The catalytic hydrogenation of an aromatic sulfonyl chloride was investigated in continuous and semi-batch mode processes using a Robinson-Mahoney stationary basket reactor. A complete experimental unit was designed and built. The operating and analytical procedures have been developed and the methodologies to gather the kinetic data have been described. Hydrogenation reactions were conducted at a reaction pressure of 364.7 psia, at three different reaction temperatures: 85°C, 97°C and 110°C, at five different residence times: 0.6 (only at 110°C), 1.0, 1.5, 2.0, 3.1 hr, with the hydrogen to the aromatic sulfonyl chloride molar ratio: 8.0 mol/mol and hydrogen to argon molar ratio: 3.0 mol/mol. Intrinsic reaction rates of the reacting species were obtained on the surface of a commercial 1 wt% palladium on charcoal catalyst.

The conversion and molar yield profiles of the reacting species with respect to process time suggest a deactivation of the 1 wt% palladium on charcoal catalyst. Kinetic data collected in a continuous process mode show that the catalyst is deactivated during an experiment when the process time equal to two to three times the residence time of the liquid within the reactor. XRD analysis shows that the active sites are blocked and an amorphous layer was formed on the surface of the palladium catalyst. Semi-Batch mode experimental data were obtained at 110°C after 8 hours of reaction time for several aromatic sulfonyl chlorides.

A kinetic model has been developed, which includes adsorption of individual components and surface reactions as well as rate equations of the Hougen-Watson type. A hyperbolic deactivation function expressed in term of process time is implemented in the Hougen-Watson equation rates. The mathematical model consists of non-linear and simultaneous differential equations with multiple variables. The kinetic parameters were estimated from the minimization of a multi-response objective function by means of a sequential quadratic program, which includes a quasi-Newton algorithm. The statistical analysis was based on the t- and F-tests and the simulated results were compared to the experimental data.