Ice-water interface

The molecular structure and dynamics of the ice/water interface are of interest, for example, in understanding phenomena like frost heaving, freezing (and the inhibition of freezing) in biological systems, and the growth mechanisms of ice crystals. In a series of simulations, Haymet and coworkers (see Refs. 193-196) studied the density variation, the orientational order and the layer-dependence of the mobilitity of water molecules. The ice/water basal interface is found to be a relatively broad interface of about... 

More recently, simulation studies focused on surface melting [198] and on the molecular-scale growth kinetics and its anisotropy at ice-water interfaces [199-204]. Essmann and Geiger [202] compared the simulated structure of vapor-deposited amorphous ice with neutron scattering data and found that the simulated structure is between the structures of high and low density amorphous ice. Nada and Furukawa [204] observed different growth mechanisms for different surfaces, namely layer-by-layer growth kinetics for the basal face and what the authors call a collected-molecule process for the prismatic system. 

In the latest experimental work131 not only the heat conductivities of ice and water, but also the latent heat of fusion were considered, but convection still disregarded. The importance of the refinements of the theory is clear from the comparison of the most recent value for the ysl of the ice - water interface, namely 29 erg/cm2, with the early result129) of 7si = 41 erg/cm2. The probable limits of error were given as 9 erg/cm2 in the early, and as 2 erg/cm2 in the later paper the former estimate appears to be too optimistic. For the interface of solid and liquid lead, 7S) = 76 erg/cm2 was calculated130). 

Figure 6.4. Block diagram of energy flows (calm 3 da 3) at the snow-ice-water interface.

O.A. Karim et al., The ice/water interface A molecular dynamics simulation using the simple point charge model. J. Chem. Phys. 92, 4634 1635 (1990)... 

With their method, Skapski et al. [ 1041 measured the surface stress of the ice-water interface to be 120 mN/m. This agrees with a more recent result of Hansen et al who report a value of I30 mN/m for the ice-water interface [ 105]. By using a combination of NMR and calorimetry to detect the melting, they extended the method to porous solids instead of a wedge. A surface stress around 120 mN/m is, however, surprisingly high. Earlier experimental and theoretical results were in the range of 10-35 mJ/m 1106. ... 

The calculations performed to date suggest that (i) the details of the short-range wall-water potentials dominate the average orientation of the contact layer of water (in combination with the strong water-water and charged wall-water interactions) and (ii) beyond the first layer the structure is insensitive to these details. The layered solvent structure extends approximately 15 A into the bulk of the liquid, representing about 4 layers of water molecules, a distance remarkably similar to that found in earlier simulations of an ice / water interface. The potential of zero charge is calculated to be -32 mV. The differential capacitance can be calculated, as discussed by Booth et al7... 

Abstract A molecular dynamics study of the Ih ice/water interface and the behavior of solute ions at the basal ice/water interface is presented. The excess stress at the interface of pure ice and pure water is discussed. We compare the solvation of Na+ and Cl- solute ions in bulk water and ice, and discuss their behavior at the ice/vacuum interface. The free energy profiles for Na+ and Cl ions across the ice/water interface is estimated from calculations of potential of mean force. 

Keywords Ice/water interface, interfacial excess stress, hydration structure,... 

Among the different two-phase liquid/crystal systems, ice/water interfaces are of great interest because of their fundamental presence in nature and importance in chemical, biological, environmental and atmospheric processes [14]. Systematic studies of ice/water interfaces by molecular dynamics simulations began in 1987, when Karim and Haymet [15] simulated for the first time the two-phase coexistence using the SPC model of water molecules. Since 1987, ice/water interfaces were studied with TIP4P [16], CF1 [17], SPC/E [18, 19] and six-site [20] models of water molecules. 

We simulated the SPC/E basal ice/water interface with Na+ and Cl- ions at a temperature of 225 K, which was estimated as the melting point for the SPC/E model of water in our previous study of two-phase coexistence [19]. We have used a collection of 2304 rigid water molecules plus a single solute ion in the NVT ensemble. The time step was chosen to be 1.5 fs. 

Figure 1. Mass-density profiles for ice/water interfaces (solid lines) basal interface (a) and (2110) interface (b). Translational order parameter f(z) changing from 1 in crystal bulk region to 0 in liquid phase is shown by short-dashed lines.

Other order parameters can be used to characterize ice/water interfaces via change in average density [17, 19] and diffusivity [17] across the interface. Local order parameters can also be defined which depend only on the position of a reference molecule. It was shown in [40] that the local tetrahedral order parameter, which reflects the change in tetrahedral environment, leads to a 10-90 width of 11 A for the basal ice-water interface which is in good agreement with the estimated value from the translational order parameter. 

Figure 2. Running averages of the excess stress a for basal ice surface (short-dashed line) and basal ice/water interface (long-dashed line). For comparison it is shown that in bulk ice (solid line) the excess stress is absent.

---