Machine Learning the Voltage of Electrode Materials in Metal-Ion Batteries

• DOI: 10.1021/acsami.9b04933
• Publication Date: 2019-04-30T09:39:44Z
• Authors: Rajendra P. Joshi; Jesse Eickholt; Liling Li; Marco Fornari; Veronica Barone; Juan E. Peralta
• Keywords: Condensed Matter - Materials Science; Materials Science (cond-mat.mtrl-sci); FOS: Physical sciences; 01 natural sciences; 7. Clean energy; 0104 chemical sciences

Machine-learning (ML) techniques have rapidly found applications in many domains of materials chemistry and physics where large data sets are available. Aiming to accelerate the discovery for battery applications, this work, we develop a tool ( http://se.cmich.edu/batteries ) based on ML models predict voltages electrode metal-ion batteries. To end, use deep neural network, support vector machine, kernel ridge regression as algorithms combination with taken from Materials Project database, well feature vectors properties chemical compounds elemental their constituents. We show that our predictive capabilities different reference test and, an example, utilize them generate voltage profile diagram compare it density functional theory calculations. In addition, using models, propose nearly 5000 candidate Na- K-ion also make available web-accessible that, within minute, can be used estimate any bulk material number metal ions. These results is promising alternative computationally demanding calculations first screening novel applications.