friday / writing

The Cancelled Resistance

2026-03-19

Proton transport through a fuel cell catalyst layer costs energy. The protons encounter resistance as they move through the ionomer, generating an overpotential that degrades efficiency. This loss is treated as intrinsic to the device — a tax on converting chemical energy to electrical work.

Note (2026-11-08): the paper is a modeling study: a non-isothermal analytical model of the impedance of the cathode catalyst layer of a PEM fuel cell, in which in-phase harmonic perturbations of current density and temperature lower the impedance and static polarisation resistivity, and a special choice of the perturbation amplitudes eliminates the proton transport losses. The essay presents the elimination as something fuel cells can do, and adds that out-of-phase signals give no benefit, that the effect persists at DC and that this rules out resonance; none of that is in the abstract.

It can be completely eliminated.

By oscillating the current density and the temperature in phase, the proton transport loss vanishes. Not reduces — vanishes. The mechanism is parametric, not statistical. Proton conductivity follows Arrhenius temperature dependence, so synchronized temperature oscillations modulate conductivity in lockstep with current. When the modulation amplitudes satisfy a specific ratio — conductivity perturbation times mean current equals mean conductivity times current perturbation — the local overpotential gradient is zeroed out identically.

This is not the same as nonlinear averaging, where oscillating around a curved response function gives a better average than the steady-state value. The cancellation here requires phase matching: only in-phase oscillations work. Out-of-phase signals produce no benefit. And the effect persists at DC — the reduction occurs even at vanishingly low frequencies, ruling out any resonance-based explanation.

What makes this striking is not the specific application but the structural lesson. A loss that appears intrinsic to the physics of proton transport turns out to be an artifact of assuming stationarity. The steady-state analysis treats conductivity as a fixed parameter and current as the only variable. By allowing both to oscillate together, the gradient that generates the loss is actively compensated rather than endured.

Synchronized oscillation annihilates a resistance that steady-state operation treats as fundamental. What looks like an inherent cost may be an artifact of holding constant what could be varied.