Abstract:
In the last years there has been an increasing interest in determining the homogeneous nucleation rate of ice, J, (number of critical solid clusters per unit of time and volume) which can be measured in experiments. From a computational point of view rigorous techniques to determine J are quite expensive. We have recently proposed a technique denoted as "seeding" to determine J. It combines elements of classical nucleation theory with simulations. After testing the method for simple systems as hard spheres or Lennard Jones, we have applied it to water. A number of interesting results have emerged from these simulation studies as the effect of pressure or the addition of salt on the nucleation of ice. Seeding can also be used to determine the competition between polymorphs. The interfacial free energy between ice and water is the key to understand ice nucleation. In the last part I will discuss some problems that has been discovered recently in the simulation of salts in water. It has been found that for most models the computed solubility seems to be too low. Also the impact of ions on transport properties seems to be too high. An interesting idea to overcome these difficulties is the use of scaled charges. We shall show that in general this scaling improves the description of the solution properties.