Ion-specific anomalous water diffusion in aqueous electrolytes: A machine-learned many-body force field study with...

Area of Science:

• Physical Chemistry
• Computational Chemistry
• Materials Science

Background:

• Water diffusion dynamics in electrolyte solutions exhibit ion-specific anomalies, with enhanced diffusion in chaotropic solutions and suppressed diffusion in kosmotropic solutions.
• Classical molecular dynamics simulations often struggle to accurately model these phenomena.

Purpose of the Study:

• To investigate the ion-specific anomalous diffusion of water in aqueous electrolyte solutions using advanced computational methods.
• To provide a quantitative improvement over previous simulation results for water dynamics in NaCl and CsI solutions.

Main Methods:

• Classical molecular dynamics simulations employing a many-body machine-learned force field (MACE equivariant graph neural network).
• The force field was trained on density functional theory (DFT) data using the revPBE-D3 exchange-correlation functional.
• Simulations covered NaCl and CsI aqueous solutions at concentrations from 0.89 to 3.56 mol/kg under ambient conditions.

Main Results:

• The simulations successfully reproduced the experimentally observed anomalous water diffusion in both NaCl and CsI solutions.
• Quantitative improvements were achieved compared to previous simulations, especially for NaCl solutions.
• The study identified specific ion-water interactions, such as strong Na+-water interactions and the influence of the second hydration shell, as key factors.

Conclusions:

• The machine-learned force field provides a more accurate description of water dynamics in electrolytes than previous methods.
• Anion I- primarily drives water acceleration in CsI solutions due to its weakly structured hydration shell.
• The findings offer a coherent microscopic understanding of water acceleration and retardation mechanisms in aqueous electrolytes.