New Liquid-Metal Membrane May Make Hydrogen Fuel Cell Vehicles Viable

Liquid metal membrane can simplify complex hydrogen production processes.

Electric cars are enjoying a surge in popularity due to their growing capabilities, but hydrogen fuel cells may not be far behind. Hydrogen-powered cars have been a green-energy concept for decades. The advantages are many, including clean, abundant fuel, lighter weight, faster refueling and longer range.

The main challenge to their commercial success has been the high cost and complexity of producing and storing the hydrogen fuel, which can present safety hazards. However, a new technology may offer a solution to these issues, making hydrogen cars finally viable for consumers.

How is Hydrogen Gas Produced?

The most common hydrogen production method is called steam reforming. Extremely hot steam reacts natural gas in the presence of a catalyst, resulting in a cocktail of CO, CO2 and H2. Separating these gases requires a lengthy, complicated chemical process. Engineers are developing membranes to do this job using palladium. Palladium has the unique property of high hydrogen solubility and permeance, but it’s also very expensive and fragile.

In short, cars that run on hydrogen would be great, but the process of producing hydrogen is currently too complex and expensive to be practical on a large scale. Palladium membrane techniques tackle the complexity issue, but don’t help with cost. The search for a more cost-effective alternative to the expensive palladium has had chemical engineers stumped for years—until now.


From left: Pei-Shan Yen ’16 (PhD), Ravindra Datta, professor of chemical engineering, and Nicholas Deveau ’17 (PhD) at WPI developed novel sandwiched liquid-metal membranes that could help lower the cost of hydrogen for fuel-cell-powered vehicles. (Image courtesy of Worcester Polytechnic Institute.)

Liquid-Metal Membranes for Hydrogen Separation       

Metals that are liquid at 500°C, such as gallium, can be used to create a film through which hydrogen can separate. Liquid metal membranes combine the high temperature of steam reforming with the hydrogen permeance of palladium, offering several advantages. 


Gallium. (Image source Wikimedia.)

The liquid-metal membranes are much more efficient at separating the hydrogen, and more durable than their palladium counterpart. To increase the flux of a palladium membrane, you must make it thinner. However, palladium is prone to defects in thin layers, and even a small crack renders the membrane worthless for separating gases. Naturally, a liquid film cannot form cracks or defects.

Challenges in the development of a functioning liquid-metal membrane included dealing with the high reactivity of liquid metals. Metal support structures formed intermetallic compounds with the gallium, reducing the permeability. The gallium reacted with ceramic structures, too.

To solve this problem, the team developed a list of materials that were nonreactive with liquid gallium but were wettable by it, allowing it to form a film. The resulting prototype sandwiched a 0.2mm-thick layer of liquid gallium between a layer of silicon carbide and a layer of graphite.

Testing a New Method of Hydrogen Separation 

The membrane was exposed to a hydrogen atmosphere at temperatures ranging from 480°C to 550°C, for two weeks. The test showed that this gallium membrane was 35 times more permeable to hydrogen than a comparable palladium membrane, and that it allowed greater diffusion of hydrogen than a palladium membrane. If issues eventually surface with using gallium, there are a host of other metals which are liquid at 500°C that could be tested.

Next steps for the research include scaling up the membranes, and testing their reactivity with substances present in steam reformed gases, such as carbon monoxide and sulfur, which poison palladium membranes.

This test is a big step toward the feasibility of hydrogen fuel cell research, in automotive and other applications. The research was pioneered by Ravindra Datta, professor of chemical engineering at Worcester Polytechnic Institute (WPI) and published in the Journal of the American Institute of Chemical Engineers.