Breaking bonds, breaking ground: Advancing the accuracy of computational chemistry with deep learning

Published 24 June 2025

A world-leading expert in high-level quantum chemical theories, Professor Karton collaborated with Microsoft Research to generate a massive dataset of highly accurate thermochemical properties — two orders of magnitude larger than any previous effort.

This landmark dataset provided the essential foundation for a breakthrough deep-learning density functional theory (DFT) developed by Microsoft.

Professor Karton's expertise in generating these highly accurate energies was critical for training the AI, paving the way for faster, more precise predictions in fields like drug design, solar panel recycling, next-generation batteries, and green fertilizers.

Read more about the achievement here: https://t.co/LuH7ZFgyVv