 |
|
It is well known that for reactions such as hydrotreating of heavy oil where a number of reactants are present, there are usually multiple reaction pathways that can lead to different product distributions depending upon the temperature, pressure and feed reactant concentrations. However it is less widely known that small scale fluid flow in the reactor can also significantly influence reaction outcome. Even when the feed flows are constant, gas-liquid, packed bed reactors often exhibit large flowrate fluctuations in the bed (which are called "pulses") that cause local hear and mass transfer variations of up to an order of magnitude. Our initial work on this topic, which was supported by the Amoco Oil Co., was a study that provided new insight into how pulses form. In our subsequent efforts, which were supported by the Monsanto Corp., we found that significant improvements in reaction selectivity can be gained with proper choice of reactor geometry and flow rates to produce the optimal flow and transport effects by "tuning" the pulsing frequency for a given reaction. Currently we are experimentally examining the reaction system: 1) Phenylacetylene + H => Styrene; 2) Styrene + H => Ethylbenzene; 3) Phenylacetylene + H2 => Ethylbenzene. Based on our theoretical work we are trying to maximize the intermediate (which we have designated the "desired") product, styrene, and minimize the "undesired" ethylbenzene. We have found that when pulses are present, styrene is formed faster and its maximum concentration increases.
Our continuing and proposed work will involve futher experimental and theoretical studies with three specific goals. First we need to determine how detailed physical (i.e., fluid flow) and chemical mechanisms interact to alter the product distribution. A second goal is to develop theoretical models that closely represent these detailed mechanisms. A third issue is to investigate nonisothermal systems where heat transfer fluctuations are expected to influence reaction outcome as much as mass transfer fluctuations. Process reactions such as hydrotreating have substantial temperature gradients and our studies to date have excluded hear transfer only to make the reaction less complex and more amenable to fundamental study. Instructors can contact Professor McCready for this information.
