Influence of Nanopore Size on the Performance of Salinity Gradient Power

Abstract

Retrieving energy based on a salt concentration difference between two liquid phases, such as sea water and fresh water, has emerged as one of the promising next-generation renewable energy techniques. Recent advances in nanotechnology and nanofabrication also make relevant applications attractive. According to the mechanisms involved, relevant results can be classified roughly into two technologies: pressure retarded osmosis and reverse electrodialysis (RED). The latter can be achieved by separating two liquids having different salt concentrations, for example, a fresh water and a sea water, by an ion-exchange membrane. Since the membrane attracts counterions and repels coions simultaneously, making the concentration of counterions inside the membrane higher than that of coions. Because it is easier for counterions to pass through the membrane, a net ionic flow is present so that Gibbs free energy is converted to electric energy. Considering the great potential of harvesting energy by RED, a detailed understanding of its performance under various conditions is highly desirable and necessary for design purposes. Among factors capable of influencing that performance, the size of a pore and its charged conditions are of particular significance. This is because the possible mechanisms involved, such as double layer overlapping, ion concentration polarization, and diffusioosmotic flow, are all related closely to those factors. In this study, we conduct a theoretical analysis on the influences of the size of a nanopore and its charged conditions on the performance of a RED system. Taking an aqueous KCl solution as an example, the results of numerical simulation reveal that for a positively charged nanopore, both the maximum power, P max, and the ion selectivity, η, have a local maximum as its radius, R, and its length, L, varies. A negatively charged nanopore, however, shows different behaviors. Regression analyses are performed to provide the dependences of both P max andη on R and L.