A retired PEM electrolyser stack looks like heavy industrial equipment. The steel end plates, titanium components, and compressed cell assemblies make it easy to classify as scrap when you upgrade or decommission a hydrogen plant. That classification can be expensive. Each stack contains very thin catalyst layers loaded with iridium (one of the world's rarest commercially important metals). The end-of-life question isn't just what the stack weighs as scrap. It's how much precious metal it contains, where that precious metal has migrated during the stack's life, and how to recover it safely.
PEM electrolysis splits water into oxygen and hydrogen across a solid polymer electrolyte membrane. On the anode side, the oxygen evolution reaction occurs in a very acidic, strongly oxidizing environment. Few catalysts perform effectively in that environment for thousands of hours, and iridium oxide offers the right balance of activity and durability to be the catalyst of choice for commercial PEM anodes.
Earlier-generation systems used much more iridium by today's standards. A typical loading of 2 mg/cm², along with the usual operating assumptions, equates to around 500 kg of iridium per GW of capacity. That means a 1 MW stack could contain hundreds of grams, and a multi-MW stack could contain several kg. The actual amounts in any given stack will, of course, vary by manufacturer, stack design, and catalyst specification. Iridium is also very difficult to replace via additional mining, as it is a minor byproduct of platinum group metal production (mainly in southern Africa). PGM producers can't increase iridium output without increasing primary-metal output, which drives mine economics. Recovery from retired stacks is therefore both a commercial opportunity and a way of getting that constrained material back into the supply chain.
Stack failure doesn't necessarily mean the catalyst has been used up. Membranes can thin or develop defects over time, catalyst layers can delaminate, and cell voltage can exceed acceptable operating limits. Some of the iridium may dissolve and go with the process water, especially under severe operating conditions. Much of the inventory will still be present on the membrane or elsewhere in the cell assembly; how the material is distributed matters when you dismantle it. A recovery plan focused solely on the intact catalyst-coated membranes would likely miss material trapped in the porous transport layers, flow channels, and so on. Some of the titanium will also have precious-metal coatings, though whether those coatings are worth recovering depends on the loading and the process.
The valuable material is in extremely small masses. A catalyst-coated membrane can have iridium at concentrations many times above those found in natural platinum group metal ores. The whole stack, with its heavy structural components, is not the same grade as the ore. The comparison does highlight the key point, though. If you mix a small amount of high-value catalyst into a general scrap stream, you lose the information and the ability to handle it in a way that lets you realize its value.
Conventional metal recycling focuses on bulk commodities rather than the thin precious-metal coatings bonded to the fluorinated membranes. A buyer who's pricing an intact stack essentially for its steel and titanium is likely to assign very little or no value at all to the iridium. The subsequent processing will tend to break up the precious metals into residues from which recovery is either very difficult or uneconomic.
The membrane presents a distinct handling issue. PEM systems typically use perfluorosulfonic acid polymers. If the fluorinated material is heated without appropriate controls, decomposition can occur, generating HF (hydrogen fluoride) and other hazardous, corrosive emissions. The applicable waste classification and associated legal obligations vary by jurisdiction and material condition; not every such membrane is automatically covered by the same PFAS hazardous substance rule. That said, you shouldn't assume a standard scrap furnace can or is authorized to process it. Separate the catalyst-laden fraction from the fluorinated fraction before commodity recycling to protect both the material's value and the downstream processor.
The ideal time to bring in a competent precious metal refiner is at the start of the decommissioning planning process, before you break down or sell the stack. That's when you can work with the plant and the refiner to establish a safe way to prepare and transport the stack, the required chain-of-custody documentation, and the component-by-component recovery plan. At the receiving facility, the stack can be carefully deconstructed to separate the structural metals, titanium components, gaskets, and catalyst-bearing materials. The processing will vary depending on the material and the equipment permitted by the refiner (chemical or controlled thermal treatment with appropriate emissions control). No single process fits every stack design. The aim is to concentrate the precious metals safely with respect to the fluorinated materials. Sampling is as important as the extraction. Iridium-rich particles can be non-uniformly distributed through membranes, coatings, and processing residues. A settlement assay based on a small, unrepresentative sample can be completely wrong for the whole lot. Operators must therefore require a fully documented procedure for weighing, drying, homogenizing, and splitting the sample, with retained portions available for independent analysis or use in a dispute.
Analytical methods must also be appropriate for iridium. A standard lead fire assay (so widely used for the other precious metals) should not be assumed to be an appropriate method for determining iridium in these materials. A qualified lab may use a validated nickel-sulfide collection, peroxide fusion, or another appropriate preparation method, followed by instrumental analysis. The contract must specify both the assay method and what to do if the operator's result differs from the refiner's.
A refining proposal should be transparent about its economics: the assayed metal content, the % payable to the operator, the treatment and refining charges, the price benchmark, and the expected settlement schedule. Gross metal value is not net payable value. At a multi-megawatt scale, iridium recovery can significantly affect a decommissioning project's economics. Retired PEM stacks should therefore be included in the project plan as precious metal-bearing assets that require controlled handling rather than as miscellaneous equipment waste. That approach gives the operator a better chance of recovering the value they paid to install it, avoids the stacks going down the wrong scrap route, and gets the iridium back into the next generation of hydrogen equipment.
Plant teams preparing a teardown should speak with Phoenix Refining before removing the first stack. The essential step is to establish a verified inventory, a safe processing route, and an auditable assay, so the stack's industrial exterior doesn't determine its fate.
