Metylcyclohexane conversion to light olefins

Metylcyclohexane conversion to light olefins

C.F. SCOFIELD¹, E. BENAZZI², H. CAUFFRIEZ² and C. MARCILLY²

¹departamento de Físico - Química, Instituto de química, Universidade do estado do

Rio de Janeiro (UERJ) - Rua São Francisco Xavier, 524, Maracanã - Rio de Janeiro - RJ, Brazil,

CEP: 20 550-013 Phone: [phone] or [phone] - Fax: [phone]

²Division cinétique et Catalyse, Institut Français du pétrole, 1 et 4, Avenue de Bois-Préau, 92.852

Rueil-Malmaison Cedex - France, Phone: [phone] - FAX: [phone]

(Received: November 5, 1997; Accepted: April 13, 1998)

Abstract - This study consists in the evaluation of the catalytic properties of zeolites with different structures in the conversion of methylcyclohexane to light olefins. Results obtained suggest that the steric constrictions of the catalysts used play an important role in hydrogen transfer reactions. Higher selectivities for light olefins (C₃⁼ and C₄⁼) were observed for zeolites having more closed structures, like MFI and ferrerite, when compared to those having more open ones, like beta, omega and faujasite.

Keywords: zeolites, methylcyclohexane, hydrogen transfer.

INTRODUCTION

Naphtenes are present in the majority of petroleum fractions and can be found in significant quantities in the products, as well as in the feed of the fluid catalytic cracking reaction (FCC). Their transformation into better quality molecules such as propylene, of value to the petrochemical industry, and butene for refining (alkylation or estheryfication) is a strategic point to be developed. In addition, naphtenes are key compounds in the study of hydrogen transfer reactions (Wojciechowski and Corma, 1986; Abbot and Wojciechowski, 1987; Cheng and Rajagopalan, 1989; Jaquinot, Mendes, Raatz, Marcilly, Ribeiro and Caeiro, 1990).

Information found in the literature is often related to zeolites having a faujasite structure (Ko, Chang and Tien, 1985; Lin, Gnep and Guisnet, 1989; Corma, Mocholi, Orchilles, Koermer, and Madon, 1991). However, if we consider that catalytic activity and selectivity can be greatly influenced by zeolite structure, it is of considerable interest to understand the behavior of other molecular sieves in reactions such as the cracking of naphtenes.

Thus, the aim of this study is to evaluate the selective transformation of methylcyclohexane (as a model molecule) to light olefins, on zeolitic catalysts having different frameworks and pore diameters in reaction conditions similar to FCC.

EXPERIMENTAL METHODS

The zeolites (supplied by different enterprises) were calcined in air at 550°C for six hours. Then their acidic forms were obtained by ionic exchange with NH₄OH 10N solution in a solution volume/catalyst weight ratio equal to 10. Three successive exchange operations were done, during periods of four hours each. After several washings with distilled water, the materials were heated to about 120°C (before being calcined again).

Catalytic properties of these materials were tested in the conversion reaction of methylcyclohexane, at 500°C and at atmospheric pressure. With reaction temperature, pressure and dilution (molar ratio N₂ / methylcyclohexane = 12) fixed, the different methylcyclohexane (MCHA) conversions (for distinct test reactions) were obtained by varying the weight hourly space velocity (WHSV). The powder catalyst was mixed with the Carborundum (SiC 99%; diameter = 0.340 mm) and put in a stainless steel reactor. The dilution ratio used for the mixture zeolite-Carborundum was 1/50. The reaction zone was completed with Carborundum (diameter = 1.680 mm). The weight of catalysts used varied from 0.025 g to 2.00 g, and the range of the WHSV used was 0.8-360.0 h^-1.

The catalytic tests were carried out in a unit at the Institut Français du Pétrole (IFP), and the catalysts were activated for an hour at 500°C in N₂ flow. The reagent was introduced into the reactor inlet by a piston bomb (reagent flow = 1 g.h^-1 to 8 g.h^-1). The N₂, in the feed mixture (N₂ + MCHA) was also injected in the reactor inlet and was used as carrier gas (N₂ flow = 2.4 - 21.6 l.h^-1). Periodically (at 3, 5, 10 and 20 minutes of time on stream), samples were withdrawn from the reactor effluents and analyzed by on-line gas chromatography (VARIAN, STAR 3400CX equipment). Detection was done by flame ionization (FID) and product separation was performed using a PONA capillary column. Picks were plotted and integrated by an integrator-register HEWLETT-PACKARD 1000.

RESULTS AND DISCUSSION

The catalysts used in this work belong to different zeolite structural groups. However, those having a comparable Si/Al ratio were selected since activity in the catalytic cracking reactions is normally related to solid protonic acidity (strength and number of active sites) and, consequently, to amount of aluminum (Hopkins, 1968). The catalysts tested in this work and their respective Si/Al ratios (in brackets), in decreasing order of pore openings, are:

Y(20) > OMEGA(18/11) > BETA(23) >

> MORDENITE(18) > OFFRETITE(21) >

> MFI(27) > FERRERITE(8.5).

Results of catalytic activity for the different materials, obtained by varying methylcyclohexane conversion (weight percentage) versus whsv, indicated that the FER zeolite presents extremely weak activity and that the BETA zeolite is the most active catalyst tested for the methylcyclohexane cracking reaction. A similar result was reported by T. Cheron(1993) in the study of n-hexene cracking at 300°C. Catalytic activity for the others catalysts used in the present work can be classified in the following order:

Y < MORDENITE (MOR) < MFI < OMEGA <

< OFFRETITE (OFF).

Stability of the catalytic activity was also studied, by observing methylcyclohexane conversion (w%) versus time on stream (Figure 1).

Figure 1 shows that under our working conditions, almost all zeolites tested deactivate (more or less quickly). Nevertheless, the MFI zeolite exhibited a methylcyclohexane cracking activity which was almost constant with time on stream. However, the catalytic activity on the W , MOR and FER zeolites drastically decreases after 5 minutes and continues decreasing while on stream. There is still another group of catalysts (b , Y and OFF) that undergoes intermediate deactivation.

The deactivation of zeolites is probably related to their tendency to produce coke, caused by the formation of polynuclear aromatics, by successive steps of alkylation, cyclization (Diels-Alder reaction) and dehydrogenation (by hydrogen transfer) (Venuto and Hamilton, 1967; rollman, 1977; Walsh and rollman, 1979). Therefore, the large pores of solids such as b , Y and OFF favour the formation of these cyclic and condensed products. On the contrary, coke precursors can not develop in the restrictive voids of the shape selectivity catalyst, as in the case of zeolites having a MFI framework. This partially explains the insignificant deactivation observed for the MFI zeolite tested in this work. This result is in agreement with those of Hernandez et al. (1984).

Figure 1:
Time-on-stream behaviour of catalysts in methylcycloheaxane (MCHA) conversion in w-% at 500°C.

However, we should expect similar results for catalytic activity and selectivity to the MFI and FER zeolites. We assume that differences in the stability of catalytic activity on these zeolites are not only due to differences in their pore openings, but also to the three-dimensional crystalline framework of the MFI zeolite, and the one-dimensional crystalline framework of the FER zeolite. In contrast to the MFI zeolite, a small quantity of coke formed in the one-dimensional porous framework of the FER can obstruct the extremities of its channels, provoking its deactivation. Analogously, the significant activity lost for the large pore zeolites, W and MOR, should be related not only to their large pore openings, but also to their one-dimensional microporous system that can be easily blocked. Mirodatos and Barthomeuf (1985), Guisnet et al. (1987) and Jacquinot (1989) also verified this phenomenon of highly conspicuous deactivation, particular to one-dimensional porous framework zeolites.

About 200 reaction products were detected in the methylcyclohexane conversion reaction and in the chromatographic analyzis conditions used in the present work. C_i products selectivities, as well as product interval selectivity (from Ci to C_n), were calculated as shown by the following equations:

The results of cracking product selectivities (SC₁C₆), and selectivities for products having at least seven carbon atoms (SC₇C₈+), at methylcyclohexane conversion levels of 30%, 40% and 60% at 500°C, are presented in the Figures 2 and 3.

Figure 2: Selectivities for cracking products (SC₁C₆) at different values of methylcycloheaxane (MCHA) conversion in w-% at 500°C.

Figure 3: Selectivities for products having seven or more carbon atoms (SC₇C₈+) in methylcycloheaxane (MCHA) conversion in w-% at 500°C.

Comparison of the results for the same MCHA conversion levels demonstrates that for all zeolites tested, the cracking products are the main reaction products. Results depicted in Figure 2 show that there is an enhancement of selectivity for the C₁ to C₆ product fractions (SC₁C₆) in the following order: OMEGA < Y ~ BETA << MOR ~ OFF << FER ~ MFI. The SC₁C₆ selectivity increases as the size of the pore openings of the solids used in the reaction test decreases. By contrast with this result, a nearly linear relationship seems to exist between SC₇C₈+ selectivity (Figure 3) and size of pore openings of the zeolites. These data suggest that large pore zeolites, such as OMEGA, BETA and Y, favour the formation of reaction products having seven or more carbon atoms. It is worth mentioning that this behaviour was also observed for selectivity for isomerization products (dimethyl- and ethylcyclopentane).

Examining the selectivities for light hydrocarbons separately, one can notice that selectivities for C₂ and C₃ (Figures 4 and 5, respectively) can be grouped as a function of size of pore openings of the zeolites used.

As can be seen in Figures 4 and 5, low SC₂ and SC₃ selectivities were observed for the large pore zeolites, Y OMEGA and BETA. The OFF and MOR zeolites, with 12-membered rings of oxygen, as well as the zeolites mentioned above, which have smaller pore dimensions, presented slightly higher SC₂ and SC₃ selectivities. Finally, the MFI and FER zeolites lead to the highest SC₂ and SC₃ selectivities. On the other hand, selectivities for C₄ products (Figure 6) are generally higher in the presence of multidimensional zeolites having large pore openings (Mirodatos and Barthomeuf, 1985).

Figure 4: Selectivities for C₂ products (SC₂) at different values of methylcycloheaxane (MCHA) conversion in w-% at 500°C.

Figure 5: Selectivities for C₃ products (SC₃) at different values of methylcycloheaxane (MCHA) conversion in w-% at 500°C.

Figure 6: Selectivities for C₄ products (SC₄) at different values of methylcycloheaxane (MCHA) conversion in w-% at 500°C.

Figure 7: Selectivities for light olefins C₃⁼ plus C₄⁼ (SC₃⁼C₄⁼) at different values of methylcycloheaxane (MCHA) conversion in w-% at 500°C.

Selectivity data related to the light olefins, presented in Figure 7, show a very high C₃⁼ plus C₄⁼ (SC₃⁼C₄⁼) selectivity for the MFI zeolite. For the solids OMEGA and Y, selectivity for the same compounds is considerably inferior. The other samples occupy an intermediary position in relation to the zeolites mentioned. These facts seem to indicate that selectivity for olefins increases when intracrystalline zeolite voids are restricted.

These results suggest that, in methylcyclohexane cracking at 500°C, hydrogen transfer reactions preferentially occur on zeolites having large pore openings or dimension cavities (OMEGA, Y, BETA, etc.). Hydrogen transfer (HT) is a reaction consecutive to the primary cracking (Corma and Orchilles, 1989). This reaction leads to saturation of olefins produced during cracking. Therefore, the selectivity results concerning olefinic fractions such as C₃⁼ and C₄⁼ for a given zeolite are an indication of activity in hydrogen transfer.

Hydrogen transfer is a bimolecular reaction that occurs between olefins formed during the feed cracking (MCHA, in this case). Therefore, these reactions involve bulky reaction intermediates. Consequently, hydrogen transfer should be more limited in the presence of zeolites having microporous systems composed of restricted pore diameters. Thus, arrangement of bulky bimolecular reaction intermediaries should be more difficult in the narrow pores of the 10-membered rings of oxygen MFI zeolite than in the supercages of Y zeolite. The order of magnitude of the selectivity for olefins C₃⁼ plus C₄⁼ (SC₃⁼C₄⁼), shown in Figure 7, is in agreement with this explanation.

There is still another indirect measurement of hydrogen transfer (HT), olefinicity, which is the ratio between an olefinic fraction and the total quantity of hydrocarbons for the same product fraction (paraffins + olefins). In a general way, we can say that this ratio decreases as HT increases as the result of an increase in intracrystalline voids of the zeolites used. Figure 8, which presents C₄⁼/C₄t olefinicity, confirms this supposition.

Figure 8: C₄⁼/C₄t olefinicities at different values of methylcycloheaxane (MCHA) conversion in w-% at 500°C.

It can be noticed by the data presented in Figure 8 that the lowest C₄⁼/C₄t olefinicity values are found for the Y and OMEGA zeolites. The higher olefinicity values for the MFI zeolite compared to the other zeolites used, are most probably related to the poorer ability of the former solid to promote hydrogen transfer reactions. As already mentioned, it is probably due to their narrow pores which prevent the formation of the bulky bimolecular intermediaries typical of HT reactions. These results are in agreement with those concerning selectivity for olefins C₃⁼ plus C₄⁼, SC₃⁼C₄⁼ (Figure 7).

CONCLUSIONS

Experimental evidence indicates that the large pore zeolites, and particularly those having one-dimensional framework, were quickly deactivated during the methylcyclohexane cracking reaction at 500°C.

The results presented also suggest that selectivity for olefins C₃⁼ plus C₄⁼ (SC₃⁼C₄⁼) and the C₄⁼/C₄t olefinicity can be used to evaluate the zeolites’ tendency to promote hydrogen transfer reactions. The larger the pore dimensions and/or the microporous voids of these solids (as for the zeolites Y and OMEGA), the higher the probability that hydrogen transfer reactions (HT) will occur. Consequently, selectivity for light olefins C₃⁼ plus C₄⁼ (SC₃⁼C₄⁼) and C₄⁼/C₄t olefinicity should be lower.

NOMENCLATURE

FCC Fluid catalytic cracking

WHSV Weight hourly space velocity

MOR Zeolite Mordenite

FER Zeolite Ferrerite

OFF Zeolite Offretite

w.% Weight percentage

SC_i Selectivity for products having i carbon atoms

SC_iCn Selectivity for product fractions having from i to n carbon atoms

SC_i⁼ Selectivity for olefins having i carbon atoms

MCHA Methylcyclohexane

HT Hydrogen transfer

Greek letters

W Zeolite Omega

b Zeolite Beta

• Abbot, J. and Wojciechowski, B., J. Catal., 107, 571 (1987).
• Cheng, W. C. and Rajagopalan, K., J. Catal., 119, p. 354 (1989).
• Cheron, T., D.E.A repport., Institut Français du Pétrole (1993).
• Corma, A. and Orchilles, A. V., J. Catal., 115, p. 551 (1989).
• Corma, A., Mocholi, F., Orchilles, V., Koermer, G., Madon, R. J., Appl. Catal., 67, p. 307 (1991).
• Guisnet, M., Tejada, J. and Gnep, N., Bull. Soc. Chem. Fr., 1, p. 45 (1987).
• Hernandez, F., Moudafi, L., Fajula, F., Figueras, F., 8th Int. Cong. Catal., West Berlin, Proc. 2, p. 447 (1984).
• Hopkins, P. D., J. Catal., 12, p.325 (1968).
• Jaquinot, E., Ph. D. diss., Université Pierre et Marie Curie (Paris VI), 1989.
• Jaquinot, E., Mendes, A., Raatz, F., Marcilly, C., Ribeiro, F., Caeiro, J., Appl. Catal., 60, p.101 (1990).
• Ko, S. Y., Chang, A. N., Tien, S. K. ,Tung-hai Hsueh Pao, 26, p.621 (1985).
• Lin, L., Gnep, N., Guisnet, M. R., ACS, DPC, Preprints, 34(4), p.687 (1989).
• Mirodatos, C. and Barthomeuf, D., J. Catal., 93, p. 246 (1985).
• Mostad, H. B., Riis, T. U., Ellestad, O. H., Appl. Catal., 58, p. 105 (1990).
• Rollman, l. d., J. Catal., 47, p.113 (1977).
• Venuto, P. B. and Hamilton, L. A., Ind. Eng. Chem., Prod. Res. Dev., 6, p. 190 (1967).
• Walsh, D. H. and rollman, l. d., J. Catal., 56, p.139 (1979).
• Walsh, D. H. and rollman, l. d., J. Catal., 56, p.195 (1979).
• Wojciechowski, B. and. Corma, A., Catalytic Cracking, Marcel Dekker, New York (1986).

Publication Dates

• Publication in this collection
  09 Oct 1998
• Date of issue
  June 1998

History

• Accepted
  13 Apr 1998
• Received
  05 Nov 1997