(683f) Intracrystalline Diffusion Effects in n-Alkane Hydrocracking Over Pt/H-ZSM5

Heterogeneously catalyzed processes are not solely reaction-controlled when the size of the reacting species approaches the pore dimensions of the catalytic material. Microscopic diffusion effects comprise the constrained transport of physisorbed species throughout the catalyst crystallite and give rise to peculiar product selectivities governed by the fastest diffusing species. Alkane hydrocracking is often applied as a model reaction to investigate such catalytic phenomena thanks to its well-defined reaction mechanism consisting of only a limited number of elementary reaction families (Baltanas, 1989; Thybaut, 2004; Narasimhan, 2003). After physisorption in the catalyst micropores, alkanes are dehydrogenated on a metal phase towards unsaturated alkenes which, in turn, are protonated on an acidic function yielding alkylcarbenium ions. The latter are reactive towards skeleton rearrangements and carbon-carbon bond scissions. Especially tribranched species are highly susceptible to cracking, specifically towards branched products, limiting their contribution to the eventual product distribution. Zeolites are commonly applied as active material thanks to their high hydrothermal stability, high internal surface and the possibility to easily accommodate various chemical functions, e.g., metallic and protonic functions as required for hydrocracking. A well-known medium-pore zeolite topology in hydrocracking processes is ZSM5 with a framework containing straight and zigzag channels with relatively large intersections. Its pore dimensions approach the kinetic diameters of linear and branched alkanes which ultimately induces a product distribution that is substantially different from that obtained over large-pore zeolites such as USY (Jacobs, 1981; Martens, 1991). The aim of this work is to develop a fundamental single-event microkinetic model which is able to accurately describe simultaneous reaction and diffusion as observed over a ZSM5 zeolite catalyst. Even more important than the extension of the current model for n-alkane hydrocracking towards another zeolite framework, is the development of a general methodology for the diffusion process through zeolitic frameworks which, hence, is not restricted to ZSM5 only.

Due to its relatively small pores, alkane hydrocracking over a ZSM5 catalyst is both reaction- and diffusion-controlled and, hence, requires the calculation of the intracrystalline concentration profile of each component present in the product mixture. Transport diffusion through ZSM5 is most fundamentally described by use of the self-diffusion coefficients of the individual components at low coverages. The concentration dependence of the diffusion coefficient is implemented by identifying different physisorption sites in a ZSM5 unit cell (Coppens, 1999). Pure component diffusion within a unit cell is consequently modeled as activated jumps of the adsorbate from one site to another. Diffusion in a multi-component mixture was approximated via the Stefan-Maxwell formulation which, in essence, considers the inter-species diffusivity as a measure for the inverse of the drag force between both components (Krishna, 1997). The latter balances the driving force for transport of each component through the multi-component mixture and is quantified by use of the chemical potential gradient.

The reaction-diffusion process in ZSM5 was first investigated by n-hexane hydrocracking as its limited reaction network rules out excessive cracking of dibranched isomers and the formation of tribranched species. The effect of product selectivity through differences in diffusion coefficients between the hexane isomers was observed in the peculiar monobranched isomer selectivities which lied even more in favor of 2-methylhexane. The single-event microkinetic model accounting for simultaneous diffusion and reaction as described above, accurately reproduced the experimentally observed selectivities to both 2-methyl and 3-methylhexane through assigning different diffusion coefficients for both components. The slower diffusion of 3-methylhexane through the catalyst crystallite induces a higher reactivity of the isomer compared to 2-methylhexane. The negligible yield of dibranched hexanes confirmed the strongly hindered mobility of these species throughout the crystallite.

Extension towards heavier feed molecules, i.e. n-octane and n-decane, was performed in order to incorporate tribranched species formation which, according to earlier works, is possibly limited due to transition state shape selectivity effects [ref jac mart]. In addition, dibranched hydrocarbons are no longer restricted to isomerization reactions only, as in the case of n-hexane conversion, but can also give rise to cracking towards lighter products. A commercially available ammonia-exchanged ZSM5 zeolite with a Si-Al ratio of 80 and a BET surface of approximately 4.25 10⁵ m² kg^-1, was transformed in the protonic form and provided afterwards with 0.67 wt% Pt via wetness impregnation with an aqueous solution of hexachloroplatinic acid. A total amount of 9.24 10^-3 kg catalyst was loaded in a Berty-type lab-scale CSTR. An isothermal profile and ideally mixing of the reactant gases with the catalyst was guaranteed by a magnetically driven impeller. The hydrocracking reaction temperature was varied from 463 K to 513 K while the reactor pressure ranged from 1 to 2 MPa. The n-alkane space time amounted up to 750 kg s mol^-1. The hydrogen-to-hydrocarbon molar ratio was situated between 50 and 150 ensuring gas-phase conditions at any point in the kinetic setup.

Experimental results showed low maximum isomerization yields of approximately 10 % for both n-octane and n-decane confirming the reputation of ZSM5 as a cracking catalyst (Martens, 1991). The formation of tribranched species was not evident from the experimental analyses as well as from the cracking product distribution which showed almost equal yields for both linear and branched products. Tribranched carbenium ions are highly reactive towards cracking resulting in significantly higher yields of iso-butane and iso-pentane during respectively n-octane and n-decane hydrocracking as observed over USY catalysts (Jacobs, 1981; Thybaut, 2004). The formation of lighter iso-alkanes was attributed to the cracking of dibranched species which, considering a kinetic diameter which is even slightly higher than the measured pore diameter of a ZSM5, are nearly trapped inside the zeolite framework. Corresponding low yields amounting up to 4 % and the peculiar cracking product distribution observed from experimentation, confirmed the strongly reduced mobility of dibranched hydrocarbons which consequently are limited to debranching and cracking. Intracrystalline diffusion effects were also observed in the monomethyl isomer product distribution which shows the highest yield for 2-methylheptane and 2-methylnonane, while a favored production of the 3-methylalkane was predicted from thermodynamic equilibrium calculations and was confirmed by n-octane hydrocracking experiments (Thybaut, 2004).The higher selectivity towards the 2-methylalkane could only be explained through a higher mobility compared to the other methyl isomers, and confirms the trend of the diffusivity which decreases as the branch is located more to the center of the carbon chain of the molecule. Consequently, an accurate description of simultaneous reaction and diffusion during n-octane and n-decane hydrocracking over ZSM5 urges on a careful selection of the diffusion parameters of each component in the reaction network.

References

Baltanas, M. A., K. K. Van Raemdonck, G. F. Froment, and S. R. Mohedas, “Fundamental Kinetic Modeling of Hydroisomerization and Hydrocracking on Noble-Metal-Loaded Faujasites. 1. Rate Parameters for Hydroisomerization”, Ind. Eng. Chem. Res., 28(7), 899 (1989).

Coppens, M., A. T. Bell, and A. K. Chakraborty, “Dynamic Monte-Carlo and Mean-Field Study of the Effect of Strong Adsorption Sites on Self-Diffusion in Zeolites”, Chem. Eng. Sci., 54(15-16), 3455 (1999).

Jacobs, P.A., J. A. Martens, J. Weitkamp, and H. K. Beyer, “Shape-selectivity Changes in High-silica Zeolites”, Faraday Discuss. 72, 353 (1981).

Krishna, R., and J. A. Wesselingh, “The Maxwell-Stefan Approach to Mass Transfer”, Chem. Eng. Sci., 52(6), 861 (1997).

Martens, J. A., R. Parton, L. Uytterhoeven, and P. A. Jacobs, “Selective Conversion of Decane into Branched Isomers. A Comparison of Platinum/ZSM-22, Platinum/ZSM-5 and Platinum/USY Zeolite Catalysts”, Appl. Catal., 76(1), 95 (1991).

Narasimhan, C. S. L., J. W. Thybaut, G. B. Marin, P. A. Jacobs, J. A. Martens, J. F. M. Denayer, and G. V. Baron, “Kinetic Modeling of Pore-mouth Catalysisin the Hydroconversion of n-Octane on Pt-H-ZSM-22”, J. Catal., 220(2), 399 (2003).

Thybaut, J. W., C. S. L. Narasimhan, G. B. Marin, J. F. M. Denayer, G. V. Baron, P. A. Jacobs, and J. A. Martens, “Alkylcarbenium Ion Concentrations in Zeolite Pores during Octane Hydrocracking on Pt/H-USY Zeolite”, Catal. Lett., 94(1-2), 81 (2004).