Beijing Institute of Technology quadruples fuel cell output per square centimetre

An engineer fitting a stack into a truck frame or an airframe bay works in litres and kilograms. The headline number in the Science paper is per square centimetre of cell area ^[14][1]. Divide the reported 0.75 watts by the claimed fourfold gain and the baseline being beaten is about 0.19 watts per square centimetre ^[1].

Then comes the discount. A stack has to make its rated power at end of life. That puts the working figure at 63 percent of 0.75, or roughly 0.47 watts per square centimetre ^[2]. After 30,000 stress cycles that is still about 2.5 times what a fresh conventional cell puts out ^[3].

Li Jie, an assistant professor at Beijing Institute of Technology and a co-author, said the technology enables the "power stack of a fuel cell vehicle can be made smaller and lighter, and the vehicle can get stronger continuous output and a longer driving range" ^[7]. The researchers also pointed to aerospace systems using liquid oxygen, where proton transport moves overall fuel-cell performance ^[15].

Platinum is the cost line here. The cell managed 6.9 kilowatts per gram of platinum ^[8]. At that rate a 100-kilowatt stack carries about 14.5 grams of the metal ^[4].

Supply of the composite is the second constraint. "In the lab right now, they can make about 100 grams in three days. That's enough for 10 100-kilowatt-class stacks," Li Jie said ^[9]. That is 10 grams per stack ^[5], and roughly 1 megawatt of stack material every three days, about 122 megawatts a year at bench pace ^[6]. China had about 32,000 fuel-cell vehicles in operation at the end of 2025 ^[11]. Equipping 100,000 of them with 100-kilowatt stacks needs 10 gigawatts of stack; at the bench rate that is 30,000 days of production ^[7].

The bottleneck here is narrow. Nafion and similar materials pack densely around the catalyst at the nanoscale and slow protons on their way to reaction sites ^[3]. The team's Bronsted acid-Lewis base interface adds transfer sites, and the measured effect was a tenfold rise in proton diffusion, 6.5 times the proton conductivity, and activation energy cut by more than half ^[4][5].

Nobody specifying trucks this year is buying on this result, and Li Jie drew the use-case boundary herself. "Fuel cell vehicles don't need to copy battery EVs. Battery EVs fit short trips. Fuel cells fit heavy trucks, long hauls, and cases that need quick refueling and long range," she said ^[13]. The test that would turn this into a packaging decision is a stack-level one: watts per litre and kilograms of installed mass, measured at end of life, at the duty cycle the vehicle actually runs. "Next, the team will build and test high-performance fuel cell stacks for different scenarios," Li Jie said ^[12].