Study of ice nucleation on silver iodide surface with defects

Abstract

In this work, we have considered the crystallisation behaviour of supercooled water in the presence of surface defects of varying size (surface fraction, α from 1 to 0.5). Ice nucleation on Ag exposed β-AgI (0001 plane) surface is investigated by molecular dynamics simulation at a temperature of 240 K. For systems with α > 0.67, the surface layers crystallise within 150 ns. In the system with defects, we observe two distinct stacking patterns in the layers near the surface and find that systems with AA stacking cause a monotonic decrease in the early nucleation dynamics with an increase in defect size.Where AB stacking (α = 0.833) is observed, the effect of the defect is diminished and the dynamics are similar to the plain AgI surface. This is supported by the variation in the orientational dynamics, hydrogen bond network stability, and tetrahedrality with respect to the defects. We quantify results in terms of the network topology using double-diamond cages (DDCs) and hexagonal cages (HCs). The configurations of the initially formed layers of ice strongly affect the subsequent growth even at long timescales. We assert that the retarded ice growth due to defects can be explained by the relative increase in DDCs with respect to HCs.

Atomistic Simulation of Ice Nucleation on Silver Iodide (0001) Surfaces with Defects

“…1 On most mineral dust particles, there are only a few active sites for ice nucleation, typically around defects or pits. [2][3][4] Silver iodide (AgI) particles are very effective INPs, used for rain seeding, 5 and they have been studied both experimentally (see the review by Marcolli et al 6 ) and computationally, focusing on ice nucleation on flat AgI surfaces, 7,8 AgI disks and plates, 9 the effect of surface charge distribution, 10 the effect of defect surface fraction, 11 as well as water adsorption in slit-like AgI pores. 12 In ambient conditions, both AgI crystals with the wurtzite structure (β-AgI) and the zincblende structure (γ-AgI) are stable.…”

d-SEAMS: Deferred Structural Elucidation Analysis for Molecular Simulations

“…• Topological network criteria, namely Double Diamond Cages, Hexagonal Cages and Mixed Cages [27,28] These criteria classify building blocks of Ic and Ih for deeply supercooled bulk water.…”

“…We used our code to reproduce production run results from our group [28,29] as well toy test systems inspired by the literature [6,26]. Below we demonstrate the code capabilities on two production systems.…”

“…Here we apply our code and the work-flow to a system of heterogeneous ice nucleation on a β-AgI surface with defects. The simulation parameters are taken from [28]. At an axial pressure of P zz = 0.1M P a, primarily hexagonal prism blocks are formed within the ICN in Figure 5(a).…”

A general topological network criterion for exploring the structure of icy nanoribbons and monolayers

“…Moreover, none of these parameters has been successfully applied to one-dimensional nanoribbon ices, although axial and angular order parameters for distinguishing between square and pentagonal ice nanotubes have been used [1]. There is no single straightforward classification technique for confined ordered ice-like water that minimizes human involvement and assumptions, and emphasizes automated, reproducible, quantifiable structural elucidation.Determining the connectivity of an ordered phase by using the primitive rings formed is a well-established technique of identification [22][23][24][25][26][27]. Ic (cubic ice) and Ih (hexagonal ice) are ice phases, which are similar enough that their structural differences are not trivially discernible by casual inspection.…”