(4ns) Atomistic Simulation of Materials for Energy Storage and Conversion

Developing new materials for energy storage and conversion can accelerate decarbonization of our society at scale. The vast design space of possible materials renders it difficult to experimentally synthesize and evaluate all materials for such applications. Computational research can accelerate such efforts by enabling predictions of material properties and providing atomic-scale, mechanistic insights into relevant processes. My proposed research involves developing and applying atomistic simulation methods to guide the development of new materials for energy storage and conversion, with a specific focus on:

1. Ion and electron transport at heterogeneous electrochemical interfaces.

2. Adsorption capacity, water stability, and synthesizability of metal-organic frameworks.

Initial efforts will involve developing a fundamental understanding of the features of these materials that give rise to desirable properties for target applications, using physics-based simulation techniques. Longer-term efforts will involve integrating these techniques with machine learning to design new materials from first principles. These projects leverage my unique combination of prior experience in fundamental methods development (stochastic linear algebra algorithms for quantum chemistry) and more applied research (atomistic simulations of ion migration in solid electrolytes and water adsorption in metal-organic frameworks). All methods developed by my group will be implemented in open-source software with comprehensive documentation to encourage others in the community to adopt and extend them.

Previous and Current Research:

My PhD research primarily involved developing stochastic electronic structure methods. By integrating bespoke Monte Carlo sampling algorithms with iterative eigensolvers, I achieved greater statistical efficiency than state-of-the-art methods for molecules with Hilbert spaces as large as 10²⁵ [1–4]. A core focus of this project was developing massively parallel open-source software; I received a graduate fellowship, which included both mentorship and funding, from the Molecular Sciences Software Institute to support these efforts.

I received a Schmidt Science Fellowship to pivot into a different field for my postdoc. My current research focuses on applying atomistic simulation methods to energy materials. One ongoing project focuses on ion transport in solid electrolytes. My recent publication [5] explores the use of phonon-related features, calculated using density functional theory, to predict activation energies for multivalent ion transport. Currently, I am developing a machine-learned interatomic potential and performing molecular dynamics simulations to study the coupling between tilt disorder and ion migration in the Na₃OCl ion conductor. A second area of focus is water adsorption in metal-organic frameworks (MOFs) for thermal energy storage. In collaboration with an undergraduate student, I am developing a machine-learned interatomic potential for water in MOFs and interfacing it with custom Monte Carlo software to predict water adsorption isotherms. During my postdoc, I have also been involved in three separate experimental collaborations [6–8].

Ion and electron transport at heterogeneous interfaces:

Interfaces between electrode and electrolyte materials can consist of a heterogeneous mosaic of amorphous regions, electrolyte decomposition products, point defects, and grain boundaries. This complexity renders it difficult to characterize the behavior of electrons and ions at these interfaces. My group will develop machine-learned interatomic potentials and use them to simulate heterogeneous electrolyte materials under nonequilibrium operating conditions using molecular dynamics, Monte Carlo sampling, and the Boltzmann transport equation framework. Additionally, I will leverage my prior experience in fundamental electronic structure and Monte Carlo sampling methods to pursue complementary methods development efforts. These will focus specifically on accelerating “on-the-fly” training of machine-learned potentials, automating the identification of ion transport mechanisms, and accelerating the sampling of rare events. These efforts will inform the design of new electrolyte materials with superior electrochemical stability and rate performance.

Adsorption capacity, stability, and synthesis of metal-organic frameworks

The tunability of the building blocks of metal-organic frameworks (MOFs) gives rise to a vast design space for applications in thermal energy storage and CO₂ capture. My group will leverage atomistic simulation methods to inform such design efforts from first principles. Building upon my postdoctoral work, we will develop machine-learned interatomic potentials that can describe strongly interacting adsorbates (e.g. salts, H₂O, and CO₂) in MOF pores more accurately than the classical interatomic potentials that are widely used today. Using these potentials, we will develop methods that integrate Monte Carlo sampling with machine learning to predict adsorption and transport properties for large databases of MOFs. Additionally, we will evaluate MOF stability and synthesizability by leveraging density functional theory to calculate static stability properties and classical molecular dynamics to simulate self-assembly and hydrolysis processes. These efforts will help guide the synthesis of new MOFs for target applications.

Teaching Interests:

My primary objective as an instructor is to teach students to reason when they do not know the correct answer, and to have them pinpoint and verbalize gaps in their understanding. I do so by creating a question-oriented environment in which students feel comfortable making mistakes. I use the mild speech impediment associated with my disability to address students’ potential hesitation to express vulnerability in this way. I have refined these techniques throughout my graduate career, through my experiences teaching General, Organic, and Physical Chemistry; Global Warming: Understanding the Forecast (for non-STEM majors); and Computer Science and App Design (for middle- and high-school students). I have received departmental teaching awards from both Yale University and the University of Chicago. My previous cross-disciplinary research and teaching experiences have prepared me to teach chemical engineering courses such as thermodynamics, transport, and atomistic simulation. I would be interested in teaching or developing elective courses on climate change or renewable energy if given the opportunity.

Fellowships received: Barry Goldwater Scholarship, Rhodes Scholarship, Molecular Sciences Software Institute Fellowship, Department of Energy Office of Science Graduate Student Research Fellowship, Schmidt Science Fellowship

References

[1] S. M. Greene, R. J. Webber, J. Weare, and T. C. Berkelbach. J. Chem. Theory Comput. 15, 2019, 4834–4850.

[2] S. M. Greene, R. J. Webber, J. Weare, and T. C. Berkelbach. J. Chem. Theory Comput. 16, 2020, 5572–5585.

[3] S. M. Greene, R. J. Webber, T. C. Berkelbach, and J. Weare. SIAM J. Sci. Comput. 44, 2022, A3067–A3097.

[4] S. M. Greene, R. J. Webber, J. E. T. Smith, J. Weare, and T. C. Berkelbach. J. Chem. Theory Comput. 18, 2022, 7218–7232.

[5] S. M. Greene, D. J. Siegel. Chem. Mater. (in review). ChemRxiv: 10.26434/chemrxiv-2024-dr91f.

[6] H. Hao, Y. Liu, S. M. Greene, et al. Adv. Energy Mater. 13, 2023, 2301338.

[7] J. S. Yoon, D. W. Liao, S. M. Greene, et al. ACS Appl. Mater. Interfaces 16, 2024, 18790–18799.

[8] K. Sada, S. M. Greene, et al. Angew. Chem. Int. Ed. (in review).