Extended Data Fig. 1: In situ seawater splitting of SES and conventional seawater direct splitting.

a, A schematic diagram of the lab-scale SES. b, Photos of the lab-scale SES and operation process. c, Ion chromatography tests show that the gas path can prevent seawater penetration, so the ion content in SDE is still nearly four orders of magnitude lower than that in seawater after long-term electrolysis. d, The energy consumption analysis. From the whole period, assuming that the water source is seawater, it is necessary to desalination before use in industrial alkaline electrolysis, which needs to consume at least 9–14.4 kJ \({\rm{k}}{{\rm{g}}}_{{\rm{water}}}^{-1}\), while the phase transition of SES is a spontaneous process, which saves the energy of desalination. During electrolysis, the energy input of our strategy is equivalent to industrial alkaline water electrolysis when the system conditions are the same, which has been confirmed above. e, Electrolysis durability test of conventional direct seawater (Shenzhen Bay seawater) splitting with commercial electrocatalysts. The inset shows photos of clear seawater before electrolysis, precipitation in seawater during electrolysis, and catalyst electrode dissolving and shedding in seawater after electrolysis.