(4y) Computational Heterogeneous Catalyst Design from Material Stability to Mechanistic Assessment

Recent advances in computational methods now permit theory in part to guide the design of heterogeneous catalysts. Rapid screening methods and machine learning (ML) have been leveraged to identify many promising catalysts for CO₂ reduction, ligands for homogeneous catalysis, and organic structure-directing agents (OSDAs) for zeolite synthesis. Heterogeneous catalysis will play a key role in energy and chemical industries as global needs for their products grow. Rapidly improving computational tools promise concomitant improvements in catalyst development for these industries. I am particularly interested in understanding materials stability, catalyst restructuring, and the effect that these have on reaction mechanisms in electro- and thermochemical systems. New ML tools are already allowing researchers to study extended systems beyond local active sites to design catalysts at the micrometer scale—beyond the usual atomistic limits of quantum chemical methods like density functional theory (DFT). As a principal investigator, I will lead studies that combine these ML tools with rigorous mechanistic studies using high-accuracy density functional theory (DFT) in collaboration with experimental experts who study kinetics and catalyst synthesis to guide the synthesis of active, stable, and selective catalysts. More specifically, I want to develop and apply approaches to model catalyst restructuring atomistically (such as metal nanoparticle dispersion and ripening processes, including oxidative single-atom formation); small-molecule electrosynthesis and how applied potentials affect mechanisms and active site structures; and larger length scale simulations with machine learned interatomic potentials (MLIPs) to inform mesoporous catalyst design.

During my Ph.D. with David Hibbitts at University of Florida, I used DFT to study the properties of Brønsted acid zeolite catalysts and single-atom Rh catalysts for NO_x reduction. Specifically, I used deprotonation energies (DPE) and methanol dehydration to study the effects of different Al distributions on catalysis in CHA and MFI zeolites in collaboration with Prof. Raj Gounder (Purdue) [1-3]. I also collaborated with the Gounder group on several studies examining the interactions of different OSDAs with Al in zeolites and with inorganic SDAs during synthesis [4-5]. Finally, I also studied the structure and behavior of Rh nanoparticles and single-atoms supported on γ-Al₂O₃ to understand their distinct behaviors during NO_x reduction in three-way catalysts in collaboration with Prof. Phil Christopher (UCSB) and Dr. Bean Getsoian (Ford) [6-7]. During my postdoc at MIT with Rafael Gómez-Bombarelli, I began working with high-throughput screening methods to evaluate Al distributions in zeolites synthesized using different OSDAs and predict zeolite synthesis outcomes in collaboration with Profs. Yuriy Román and Elsa Olivetti at MIT and Dr. Manuel Moliner at ITQ, with funding from BASF and ExxonMobil. This work uses a combination of computational techniques, including DFT, MLIPs, and symbolic regression with other ML techniques to predict what material an OSDA will produce during synthesis.

Teaching Interests

I have a B.S. in Chemistry, a Ph.D. in Chemical Engineering, and I am doing my post-doctoral work in Materials Science and Engineering. As such, I feel best qualified and most excited to teach core undergraduate courses such as reactor design and thermodynamics and graduate courses such as chemical kinetics and statistical mechanics. Additionally, I hope to teach graduate electives on heterogeneous catalysis at the molecular level and computational methods in the chemical sciences. As an undergraduate, I served as a teaching assistant (TA) for a junior-year chemistry lab. During my Ph.D., I spent two semesters as a TA for graduate statistical mechanics and two additional semesters as a teaching assistant for the graduate chemical engineering lab course at the University of Florida. Finally, as a postdoctoral researcher, I co-instructed a course on computational methods in materials science and chemistry, where I was responsible for the part of the course about quantum chemistry methods.

Select Publications

[1] A. Hoffman; M. DeLuca; D. Hibbitts. J. Phys. Chem. C, 2019.

[2] J. Di Iorio; A. J. Hoffman; C. Nimlos; S. Nystrom; D. Hibbitts; R. Gounder. J. Catal., 2019.

[3] A. J. Hoffman; et al. Angew. Chem. Int. Ed., 2020.

[4] C. T. Nimlos†; A. J. Hoffman†; Y. Hur; J. Di Iorio; D. Hibbitts; R. Gounder. Chem. Mater., 2020.

[5] E. E. Bickel; A. J. Hoffman; S. Lee; H. E. Snider; C. T. Nimlos; N. K. Zamiechowski; D. Hibbitts; R. Gounder. Chem. Mater., 2022.

[6] A. J. Hoffman; et al. J. Phys. Chem. C, 2021.

[7] A. J. Hoffman†; C. Asokan†; et al. ACS Catal., 2022.