A Dual‐Cation Exchange Membrane Electrolyzer for Continuous H2 Production from Seawater

Abstract Direct seawater splitting (DSS) offers an aspirational route toward green hydrogen (H2) production but remains challenging when operating in a practically continuous manner, mainly due to the difficulty in establishing the water supply–consumption balance under the interference from impurity ions. A DSS system is reported for continuous ampere‐level H2 production by coupling a dual‐cation exchange membrane (CEM) three‐compartment architecture with a circulatory electrolyte design. Monovalent‐selective CEMs decouple the transmembrane water migration from interferences of Mg2+, Ca2+, and Cl− ions while maintaining ionic neutrality during electrolysis; the self‐loop concentrated alkaline electrolyte ensures the constant gradient of water chemical potential, allowing a specific water supply–consumption balance relationship in a seawater–electrolyte–H2 sequence to be built among an expanded current range. Even paired with commercialized Ni foams, this electrolyzer (model size: 2 × 2 cm2) continuously produces H2 from flowing seawater with a rate of 7.5 mL min−1 at an industrially relevant current of 1.0 A over 100 h. More importantly, the energy consumption can be further reduced by coupling more efficient NiMo/NiFe foams (≈6.2 kWh Nm−3 H2 at 1.0 A), demonstrating the potential to further optimize the continuous DSS electrolyzer for practical applications.

(with the applied electric field), the reduction range of NaOH pH value becomes smaller in contrast to the case without electric field (ΔpH = 0.10, 0.15, respectively), meaning that the electric field-enabled Na + migration can weaken the concentration gradientenabled Na + diffusion to some degree.This is most likely due to the competitive Na + permeation in the finite ion channels.Thus, the NaOH concentration can maintain relatively constant during the electrolysis.and CNaOH exhibits a good linear relationship with the equation of qinflux = 3.8*10 −6 CNaOH (R 2 = 0.99).Also, the D value in this system can be calculated as 4.8*10 −6 based on the known surface area of the adopted membrane (0.785 cm 2 ).Thus, the relation between qinflux and CNaOH can be described as qinflux = 4.8*10 −6 SCNaOH.According to this, the conditions for a continuous DSS system (qinflux = qoutflux) involving i and CNaOH can be obtained (i = 0.051SCNaOH).

Figure S3 .
Figure S3.Schematic of the continuous DSS system for H2 production.

Figure
Figure S4.a, The current at a cell voltage of 3.0 V as a function of ΔC.The data is derived from the corresponding LSV curves.b,The Ohmic drop of the electrochemical system as a function of ΔC.Each test at a ΔC was conducted three times, and the results were averaged.

Figure S5 .
Figure S5.Variation for Cl − ion concentration of pure water in the chamber of the H-type cell, wherein another chamber is filled with 0.6 M NaCl solution.The volume of the H-type cell is 30.0 mL, and the two chambers are separated by Gore CEM.The concentration of Cl − ion is determined by ion chromatography.The results indicate that a certain amount of Cl − ions can pass through the CEM after 96 h, but the corresponding concentration is still less than that of the 0.6 M NaCl by one order of magnitude.

Figure S6 .
Figure S6.UV-Vis absorbance spectra (a) and corresponding fitting curve (b) of the modified DPD assay for determining HClO/ClO − with known concentrations of 0, 20, 40, 60, 80, and 100 mM, wherein the fitting curve was constructed with the absorbances of samples at 550 nm and the corresponding concentrations.The insert is the images for the samples stained with DPD.

Figure S7 .
Figure S7.UV-Vis absorbance spectra for the electrolytes stained with DPD after the electrolysis at 0.25 A for 1 h in the range of ΔC from 0.2 to 13.4 M. The absorbances of electrolytes at 550 nm are less than 0.2, indicating that no HClO/ClO − products exist in the electrolytes.

Figure S8. a ,
Figure S8.a, Galvanostatic electrolysis curves of the DSS system with simulated seawater at different currents with the increase of electrolysis time.Cathode/anode: Ni foams.NaOH concentration: 1.0 M. CEM: Gore.ΔC = 0.6 M. b, The corresponding FEs of HER and OER during the galvanostatic electrolysis at different currents for 1 h.

Figure S9 .
Figure S9.In situ Raman spectra of Ni foams for HER (a) and OER processes (b) performed in 1.0 M NaOH under various potentials (vs RHE), respectively.

Figure S10 .
Figure S10.Image of the DEMS test system and the corresponding illustration for the adopted electrochemical device.

Figure S11 .
Figure S11.The volume of NaOH solution (2.9 M) before and after the constant current electrolysis at 1.0 A for 120 h.The test is conducted three times, and the corresponding standard error (SD) is calculated to be 4.6 mL.

Figure S12 .
Figure S12.pH value of NaOH electrolyte (1.0 M) in the dual-CEM three-compartment electrolyzer as the function of time with/without applied electric field (galvanostatic electrolysis at 1.0 A), wherein the seawater was continuously purged into the seawater chamber.Without the applied electric field, the pH value of NaOH electrolyte slightly reduces with the increase of the time (initial stage: 13.84; after 100 h: 13.69), indicative of the slight decrease of NaOH concentration.However, in the practical system

Figure S13 .
Figure S13.The optical image of the natural seawater from Qingdao Bay, Qingdao, China.Before use, the insoluble impurities of natural seawater were removed by filtering with polyethersulfone (PES) filtration membrane (pore size: 0.45 μm, ANPEL Laboratory Technologies).

Figure S14 .
Figure S14.The change for Mg 2+ /Ca 2+ ion concentration of pure water in the chamber of the H-type cell, wherein another chamber is filled with natural seawater.The volume of the H-type cell is 30.0 mL, and the two chambers are separated by Gore CEM.The concentration of Mg 2+ /Ca 2+ ions were determined by ion chromatography.The results indicate that very low concentrations of Mg 2+ /Ca 2+ ions can be detected in the pure water after standing for 96 h.

Figure S15 .
Figure S15.Galvanostatic electrolysis curves of the systems assembled with different CEMs at 0.25 A. NaOH concentration: 1.0 M.

Figure S16 .
Figure S16.Galvanostatic electrolysis curves of the DSS systems assembled with different CEMs at 4.0 V for 1 h.

Figure S17 .
Figure S17.X-ray diffraction (XRD) pattern for the white precipitation over the Selemion CEM, wherein the precipitation was collected by constant voltage electrolysis at 4.0 V for 3 h.

Figure S18 .
Figure S18.The images of the Neosepta, Fumasep, and Selemion CEMs at the initial stage, after the long-term electrolysis (1.0A for 100 h), and washed by acid solution, as well as their calculated dissolved fraction after the electrolysis.

Figure
Figure S19.a, b.Images for the external (a) and internal surfaces (b) of Selemion CEMs after the potentiostatic electrolysis at 4.0 V for 1 h, respectively.

Figure S20 .
Figure S20.Function of qinflux versus CNaOH in the system with seawater as the water source.The corresponding fitting curve of qinflux

Figure
Figure S21.a, Ohmic drops of the DSS system at the initial stage and after the electrolysis at 1.0 A for 120 h.b, c, The pristine Gore CEM (b) and the Gore CEM after the electrolysis at 1.0 A for 120 h (c).As shown in Figure S20c, a little precipitation can observed on the surface of the membrane.

Figure S22 .
Figure S22.The attenuated total reflection (ATR) spectra of the Gore membrane before and after the electrolysis.

Figure S23 .
Figure S23.The dry weight of the Gore membrane before and after the electrolysis as well as the calculated dissolved fraction after the electrolysis.

Figure S24 .
Figure S24.XRD patterns of pristine Ni foam and the Ni foams used as cathode (HER) and anode (OER) after the DSS in natural seawater at 1.0 A for 120 h.

Figure S25 .
Figure S25.The volume of NaOH solution (2.9 M) before and after the continuous DSS at 1 A for 120 h.The test is conducted three times, and the corresponding standard error (SD) is calculated to be 4.5 mL.

Figure S26 .
Figure S26.HER/OER FEs of the DSS system coupled with NiMo and NiFe foams under galvanostatic electrolysis at 1.0 A. It can be found that, the NiMo foam||NiFe foam system can also render ~100% Faradaic efficiencies for HER and OER, further demonstrating the practical superiority of our electrolyzer design concept.

Figure S28 .
Figure S28.The ohmic drop of the system with different membrane sizes of 2 × 2, 5 × 5, and 10 × 10 cm 2 .The tests were conducted with seawater and 1.0 M NaOH solution.

Figure S30 .
Figure S30.HER/OER FEs of scaled-up DSS systems under galvanostatic electrolysis at 1.0 A for 1 h.

Table S2 .
Concentration of the major constituents in seawater employed in this work.

Table S3 .
Comparison of H2 production performance of our DSS system with the state-of-art electrolysis systems in seawater.