Gordon Cooper didn’t carry extra batteries on Gemini 5. He carried hydrogen. The mission ran for nearly eight days on a proton exchange membrane fuel cell — a device that had never flown before, built by a team that wasn’t entirely sure it would survive re-entry vibration. It did. And the water it produced as a byproduct? Clean enough to drink on orbit.
That was 1965. The core principle hasn’t moved an inch since.
Strip Away the Jargon
Hydrogen goes in one side. Air goes in the other. Electricity comes out. Water comes out too. Nothing burns. Nothing spins. The cell just sits there doing chemistry, quietly, continuously, for as long as you keep feeding it fuel. The reason this works comes down to one component. The membrane.
What the Membrane Does
It’s called Nafion. Thin, slightly waxy, maybe a hundred microns thick. It lets protons through and stops electrons cold.
Hydrogen arrives at the anode. A platinum catalyst breaks each H₂ molecule apart — two protons, two electrons. The protons migrate through the membrane. The electrons try to follow and can’t. Blocked. So they exit through the only available path: the external circuit. Your wire. Your load. Whatever you’ve connected to the terminals. That forced detour is the electricity.
On the cathode side, oxygen from incoming air is waiting. The protons arrive through the membrane. The electrons complete their journey through the circuit. Both meet the oxygen. Water forms, exits as vapour.
H₂ + ½O₂ → H₂O. No carbon anywhere in that equation. No combustion. Nothing worth worrying about coming out the exhaust.
The Problem Nobody Mentions First
Nafion is temperamental. It conducts protons through a hydration-dependent mechanism — protons hop along water molecules embedded in the polymer’s sulphonic acid sites. Take the water away and conductivity drops fast. Run it too wet and liquid water blocks reactant access to the electrode surface.
The operating window that avoids both problems sits between 60°C and 80°C. Not because that temperature produces the best reaction kinetics — higher would — but because it’s where the membrane stays cooperative. The U.S. Department of Energy identifies water and thermal management as primary design constraints. Entire subsystems exist in commercial stacks just to keep the membrane in that window.
One Cell Isn’t Enough
A single pem fuel cell produces 0.6–0.7 volts under real load. The theoretical ceiling is 1.23 V — activation losses and ohmic resistance eat the difference before a single electron reaches your terminals.
So engineers stack them. Cells in series, separated by bipolar plates that distribute gas and carry current between layers. A hundred cells gives roughly 65 volts. Scale the active electrode area and current rises with it. Toyota’s Mirai runs two stacks totalling 128 kW. Warehouse forklifts — Amazon operates thousands — use smaller stacks sized for an 8-hour shift with a 3-minute hydrogen refill.
Same membrane. Same electrode structure. Same physics. Just more of it, or less, depending on what you need.
The Efficiency Conversation — Done Honestly
40–60% electrical efficiency under real operating conditions. Recover waste heat for building services or industrial processes and total system efficiency crosses 80%. A petrol engine converts 25–30% of fuel energy into useful work — the rest leaves through the exhaust.
That gap is real. But efficiency at the cell says nothing about upstream emissions. Grey hydrogen — produced from natural gas, which is how most hydrogen is made today — carries a carbon footprint of around 10–12 kg CO₂ per kg H₂. Green hydrogen, produced by electrolysers running on wind or solar, closes that gap entirely. The cell produces no carbon. The question is always what went into making the hydrogen.
Where It Stands in India Right Now
The National Green Hydrogen Mission targets 5 million metric tonnes of annual green hydrogen production by 2030, alongside 5 GW of annual electrolyser manufacturing capacity. Those numbers are government policy now, not aspiration.
PEM electrolysers and pem fuel cells share their core architecture — same membrane, same electrode assembly, same stack design. Expertise transfers between them. Engineering institutions are reflecting this in their lab infrastructure, moving from simulations to actual hardware: polarisation curve characterisation, stack warm-up behaviour, gas crossover measurement. Five years ago a working fuel cell test setup in an Indian university was unusual. That’s changed.
What It Comes Down To
The PEM fuel cell works because one polymer membrane passes protons and refuses electrons. That single property forces electrons through an external circuit — and that circuit is your power source. Stack enough cells, manage the membrane’s hydration, keep the platinum catalyst clean, and you have a scalable, silent, zero-emission power source.
It flew in 1965. It drives on public roads today. The physics was always sound. What’s catching up now is everything around it — hydrogen supply chains, cost curves, infrastructure. The cell itself was never the problem.