Methanol Conversion and Propene Oligomerization Productivity of

Catalysis Research Unit, Department of Chemical Engineering, University of Cape Town, Rondebosch 7700, Republic of South Africa. Ind. Eng. Chem...
1 downloads 0 Views 143KB Size
Ind. Eng. Chem. Res. 1996, 35, 697-702

697

Methanol Conversion and Propene Oligomerization Productivity of Dealuminated Large-Port Mordenites Miles J. van Niekerk Department of Chemical Engineering, The Queen’s University of Belfast, Belfast BT9 5AG, Northern Ireland

Cyril T. O’Connor*,† and Jack C. Q. Fletcher† Catalysis Research Unit, Department of Chemical Engineering, University of Cape Town, Rondebosch 7700, Republic of South Africa

Four different types of large-port mordenite were studied. Three of these catalyst samples were dealuminated by treatment with nitric acid, and the fourth type was a series of commercially available dealuminated mordenites. The methanol conversion and propene oligomerization productivities and selectivities of these dealuminated mordenites were investigated on a laboratory scale at typical industrial reaction temperatures and pressures. The optimum catalyst morphology and degree of dealumination needed for maximum productivity were found to be similar for both methanol conversion and propene oligomerization. High catalytic productivities were obtained with mordenite which had been hydrothermally dealuminated or with small crystallites which had been synthesized or dealuminated in such a way as to minimize the amount of extra-framework aluminum in the mordenite pores. Introduction It has been reported that dealuminated mordenites show improved catalytic performance for a number of chemical reactions (Fernandes et al., 1994; Pardillos et al., 1989; Niwa et al., 1988; Karge and Weitkamp, 1986; Haas et al., 1985). Mordenite may be synthesized with Si/Al ratios ranging between 4.5 and 17 (Chumbhale et al., 1992; Itabashi et al., 1986), and further dealumination is often accomplished by acid leaching, steam treatment or heat treatments (Raatz et al., 1983; Beyer et al., 1984; Van Geem et al., 1988; Goovaerts et al., 1989). Different optimum aluminum contents have been reported for various reactions over dealuminated mordenites. In this study, the conversion of methanol to light olefins and the oligomerization of propene have been investigated for four different series of dealuminated mordenites in order to establish the type and extent of dealumination required for maximum catalyst productivity and the optimum catalyst morphologies required for these two reactions. Experimental Section Catalysts. The catalysts used in the study were Zeocat mordenite (prefix ZM), Norton mordenite Zeolon 900 Na (prefix NM), and two batches of mordenite synthesized at different stirring speeds (prefixes S1 and S2). The mordenite samples obtained from Zeocat, viz., ZM510, ZM760, and ZM980, had been dealuminated by a combination of hydrothermal and acid treatments. The Norton mordenite was obtained as 1/16 in. binderless extrudates in the sodium form. In the manufacture of the extrudates it is possible that a viscosity modifier such as (hydroxymethyl)cellulose was used (this would be removed by combustion during calcination). The NM material was crushed and sieved, and particles of size less than 75 µm were used in this study. Two batches of mordenite were synthesized in a mechanically stirred * Author to whom all correspondence should be addressed. † Fax: 0027 21 6502516.

0888-5885/96/2635-0697$12.00/0

autoclave at 190 °C under autothermal pressure as described by Itabashi et al. (1986). The first batch was synthesized at an autoclave impeller speed of 300 rpm (S1) and the second at an impeller speed of 30 rpm (S2). The molar ratio of both synthesis mixtures was Na2O/ Al2O3/SiO2/H2O ) 1.7/1/11.5/230, and the synthesis time was 48 h. Hydrogen form samples (denoted -H) were obtained by calcination of the ammonium forms in air at 400 °C for 12 h. Dealumination Procedure. The dealumination procedure used was that described by Karge et al. (1984). Samples designated -01 were prepared by stirring 4 g of sodium mordenite in 53 mL of 6 N HNO3 for 4 hours at room temperature, washing, drying, and crushing and repeating these treatments once. More severely dealuminated samples were prepared by multiple cycles of refluxing 4 g of sodium mordenite in 53 mL of 6 N HNO3 for 4 h and then washing, drying, and crushing as described above. Samples undergoing two and four such refluxing cycles are designated -02 and -03, respectively. Catalyst Characterization. Catalyst aluminum contents were determined by atomic absorption spectroscopy. For this purpose, samples were prepared by dissolving 100 mg of the catalyst in hydrofluoric/ hydrochloric acid (5 mL of 40% HF, 5 mL of 31% HCl) at room temperature. After dissolution 20 mL of a saturated boric acid solution was added to complex the excess hydrofluoric acid, and the resulting solution was made up to 50 mL using distilled water. Standards were prepared using hydrofluoric, hydrochloric, and boric acid concentrations corresponding to those used for sample preparation. Framework aluminum is generally taken as the tetrahedral aluminum as determined by 27Al MAS NMR (Van Geem et al., 1988; Goovaerts et al., 1989; Sawa et al., 1990). 27Al MAS NMR spectra were recorded using a Bruker AM 300 spectrometer at a spin rate of 5000 Hz, a pulse width of 1.5 µs, a pulse angle of 30°, and a relaxation delay of 0.1 s and employing AlCl3(H2O)6 as an external standard at 0 ppm. Prior to NMR analysis, © 1996 American Chemical Society

698

Ind. Eng. Chem. Res., Vol. 35, No. 3, 1996

samples were calcined in air at 400 °C for 12 h and allowed to equilibrate under ambient conditions. Ammonia temperature-programmed desorption (NH3 TPD) spectra were recorded in the range 100-600 °C (0.5 g of sample, 10 °C/min temperature ramp, 60 mL/ min helium carrier). After ammonia adsorption, the catalyst samples were flushed in flowing helium for 2 h prior to ramping the temperature. All the TPD spectra displayed two characteristic desorption peaks, and for integration purposes these peaks were separated by dropping a perpendicular line from the trough between the two peaks to the baseline of the spectra. Electron microscopy showed that the catalyst crystallite sizes and morphologies varied considerably from one catalyst to another. The NM samples consisted of ca. 15 µm intergrown crystals. The ZM samples consisted of 10-50 µm agglomerates of ca. 2.5 µm crystals, some of which were intergrown. The S1 samples consisted of 3-10 µm agglomerates of 0.5 µm crystals which were uniform in size and shape. The S2 mordenite consisted of two different types of agglomerates, both of which were made up of long needle-shaped crystals. The first agglomerate type consisted of tightly packed bundles of crystals about 1 µm long and 0.1 µm in diameter. The second agglomerate type, which made up much of the material, consisted of randomly orientated crystals which were about 1.5-2.0 µm long and 0.5 µm in diameter. None of the catalyst morphologies were observed to change, even after severe dealumination. Conversion of Methanol to Light Olefins. The methanol conversion experiments were carried out in a fixed-bed borosilicate glass reactor with an internal diameter of 16 mm. High-purity nitrogen was used as a carrier gas for the methanol feed. The nitrogen stream, controlled by a Brooks mass flow controller, was saturated with methanol using a double-stage saturator. The reaction temperature used was 400 °C, the methanol space velocity 1.0 g/g/h, and the methanol partial pressure 22 kPa. Catalyst samples were calcined in air (60 mL/min) at 400 °C for 12 h and then purged in flowing dry nitrogen (60 mL/min). Reaction products were sampled on-line and were separated on a Supelco DH150 capillary column and detected by flame ionization. Catalyst performances were compared as a function of the catalyst utilization value (CUV), which is defined as the mass of methanol converted to hydrocarbons (i.e., excluding dimethyl ether) before the oxygenate conversion level drops below 50%. Propene Oligomerization. The reactor configuration in which the oligomerization studies were carried out has been described by Schwarz et al. (1991). The system consisted of a high-pressure fixed-bed stainless steel reactor to which a propene/propane mixture (containing approximately 87% propene) was fed at a weight hourly space velocity of 12 g/g/h using a high-pressure diaphragm pump. The reaction products were separated into liquid and gas samples which were analyzed using a 30 m megabore column with a DB1 coating and a 2 m stainless steel packed column with an OV101 packing, respectively. Reaction experiments were carried out at 5 MPa using 1.0 g of catalyst which had been calcined in air (60 mL/ min) at 400 °C for 12 h and then cooled to the initial reaction temperature in flowing dry nitrogen (60 mL/ min). As some catalysts were inactive at 200 °C (typical low-temperature industrial oligomerization), the catalyst samples were run at a stepped temperature pro-

gram. The reactor temperature was held at 200 °C until the first drop of liquid was collected and then maintained at this temperature for 1 h. The temperature was then increased to 250 °C over 1 h and this temperature maintained for 1 h. The temperature was increased to 300 °C over 1 h and maintained for 1 h and finally increased to 350 °C over 1 h and maintained at this temperature for 2 h. The liquid and gas products were sampled every hour. Catalyst performances were compared as a function of initial liquid production, which is defined as the mass of liquid product produced per gram of catalyst during the first 5 h of the run (the industrial temperature range usually used for olefin oligomerization: 200-300 °C). Coke Analysis. Thermogravimetric analysis of the coked catalyst was carried out using a Stanton-Redcroft STA 780 thermal analyzer. Nitrogen and air were used to remove coke deposits over a temperature range of 30-500 °C at a heating rate of 10 °C/min. The coke removed between 350 and 500 °C using nitrogen was classified as “high boiling point hydrocarbons” and the coke removed subsequently, using air at 500 °C, as graphitic coke. Results Methanol Conversion over Mordenite. The initial methanol conversion levels of all the catalysts were above 95%. The reactivity trends of the S1 and S2 series of catalysts were similar in that the mildly dealuminated samples were less productive than the untreated samples and the severely dealuminated catalysts had the highest CUVs. None of the NM samples showed significant methanol conversion productivity after the first product sample was taken. Of the mordenites investigated, the ZM samples were the most productive, with their CUVs increasing with increasing dealumination (Table 1). The product selectivities of the catalysts changed significantly as the samples deactivated. The initial reaction products at complete oxygenate conversion consisted of mostly C1-C4 paraffins. As the catalyst samples deactivated and the oxygenate conversion dropped, the selectivity shifted to mostly C2-C4 olefins. The most productive samples (ZM and S1-02/03) produced more high molecular weight hydrocarbons than the less active catalysts, with the combined C9+ and aromatic selectivity rising to 34% (Table 1). The amount of graphitic coke deposited on the catalyst samples varied between 6.1% and 15% and was generally higher for the more productive catalysts. Propene Oligomerization over Mordenite. The productivities of the various mordenite catalysts for propene oligomerization are reported in Table 2. The mildly dealuminated NM, S1, and S2 samples were generally less productive than the unmodified catalysts, and the more severely dealuminated samples were the most productive of the mordenite samples investigated. The severely dealuminated S1 samples had much higher initial productivities than the other mordenite samples investigated, some of which produced no liquid product until the reaction temperature was ramped to 350 °C. The oligomer product spectra of most of the catalysts followed a general trend. At low temperatures (200250 °C) the liquid product had a high fraction of typical distillate hydrocarbons (C12+ fraction) and the oligomer groupings could be clearly seen in the GC traces. At 300 °C the C12+ fraction decreased and the oligomer groupings became indistinct. There were no significant

Ind. Eng. Chem. Res., Vol. 35, No. 3, 1996 699 Table 1. Methanol Conversion Performance of Dealuminated Mordenites catalyst/variable

total Al (wt %)

tetrahedral Al (wt %)

acidity NH3 TPD (mmol/g of cat.)

methanol conv. CUV (g/g of cat.)

C22-C42(sel.) (wt %)

MTO C9+ Sel. (wt %)

graphitic coke (g/100 g of cat.)

NM-H NM-01 NM-02 NM-03

6.34 5.82 4.88 3.54

6.19 4.19 3.02 2.30

1.77 1.65 1.17 0.87

0.85 0.85 0.85 0.85

9.5 5.7 5.1 9.1

3 7 4 5

7.4 8.5 9.1 10.0

ZM-510 ZM-760 ZM-980

4.01 1.08 0.52

3.17 0.77 0.43

1.05 0.32 0.12

2.6 8.9 32.9

27.6 28.4 28.2

16 28 23

11.5 15.0 13.6

S1-H S1-01 S1-02 S1-03

6.73 6.08 0.91 0.87

6.53 3.89 0.89 0.86

2.20 1.17 0.35 0.29

1.3 1.0 4.9 2.7

43.7 9.3 16.5 15.8

4 4 34 32

9.1 9.2 11.0 11.6

S2-H S2-01 S2-02 S2-03

6.70 6.40 4.17 2.77

5.89 4.96 3.21 2.33

1.66 1.70 0.95 0.58

1.1 0.93 0.91 1.3

17.3 19.5 12.6 19.3

4 3 3 4

6.1 6.4 7.2 9.0

Table 2. Propene Oligomerization Performance of Dealuminated Mordenites catalyst

tetrahedral aluminum (wt %)

total Al/unit cell

EFAL/pore

oligomeriz. productivity (g/g of cat.)

oligomeriz. C12+ select. (wt %)

graphitic coke (g/100 g of cat.)

NM- H NM-01 NM-02 NM-03

6.19 4.19 3.02 2.30

6.76 6.13 5.01 3.53

4159 45860 50929 32991

5.54 0.00 0.00 0.90

57

12.3 7.1 7.4 8.1

ZM-510 ZM-760 ZM-980

3.17 0.77 0.43

4.01 1.00 0.48

2815 939 294

4.70 7.18 7.35

60 64 58

14.1 18.1 12.8

S1-H S1-01 S1-02 S1-03

6.53 3.89 0.89 0.86

7.18 6.38 0.84 0.80

115 1228 11 7

6.71 4.18 15.2 12.7

49 63 47 56

17.6 8.5 11.8 10.9

S2-H S2-01 S2-02 S2-03

5.89 4.96 3.21 2.33

7.15 6.77 4.18 2.68

2313 4080 2569 1146

3.55 2.58 0.00 7.00

80

4.9 3.7 5.3 8.8

differences in the C12+ selectivities between the different types of mordenite except for the S2-02 sample, which produced an 80% C12+ fraction. The amount of graphitic coke formed on the various samples varied between 3.7% and 18.1%. Discussion From Tables 1 and 2, it can be seen that for methanol conversion and propene oligomerization, the catalytic productivity of dealuminated mordenite is not simply a function of catalyst acidity. Although several authors (Scherzer et al., 1984; Karge and Dondur, 1990; Ghosh and Curthoys, 1984; Kiovsky et al., 1978) have reported increased acid site strength on dealumination, the rate of coke formation for a number of reactions is found to decrease with reduction in aluminum content (Karge et al., 1985; Karge and Boldingh, 1988; Itoh et al., 1982; Haas et al., 1985), indicating that coke formation is probably also dependent on acid site concentration. In agreement with these findings, several authors have found increasing methanol conversion lifetimes with increasing mordenite dealumination (Bandiera et al., 1984; Sawa et al., 1989; Meyers et al., 1988). The dependence of coking on acid site concentration does not explain the differences in the productivity trends of the NM, S1, and S2 materials. These differences are brought about by differences in crystallite size, the small crystallite S1 (0.5 µm) material being more productive than the S2 material (1.5 µm), which is more productive than the NM material (15 µm). The ZM material cannot be compared to the NM, S1, and S2

catalysts due to the different dealumination procedure employed for this material. The hydrothermal conditions used in the dealumination of the ZM samples are conducive to silicon migration, resulting in the replacement of lost aluminum by silicon. This phenomenon stops the unit cell contraction, which was seen for the NM, S1, and S2 samples (van Niekerk et al., 1992), and thus prevents the pore diameters from decreasing. A reduction in unit cell size, and hence pore diameter, as seen in the acid dealuminated samples (van Niekerk et al., 1992) reduces reactant and product diffusivities and results in increased product residence times. Among these products are coke precursors, so this results in increased coking and reduced catalyst lifetime. As may be expected, the diffusivity of hydrocarbon molecules within the pores of a zeolite crystal is unaffected by crystallite size (Post, 1991). The rate of hydrocarbon adsorption/desorption, however, has been found to increase with decreasing crystallite size (Lin et al., 1989; Ruthven and Eic, 1988; Shah et al., 1988; Chon and Park, 1988). Herrmann et al. (1987) found decreased conversion with increasing crystallite size of ZSM-5 for methanol amination, methanol conversion, and hexane cracking and ascribed this phenomenon to underutilization of the interior regions of the larger crystallites. Similar findings were reported by Petrik et al. (1995) for propene oligomerization over ZSM-5. Although the influence of crystallite size may be used to explain the differences in productivity between the NM, S1, and S2 samples, the reduced CUVs of the samples containing between 0.8 and 1.7 mmol/g of

700

Ind. Eng. Chem. Res., Vol. 35, No. 3, 1996

acidity require a different explanation. Samples in this range all contain significant amounts of octahedral aluminum within the channel structure. These aluminum species reduce the effective channel diameter and therefore reduce reactant and product diffusivities, which results in increased reactant and product residence times within the catalyst pores as described above. This increase in hydrocarbon residence time increases the amount of reaction products which are converted to coke, thereby reducing the catalyst CUV and lifetime. Configurational diffusion occurs in situations where the structural dimensions of the pores approach those of the molecules (Weisz, 1973). Under such conditions, small changes in pore diameter may reduce the diffusivity by several orders of magnitude (Chen et al., 1988). That small changes in the size of pore openings strongly affect the catalytic performance of mordenite was shown by Niwa et al. (1988), who demonstrated that reducing the pore mouth opening of mordenite from 6-7 to 5-6 Å resulted in large reductions in the cracking rates of various C8 paraffins. Similar results were reported by Peeters et al. (1984) for the sorption capacity of various gases (Xe, Kr, Ar, N2, O2, H2). In light of these results it is reasonable to assume that the deposition of extra-framework aluminum within the mordenite pores would result in reduced reaction product (and thus coke precursor) diffusivities as also suggested by Scherzer (1984) and therefore longer residence times of these products within the catalyst pores. The ionic diameter of Al3+ is 1.06 Å and, depending on its position in the mordenite pore, could reduce the pore diameter by as much as 17%. This calculation is made using only the ionic radius of the aluminum ion and does not take into account the contribution of any of the oxygen atoms which would be bonded to the aluminum atom. In addition, Sawa et al. (1992) suggested that extra-framework aluminum species may be occluded as blocks or clusters in the zeolite crystal, a situation which would result in a further reduction in reactant and product diffusivities. As is the case with different crystallite sizes, samples which have physical characteristics which cause increased product residence times have increased rates of coke formation (series nature of coking reaction) and therefore have reduced catalytic lifetimes and CUVs. The approximate amount of extra-framework aluminum per pore (EFALpore) may be calculated using the average length of a pore in the crystallite (Lcrystallite) (side length for cubic crystals and length for needle-shaped crystals), the average length of a unit cell (Luc, approximately 7.5 Å), and the number of extra-framework aluminum atoms per unit cell (EFALuc), which may be calculated from the total aluminum content and the relative amounts of octahedral and tetrahedral aluminum: EFALpore ) EFALuc × Lcrystallite/Luc. The use of this parameter, viz., the number of extraframework aluminum per pore, takes into account the effect of extra-framework material as well as that of crystallite size (Figures 1 and 2), both of which increase the reactant and product residence times within the mordenite pores and thus the coking rates of these samples. The more productive samples containing less extra-framework aluminum per pore had increased selectivities toward the higher molecular weight hydrocarbons, indicating that the reduced hydrocarbon residence times in these samples allowed the escape of these larger molecules before they could be converted to coke.

Figure 1. Methanol conversion CUV as a function of extraframework Al/pore.

Figure 2. Propene oligomerization liquid production as a function extra-framework Al/pore.

Acid dealuminated samples containing above ca. 2000 extra-framework aluminum atoms/pore were quite inactive for both methanol conversion and propene oligomerization. For NM, S1, and S2 samples containing less than ca. 2000 extra-framework aluminum atoms/ pore, the methanol conversion CUV and propene oligomerization productivity increased with decreasing the number of extra-framework aluminum atoms per pore. An exception to this was the S1-03 sample which was less productive than the S1-02 material. This drop in catalytic productivity is expected considering that the S1-03 was the most severely acid dealuminated sample studied, exhibiting reduced unit cell dimensions and severely reduced cyclohexane adsorption levels (van Niekerk et al., 1992). Sawa et al. (1989) also found that mordenite which had been extensively and severely dealuminated by acid treatment deactivated more rapidly than material which was dealuminated under milder conditions. Considering the reduced cyclohexane adsorption levels of this sample, it is likely that partial collapse of the pores has occurred; reducing reactant and product diffusion rates which would result in an increased coking rate and hence a reduced CUV. It has also been suggested that in the so-called “T-jump” mechanism for filling of vacancies following acid dealumination, whole cages may collapse, with the residual

Ind. Eng. Chem. Res., Vol. 35, No. 3, 1996 701

amorphous material creating severe diffusional restrictions within the pores. Unlike with the methanol conversion reaction, the initial propene oligomerization productivity trend of the ZM samples was similar to that of the NM, S1, and S2 samples when plotted as a function of the amount of extra-framework aluminum per channel. The reaction temperatures used for propene oligomerization were much lower than those used for methanol conversion (200-300 °C compared to 400 °C), resulting in a reduced Thiele modulus. Under such conditions, the additional diffusional restrictions created by a reduction in pore diameter in the case of the acid-treated samples are of less importance than the reaction rate, and hence the reduced diffusional restrictions imposed by the pore system of the ZM samples do not have the marked effect that they would have at the higher temperature methanol conversion reaction conditions. Conclusions The trends in catalytic productivity of the dealuminated mordenites were the same for both methanol conversion and propene oligomerization. Reducing mordenite crystallite size improves its catalytic productivity. The dominant factor affecting the catalytic productivity of the various mordenite samples seems to be the reactant and product residence times within the zeolite pores. Increased residence times lead to higher coking rates due to the series nature of the coking reactions. Residence times are affected by both crystallite size and occlusion of extra-framework material within the zeolite pores. This material may be silicon species deposited during synthesis or extra-framework aluminum species deposited during the dealumination process. If dealumination is accomplished using mineral acid treatment only, excessive dealumination (greater than 80% aluminum removal) is not desirable, as this results in collapse of the crystal structure, which leads to reduced methanol conversion and propene oligomerization catalytic productivity. Dealumination accomplished using hydrothermal treatment, which facilitates the replacement of extracted aluminum by silicon, is desirable as catalysts prepared in this way have increased methanol conversion productivities. Minimization of the amount of extra-framework aluminum may be difficult if only mild catalyst dealumination is required unless a dilute acid solution is used in the dealumination procedure. In comparison to ZSM5, which is the commercially favored catalyst for methanol conversion to light olefins and olefin oligomerization (although at different Si/Al ratios), even the highly dealuminated ZM samples exhibit severely reduced lifetimes and the mordenite catalysts are far more susceptible to coke formation than ZSM-5. Literature Cited Bandiera, J.; Hamon, C.; Naccache, C. Modified mordenites for catalytic conversion of methanol. In Proceedings of the sixth international zeolite conference; Olson, D., Bisio, A., Eds.; Butterworth: London, 1984; pp 337-344. Beyer, H. K.; Belanykaja, I. M.; Mishin, I. V.; Borbely, G. Structural peculiarities and stabilization phenomena of aluminum deficient mordenites. In Structure and Reactivity of Modified Zeolites; Jacobs, P. A., Jaeger, N. I., Jiru, P., Schulzekloff, G., Kazansky, V. B., Eds.; Elsevier: Amsterdam, The Netherlands, 1984; pp 133-140.

Chen, N. Y.; Garwood, W. E.; Dwyer, F. G. In Shape selective catalysis in industrial applications; Marcel Dekker Inc.: New York, 1988; p 56. Chon, H.; Park, D. H. Diffusion of cyclohexanes in zsm-5 zeolites. J. Catal. 1988, 114, 1-7. Chumbhale, V. R.; Chandwadkar, A. J.; Rao, B. S. Characterization of siliceous mordenite obtained by direct synthesis or by dealumination. Zeolites 1992, 12, 63-69. Fernandes, L. D.; Bartl, P. E.; Monteiro, J. L. F.; daSilva, J. G.; deMenezes, S. C.; Cardoso, M. J. B. Zeolites 1994, 14, 533540. Ghosh, A. K.; Curthoys, G. Electronegativity and Bronsted acidity of mordenites. J. Catal. 1984, 86, 454-456. Goovaerts, F.; Vansant, E. F.; Philippaerts, J.; De Hulsters, P.; Gelan, J. Initial cracking properties and physicochemical characterization of acid-leached small-port (sp) and large-port (lp) mordenites by pulse normal-hexane cracking, infrared and Al-27 magic angle spinning nuclear magnetic-resonance spectroscopy. J. Chem. Soc., Faraday Trans. 1 1989, 85, 3675-3685. Haas, J.; Fetting, F.; Gubicza, L. Investigation of the deactivation of mordenite catalysts by coke deposition and their regeneration. Acta Phys. Chem. 1985, 31, 659 Herrmann, C.; Haas, J.; Fetting, F. Effect of the crystal size on the activity of zsm-5 catalysts in various reactions. Appl. Catal. 1987, 35, 299-310. Itabashi, K.; Fukushima, T.; Igawa, K. Synthesis and characteristic properties of siliceous mordenite. Zeolites 1986, 6, 30-34. Itoh, H.; Hattori, T.; Murakami, Y. Product distribution in the conversion of methanol on partially ion-exchanged mordenites. Appl. Catal. 1982, 2, 19-37. Karge, H. G.; Weitkamp, J. Studies on dealuminated mordenite catalysts. Chem.-Ing.-Tech. 1986, 58, 946-959. Karge, H. G.; Boldingh, E. P. Spectroscopic investigations on deactivation of zeolites during reactions of olefins. Catal. Today 1988, 3, 379. Karge, H. G.; Dondur, V. Investigation of the distribution of acidity in zeolites by temperature-programmed desorption of probe molecules. 1. Dealuminated mordenites. J. Phys. Chem. 1990, 94, 765-772. Karge, H. G.; Wada, Y.; Weitkamp, J.; Ernst, S.; Girrbach, U.; Beyer, H. K. A comparative-study of pentasil zeolites and dealuminated mordenites as catalysts for the disproportionation of ethylbenzene. In Catalysis on the Energy Scene; Kaliaguine, S., Mahay, A., Eds.; Elsevier: Amsterdam, The Netherlands, 1984; pp 101-111. Karge, H. G.; Boldingh, E. P.; Lange, J.-P.; Gutsze, A. Studies of coke formation on dealuminated mordenites by in-situ IR and EPR measurements. Acta Phys. Chem. 1985, 31, 639-648. Kiovsky, J. R.; Goyette, W. J.; Notermann, T. M. Acid site promotion of mordenite. J. Catal. 1978, 52, 25. Lin, D. H.; Ducarme, V.; Coudurier, G.; Vedrine, J. C. Adsorption and diffusion of different hydrocarbons in mfi zeolite of varying crystallite sizes. In Zeolites as catalysts, sorbents and detergent builders; Karge, H. G., Weitkamp, J., Eds.; Elsevier: Amsterdam, The Netherlands, 1989; pp 615-623. Meyers, B. L.; Fleisch, T. H.; Ray, J. T.; Miller, G. J.; Hall, J. B. A multitechnique characterization of dealuminated mordenites. J. Catal. 1988, 110, 82-95. Niwa, M.; Sawa, M.; Murakami, Y. In Proceedings of the 9th International Catalysis Congress; Phillips, M. J., Ternan, M., Eds.; Chemical Institute of Canada: Ottawa, 1988; p 380. Pardillos, J.; Coq, B.; Figueras, F. Isomerization of ortho-dichlorobenzene over H-mordeniteseffect of the silicon-to-aluminum ratio. Appl. Catal. 1989, 51, 285-293. Peeters, G.; Thijs, A.; Fansant, E. F.; De Bievre, P. Pore-size engineering by modifying the zeolitic pore system of mordenite LP. In Proceedings of the sixth international zeolite conference; Olson, D., Bisio, A., Eds.; Butterworth: London, 1984; pp 651659. Petrik, L. F.; O’Connor, C. T.; Schwarz, S. The influence of various parameters on the morphology and crystal size of ZSM-5 and the relationship between morphology and crystal size and propene oligomerization activity. In Proceedings of the Zeocat Conference; Beyer, H. K., Kirisci, I., Nagy, L. B., Eds.; Elsevier: Amsterdam, The Netherlands, 1995; pp 517-524. Post, M. F. In Introduction to zeolite science and practice; van Bekkum, H., Flanigen, E. M., Jansen, J. C., Eds.; Elsevier: Amsterdam, The Netherlands, 1991; p 391.

702

Ind. Eng. Chem. Res., Vol. 35, No. 3, 1996

Raatz, F.; Freund, E.; Marcilly, C. Study of small-port and largeport mordenite modifications. 1. Preparation of the HM forms. J. Chem. Soc., Faraday Trans. 1 1983, 79, 2299-2309. Ruthven, D. M.; Eic, M. Intracrystalline diffusion in zeolites. In Perspectives in molecular sieve science; Flank, W. H., Whyte, T. E., Jr., Eds.; American Chemical Society: Washington, DC, 1988; pp 362-375. Sawa, M.; Niwa, M.; Murakami, Y. Acid-leached dealuminated mordeniteseffect of acid concentration on catalyst life in methanol conversion. Appl. Catal. 1989, 53, 169-181. Sawa, M.; Niwa, M.; Murakami, Y. Relationship between acid amount and framework aluminum content in mordenite. Zeolites 1990, 10, 532-538. Sawa, M.; Niwa, M.; Murakami, Y. Change of pore-opening structure of mordenite upon dealumination by hydrochloric-acid. Zeolites 1992, 12, 175-179. Scherzer, J. The preparation and characterization of aluminumdeficient zeolites. In Catalytic materials: relationship between structure and reactivity; Whyte, T. E., Dallabetta, R. A., Derouane, E. G., Baker, R. T. K., Eds.; American Chemical Society: Washington, DC, 1984; pp 157-200. Schwarz, S.; Kojima, M.; O’Connor, C. T. Effect of stirring, extrusion and pelletization on high-pressure propene oligomerization and xylene isomerization over ZSM-5. Appl. Catal. 1991, 68, 81-96.

Shah, D. B.; Hayhurst, D. T.; Evanina, G.; Guo, C. J. Sorption and diffusion of benzene in hzsm-5 and silicalite crystals. AIChE J. 1988, 34, 1713-1717. Van Geem, P. C.; Scholle, K. F. M.; van der Velden, G. P. M.; Veeman, W. S. Initial cracking properties and physicochemical characterization of acid-leached small-port (sp) and large-port (lp) mordenites by pulse normal-hexane cracking, infrared and Al-27 magic angle spinning nuclear magnetic-resonance spectroscopy. J. Phys. Chem. 1988, 92, 1585-1589. van Niekerk, M. J.; Fletcher, J. C. Q.; O’Connor, C. T. Characterization of dealuminated large-port mordenites. J. Catal. 1992, 138, 150-163. Weisz, P. V. ZeolitessNew horizons in catalysis. CHEMTECH 1973, 3, 498-505.

Received for review June 9, 1995 Revised manuscript received October 12, 1995 Accepted December 19, 1995X IE9503454

X Abstract published in Advance ACS Abstracts, February 15, 1996.