For decades, water electrolysis research has operated under the assumption that small, fast-departing bubbles are ideal for efficiency. However, a new study published in the journal eScience challenges this long-held belief, revealing that under high-current conditions, bubbles that merge and depart later can actually improve the hydrogen evolution reaction (HER). The findings suggest that larger departing bubbles are not beneficial simply because of their size, but because their coalescence effectively clears tiny bubbles stuck to the electrode and stirs the surrounding liquid, enhancing overall performance.
The research, conducted by a team from East China University of Science and Technology and Southern University of Science and Technology, explored how electrolyte composition influences bubble coalescence and its impact on HER efficiency in both acidic and alkaline water electrolysis. The team discovered that in systems where coalescence was promoted, HER efficiency increased by up to 30% compared to systems where coalescence was inhibited. This significant improvement offers a new perspective on mitigating energy losses in hydrogen production, a critical component in the transition to green hydrogen for decarbonizing industries such as chemical manufacturing, transportation, and steelmaking.
Conventional strategies for improving electrolysis efficiency have focused on making bubbles detach earlier and at smaller sizes through surface design, wettability control, or external fields. However, at high current densities, bubble-bubble interactions become dominant, and the new research indicates that promoting coalescence can be more effective. The team used a three-electrode electrolytic cell with a platinum disk electrode, employing electrochemical measurements, high-speed imaging, and numerical simulations to observe bubble behavior. In sulfuric acid, bubbles readily coalesced, but when perchloric acid or sodium sulfate was added, coalescence was suppressed, leading to smaller departure sizes. Surprisingly, the smaller bubbles did not improve performance; at -40 mA, adding perchloric acid reduced bubble size but caused a 20% drop in HER efficiency, and at -60 mA, the performance gap reached 30%.
The mechanistic analysis revealed that a just-detached bubble can linger above the electrode and continuously merge with surface-anchored microbubbles. This late departure pulls microbubbles away at sizes below 10 μm, freeing active sites before they become blocked. Additionally, coalescence generates local flows exceeding 1 m/s, which disrupts the stagnant interfacial layer and improves heat and mass transfer. In alkaline media, where coalescence is naturally suppressed, the addition of hydrophobic polystyrene microparticles promoted coalescence and improved efficiency by 2-6%.
The authors emphasize that this work shifts the key question in bubble management from how to make bubbles smaller to how bubbles interact after formation. Bubble coalescence acts as a self-driven cleaning and mixing process at the electrode surface, removing microbubbles early, reopening reaction sites, and bringing fresh electrolyte into a region where transport is usually slow. This explains why larger departing bubbles can signal better performance under high-current conditions.
These findings suggest a new design principle for gas-evolving electrochemical systems. In acidic systems, where bubbles already merge easily, electrodes or flow fields could be designed to increase useful bubble collisions. In alkaline water electrolysis, seawater electrolysis, and chlor-alkali processes, where coalescence is often inhibited, electrolyte additives or particle-assisted strategies may help restore beneficial merging. The study also points to broader applications in industrial electrolysis, where surface bubble removal and interfacial transport remain major limits. By treating coalescence as a controllable tool, future devices may reduce energy loss without relying solely on catalyst or electrode-surface improvements.
For more details, the full article is available at https://doi.org/10.1016/j.esci.2025.100472.
