Accessing Ice–Water Interfaces with Critical Curvature for Investigating Ice Nucleation at Low Supercooling

Abstract Formation of critical nuclei is the essential step of first-order phase transitions such as the crystallization of supercooled water. However, probing these nanometer-sized critical nuclei remains extremely challenging in both experiments and simulations because they form rarely and exist transiently. Here we introduce a general simulation approach, also amenable to experimental realization, that enables spatiotemporally controlled formation and long-term stabilization of critical-sized ice nuclei. By covering a crystalline ice substrate with graphene membranes containing a single nanopore, we show that a spherical-cap ice nucleus forms barrierlessly on the nanopore and coexists with supercooled water over macroscopic time scales at temperatures above a pore-size-dependent threshold but rapidly grows once the temperature falls below this threshold. The resulting (meta)stable ice–water interface provides direct access to the thermodynamic and kinetic properties of critical ice nuclei, allowing the critical radius, interfacial free energy, molecular attachment rate, and nucleation rate to be determined using standard, unbiased simulations without any additional rare-event sampling. Beyond simulations, our results suggest a feasible experimental strategy for directly probing stabilized critical nuclei and their associated ice–water interfaces at low supercooling conditions that are otherwise difficult to access. This framework overcomes a long-standing barrier in both simulation and experimental studies of ice nucleation and is readily extendable to other first-order phase-transition processes.