Structural and Dynamical Properties of Aqueous Solutions at Interfaces

Understanding the mechanism of ion and water adsorption at aqueous interfaces is relevant to a number of fields in science including heterogeneous and photo-catalysis, electrochemistry, and nanofluidics, among others. Concerning the field of nanofluidics, there have been remarkable developments recently, largely driven by the fabrication of two-dimensional materials, as well as by achieving control of liquid and ionic flow at the nano- and Ångstrom- scales. Osmotic transport in nanoconfined aqueous electrolytes provides alternative venues for water desalination and for harnessing the energy released upon mixing salty and fresh water, so-called “blue energy”. It is extremely challenging for experiments to provide atomic resolution into the structure and transport at the nanomembrane/fluid interface. Moreover, as dimensions are shrunk down to the atomic scale, it has become clear that the role of the electronic structure is increasingly relevant. Hence, understanding the interplay between the electronic structure and transport is key to develop efficient membranes for osmotic energy conversion.

Combining statistical mechanics approaches based on sampling by means of molecular dynamics simulations with the description of fluid dynamics according to the Stokes equation, we are able to address the diffusio-osmotic transport in nanufluidics systems, considering hydrophobic particles, concentrations of ions, and condition of electrified interfaces. To this purpose, we will carry out AIMD studies of aqueous systems at the interface with two-dimensional materials that exhibit desirable and novel electronic structure properties for water desalination and osmotic energy conversion. Because the number of possible combinations between aqueous interfaces and two-dimensional materials is close to being limitless, two-dimensional materials and van der Waals heterostructures represent an ideal playground to explore novel transport phenomena that are inherent to their electronic structure.

However, AIMD often struggles to reach the time scales and length scales required to properly converge some dynamic properties. On the other hand, traditional force field approaches are complicated to develop and often not accurate enough. Significant progress has been offered in the recent years by the development of a large number of approaches exploiting machine learning procedures. The resulting models can then be applied to widely explore the phase space at a nominally ab-initio level of accuracy. This approach enables the reliable description of aqueous interfaces under different working conditions using an efficient atom-based force fields which provide DFT-level accuracy.