A PEM fuel cell converts hydrogen to electricity through proton transport across a membrane. The cathode catalyst layer — where oxygen reduction occurs — has internal resistance from proton transport losses. This resistance dissipates energy and reduces efficiency. The standard approach to minimizing it: better catalysts, thinner membranes, optimized microstructure.
An unexpected approach: oscillate. When current density and temperature are perturbed harmonically and in phase, the impedance of the cathode catalyst layer decreases. The proton transport losses that constitute most of the layer's resistance are reduced because the oscillation changes the average operating point — the nonlinear dependence of proton conductivity on temperature and current means that the mean of the oscillating state is not the same as the steady state at the mean conditions.
This is Jensen's inequality applied to electrochemistry. For a convex function, the function of the average is less than the average of the function. If proton transport losses depend convexly on temperature, then oscillating the temperature around a mean value produces lower average losses than operating at the constant mean. The oscillation exploits the nonlinearity of the transport physics.
At carefully calibrated amplitudes, the model predicts the proton transport losses can be eliminated entirely — the oscillation amplitude that maximizes the Jensen's inequality benefit happens to match the loss magnitude. Whether this is achievable in practice (thermal management of rapid oscillations in a real fuel cell) is another question, but the theoretical mechanism is clean.
The through-claim: steady-state operation is not necessarily optimal for nonlinear systems. The best operating point might not be a point at all but a cycle — a periodic orbit that exploits the system's nonlinearities to achieve lower average losses than any fixed condition.
(arXiv:2603.17709)