(698g) Ultra Thin Layer-By-Layer Membranes: Insights from Atomistic Molecular Simulations

Authors: 
liyana-Arachchi, T., Pennsylvania State University
Colina, C. M., Pennsylvania State University
Sturnfield, J., The Dow Chemical Company

Reverse osmosis (RO) membranes based on polyamide (PA) thin film composites represents one of the most energy efficient membrane technologies used for producing potable water from seawater and brackish water nowadays. Specifically, cross-linked PA-RO membranes synthesized via interfacial polymerization (IP) of trimesoyl chloride (TMC) and m-phenylenediamine (MPD) are one of the most common membranes used in RO applications. However, typical RO membranes synthesized via IP have several drawbacks which impact the efficiency of the membrane. In particular, IP membranes lack control over (1) film thickness, (2) surface roughness, (3) cross-linkage, and (4) local chemical composition, which influence the membrane performance and efficiency. In an effort to overcome these challenges, ultrathin membranes synthesized via molecular layer-by-layer (mLbL) deposition have been introduced as an alternative to control structural and chemical properties.

The mLbL process has major challenge for commercial application, but does provide information that can be used for insights into the formation of RO membranes and their transport properties. Still these ultra-thin mLbyL membranes for RO applications have a complex structure and morphology which in addition to their thinness is very difficult to characterize experimentally. Atomistic molecular simulations can be used to examine local structural and chemical properties at a molecular level that support, complement and dramatically enhance these experimental methods to drive the discovery of new materials These computational analyses can provide information about pore structures, molecular orientations, local densities, thickness, and chemical compositions among others, offering a unique insight into understanding mLbL-RO membranes for several industrial applications.

In this work, we apply a general simulated polymerization algorithm developed in our group [Theor. Chem. Acc. 2013, 132, 1334] to synthesize and characterize in silico, RO membranes via mLbL deposition on silicon substrates. mLbL in silico membrane synthesis, consists of ‘building’ the membrane one  molecular layer at  a time via substrate deposition, “reaction”, and removal of excess solvated monomers. This approach utilizes a simulated polymerization algorithm to form networks by connecting repeat units of TMC and MPD deposited on a silicon substrate. The resulting simulated structures for the first ten layers are characterized by analyzing densities and densities profiles, membrane thickness, molecular orientations, pore size distributions, and local chemical compositions.

*Corresponding Author. E-mail: colina@matse.psu.edu