(478b) Microkinetic Modelling of Toluene and o-Xylene Hydrogenation On a Pt Catalyst | AIChE

(478b) Microkinetic Modelling of Toluene and o-Xylene Hydrogenation On a Pt Catalyst


Thybaut, J. W. - Presenter, Ghent University
Bera, T. - Presenter, Ghent University

The Single-Event MicroKinetic
(SEMK) methodology, which had been successfully applied to benzene
hydrogenation [1], has now been extended towards substituted aromatics, i.e.,
toluene and o-xylene. In addition to the number of unsaturated nearest
neighbour carbon atoms, H-atom addition rate coefficients were assumed to
depend on the carbon atom type, i.e., secondary or tertiary. A simultaneous
regression of the SEMK model to an experimental data set consisting of 39
toluene and 37 o-xylene hydrogenation experiments measured
at temperatures in the range from 423 K to 498 K, aromatic inlet partial
pressures in the range from 10 kPa to 60 kPa and hydrogen inlet partial
pressures from 100 kPa to 600 kPa on Pt catalyst resulted in activation energies of
H-additions to tertiary carbon atoms that are 11 kJ mol-1 higher
than to secondary carbon atoms. This can be related to the steric hindrance
experienced during H addition to a carbon atom bearing a substituent.

In steady-state experimentation of aromatic
hydrogenation no direct information is obtained on the intermediates involved
in the complex surface reaction network. Benzene hydrogenation involves 13
surface species interconnected by 20 H-additions and abstractions, while for
toluene and o-xylene, these numbers amount to 40 and 104, and 36 and 96
respectively. Six different reaction families are involved with potentially
different activation energies and reaction enthalpies. In contrast to typical kinetic models
for aromatic hydrogenation, the SEMK methodology specifically accounts for all
partially hydrogenated species in the reaction network, including the position
in which the hydrogen atoms are added. It provides a unique insight in
potential dominant reaction pathways, most abundant surface intermediates as
well as the identification of a rate-determining step. For simulation purposes, reactant
chemisorption and product desorption were assumed to be quasi equilibrated,
while the pseudo-steady state approximation was applied for the other
hydrocarbon surface intermediates. By virtue of statistical thermodynamics and
the implementation of thermodynamic consistency in the model, the ultimate
number of adjustable parameters amounted to 13, i.e., 6 rate coefficients, 3
surface equilibrium coefficients and 4 chemisorption equilibrium coefficients.

The effect of the operating conditions on the aromatic
hydrogenation behavior is consistent with the literature [2]. A maximum in the
conversion as a function of the temperature is observed for all components, see
Figure 1. The increase of the hydrogenation rate coefficient with the
temperature is gradually overcompensated by the decrease of the surface
concentrations of the hydrocarbon species. An increase in the hydrogen inlet
partial pressure enhances the cycloalkane outlet flow rate, while the opposite
is observed for the aromatic inlet partial pressure [1-2]. Aromatic
hydrogenation conversions and, hence, rates, decreased in the following order:
benzene > toluene > o-xylene.

The activation energies are in the range of 59 to 73 kJ
mol-1 and the surface reaction enthalpies correspond to slightly
endothermic H addition reactions consistent with literature [1-5]. This indicates
that on Pt the aromaticity is lost upon adsorption of the aromatic component on
the surface.  Chemisorption enthalpies were estimated very close to each other
for toluene and o-xylene. As a result, according to the SEMK model, differences
in hydrogenation rates between the investigated monoaromatic components are
primarily due to differences in H addition rate coefficients rather than in
chemisorption strength. The simulated temperature effect is more pronounced
than what is actually observed. A refinement in the fundamentals of the
preexponential factor calculation is expected to further enhance the agreement
between model simulations and experimental data.


[1]            Bera, T.; Thybaut, J. W.; Marin, G. B.,
Ind. Eng. Chem. Res., In press,  doi: 10.1021/ie200541q,

[2]            Thybaut, J. W.; Saeys, M.; Marin, G. B.,
Chem. Eng. J., (2002) 90, 117.

[3]            Lin, S. D.; Vannice, M. A., J.Catal.,
(1993) 143, 563.

[4]            Saeys, M.; Reyniers, M. F.; Thybaut, J.
W.; Neurock, M.; Marin, G. B., J.Catal., (2005) 236, 129.

[5]            Cooper, B. H.; Donnis, B. B. L., Appl.
Catal. A Gen., (1996) 137, 203.

" src="https://www.aiche.org/sites/default/files/aiche-proceedings/conferences/..." height="70" class="documentimage">