(4qi) Physically Informed Material Design for Sustainable Energy Production and Storage

My research interest lies in the development of efficient, climate friendly chemical processes and energy generation through computational material modeling and integrated process performance optimization. My research lab will identify key process material compositions leading to optimal chemical reactions for any given process and higher energy and chemical production efficiencies. My coupling of computational ab intitio techniques like density functional theory, process and technoeconomic modeling, and experimental data collection techniques like thermogravimetric analysis and adiabatic calorimetry enables high fidelity system models that span many orders of spatiotemporal magnitude. Therefore, I can globally optimize the highly interdependent atomistic chemical behavior and plant performance. Further, I will use multilevel statistical approaches, based on Bayesian inference and machine learning techniques, to integrate computational and experimental results and inform phase space exploration, minimizing the number of required experiments/simulations while guiding the search for optimal materials composition and plant design.

Doctoral Research

My doctoral research focused on the optimization of the solar thermochemical hydrogen production (STCH) cycle. There are several hydrogen production pathways including electrolysis, thermolysis, steam reforming, photocatalysis, and thermochemical processes. Among these, STCH via H₂O splitting has the potential for high solar energy conversion efficiencies because it uses radiation from the entire solar spectrum in the form of heat to drive the endothermic H₂O splitting reaction. Furthermore, unlike electrolysis or photocatalysis there is no conversion of energy to electricity needed, which incurs a large efficiency loss. Lastly, STCH operates at significantly lower temperatures (~1500°C cooler) than the direct thermolysis of water. Using STCH as an exemplar cycle, I developed and successfully implemented a state-of-the-art reduction thermodynamic modeling technique, the CrossFit Compound Energy Formalism (CF-CEF). The CF-CEF is a more expansive and accuracy thermodynamic modeling algorithm as compared to current common practices in the field. Furthermore, coupling the CF-CEF with Bayesian Inference techniques has led to high throughput material identification and data selection techniques significantly decreasing the time to material identification.

Postdoctoral Research

In my postdoctoral studies I worked on the computational examination of surface plasma species reduction kinetics. The kinetics of iron ore reduction by plasma species involves several key steps, namely surface impingement of the reducing species (i.e. H₂, H₂⁺, H₂^-, H^-, H⁺, etc.) on the surface, either adhesion or rebounding of the species, once adhered surface reactions to form H₂O, H₂O desorption, and diffusion of O atoms from the bulk to the surface where it can react. In this project I built hierarchical computational models of these reaction steps. Because these steps intrinsically involve reduction/oxidation reactions, classical force-field method, which are constructed for specific oxidation states and coordination environments, are insufficient in modeling the behavior. Therefore, I deployed a combination of density functional theory calculations and Machine Learning (ML) techniques to understand these steps. Two ML models have been under development for accelerating the DFT calculations: 1) a simpler Gaussian Process (GP) method, and 2) a graphical neural network, both of which predict system energies and atomic forces from microscopic geometries. The GP predicts the energy of a system of atoms as a summation of atomic contributions, where the contributions are calculated from an GP interpolation of reference atomic configurations. GP regression is a non-parametric Bayesian approach that preforms both prediction and estimation of uncertainty. Thus, I have access to on-the-fly estimates of the error.

Future Research

My future work is aimed at applying my skills to optimize material performance in energy generation and storage specific to chemical looping with inorganic materials. Energy generation and storge comes in many forms such as the common low temperature (400-800 K) Chemical Looping (CL) cycles, Combustion (CLC), Reforming (CLR), Gasification (CLG), and Partial Oxidation (CLPO), and H₂ storage (CLHS). Likewise, high temperature (1000-1500K) Solar Thermochemical applications exist, such as Solar Thermochemical CO₂ and H₂O splitting, and Solar Thermochemical Energy Storage. The extensive but not all-inclusive list of chemical processes all rely on the chemical reactions involving an inorganic material. The selection and optimization of these materials directly affects the efficiency of energy production and thus its cost of production. Investigating all these processes across a wide range of material families and compositions would normally exceed a lifetime of work. However, my unique combination of theoretical, experimental, and statistical approaches will lead to significantly reduced time to material identification and optimization for a given chemical looping process as compared to a trial-and-error approach.

Steven Wilson Teaching Interests

I believe that everyone can learn any subject they desire. While some people may absorb knowledge at different rates, everyone is capable of achieving whatever academic dreams they set their mind to. The job of the professor is to nurture and aid the student in achieving their academic goals through guided lectures and hands-on laboratory work. Chemical engineering, especially, requires the finesse of both in-class and laboratory work to root the fundamentals of the subject within the student. In this way, the students will not just regurgitate memorized facts but have a firm “instinctual” knowledge of the subject creating a critical thinker and stout problem solver. Through my four years as a Navy Nuclear Propulsion Plant instructor and three years as an undergraduate teaching assistant I have developed a unique skill set in helping students tie theoretical knowledge to practical application. These teaching skills, coupled with my over decade long experience of mentoring junior sailors and undergraduate and master students has prepared me extensively for the responsibility of training students to master their subject and solve the global current and future engineering challenges.

My teaching interests focus on understanding chemical reactions and then designing materials and systems which facilitate the most efficient outcome for a given process. The fundamentals of chemical and process reactions stem from the thermodynamics that control them and the statistical results of the models we use as scientists to understand them. Therefore, I am particularly interested in teaching Thermodynamics at the undergraduate and graduate level. Additionally, I enjoy optimization and modeling of complex problems to elucidate intrinsic properties of a given material or process. To this end, I am also interested in teaching Numerical Methods or Statistical Analysis. The crossroads of understanding chemical reactions and mathematically modeling them has led me, in my academic career, to computational quantum mechanical modeling. I would like to offer students the opportunity to learn about this useful tool by developing a graduate level course on density functional theory. This class would feature not only the fundamental quantum chemical approaches but also hands on simulations, with an emphasis on extracting or relating simulations results to real world properties, such as chemical potentials, diffusivity constants, and reaction coefficients.

Select Publications

1. Wilson, S. A., Sarsman P. W, Stechel, E. B Muhich, C. L., Extracting Metal Oxide Redox Thermodynamics from TGA Measurements Requires Moving Beyond the Linearized Van 't Hoff Approach. Frontiers in Energy Research 2024 In Press.
2. Tran, J. T.; Warren, K. J.; Wilson, S. A.; Muhich, C. L.; Musgrave, C. B.; Weimer, A. W., An Updated Review and Perspective on Efficient Hydrogen Generation Via Solar Thermal Water Splitting. WIREs Energy and Environment 2024, 13, e528.
3. Wilson, S. A.; Muhich, C. L., A Bayesian Method for Selecting Data Points for Thermodynamic Modeling of Off-Stoichiometric Metal Oxides. Journal of Materials Chemistry A 2024, 12, 13328-13337.
4. Wilson, S. A.; Stechel, E. B.; Muhich, C. L., Overcoming Significant Challenges in Extracting Off-Stoichiometric Thermodynamics Using the Compound Energy Formalism through Complementary Use of Experimental and First Principles Data: A Case Study of Ba1-Xsrxfeo3-Δ. Solid State Ionics 2023, 390, 116115.
5. Wilson, S. A.; Muhich, C. L., Fast Identification, and Construction of Adsorbate-Adsorbent Geometries for High Throughput Computational Applications: The Automatic Surface Adsorbate Structure Provider (Asap) Algorithm. Computational and Theoretical Chemistry 2022, 1216, 113830.
6. Wilson, S. A.; Stechel, E. B.; Ermanoski, I.; Muhich, C. L., Substituted Alpo-5 Zeolites as Promising O2 Sorption Pump Materials: A Density Functional Theory Study. The Journal of Physical Chemistry C 2021, 125, 1269-1281.
7. Arifin, D.; Ambrosini, A.; Wilson, S. A.; Mandal, B.; Muhich, C. L.; Weimer, A. W., Investigation of Zr, Gd/Zr, and Pr/Zr – Doped Ceria for the Redox Splitting of Water. International Journal of Hydrogen Energy 2020, 45, 160-174.