(89f) Mechanistic Modeling of Autoxidation of Ethyl Benzene

Mechanistic
modeling of autoxidation of ethyl benzene

Siddharth
R. Jain ^a, Kaushik Basak ^b, R. Vinu ^c

^a
Department
of Chemical Engineering, Indian Institute of Technology Madras, Chennai ?
600036, E-mail: sid.jain1991@gmail.com

^b
Shell Technology Center Bangalore (P&T), Shell India Markets Pvt. Ltd., RMZ
Centennial Campus B, # 8B Kundalahalli Main Road, Bengaluru 560048, India,
E-mail: k.basak@shell.com

^c
Department of Chemical Engineering, Indian Institute of Technology Madras,
Chennai ? 600036, E-mail: vinu@iitm.ac.in

Liquid
phase autoxidation of hydrocarbons has industrial importance especially for
alkyl benzenes as these form key building blocks for the production of phenol,
acetone, polyethylene terephthalate, styrene and propylene oxide, etc. Autoxidation
proceeds via free radical mechanism and chemical kinetics developed so far are
empirical in nature. Increasing the yield and selectivity today even poses a
scientific challenge owing to its complex nature. Oxidation of alkyl benzenes involves
autocatalytic chain reactions, including the formation and consumption of alkyl
peroxy, alkoxy and alkyl radicals in a series of initiation, propagation and
termination reactions. The desired intermediates further react and form
undesirable products, consequently drawing a sense of balance in keeping lower
conversions and higher selectivity.

A
mechanistic/microkinetic model for liquid phase autoxidation shall provide key
pathways for the selective production of desired products and intermediates at
higher conversions. The kinetic model was developed with heuristic approach
with plausible reaction products for liquid phase oxidation of ethyl benzene
(EB) at 400 K. Abridged mechanism of key reaction sequences are shown in Figure
1, where short lived alkyl peroxy radicals are formed by addition of molecular
oxygen over the alkyl radicals, and the former abstracts hydrogen from the
substrate to form the desired products. Mole balances of all the species
involved in the process were coupled with batch, semi-batch and stirred tank
reactor models to apprehend the details of the microkinetic modeling. The system
of differential-algebraic equations was solved to understand the time evolution
of intermediates and final products. Importantly, the rates of free radical
species were also included (quasi steady state assumption for free radicals
used usually was ridiculed) in the study. A majority of the rate coefficients
of elementary steps were taken from the literature, while a few were
numerically fitted to match the product profiles. The proposed mechanism was
then validated with the experimental outcomes from a pressurized reactor vessel.
As a part of intensifying the overall process selectivity, a number of catalysts
(based on cobalt, nickel, barium and quaternary ammonium salts) were utilized
to comprehend the competing reactions from the proposed mechanism to improve
the selectivity of desired products like 1-phenyl-ethylhydroperoxide (ethylbenzene
hydroperoxide, EBHP), acetophenone (methyl phenyl ketone, MPK), and
1-phenylethanol (methyl phenyl carbinol, MPC). Statistical design of
experiments was carried out and the effects of temperature, initiator and
passivator concentration at various levels were analyzed. Interesting results
from the study shall be discussed in the presentation.