The key point is simple: performance loss is rarely a single failure event. It’s a collection of small, interacting changes,chemical, mechanical, and transport-related, that accumulate until voltage rises, efficiency drops, gas purity suffers, and maintenance becomes inevitable. That is why monitoring is not an academic luxury. It is the shortest path to understanding which mechanism dominates, where it happens, and how to stop it.

That is why I believe the most powerful question in hydrogen R&D is not “how high can we push the current density?” but rather: “What happens after 1,000… 10,000… 30,000 hours, and how do we know early enough to do something about it?”

Alkaline and Anion exchange electrolysis: the cost promise comes with a durability bill

At CIC energiGUNE, we place a strong emphasis on alkaline water electrolysis (AWE) and anion-exchange membrane water electrolysis (AEMEL), because these platforms offer a compelling cost pathway: non-precious catalysts, less dependency on critical raw materials, and potentially simpler stack architectures.

But the price advantage only becomes real if durability follows.

AEM electrolysis, in particular, sits in an interesting “middle ground”: it wants to combine the compactness and differential-pressure operation of PEM (Proton Exchange Membrane)-style MEAs (Membrane Electrode Assembly) with the catalyst flexibility of alkaline chemistry. The challenge is that AEM systems concentrate a lot of complexity in the catalyst layer and ionomer: the ionomer must bind catalyst particles, provide OH⁻ conductivity pathways, and survive strongly alkaline and oxidative conditions, especially near the anode. Recent literature is increasingly explicit that the ionomer in the catalyst layer can be the limiting component, not just the membrane itself.

And then there is classic alkaline electrolysis, where the separator/diaphragm and the two-phase flow environment are decisive. Alkaline systems are often seen as “mature,” but modern AWE is evolving quickly toward zero-gap and narrow-gap designs, higher current densities, and thinner separators, all of which raise the stakes for controlled gas removal, crossover management, and mechanical robustness.

What actually degrades? Three families of failure modes you can’t ignore

1. The ion-conducting barrier (membrane, diaphragm, separator)

This component has one job: allow ions through while keeping gases apart. In the real world it also has to survive pressure gradients, temperature swings, chemical attack, swelling/shrinkage cycles, and impurity exposure.

In alkaline and AEM systems, a recurring trade-off appears again and again:

• Thinner = lower ohmic losses, but higher vulnerability to pinholes, mechanical creep, and crossover risk.
• Smaller pores / higher tortuosity = lower crossover, but higher resistance and more sensitivity to wetting and bubble management.

In other words, separator engineering is not “passive”: it’s a central design lever, and a common failure origin.

2. The catalyst layer and its interfaces (where performance is born and where it can quietly die)

In AEM electrolysis, durability is often determined within the catalyst layer, including the catalyst, ionomer, pore structure, and transport layers. Even if the catalyst is stable, the system can degrade if the ionomer loses conductivity, detaches from the catalyst surface, or chemically fragments under oxidative stress.

For PEM, the story is different but equally instructive: expensive iridium-based anodes remain the benchmark, yet operando studies show that catalysts restructure under oxygen evolution, forming highly disordered active oxide motifs that evolve with potential and current. This matters because it reminds us of a universal principle: the active state is often not the as-manufactured state, and degradation sometimes starts as “normal activation” that slowly drifts too far.

3. Two-phase transport (bubbles are not just a nuisance; they are a degradation driver)

This is where alkaline and AEM really demand attention. Gas evolution creates bubbles everywhere, inside porous electrodes, at catalyst surfaces, in channels, and within small gaps. If you manage bubbles well, you improve efficiency. If you manage them poorly, you create local dry-out/flooding, current hotspots, and uneven aging.

The takeaway from several studies is blunt: bubble physics is lifetime physics. Two-phase flow determines local resistances, local concentrations, and local mechanical stress, and those local conditions decide which component fails first.

Why monitoring changes the game: from post-mortem to prevention

Traditional durability work often follows a painful script: run for hundreds of hours, observe a voltage rise, disassemble the cell, and try to guess the cause. The problem is that many degradation pathways leave similar electrical fingerprints.

That is why fundamental and applied research in water electrolyzers is moving toward multi-modal, operando, and actionable monitoring, combining:

• Electrochemical tools (polarization breakdown, EIS, accelerated stress tests)
• Materials analysis (electron microscopy, in situ  and in operando spectroscopy, chemical and structural probes)
• Imaging and spatially resolved methods (X-ray / neutron techniques to map gas and water)

These approaches are especially important for AEM and alkaline technologies because performance is often dominated by transport + interfaces, not just catalyst intrinsic activity.

Fuel cells: a mirror that helps us design better electrolyzers (and vice versa)

Fuel cells and electrolyzers share a deep symmetry: they are MEA-based devices where ionomers, catalysts, and water management decide performance and lifetime.

Fuel cell research has taught the community hard lessons about chemical degradation of polymer electrolytes and ionomers, interface stability under cycling and the long lag between “looks stable” and “fails suddenly”.

Those lessons translate directly into AEM electrolysis, where ionomer stability and electrode–ionomer integrity are emerging as central bottlenecks. The benefit of viewing both technologies together is that monitoring strategies, especially those that separate membrane effects from electrode/ionomer effects, can be transferred, adapted, and accelerated.

The bottom line

If we want electrolyzers and fuel cells to become infrastructure, we must treat durability as a design parameter, not as an afterthought. And to do that, we need to observe degradation while it is happening, link it to local operating conditions, and transform that knowledge into smarter materials and architectures, particularly in alkaline and AEM systems, where bubble management, separator properties, and ionomer behavior tightly couple into lifetime.

At CIC energiGUNE, we can support partners with a state-of-the-art, complementary toolbox to monitor MEA degradation as it happens and then explain it at the materials level. We combine in situ/operando electrochemical diagnostics with advanced structural and chemical analysis. This includes surface-sensitive tools such as XPS to track chemical/electronic changes and contamination at interfaces, HRTEM to resolve nanoscale restructuring, dissolution, and redeposition phenomena in catalysts and ionomers, and X-ray tomography to visualize microstructural evolution and transport-layer changes.

A particularly powerful asset in this context is our electrochemical mass spectrometry (ECMS/DEMS) platform, which enables real-time detection and quantification of gaseous products and parasitic species during operation. By directly coupling electrochemical cells to mass spectrometry, we can identify subtle changes in gas evolution, crossover, side reactions, and degradation by-products long before they become visible in voltage drift alone. This capability is especially relevant for polymer electrolyte systems, where two-phase transport, ionomer instability, and interfacial reactions can silently reshape performance. The combination of ECMS with structural and chemical post-analysis provides a rare level of mechanistic resolution, allowing partners not only to observe degradation, but to understand its origin and kinetics under realistic operating conditions.

Together, these techniques allow us to connect device-level drift to the evolution of specific MEA components so partners can turn durability observations into concrete design and validation steps.