Research

Engineering Sciences

Title :

Development of deep neural network potentials to accurately model water permeation through nanoporous 2D materials

Area of research :

Engineering Sciences

Focus area :

Computational science

Principal Investigator :

Dr. Kousika A, Indian Institute of science, Bangalore, Karnataka

Timeline Start Year :

2024

Timeline End Year :

2026

Contact info :

Details

Executive Summary :

The increasing global population, industrialization, and climate change have led to water shortages, necessitating innovative water desalination technologies. 2D materials like graphene, molybdenum disulfide (Mos2), and hexagonal boron nitride (hBN) are emerging as potential alternatives for seawater desalination. Nanoporous Mos2 is particularly attractive due to its superior properties, such as high water permeability and ion rejection rate. It is also mechanically stronger than graphene, making it more suitable for membranes. Experimental studies have been conducted on Mos2 nanosheets for nanofiltration, reverse osmosis, forward osmosis, and gas separations. Classical molecular dynamics (MD) simulations have been used to understand the design of Mos2 membranes, providing valuable information on water permeation and salt rejection. However, classical MD potentials often model water as a rigid nonpolarizable molecule, which can vary significantly with different models. To overcome this limitation, machine-learning (ML) potentials have been explored to describe water-2D material interactions with the accuracy of ab-initio methods and the efficiency of classical MD. This work aims to develop ML potentials for water-Mos2 systems using density functional theory calculations and apply them to water permeation through nanoporous Mos2.

Total Budget (INR):

Organizations involved