Controlling ion transport in thin film composite (TFC) and ion exchange membranes to enable the use of impaired (salty) water in water electrolyzers

The water used in electrolyzers for green hydrogen gas production currently must be highly purified to remove salt ions, especially chloride which can react at the anode to form toxic and highly oxidizing chlorate gas. The cost of treating the water to these high levels is not a large part of operational costs, but the need for ultrapure water increases the complexity of the operation and capital costs. We are investigating different approaches for enabling the use of salty water, such as seawater, in water electrolyzers (WE). One general approach is based on using relatively inexpensive thin film composition (TFC) membranes produced in large quantities for seawater desalination by reverse osmosis (RO), compared to ion exchange (IX) membranes in conventional WE. TFC membranes are highly effective in containing salt ions (e.g. Na⁺ and Cl^–) but they are relatively permeable to water ions (H⁺ and OH^–). Predicting water and salt ion transport in TFC membranes based on past research is difficult as operational conditions for WE are much different than for RO. In WE ion motion is predominantly driven by the electric field which will greatly enhance ion motion in the direction of the field favorable for charged ions, and large pH gradients rapidly develop that are large driving forces for water ion transport, compared to RO operational conditions where ion transport is driven by concentration gradients, the pH is near neutral, and there is a high water molecule flux due to the high pressure gradients.

Ohmic losses using certain TFC membranes for WE can be reduced to levels comparable to IX membranes and thus the energy needed for water splitting are similar for both systems. TFC membranes are composed of three layers: the polyamide active layer, the polysulfone supporting layer, and a polyester structural support layer (PFC). The PFC layer is not needed for WE due to the use of little or no pressure differences between the layers. However, the PFC layer can contribute to ohmic losses and thus it should be removed for WE applications. The large pH gradient using a TFC membrane drives water ions through the membrane, but this pH difference adds an energy loss (Nernst overpotential). Enlarging the pores in the active layer increases undesirable chloride ion transport through the membrane. However, we invented a novel way to improve reduce chloride ion transport by adding on a second active layer to the opposing face of the TFC membrane, creating a two-active layer membrane. Another approach to using impaired water is to use a conventional cation exchange membrane (CEM) together with a non-aqueous anode chamber (vapor anolyte) so that ion flux from the anode dominates ion transport and substantially reduces chloride ion transport to the anode. However, this configuration does not avoid the use of expensive IX membranes. Overall, our results show that there are innovative ways to manage ion transport to enable the use of salty solutions in WE through suitable changes in types of membranes, architectures, and operational conditions.