The kinetics of the ice-water interface from ab initio machine learning simulations.

Molecular simulations employing empirical force fields have provided valuable knowledge about the ice growth process in the past decade. The development of novel computational techniques allows us to study this process, which requires long simulations of relatively large systems, with ab initio accuracy. In this work, we use a neural-network potential for water trained on the revised Perdew-Burke-Ernzerhof functional to describe the kinetics of the ice-water interface. We study both ice melting and growth processes. Our results for the ice growth rate are in reasonable agreement with previous experiments and simulations. We find that the kinetics of ice melting presents a different behavior (monotonic) than that of ice growth (non-monotonic). In particular, a maximum ice growth rate of 6.5 Å/ns is found at 14 K of supercooling. The effect of the surface structure is explored by investigating the basal and primary and secondary prismatic facets. We use the Wilson-Frenkel relation to explain these results in terms of the mobility of molecules and the thermodynamic driving force. Moreover, we study the effect of pressure by complementing the standard isobar with simulations at a negative pressure (-1000 bar) and at a high pressure (2000 bar). We find that prismatic facets grow faster than the basal one and that pressure does not play an important role when the speed of the interface is considered as a function of the difference between the melting temperature and the actual one, i.e., to the degree of either supercooling or overheating.

Minimum in the pressure dependence of the interfacial free energy between ice Ih and water.

Despite the importance of ice nucleation, this process has been barely explored at negative pressures. Here, we study homogeneous ice nucleation in stretched water by means of molecular dynamics seeding simulations using the TIP4P/Ice model. We observe that the critical nucleus size, interfacial free energy, free energy barrier, and nucleation rate barely change between isobars from -2600 to 500 bars when they are represented as a function of supercooling. This allows us to identify universal empirical expressions for homogeneous ice nucleation in the pressure range from -2600 to 500 bars. We show that this universal behavior arises from the pressure dependence of the interfacial free energy, which we compute by means of the mold integration technique, finding a shallow minimum around -2000 bars. Likewise, we show that the change in the interfacial free energy with pressure is proportional to the excess entropy and the slope of the melting line, exhibiting in the latter a reentrant behavior also at the same negative pressure. Finally, we estimate the excess internal energy and the excess entropy of the ice Ih-water interface.

On the thermodynamics of curved interfaces and the nucleation of hard spheres in a finite system.

We determine, for hard spheres, the Helmholtz free energy of a liquid that contains a solid cluster as a function of the size of the solid cluster by means of the formalism of the thermodynamics of curved interfaces. This is done at the constant total number of particles, volume, and temperature. We show that under certain conditions, one may have several local minima in the free energy profile, one for the homogeneous liquid and others for the spherical, cylindrical, and planar solid clusters surrounded by liquid. The variation of the interfacial free energy with the radius of the solid cluster and the distance between equimolar and tension surfaces are inputs from simulation results of nucleation studies. This is possible because stable solid clusters in the canonical ensemble become critical in the isothermal-isobaric ensemble. At each local minimum, we find no difference in chemical potential between the phases. At local maxima, we also find equal chemical potential, albeit in this case the nucleus is unstable. Moreover, the theory allows us to describe the stable solid clusters found in simulations. Therefore, we can use it for any combination of the total number of particles, volume, and global density as long as a minimum in the Helmholtz free energy occurs. We also study under which conditions the absolute minimum in the free energy corresponds to a homogeneous liquid or to a heterogeneous system having either spherical, cylindrical, or planar geometry. This work shows that the thermodynamics of curved interfaces at equilibrium can be used to describe nucleation.

Homogeneous nucleation of NaCl in supersaturated solutions.

The seeding method is an approximate approach to investigate nucleation that combines molecular dynamics simulations with classical nucleation theory. Recently, this technique has been successfully implemented in a broad range of nucleation studies. However, its accuracy is subject to the arbitrary choice of the order parameter threshold used to distinguish liquid-like from solid-like molecules. We revisit here the crystallization of NaCl from a supersaturated brine solution and show that consistency between seeding and rigorous methods, like Forward Flux Sampling (from previous work) or spontaneous crystallization (from this work), is achieved by following a mislabelling criterion to select such threshold (i.e. equaling the fraction of the mislabelled particles in the bulk parent and nucleating phases). This work supports the use of seeding to obtain fast and reasonably accurate nucleation rate estimates and the mislabelling criterion as one giving the relevant cluster size for classical nucleation theory in crystallization studies.

The Young-Laplace equation for a solid-liquid interface.

The application of the Young-Laplace equation to a solid-liquid interface is considered. Computer simulations show that the pressure inside a solid cluster of hard spheres is smaller than the external pressure of the liquid (both for small and large clusters). This would suggest a negative value for the interfacial free energy. We show that in a Gibbsian description of the thermodynamics of a curved solid-liquid interface in equilibrium, the choice of the thermodynamic (rather than mechanical) pressure is required, as suggested by Tolman for the liquid-gas scenario. With this definition, the interfacial free energy is positive, and the values obtained are in excellent agreement with previous results from nucleation studies. Although, for a curved fluid-fluid interface, there is no distinction between mechanical and thermal pressures (for a sufficiently large inner phase), in the solid-liquid interface, they do not coincide, as hypothesized by Gibbs.

Interfacial free energy of a liquid-solid interface: Its change with curvature.

We analyze the changes in the interfacial free energy between a spherical solid cluster and a fluid due to the change of the radius of the solid. Interfacial free energies from nucleation studies using the seeding technique for four different systems, being hard spheres, Lennard-Jones, and two models of water (mW and TIP4P/ICE), were plotted as a function of the inverse of the radius of the solid cluster. In all cases, the interfacial free energy was a linear function of the inverse of the radius of the solid cluster and this is consistent with Tolman's equation. This linear behavior is shown not only in isotherms but also along isobars. The effect of curvature on the interfacial free energy is more pronounced in water, followed by hard spheres, and smaller for Lennard-Jones particles. We show that it is possible to estimate nucleation rates of Lennard-Jones particles at different pressures by using information from simple NpT simulations and taking into account the variation of the interfacial free energy with the radius of the solid cluster. Neglecting the effects of the radius on the interfacial free energy (capillarity approximation) leads to incorrect values of the nucleation rate. For the Lennard-Jones system, the homogeneous nucleation curve is not parallel to the melting curve as was found for water in previous work. This is due to the increase in the interfacial free energy along the coexistence curve as the pressure increases. This work presents a simple and relatively straightforward way to approximately estimate nucleation rates.

Ice growth rate: Temperature dependence and effect of heat dissipation.

The transformation of liquid water into solid ice is arguably the most important phase transition on Earth. A key aspect of such transformation is the speed with which ice grows once it is nucleated. There are contradictory experimental results as to whether the ice growth rate shows a maximum on cooling. Previous simulation results point to the existence of such a maximum. However, simulations were performed at constant temperature with the aid of a thermostat that dissipates the heat released at the ice-water interface unrealistically fast. Here, we perform simulations of ice growth without any thermostat. Large systems are required to perform these simulations at constant overall thermodynamic conditions (pressure and temperature). We obtain the same growth rate as in previous thermostatted simulations. This implies that the dynamics of ice growth is not affected by heat dissipation. Our results strongly support the experiments predicting the existence of a maximum in the ice growth rate. By using the Wilson-Frenkel kinetic theory, we argue that such maximum is due to a competition between an increasing crystallization thermodynamic driving force and a decreasing molecular mobility on cooling.