NSF Award Supports Development of Open-Source Platform to Accelerate Electrolyte and Energy Materials Research

Nav Nidhi Rajput, an assistant professor in the Department of Materials Science and Chemical Engineering,has received a $589,000 grant from the National Science Foundation's Cyberinfrastructure for Sustained Scientific Innovation (CSSI) program to develop a platform designed to speed up the discovery of advanced energy materials.

The project aims to build a free, open-source computer tool. This tool will help scientists design better batteries, supercapacitors and other energy technologies more quickly and accurately. The project will expand MISPR (Materials Informatics for Structure-Property Relationships), a community-driven software framework developed by Rajput and her team to model complex liquid systems and interfaces that are central to modern energy technologies.

"Machine learning has become an important tool in modern chemistry and material science, and it is also a critical component in the MISPR infrastructure," said Tengfei Ma, a team member of the project. "Our machine learning team integrates commonly used machine learning technologies into the infrastructure to accelerate the electrolyte property prediction, and more importantly we will also develop new models to better understand the intermolecular interactions and model electrode-electrolyte interfaces."

Currently, it is difficult to model and predict how liquid electrolytes, the chemicals inside batteries that move charge, behave. This is especially true when they interact with solid materials. Most existing tools focus on solid materials and don't work well for complex liquids or liquid-solid interfaces. This in turn slows down research and innovation.

"Electrolyte behavior is incredibly complex because molecules can arrange themselves in many different stable configurations in solution," Rajput explained. "MISPR allows us to move beyond studying a single structure and instead analyze a spectrum of possible molecular environments. That provides a much more realistic picture of how these systems actually behave."

MISPR fixes this problem by combining physics-based simulations, computer simulations of liquids and machine learning. The physics-based simulations help to understand how atoms and molecules behave, while the computer simulations and machine learning help spot patterns to make predictions faster.

"It completely changes the way we look at liquid solutions, where we currently only focus on one static solvation environment and how we think about molecules interacting with each other in liquid solutions," Rajput said. "This helps us in broadening our understanding to something called a spectra of solvation structure, where we can accurately identify multiple different stable solvation environments and liquid."

Beyond the science, the project emphasizes community building. The MISPR platform will be distributed as free, open-source software through widely accessible platforms such as GitHub and PyPI, allowing researchers worldwide to contribute to and benefit from the framework.

"The real world impact is that we will have the understanding of how molecules interact in the solution accurately," Rajput said.

A network of collaborators from universities and national laboratories will serve as early adopters of the platform, helping to expand its capabilities and ensure its long-term sustainability. Training workshops and educational resources will also be developed to support students and researchers learning to apply data-driven approaches in materials discovery.

By creating an integrated computational infrastructure for studying electrolytes and interfaces, the project aims to accelerate research across a range of technologies.

"This effort is about building a shared scientific infrastructure," Rajput said. "By making these tools open and accessible, we hope to enable researchers around the world to explore complex chemical systems much more efficiently."

- Angelina Livigni