This article describes how researchers from Tokyo University of Science have applied machine learning to optimize the composition of transition metal oxides for sodium-ion batteries, an alternative to lithium-ion technology. They trained models on data from 100 sodium half-cell samples, allowing them to predict optimal compositions for high energy density. The machine learning model identified a specific composition—Na[Mn 0.36 Ni 0.44 Ti 0.15 Fe 0.05]O2—as particularly effective, which they validated experimentally. This approach, combining computational predictions with experimental testing, could streamline the discovery of more efficient, sustainable battery materials across the energy storage industry.
For more details, please continue reading the full article under the following link:
Please consult also the Quantum Server Marketplace platform for the outsourcing of computational science R&D projects to external expert consultants through remote collaborations:
#materials #materialsscience #materialsengineering #computationalchemistry #modelling #chemistry #researchanddevelopment #research #MaterialsSquare #ComputationalChemistry #Tutorial #DFT #simulationsoftware #simulation