The global demand for hydrogen has been rising in recent years, with a 22% increase to 115 million tons per year expected by 2030. (1,2) Hydrogen gas is currently produced primarily via steam-methane reforming, which results in the generation of carbon monoxide and carbon dioxide. (3,4) Water electrolysis driven by renewable electricity (e.g., wind or solar power) for hydrogen production is a cleaner alternative. (5,6) However, ultrapure water is usually needed for water electrolysis, which is often not readily available in areas where renewable energy sources are abundant, like coastal regions, as salt water is prevalent in those regions. (7,8) Direct seawater electrolysis is problematic due to the evolution of corrosive chlorine gas at the anode, which can shorten the life span of the electrolyzer setup. (9,10) One potential solution is to pretreat seawater to produce ultrapure water for electrolysis, but this increases the complexity of hydrogen production. (11,12)

An asymmetric water electrolyzer configuration has recently been proposed that uses fully oxidized salts (e.g., sodium perchlorate) as contained anolyte and seawater as the catholyte to prevent chlorine evolution. (13) Low-cost (less than $10 m^–2) (13) polyamide (PA) thin-film composite (TFC) reverse osmosis membranes were used instead of more expensive cation exchange membranes ($250 to $500 m^–2) (13) commonly used in acidic water electrolyzers to control the ion exchange between the catholyte and anolyte. (13,14) These PA TFC membranes exhibit high resistance to salt permeation due to steric exclusion. (15,16) Thus, applying TFC membranes to seawater electrolysis helps to limit the passage of chloride ions (Cl^–) into the anolyte and the leakage of the anolyte. Water ions, H⁺ and OH^–, can easily transport through TFC membranes, which can decrease the membrane charge transfer resistance. (17,18)

One major concern is that the small amount of Cl^– ions passing through TFC membranes during seawater electrolysis still needs to be decreased to minimize chlorine generation. (13,19) Another concern is the energy loss due to the generation of a pH gradient when using TFC membranes and salty water. During electrolysis, H⁺ ions are generated in the anode and OH^– ions are produced in the cathode. Although TFC membranes favor the transport of water ions over salt ions, some cross-membrane charge is still carried by salt ions, increasing the concentrations of H⁺ in the anolyte and OH^– in the catholyte. This pH difference can contribute to the voltage increment during electrolysis, referred to as the Nernstian overpotential. (20,21) Using a membrane with a more selective PA active layer with negative charge was recently shown to moderately reduce the permeation of Cl^– ions but have little effect on the transport of counterions (Na⁺) during electrolysis. (22) The Nernstian overpotential remained nearly constant and thus did not further decrease the energy demands. Therefore, other methods are needed to further reduce Cl^– ion transport in TFC membranes, ideally along with the deceleration of Na+ ion permeation.

In this study, we examined the use of a double-polyamide TFC membrane structure for reducing the permeation of chloride ions during water electrolysis (Figure 1). It was hypothesized that the negatively charged polyamide facing the alkaline catholyte would decrease the transport of negatively charged Cl^– ions into the anolyte, and the positively charged polyamide facing the acidic anolyte would reduce the transport of Na⁺ ions into the catholyte. Hence, having a polyamide layer on both sides of the membrane would favor the transport of water ions due to their higher permeability in the PA layer. In addition, the rapid water ion transport can be maintained through very fast water ion association to form water molecules.