Investigating the properties of the water−vapor and water−graphite interfaces by PC-SAFT based density functional theory: A comparative study of different association functionals and PC-SAFT water parameter sets

Abstract

In this work, we apply the perturbed-chain statistical associating fluid theory (PC-SAFT) based classical density functional theory to investigate the thermodynamics and structures of the water−vapor and water−graphite interfaces. For the water−vapor interface, we examine the surface tension as well as structural characteristics such as density profiles and hydrogen bonding distributions along the interface. For the water−graphite interface, we investigate these same structural features at a contact angle of 90° at 300 K. Specifically, we focus on comparing the results obtained from the three association functionals and six PC-SAFT water parameter sets from literature with experimental data and molecular simulation results. We find that the three association functionals exhibit qualitatively similar behavior in the calculation of surface tension, and it is more influenced by the parameter set rather than the association functional. The optimal functional or parameter set for the calculation of density profile of the vapor − liquid interface remains inconclusive, whereas preferable parameter sets and functionals have been identified for the water−graphite interface. Furthermore, the three functionals demonstrate a remarkably similar and qualitatively accurate depiction of the hydrogen bonding structures at both interfaces when compared with molecular simulation results. The level of quantitative accuracy, however, relies on whether the parameter set accurately reproduces the average number of hydrogen bonds of bulk water as observed in this particular set of molecular simulation results. The results of the present work offer valuable guidance for further investigation into the adsorption and wetting behavior of water on graphite-like surface, as well as the properties of water confined at the nanoscale.