Materials science × machine learning
Connecting structure, models,
and materials discovery.
I develop machine learning and multiscale simulation workflows to understand materials—and make better decisions about which ones to explore next.
Postdoctoral Appointee / Sandia National Laboratories
Albuquerque, New Mexico
01 / Research
My work connects learning algorithms with physical models, from molecular separations and battery electrolytes to the growth of thin films.
Metal–organic frameworks offer a vast space of porous materials. I combine graph neural networks, uncertainty estimates, and multi-objective Bayesian optimization to screen for favorable separation performance.
Graph neural networks · Bayesian optimization
A candidate pool of 174,758 LLZO–polymer microstructures provides a testing ground for surrogate-guided screening. The work connects transport and mechanical tradeoffs with interpretable microstructural features.
Multi-objective optimization · Microstructure
Surface binding energies vary with chemistry and structure. I investigate how atomistically resolved distributions in Mo–S systems propagate through sputtering and deposition models, with experimental sputter yields as a benchmark.
Atomistic simulation · Multiscale modeling
Research foundations
My doctoral research developed applicability-domain methods for materials machine learning and high-throughput atomistic workflows for metallic glass forming ability. These questions—what a model knows, and what a simulation reveals—continue to guide my work.
02 / Selected publications
Journal of Materials Chemistry A
Graph neural networks and multi-objective Bayesian optimization for screening metal–organic frameworks with favorable separation performance.
npj Computational Materials
A kernel-density-estimation approach to understanding where a machine learning model's predictions and uncertainty estimates are reliable.
Journal of Materiomics
Connecting elemental descriptors and molecular-simulation features to the prediction of metallic glass critical cooling rates.
Explore the complete record, including preprints and ongoing work, on the publications page or Google Scholar.
03 / Background
I’m a materials scientist and engineer at the Center for Integrated Nanotechnologies at Sandia National Laboratories. My work brings together materials physics, scientific computing, and machine learning.
I earned my Ph.D. at the University of Wisconsin–Madison and my engineering degree at Fort Lewis College. Alongside research, I’ve helped build and administer scientific computing clusters and supported students and colleagues in using research software.
Full experience & credentialsCenter for Integrated Nanotechnologies
Sandia National Laboratories
Computational Materials Group
University of Wisconsin–Madison
University of Wisconsin–Madison
Distributed minor in Machine Learning
University of Wisconsin–Madison
Fort Lewis College
Electro/Mechanical Engineering concentration · Mathematics minor
Service & community
Computational Materials Science and Engineering Committee, The Minerals, Metals & Materials Society (TMS).
Assisted participants in the in-person Software Development in Materials Science and Engineering workshop at TMS 2026.
Professional development
TMS · Online course
–
NVIDIA · Workshop
04 / Contact
For research collaborations and professional inquiries,
get in touch by email.
At Sandia: leschul@sandia.gov
Albuquerque, New Mexico · +1 (806) 678-6904