(720e) Coupling Electrodialysis With a Novel Carbon Nanotube Topology Structure for Desalination: An Idea Demonstration

Carbon nanotube (CNT) has been one of the research spotlights in recent years. By virtue of the several orders of magnitude faster water flow than macroscopic hydrodynamics and the impressive anti-fouling properties, aligned CNT membrane has become a super-star candidate for the next-generation reverse-osmosis (RO) desalination membrane materials. However, the shadow casting over the promising application of CNTs lies in its lack of intrinsic water/ion selectivity. Despite many significant research attempts by using steric hindrance or electrostatic interactions, researchers failed to achieve high-flux and high-selectivity at the same time, either because of the difficulty to synthesize small-diameter CNTs or the reduction of water flux once entrance of CNTs is functionalized. To obtain both high selectivity and flux, we design a novel membrane, which is filled with honeycomb-shaped CNT, and demonstrate the feasibility of desalination with high flux and selectivity by using molecular dynamics simulation. In operation, an external electric field vertical to the flow direction is applied to serve as the driving force for the spatial segregation of ions and water. In our honeycomb-CNT model, the diameters of all the CNTs are identical, ranging from 1.3 nm to 2.0 nm in different simulations. Each separation stage contains two Y-junctions, with a unit length as long as 5 nm. There are four CNT entrances and outlets on the graphene layer which functions as the membrane surface, resulting in the pore density around 6.7×10¹² pores/cm². Concentrated streams, consisting of Na⁺ and Cl^-, flow through the two outlets near anode and cathode, respectively, while the dilute streams flow through the two outlet streams in the middle. The ion concentration of the incoming stream is set to be 1-2 mol/L and the hydrostatic pressure at different side of the membrane is 400 MPa to enable more permeation events observed in 60 ns simulation. Under the electric field of 0.2-0.6 V/nm, desalination is successfully realized: the ion permeation ratio, defined as the ratio between ion concentrations of the dilute and the incoming streams, is close to 0%, indicating pure water is obtained in the middle outlet. In addition, the selectivity, defined as the ratio between ion concentrations of the concentrate and the dilute streams, can be as high as 5.5±1.1. For better understanding of separation mechanism, further analysis was carried out. It is demonstrated that, first of all, low ion flux in the dilute streams is attributed to both steric hindrance and electric-field driven migration, which is consistent with our initial guess. Secondly, free energy barrier for ions to move upwards or downwards at the Y-junction is approximately 5 kcal/mol, which is the major driving force for the spatial distribution of ions. Thirdly, the importance of vertical ion transport is further demonstrated in the aligned CNT model, in which the lack of vertical connectivity leads to poor ion rejection performance even if the vertical electric field is applied. Finally, faster water transport in the concentration streams is subjected to examination, which probably resulted from the electro-osmosis effect. In conclusion, this hybrid approach can solve problems caused by both traditional electrodialysis (ED) and CNT membrane. On the one hand, CNT membrane is equipped with additional driving force for high selectivity. On the other hand, the lower ion concentration inside CNT membrane will significantly reduce the energy consumption of ED desalination. Notably, our model is not far from experimental practices. For example, techniques to synthesize sub-2-nm or Y-shaped CNTs are currently available, and applying electric field is much easier than charge modification. Therefore, our simulation provides an insight into the design of a more complex CNT membrane to achieve desalination.