The Pre-Catalyst Illusion

Pick up almost any catalysis journal from the past five years and you will find dozens of papers celebrating newly synthesized nickel-iron phosphides, sulfides, selenides, or bimetallic metal-organic frameworks (MOFs) for the alkaline Oxygen Evolution Reaction (OER). Authors routinely dedicate pages to characterizing pristine crystal structures with XRD and high-resolution TEM, only to credit that initial architecture for the low overpotential measured hours later.

Electrochemically speaking, that narrative does not hold up.

The moment a nickel-iron chalcogenide, pnictide, or coordination polymer enters 1.0 M KOH under an oxidative anodic potential, bulk thermodynamics takes over. Non-metal anions (P, S, Se) and organic linkers rapidly oxidize and leach into the electrolyte. Driven by the harsh alkaline bias, the surface undergoes irreversible structural collapse and reorganizes into an amorphous or disordered nickel-iron oxyhydroxide (Ni₁₋ₓFeₓOOH) layer.

What we spend weeks synthesizing at the bench is not the true catalyst, but a sacrificial pre-catalyst. The actual workhorse splitting water is the reconstructed oxyhydroxide skin formed in situ during the first few cyclic voltammetry sweeps.

The Ni-Fe Synergy and the Mechanism Shift

The electrocatalysis community remains committed to the nickel-iron pair for a simple reason: in alkaline media, the electronic synergy between Ni and Fe routinely outperforms noble metal benchmarks like IrO₂ and RuO₂ at a fraction of the cost.

Pure nickel hydroxide cycles through the classic Bode scheme, transitioning from α/β-Ni(OH)₂ to oxidized γ/β-NiOOH, yet it remains a mediocre OER catalyst on its own. Introduce even trace amounts of iron into that NiOOH host lattice, however, and the overpotential drops by more than 100 mV. Substitutional Fe³⁺ cations occupying edge and defect sites within the NiO₆ octahedral sheets induce strong local lattice strain and partial charge transfer. The conductive NiOOH host acts as an electronic reservoir that stabilizes high-valent iron (Fe⁴⁺) and nickel (Ni³⁺/⁴⁺) centers while tuning metal 3d and oxygen 2p orbital hybridization.

Crucially, deep surface reconstruction often shifts the reaction pathway away from the conventional Adsorbate Evolution Mechanism (AEM), where catalysis stays confined to metal centers, toward the Lattice Oxygen Mediated (LOM) mechanism. Along the LOM pathway, strongly covalent metal-oxygen bonds allow oxygen atoms directly from the NiFeOOH crystal lattice to couple with incoming hydroxide intermediates. This bypasses the thermodynamic scaling relations that bottleneck conventional catalysts, but it comes at a steep structural price.

When Reconstruction Turns Destructive

This brings us to the central paradox of modern OER catalyst design: the very mechanisms that unlock peak catalytic activity also accelerate material failure.

When a reconstructed Ni₁₋ₓFeₓOOH layer operates via the LOM pathway, it continuously cycles oxygen vacancies in and out of its own structural backbone. At academic benchmark currents (10 mA cm⁻²) and room temperature, the lattice heals itself fast enough to appear stable over a 24-hour run. Push that same electrode to industrial alkaline electrolyzer conditions (500 to 1,000 mA cm⁻² in 6.0 M KOH at 80 °C), and the balance falls apart.

Under intense oxidative polarization, the high-valent iron centers driving OER activity become thermodynamically unstable and oxidize further into soluble ferrate (FeO₄²⁻) species that dissolve straight into the bulk electrolyte. Meanwhile, the remaining iron ions migrate out of the active NiOOH host matrix and segregate into catalytically sluggish FeOₓ or α-FeOOH islands. Worse still, if the starting pre-catalyst is overly porous or leaches anions too aggressively, reconstruction does not stop at a thin, conductive surface skin. It eats deep into the bulk core, severing electrical contact with the underlying current collector and triggering mechanical spallation as oxygen gas bubbles nucleate inside the crumbling pores. Once the active iron sites bleed out of the reconstructed layer, the surface reverts to an iron-depleted β/γ-NiOOH crust and the cell voltage climbs irreversibly.

Engineering the Reconstruction Boundary

If earth-abundant Ni-Fe anodes are going to survive multi-year industrial operation, we have to stop treating surface reconstruction as a fortunate accident and start engineering its limits. Three design principles are proving particularly effective:

• Self-limiting core-shell architectures: Rather than building fully sacrificial MOFs that convert entirely into poorly conductive hydroxides, we need dense intermetallic or alloy cores capped by controlled, nanometer-thin reconstruction zones. Arresting anion leaching after 2 to 5 nanometers preserves the metallic conductivity and mechanical anchoring of the core while presenting a high-density NiFeOOH skin to the electrolyte.
• Cationic lattice locking: Incorporating high-valence or oxyphilic third cations such as Cr³⁺/⁶⁺, V⁵⁺, Mo⁶⁺, or W⁶⁺ provides an electronic buffer within the oxyhydroxide lattice. These dopants modulate metal-oxygen covalency, curb runaway lattice oxygen loss, and raise the kinetic barrier for iron demetallation during high-current operation.
• Dynamic dissolution-redeposition equilibrium: Instead of fighting iron dissolution strictly within the solid state, we can manage it across the solid-liquid interface. Maintaining a tightly controlled chemical potential of dissolved iron species in the circulating KOH loop allows leached Fe ions to continuously redeposit onto exposed NiOOH edge sites, turning a degrading surface into a self-healing catalytic cycle.

Retiring the 24-Hour Beaker Test

Moving these materials from academic papers into gigawatt-scale electrolyzers also demands more honest testing protocols. Claiming "exceptional long-term stability" from a 24-hour chronopotentiometry curve at 10 mA cm⁻² in a standard glass H-cell tells us almost nothing about commercial survival. That is especially true when hot KOH etches laboratory glassware, releasing trace silicates and adventitious iron that deposit on the anode and artificially mask degradation.

Until we routinely pair operando spectroscopy (such as XAS and Raman) with online or downstream ICP-MS to track real-time metal dissolution at industrial current densities and temperatures, we are only measuring half the picture.

I want to put a question to colleagues working across electrocatalysis and electrolyzer scale-up: To keep Ni-Fe anodes alive above 500 mA cm⁻², is it more practical to engineer multi-element doped lattices that suppress the LOM pathway and lock Fe in place, or to accept dynamic Fe leaching and dose trace iron directly into industrial KOH feed loops? Let’s discuss in the comments.

#Electrocatalysis #WaterSplitting #GreenHydrogen #OxygenEvolutionReaction #NickelIron #SurfaceReconstruction #Electrochemistry #MaterialsScience #CleanEnergy #SpringerNature