In our investigation of glass formation in metallic nanoparticles, we used an implicit solvent method for the solvent-nanoparticle interaction. Ideally, we would like to treat that interaction explicitly to probe the mechanism of solvent-nanoparticle interaction and to investigate the role capping agents play in that interaction. To facilitate these investigations we require a computationally efficient model for the water-metal and water-water interactions in such a system. One fast water model that can be used as a starting point for developing water-metal interactions is the soft sticky dipole model (SSD) of Chandra and Ichiye  as extended (SSD/E) by Fennell and Gezelter.[13-14] This model has the advantage of being a single point model requiring only one distance calculation for each pair of molecules and uses a “sticky” potential to mimic the tetrahedral hydrogen bonding in water. Our new model for the water-metal interaction uses a Morse potential for the van der Waals interaction similar to that of Spohr [15]  and a tetrahedral modulating function similar in principle to the tetrahedral potential in the SSD model. The total potential is then given by

This modulating function is designed to be attractive at the lone pair sites and repulsive at hydrogen sites. This is illustrated by the attractive cross-section in the cartoon at the right. The parameters for this potential are calculated using single-point ab-initio Density Functional Theory calculations for the water metal interaction conducted in the Schrödinger quantum chemistry package Jaguar. Once the potential energy surface is mapped, it can be fit using the functional form presented above.