Multiscale modeling of traditional and dynamic heterogeneous catalytic processes
Date
2022
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Heterogeneous catalysis is a key element in a global energy strategy that enables a lower energy route for the many chemical reactions essential to food production, fuels, pharmaceuticals, and other essential materials in our everyday lives. Modeling catalytic systems can inform the design of experiments, identify catalyst properties that optimize quantities-of-interest (QoI) for the discovery of new catalysts, and link catalyst properties to commercial reactor design. A comprehensive understanding of these reactions, achieved through multiscale modeling, is a vital step in establishing viable commercial processes. Additionally, a thorough understanding of the key elements comprising an effective model, efficiently and accurately parameterizing the model, and properly understanding and accounting for parametric uncertainty in models are essential for maximizing and quantifying the models' predictive ability. ☐ Chapter 2 of this thesis provides a detailed discussion of the theory and process of building microkinetic models, parameterization via statistical thermodynamics, and analysis of model results. Chapters 3 & 4 present two pieces of software developed to speed parametrization, minimize errors, and improve efficiency. The Python Multiscale Thermodynamic Toolbox (pMuTT) streamlines the statistical thermodynamic computations required to convert density functional theory data to thermochemical and kinetic parameters and create the necessary input files for microkinetic and kinetic Monte Carlo models. This both streamlines this work as well as reduces errors in statistical thermodynamic calculations and input file creation. The 2nd software is the Python Group Additivity (pGrAdd) which uses the semi-empirical group additivity method to estimate thermochemical properties of adsorbates without the use of computationally expense density functional theory. Model thermodynamic data can now be parameterized in minutes vs. days, allowing for rapid model deployment and results. Chapter 5 addresses the impact of parametric uncertainty on the microkinetic model's quantities of interest. We use group additivity as the source of parametric uncertainty and establish a probabilistic framework for assessing parametric uncertainty from any source. In Chapters 6, 7, & 8, we model several reactions of interest. Chapter 6 examines the dehydrogenation of ethane and propane to ethylene and propylene from shale gas to optimize this rapidly growing resource. Conversion, selectivity, yield, dominant reaction paths, and rate-determining steps are identified for each reaction. Chapter 7 takes a new look at the production of ammonia, the largest volume and highly energy-intensive process using dynamic catalysis. Dynamic catalysis manipulates the binding energy of the catalyst to achieve orders-of-magnitude higher reaction rates beyond the Sabatier limit and exceed equilibrium conversion. For the first time, we demonstrate this on a real-world comprehensive ammonia microkinetic model using a custom Matlab tool. The work demonstrates we can operate the process at the same rates/conversions as a commercial Haber-Bosch process but at lower temperature and an order-of-magnitude lower pressure paving the way for significantly reduced energy costs. Finally, Chapter 8 tackles a need for portable power by modeling a thermoelectric generator as an alternative portable power source to batteries for electronically enabled soldiers. It combines a reduced mechanism kinetic model for the combustion of propane with constitutive heat transfer equations to model a microburner combusting high energy density propane with a commercial thermoelectric device and demonstrate the optimum operating envelope for such a device.
Description
Keywords
Catalysis processes, Dynamics, Modeling, Heterogeneous catalysis