Sodium palladium sulfide backup electrocatalyst for electrochemical CO2 reduction

Na2PdS2 — sodium palladium disulfide — is a polar acentric alkali-palladium sulfide that Lattice Graph has characterized computationally and advanced as a backup composition within the alkali-palladium chalcogenide CO2-reduction electrocatalyst family, where the lead compound is the potassium analog K2PdS2. Its role is honest and deliberate: it broadens the defensive perimeter of the family patent by establishing that the sodium-bearing sulfide is independently dynamically stable and shares the dielectric character relevant to electrocatalytic performance, while the lead composition moves through the primary experimental validation queue. That is a legitimate and strategically important function — a well-constructed patent family requires more than a single exemplary composition, and a backup that is computationally validated to the same standard as the lead compound is a credible, enforceable claim rather than a speculative placeholder. The timing matters for the CO2 reduction reaction (CO2RR) space broadly. Regulatory and industrial pressure on carbon capture and utilization is accelerating, and electrochemical CO2 reduction to useful feedstocks — carbon monoxide, formate, ethylene, methanol — is increasingly viewed as a near-term deployment path rather than a research curiosity. The dominant cathode materials today (copper, silver, gold) are well-understood but face cost, selectivity, and stability limits. Transition-metal chalcogenides, particularly those with polar crystal symmetry and tunable electronic structure, represent a structurally distinct design space that has received far less systematic patent coverage, creating genuine whitespace. Lattice Graph's family-level approach — validating multiple alkali-palladium sulfide compositions computationally before any one is committed to expensive synthesis — is precisely the kind of staged de-risking that turns a discovery pipeline into a durable IP portfolio. The honest caveat is that Na2PdS2 is not the lead candidate and is not ahead of K2PdS2 in the experimental queue. Its bandgap has not yet been resolved to the precision needed for confident CO2RR activity prediction — the PBE-level density functional estimate of approximately 0.94 eV is a starting point, not a confirmed value, and the HSE06 hybrid-functional refinement that will provide a more reliable number is in progress under a corrected computational pipeline. A CO2RR coupon measurement — the definitive electrochemical validation gate — has not yet been run. Buyers should price this asset as what it is: a computationally well-characterized, freedom-to-operate-clean backup composition with confirmed dynamic stability and dielectric character, positioned to capture defensive breadth and to serve as an insurance policy if the lead compound encounters synthesis or performance obstacles.

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.

Na2PdS2 belongs to the alkali-palladium chalcogenide class, crystallizing in a polar acentric space group — a structural symmetry that breaks inversion, which matters for electrocatalysis because polar surfaces can present net dipole moments that influence adsorbate orientation, local electric field distribution, and reaction-intermediate stabilization at the electrode-electrolyte interface. The compound is sourced from the Materials Project database (mp-10223), providing a well-curated starting geometry for all downstream calculations. The chalcogenide framework consists of palladium coordinated by sulfur, with sodium occupying interstitial sites; the result is a layered-type architecture that is structurally distinct from the selenide and telluride analogs, which are explicitly distinguished in the claim scope. Dynamic stability — the question of whether the crystal will hold together under realistic phonon excitations rather than spontaneously distort — is the first and most critical computational gate. Two independent machine-learning interatomic potentials, MACE and CHGNet, were applied to the Na2PdS2 structure. Both returned positive phonon frequencies across the entire Brillouin zone, with the minimum phonon frequency computed at approximately +0.0712 THz under the MACE potential and a corresponding positive result from CHGNet. The absence of imaginary (negative) phonon modes in either potential is strong evidence that the structure sits in a genuine energy minimum rather than a saddle point, meaning the material is expected to be synthesizable and thermally robust under operating conditions. This cross-potential consensus is the cornerstone of Lattice Graph's validation protocol — agreement between structurally distinct ML potentials trained on different datasets is materially more reliable than a single-potential result, because it rules out artifacts specific to any one model's training distribution. Density functional perturbation theory (DFPT) calculations were carried out on the phonon-stable structure, yielding a high-frequency (optical) dielectric constant epsilon_inf of approximately 7.34. This quantity captures the purely electronic polarizability of the lattice — distinct from the ionic contribution — and is diagnostically relevant for two reasons. First, it confirms that the material is a genuine dielectric rather than a near-metal, consistent with the computed PBE electronic bandgap of approximately 0.94 eV. Second, a moderately high epsilon_inf in a polar material correlates with strong electron-phonon coupling and significant polaronic character, which can influence charge-carrier mobility and the dielectric screening of reaction intermediates at the electrode surface. A soft bond-valence proxy calculation (using the bond valence sum method as an ionic migration surrogate) was also completed, providing a preliminary assessment of sodium-ion mobility within the structure — relevant context if the material were ever considered in a mixed ionic-electronic conductor role, though the primary intended application here is as an electronic cathode for CO2RR. The electronic structure picture is incomplete in one important respect: the PBE-level bandgap of ~0.94 eV is known to systematically underestimate true gaps in transition-metal sulfides due to self-interaction error in the exchange-correlation functional. An HSE06 hybrid-functional calculation, which mixes a fraction of exact Hartree-Fock exchange to partially correct this error, is in progress under a fixed-occupations protocol designed to handle the near-metallic density of states that can cause convergence difficulties in PBE-derived structures. Until the HSE06 result is available, the true gap of Na2PdS2 could be anywhere from slightly above 0.94 eV (if PBE is fortuitously close) to substantially larger. The gap value matters because CO2RR cathode operation requires the material to be sufficiently semiconducting to support electrochemical reduction under applied bias without short-circuiting through excessive dark conductivity, while also being narrow enough to absorb light or couple efficiently to an external circuit. Two DFT source calculations have been completed to date, establishing the structural and dielectric baseline; the HSE06 refinement and the first electrochemical coupon measurement remain as open validation gates.

The electrochemical CO2 reduction market sits at the intersection of carbon capture, green chemistry, and industrial feedstock production. Demand is driven by corporate net-zero commitments, emerging carbon markets, and policy mandates in the EU, US, and increasingly Asia-Pacific that require industrial emitters to either sequester or valorize their CO2 output. The near-term commercial sweet spot is CO2 reduction to CO (for Fischer-Tropsch synthesis or syngas applications) and to formate/formic acid (a hydrogen carrier and chemical feedstock), with ethylene and ethanol as higher-value targets that remain more technically challenging. Addressable market estimates for CO2 electrolyzer components — including cathode materials, membrane electrode assemblies, and stack hardware — sit in the range of $0.5 to $1 billion over the next decade for materials suppliers specifically, though the broader CO2 utilization market that creates demand for those components is substantially larger. The relevant customer segment for a novel cathode material is the CO2 utilization developer: companies building and selling CO2 electrolyzer systems, typically at the pilot-to-commercial scale, who need cathode materials that outperform copper and silver on one or more axes — selectivity toward a target product, stability over thousands of hours of operation, or lower cost of goods. A second customer segment is the major chemical company or industrial gas producer that is building internal CO2 utilization capacity and may prefer to license a proprietary cathode material rather than rely on commodity metals. Licensing economics for electrocatalyst materials typically follow a royalty-on-sales model, with rates in the 2–5% range on cathode material cost, or a milestone-plus-royalty structure tied to electrolyzer deployment scale. Na2PdS2 specifically carries a "Na-bearing low-PGM CO2RR backup" positioning, which captures two cost arguments relative to incumbents. First, the sodium content reduces the overall platinum-group metal (PGM) loading compared to a pure-palladium catalyst, since sodium is earth-abundant and inexpensive. Second, palladium — while itself a PGM — is present in a sulfide coordination environment that may dilute effective Pd site density relative to metallic Pd catalysts, potentially reducing material cost per unit of active surface area. These are structural arguments that require experimental confirmation, particularly through turnover frequency and selectivity measurements, before they can be translated into a commercial cost narrative. The asset's value to a buyer today is primarily as portfolio breadth within the alkali-palladium chalcogenide family rather than as a standalone commercial proposition.

Na-bearing low-PGM CO2RR backup

sulfide + CO2RR use

The most natural acquirers for this asset are companies already active in the CO2 electrolyzer or carbon utilization space who are building or consolidating cathode material IP. This includes electrolyzer stack developers who want proprietary cathode materials that are not available to competitors, specialty chemicals companies building CO2-to-formate or CO2-to-CO capacity, and advanced materials suppliers seeking to enter the CO2RR cathode market with a differentiated, IP-protected composition. The asset is also of interest to any party acquiring the broader alkali-palladium chalcogenide family, where Na2PdS2 serves as a backup member that strengthens the family's enforceability and breadth; in that context, the sodium analog would typically be bundled rather than transacted separately. A secondary buyer profile is the academic or government laboratory that is building a CO2RR materials portfolio and wants to license composition rights before committing to synthesis and testing — particularly given the clean FTO status and the relatively modest experimental investment needed to close the two remaining validation gates. Industrial R&D groups at major chemical companies (BASF, Dow, Mitsubishi Chemical, Sumitomo) with internal carbon utilization programs represent a third segment, as these organizations often prefer to license computationally pre-screened compositions at an early stage rather than run their own discovery campaigns. The asset's backup status means the licensing conversation will naturally be bundled with the lead K2PdS2 compound, and a buyer should expect that the most favorable terms are available for acquiring the full alkali-palladium chalcogenide family together.

The principal risk is bandgap uncertainty. The PBE-level gap of ~0.94 eV is suggestive of a narrow-gap semiconductor, but if the HSE06 refinement returns a substantially larger value — or, conversely, reveals near-metallic behavior — the material's suitability as a CO2RR cathode under standard operating conditions would require reassessment. A gap that is too large would impair charge injection under modest applied bias; a gap that is too small would make the material behave more like a conductor than a semiconductor, potentially reducing selectivity. This uncertainty is resolvable in the near term at low cost, and it is the next computational step in the pipeline. The second risk is the absence of experimental electrochemical data: dynamic stability and dielectric character are necessary but not sufficient conditions for CO2RR activity, and the material could in principle be stable and dielectrically well-characterized while still being inactive or poorly selective for CO2 reduction. The coupon measurement is the gate that resolves this, and its absence means the asset carries meaningful performance uncertainty that any buyer must price. De-risking follows a clear sequence: complete the HSE06 gap calculation (weeks, no synthesis required), synthesize a phase-pure Na2PdS2 sample (the mp-10223 structure is a known compound, which reduces synthesis uncertainty), and run an electrochemical CO2RR coupon in a standard H-cell or flow-cell configuration to establish onset potential, Faradaic efficiency, and selectivity. A buyer with existing CO2RR infrastructure could compress this timeline significantly. The palladium content is a cost risk at commercial scale, but it is common to the entire alkali-palladium chalcogenide family and would be addressed at the family level rather than for this asset specifically.