Abstract

Water confined at the nanoscale is ubiquitous in geological, biological and technological environments, yet its molecular-scale behaviour remains poorly understood. Here, I will discuss our work combining first-principles electronic structure theory, machine learning and advanced statistical sampling to investigate water under extreme confinement. We find that monolayer water exhibits remarkably rich phase behaviour, including unconventional crystalline ices, a hexatic-like phase intermediate between a solid and a liquid, and a superionic phase with exceptionally high proton conductivity. Nuclear quantum effects strongly enhance proton transport and lower the conditions required to access the superionic regime. More generally, we show that subtle changes in confinement geometry and the surrounding material can strongly modify the hydrogen-bond network and phase behaviour of water. I will conclude by discussing methodological advances that are enabling increasingly realistic first-principles simulations of nanoconfined water and bringing direct comparison with experiment within reach.