(305e) Single-Event Microkinetics of Aromatics Hydrogenation on a Pt Catalyst

A fundamental single-event microkinetic (SEMK) model for the hydrogenation of aromatic components on a Pt catalyst has been developed. It is based on the Horiuti-Polanyi mechanism considering atomic hydrogen addition steps to the (partially hydrogenated) aromatic species on the catalyst surface. The reaction network used accounts for the position at which the hydrogen atoms are added to the ring. Hence, the SEMK model for aromatics hydrogenation can provide much more insight in the surface concentrations of the reactive intermediates than other models assuming the existence of a most abundant surface intermediate or not accounting for the position at which the hydrogen atoms are added. Quantumchemical calculations have provided guidelines for kinetic parameter definition (Saeys et al. 2005). The current model validation with benzene hydrogenation data illustrates the model's adequacy. Originally, the SEMK methodology has been applied to thermal and acid catalyzed reactions. Recently the methodology has been extended to metal catalyzed reactions using Fischer Tropsch synthesis as example reaction. The current work demonstrates the adequacy of the SEMK methodology for the hydrogenation of aromatic components on a noble metal such as Pt. The model is being validated for benzene hydrogenation and is ready for validation with any other monoaromatic component.

The reaction network is generated using a computerized algorithm. A new type of species has been defined as ?metalrings' and represents the (partially hydrogenated) aromatic species on the catalyst surface. While for benzene the reaction network is limited to 13 metalrings and 20 H-additions and abstractions, vide Figure 1, its size increases to 40 metalrings and 104 H-additions and abstractions for toluene. A characteristic of aromatic hydrogenation kinetic modeling is that many elementary steps occur in the formation of the cycloalkane out of the aromatic component. In steady-state experimentation no direct information is obtained on these reactive intermediates. In order to construct a kinetic model with a reasonable number of adjustable parameters, information with respect to the surface intermediates was obtained from quantumchemical calculations. Activation energies and surface reaction enthalpies for benzene hydrogenation have been calculated using DFT methods (Saeys et al., 2005). The number of unsaturated nearest carbon atoms to the carbon atom to which the H-atom is being added was found to be the distinctive feature between the values obtained for the activation energies and reaction enthalpies of the various H-atom additions. In addition the activation energy and reaction enthalpy of H-addition are also assumed to depend on the secondary or tertiary character of the carbon atom involved. No rate-determining step is assumed. All H-atom additions and abstractions are considered to be non quasi equilibrated. Reactant chemisorption and product desorption was assumed to be quasi equilibrated. The total number of parameters in the model for benzene hydrogenation amounts to 9, i.e., 3 rate coefficients and 3 surface reaction equilibrium coefficients and chemisorption equilibrium coefficients for the aromatic component, the cycloalkane and hydrogen. When accounting for their temperature dependence this number doubles to 18, however, order of magnitude calculations have been performed for the preexponential factors of rate, surface reaction equilibrium and chemisorption coefficients. These calculations were based on reasonable assumptions on the mobilities of the species involved in the elementary step considered (Thybaut et al., 2002). E.g., benzene chemisorption was assumed to lead to a significant loss of translational freedom, while the hydrogen mobility on the catalyst surface was assumed to be high. An advanced reparameterization technique during the estimation of the remaining adjustable parameters allowed the assessment of these assumptions regarding the species' mobilities. Thermodynamic constraints allow calculating one of the surface reaction enthalpies and one of the chemisorption enthalpies from the remaining parameter values and the overall reaction enthalpy. Hence 7 adjustable parameters remain to be estimated from regression to experimental data. Kinetic parameter estimation leads to values which are in agreement with those reported in the literature (Lin and Vannice, 1993; Thybaut et al., 2002), i.e., activation energies in the range of 60 to 70 kJ mol-1 and surface reaction enthalpies close to 0 kJ mol-1. The benzene chemisorption enthalpy amounted to -67 kJ mol-1, while that of hydrogen amounted to -47 kJ mol-1. The model adequacy is illustrated in the parity diagram for the methylcyclohexane outlet flow rate and in the simulation of the temperature effect on the benzene hydrogenation at various total pressures. The peculiar temperature effect is related to the evolution of the surface concentrations with the temperature. Benzene and other hydrocarbon concentrations are high, i.e., occupying more than half of the active sites at lower temperatures but steadily decrease at higher temperatures. Roughly one quarter of the active sites is taken by hydrogen while the remaining fraction is free.

references Saeys, M., Reyniers, M.-F., Neurock, M. and Marin, G.B. Ab initio reaction path analysis of benzene hydrogenation to cyclohexane on Pt(111) J. Phys. Chem. B 109(6) 2064-2073 (2005) Thybaut, J.W., Saeys, M. and Marin, G.B. Hydrogenation kinetics of toluene on Pt/ZSM-22 Chem. Eng. J. 90(1-2) 117-129 (2002) Lin, S.D. and Vannice, M.A., Hydrogenation of aromatic hydrocarbons over supported Pt catalysts .3. reaction models for metal surfaces and acidic sites on oxide supports J. Catal. 143(2) 563-572 (1993).