Proton exchange membrane (PEM) water electrolyzers are among the most promising routes to clean hydrogen, but their poor durability remains a central obstacle to large-scale commercialization. Because so many degradation processes unfold simultaneously inside a working cell, experiments alone struggle to quantify how much each mechanism contributes to overall performance loss. A new open-access study published in Carbon Neutrality on 28 October 2026 addresses this gap with a mechanistic model that simulates the entire degradation history of a PEM electrolyzer, from startup to failure, and identifies which mechanism kills the device under different operating conditions.
The model, developed by Boshi Xu, Qiang Liao and colleagues, describes three dominant degradation processes: chemical degradation of the ionomer, dissolution of the iridium dioxide (IrO2) catalyst, and passivation and corrosion of titanium components. Rather than treating these in isolation, the researchers coupled all three sub-models into a three-dimensional full-electrolyzer multiphysics simulation that resolves temperature, charge transfer, electrochemical reactions, liquid-gas flow and water transport across the membrane. This coupling matters because degradation rates depend on local physical fields such as current density, temperature and membrane water content, which themselves change as the cell degrades.
Each sub-model rests on established chemistry. For the catalyst layer, the team adopted the widely accepted framework in which IrO2 forms an IrO2(OH) intermediate during the oxygen evolution reaction; most of this intermediate cycles back to IrO2 without mass loss, but a small fraction branches into soluble trivalent iridium species that permanently remove catalyst. For the ionomer, oxygen permeating the membrane reacts at the cathode to form hydrogen peroxide, and trace Fe2+ impurities in the feedwater drive a Fenton reaction that generates hydroxyl radicals. These radicals cleave the side chains and then the backbone of the polymer, destroying the sulfonate groups that conduct protons. For titanium bipolar plates and porous transport layers, the model tracks the growth of a semiconducting TiO2 passivation film, which raises contact resistance, alongside its dissolution in acidic fluoride-containing environments.
To anchor the model in reality, the researchers ran a 300-hour durability test on an electrolyzer with a 2.6 by 2.6 centimeter active area, operating at 80 degrees Celsius and 1.5 amperes per square centimeter, with electrochemical and physical characterization before and after. Fluoride released by ionomer breakdown appeared in the very first sample at 50 hours, confirming that degradation begins immediately after startup, and stabilized at 0.35 to 0.38 micrograms per hour per square centimeter after 100 hours. Scanning electron microscopy showed the membrane thinning unevenly from an initial 127 micrometers to as little as 121 micrometers in places, reflecting spatial variations in local current density, temperature and water distribution.
The experimental signatures of each mechanism matched the model’s assumptions. Electrochemical impedance spectra showed higher high-frequency resistance at end of test, indicating reduced proton conductivity, and higher charge transfer resistance attributable to the anode catalyst. Cyclic voltammetry revealed declining peak currents and a drop in electrochemically active surface area from 113 to 17 square meters per gram, evidence of substantial catalyst loss. Contact resistance between the bipolar plate and porous transport layer surged from 0.602 to 2.235 milliohms at 2 megapascals, roughly 3.5 times its initial value. Fitted against these data, the simulated voltage evolution tracked the experiment with a maximum absolute error of 20 millivolts, and simulated and experimental degradation rates came out at 173 and 178 microvolts per hour, respectively.
With the validated model in hand, the team explored how operating conditions shape lifetime. Simulations across 50 to 90 degrees Celsius showed that hotter operation shortens life dramatically: predicted lifetime fell from 7000 hours at 50 degrees to 1445 hours at 90 degrees, while the degradation rate climbed from 25 to 76 microvolts per hour. The failure mechanism also shifted. At 50 to 80 degrees, failure occurred when the operating voltage reached the criterion of 110 percent of its initial value, whereas at 90 degrees the cell failed because the catalyst particles had dissolved to 80 percent of their initial size, with the voltage still well below the threshold. Ionomer degradation dominated overall at every temperature, contributing 42 to 61 percent, while the share of IrO2 dissolution rose from 13 percent at 50 degrees to 38 percent at 90 degrees.
Current density proved even more punishing. Raising the current density from 1 to 3 amperes per square centimeters cut the simulated lifetime from 6440 hours to just 520 hours and accelerated the degradation rate from 16.7 to 259.6 microvolts per hour, roughly a fifteenfold increase. The early-life behavior was dominated by titanium: at 3 amperes per square centimeter, Ti degradation accounted for 58 percent of degradation in the initial period, as the passivation film grew rapidly before stabilizing at a constant thickness, in some cases within as little as 60 hours. Above 1.5 amperes per square centimeter, failure was governed by the catalyst particle-size criterion rather than voltage, and at the highest current densities the IrO2 contribution became comparable to that of ionomer degradation.
The model also probed structural and water-quality factors. Trace Fe2+ impurities in the feedwater shortened predicted lifetime from 3440 hours at 0.01 ppm to 3120 hours at 5 ppm, with the effect saturating beyond roughly 0.1 ppm; above that threshold, ionomer degradation governed failure. Shrinking the initial catalyst particle size from 50 to 10 nanometers reduced lifetime from 3470 to 2660 hours and raised the degradation rate from 47.8 to 57.7 microvolts per hour, because small particles with high surface free energy dissolve faster. Lowering the anode catalyst loading from 2 to 0.1 milligrams per square centimeter, the range spanning commercial to state-of-the-art electrodes, cut lifetime from 3210 to 1000 hours, with catalyst dissolution becoming the dominant degradation pathway at the lowest loading.
The authors are careful to note limitations. The reported contribution percentages are relative indices rather than a strict mathematical decomposition, since the three mechanisms are fully coupled through the multiphysics framework. The model covers only chemical and electrochemical degradation, omitting mechanical failure modes, and it examines each operating and structural factor individually rather than their nonlinear interactions. Lifetime projections rest on a membrane-thinning integrity criterion rather than full gas-crossover safety modeling. Still, by capturing the stage-dependent character of degradation, rapid early voltage rise driven by titanium passivation and ionomer breakdown, followed by slower, catalyst-dominated decline, the framework offers engineers a quantitative tool for predicting electrolyzer lifetime and identifying, before the hardware fails, which mechanism will end its service life.
Subject of Research: Mechanistic modeling of long-term degradation and failure in PEM water electrolyzers
Article Title: Mechanism model for long-term degradation failure of PEM water electrolyzers
Article References: Xu, B., Yang, Y., Wang, X., Tao, D., Wang, Y., Perwez, A., Li, J., Zhu, X., & Liao, Q. (2026). Mechanism model for long-term degradation failure of PEM water electrolyzers. Carbon Neutrality, 5(1), Article 32. https://doi.org/10.1007/s43979-026-00187-6
Image Credits: AI Generated
DOI: 10.1007/s43979-026-00187-6
Keywords: PEM water electrolysis, hydrogen, degradation, IrO2 dissolution, ionomer degradation, titanium passivation, durability, mechanistic model, electrolyzer lifetime, current density, operating temperature, water purity
News Source: Denise Maddox. (October 29, 2026). New Model Tracks How PEM Electrolyzers Fail Over Thousands of Hours. Scienmag.
