Science Route · continuation manual · traveller: iridium species · route: catalyst → dissolution → membrane transport → metallic filament → parasitic pathway → degradation evidence.

Wait, What? A catalyst can leave home

Iridium is used in proton-exchange-membrane water electrolysers because the oxygen-producing side of the device is a demanding place for a catalyst. It is tempting to picture the iridium as a permanent solid layer: electricity comes in, water reacts, gases leave, and the catalyst simply keeps doing its job.

But atoms and ions do not respect our diagrams. Under electrochemical stress, some iridium can leave its original material. In a 2026 study of a PEM water-electrolyser stack operated for thousands of hours, researchers found metallic nanocrystalline iridium filaments extending across the polymer membrane. Those filaments created unwanted electrically conductive paths through a part of the device that is supposed to separate functions.

The startling idea is not merely that a catalyst degrades. It is that material can move, change chemical state, reappear somewhere else and create a new failure mechanism.

Worth My While

This route is useful far beyond electrolysers. It teaches how to read a modern materials-failure story. First identify what was directly observed. Then separate the proposed transport mechanism from the observation. Finally ask how a microscopic structure changes a macroscopic property such as efficiency, resistance or lifetime.

The Big Question

How can iridium leave a PEM-electrolyser catalyst, cross a membrane and appear as a metallic filament—and what parts of that story are observations rather than inferred mechanism?

Quick Answer

Researchers examining a PEM water-electrolyser stack after prolonged operation directly identified metallic nanocrystalline iridium filaments crossing the membrane from the anode side towards the cathode. Imaging and spectroscopy established their location, morphology and composition, while electrochemical measurements linked them to parasitic current pathways.

The proposed formation mechanism is more cautious. Iridium species are thought to dissolve from the anode-side catalyst, migrate into or through the membrane, then be reduced back to metallic iridium where local conditions favour deposition. Hydrogen crossing the membrane is one proposed reducing influence. The study found that filaments formed beneath particular structural features, suggesting that local geometry and chemistry can create preferred routes. That mechanism is evidence-supported, but it is still a model built from multiple observations rather than a film of one labelled iridium ion being watched continuously from start to finish.

Primary → Secondary → JC → Edge

Primary: a solid object can lose tiny pieces even when the object still looks intact. In advanced materials, those pieces may be atoms or ions too small to see directly with ordinary eyes.

Secondary: oxidation can convert a metal atom into an ion, while reduction can convert an ion back towards a metallic state. Ions can move through materials under concentration and electric-field gradients.

JC: catalyst dissolution, ionic transport, membrane chemistry, local electrochemical potential and reduction reactions can be coupled. A failure path can therefore involve more than one redox step and more than one physical region.

Edge: the device is spatially heterogeneous. Weld seams, interfaces, local stresses, hydration and gas crossover can create microenvironments different from the average membrane. A failure mechanism that is rare in bulk material can become important where those conditions concentrate.

Follow One Iridium Species

1. It begins as part of an oxygen-evolution catalyst

On the anode side of a PEM water electrolyser, the catalyst assists the reaction that produces oxygen from water. Iridium-based materials are valued because they can function in the strongly oxidising, acidic environment of this device better than many alternatives. Yet “stable” does not mean “unchanging forever”. Some catalyst material can dissolve or restructure during operation.

2. A solid-bound atom may become a mobile ionic species

If an iridium atom leaves the solid in an oxidised ionic form, its scientific identity has changed. It is no longer simply part of a metallic or oxide lattice. Its charge, hydration and interaction with the ion-conducting membrane determine what routes are now available. That distinction matters: “iridium moved” is too vague. A material can move as a particle, a dissolved species or an atomically dispersed ion, and those routes imply different mechanisms.

3. The membrane is not an empty wall

A PEM membrane conducts protons while helping keep the electrochemical compartments functionally separated. At molecular scale it is a hydrated polymer with nanoscale pathways and chemical groups. Foreign ionic species may interact with that environment. Their transport depends on local chemistry and structure, not on a simple straight tunnel from anode to cathode.

4. Somewhere along the route, ionic iridium can become metallic again

The 2026 work proposed that dissolved iridium species can be reduced, with hydrogen crossover helping provide a chemically reducing environment. Once metallic nuclei form, additional material may accumulate. Repetition can turn atomic-scale deposition into a filament large enough to become an electrically meaningful structure.

5. A filament changes the circuit

A membrane should separate pathways that the device needs to control. A metallic filament crossing it can provide an unwanted electronic route. That does not mean the whole cell becomes an immediate short circuit. It means some current can travel through a parasitic pathway, lowering efficiency and contributing to degradation. The microscopic traveller has become a circuit-level problem.

How Do We Know?

No single instrument tells the whole story. X-ray computed tomography can reveal three-dimensional internal structures without reducing the entire sample to a flat surface. Focused-ion-beam scanning electron microscopy can expose fine cross-sections, while energy-dispersive X-ray spectroscopy helps identify elemental composition. Raman measurements add chemical information. Electrochemical impedance measurements test how those structural changes affect electrical behaviour.

The important methodological lesson is triangulation. A bright feature in an image is not automatically iridium. An elemental signal is not automatically a continuous conductor. An electrical change is not automatically caused by the feature you noticed. Confidence grows when morphology, composition, spatial continuity and electrical behaviour point towards the same explanation.

Nature Catalysis reported that the stack had been assessed under industrially relevant conditions for 5,000 hours with repeated start-up and shut-down events. The filaments were observed beneath particular weld-seam locations and extended through the membrane. The study proposed a mechanism involving dissolved Ir species and reduction associated with hydrogen crossover. The word proposed matters: the filament itself was observed; every intermediate step in its formation was not individually watched in real time.

Observation vs Inference

• Observation: cross-plane filamentary structures were imaged in the membrane.
• Observation: elemental and spectroscopic evidence identified metallic iridium in those structures.
• Observation: the filaments were spatially associated with particular stack features and parasitic current behaviour.
• Inference: anode iridium first dissolved as mobile ionic species.
• Inference: those species migrated through the membrane along the route proposed.
• Inference: hydrogen crossover supplied the local reducing conditions responsible for metallic redeposition.

The inferences are scientifically reasonable because they connect several observations and known electrochemical behaviours. They are not weakened by being called inferences. Science becomes stronger when the boundary between measured structure and reconstructed mechanism is explicit.

Misconception Repair

“A catalyst is not consumed, so none of it can move.” A catalyst is not consumed stoichiometrically by the ideal overall reaction, but a real catalyst can dissolve, restructure, detach or change chemical state.

“If iridium is detected in the membrane, the full transport mechanism is proven.” No. Location is a powerful constraint, but the sequence of dissolution, migration, nucleation and reduction must be supported by additional evidence and modelling.

“One filament means the whole membrane has failed everywhere.” No. Local defects and conductive paths can matter before global failure, but their abundance, continuity and electrical consequences have to be measured.

“The more iridium a device uses, the safer it must be from degradation.” Not necessarily. Loading, dispersion, catalyst support, local operating conditions and material interfaces all matter. More material is not the same as better retention.

Worked Reasoning: The Same Voltage Drift, Two Different Stories

Imagine two PEM electrolysers both require gradually increasing voltage to sustain the same output. In Device A, catalyst particles grow or detach at the anode. In Device B, iridium migrates and forms conductive filaments through the membrane.

The macroscopic symptom—worsening efficiency—does not identify the microscopic cause. To discriminate between the stories, a researcher needs spatial and chemical evidence. Does the membrane contain metallic iridium? Is there a continuous filament? Where is it located relative to structural features? Does impedance behaviour indicate an electronic leakage path? Does the anode show corresponding material loss?

This is why degradation science cannot stop at a performance graph. A graph tells you that something changed. A mechanism requires a chain connecting performance to material state.

Checkpoints

1. What chemical change can make an iridium species more mobile than when it is bound in a solid catalyst?
2. Why does imaging a filament not by itself prove how the filament formed?
3. Why can a nanoscale or microscale filament matter to whole-device efficiency?
4. Why is the membrane better described as a functional material than an empty separator?

Checkpoint Answers

1. Dissolution and oxidation can create ionic iridium species that interact with the hydrated membrane and can be transported.
2. Imaging gives the final or sampled structure; the intermediate dissolution, migration and reduction steps require other evidence and inference.
3. A conductive filament can create an unwanted electronic route across a layer whose electrical separation is essential to device function.
4. The membrane has molecular structure, hydration, fixed chemical groups and selective transport behaviour; those properties influence what can cross it.

Can You Explain WHY?

• Why might repeated start-up and shut-down matter even if average operating performance looks acceptable?
• Why is a preferred filament location beneath a structural feature scientifically interesting?
• Why should catalyst-retention measurements be connected to membrane analysis rather than studied in isolation?
• Why is a proposed reduction mechanism stronger when it also explains where metallic iridium appears?

Singapore and the World

Hydrogen technologies are often discussed at the scale of energy systems, shipping fuels or industrial decarbonisation. This iridium route reminds us that those large ambitions can depend on nanoscale reliability. A country or company can plan gigawatts of electrolysis, but the economics still depend on how long membranes, catalysts and interfaces perform before repair or replacement.

For Singapore, where imported energy carriers, maritime systems and advanced manufacturing all matter, the useful lesson is not that one electrolyser design should be chosen. It is that materials durability belongs inside system-level thinking. Precious-metal supply, stack lifetime, efficiency and maintenance are linked rather than separate columns in a spreadsheet.

Deep Science Window: A Failure Path Can Reverse Chemical Identity

The route contains a beautiful chemical symmetry. An iridium atom can begin in a solid catalyst environment, become an oxidised mobile species, travel, then be reduced and return to a metallic state elsewhere. The beginning and end may both contain “iridium”, but they are not the same scientific object. Oxidation state, bonding, location and function have changed.

That is why materials science preserves state. Saying “iridium moved from A to B” loses the mechanism. A better chain is: catalyst-bound iridium → dissolution/oxidation → mobile species → transport through a hydrated polymer environment → reduction/nucleation → metallic filament → parasitic electronic conduction.

Counterexamples and Model Limits

Iridium migration is not the only degradation route in PEM electrolysers. Catalyst-layer restructuring, membrane chemical damage, mechanical defects, transport limitations and contamination can also reduce performance. A filament mechanism observed in one stack architecture does not imply that every PEM electrolyser fails in the same way.

Nor does spatial correlation with a weld seam prove that the seam alone caused the filament. The feature may change local pressure, contact, transport or chemistry in a way that favours nucleation. Each link needs separate testing.

Evidence Boundaries

This page is educational and non-operational. It does not provide electrolyser construction parameters, high-current operating procedures, catalyst fabrication recipes or hydrogen-handling instructions. It also does not claim that metallic iridium filaments are the dominant lifetime limit in every commercial stack. The evidence establishes a real degradation pathway in the studied system and offers a mechanism to test in others.

KNOW → CONNECT → EXPLAIN → APPLY → CHECK

• KNOW: catalyst atoms can change state and leave a solid.
• CONNECT: dissolution links electrocatalysis to membrane transport and materials failure.
• EXPLAIN: transported iridium species may be reduced and accumulate into metallic filaments that create parasitic current pathways.
• APPLY: when performance degrades, ask which structural and chemical observations connect the symptom to a mechanism.
• CHECK: keep the observed filament separate from the inferred sequence that produced it.

eduKateAI Direction Graph

PEM electrolyser → anode catalyst → iridium dissolution → ionic state → membrane transport → local reduction → metallic nucleation → filament growth → parasitic electronic path → efficiency loss → degradation diagnosis → catalyst-retention and membrane-owner routes.

Where to Go Next

Hand the underlying mechanisms back to their specialist owners: electrochemistry for redox and electrode kinetics, polymer science for membrane transport, materials science for nucleation and microstructure, and hydrogen engineering for complete stack operation. This Science Route owns only the traveller story that connects them.

Authoritative Sources

• Nature Catalysis (26 February 2026): Degradation through filaments
• Nature Communications (2026): Dual-interface stabilization of low-iridium anodes for durable PEM water electrolysis
• U.S. Department of Energy H2NEW: electrolyser durability, performance and manufacturing research

Teaching Guide for Parents, Tutors and Teachers

Begin with a familiar distinction: a bridge can fail because material disappears from one place or because unwanted material appears in another. Then ask the learner to trace both sides of the iridium story. What left the catalyst? What appeared in the membrane? What evidence would connect the two?

For younger learners, focus on the idea that materials can change and migrate. For Secondary students, use oxidation and reduction to explain how a metal can become an ion and later return to metal. For JC students, introduce transport through a membrane and distinguish local chemical environment from bulk conditions. For advanced learners, require an evidence map with separate boxes labelled observation, inference and alternative explanation.

A strong diagnostic question is: “If you found iridium halfway through the membrane, what would you know for sure, and what would you still need to prove?” The best answers preserve the difference between composition, origin, pathway and cause.