No abstract

Abstract:Coupling the dehydrogenation of ethylbenzene to styrene with the hydrogenation of nitrobenzene to aniline in a catalytic fixed bed membrane reactor has the potential for significantly improving both processes (Abo-Ghander et al. 2008). In a continuing effort to realize this potential, an optimal design is sought for a co-current flow, catalytic membrane reactor configuration. To achieve this objective, two conflicting objective functions, namely: the yield of styrene on the dehydrogenation side and the conversion of nitrobenzene on the hydrogenation side, have been considered. The total number of the decision variables considered in the optimization problem is twelve, representing a set of operational and dimensional parameters. The problem has been solved numerically by two deterministic multi-objective optimization approaches: the normalized normal constraint method and the normal boundary intersection method. It was found that the integrating reactor can be run to produce a maximum styrene yield of 97% when production of styrene is emphasized and a maximum of 80% of nitrobenzene conversion when nitrobenzene conversion is emphasized. The resulting sets of Pareto optimal solutions obtained by both techniques have been found to be identical. In addition, * Corresponding

The use of Delplots to deduce the key features of reaction networks using nonisothermal kinetics data was examined. Using Delplots, a product’s network rank, i.e., the number of reaction steps required for its formation from a specified reactant “A,” is generally obtained by extrapolating plots of yi /x A r vs x A to x A = 0, where at isothermal conditions, contact time was varied to provide the range of conversion supporting the extrapolation. The presently described work addressed the common experimentalists’ technique of using temperature, rather than contact time, to provide the range of conversion. To assess any uncertainties thus introduced, the effect of changing the temperature of kinetic measurements has been addressed for the parallel-series reaction network ; with B 0 = 0. The relative activation energies of the key reactions were varied by ±6 kcal/mol with respect to that for k 1, and temperature was varied between 200 and 1000 K. The resulting Delplot information can appear to suggest different reaction networks if the activation energy difference is too large and the temperature range too wide. The Delplot method classifies species B to be a primary product at low temperature when E 2 > E 1, while it appears to be a secondary product when E 2 < E 1. We suggest, as rough guidelines, that varying temperature to provide variations in conversion in the kinetic study is reasonable for E 1 ∼ 50 kcal/mol if the activation energy difference E 21 is in between 3 and −3 kcal/mol.

This document contains the post-print pdf-version of the refereed paper:"Heterogeneous modeling of an autothermal membrane reactor coupling dehydrogenation of ethylbenzene to styrene with hydrogenation of nitrobenzene to aniline: Fickian diffusion model."

No abstract