Work

High-Throughput Computation of Li-based Battery Material Databases: Chemistry-Processing-Property Relationships

Year: 2022

Type: preprint

Abstract: In this paper, we incorporate processing into the data driven property modeling of Li-based spinel battery materials. When considering spinel compounds of the form LiMe2O4 with Me as a metal or metals, there are 125,000 possible combinations assuming a maximum of three metallic elements. In this work, we focus on capacity for our predicted property, and increase the number of possible combinations to two million by incorporating processing into the modeling. Due to the non-linear relationship between processing and property, as well as ensuring proper sensitivity, a new approach which tracks the change with changing processing is introduced. This work provides an invaluable tool for guiding the next generation of battery experiments, providing a significant narrowing of the infinite design space.

Source: arXiv (Cornell University)

DOI: 10.48550/arxiv.2203.08699

Authors: Broderick, Scott R., Miyamoto, Kaito, Rajan, Krishna

Institutions: University at Buffalo, State University of New York, Toyota Motor North America (United States), Toyota Motor North America Research & Development (United States)

Language: English

Topic: Machine Learning in Materials Science

Subfield: Materials Chemistry

Field: Materials Science

Domain: Physical Sciences

Open Access status: green

Funders: National Science Foundation, University at Buffalo

Award: CIF21 DIBBs: EI: Data Laboratory for Materials Engineering

OpenAlex ID: W4221150792

Cites:

Cited by:

Related to: 10