Lizhen Wu†
a,
Yifan Xu†b,
Qing Wanga,
Xiaohong Zoua,
Zhefei Pan*cd,
Michael K. H. Leung
*b and
Liang An
*ae
aDepartment of Mechanical Engineering, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong SAR, China. E-mail: liang.an@polyu.edu.hk
bAbility R&D Energy Research Centre, School of Energy and Environment, City University of Hong Kong, Kowloon, Hong Kong SAR, China. E-mail: mkh.leung@cityu.edu.hk
cKey Laboratory of Low-grade Energy Utilization Technologies and Systems (Chongqing University), Ministry of Education, Chongqing, 400030, China. E-mail: zhefei.pan@cqu.edu.cn
dInstitute of Engineering Thermophysics, Chongqing University, Chongqing 400030, China
eResearch Institute for Smart Energy, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong SAR, China
First published on 28th April 2025
Direct seawater electrolysis (DSE) is a promising technology for sustainable hydrogen production, utilizing abundant marine resources. However, industrialization of DSE faces significant long-term stability challenges due to the complex composition of seawater, which contains various ions and microorganisms that can lead to both chemical and physical degradation of the electrolysis system. For instance, the presence of chloride ions (Cl−) hinders the desired oxygen evolution reaction (OER) because competing chlorine evolution reactions (CER) occur and adversely impact electrode materials, resulting in low system efficiency and poor longevity. To enhance long-term stability of DSE, researchers are investigating robust electrocatalysts and advanced surface modifications that improve protection against corrosive environments and enhance selectivity. Innovative electrode designs are also being developed to manage bubble transport and decrease precipitation. Additionally, the design of electrolysis cells, such as bipolar membrane cells, offers a viable solution by minimizing Cl− transport and corrosive environment. With an increasing number of offshore renewable energy projects, the integration of effective DSE technologies in the offshore environment is critical. This review provides the state-of-the-art of electrodes, cells and systems, contributing to the development of DSE for long-term stable operation.
Broader contextDirect seawater electrolysis (DSE) is emerging as a critical technology for sustainable hydrogen production, leveraging the abundant supply of seawater to address global energy demands and reduce reliance on fossil fuels. Despite its potential, the widespread application of DSE faces significant challenges due to the complex chemical composition of seawater, particularly the presence of chloride ions (Cl−), which can lead to competing reactions like chlorine evolution that degrade the performance of electrocatalysts. Ongoing research focuses on developing robust electrocatalysts, advanced surface modifications, and innovative electrode designs, including novel cell configurations, to enhance the resilience and efficiency of electrolysis systems. As offshore renewable energy projects increase, integrating DSE technologies is vital for sustainable hydrogen production and achieving environmental sustainability goals, paving the way for a more resilient and renewable energy future. |
SWE is generally categorized into two types: indirect seawater electrolysis (ISE) and direct seawater electrolysis (DSE).16,17 DSE and ISE differ primarily in their pre-treatment requirements. ISE necessitates desalination methods, such as reverse osmosis or electrodialysis, to convert seawater into highly pure water before electrolysis, leading to additional costs and energy losses (up to 7% of the total system cost).18,19 In contrast, DSE directly utilizes untreated seawater, adding alkaline electrolyte (e.g., 1 M KOH) to suppress the chlorine evolution reaction (CER) and enhance OER.20 This approach results in alkaline effluent with residual ions (e.g., Cl−, Mg2+, Ca2+) and unreacted bases, complicating discharge management. Strategies for managing DSE effluent often involve diluting it with seawater or employing acid and base neutralization. However, these approaches can be detrimental to the marine environment and result in the chemical energy from the neutralization process being released as heat, offering little economic advantage.21,22 Alternatively, the recirculation systems can filter precipitates (e.g., Mg(OH)2, Ca(OH)2) to minimize waste but require efficient filtration and periodic alkaline replenishment.23 Resource recovery options exist, such as extracting Mg(OH)2 and Ca(OH)2 for industrial uses (e.g. construction materials, CO2 sequestration)24 and hypochlorite for disinfection,25 enhancing process economics. DSE is increasingly favored over ISE due to advancements in electrocatalysis and electrolysis cell design,26 demonstrating competitive hydrogen production from offshore wind farms at distances over 40–50 km.27 Moreover, DSE's compatibility with fuel cells enables integrated energy conversion and storage, offering a pathway to harness surplus renewable electricity while providing clean drinking water and renewable energy to arid regions, supporting future energy transitions.28 However, the industrialization of DSE still faces challenges, particularly concerning catalyst efficiency and system stability, which are critical for large-scale hydrogen production.29 There exists a considerable body of research focused on three main categories of seawater electrolysis: natural seawater electrolysis (NSE), characterized by a pH of approximately 8.1 and without the addition of strong bases like NaOH or KOH;11 alkaline seawater electrolysis (ASE), wherein NaOH or KOH is added to adjust the alkalinity of the seawater;30 and simulated seawater electrolysis (SSE), which typically involves a 0.5 M NaCl solution, potentially modified with NaOH or KOH.31 Given that above techniques use either untreated seawater or only simple physical pre-treatment (e.g., filtration, disinfection), NSE and ASE can be categorized collectively as DSE. Both DSE and SSE-related research aims to provide valuable technical guidance for DSE.
In recent years, advancements in novel electrode materials,12,32 electrolysis cells,33,34 and system design35,36 have propelled the development of DSE (Fig. 1). The interplay between catalyst activity, stability, and selectivity must be carefully balanced, and various types of electrolysis cells require optimization based on specific electrode and catalyst design criteria.28 Additionally, optimizing the coupling configurations of renewable energy sources with electrolysis cells and auxiliary machines is crucial for enhancing overall system stability.
![]() | ||
Fig. 1 Possible challenges about long-term stability in DSE. Reproduced with permission.37 Copyright 2015, The Royal Society of Chemistry. Reproduced with permission.38 Copyright 2024, American Chemical Society. Reproduced with permission.39 Copyright 2024, Springer Nature. Reproduced with permission.40 Copyright 2024, Elsevier. Reproduced with permission.41 Copyright 2019, American Chemical Society. Reproduced with permission.42 Copyright 2023, ASME. |
In this review, we aim to elucidate the mechanisms underlying the poor stability associated with DSE technologies. We will first provide a detailed overview of the challenges that influence the long-term stability of DSE technologies. Subsequently, we will discuss effective approaches for enhancing the long-term stability of both OER and HER electrodes, encompassing material development, surface modifications, and structural optimization. Furthermore, we will conduct a thorough evaluation of the diverse electrolysis cell designs employed in DSE, including membrane-based and membrane-less configurations, as well as low- and high-temperature electrolysis systems. Additionally, we will analyze stability improvement strategies tailored for SWE systems coupled with intermittent renewable energy sources. Finally, we will summarize the critical stability challenges that require attention in the future development of DSE systems.
![]() | ||
Fig. 2 (a) The Pourbaix diagram of the DSE. (b) Maximum permitted overpotentials for OER electrocatalysts. Reproduced with permission.44 Copyright 2024, The Royal Society of Chemistry. (c) Pourbaix diagram of HER and the competitive reduction of impurity cations in seawater. Reproduced with permission.51 Copyright 2021, The Royal Society of Chemistry. (d) Main long-term stability challenges for DSE: chemical degradation and physical/mechanical degradation. |
Acidic:
Anode: 2H2O → O2 + 4H+ + 4e− (E° = 1.23 V) |
Cathode: 4H+ + 4e− → 2H2 (E° = 0.00 V) |
Alkaline:
Anode: 4OH− → O2 + 2H2O + 4e− (E° = 0.40 V) |
Cathode: 4H2O + 4e− → 2H2 + 4OH− (E° = −0.83 V) |
However, due to the complex composition of seawater, challenges associated with DSE including potential side reactions, Cl− (∼0.54 M) plays a major hindering role in the electrolysis process.54 In the context of the reaction dynamics, the substantial presence of Cl−, paired with the limited availability of OH− and the analogous thermodynamic overpotentials associated with the CER and the OER, positions CER as a formidable side reaction at the anode. Under acidic or alkaline conditions at room temperature, competitive CER are described respectively:55
Acidic:
2Cl− → Cl2 + 2e− (E° = 1.36 V) |
Alkaline:
Cl− + 2OH− → ClO− + H2O + 2e− (E° = 0.89 V) |
It is important to highlight that, according to thermodynamics, the OER is somewhat superior to the CER across a generalized pH range.28 Dionigi et al.46 conducted an in-depth analysis of Pourbaix plots that incorporated chloride and oxygen chemistry (Fig. 2(a)). They discovered that the potential difference between the CER and the OER is influenced by the pH of the electrolyte. Specifically, at pH < 3, the potential difference ranges from 180 to 350 mV, whereas in alkaline conditions (pH > 7.5), this potential difference can reach as high as 490 mV (Fig. 2(b)). This increased potential difference favors the selective occurrence of the OER.56 However, the kinetics of the four-electron OER is considerably slower than that of the two-electron CER. This disparity leads to significant overpotentials and reduces the thermodynamic advantage of OER, particularly in acidic conditions where the potential gap is at its narrowest.12 The result indicates that the CER is more advantageous compared to the OER, especially at high current densities. Therefore, based on the alkaline conditions inhibiting CER, alkaline water electrolysis (AWE) cell and anion exchange membrane water electrolysis (AEMWE) cell appear to hold greater promise for DSE compared to proton exchange membrane water electrolysis (PEMWE) cell.45 In addition, this may be the reason why some of the studies have taken ASE. The primary difference between NSE and ASE is that ASE employs natural seawater as the electrolyte and is supplemented with an alkaline additive (typically 1 M NaOH/KOH), precisely controlling the operating potential of ASE to shift the electrolysis conditions toward those of alkaline water electrolysis, which enhances the competitiveness of OH−.56 This effectively suppresses the competitive reactions and corrosive effects, thereby significantly improving the durability of electrolyzers. Unlike conventional water electrolysis where only H2O is consumed, enabling acids and bases to persist within the electrolyzer over extended operation to maintain required pH conditions (with only periodic replenishment of high-purity water needed), ASE systems face inherent limitations. The continuous replenishment of natural seawater in ASE leads to progressive ion enrichment in the electrolyte. Primarily, Ca2+ and Mg2+ ions persistently deplete OH− through precipitation formation, gradually lowering environmental pH and generating difficult-to-remove precipitates.48,49 Consequently, ASE operation necessitates a flowing electrolyte system with continuous alkaline seawater renewal, posing substantial commercialization challenges—particularly considering that DSE is economically viable mainly when deployed on offshore wind platforms where routine maintenance proves impractical.35,36 Additionally, the high price of strong alkalis (e.g., $800 per ton for KOH)1 is also a barrier to the widespread practical application of ASE. NSE has an undeniable cost advantage in practical applications. We consider that ASE is an improvement of NSE in DSE, and although the cost aspect needs to be improved, all related research aims to provide an effective indirect solution for the lasting development of DSE.
Catalysts | Electrolyte | j [mA cm−2] | η [mV] | Stability | Ref. |
---|---|---|---|---|---|
OER | |||||
NSE | |||||
Cr2O3–CoOx | Seawater | 400 | 760 | 230 h@160 mA cm−2 | 76 |
Mo3Se4–NiSe | Seawater | 10 | 166 | 50 h@500 mA cm−2 | 77 |
PtPd–Ti | Seawater | 130 | 600 | 12 h@2.0 V (vs. RHE) | 78 |
Zr–Co3O4/CP | Seawater | 100 | 570 | 160 h@100 mA cm−2 | 79 |
CuS@CoOOH | Seawater | 10 | 290 | 72 h@10 mA cm−2 | 80 |
HPS-NiMo | Seawater | 500 | 595 | 120 h@1.8 V (vs. RHE) | 81 |
P/RP-SNCF | Seawater | 10 | 340 | 100 h@10 mA cm−2 | 82 |
CoFeOF/nickel foam (NF) | Seawater | 100 | 280 | 145 h@400 mA cm−2 | 83 |
ASE | |||||
NiFe/NiSx-NF | 1.0 M KOH + seawater | 400 | 560 | 1000 h@400 mA cm−2 | 84 |
MnCo/NiSe/NF | 1.0 M KOH + seawater | 1000 | 460.2 | 200 h@500 mA cm−2 | 85 |
CoFe–Ni2P/NF | 6.0 M KOH + seawater | 500 | 304 | 600 h@500 mA cm−2 | 86 |
RuMoNi/NF | 1.0 M KOH + seawater | 1000 | 470 | 3000 h@500 mA cm−2 | 87 |
Ru/NiFeOOH | 1.0 M KOH + seawater | 500 | 330 | 400 h@100 mA cm−2 | 88 |
NiFeO–CeO2/NF | 1.0 M KOH + seawater | 1000 | 408 | 200 h@1000 mA cm−2 | 89 |
Ni3S2@NiFe LDH/NF | 1.0 M KOH + seawater | 1000 | 340 | 100 h@200 mA cm−2 | 90 |
NiMoN@NiFeN/NF | 1.0 M KOH + seawater | 1000 | 398 | 100 h@500 mA cm−2 | 91 |
Fe–Ni2Pv−x | 1.0 M KOH + seawater | 100 | 180 | 100 h@100 mA cm−2 | 92 |
Co/P–Fe3O4@IF | 1.0 M KOH + seawater | 100 | 290 | 100 h@500 mA cm−2 | 93 |
(NiFe)C2O4/NF | 1.0 M KOH + seawater | 100 | 280 | 600 h@1000 mA cm−2 | 94 |
NiFe–CuCo LDH | 6.0 M KOH + seawater | 100 | 259 | 500 h@500 mA cm−2 | 95 |
Fe2P/Ni3N/NF | 1.0 M KOH + seawater | 1000 | 340 | 40 h@500 mA cm−2 | 96 |
Ru–Ni(Fe)P2/NF | 1.0 M KOH + seawater | 1000 | 361 | 50 h@600 mA cm−2 | 97 |
Fe2P/Ni1.5Co1.5N/N2P | 1.0 M KOH + seawater | 500 | 307 | 40 h@100 mA cm−2 | 98 |
Ni-doped FeOOH | 1.0 M KOH + seawater | 200 | 350 | 80 h@100 mA cm−2 | 99 |
Fe–Ni–O–N/NFF | 1.0 M KOH + seawater | 1000 | 289 | 40 h@500 mA cm−2 | 100 |
S-(Ni,Fe)OOH | 1.0 M KOH + seawater | 1000 | 462 | 100 h@100 mA cm−2 | 101 |
Ni2P–Fe2P/NF | 1.0 M KOH + seawater | 1000 | 431 | 48 h@100 mA cm−2 | 102 |
SSE | |||||
NiFe-LDH/NF | 1.0 M NaOH + 0.5 M NaCl | 100 | 270 | 180 h@400 mA cm−2 | 103 |
Ir/CoFe-LDH | 6.0 M NaOH + 2.8 M NaCl | 10 | 202 | 1000 h@800 mA cm−2 | 104 |
Co3−xPdxO4 | 1 M PBS + 0.5 M NaCl | 10 | 370 | 250 h@100 mA cm−2 | 105 |
Se_NiFe LDH | 1.0 M NaOH + 1.0 M NaCl | 400 | 670 | 600 h@100 mA cm−2 | 106 |
NiCoHPi@Ni3N/NF | 1.0 M KOH + 0.5 M NaCl | 100 | 365 | 120 h@200 mA cm−2 | 107 |
NiMoFe/NM | 1.0 M NaOH + 0.5 M NaCl | 100 | 241 | 1500 h@100 mA cm−2 | 108 |
FeCr–Ni3S2 | 1.0 M KOH + 0.5 M NaCl | 500 | 286 | 500 h@500 mA cm−2 | 109 |
F-NiFe-LDH | 1.0 M KOH + 0.5 M NaCl | 500 | 306 | 1000 h@1000 mA cm−2 | 110 |
HER | |||||
NSE | |||||
Mo5N6 | Seawater | 10 | 260 | 100 h@20 mA cm−2 | 111 |
CoxMo2−xC/MXene/NC | Seawater | 40 | 440 | 225 h@4 mA cm−2 | 63 |
Pt@mh-3D MXene | Seawater | 30 | 360 | 250 h@10 mA cm−2 | 112 |
Ti/NiCo | Seawater | 120 | 1000 | 10 h@-1.0 V (vs. RHE) | 113 |
PF–NiCoP/NF | Seawater | 10 | 287 | 20 h@-0.29 V (vs. RHE) | 114 |
Ptat–CoP MNSs/CFC | Seawater | 10 | 300 | 24 h@10 mA cm−2 | 115 |
Ni5P4 | Seawater | 10 | 144 | 10 h@100 mA cm−2 | 116 |
PtRuMo | Seawater | 100 | 800 | 20 h@-0.8 V (vs. RHE) | 117 |
CoMoC/Mxene/NC | Seawater | 15 | 500 | 225 h@-0.5 V (vs. RHE) | 63 |
Pt@mh-3D MXene | Seawater | 10 | 280 | 250 h@10 mA cm−2 | 112 |
PtNb–Nb2O5 | Seawater | 500 | 440 | 360 h@-0.44 V (vs.RHE) | 118 |
Pt/WO2 | Seawater | 10 | 290 | 500 h@100 mA cm−2 | 119 |
ASE | |||||
NiCoP–Cr2O3 | 1.0 M NaOH + seawater | 4000 | 275 | 10![]() |
120 |
F-FeCoPv@IF | 1.0 M KOH + seawater | 1000 | 210 | 20 h@100 mA cm−2 | 121 |
Cu2S@NiS@Ni/NiMo | 1.0 M KOH + seawater | 1000 | 250 | 2000 h@500 mA cm−2 | 122 |
Ru/MoO2@NiMoO4 | 1.0 M KOH + seawater | 1000 | 184 | 50 h@1000 mA cm−2 | 20 |
Ru/Cd0.02Se4 | 1.0 M KOH + seawater | 10 | 6.3 | 50 h@10 mA cm−2 | 123 |
Pt–Ni@NiMoN | 1.0 M KOH + seawater | 10 | 11 | 80 h@200 mA cm−2 | 124 |
Pt–Ni3S2/Co9S8-Sv | 1.0 M KOH + seawater | 18 | 10 | 300 h@100 mA cm−2 | 125 |
HW-NiMoN-2h | 1.0 M KOH + seawater | 1000 | 130 | 70 h@1000 mA cm−2 | 126 |
Pt–Ni3N@V2O3/NF | 1.0 M KOH + seawater | 10 | 21 | 500 h@500 mA cm−2 | 127 |
HPS–NiMo | 1.0 M KOH + seawater | 10 | 34 | 240 h@500 mA cm−2 | 128 |
Co/P–Fe3O4@IF | 1.0 M KOH + seawater | 500 | 261 | 100 h@500 mA cm−2 | 76 |
SA-MoO2/Ni3(PO4)2/NF | 1.0 M KOH + seawater | 10 | 46 | 10 h@100 mA cm−2 | 129 |
Ni2P–Fe2P | 1.0 M KOH + seawater | 1000 | 389 | 40 h@500 mA cm−2 | 102 |
NiMoN@NiFeN | 1.0 M KOH + seawater | 1000 | 218 | 100 h@500 mA cm−2 | 130 |
SSE | |||||
Ni–Co@Fe–Co PBA | 1 M KOH + 0.5 M NaCl | 183 | 10 | 24 h@50 mA cm−2 | 131 |
NiCoPv@NF | 1 M KOH + 0.5 M NaCl | 1000 | 237 | 50 h@100 mA cm−2 | 132 |
B,V-Ni2P | 1 M KOH + 0.5 M NaCl | 100 | 162 | 100 h@500 mA cm−2 | 133 |
NiS–FeS@IF | 1 M KOH + 0.5 M NaCl | 500 | 322 | 500 h@1000 mA cm−2 | 134 |
Ni-SN@C | 1 M KOH + 0.5 M NaCl | 10 | 23 | 24 h@10 mA cm−2 | 135 |
a-NiCoS/c-CeOx | 1 M KOH + 0.5 M NaCl | 10 | 79 | 50 h@50 mA cm−2 | 136 |
![]() | ||
Fig. 3 The element content changes in the surface oxide compositions of the different elements doped manganese dioxide-coated electrodes before and after electrolysis: (a) Fe and (b) V. (c) The durability of the different elements doped manganese dioxide-coated electrodes. (a)–(c) Reproduced with permission.138 Copyright 2012, Elsevier. (d) TEM image of the NiFe LDH/FeOOH sample. (e) The durability test of the NiFe LDH/FeOOH in alkaline NaCl solutions. (d) and (e) Reproduced with permission.139 Copyright 2021, American Chemical Society. (f) The durability test and the SEM images of Pt–Ru–Mo-decorated Ti mesh. Reproduced with permission.117 Copyright 2021, The Royal Society of Chemistry. (g) The durability test of Mo5N6. (h) The N K-edge XANES of Mo5N6 and MoN before and after HER. (i) Mo LIII edge of Mo5N6 before and after HER. Reproduced with permission.111 (g)–(i) Copyright 2018, American Chemical Society. |
In the field of HER catalyst, alloying platinum-based materials with transition metals such as Fe, Ni, or Mo has been shown to potentially improve their corrosion resistance. Li et al.117 showcased a Pt–Ru–M (where M denotes Cr, Fe, Co, Ni, or Mo) alloy on a titanium mesh, achieving commendable stability exceeding 172 hours for seawater HER (Fig. 3(f)). They elucidated that this enhanced stability was due to the preferential dissolution of the M species over that of Pt and Ru, given that M species were more thermodynamically inclined to dissolve in the presence of corrosive Cl−. Jin et al.111 reported that nitrogen-rich Mo5N6 exhibits impressive durability, lasting 100 hours at an overpotential of 300 mV in seawater (Fig. 3(g)). The material's structural stability is attributed to its abundant strong metal–nitrogen bonds and the high valence state of molybdenum, which enhances its resistance to detrimental Cl− (Fig. 3(h) and (i)).
There are two main ways to carry out the electrode surface modification, including the introduction of other anion ions84 by electrolyte additives/in situ generation and formation of a built-in electric field layer at the electrode–electrolyte interface to repel Cl− and the construction of Cl− a protective layer or blocking layer to decrease the local concentration of Cl− near the electrode. Firstly, Ma et al.103 offered a fresh perspective on surface sulfate additives in enhancing the stability of Ni-based electrodes in DSE. The presence of SO42− as an additive in the electrolyte tends to be preferentially adsorbed onto the anode surface. This adsorption creates an electrostatic repulsive force that effectively pushes Cl− away from the bulk solution. This repulsive influence of SO42− is also observed with NiFe LDH nanoarrays on nickel foam anodes, leading to stable performance at a current density of 400 mA cm−2 over 500 h in real seawater conditions. This stability is reported to be 3 to 5 times greater compared to operations conducted in an electrolyte lacking Na2SO4. Fan et al.142 developed a stable earth-rich LDH electrocatalyst for seawater electrolysis, achieving 2800 hours of operation at 2.0V versus RHE (Fig. 4(a)). Intercalating carbonate ions and anchoring graphene quantum dots improved corrosion resistance by reducing chloride ion adsorption (Fig. 4(b)). In addition, Liu et al.143 described a cobalt ferricyanide/cobalt phosphide (CoFePBA/Co2P) anode that can electrolyze alkaline seawater for more than 1000 hours at 1000 mA cm−2 (Fig. 4(c)) without corroding due to its self-generated PO43− and Fe(CN)63− layers repelling Cl− through electrostatic repulsion during operation (Fig. 4(d)). Tsao et al.106 electrodeposited NiFe LDH onto Se_NiFe foam. During the OER process in alkaline seawater, Se species were in situ oxidized to SeO3− or SeO4−, which could diffuse into the NiFe LDH layer, effectively repelling Cl− through electrostatic repulsion (Fig. 4(e)). Gong et al.144 developed a NiFe@DG core–shell nanocatalyst with a defective graphene coating that creates a built-in electric field at the electrode–electrolyte interface, repelling corrosive Cl− and protecting active metal sites. The electrolysis cell with NiFe@DG||Pt/C showed excellent performance in seawater with 1 M KOH, achieving current densities of 10 mA cm−2 and 100 mA cm−2 at low voltages of 1.496 V and 1.602 V, respectively, while operating stably for 1000 hours (Fig. 4(f)). Zhang et al.145 also created a strong built-in electric field at the (Ni, Fe) OOH/Ni12P5 interface, improving oxygen-evolution kinetics while reducing Cl− adsorption.
![]() | ||
Fig. 4 (a) Stability test of CoFe–Ci@GQDs/NF at 2.0![]() ![]() |
For the construction of a protective layer or blocking layer, Nocera's research group reports that coatings such as Fe2O3, MnO2 and PbO2 will help to inhibit Cl−.147 Wang et al.105 added a MnO2 protective layer to nickel foam to prevent Cl− corrosion, then deposited Co3−xPdxO4 on the modified substrate. After 450 hours of continuous operation at 200 mA cm−2 in natural seawater, the potential remained stable (Fig. 4(g)). Keisuke Obata et al.146 found that a CeOx layer deposited at the anode protects the NiFeOx electrocatalyst from iron loss, ensuring stable performance. This layer permits OH− and O2 to pass while blocking Cl−, without impairing the underlying activity of OER catalyst (Fig. 4(h)).
The primary challenges associated with HER electrocatalysts include corrosion resulting from the strong interactions between Cl− and Pt, as well as the precipitation of Mg2+/Ca2+ at the cathode, both of which significantly compromise performance and long-term stability. Surface modifications can be employed to mitigate the effects of Cl− corrosion and prevent the precipitation. Xu et al.122 developed a Cu2S@NiS@Ni/NiMo electrode demonstrating high HER activity with a 250 mV overpotential at 1000 mA cm−2 and exceptional durability over 2000 h at 500 mA cm−2 in 1 M NaOH and seawater (Fig. 5(a)). The NiS layer provides sulfur for protective polyanion-rich coatings, effectively resisting Cl− corrosion (Fig. 5(b)). Additionally, Ni3N@C/NF, featuring Ni3N nanosheets coated with a carbon shell, protects Ni3N from poisoning and corrosion by Cl−, achieving the sustained performance over 100 h in 1 M KOH and seawater (Fig. 5(c)).148 For the improvement of precipitation, OH− can be captured or limited, or Mg2+/Ca2+ can be repelled from reaching the electrode surface. Guo et al.11 showed that OH− can be captured by an introduced Lewis acid layer (creating a dynamic localized acidic environment) (Fig. 5(d)), such as Cr2O3, TiO2 and V2O3 layer proposed by Hu et al.127 Bao et al.119 exploited the in situ formation of hydrotungsten bronze (HxWOy) due to the reversible hydrogen insertion/excretion behavior of WO2 to act as a proton ‘sponge’ to store H+ and create an acid-like local environment (Fig. 5(e)), culminating in the proposal of a Pt/WO2 catalyst and realizing the long-term stability of >500 hours at 100 mA cm−2 in natural seawater electrolysis (Fig. 5(f)).
![]() | ||
Fig. 5 (a) Stability test of BDD//Cu2S@NiS@Ni/NiMo at 500 mA cm−2 in 1 M NaOH + seawater. (b) Time-dependent mass loss measurement of Cu2S@NiS@Ni/NiMo under 400 mA cm−2. (a) and (b) Reproduced with permission.122 Copyright 2023, Wiley. (c) The preparation process of NixFeyN@C/NF and stability tests at 100 and 500 mA cm−2. Reproduced with permission.148 Copyright 2021, The Royal Society of Chemistry. (d) Schematic illustrating the generation of a local alkaline microenvironment at the Lewis acid-modified cathode, enhancing HER while inhibiting precipitate formation. Reproduced with permission.11 Copyright 2023, Springer Nature. (e) Schematic comparison of the HER mechanisms on Pt/C and Pt/WO2. (f) Stability testing at 100 mA cm−2 over Pt/WO2 and Pt/C in natural seawater under Ar. Reproduced with permission.119 Copyright 2024, American Chemical Society. |
It was found that a heterogeneous bimetallic phosphide, Ni2P–Fe2P, on Ni foam is an effective bifunctional catalyst for seawater electrolysis through an in situ growth-ion exchange phosphidation method. The resulting porous, ultrathin nanosheets (7.4 nm) exhibited hydrophilic properties (Fig. 6(a)), enhancing specific surface area (Fig. 6(b)), electrolyte diffusion, and bubble release, which improved the electrocatalyst's stability at high current densities.102 In addition, a similar function was achieved by the multilayer crystalline-amorphous heterostructured electrode material NF/(CoMo)0.85Se@FeOOH proposed by Li et al.153 (Fig. 6(c)), featuring a porous architecture for enhanced active site exposure and mass transfer. This also emphasizes the importance of constructing porous structures and achieving ideal surface superhydrophobicity and superhydrophilicity to ensure adequate exposure to active sites and efficient mass transfer.154 The multilayered architecture, featuring a (CoMo)0.85Se core with a FeOOH shell and an in situ formed transition metal (oxy)hydroxide outer layer enriched with polyatomic anions (MoOxn− and SeOxn−), demonstrates excellent mechanical stability and resistance to chloride corrosion in harsh seawater conditions. Liu et al.155 proposed a design, Os–Ni4Mo/MoO2 micropillar arrays (Fig. 6(d)) with strong metal-support interaction (MSI), for seawater electrolysis. Strong MSI between Os and Ni4Mo/MoO2 enhances the catalyst's surface electronic structure, lowering the reaction barrier and boosting catalytic activity. In addition, this design not only enhances electron and mass transfer, but also creates a dual Cl− repelling layer using electrostatic force to protect active sites from Cl− attack in seawater oxidation, consisting of strong Os–Cl adsorption and an in situ-formed MoO42− layer. Consequently, it shows excellent stability, with a degradation of just 0.37 μV h−1 after 2500 hours of seawater oxidation. For the precipitation, Liang et al.156 developed a novel honeycomb-like 3D structure with longitudinal micro-channels for a microscopic bubble/precipitate traffic system, allowing for robust anti-precipitation in seawater (Fig. 6(e)). This ordered cathode generates uniform small-sized bubbles that aid in self-cleansing (Fig. 6(e)). Liu et al.157 designed a seawater HER electrode with a Ni(OH)2 nanofiltration membrane grown on nickel foam at room temperature (Fig. 6(f)). The positively charged Ni(OH)2 membrane, featuring nanometer-scale cracks, selectively hinders the transfer of Mg2+/Ca2+, reducing precipitation by 98.3% (Fig. 6(g)). This design ensures rapid of transfer OH− and H2O for enhanced HER activity and stability than the Nikel foam electrode. In terms of electrode–membrane interface, Frisch et al.152 adopted a combined catalyst coated membrane (CCM) and catalyst coated substrate (CCS) approach (a lower (2 mgLDH cm−2) and a higher (8 mgLDH cm−2) amount on the membrane (CCM) and on the PTL (CCS), respectively.), enhancing catalyst adhesion and reducing delamination at the interface. Since there are fewer studies on the improvement of membrane–electrode interface for DSE, we can draw on the relevant strategies in AEMWE. Wan et al.158 proposed a swelling-assisted transfer strategy to build ordered anode catalyst layers (ACL) on AEM (Fig. 6(h)). Specifically, a three-dimensional interlocked ACL/AEM interface, formed by direct membrane deposition method, can be perfectly transferred to the ordered ACL to AEM, thus permitting vertically orientated through-hole ACL structures and aligned ionic layer layers for OH− transfer. Stable operation was achieved for 700 h in pure water-feeding mode at ∼1.7 V with a current density of 1000 mA cm−2 due to the strong adhesion between the membrane and CL. Hu et al.159 presented an easy method to prepare patterned membranes by casting a polymer solution to the surface of commercially available monocrystalline silicon plates that have a pyramidal pattern on their surface (Fig. 6(i)). The prepared membranes show 39% permeability and 23% enhancement in electrochemical surface area as compared to flat membranes with the same catalyst loadings.
![]() | ||
Fig. 6 (a) Schematic illustration of the formation of Ni2P–Fe2P/NF and chronopotentiometric curves at 100 and 500 mA cm−2 for overall water/sweater splitting in 1 M KOH seawater. (b) SEM images of Ni2P–Fe2P/NF. (a) and (b) Reproduced with permission.102 Copyright 2020, Wiley. (c) Synthetic procedure for the NF/(CoMo)0.85Se@FeOOH nanosheet array and wetting characteristics of bare NF, NF/(CoMo)0.85Se, and NF/(CoMo)0.85Se@FeOOH. Reproduced with permission.153 Copyright 2023, American Chemical Society. (d) Schematic illustration of dual Cl− repelling layer and chronopotentiometry curves of Os–Ni4Mo/MoO2 for OER. Reproduced with permission.155 Copyright 2024, Wiley. (e) SEM images of the honeycomb-like 3D electrode and its mechanism of repelling precipitation. Reproduced with permission.156 Copyright 2024, Elsevier. (f) The number of Mg2+, Cl−, and H2O passing through the charged and uncharged Ni(OH)2 nanosheet. (g) The mass of precipitation on the Ni(OH)2–Pt–NF and Pt–NF electrodes. Reproduced with permission.157 Copyright 2023, The Royal Society of Chemistry. (h) Details of the novel swell-assisted transfer method. and schematic illustration of the general fabrication method of the 3D-ordered ACL in an MEA. Reproduced with permission.158 Copyright 2024, The Royal Society of Chemistry. (i) Preparation of patterned membrane and schematic diagram of membrane electrode with patterned membrane. Reproduced with permission.159 Copyright 2024, American Chemical Society. |
![]() | ||
Fig. 7 (a) The membrane-less electrolysis cell separates the produced gas utilizing Segré–Silberberg effect. Reproduced with permission.37 Copyright 2015, The Royal Society of Chemistry. (b) Mechanism of membrane-free divergent electrode-flow-through electrolysis. Reproduced with permission.161 Copyright 2015, Elsevier. (c) Membrane-less electrolysis cell utilizing angled mesh flow-through electrodes and a single-component device structure. (d) Solution resistance as a function of angle between electrodes and buildup of bubbles on the electrodes in stagnant solution and in a flowing electrolyte with fluid velocity of 13.2 cm s−1 during electrolysis at 2.5 V. (c) and (d) Reproduced with permission.162 Copyright 2016, The Electrochemical Society. (e) Detailed schematic of a passive membrane-less electrolysis cell utilizing buoyancy-driven product separation. (f) The comparison of hydrogen cross-over percentage is presented for both symmetric and asymmetric electrodes under different angles, with operations conducted at a current density of 20 mA cm−2. Reproduced with permission.163 Copyright 2023, The Royal Society of Chemistry. (g) Separation of gas products at an intermediate position and at the exit when the flow rate varies. Reproduced with permission.164 Copyright 2021, American Chemical Society. (h) Design concept for capillary-fed electrolysis cell and its 30 days stability test. Reproduced with permission.165 Copyright 2022, Springer Nature. |
In DSE, modifications to the electrode structure have been described in detail in Section 3.2. Next, we focus on the impact of operating modes and other key components on long-term stability. The typical mode of operation is to use a symmetrical feed of alkalized seawater, however, asymmetrical feeding may avoid the competitive CER at the anode. Strasser et al.166 describe an electrolysis cell that circulates neutral seawater at the cathode and pure KOH at the anode, demonstrating excellent corrosion resistance and OER selectivity. Starting at 1.7 V with a current density of 200 mA cm−2, the voltage rose by 8–10 mV every 12 hours, accumulating to about 100 mV over 100 hours (Fig. 8(a)). While the NiFe-LDH electrodes remained stable, the degradation was linked to the anode catalyst, collector, or membrane components. Although few studies have been done on the key components such as flow fields in DSE, there have also been some studies into structural designs to facilitate bubble removal in AMEWE using KOH or water.167 Duan et al.168 introduced a novel concave–convex mastoid polar plate that enhances gas–liquid distribution uniformity in the flow channel. Using a three-dimensional model for numerical simulations, they optimized the structure with a multi-objective genetic algorithm. The optimized design achieved an 8.40% improvement in electrolyte flow rate uniformity while maintaining a lower pressure drop compared to the original structure (Fig. 8(b)), which can be beneficial for the long-term stable operation.3 In addition, Shi et al.169 designed a pH asymmetric electrolysis cells (Fig. 8(c)) incorporating a Na+ exchange membrane for direct seawater decomposition to prevent Cl− corrosion by blocking Cl− transport to the anode, and mitigated Mg2+ and Ca2+ precipitation in near-neutral seawater (pH < 9.5). The output cathode electrolyte is maintained at pH 8.5 at 100 mA cm−2, showing a stable voltage response. After 120 hours of water-splitting, only a slight voltage increase is needed to reach 100 mA cm−2 after refreshing the solution, confirming the system's durability (Fig. 8(d)). Alternatives to the anode include electronic mediators or small organic molecules that can replace slow OERs, enabling reactions at low voltages to avoid OERs and potentially eliminate chlorine emissions,44,170 which is not what the review is exploring.
![]() | ||
Fig. 8 (a) Symmetric alkalinized 0.5 M NaCl feed and asymmetric 0.5 M NaCl feed at the cathode and 0.5 M KOH feed at the anode. Reproduced with permission.166 Copyright 2020, The Royal Society of Chemistry. (b) Machine learning aids the optimization process of flow field in AWE.168 Copyright 2024, Elsevier. (c) Asymmetric electrolysis cell with Na+ exchange membrane. (d) Long-term stability test of the asymmetric electrolysis cell with Na+ exchange membrane at 100 mA cm−2 and Cl− concentration in anode electrolytes. (c) and (d) Reproduced with permission.169 Copyright 2023, Springer Nature. (e) Schematic of an electrolysis cell with anode: seawater vapor (80% relative humidity) and cathode: dry N2 and current density versus time at an applied voltage of 1.6 V for different feedstock at the electrolysis tank anode. Reproduced with permission.171 Copyright 2016, The Royal Society of Chemistry. (f) Schematic of an electrolysis cell using a vapor feed at the anode and saltwater at the cathode and the effect of Na+ in solution on the overpotential caused by pH gradients. Reproduced with permission.172 Copyright 2021, The Royal Society of Chemistry. (g) Schematic of an electrolysis cell with BPM. Reproduced with permission.33 Copyright 2023, Elsevier. (h) Mechanism of seawater acidification via synergistic effects between inorganic precipitation on the cathode surface and proton flux from BPM. (i) pH variation of bulk seawater during 1000 hours of BPMWE operation at 20 mA cm−2, along with changes in concentrations of Mg2+ and Ca2+. (h) and (i) Reproduced with permission.173 Copyright 2023, The Electrochemical Society. |
Type | Advantages | Challenges |
---|---|---|
AWE3,9 | 1. Mature, commercially available | 1. Low efficiency and current density |
2. Lower material costs | 2. Gas crossover | |
3. Robust design | 3. Slow response to variable conditions | |
AEMWE3,9,45 | 1. High efficiency and current density | 1. Limited membrane durability |
2. Compact, modular design | 2. Low OH− conductivity | |
3. High-purity hydrogen production | 3. Precipitation blocking membrane and electrodes | |
4. Lower material costs | ||
PEMWE4,174,181 | 1. High efficiency and current density | 1. Expensive catalysts and components |
2. Compact, modular design | 2. More prone to CER | |
3. High-purity hydrogen production | 3. Ca2+/Mg2+, etc. displace H+ in membrane175 | |
BPMWE177,179 | 1. Separate anode/cathode environments | 1. Less mature |
2. Flexible anode/cathode pH control | 2. Complex membrane fabrication | |
3. Reduce Cl2 and precipitation |
![]() | ||
Fig. 9 (a) Examples of some typical degradation pathways of membrane materials. Reproduced with permission.28 Copyright 2020, Elsevier. (b) The schematic structure of high-molecular-weight hexyltrimethylammonium-tethered polycarbazoles (HQPC-TMAs) is presented, along with the electrolysis performance of HQPC-TMA-2.4 in 1.0 M KOH and 1.0 M KOH + seawater. Additionally, a 1000-hour durability test of the 1.0 M KOH electrolyzer, operated at 1.0 A cm−2 and 60 °C.184 Copyright 2024, The Royal Society of Chemistry. (c) Molecular structures of AEM: branched poly(terphenyl piperidinium) (b-PTP), AEI: Sustainion XB-7, and CEI: Nafion, along with a schematic representation of the various catalyst-PTL interfaces and stability tests at 10 A cm−2 (Pt/C as the cathode. The operational conditions included a cell temperature of 80 °C and 1 M KOH as the electrolyte). Reproduced with permission.185 Copyright 2024, Wiley. (d) Chemical structures of ionomers (PTPA, PBPA, and PFTA-20) and an illustration of CCS || CCM MEA configuration is presented. Additionally, a long-term durability test of AEMWE was conducted at a constant current density of 1.0 A cm−2 at 80 °C. Reproduced with permission.186 Copyright 2025, Elsevier. |
Ion | Concentration (mol kg−1 (H2O)) | Effect |
---|---|---|
Cl− | 0.56576 | CER (corrosion),45 reducing the conductivity166 |
Na+ | 0.48616 | No significant adverse effects reported |
Mg2+ | 0.05475 | Precipitate covering the membrane28 |
SO42− | 0.02927 | Affecting the OH− transport and reducing the conductivity45 |
Ca2+ | 0.01065 | Precipitate covering the membrane28 |
K+ | 0.01058 | No significant adverse effects reported |
HCO3− | 0.00183 | Concentration polarization,183 reducing the conductivity187 |
Br− | 0.00087 | Corrosion and poison29 |
CO32− | 0.00027 | Concentration polarization,183 reducing the conductivity187 |
F− | 0.00007 | Corrosion and poison29 |
Biofouling | NA | Poison and covering membrane188 |
OH− | 0.00001 | No adverse effects reported |
Improved electrode design strategies for degradation (bubble/precipitation coverage, CER corrosion, etc.) are also beneficial to the performance of the membrane in long-term operation, as described in detail in Section 3. In addition, some specific strategies have been proposed to address the above membrane degradation. Firstly, Kim et al.184 developed high-molecular-weight hexyltrimethylammonium-tethered polycarbazoles (HQPC-TMA-x) membrane that exhibited high ionic conductivity, mechanical robustness, and alkaline stability. HQPC-TMA-x also helps mitigate ionomer adsorption issues on electrodes due to the polycarbazole backbone. It demonstrated outstanding performance in both pure water and direct seawater electrolysis. HQPC-TMA-2.4 also showed impressive durability, maintaining a high current density of 1.0 A cm−2 for 1000 hours, with minimal irreversible degradation rates of 52 and 6 μV h−1 for platinum group metal (PGM)-free cells, respectively (Fig. 9(b)). In addition, from the point of view of enhanced membrane conductivity (strong OH− transport) and strong adhesion of the ionomer, Zheng et al.185 used a branched poly(terphenyl piperidinium) (b-PTP) AEM189 that proved to be robust, and a blended ionomer combining strong adhesive Nafion ionomer and strong OH− transport capacity of the Sustainion ionomer, which allowed the AEMWE to operate for more than 800 hours at 10 A cm−2 (Fig. 9(c)). Li et al.186 designed an anion exchange ionomer (AEI) featuring a non-rotatable fluorenyl backbone that minimizes phenyl adsorption and enhances oxidation stability, resulting in 1800 h with a low voltage of 1.6 V at 1.0 A cm−2. Notably, they proposed a hybrid MEA design, combining a coated-catalyst substrate anode and a coated-catalyst membrane cathode, addressing membrane swelling and enhancing MEA stability through strong bonding between the anode catalyst layer and the porous substrate (Fig. 9(d)). Although the above studies did not directly conduct long-term stability tests under seawater electrolysis conditions, they still provide guidance for the design of durable membranes and ionomers for DSE.
![]() | ||
Fig. 10 (a) Key components of the FOWS cell for use in contaminated water sources and their operation process. (b) The rates of water influx (red) and outflux (blue). (c) Stability test at 250 mA with a 0.8 M NaPi inner electrolyte solution and 0.6 M NaCl outer electrolyte solution. (a)–(c) Reproduced with permission.196 Copyright 2020, National Academy of Sciences. (d) The membrane divides the electrodes into two chambers, with desalinated water collected from the catholyte. Reproduced with permission.197 Copyright 2021, Elsevier. (e) Actual picture of the integrated stack of the AWE with porous PTFE membrane. Reproduced with permission.34 Copyright 2022, Springer Nature. (f) Electrolysis of seawater using a water phase-transition migration mechanism through porous PTFE membrane. (g) The average ion concentrations in the innovative DSE cell and Xinghua Bay seawater. (h) Water migration behavior of the porous PTFE membrane before and after 10 days of use. Reproduced with permission.39 Copyright 2024, Springer Nature. |
Based on the above research about the integration of in situ seawater purification technologies and water electrolysis technologies, Xie et al.34 developed an innovative DSE system for hydrogen production, operating at 250 mA cm−2 for over 3200 hours. This system integrates an in situ water purification mechanism that uses a self-driven phase transition process and seawater electrolysis. It features a hydrophobic, porous polytetrafluoroethylene (PTFE) membrane as a gas pathway, alongside a concentrated KOH solution acting as a self-dampening electrolyte (Fig. 10(e)). The difference in vapor pressure between seawater and the KOH allows for the evaporation of seawater, which then diffuses through the membrane and is re-liquefied, effectively filtering out contaminants like Cl−, Mg2+, and SO42−. This results in a continuous supply of pure water for hydrogen production from seawater. Due to the efficient isolation of impurities in the electrolysis cells, the integrated cell stack utilizes commercial MoNi/Ni foam for the anode and PtNi mesh for the cathode (Fig. 10(f)). Additionally, in collaboration with China Dongfang Electric Corporation, the team developed the “Dongfu-1,” the first offshore wind-driven hydrogen production platform without desalination.39 During sea trials in May 2023, it produced 1.3 N m3 h−1 of hydrogen, exceeding the target energy consumption of 5 kW h N m−3, and operated for over 240 hours with over 99.99% ionic barrier efficiency against seawater impurities (Fig. 10(g)) and there is no change in water migration in PTFE membranes (Fig. 10(h)), ensuring hydrogen purity levels of 99.9% to 99.99%.
For large-scale DSE, there is still a trade-off between economic factors and performance stability for different electrode materials during the transition from laboratory scale to industrial scale. Specifically, commercial Pt-based electrocatalysts can only operate in natural seawater for less than 1 hour due to toxic substances,198 while commercial RANEY® Ni suffers severe Cl− corrosion.199 Commercial IrO2200 and RuO2201 also face selectivity and durability challenges. Therefore, directly applying existing commercial catalysts to DSE for large-scale applications is impractical. Most ampere-level DSE electrodes currently employ self-supported electrodes developed on metal foam substrates. Metal foams act not only as conductors and carriers but also enable their own metal ions to replace metal salt precursors, directly participating in electrocatalyst assembly. This facilitates the growth of diverse nanostructures (e.g., nanowires, nanosheets, nanoarrays), effectively enhancing catalytic activity while reducing costs and stability uncertainties associated with ionomer binders.202 Taking nickel foam (NF) as an example, although its intrinsic DSE catalytic activity is modest (η = 573 at 10 mA cm−2),203 its catalytic performance and durability can be dramatically improved through catalyst loading and modification (see Table 1). Combined with low cost, NF emerges as a strong candidate for DSE electrodes. Additionally, metal meshes/felts offer the advantages of lower cost, higher mechanical strength, and superior chemical stability.204 Relative studies have explored stainless steel mesh (SSM) substrates as electrodes for various electrochemical reactions. For example, Lyu et al.205 investigated commercial 316 SS and 304 SS as OER electrodes in natural seawater electrolysis. They found that 304 SS exhibited inferior corrosion resistance compared to 316 SS in neutral and mildly alkaline seawater electrolytes due to direct metal dissolution and CER. Increasing Mo content in 304/316 SS can enhance their stability and performance in seawater electrolysis. Notably, SSM has demonstrated performance surpassing commercial RuO2 and IrO2.202 These cost-effective, durable substrates also enable the development of 3D electrodes with controlled microstructures, enhancing local mass transfer to accommodate industrial-level high current densities while maintaining catalytic activity. We strongly recommend utilizing these advantageous catalytic substrates in DSE electrode R&D to further improve performance, selectivity, and long-term durability.
Compared to catalyst design, DSE device engineering presents greater challenges. Integrated DSE systems require balancing catalyst cost with prolonged durability. While academia has extensively studied seawater electrolysis catalysts and proposed numerous strategies to address harsh operating conditions, two critical gaps remain: (1) most studies overlook practical challenges of real seawater in application scenarios; (2) practical implementation demands innovative catalyst engineering and optimization of catalyst layer characteristics (uniformity, thickness) alongside other components like porous transport layers (PTL material properties, hydrophilicity, thickness). Furthermore, catalyst synthesis methods profoundly impact scalable production, essentially requiring a paradigm shift from materials chemistry to chemical engineering. Despite catalysis being a well-established field, knowledge about catalyst mass production remains scarce, though it ultimately determines process feasibility. Industrially, catalyst mass-production formulas remain closely guarded secrets, still dominated by empirical approaches. Taking widely studied NiFe-LDH in DSE as an example, electrochemical deposition and in situ growth (e.g., hydrothermal methods) offer simple and scalable fabrication. However, these methods suffer from time/energy intensiveness, typically requiring hours (even the fastest electrodeposition takes 3–15 min206). Notably, Tian et al. developed an ultrafast method by immersing Prussian blue analogs (PBA) chemically deposited on NF into 1 M KOH solution, achieving NiFe-LDH nanoparticle immobilization on NF within 10–90 s without post-treatment. The NiFe-LDH-20s/NF electrode demonstrates exceptional activity with low overpotential (∼0.240 V at 10 mA cm−2), small Tafel slope (38 mV dec−1), and stable performance during 15-h multi-step chronopotentiometric testing (50–500 mA cm−2).207 With advancing catalytic technologies, academia urgently needs to deepen understanding of catalyst scale-up to comprehensively bridge synthesis–mechanism–process relationships. This requires interdisciplinary expertise spanning chemistry (synthesis, physics, analysis), materials science, and chemical/mechanical engineering (reactor design). We therefore advocate for cross-disciplinary collaborative development in the transition of DSE technologies from laboratory to industrial scale. A comparison table of key performance indicators for different types of electrolysis cells (AWE, PEM, AEMWE and BPMWE) is given for reference (Table 4).
Anode | Cathode | Electrolyte | T (°C) | Performance | Ref. |
---|---|---|---|---|---|
AWE | |||||
MHCM-z-BCC | NiMoS | Neutral phosphate-buffered seawater | 25 | 2.1 V@10 mA cm−2, current density dropped to 100% after 10 h | 208 |
NiNS | NiNS | Neutral phosphate-buffered seawater | 25 | 1.8 V@48.3 mA cm−2 | 209 |
Co–Fe2P | Co–Fe2P | 0.5 M NaCl + 1.0 M KOH | 25 | 1.69 V@100 mA cm−2, 22 h@100 mA cm−2 | 210 |
SSM | Ni–MoN | 1 M KOH + seawater | 25 | 1.783 V@500 mA cm−2, 100 h@500 mA cm−2 | 211 |
NCMS/NiO | NCMS/NiO | 5 M KOH + seawater | 80 | 1.98 V@1000 mA cm−2, 30 d@100 mA cm−2 | 212 |
S-(Ni,Fe)OOH | NiMoN | 1 M KOH + seawater | 25 | 1.837 V@500 mA cm−2, 100 h@500 mA cm−2 | 101 |
Mo–Ni3S2/NF | PtNi mesh | 30 wt% KOH (inner) + seawater (outer) | 25 | 3200 h@250 mA cm−2 | 34 |
NiFeP@Ag | Cu2S@Ni | 1 M KOH + seawater | 25 | 1.95 V@400 mA cm−2, 1200 h@400 mA cm−2 | 213 |
PEMWE | |||||
FeOx | FeOx | 0.6 M NaCl + 0.1 M KOH | 25 | 30 min | 214 |
Ir black | Pt/C | Anode: water vapor | 80 | 2.37 V@500 mA cm−2, 1.5 h@500 mA cm−2 | 172 |
Cathode: 0.5 M NaCl | |||||
(NiFe)C2O4/NF | Pt/C | 1 M KOH + seawater | 25 | 2.4 V@500 mA cm−2, 150 h@500 mA cm−2 | 94 |
CoOx–Cr2O3 | CoOx–Cr2O3 | Seawater | 60 | 1.87 V@1000 mA cm−2, 100 h@500 mA cm−2 | 11 |
AEMWE | |||||
Pt/C | Co3−xPdxO4 on MnO2-NF | Anode: 0.5 M NaCl + 1.0 M PBS | 25 | 65 h@100 mA cm−2 | 105 |
Cathode: humidified N2 | |||||
Pb2Ru2O7−x | Pt/C | 0.6 M NaCl + 0.1 M NaOH | 25 | 1.8 V@275 mA cm−2, 5 h@200 mA cm−2 | 140 |
Fe, P-NiSe2 NFs | Fe, P-NiSe2 NFs | Anode: N2 | 25 | 1.7 V@305 mA cm−2, 1.8 V@200 h | 215 |
Cathode: seawater | |||||
NiFe-LDH | Pt/C | Anode: 0.5 M NaCl | 25 | 12 h@1.7 V | 166 |
Cathode: 0.5 M KOH | |||||
RuMoNi | RuMoNi | 1 M KOH + seawater | 80 | 1.72 V@1000 mA cm−2, 240 h@500 mA cm−2 | 87 |
Ni–FeOOH | Pt/C | 1 M KOH + seawater | 25 | 1.7 V@469 mA cm−2, 15 h@500 mA cm−2 | 99 |
NiFe-LDH/Ni@NixSy | CoP/C/Ni@NIxPy | 1 M KOH + seawater | 60 | 2 V@1000 mA cm−2, 100 h | 152 |
BPMWE | |||||
Ir@Ti fiber | Pt@Ti fiber | Anode: 0.5 M KOH | 25 | 1000 h@100 mA cm−2 | 173 |
Cathode: seawater | |||||
Pt@Ti mesh | Pt@Ti mesh | Anode: 0.1 M NaOH | — | Cell voltage was changed from −3.0 to –3.5 V (100 h@10 mA cm−2) | 216 |
Cathode: sea salt solution | |||||
IrOx nanoparticles | Pt black | Seawater for both electrolytes | RT | Cell voltage increased by 0.90 V (>6 h operation; 250 mA cm−2) | 33 |
Porous Pt | Porous Pt | Anode: 0.5 M NaOH | 25 | Cell voltage was stable at −2.3 V (100 h; 20 mA cm−2) | 158 |
Cathode: seawater |
![]() | ||
Fig. 11 (a) Illustration of the antifouling mechanism for Cu–Ti composite and corrosion processes in Cu–Fe laser-cladded coatings. Reproduced with permission.225 Copyright 2021, Elsevier. (b) Schematic representation of corrosion processes in Cu–Fe laser-cladded coatings. Reproduced with permission.226 Copyright 2023, Elsevier. (c) Integrated electrolysis system configurations: onshore, offshore, and distributed. Reproduced with permission.233 Copyright 2024, Elsevier. (d) Illustration of the seawater chamber connected to the ocean. (e) Illustration of the overall system consisting of power supply, electrolysis module, and auxiliary module. (f) The diagram shows the wind turbine network, fluctuation in turbine power with wind speed, and pressure and stress distribution of the floating platform in a fluctuating environment. (d)–(f) Reproduced with permission.39 Copyright 2024, Springer Nature. |
(a) Onshore:35 in this scenario, electricity generated offshore is transmitted through undersea cables to offshore substations, where the voltage is increased. The electricity is then further conveyed via high-voltage cables to an onshore electrolysis cell for large-scale hydrogen production. While this configuration may facilitate the construction and upkeep of electrolysis cells, it could lead to greater energy losses due to the lengthy transmission distances and the challenges associated with managing high-voltage power offshore. In addition to maintaining plant components and systems, ensuring the stable operation of electrolysis cells is a significant challenge. Renewable energy sources, like wind and solar, experience variability due to seasonal changes in wind patterns and solar intensity. This fluctuation in power input can lead to frequent starts and stops of the entire system, potentially causing harm to the electrodes of the electrolysis cells and diminishing their longevity.234 Therefore, there is an analysis indicating that a single 7 MW cell stack undergoes 87 cycles per day. However, by utilizing three 2.33 MW cell stacks, this cycling can be notably reduced to just 35 cycles per day. This reduction could potentially enhance the longevity of the electrolysis cells, as frequent cycling contributes to the degradation of cell performance. On the downside, using multiple stacks may lead to higher capital costs, although the cost per kW for electrolysis cell stacks tends to decrease as their power capacity increases.235
(b) Offshore:236 in this arrangement, electricity produced by offshore wind turbines is sent to a proximate offshore electrolysis cell. This system converts electricity into pressurized hydrogen, which is subsequently transported to the mainland through pipelines. This approach minimizes energy losses and may reduce the costs associated with long-distance electrical transmission. However, oxygen and hydrogen bubble dynamics, as well as the performance of liquid–gas separators, can be affected by the orientation of the equipment due to the motion of the offshore platform. Be aware that bubbles can cause physical/mechanical degradation in the electrolysis cells, Liu et al.39 have proposed an innovative floating electrolysis system that harnesses wind and wave energy for direct electrolysis using dynamic seawater. The heart of this system is an electrolysis module developed by the Xie's group, which leverages the principle of aqueous phase transfer and integrates various modules, including energy storage, current conversion, hydrogen detection, and transport, for enhanced synergy. The electrolysis cell is designed to float within a seawater tank, addressing seawater fluctuations to improve system stability. To ensure the platform remains stable against displacement from cables or adverse weather conditions, it features counterweight modules and four anchoring points. Additionally, seawater tanks connected to the ocean via controlled valves help mitigate wave impact while facilitating direct interaction with the electrolysis cells. Weighing about 48.5 tons, the platform is engineered to withstand wave forces. A sealed connection between the seawater tank and the electrolysis cells prevents sloshing and overflowing into the floating structure. The collaborative functionality of the offshore wind energy storage module, hydrogen detection module, and other auxiliary components ensures reliable and efficient operation of the system.
(c) Distributed:237 this configuration entails the installation of small-scale electrolysis cells directly on each wind turbine, enabling immediate conversion of electricity into hydrogen at the generation site. The produced hydrogen is then compressed and conveyed to the mainland. Individual (hydrogen) produced by each unit is transported through risers and can be collected in subsea manifolds for output. By localizing production, this setup decreases transmission energy losses and enhances the overall efficiency of hydrogen generation. The primary benefit of this configuration is that if one electrolysis cell malfunctions, the remaining wind turbines can still effectively generate hydrogen.238 Relative strategies are employed to enhance the stability of systems like floating offshore wind (FOW) platforms (Fig. 11(d) and 11(e)). These strategies must account for adequate space for the electrolysis cell system and ensure robust mechanical stability to maintain safety and performance once the electrolysis facility is integrated into the foundation (Fig. 11(f)).239
For the three configurations mentioned above, in the case of centralized onshore electrolysis, energy is transmitted to shore via a submarine high-voltage direct current (HVDC) cable. This approach is practical for nearby offshore locations, but it faces challenges such as energy losses and limited flexibility for future expansion. Conversely, both decentralized and centralized offshore systems utilize submarine hydrogen pipelines for energy transfer, which offer advantages over high-voltage cables. Hydrogen pipelines provide greater capacity for expansion and are generally more economical, as they are not constrained by the transmission limits of a single cable. Decentralized offshore systems are particularly advantageous due to their modular nature, allowing for continuous hydrogen production even if one electrolysis cell or turbine fails. However, they face challenges related to operational complexity and maintenance, necessitating further validation in offshore conditions.238 On the other hand, centralized offshore systems can compete effectively, particularly in terms of simplified maintenance for individual turbines. While centralized systems are generally less complex, a significant drawback is that hydrogen production ceases if a failure occurs in the system. Furthermore, economic analyses240 reveal that distributed hydrogen production is the most cost-effective option, with a levelized cost of hydrogen (LCoH2) at $13.34 per kg, followed closely by centralized production at $13.66 per kg, and land-based production at $14.10 per kg. Rogeau et al.241 also indicate that offshore electrolysis is significantly cheaper in terms of LCoH2, particularly in deep-water scenarios (over 100 m), where the decentralized option is expected to gain favor over centralized systems in the coming years. However, these findings require further investigation and discussion to be fully validated while considering long-term stability and economy of system.
Finally, we analyzed and projected the future of three prominent seawater hydrogen production demonstration projects in China. The 100 kW-level seawater electrolysis system (20 N m3 h−1) developed by the Dalian Institute of Chemical Physics (DICP) achieves a cell voltage of 1.59 V and DC power consumption of 3.80 kW h N−1 m−3 h−1 H2−1 at a current density of 3000 A m−2 through high-performance electrodes and wide-power-adaptation processes. The hydrogen purity exceeds 99.999%, with successful adaptation to offshore wind power fluctuations.242 Xie Heping's team pioneered the phase transition migration-driven in situ direct seawater electrolysis technology. Utilizing PTFE membrane gas–liquid isolation and vapor pressure differential mechanisms, this desalination-free hydrogen production method was validated by a 10 N m3 h−1 prototype operating continuously for 240 hours in Xinghua Bay, Fujian. A 1.2 N m3 h−1 floating platform for seawater hydrogen production via renewable energy achieved the first direct integration with offshore wind power under challenging conditions (wind scale 3–8, wave height 0.3–0.9 m) in Xinghua Bay, maintaining stable operation for 10 days.243 The “Integrated Offshore Green Hydrogen-Methanol-Ammonia Production System” jointly developed by National Energy Group Hydrogen Technology Co., Ltd, Yantai CIMC Raffles Offshore Engineering Co., Ltd, and others has received China Classification Society (CCS) certification. Combining off-grid PEM/alkaline electrolysis with green ammonia/methanol storage technologies, it represents China's first marine hydrogen demonstration project featuring offshore mobile platform-based green hydrogen production coupled with hydrogen-based chemical processes.244
In pursuit of longer-term stability, many catalysts have been tested in alkaline natural seawater and demonstrated significant performance, but there has been no further discussion on how these catalysts function in natural seawater. Compared to alkaline seawater solutions, natural seawater has lower conductivity, and its pH is closer to neutral or weakly alkaline. As a result, under the same conditions, NSE of DSE encounters a higher overpotential than ASE of DSE, making it difficult to achieve sufficient industrial-grade current density within the OER potential range at this pH. Additionally, due to the high overpotentials and the complex and harsh electrolyte environment of natural seawater, NSE electrodes are subjected to more intense chemical and electrochemical corrosion. It is important to note that the design of long-durability NSE cathodes also requires careful modulation around the key pH-related ion (OH−) and the development of NSE anode electrode catalytic materials requires greater emphasis on enhancing the water dissociation capability of electrocatalysts under neutral or weak alkaline conditions, improving OER selectivity and increasing the corrosion resistance of electrode materials. Generally, NSE exhibits poorer performance and durability compared to ASE and SSE. However, some high-performing catalysts provide effective guidance, such as the Cr2O3–CoOx, introducing a Lewis acid layer (Cr2O3) was introduced onto the catalyst surface (CoOx) to split H2O and generate localized surface alkalinity, thereby suppressing the release of chlorine gas captured at the anode. Additionally, the strong binding between OH− and the Lewis acid layer reduced the capture of OH− by Mg2+ and Ca2+ cations present in the seawater electrolyte, enabling anti-precipitation at the cathode.
Moreover, the design of electrolysis cells, whether membrane-less systems or membrane-based systems—plays a crucial role in enhancing operational stability. Membrane-less electrolyzers, leveraging hydrodynamic control (e.g., Segré–Silberberg effect, asymmetric electrodes) or capillary-fed designs, minimize gas crossover but face scalability limitations and require highly selective catalysts. Membrane-based systems, such as mature AMEWE, the stability of AEMs significantly impacts the long-term operation of DSE. Membrane degradation in DSE can be attributed to basic and seawater effects. Strategies to enhance membrane performance include developing robust high-molecular-weight polycarbazole membranes and blending ionomers that improve adhesion and ionic conductivity. Notable advances include hybrid MEA designs that strengthen bonding between catalyst layers and membranes. In addition, BPMWE have shown particular promise due to their ability to restrict chloride ion transport (achieving <0.005% Cl− crossover), thereby minimizing corrosive side reactions. However, further optimization is still needed to adapt to asymmetric seawater feed to cross the laboratory to commercial level. For integrated systems, such as phase-transition PTFE membrane designs, demonstrate promise by combining in situ purification with electrolysis, achieving >3200 hours of stable operation and >99.99% impurity rejection. Morever, it can be believed that the synergizing hydrogen production with byproduct valorization (e.g., Mg(OH)2 for cement) could further enhance economic viability in the future.
Looking ahead, the future of DSE also aligns with the broader transition towards offshore renewable energy generation. The successful integration of advanced DSE technologies within maritime environments is pivotal. This will necessitate not only continued innovations in electrolysis cell designs and materials but also the establishment of robust support systems for hydrogen purification and transport. System-level integration—including consideration of auxiliary components like separators, compressors, and energy management systems—will be essential for maximizing efficiencies and ensuring reliable hydrogen production. With ongoing developments in offshore wind projects incentivizing hydrogen production directly at sea, three configurations (onshore, offshore, distributed) for hydrogen production using offshore electrolysis are compared. Although they have their pros and cons, there's significant potential for decentralized production capabilities to emerge, minimizing logistical barriers and improving economic viability. Several stability challenges arise in electrolysis systems due to the complex composition of seawater, intermittent fluctuations in renewable energy, and dynamic marine environmental disturbances, such as waves, temperature variations, and mechanical stresses. These factors can adversely affect the long-term stability of these systems, highlighting the need for innovations in dynamic control and buffering strategies: (1) to stabilize operating temperatures in offshore environments, it is essential to design integrated phase-change materials (PCMs) or incorporate active cooling and heating loops. (2) Additionally, real-time platform stabilization technologies, such as dynamic counterweight systems or thrusters, can mitigate wave-induced oscillations, reducing mechanical stresses on electrolysis modules and gas–liquid separators. (3) Implementing hybrid energy storage solutions can buffer intermittent power inputs and minimize cycling degradation. (4) Real-time monitoring through embedded sensor networks will enable tracking of temperature variations and environmental changes, allowing for AI-driven predictive maintenance. (5) At the laboratory scale, we can pre-simulate waves, temperature variations, and conduct accelerated aging studies alongside intermittent power conditions. Characterizing electrode and membrane electrode assembly and linking them to electrochemical signals will contribute to building a database that enables AI-driven alarm signals to identify maintenance needs under actual operating conditions.
Overall, while the path to efficient and stable seawater electrolysis remains fraught with challenges, ongoing research and innovation position this technology as a central player in the quest for clean hydrogen production. By addressing the multifaceted aspects of stability and integration, DSE could potentially transform the renewable energy landscape and contribute substantially to global decarbonization efforts. Continued collaborative efforts among researchers, engineers, and industry stakeholders will be crucial to realizing the full potential of DSE in the near future.
Footnote |
† These authors contributed equally to this work. |
This journal is © The Royal Society of Chemistry 2025 |