(739b) Computational Investigation of Water Desalination across Nanofiltration Membranes Using Advanced Sampling Techniques | AIChE

(739b) Computational Investigation of Water Desalination across Nanofiltration Membranes Using Advanced Sampling Techniques

Authors 

Malmir, H. - Presenter, Yale University
Epsztein, R., Yale University
Elimelech, M., Yale University
Freshwater availability and sustainability is a major challenge facing humanity in upcoming decades. A potent approach to address this challenge is seawater desalination, which, despite its several outstanding issues, has the potential to provide a sustainable framework for freshwater production. Currently, membrane processes such as reverse osmosis (RO) and nano filtration (NF) are the most widely used desalination approaches both in terms of installed capacity and annual growth. Membrane separation normally requires applying driving forces such as pressure or electric field in order to overcome the natural osmotic pressure gradient between pure and salty water. There is, however, significant room for improvement of the existing technologies both in terms of energy efficiency and economic affordability, making studies of filtration an active area of research.

In this project, we use molecular dynamics (MD) simulations and advanced sampling techniques to investigate ion transport across nano-porous graphene membranes. The two most important metrics for efficiency of a water desalination process are the water flow rate and the salt rejection. We are particularly interested in using MD and a path sampling technique known as forward-flux sampling (FFS) to predict the flux of water molecules and various anions such as Chloride (Cl-) and Nitrate (NO3-) and to inspect their dependence on process variables such as temperature and the applied pressure gradient. We will also investigate the effect of functionalization and hydrophobicity of nanopores on the flow rate and salt rejection. Using an advanced sampling technique such as FFS enables us to estimate fluxes that are too small to quantify in a conventional MD simulation, and can therefore be utilized for rational design of more efficient filtration processes.