Vanadium monophosphide catalyst for earth-abundant hydrogen evolution

Platinum-group metals dominate commercial hydrogen-evolution cathodes for a simple reason: the Sabatier volcano places platinum near the thermodynamic optimum for hydrogen binding, with a Gibbs free energy of adsorption close to zero. The problem is cost and scarcity. Platinum at cathode loadings of several milligrams per square centimeter represents a structural barrier to terawatt-scale green hydrogen, and the entire electrolyzer industry has accepted that constraint largely because credible earth-abundant alternatives have been slow to emerge at the same combination of activity, stability, and scalability. Vanadium monophosphide in its Pnma ground-state polymorph, with the (110) surface preferentially exposed, changes that calculus. The thesis here is not that VP is a marginal cost reduction on a comparable material. The thesis is that the (110)-faceted Pnma polymorph delivers a hydrogen adsorption free energy within the range where meaningful exchange-current density is expected — between +0.036 eV and +0.098 eV depending on which machine-learning potential is used — and does so without any carbon support, without any precious-metal co-catalyst, and without proprietary precursor chemistry. Vanadium and phosphorus are both abundant, inexpensive, and produced at scale for existing industrial purposes. A support-free architecture also simplifies electrode manufacturing and eliminates degradation pathways associated with carbon corrosion under the oxidizing conditions of system startup and shutdown. The asset sits within the broader catalysts and energy-conversion materials portfolio as a composition-plus-device-use claim: a specifically crystallographically defined material in a specifically defined application. The polymorph specificity is a feature, not a limitation. It allows the claim to be defensible and freedom-to-operate clean while still covering the commercially relevant form. The timing is favorable: electrolyzer OEMs are under active margin pressure to replace platinum cathodes, supply-chain regulators in the EU and US are scrutinizing PGM dependence, and the materials-by-design approach taken here means the computational case for this composition was built before the experimental campaign rather than after.

Each candidate is validated by multiple independent machine-learning interatomic potentials. A material advances only when the engines agree on phonon (dynamic) stability — disagreement is surfaced, not hidden.

Vanadium monophosphide crystallizes in the orthorhombic Pnma space group (No. 62) under ambient synthesis conditions, and this polymorph is the thermodynamic ground state on the VP phase diagram. The Pnma structure features edge-sharing VP6 octahedra arranged in a layered motif that exposes vanadium-phosphorus mixed-termination surfaces with favorable surface-charge density for proton-coupled electron transfer. The (110) surface cut of this structure presents a particularly attractive geometry for the Volmer-Heyrovsky hydrogen-evolution mechanism: the surface vanadium sites offer moderate d-band occupancy, and the adjacent phosphorus atoms provide a secondary adsorption locus that helps tune the effective binding energy away from the extremes that plague many transition-metal surfaces. Hydrogen adsorption free energy (dG_H) was calculated using two independent machine-learning interatomic potentials, MACE and CHGNet, yielding values of +0.036 eV and +0.098 eV respectively. Both potentials agree directionally — the material binds hydrogen slightly endothermically, placing it on the product-side slope of the Sabatier volcano near the optimum — and the numerical spread of roughly 60 meV is within the expected inter-model uncertainty for this class of calculation. This is what the field calls a thermodynamically promising adsorption regime: the surface does not bind hydrogen too tightly (which would poison active sites and suppress turnover) nor too weakly (which would limit proton capture). The agreement across two independently parameterized models meaningfully reduces the risk that the favorable dG_H is an artifact of one potential's training-set bias. Dynamic stability — the question of whether the crystal will sustain oscillatory atomic motion without structural collapse — is addressed through phonon calculations on the Pnma polymorph specifically. Three independent machine-learning potentials (MACE, CHGNet, and ORB) all return positive phonon frequencies across the Brillouin zone for the ground-state structure, meaning no imaginary vibrational modes are present and the crystal occupies a true local energy minimum. Additionally, bulk energy calculations across four independent potentials show consensus within approximately 0.019 eV per atom, the tightest agreement seen across the broader pipeline for VP compositions. This degree of multi-model concordance is unusual and indicates that the Pnma structure is robustly described by the current generation of universal machine-learning potentials — a meaningful signal about the quality of the computational prediction. Two independent DFT source calculations underpin these results. A critical piece of intellectual honesty: alternative VP polymorphs that sit off the convex hull — meaning they are metastable or thermodynamically disfavored — were also evaluated and returned imaginary phonon modes in at least one potential. Those off-hull polymorphs are explicitly excluded from stability claims and from the patent assertions. This is not a weakness in the case for Pnma VP; it is the correct scientific behavior, and it sharpens the claim by ensuring every asserted composition is backed by genuine computational evidence. The remaining open validation gate is laboratory fabrication of a (110)-enriched coupon — synthesis conditions that favor Pnma over competing phases and that produce measurable (110) surface enrichment via X-ray diffraction or electron backscatter — followed by rotating-disk electrode polarization curves to confirm that the experimental exchange-current density and Tafel slope are consistent with the predicted dG_H.

The hydrogen-evolution reaction cathode market sits at the intersection of the broader green-hydrogen equipment market and the electrochemical-materials supply chain. Proton-exchange membrane electrolyzers currently dominate the high-efficiency segment, and PEM cathodes are almost universally platinized. As installed electrolyzer capacity grows toward multi-gigawatt annual installations — driven by policy incentives in the US Inflation Reduction Act, the EU Hydrogen Strategy, and analogous programs in Japan, South Korea, and China — the cumulative demand for platinum cathode material represents a supply and cost constraint that OEMs and project developers have already identified as a strategic vulnerability. Addressable cathode catalyst spend in the electrolyzer market is estimated at roughly $1 to $2 billion at moderate market penetration, with the range reflecting uncertainty about the pace of capacity build-out and the degree to which alkaline electrolyzers (which already use nickel-based cathodes) versus PEM systems capture share. Customers in this space are electrolyzer OEMs — companies like Nel, ITM Power, Plug Power, and their Asian counterparts — who source or develop electrode materials internally and who have both the technical capability to evaluate a new cathode composition and the commercial incentive to qualify one that reduces bill-of-materials cost. Licensing economics in this segment typically track a royalty on cathode active-material value or on electrolyzer stack revenue, rather than a per-kilogram material price. A composition-plus-device-use claim allows royalty capture both from materials manufacturers who supply the cathode powder and from stack integrators who incorporate a defined electrode architecture. The support-free feature of this formulation also reduces the number of supply-chain nodes that would be covered by the claim, making licensing negotiations somewhat simpler than for composite electrode systems.

earth-abundant V, PGM-free

Pnma ground state + (110) + support-free

The primary strategic buyer or licensee is an electrolyzer OEM with an active cathode development program — particularly one that has publicly committed to reducing PGM loading in its next-generation stacks and is therefore actively scanning for validated alternatives. Companies in this category include both the established Western PEM stack manufacturers and the rapidly scaling Asian electrolyzer producers who are building out capacity for domestic hydrogen programs. A secondary buyer category is a transition-metal phosphide materials company or a specialty chemicals firm with existing vanadium processing capability, for whom a composition-plus-device-use license would enable entry into the cathode materials supply chain with a differentiated, patented product. In either case, the value proposition is the same: a computationally validated, crystallographically specific, PGM-free cathode composition with a clean freedom-to-operate position and a defined experimental validation roadmap. Defensive value also exists for any company currently operating in the transition-metal phosphide catalyst space that wants to ensure freedom to develop VP-based electrodes without risk of infringement if a competitor acquires this asset. The support-free and polymorph-specific framing means this is not a broad blocking patent on all VP use, but it does cover the most computationally attractive form of VP for HER, which makes it strategically relevant for any player who might independently arrive at Pnma VP through their own R&D. The asset's position within the broader catalysts and energy-conversion materials portfolio also creates opportunity for package licensing alongside related phosphide compositions, which could be attractive to a buyer seeking a more comprehensive position in the PGM-free HER cathode space.

The principal technical risk is the gap between computational prediction and experimental measurement. The dG_H values from MACE and CHGNet are in agreement but span a 60 meV range, and neither has been benchmarked against explicit electrochemical measurements on a synthesized Pnma VP (110) surface. If the actual experimental dG_H falls outside the near-optimal window — due to surface reconstruction, adsorbate-induced relaxation, or coverage effects not captured in the single-hydrogen slab model — the activity advantage over cheaper non-PGM alternatives could be smaller than predicted. A related risk is acid stability: vanadium phosphides can be susceptible to dissolution under the acidic conditions of PEM operation, and stability data under potential cycling is entirely absent at this stage. These risks are addressable through the coupon fabrication and rotating-disk electrode measurement that constitutes the open validation gate, followed by accelerated durability testing under PEM-relevant conditions. The commercial risk is timing: the competitive landscape for PGM-free HER cathodes is active, with several academic groups and at least a few companies pursuing MoP, CoP, and nickel-based alternatives at higher technology-readiness levels. If a competitor reaches commercial qualification of a non-VP phosphide cathode before this asset completes experimental validation, the addressable royalty base narrows to customers who specifically want VP or who encounter IP barriers around the competing materials. The polymorph-specific, support-free claim architecture provides some insulation from this scenario — it is unlikely a competitor's MoP or CoP cathode would read on a VP Pnma (110) claim — but it also means the asset's value is contingent on VP specifically being adopted, rather than on phosphide cathodes broadly being adopted. Accelerating the experimental program and coupling it to an OEM co-development agreement would substantially reduce both technical and commercial risk simultaneously.