PEM catalysts ·

Ultra-Low Iridium PEM Catalyst: OXYGN-M Test Data

Roll-to-roll coating machine with a precision coating head, rollers and dryer section for continuous electrode production.

Iridium is the uncomfortable line item in every PEM electrolyser bill of materials. It is one of the rarest elements in the Earth's crust, it is recovered only as a by-product of other precious-metal mining, and global output is measured in single-digit tonnes a year. Yet a conventional PEM anode still carries around 2 mg of iridium per cm². If PEM is to reach gigawatt scale, that loading has to fall by an order of magnitude.

Many catalysts promise this. The question buyers should ask is simple: at what loading, at what current density, and for how many hours? Here are our answers for OXYGN-M™, the ultra-low iridium anode catalyst we developed for PEM water electrolysis.

0.2 mg Ir/cm²: ten times less iridium

At only 0.2 mg Ir/cm², a tenth of the conventional loading, OXYGN-M electrodes deliver:

  • 1.86 V at 2 A/cm² at 80 °C
  • more than 1,000 hours at a constant 2 A/cm² without degradation

That is the operating point that matters for modern PEM stacks: high current density, at a loading low enough to take iridium off the list of supply risks. We presented these results at the International Conference on Electrolysis (ICE 2025). The full peer-reviewed paper on the 0.2 mg results is in revision, and we will link it here as soon as it is published.

The design principle, peer-reviewed

The route to ultra-low loading is catalyst-layer design, not just a more active powder. In a study with the German Aerospace Center (DLR) and the University of Stuttgart, published in the Journal of Power Sources in 2026, we describe what the authors call a self-governed catalyst layer. During operation, the anode layer organises itself into a well-connected catalyst network with a favourable pore structure. More of the iridium you pay for does useful work, and the layer stays stable instead of slowly losing active area. Microscopy before and after operation showed better access to active sites and more electronically conductive regions in the layer.

The paper benchmarked the self-governed layer against a conventional 2 mg Ir/cm² anode under identical conditions (80 °C, ambient pressure, a standard PEM membrane): it matched the benchmark with a fraction of the iridium and ran stably in long-term operation. Our product work applies the same design principle at 0.2 mg Ir/cm², the loading behind the results above.

The published abstract also notes that CCMs with OXYGN-M are commercially available from CENmat. That is unusual: this is not a lab curiosity waiting for a separate scale-up programme.

Read the paper: S. S. Ambu, T. Morawietz, B. Kimmel, M. Goll, A. S. Gago, S. S. Hosseiny, K. A. Friedrich, "Self-governed catalyst layer architecture as a key design strategy for low iridium proton exchange membrane water electrolyzers", Journal of Power Sources 672 (2026) 239614. doi.org/10.1016/j.jpowsour.2026.239614

What 0.2 mg means for a stack

Take a stack running at 2 A/cm² and about 2 V. At 2 mg Ir/cm², each megawatt needs roughly 0.5 kg of iridium on the anode. At 0.2 mg Ir/cm², that falls to about 50 g. Operating point, active area and yield change the exact figure, and our AEM vs PEM iridium calculator lets you plug in your own numbers. At gigawatt scale, a tenfold cut in anode iridium separates a supply-constrained technology from a scalable one.

Lower loading also means thinner catalyst layers, which are easier to coat evenly. It has limits: very thin anodes can lose contact with the porous transport layer or become fragile. That is exactly why the durability data matter. They show those failure modes did not appear over more than 1,000 hours at full current.

Five questions to ask any ultra-low iridium supplier

  1. At what loading was the headline voltage measured, and at what current density? A good number at 1 A/cm² says little about 2–3 A/cm² operation.
  2. How long was the durability test, and was it at constant current? Constant current at 2 A/cm² is far harder than a test at low load.
  3. Was it benchmarked against a conventional anode in the same hardware? Without a side-by-side reference, absolute voltages depend heavily on the cell and membrane.
  4. Is the result peer-reviewed or independently reproduced?
  5. Can you buy it as a CCM, not just as a powder?

We try to answer all five for our own materials. For the degradation mechanisms behind these questions, see our note on reducing iridium loading, and for a second peer-reviewed ultra-low iridium result, our summary of the 2.82 A/cm² paper.

Getting OXYGN-M

OXYGN-M is available as catalyst powder and as ready-to-use catalyst-coated membranes for PEM electrolysis. Specifications and formats are on the OXYGN PEM product page. To test it in your own cell, request samples and tell us your operating window; we will suggest a starting loading.

Sources

  1. CENmat OXYGN-M results at 0.2 mg Ir/cm², presented at the International Conference on Electrolysis (ICE 2025), 25 Aug 2025; peer-reviewed manuscript in revision
  2. S. S. Ambu et al., Journal of Power Sources 672 (2026) 239614

Ready to discuss your project? CENmat develops AEM and PEM electrolysis from material to system: AionFLX™ membranes and ionomers, OXYGN and H2GN catalysts and electrodes, ultra-low iridium PEM catalysts and CCMs, and HYScale stacks and systems. Request samples, get datasheets, get in touch or keep reading the CENmat blog.