THE EFFECT OF HYDROGEN ON COKE REMOVAL FROM Ga/HZMS5 CATALYSTS

The Effect of hydrogen on coke removal from Ga/HZMS5 catalysts

M.G.F. Rodrigues ¹ , P. Magnoux ² , M. Guisnet ² and V.R. Choudhary ³

¹Universidade Federal da Paraíba, Departamento de Engenharia Química, avenida Aprígio Veloso, 882

Bodocongó, 58.109-000 Campina Grande - Paraíba, Brazil.

²URA CNRS 350, Catalyse en Chimie Organique, Université de Poitiers, 40, avenue du Recteur Pineau,

86022 Poitiers, France.

³National Chemical Laboratory, Chemical Engineering Division, 411.008 Pune, India.

(Received: March 5, 1997; Accepted: August 5, 1997)

Abstract:The presence of gallium on HZMS5 zeolites increases the rate of coke elimination in their regeneration in hydrogen. The gallium species should contribute to the hydrogenation of the coke molecules and the acid sites should favour their crackings. However, the location of the coke molecules (near the outer surface and near the gallium species) instead of their nature seems to be the determining parameter for their elimination.

Keywords: Ga/HZSM5 catalysts, coke, regeneration in hydrogen.

Introduction

The deposit of carbonaceous compounds (coke) inside the pores or on the outer surface of the zeolites is the main cause of their deactivation during the transformation of organic reactants (Rollmann and Walsh, 1982; Derouane, 1985; Bhatia et al., 1990; Guisnet and Magnoux, 1990; Guisnet and Magnoux, 1992). The cost of this deactivation is very high and great efforts have been and are being made to find methods for, i) limiting the formation of coke and its effects on zeolite activity, and ii) regenerating the activity. Hydrogen can limit the deactivation and the rate of coke formation on acid catalysts (Bearez et al., 1983; Gnep and Guisnet, 1981; Chen and Garwood, 1977); this is obviously the case when there is a change from an acid to a bifunctional mechanism. Coke formation and deactivation are very rapid during alkane cracking on acid zeolites but very slow during hydrocracking on Pt acid zeolites (Kouwenhoven, 1973; Guisnet and Perot, 1984). The bifunctional Ga/HZSM5 catalysts are active and selective for the aromatization of lower paraffins (Ono, 1992; Guisnet and Gnep, 1992); nevertheless, the reaction of aromatization is accompanied by coke formation which causes the deactivation of these catalysts (Meriaudeau and Naccache, 1991; Abdul Hamid et al., 1994; Chang et al., 1995). Coke is generally removed from the catalysts by oxidative treatment in air flow. Nevertheless, structural alteration of the catalyst may result from its being exposed to steam produced by the burning of coke during oxidative regeneration of the deactivated catalyst (Rollmann and Walsh, 1982). On Ga/HZSM5, dealumination and degalliation of the catalyst can be observed during oxidative regeneration of the aged catalysts (Abdul Hamid et al., 1994; Chang et al., 1995; Abdul Hamid, Yarne et al., 1994). In order to limit the effect of oxidative regeneration on the catalyst properties, it is useful to remove the coke with another treatment such as hydrogen treatment. In fact, the aromatization of propane, or other light paraffins, is accompanied by hydrogen production and gallium species are able to catalyze ethene hydrogenation into ethane (Lukyanov et al., 1994). On the other hand, the aromatization of propane increases when the reaction is carried out in hydrogen due to the inhibition effect of the hydrogen on coke formation (Rodrigues, 1996). The aim of this paper is to demonstrate that coke deposited on the Ga/HZSM5 catalyst during the aromatization of propene can be removed by hydrogen treatment. We'll also examine the effect of the location of coke on its elimination.

Experimental methods

The HZSM5 zeolite (Si/Al = 40) was synthesized according to the method of Guth and Caullet (Chen and Garwood, 1977) and the Ga/HZSM5 catalyst (2 wt % Ga) was prepared by impregnation of this zeolite with Ga(NO₃)₃. Prior to use, the HZSM5 zeolite was pretreated for 10 hours at 530°C in dry nitrogen flow and the Ga/HZSM5 catalysts were treated for 10 hours at 600°C in dry hydrogen flow.

Coke formation resulting from propene aromatization was carried out under the following conditions: T = 350°C and 530°C, P_propene = 1 bar, WHSV (Weight of reactant injected per weight of catalyst per hour) = 1.7 h^-1. After coking, catalysts were treated for 12 hours at 600°C in dry hydrogen flow (P_H2 = 1 bar). The experimental methods used to recover and analyze the coke components have already been described (Gnep and Guisnet, 1981): dissolution of the zeolite in a 40% hydrofluoric acid solution followed by extraction with methylene chloride and analysis of the extracts by Gas Chromatography (GC) and Gas Chromatography coupled with Mass Spectrometry (GC/MS). The nonsoluble coke in methylene chloride was characterized by its H/C atomic ratio.

Results and discussion

HZSM5 catalysts were coked under propene transformation at 350°C and 530°C in order to obtain 3.3 and 2.8 wt % coke, respectively. The Ga/HZSM5 catalyst was coked under propene transformation at 350°C (3.7 and 8.9 wt % coke) and at 530°C (3 wt % C) and also under propane aromatization at 530°C (3 wt % C).

Coke composition was established on these catalysts. At 350°C and for low coke content, coke was already soluble in the methylene chloride for all samples. A very low amount of insoluble coke was found in the methylene chloride (< 10 %) for the sample containing 8.9 wt % C. The major compounds of soluble coke were the alkylnaphthalenes (80 %), accompanied by alkylfluorenes, alkylanthracenes and/or akylphenanthrenes. These compounds, due to their size, were retained inside the zeolite pores (trapping). The mode of formation of these compounds is the classical one which involves a succession of alkylation, cyclization and hydrogen transfer (and dehydrogenation with Ga/HZSM5) steps.

At 530°C, the coke was very polyaromatic and insoluble in methylene chloride. The H/C ratio of these molecules was close to 0.45 and practically independent of the catalysts. Insoluble coke, located on the outer surface of the zeolite, resulted from the secondary transformation of coke molecules trapped at channel intersections close to the outer surface of the zeolite crystallites with overflow onto this surface (Guisnet and Perot, 1984). A second mode of formation of insoluble coke was the dehydrogenative coupling of aromatic molecules trapped in the zeolite pore (Ono, 1992). These very polyaromatic molecules were most likely located inside the linear channels of the HZSM5 zeolite. In the HZSM5 zeolite, whatever the temperature required for coke formation, only 15 to 20 % of the coke could be removed by hydrogen treatment. In the Ga/HZSM5 coked at 350°C, whatever the coke content (3.7 or 8.9 wt %), 65 % was removed in hydrogen; on the other hand, 90 % could be eliminated from the Ga/HZSM5 coked at 530°C. The activity of the catalyst could be recovered after this treatment.

Figure 1: Total transformation of propane and transformation into aromatics at 530°C in the Ga/HZSM5. Effect of hydrogen treatment.

Total and aromatization activities could be recovered on a Ga/HZSM5 sample coked with propane (3 wt %) at 530°C and treated in hydrogen during 12 hours at 600°C (Figure 1). Coke content after hydrogen treatment was close to 0.3 wt %.

From the sample Ga/HZSM5 coked from propane transformation at 350°C (8.9 wt %), coke was analyzed before and after the hydrogen treatment. Initially, it was made up of 20 % insoluble coke and 80 % bi and tri-aromatic compounds. After hydrogen treatment, the coke content was close to 3 % and only very polyaromatic compounds, insoluble in methylene chloride, were found.

In the Ga/HZSM5 catalysts, the coke elimination in hydrogen can be schematized as follows in Figure 2.

The first route (1) consists of partially hydrogenating the coke molecules on the gallium species, then cracking these hydrogenated compounds on acid sites.

The second (2) is a parallel route which consists of polycondensating the coke molecules on acid sites by dehydrogenative coupling, or by successive reactions of alkylation, cyclization, and hydrogen transfer (or dehydrogenation) between a coke molecule and the cracking products formed in the first route (1). This second route leads to the formation of very polyaromatic molecules which are insoluble in methylene chloride.

The desorption of the cracking products from the pores of the zeolite could be the limiting step in the process of coke elimination in hydrogen treatment and the coke location rather than its nature seemed to be an important parameter for its elimination.

Thus at 350°C, the coke molecules were preferentially located in the zeolite pores and although their degree of aromaticity was lower than that of the coke molecules formed at 530°C (Rodrigues et al., 1996), their elimination in hydrogen treatment was more difficult.

Various hypotheses can be advanced:

i) the gallium species dispersion inside the HZSM5 zeolite is low and coke hydrogenation is incomplete. This favours the formation of very polyaromatic molecules (route 2).

ii) the desorption of the cracked products is slow and the probability of their encountering a coke molecule and an acid site is high. This also favours the formation of very polyaromatic compounds and, thus, the second route.

On the other hand, at 530°C, the desorption of the cracked products is much faster owing to the localization of the coke (near or on the outer surface of the zeolite crystallites). The probability that these products encounter a coke molecule and an acid site becomes lower, and favours the first route (1). But, it's easy to imagine that the gallium species, active for hydrogenating the coke, are located on the surface of the zeolite crystallites or in the pores close to the surface.

Figure 2: Ga/HZSM5 catalysts.

Conclusion

With hydrogen treatment at 600°C, the gallium species deposited on the HZSM5 zeolite seem to be able to dehydrogenate the aromatic and polyaromatic molecules which form the coke. It is then possible to eliminate the coke. On these catalysts, the location of the gallium species and of coke, instead of their composition and nature, seems to be the determining parameter for coke elimination in hydrogen treatment. The very polyaromatic molecules, formed at 530°C by propene transformation on Ga/HZSM5 and located on the outer surface of the zeolite can be eliminated more easily by treatment of the catalyst in hydrogen than the molecules are less polyaromatic and located in the pores of the zeolite. This elimination is related to the presence of the gallium species near the coke molecules and to a fast desorption of the cracked products formed during coke elimination.

Acknowledgments

This work was carried out with the financial support of the Indo-French Center for the Promotion of Advanced Research / Centre Franco-Indien pour la Promotion de la Recherche Avancée. M.G.F. Rodrigues thanks the CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico) for the financial support provided during her Ph.D. research.

References

Abdul Hamid, S.B.; Derouane, E.G.; Meriaudeau, P. and Naccache, C., Studies in Surface Science and Catalysis, 88, 185 (1994).

Abdul Hamid, S.B.; Derouane, E.G.; Demortier, G.; Riga, J. and Yarne, M.A., Appl. Catal. A: General, 108, 85 (1994).

Bearez, C.; Chevalier, F. and Guisnet, M., React. Kinet. Catal. Lett., 22(3-4), 45 (1983).

Bhatias, S.; Beltramini, J. and Do, D.D., Catal. Rev. Sci. Eng., 31(4), 431 (1990).

Chang, C.; Tan, C. and Zhuo, R., 21th National Meeting, Advanced Catalysis Series (ACS), Washington, 40, 581 (1995).

Chen, N.Y. and Garwood, W.E., J. Catal., 50, 252 (1977).

Derouane, E.G., Catalysis by Acids and Bases, Studies in Surface Sciences and Catalysts, B. Imelik, N. Naccache, G. Coudurier, Y. Ben Taarit and J.C. Vedrine (Eds), Elsevier, Amsterdam, vol. 20, p.221 (1985).

Gnep, N.S. and Guisnet, M., Appl. Catal., 1, 329 (1981).

Guisnet, M. and Gnep, N.S., Appl. Catal. A: General, 89, 1 (1992).

Guisnet, M. and Magnoux, P., Appl. Catal., 54, 431 (1990).

Guisnet, M. and Magnoux, P., Zeolite Microporous Solids: Synthesis, Structure and Reactivity, NATO ASI Series E, E.G. Derouane, F. Lemos, C. Naccache and F.R. Ribeiro (Eds.), Kluwer Academic Publishers, Dordrec, 352, p.437 (1992).

Guisnet, M. and Perot, G., Zeolite Science and Technology, NATO ASI Series E, F.R. Ribeiro et al. (Eds.), Martinus Nighoff Publishers, The Hague, p.397 (1984).

Kouwenhoven, H.W., Molecular Sieves, W.N. Meier and J.B. Vytterhoeven (Eds.), Advanced Chemical Series (ACS), Washington, 121, p.529 (1973).

Lukyanov, D.B.; Gnep, N.S. and Guisnet, M., Ind. Eng. Chem. Res., 34, 516 (1994).

Meriaudeau, P. and Naccache, C., Catalysis Deactivation, Ch. Barthomew and J. Butt (Eds.), Studies in Surface Science and Catalysis, vol. 68, 767 (1991).

Ono, Y., Catalysis Review - Science Engineering, 34(3), 179 (1992).

Rodrigues, M.G.F., Ph.D. diss., University of Poitiers, France (1996).

Rodrigues, M.G.F.; Guisnet, M.; Magnoux, P.; Barre, M. and Choudhary, V.R., J. Chim. Phys., 93, 337 (1996).

Rollmann, L.D. and Walsh, D.E., Progress in Catalyst Deactivation, NATO ASI Series E, J.L. Figuereido (Ed.), Martinus Nighoff Publishers, The Hague, vol. 54, p.81 (1982).

Publication Dates

• Publication in this collection
  09 Oct 1998
• Date of issue
  Sept 1997

History

• Accepted
  05 Aug 1997
• Received
  05 Mar 1997