Technology explainer
How Do Hydrogen Fuel Cells Actually Generate Electricity?
A PEM fuel cell splits hydrogen at the anode, sends electrons through a circuit and protons through a membrane, then forms water with oxygen at the cathode. Catalyst activity, platinum cost, water management, durability, and hydrogen production determine real performance.
Short answer: a hydrogen fuel cell separates one chemical reaction into two electrochemical half-reactions. Hydrogen releases electrons at the anode; the electrons power an external circuit while ions cross an electrolyte; oxygen then combines with them at the cathode to form water. The cell produces electricity continuously while fuel and air are supplied.
The parts of a PEM fuel cell
The common proton-exchange-membrane fuel cell contains two porous electrodes separated by a polymer electrolyte membrane. Catalyst layers sit beside the membrane, while flow fields distribute hydrogen and air and remove water and heat.
| Part | Function |
|---|---|
| Anode | Splits hydrogen into protons and electrons |
| Electrolyte membrane | Conducts protons while blocking electrons and separating gases |
| Cathode | Combines oxygen, protons, and electrons to form water |
| Gas-diffusion layers | Spread gases, conduct electricity, and help manage water |
| Bipolar plates | Distribute gases, collect current, remove heat, and connect cells |
How electricity is generated
- Hydrogen reaches the anode catalyst.
- The hydrogen oxidation reaction occurs: H2 → 2H+ + 2e−.
- Protons cross the membrane. Electrons cannot, so they travel through the external circuit and perform useful work.
- Oxygen enters the cathode from air.
- The oxygen reduction reaction occurs: ½O2 + 2H+ + 2e− → H2O.
The net reaction is H2 + ½O2 → H2O, with electrical energy and heat released. The cell is electrochemical rather than a heat engine, although practical losses still limit efficiency.
Why one cell is not enough
A single cell operates at less than one volt under load. Engineers connect many cells in series as a stack to raise voltage. The system also needs pumps, blowers, humidification, cooling, sensors, valves, controls, hydrogen recirculation, and power electronics.
Stack output is therefore not the same as net system output. Auxiliary equipment consumes part of the generated power.
Why the cathode is the bottleneck
Hydrogen oxidation at the anode is comparatively fast on platinum. Oxygen reduction at the cathode involves several electron and proton transfers and the breaking and forming of oxygen bonds. It is much slower and causes a large voltage loss.
This is why PEM fuel cells use most of their platinum-group-metal loading at the cathode. Reducing that loading without sacrificing activity and life is central to cost reduction.
What the catalyst must do
Platinum surfaces adsorb oxygen-containing intermediates strongly enough to enable reaction but not so strongly that sites remain blocked. Alloying platinum with cobalt, nickel, or other metals can tune its electronic structure and increase activity.
Nanoparticles expose more surface per gram than bulk metal. They are dispersed on conductive carbon so gases, protons, and electrons can reach active sites. The catalyst layer must also balance pores and water: flooding blocks oxygen transport, while excessive drying impedes proton conduction.
The activity-durability tradeoff
| Degradation process | What happens |
|---|---|
| Particle coarsening | Small particles migrate or dissolve and redeposit into larger ones, reducing active area |
| Metal dissolution | Alloying elements leave the particle, changing activity and contaminating the membrane |
| Carbon corrosion | The support oxidizes during harsh voltage conditions, losing electrical contact |
| Platinum dissolution | Voltage cycling moves platinum and reduces accessible catalyst |
| Membrane attack | Chemical radicals, heat, dryness, or stress thin or puncture the polymer |
| Contamination | Carbon monoxide, sulfur, ammonia, and air impurities poison or disrupt the cell |
A highly active nanostructure may be unstable because small particles have high surface energy. Stabilizing the structure can protect durability but hide active surface. Good design must achieve both.
A carbon-cage approach
A 2026 study reported a nanostructured carbon support that confined platinum-cobalt particles and limited clumping even after high-temperature processing. The catalyst retained 85 percent of its initial performance after 150,000 accelerated stress cycles. Read A Nanocarbon Cage Could Make Hydrogen Fuel Cells Cheap Enough for Data Centers.
This is promising catalyst evidence, not proof of a complete commercial system's lifetime or cost. Full stacks must validate mass production, impurity tolerance, water management, start-stop behavior, freeze operation, and sustained real-world load cycles.
Fuel cell versus battery
| Feature | Fuel cell | Battery |
|---|---|---|
| Energy location | Stored externally as hydrogen | Stored inside electrode materials |
| Replenishment | Refuel the tank | Recharge electrochemically |
| Operating duration | Depends on fuel supply | Depends on installed battery capacity |
| System needs | Tanks, valves, fuel delivery, and often compression | Battery management and charging equipment |
A fuel cell is not “a battery that never runs out.” It runs only while reactants are supplied, and its external fuel infrastructure is part of the technology.
Where the hydrogen comes from
The exhaust of a hydrogen fuel cell is mainly water and heat, with no carbon dioxide produced inside the cell. But lifecycle emissions depend on hydrogen production. Hydrogen made from fossil gas without effective carbon capture can carry substantial emissions. Electrolytic hydrogen can be low-carbon when powered by additional low-carbon electricity.
Compression, liquefaction, storage, transport, leakage, and dispensing also require energy. Climate comparisons must use well-to-wheel or full lifecycle boundaries, not tailpipe emissions alone.
Efficiency and heat
PEM systems commonly convert a substantial fraction of hydrogen's chemical energy into electricity, with the remainder mostly heat. Exact figures depend on load, system boundary, and whether lower or higher heating value is used. Combined heat and power can raise total useful-energy efficiency if nearby demand can use the heat.
Producing hydrogen with electricity and later converting it back loses more energy than using electricity directly or storing it in a battery. Hydrogen can still be useful where long duration, low mass, fast refueling, industrial feedstock, or chemical storage outweighs the efficiency penalty.
Where fuel cells fit
- Backup and distributed power where quiet operation and long runtime matter
- Vehicles with high utilization or difficult battery weight and charging constraints
- Combined heat and power for buildings and industry
- Remote systems where fuel logistics are workable
- Data centres seeking dispatchable low-local-pollution power
The best choice depends on hydrogen price and carbon intensity, utilization, reliability requirements, grid conditions, space, noise, heat use, and competing batteries or generators.
The mental model
The membrane forces hydrogen's electrons to take the long route through a circuit while protons take the short route through the electrolyte. Their reunion with oxygen makes water. The chemistry is simple; the engineering challenge is making every reaction site accessible, fast, hydrated, clean, affordable, and durable for thousands of operating hours.
First appeared in
A Nanocarbon Cage Could Make Hydrogen Fuel Cells Cheap Enough for Data Centers
