. Whereas ITQ-2, produced by the delamination of MCM-22 precursor, showed a poor p-xylene selectivity of 23% at 1.7% of the level of toluene conversion, a high p-xylene selectivity of 80% at 0.6% of the level of toluene conversion was obtained by poisoning the external surface of ITQ-2 with collidine. Similarly, MCM-22 treated with collidine showed a high p-xylene selectivity of 74% at 3.8% of the level of toluene conversion. We concluded that both interlayer and intralayer 10-membered ring (10MR) micropores in the MWW structure caused the shape-selective formation of p-xylene in the alkylation of toluene with methanol. # 2006 Elsevier B.V. All rights reserved. Keywords: MCM-22 zeolite; ITQ-2 zeolite; Alkylation of toluene with methanol; Shape selectivity; Collidine

1. Introduction MCM-22 zeolite, which is designed as an MWW structure [1] and has the same topology as PSH-3 [2], SSZ-25 [3], ERB-1 [4], and ITQ-1 [5] can be obtained by the hydrothermal synthesis using hexamethyleneimine (HMI) as structuredirecting agent and further calcination. This type of zeolite has recently been applied to commercial processes for ethylbenzene and cumene production by the alkylation of benzene with ethylene and propylene [6,7], respectively. The microporous structure of MCM-22 has been investigated by various techniques, including X-ray diffraction (XRD) [8] and catalytic reactions [9,10]. Two types of independent 10membered ring (10MR) pore systems have been found; one is constituted by two-dimensional sinusoidal channels, the other by micropores possessing 12MR supercages with dimensions of 0.71 nm  0.71 nm  1.81 nm [9]. The two-dimensional sinusoidal 10MR channels, which exist in the MWW sheets, have been designated ‘‘interlayer micropores’’. The 10MR

* Corresponding author. Tel.: +81 45 924 5265; fax: +81 45 924 5282. E-mail address: [email protected] (S. Inagaki). 0926-860X/$ – see front matter # 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.apcata.2006.10.036

micropores with 12MR supercages, which form between stacked MWW sheets through the calcination of MCM-22 precursor (which causes the removal of HMI molecules in the MCM-22 precursor and dehydration condensation between facing silanols on the sheets), have been designated ‘‘intralayer micropores’’. ITQ-2 zeolite [11], prepared by swelling and exfoliating MCM-22 precursor, possesses an extremely large external surface area. Argon adsorption and high-resolution transmission electron microscopy (HR-TEM) experiments have confirmed that ITQ-2 is an exfoliating structure with external 12MR cups [12]. The extremely large external surface area of ITQ-2, as compared with MCM-22, results in high catalytic activities for the reactions of larger molecules inaccessible to the 10MR micropore system [13,14]. In the case of unimolecular reactions, Laforge et al. [15] have recently reported that both protonic sites on the external surface and the two types of 10MR micropores in the MWW structure contribute to the isomerization of m-xylene at 350 8C. Roles for acid sites on the external surface of MCM-22, and for intralayer and interlayer micropores, have been characterized in the case of bimolecular reactions such as alkylation of aromatics with alkenes or alcohols. Corma et al. have reported that the

S. Inagaki et al. / Applied Catalysis A: General 318 (2007) 22–27

alkylation of benzene with ethylene in liquid phase at 160 8C predominantly occurs at the acid sites on the external surface of H+-MCM-22 catalyst [16], but that the high selectivity to pxylene in the disproportionation of toluene at 300 8C results from the occurrence of the reaction in the interlayer 10MR micropores with 12MR supercages [17]. Wu et al. [17] have suggested that the disproportionation of toluene cannot occur in the intralayer 10MR micropores, wherein an intermediate consisting of two toluene molecules cannot easily be formed. In this study, the alkylation of toluene with methanol in vapor phase was investigated at 250 8C over MCM-22 and ITQ-2. We will discuss the contribution of each type of 10MR micropore to the catalytic activities and selectivities of MWW-type zeolites. 2. Experimental 2.1. Preparation of MWW-type zeolite catalysts MCM-22 precursor having an MWW-type framework was hydrothermally synthesized following a previous report [13]. A parent mixture of fumed silica (99.8%; Aldrich), NaAlO2 (Al2O3, 36.5 wt%; Na2O, 33.0 wt%; Kanto Chem.), NaOH pellets (96.0% purity; Kokusan Chem.), HMI (>97.0% purity; Kanto Chem.), and distilled water was prepared and stirred vigorously for 30 min at room temperature. The composition of the mixture was SiO2:Na2O:Al2O3:H2O:HMI = 1.0:0.075: 0.028:44:0.50. Crystallization was carried out in a 100 cm3 autoclave at 150 8C for 7 days under rotating conditions (rotating speed = 20 rpm). Since the formation of MCM-22 zeolite requires the calcination of as-synthesized products, MCM-22(P), the precursor was calcined in airflow at 540 8C for 12 h (heating rate = 5 8C min1). ITQ-2, which is obtained by the delamination of MCM-22 precursor, was prepared according to the procedure described in Corma et al. [11] as follows. First, 16.9 g of hexadecyltrimethylammonium bromide (C16TMABr; Merck) was dissolved in 74.5 g of an aqueous solution containing 10 wt% of tetrapropylammonium hydroxide (TPAOH; Acros), and then 3.0 g of MCM-22(P) was suspended in the solution. The suspension was refluxed for 18 h at 80 8C, and then treated in an ultrasound bath for 1 h. After a few drops of concentrated hydrochloric acid were added to the resulting slurry in order to bring the pH value slightly below 2, solid product was collected by centrifugation. The organic materials in the solid were removed by calcination in airflow at 540 8C for 12 h (heating rate = 5 8C min1), yielding ITQ-2. The calcined product was treated in a 1.0 N aqueous NH4NO3 solution at 80 8C for 1 h with stirring and then filtered. This ionexchange procedure was repeated four times. NH4+-exchanged zeolites were calcined in a furnace at 540 8C for 12 h (heating rate = 5 8C min1), and then H+-formed zeolites were obtained. 2.2. Characterization The crystallinity and phase purity of all as-synthesized and calcined products were analyzed by XRD on an RINT 2100 (Rigaku) using Cu Ka radiation at 40 kV and 20 mA.

23

The textural properties of the MWW-type zeolites obtained were determined using N2 adsorption at 196 8C on an AUTOSORB-1 (Quantachrome Inst.). The sample was evacuated at 350 8C for 3 h prior to the measurement. The external surface area and micropore volume were determined from the adsorption branch using the t-plot method [18]. The number of acid sites was measured using temperatureprogrammed desorption (TPD) of ammonia (AT-1; Bel-Japan, Inc.). The sample was evacuated at 500 8C prior to measurement. The heating rate during the collection of TPD data was 1 8C min1. Bro¨nsted and Lewis acid sites on the MWW-type zeolites were characterized by pyridine adsorption using Fourier transform infrared spectroscopy (FT-IR, FT-6100; JASCO). Self-supporting pressed wafers (diameter: 2 cm, weight: ca. 20 mg) were prepared, then activated by heating under dynamic vacuum for 1 h at 500 8C in an IR cell that allowed in situ thermal treatments and low temperature gas dosage. After cooling the IR cell to 150 8C, introducing pyridine vapor (ca. 2.6 kPa), and evacuating the gases, we recorded the IR spectra of the wafers on an FT-IR with a resolution of 4 cm1. Nuclear magnetic resonance (NMR) measurements were carried out using CMX400 (JEOL) at a magnification field of 9.4 T. 27Al magic angle spinning (MAS) NMR spectra for the products was recorded at 104.17 MHz with a pulse width of 3.0 ms, a pulse interval of 5.0 s, and a spinning rate of about 5 kHz. 2.3. Determination of the number of acid sites on the external surface of the zeolites The number of Bro¨nsted acid sites on the external surface of the MWW-type zeolites investigated was determined by the adsorption of 2,4,6-trimethylpyridine (collidine) at 250 8C. The cross section of the collidine is 0.62 nm  0.56 nm [19], larger than the 10MR micropores of the MWW structure. Moreover, bulky bases such as collidine, 2,4- and 2,6-diemthylquinoline show no strong interaction with Lewis acid sites of zeolites due to steric constraints [19,20]. A pulse method was employed to poison the zeolite catalyst with collidine in this experiment. A catalyst weighing 20 mg was packed in the fixed bed of a stainless-tube microreactor (inner diameter: 4 mm) and heated at 400 8C for 1 h in a stream of helium (30 cm3 min1). After the temperature was decreased to 250 8C, 0.2 ml of collidine was pulsed into the catalyst-bed with a stream of helium (30 cm3 min1). Surplus collidine was detected using a gas chromatograph (GC-8A; Shimadzu) equipped with a thermal conductivity detector. 2.4. Alkylation of toluene with methanol The alkylation of toluene with methanol was performed at atmospheric pressure in a quartz-tube microreactor with an 8 mm inner diameter. Before the reaction, 10 or 20 mg of catalyst was packed in a fixed bed of the reactor with 2.0 g of silicon nitride particles as a diluent and heated at 350 8C for 1 h in a stream of argon. After the temperature was decreased to

24

S. Inagaki et al. / Applied Catalysis A: General 318 (2007) 22–27

250 8C, an equimolar mixture of toluene and methanol was fed with argon. The partial pressures of both toluene and methanol were 9 kPa. In some cases, the acid sites on the external surfaces of the zeolite catalysts were poisoned by introducing collidine into the reactor in a pulse-wise manner before the reactant mixture was fed. The catalysts poisoned with collidine were designated P-MCM-22 and P-ITQ-2. The argon stream contained methane (5%) as an internal standard material. A packing column with Bentone 34 (5%) + DIDP (5%) (GL Science) was used to separate the reaction products. The reactants and products were analyzed using a gas chromatograph (GC-8A; Shimadzu) equipped with a flame ionization detector. The levels of conversion of toluene and methanol were calculated by the following equations:   F Tol;out conversion of toluene ð%Þ ¼ 1   100; (1) F Tol;in and  conversion of methanol ð%Þ ¼

1

F MeOH;out F MeOH;in

  100; (2)

where F Tol,in is the inlet flow of toluene (mol min1), F Tol,out is the outlet flow of toluene (mol min1), F MeOH,in is the inlet flow of methanol (mol min1), and F MeOH,out is the outlet flow of methanol (mol min1). The yield of dimethyl ether (DME) was determined by the following equation: yield of DME ð%Þ ¼

F DME;out  2  100; F MeOH;in

(3)

Fig. 1. Shape selectivity of MWW-type aluminosilicate zeolites in the alkylation of toluene with methanol. XRD patterns for (a) product crystallized at 150 8C for 7 days, (b) calcined MCM-22 zeolite, and (c) calcined ITQ-2 zeolite.

where FAromatic,out is the outlet flow of each aromatic compound (mol min1), including isomers of xylenes and trimethylbenzenes. 3. Results and discussion

where F DME,out is the outlet flow of DME (mol min1). The product distribution in the alkylated products was determined based on the number of benzene-rings as follows:

3.1. Physicochemical properties of H+-MCM-22 and H+ITQ-2 zeolites

product distribution of each aromatic compound ð%Þ

Fig. 1 shows the XRD patterns for as-made products crystallized at 150 8C for 7 days, as well as for calcined and exfoliated products. MCM-22(P) was formed after 7 days of crystallization, as shown in Fig. 1a. Fig. 1b shows the formation

¼

F Aromatic;out  100; F Tol;in

(4)

Fig. 2. Nitrogen adsorption–desorption isotherms at 196 8C for (a) MCM-22 and (b) ITQ-2 zeolites. Open symbols, adsorption branch; closed symbols, desorption branch.

S. Inagaki et al. / Applied Catalysis A: General 318 (2007) 22–27

25

of MCM-22 zeolite by calcination. As shown in Fig. 1c, calcined ITQ-2 produced weakened and broadened XRD reflection peaks compared with those of MCM-22(P), suggesting that the swelling and exfoliating treatments as well as the successive calcination of the MCM-22 precursor yielded ITQ-2, in agreement with the previous report [11]. Table 1 lists the micropore volumes and external surface areas for MCM-22 and ITQ-2. The remarkable decrease in micropore volume from 0.160 cm3 g1 for MCM-22 to 0.047 cm3 g1 for ITQ-2 was caused by the disappearance of the intralayer micropores with 12MR supercages due to infinite expansion of the intralayers. Similarly, since the intralayer micropores of MCM-22 became part of the external surfaces of the delaminated products, ITQ-2 showed a larger external surface area (592 m2 g1) than that of MCM-22 (110 m2 g1) (see Fig. 2). Table 1 also lists the total numbers of acid sites and the numbers of external acid sites on H+-MCM-22 and H+-ITQ-2. The increase in the number of acid sites on the external surface from 0.14 mmol g1 for MCM-22 to 0.28 mmol g1 for ITQ-2 resulted from the increase in external surface area. On the other hand, the total number of acid sites of ITQ-2, 0.53 mmol g1, was much less than that of MCM-22, 1.13 mmol g1. We deduce that aluminum species extracted from the MWW structure in the course of the delamination treatment were dissolved into the acidic solution after the drops of concentrated hydrochloric acid were added. The acid properties of MCM-22 and ITQ-2 were examined by the FT-IR spectra using the pyridine-adsorption technique, as shown in Fig. 3. The relative intensities of the two peaks at 1540 and 1450 cm1, corresponding to pyridine adsorbed on Bro¨nsted and Lewis acid sites, respectively, were different for ITQ-2 and MCM-22; the relative number of Bro¨nsted acid sites on ITQ-2 was smaller than the number on MCM-22. It is well known that Lewis acid sites on zeolite catalysts often originate from hexa-coordinated aluminum species, which give a peak appearing at 0 ppm in the 27Al MAS NMR spectra of zeolites. As shown in Fig. 4b and c, the 27 Al MAS NMR spectra for both MCM-22 and ITQ-2 showed a peak appearing at 0 ppm, corresponding to the hexa-coordinated aluminum; dealumination therefore occurred in the course of the transformation of MCM-22 precursor into MCM-22 or ITQ-2. 3.2. p-Xylene selectivity in the alkylation of toluene with methanol on MWW-type zeolites Four kinds of MWW-type zeolite catalysts, MCM-22, PMCM-22, ITQ-2, and P-ITQ-2, were used for the alkylation of Table 1 Textural and acid properties of MCM-22 and ITQ-2 zeolites Catalyst Fig. 3. 27Al MAS NMR spectra for (a) MCM-22 precursor, (b) MCM-22 zeolite, and (c) ITQ-2 zeolite.

MCM-22 ITQ-2 a

External surface area (m2 g1)

110 592

Micropore volume (cm3 g1)

0.160 0.047

Number of acid sites (mmol g1) Totala

Externalb

1.13 0.53

0.14 0.28

The number of acid sites was determined by NH3-TPD. The number of Bro¨nsted acid sites on the external surface was determined by collidine adsorption. b

26

S. Inagaki et al. / Applied Catalysis A: General 318 (2007) 22–27

Fig. 4. FT-IR spectra for (a) MCM-22 and (b) ITQ-2 zeolites after adsorbing pyridine and evacuating.

toluene with methanol to investigate whether this bimolecular reaction takes place on the acid sites in the intralayer 10MR micropores. Table 2 lists the catalytic activities of the four kinds of MWW-type zeolite catalysts as well as the product distributions. In the case of the reaction at 0.69 g h mol-methanol1 of W/ F on fresh MCM-22 catalyst, the major product in alkylated products was p-xylene with 44% selectivity. The level of conversion of toluene decreased slightly from 1.8% at 5 min of

time on stream (TOS) to 1.4% at 65 min. The slight decrease in the conversion of toluene may have been caused by coke deposition during the reaction, as the color of the catalyst changed from white to black over the course of the reaction. However, the distributions of alkylated products were almost constant during the reaction. On the other MWW-type zeolite catalysts, the deactivation of catalysts during the reaction likewise hardly affected product distribution in the alkylated products, as shown in Table 2. Hereafter, we will discuss the product distribution in the alkylated products at 5 min of TOS. MCM-22 zeolite showed a high selectivity to p-xylene at the values of W/F from 0.69 to 2.08 g h mol-methanol1. On the other hand, ITQ-2 showed a high selectivity of 56% to o-xylene, in comparison with 24% selectivity to p-xylene at 2.3% of the level of conversion. As ITQ-2 possessed a larger number of Bro¨nsted acid sites on the external surface (0.28 mmol g1) than MCM-22 (0.14 mmol g1), we conclude that nonselective alkylation of toluene with methanol can occur on the external surfaces of MWW-type zeolites. As collidine selectively poisons the acid sites on the external surface of ITQ-2, the remaining acid sites are assumed to be located mainly in the intralayer 10MR micropores. The selectivity of ITQ-2 to p-xylene was increased from ca. 24 to 80% by poisoning it with collidine. Therefore, we suppose that p-xylene can be selectively produced in the intralayer 10MR micropores of MWW structure. Similarly, P-MCM-22, which is MCM-22 poisoned with collidine but possessing unpoisoned active sites in both intra- and interlayer 10MR micropores, showed a high selectivity of 74% to p-xylene at 3.8% of the level of conversion. We conclude that shape-selective alkylation to pxylene also occurred in the interlayer 10MR micropores of MWW structure. While both m- and o-xylene might be produced in the interlayer 10MR micropores of P-MCM-22, the selective formation of p-xylene was caused by the shape selectivity of the product limitation, which is explained by the difference in

Fig. 5. Plausible scheme of the alkylation of toluene with methanol on MWW-type zeolites.

S. Inagaki et al. / Applied Catalysis A: General 318 (2007) 22–27

27

Table 2 Catalytic activities and product distributions in the alkylation of toluene with methanol over MWW-type zeolites Catalyst

MCM-22

W/F (g h) (mol-methanol)1

0.69 1.35 2.08

Time on stream (min)

Conversion of toluene (%)

Conversion of methanol (%)

Yield of DME (%)

Product distribution in alkylated products (%) p-X

m-X

o-X

1,3,5TMB

1,2,4TMB

1,2,3TMB

5 65 5 95 5 65

1.8 1.4 3.1 2.3 5.4 4.4

36.9 32.5 49.9 44.5 65.8 62.1

34.9 31.0 46.5 42.1 59.8 57.4

44.0 43.5 41.8 41.0 40.4 39.9

14.7 14.0 14.3 13.9 15.2 14.9

37.9 39.5 36.1 39.2 32.5 34.8

0.0 0.0 0.0 0.0 0.0 0.0

3.4 3.0 5.1 3.9 8.9 6.9

0.0 0.0 2.6 2.0 3.1 3.5

P-MCM-22a

2.13

5 95

3.8 2.9

54.0 46.7

50.2 43.7

74.1 78.5

11.7 11.0

6.3 5.3

0.0 0.0

7.9 5.2

0.0 0.0

ITQ-2

2.26

5 95

2.3 1.7

33.6 27.9

31.1 26.0

23.6 23.0

13.4 12.6

56.2 59.1

0.0 0.0

2.7 2.1

4.0 3.2

P-ITQ-2a

4.29

5 95

0.6 0.5

34.9 36.0

34.3 35.5

80.0 83.0

10.3 8.7

9.7 8.3

0.0 0.0

0.0 0.0

0.0 0.0

Reaction conditions: temperature, 250 8C; weight of catalyst, 10 or 20 mg; partial pressure of methanol, 9 kPa; partial pressure of toluene, 9 kPa. DME, dimethyl ether; X, xylene; TMB, trimethylbenzene. a P-MCM-22 and P-ITQ-2 represent the catalysts poisoned with collidine.

diffusion rates of xylene isomers in the interlayer 10MR micropores, similarly to the 10MR micropores of ZSM-5 zeolite [21]. At the same level of conversion, as the selectivity to mxylene was decreased from 14.3% to 11.7% by poisoning MCM22 with collidine, nonselective alkylation of toluene and xylene isomerization might occur to give m-xylene on the external acid sites of the MWW structure. Fig. 5 shows a plausible scheme of the alkylation of toluene with methanol on MWW-type zeolites. The high selectivity of p-xylene is caused by shape selectivity originating from both 10MR micropores in the MWW structure. This contention is supported by the high selectivity to p-xylene observed in PITQ-2 and P-MCM-22 possessing unpoisoned active sites in both kinds of 10MR micropores. In particular, since the interlayer 10MR micropores have supercages, the shape selective formation of p-xylene in the interlayer 10MR micropores is probably caused by the diffusion limitations of m- and o-xylenes. On the other hand, as ITQ-2 indicated a high selectivity to o-xylene, the external surface of MWW-type zeolites showed nonselective alkylation of toluene with methanol. In addition, P-ITQ-2 hardly produced any trimethylbenzenes (TMBs) under the reaction conditions investigated in this study, whereas the other MWW-type zeolite catalysts yielded TMBs. These results suggest that TMBs formed on the external surface and in the interlayer micropores with 12MR supercages, which provide sufficient space to allow the successive alkylation of xylenes with methanol. 4. Conclusions We investigated the contributions of acid sites in the intra- and interlayer 10MR micropores of MCM-22 and ITQ-2 zeolites to the alkylation of toluene with methanol at 250 8C by selectively poisoning the acid sites on external surfaces with collidine. We concluded that the high selectivity of p-xylene in MCM-22 zeolite is caused by shape selectivity originating from the

interlayer 10MR micropores with 12MR supercages, as well as the intralayer sinusoidal 10MR micropores in the MWW structure. On the other hand, nonselective alkylation of toluene with methanol occurs on the acid sites on the external surface of MWW structures. References [1] M. Rubin, P. Chu, U.S. Patent 4,954,325 (1990). [2] L. Puppe, J. Weisser, U.S. Patent 4,439,409 (1984). [3] S.I. Zones, D.I. Holtermann, R.A. Innes, T.A. Pecoraro, D.S. Santilli, J.N. Ziemer, U.S. Patent 4,826,667 (1989). [4] G. Bellussi, G. Perego, M.G. Clerici, A. Giusti, European Patent Application 293,032 (1988). [5] M.A. Camblor, A. Corma, M.-J. Dı´az-Caban˜as, Ch. Baerlocher, J. Phys. Chem. B 102 (1998) 44–51. [6] P. Chu, M.E. Landis, Q.N. Le, U.S. Patent 5,334,795 (1994). [7] A. Corma, V. Martı´nez-Soria, E. Schnoeveld, J. Catal. 192 (2000) 163–173. [8] M.E. Leonowicz, J.A. Lawton, S.L. Lawton, M.K. Rubin, Science 264 (1994) 1910–1913. [9] W. Souverijns, W. Verrelst, G. Vanbutsele, J.A. Martens, P.A. Jacobs, J. Chem. Soc. Chem. Commun. (1994) 1671–1672. [10] A. Corma, C. Corell, F. Llopis, A. Martine´z, J. Pe´rez-Pariente, Appl. Catal. A 115 (1994) 121–134. [11] A. Corma, V. Forne´s, S.B. Pergher, Th.L.M. Maesen, J.G. Buglass, Nature 396 (1998) 353–356. [12] A. Corma, V. Forne´s, J.M. Guil, S. Pergher, Th.L.M. Maesen, J.G. Buglass, Microporous Mesoporous Mater. 38 (2000) 301–308. [13] A. Corma, V. Forne´s, J. Martı´nez-Triguero, S.B. Pergher, J. Catal. 186 (1999) 57–63. [14] A. Corma, V. Forne´s, Stud. Surf. Sci. Catal. 135 (2001) 73–82. [15] (a) S. Laforge, D. Martin, J.L. Paillaud, M. Guisnet, J. Catal. 220 (2003) 92–103; (b) S. Laforge, D. Martin, M. Guisnet, Microporous Mesoporous Mater. 67 (2004) 235–244. [16] A. Corma, V. Martı´nez-Soria, E. Schnoeveld, J. Catal. 192 (2000) 163–173. [17] P. Wu, T. Komatsu, T. Yashima, Microporous Mesoporous Mater. 22 (1998) 343–356. [18] R.S. Mikhail, S. Brunauer, E.E. Bodor, J. Colloid Interface Sci. 26 (1968) 45. [19] H. Du, D.H. Olson, J. Phys. Chem. B 106 (2002) 395–400. [20] J.-H. Kim, A. Ishida, M. Niwa, React. Kinet. Catal. Lett. 67 (1999) 281–287. [21] J.-H. Kim, T. Kunieda, M. Niwa, J. Catal. 173 (1998) 433–439.