Heterogeneous ice nucleation on model substrates

TL;DR

The study investigates heterogeneous ice nucleation on model substrates with simple cubic, body-centered cubic, and face-centered cubic lattices exposing specific planes, using the mW water model to observe spontaneous nucleation. By computing heterogeneous nucleation rates $J_{het}$ from induction times and applying Classical Nucleation Theory, the authors extract the nucleation barrier $ riangle G^c_{het}$, the contact angle $ heta$, and the kinetic pre-factor, validating CNT down to small critical clusters. They find that certain planes, notably fcc-01$ar{1}$ and ice-like substrates, markedly promote nucleation and that the underlying lattice strongly influences nucleation efficiency and its sensitivity to lattice parameters. Structural analysis via RDF and $Q_6$ shows a direct link between interfacial water structuring and nucleation propensity, while microscopic trajectories confirm basal-plane nucleation and stacking faults, aligning with CNT-based predictions. Overall, the work provides a qualitative framework and a practical methodology to assess and compare ice-nucleating abilities of generic substrates, with implications for atmospheric science and materials design.

Abstract

Ice nucleation is greatly important in areas as diverse as climate change, cryobiology, geology or food industry. Predicting the ability of a substrate to induce the nucleation of ice from supercooled water is a difficult problem. Here, we use molecular simulations to analyse how the ice nucleating ability is affected by the substrate lattice structure and orientation. We focus on different model lattices: simple cubic, body centred cubic and face centred cubic, and assess their ability to induce ice nucleation by calculating nucleation rates. Several orientations are studied for the case of the face centred cubic lattice. Curiously, a hexagonal symmetry does not guarantee a better ice nucleating ability. By comparing the body centred cubic and the cubic lattices we determined that there is a significant role of the underlying crystal plane(s) on ice nucleation. The structure of the liquid layer adjacent to the substrate reveals that more efficient nucleants induce a more structured liquid. The most efficient substrates present a strong sensitivity of their ice nucleating ability to the lattice parameters. Introducing a novel methodological approach, we use Classical Nucleation Theory to estimate the contact angle of the ice nucleus on the studied substrates from the calculated nucleation rates. The method also provides the nucleation free energy barrier height, the kinetic pre-factor and the critical cluster size. The latter is in agreement with the nucleus size obtained through a microscopic analysis of the nucleation trajectories, which supports the validity of Classical Nucleation Theory down to small critical clusters.

Figures (12)

• Figure 1: Plane exposed to the liquid for the different substrates under study. In (a) the 100 orientation of the sc and bcc lattices. In (b), (c) and (d) the 100, 111 and 01$\bar{1}$ planes of the fcc lattice, respectively.
• Figure 2: a) Snapshot of a typical configuration of a liquid in contact with an sc-100-2.7 substrate. After simulating 15.1 ns at 221.5 K an ice nucleus appears (shown in (b)) and the whole system quickly turns into ice (shown in (c)).
• Figure 3: Contact angle of a nucleus on a substrate as a balance of the interfacial free energies involved.
• Figure 4: Potential energy vs time for 10 different trajectories at 221.5 K on an sc-100-2.7 substrate. The sudden steps are due to nucleation and growth of ice.
• Figure 5: Decimal logarithm of the nucleation rate versus temperature for different substrates having an nnd of 2.7 Å (see legend).