Machine Learning-Aided Prediction of Na Solid-State Interfaces
Li, B., Jeon, H., Persson, K.A. "Machine Learning-Aided Prediction of Na Solid-State Interfaces." In preparation.
CV
Education
University of California, Berkeley Ph.D. Materials Science & Engineering (Aug 2021 – May 2026)
- Advisor: Prof. Kristin A. Persson, Lawrence Berkeley National Laboratory
University of California, Berkeley M.S. Materials Science & Engineering (Dec 2023)
- GPA: 3.80, Graduate Certificate in Applied Data Science
The Pennsylvania State University B.S. Materials Science & Engineering (May 2021)
- Magna cum laude, GPA: 3.87, Minor: Applied Mathematics, Dean’s List (all semesters)
Research Experience
Graduate Student Researcher, Persson Group
Lawrence Berkeley National Laboratory / UC Berkeley Aug 2021 – Present
AI/ML Algorithm Development
- Developed and benchmarked MLIPs (MACE, ACE, GRACE, NequIP, M3GNet, CHGNet, DeepMD-kit) for reactive disordered systems
- Identified ACE-based descriptors outperform spherical-harmonic architectures for amorphous morphology prediction, 100× faster inference on CPUs while maintaining accuracy for million-atom simulations
Interface & Surface Reaction Modeling
- Discovered 4-layer heterogeneous SEI morphology (LixP, nanocrystalline Li2S, disordered LixP-Li2S, amorphous electrolyte) invisible to in-situ XPS/SEM
- Explained passivation via electronic bandgap analysis: Li2S acts as insulating barrier while LixP phases are semiconducting (1.1–1.8 eV)
- Distinguished SEI from mixed-conductor interphase (MCI) based on NaxP percolation pathways
Million-Atom Reactive Simulations
- Executed 1M-atom, 10 ns reactive interface simulations, equivalent to millions of CPU-hours of AIMD, to characterize SEI vs. MCI formation kinetics
Open-Source Development (MPMorph/Atomate2)
- Architected core MPMorph module for automated melt-quench MD; integrated Packmol backend for molecular packing
- Adopted by 50+ researchers across NERSC matgen allocation; maintained with 90% unit test coverage and CI/CD
Active Learning Framework
- Developed pipeline combining structural generation with MaxVol and DIRECT sampling
High-Throughput Data Analytics
- Managed scalable data pipelines across 3 MongoDB databases (50k+ entries each)
- Contributed 15k entries to Materials Project R2SCAN database (public release v2025.09.25)
Technical Leadership
- Lead representative for DOE BMR and ESRA consortiums; present quarterly to multi-institutional teams
- Mentored 5+ PhD students and postdocs across 5 consortiums on MLIP development and deployment
Undergraduate Research Assistant, Dabo Group
The Pennsylvania State University Sept 2018 – May 2021
- Executed 1000+ DFT calculations (Quantum ESPRESSO) to model surface adsorption and charge transfer kinetics for Pt–Pd nanoparticle electrocatalysts (DOE-funded)
- Developed cluster basis sets with ICET for cluster expansion; performed Monte Carlo simulations to predict surface morphology
- First-author contribution demonstrating thermodynamic favorability of Pd channel formation; published in Physical Review B
Research Assistant, Lemke Group
University of Hong Kong Sept 2016 – Dec 2017
- Modeled catalytic NH₃ production on Fe–Ni alloy clusters (Awaruite, Tetrataenite) for prebiotic nitrogen fixation
- Classical modeling using Finnis–Sinclair, Lennard–Jones, and Sutton–Chen potentials (n ≤ 24 atoms)
- Presented at 2017 AGU Fall Meeting
Technical Skills
AI/ML & Algorithm Development PyTorch, TensorFlow, JAX | MLIPs: MACE, ACE, GRACE, NequIP, Allegro, M3GNet, CHGNet, DeepMD-kit | Active Learning (MaxVol, DIRECT) | GPU/CPU Deployment
Atomistic Simulation & Electronic Structure DFT (VASP, Quantum ESPRESSO, PBEsol, r²SCAN) | AIMD | Classical MD (LAMMPS) | Bandstructure & DOS | Interface Reactivity | Amorphous/Disordered Systems
Software Engineering Python | atomate2/pymatgen (core contributor) | Git/GitHub | Unit Testing (90% coverage) | Kokkos-enabled LAMMPS | Docker | AI-Assisted Dev (Claude Code, Cursor)
Data & HPC MongoDB (150k+ entries) | Pandas, NumPy, SciPy | NERSC Perlmutter, NREL Kestrel | SLURM | CUDA/Kokkos Parallelization
Visualization OVITO Pro | VESTA | Trajectory Analysis | Electronic Structure Visualization
Languages English (fluent) | Mandarin (fluent) | Cantonese (fluent)
Publications
The Role of Machine Learning Interatomic Potentials in Developing the Next Generation of Solid-State Batteries
Phuthi, M.K., Wei, G., Li, B., Majumdar, S., Kolluru, V.S.C., Kumar, N., Blau, S., Canepa, P., Chan, M.K.Y., Gomez-Bombarelli, R., Persson, K.A., Ceder, G., Ong, S.P. "The Role of Machine Learning Interatomic Potentials in Developing the Next Generation of Solid-State Batteries." In preparation.
Towards an Atomistic Understanding of Solid-State Reaction Interfaces Using Machine-Learned Interatomic Potentials
Li, B., Karan, V., Persson, K.A. (2025). "Towards an Atomistic Understanding of Solid-State Reaction Interfaces Using Machine-Learned Interatomic Potentials." arXiv. DOI: arxiv.2511.20976.
An Atomistic Study of Reactivity in Solid State Electrolyte Interphase Formation for Li/Li₇P₃S₁₁
Li, B., Karan, V., Kaplan, A.D., Wen, M., Persson, K.A. (2025). "An Atomistic Study of Reactivity in Solid State Electrolyte Interphase Formation for Li/Li7P3S11." The Journal of Physical Chemistry C. Accepted.
Atomate2: Modular Workflows for Materials Science
Ganose, A.M., Sahasrabuddhe, H., Asta, M., Beck, K., Biswas, T., Bonkowski, A., ..., Li, B., ..., Jain, A. (2025). "Atomate2: Modular Workflows for Materials Science." Digital Discovery, 4(7), 1944–1973.
A Cellular Automaton Simulation for Predicting Phase Evolution in Solid-State Reactions
Gallant, M.C., McDermott, M.J., Li, B., Persson, K.A. (2024). "A Cellular Automaton Simulation for Predicting Phase Evolution in Solid-State Reactions." Chemistry of Materials, 37(1), 210–223.
Quantifying Multipoint Ordering in Alloys
Goff, J., Li, B., Sinnott, S., Dabo, I. (2021). "Quantifying Multipoint Ordering in Alloys." Physical Review B, 104, 054109.
Talks & Presentations
Machine Learning Aided Atomistic Modeling of Solid State Electrolyte Interphase Formation for Li/Li₇P₃S₁₁
Oral Presentation at AI4X 2025 International Conference, Singapore
An Atomistic Study of Reactivity in Solid State Electrolyte Interphase Formation for Li/Li₇P₃S₁₁ Systems
Oral Presentation at Materials Research Society (MRS) Spring Meeting, Seattle, WA, USA
Awaruite and Tetrataenite Driven NH₃ Synthesis
Poster Presentation at American Geophysical Union (AGU) Fall Meeting, San Francisco, CA, USA
Teaching
Honors & Awards
| Award | Year |
|---|---|
| UCB MSE Corrosion Award | 2025–2026 |
| DOE Energy Storage Research Alliance (ESRA) | 2023–Present |
| DOE Advanced Battery Materials Research (BMR) | 2021–Present |
| Sam Zerfoss Memorial Scholarship | 2019–2021 |
| Matthew J. Wilson Honors Scholarship | 2020–2021 |
| The Sobota Family Scholarship | 2020 |
| Edward S. Sproles, Jr. Scholarship Fund | 2020 |
| John and Elizabeth Holmes Teas Merit Scholarship | 2019 |
| Virginia S. and Phillip L. Walker, Jr. Scholarship | 2018 |
Certifications
- Graduate Certificate in Applied Data Science
- International Baccalaureate (IB) Bilingual Diploma