A collaboration between researchers at Cornell University and the City University of Hong Kong has produced a new computational tool that could help overcome a surprisingly persistent obstacle to efficient green hydrogen production: bubbles.

Green hydrogen is made by splitting water using electricity from renewable sources. It is seen as a promising alternative to fossil fuels, especially for hard-to-decarbonise industries such as steelmaking and chemical manufacturing. But the electrolysis process itself creates a problem: as electricity passes through an electrolyzer, hydrogen and oxygen bubbles form on the electrodes where the chemical reactions occur.

According to the Cornell Chronicle, which reported the findings, those bubbles can cover active reaction sites, prevent liquid from reaching the electrode surface and interfere with the movement of electrically charged particles. The system then has to consume additional energy to keep working. The problem becomes more serious as electrolyzers are scaled up and operated at higher current densities to increase hydrogen output.

Scientists have long known that bubbles reduce efficiency, and researchers have tried modifying electrode shapes, changing liquid flows and altering surface properties to encourage bubbles to detach. What has been missing, according to Lenan Zhang, assistant professor in Cornell's Sibley School of Mechanical and Aerospace Engineering and one of the study's leaders, is a reliable way to quantify precisely how much energy bubbles waste and to determine when intervention becomes necessary.

Zhang told the Cornell Chronicle that the difficulty arises because the physical processes involved span enormous differences in scale, from electrochemical reactions occurring across distances smaller than a nanometre to the centimetre-scale dimensions of an electrolyzer. "Experimentally it's really challenging to isolate the impact of bubbles on each individual scale," Zhang said.

To address that challenge, the team created a 3D mathematical model that can simulate linked processes spanning eight orders of magnitude in scale. Using a novel approach to depict the tiniest electrochemical reactions, they cut the needed computing power by about 1,000 times while maintaining numerical precision, according to the Cornell Chronicle.

The result, Zhang said, is akin to a digital twin of an electrolyzer—a simulated system capable of replicating actual electrode shapes, materials, and operational parameters, and exposing microscopic phenomena that standard experiments cannot readily detect. By applying the simulation to study individual hydrogen bubbles forming on an electrode within an alkaline electrolyzer, the team found that blocking the active electrode surface area caused the greatest bubble-related energy loss. Furthermore, they pinpointed a threshold bubble radius; when a bubble surpasses this size, the extra voltage needed to maintain electrolysis rises exponentially. Under certain modeled scenarios, the total additional voltage can exceed its bubble-free value by more than 50%, according to the study.

The findings were published Nov. 9 in the journal EES Catalysis.

The critical radius depends on electrode geometry and how readily liquid wets the electrode surface, rather than on operating current density. That finding suggests a practical design strategy, the researchers said: design electrode surfaces so that gas bubbles come off before they hit the critical size. The team created a new design metric that compares the bubble detachment radius to the critical radius. Their computer models show that if this ratio stays below one, the extra voltage needed because of a single bubble can be kept within 3%., associate professor of mechanical engineering at City University of Hong Kong. Muhan Zhang of the City University of Hong Kong and visiting undergraduate student Kaiyu Li of Cornell, who contributed equally to the research, are co-first authors.

The current study focuses on a single bubble, but Zhang and Li are extending the model to investigate interactions among multiple bubbles, which can grow, merge and influence one another. The researchers also envision combining their modelling framework with computational optimization and additive manufacturing to create electrodes designed specifically to minimise energy losses.

Ultimately, the researchers hope to develop their computational framework into broadly accessible software that engineers can use to evaluate materials, electrode structures and operating conditions.

For university students, particularly those studying engineering, materials science, chemistry or renewable energy, the research highlights a concrete, scaleable design challenge in the green hydrogen field. Understanding how microscopic bubble dynamics affect macroscopic energy efficiency could inform future work on electrolyzer design, a key area as the world looks to decarbonise heavy industry.