Theoretical and experimental investigation of heterogeneous catalytic reactions: Two case studies

Brian R Kromer, Purdue University

Abstract

The overall goal of this work is to utilize multi scale modeling techniques as well as various experimental methods to understand the fundamentals of two important catalytic reactions. The reactions studied were the combustion of methane on PdO catalysts and the reduction of NOx compounds in diesel exhaust using a Pt/Ba/Al2O3 catalyst. Both reactions are important from an environmental standpoint due to strict regulations on reducing the emissions of harmful NOx compounds released into the atmosphere. The goal of the study on methane combustion was to understand the fundamental nature of the chemistry that actually occurs on the catalyst surface. To accomplish this, the reaction steps and catalyst surface were modeled through the use of density functional theory (DFT). Density functional theory is a quantum mechanics based formalism used to calculate the wavefunctions and energies of various quantum states that the system can be in. DFT can be used to examine various stable adsorption states on the surface as well as study the fundamentals of the reaction mechanism through the use of transition state calculations and statistical mechanics. DFT was used to examine the plausibility of the commonly proposed Mars-van Krevelen mechanism for methane combustion through the use of various adsorption calculations as well as reaction pathway studies. The most important step in this mechanism is the abstraction of the first hydrogen from the methane molecule. This step is well known as the rate limiting step in the reaction mechanism. The results show that methane will activate on the saturated PdO(100) surface and does not necessarily require an oxygen vacancy or a defect in the surface structure. The activation energy for this step was found to be 28.5 kcal mol-1, which is within the range seen in the experimental literature. Besides the rate determining step, other important steps such as oxygen dissociation were investigated. Also, a large library of adsorption calculations were completed in order to consider all of the intermediate species thought to be involved in the reaction. The results ultimately show that the Mars-van Krevelen mechanism is not sufficient to describe the complex nature of the methane combustion reaction on PdO surfaces. Species such as O 2, H2O, CO2, and CH4 were all shown to behave differently than the mechanism proposes. The results also indicate the there are multiple pathways that the reaction steps can occur through and that the mechanism is most likely too complicated to be described by a simple set of elementary steps. In contrast to the atomistic modeling of the methane combustion work, the NOx reduction in diesel exhaust was studied through the use of macro scale reactor modeling with the purpose of being able to predict the behavior of the catalytic reactor under varying operating conditions. The goal of the work was to be able to develop a simple model that could be solved in real time in order to be able to control the operation of the engine/catalyst system in an actual on road application. The catalyst studied was Pt/Ba/Al 2O3 which is known as a NOx trap. It is operated in a cyclic manner in which NO is oxidized to NO2 and NOx is stored on the catalyst surface under an oxidizing environment. Once the surface is saturated, a short burst of a reducing atmosphere is used to desorb and reduce the NO x to N2 and H2O. One dimensional models were developed to describe each of the steps in the cycle. The storage model included mechanisms for describing diffusion into the bulk of the catalyst as well as Pt/Ba proximity effects. The results show that the simple modeling approach used worked very well for describing the NOx trap reactor under various inlet concentrations as well as varying temperatures. The model is able to make accurate predictions as to when the storage component of the catalyst is saturated. The separate models were also designed for easy coupling so that a model describing the entire cycle can be put together. In addition to the modeling, some experimental scanning tunneling microscopy (STM) studies were performed on the NO oxidation reaction. The goal of this work was to use STM to understand the structural changes of the catalyst surface after being exposed to oxidizing conditions. The catalyst used was a single crystal Pt(100) surface. The studies were performed to support the hypothesis that larger particles are more active toward NO oxidation because the large particles consist mostly of stable low index planes of the catalyst and are resistant to forming a surface oxide. The results confirm that the surface does not go under any major topological changes even in the presence of extreme oxidizing environments which support the hypothesis from other reaction studies.

Degree

Ph.D.

Advisors

Thomson, Purdue University.

Subject Area

Chemical engineering

COinS