Deep learning neural network potential for simulating gaseous adsorption in metal–organic frameworks

Abstract

This study proposes ab initio neural network force fields with physically motivated features to offer superior accuracy in describing adsorbate–adsorbent interactions of nonpolar (CO₂) and polar (H₂O and CO) molecules in metal–organic frameworks with open-metal sites. Effects of the neural network architecture and features are also investigated for developing accurate models.