Effectiveness of a partially wetted catalyst for bimolecular reaction kinetics

Abstract

A large number of studies on the effectiveness of a partially wetted catalyst particle have appeared in the literature (for example, Mills and Dudukovic, 1979;Tan and Smith, 1980;Herskowitz, 1981;Goto et al., 1981;Sakornwimon and Sylvester, 1982;Capra et al., 1982;Ring and Missen, 1986). With the exception of work by Goto et al., all these earlier studies considered only a bimolecular reaction with pseudofirst-order kinetics; i.e., first order with respect to the limiting reactant and zero order for the excess reactant. Goto et al. used pseudo-nth-order kinetics. Recently, Beaudry et al. (1987) and Harold and Ng (1 987) demonstrated that pseudofirst-order kinetics can be misleading in that depletion of the supposedly abundant (zeroorder) reactant may occur within the pellet. Also recently, Yentekakis and Vayenas (1987) suggested that bimolecular kinetics, first order with respect to each reactant, is essential for hydrodesulfurization processes.Since zero-order kinetics in actuality loses its physical meaning when the zero-order reactant is no longer in excess and other kinetic expressions are more appropriate for some commercial processes, the objective of this work is to examine the behavior of a single reaction that is first order with respect to both gaseous and liquid reactants. Emphasis is placed on elucidating the interplay between internal diffusion, reaction, and external mass transfer on the partially wetted surface, and to examine the impact of the interplay on catalyst effectiveness. Model DevelopmentA two-dimensional model is developed to describe the isothermal, irreversible reaction between a dissolved gas reactant A and a nonvolatile liquid reactant B within a partially wetted, uniformly active porous catalytic pellet. As illustrated in Figure 1, the pellet has a square cross section with each side of length S and is externally wetted about the four corners. Each of the four identical liquid films is of a uniform depth and covers an equal Correspondence concerning this paper shwld be addrcsscd to K. M. Ng or M. P. Harold. length in both directions from the corresponding corner. The desired wetting efficiency is obtained by adjusting this length from zero to its maximum value of S / 2 , which corresponds to complete wetting. Clearly, this fixed wetting pattern is idealized, as other flow features are present in a trickle-bed reactor (Ng and Chu, 1987). The liquid phase contains the reactants, reaction product(s), and possibly an inert solvent. A gaseous environment, with a constant partial pressure of reactant A, pbA, completely surrounds the wetted pellet.The liquid phase catalytic reaction is given by + uB(,, -Products,,, where g and 1 represent gas and liquid, respectively. The intrinsic reaction rate r is assumed to be first order with respect to A and B; i.e., r = k,CACB Species A and B respectively satisfy the following dimensionless diffusion-reaction equations:The dimensionless variables in Eqs. 2 and 3 are defined as follows:

Experimental study of reaction in a partially wetted catalytic pellet

“…The maximum rate occurs at such low wetting efficiencies in the pure AMS experiments because of the large excess of AMS in the liquid phase. These features confirm the model predictions of Harold and Ng (1987) and Funk et al (1988Funk et al ( , 1989).…”

“…As the flow rate is increased from this point, the rate declines monotonically for the two hydrogen gas experiments. These data clearly demonstrate the effectiveness enhancement by partial wetting phenomenon (Harold and Ng, 1987;Funk et al, 1988Funk et al, , 1989. In effect, by covering more of the surface with the flowing liquid, the supply rate of the limiting reactant hydrogen is reduced, since the supply of hydrogen into the liquid-filled pores is deterred by the flowing liquid film.…”

“…For reasons discussed earlier, these models are incapable of predicting an overall reaction rate maximum at an intermediate liquid flow rate. This led to the development of models with more realistic bimolecular rate expressions such as a first-order dependence on both the gas reactant and the liquid reactant (Gabitto et al, 1986;Yentekakis and Vayenas, 1987;Funk et al, 1988). Funk et al showed that under certain conditions there may be regions within the catalyst with different limiting reactants and that both interparticle and intraparticle mass transport can be rate-limiting in these regions.…”

“…The novel design simulates the local environment within a trickle-bed reactor, and permits the direct measurement of the degree of wetting under reaction conditions. Experimental data of the hydrogenation of a-methylstyrene ( A M S ) over a Pd/A1203 pellet provide unequivocal evidence of effectiveness enhancement by partial wetting, a predicted single-pellet phenomenon (e.g., Harold and Ng, 1987;Funk et al, 1988Funk et al, , 1989. The overall reaction rate is an increasing (decreasing) function of the degree qf wetting i f the overall limiting reactant is more effectively supplied through the wetled (nonwetted) part of the surface.…”

Solving linear diffusion‐reaction networks in porous catalyst particles using BEM

“…One such case occurs in a trickle-bed reactor when the external surface of the catalyst is only partly covered by the flowing liquid (rivulet). This phenomenon is called the partial wetting effect and has been the subject of active research for over a decade (Ramachandran and Smith, 1979;Mills and Dudukovii, 1979;Yentekakis and Vyenas, 1987;Funk et al, 1988;Harold, 1993, among others). Partial wetting Ieads to particle scale nonuniformitics that have an important bearing on the reactor performance.…”