Accelerated screening of cellulose additives via supervised latent representations

Abstract

We combined chemical descriptors with atomistic simulations to construct a physics-informed latent representation that predicts the effects of molecular and polymeric additives on regenerated cellulose films.