Neodymium-based ternary silicide (NdNiSi) for cryogenic and superconducting device applications

NdNiSi is the third arm of a three-genus rare-earth silicide superconductor candidate portfolio covering the 1:1:1 ternary RE-transition-metal-silicon structural family. The strategic rationale for this arm is straightforward: neodymium and its near-neighbors occupy a distinct corner of rare-earth-silicide phase space, and the patent landscape around those specific compositions — while having academic prior art on the compounds themselves — shows meaningful whitespace on the application side. Lattice Graph has assembled a method-of-screening and device-use claim set that does not require ownership of the underlying composition; instead, it claims the computational workflow that identifies phonon-stable members of this family as candidates for superconducting or cryogenic device integration, and the use of qualifying members in device contexts. The timing logic is straightforward. The broader superconducting electronics sector is under sustained pressure to move away from niobium-centric materials as quantum computing and cryogenic sensing scale beyond the laboratory. The rare-earth-silicide family has attracted renewed attention precisely because it sits at the intersection of structural tunability, comparatively accessible synthesis, and compatibility with silicon-adjacent processing. The Nd-anchored arm of this genus is not being positioned as a flagship composition-of-matter play; it is a carefully bounded defensive and complementary dossier that rounds out the full claimed genus, ensuring that any licensee taking the full portfolio has coverage across all three rare-earth arms without gaps. Candidly, this is a supporting rather than lead asset. Its value is additive within the broader portfolio: a buyer licensing the complete rare-earth silicide superconductor candidate portfolio receives coherent genus-level coverage that includes the Nd and related rare-earth arm, making the overall package harder to design around. Considered alone, the asset's commercial case rests on method-of-use and device-use claims in a market segment where such claims can still command licensing fees from device manufacturers who need freedom to operate on their specific material selection and screening workflows.

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.

NdNiSi adopts a layered 1:1:1 ternary structure, with the relevant polymorphs falling into the ThCr2Si2-derived family, the PbFCl-type arrangement, and the CeFeSi-type layered variant — all of which are structurally related and share the key feature of rare-earth layers interleaved with transition-metal-silicide slabs. This structural motif is well established in the literature as a host for a range of electronic ground states, including superconductivity, magnetic ordering, and heavy-fermion behavior, depending on the rare-earth and transition-metal choice. Neodymium's partially filled 4f shell introduces additional complexity relative to lighter rare earths, making the electronic structure strongly dependent on the precise crystallographic site symmetry and the degree of hybridization between Nd-f states and the Ni-d / Si-p manifold. The computational validation for this arm was conducted using Lattice Graph's four-engine machine-learning interatomic potential (MLIP) consensus protocol, running MACE, CHGNet, MatterSim, and ORB as independent evaluators. The phonon stability assessment returned a majority-stable verdict: a majority of the four independent potentials find the structure dynamically stable with no imaginary phonon modes at the equilibrium geometry. This is a meaningful but honestly bounded result — it is not a full-consensus four-out-of-four agreement, which is the higher bar Lattice Graph requires for its lead candidates. Majority stability across four structurally diverse MLIPs trained on different datasets is still strong evidence that the compound is not dynamically unstable, and distinguishes NdNiSi from the large population of hypothetical candidates that fail even a single-MLIP screen. However, the partial consensus means the result carries more uncertainty than the lead genus members and warrants confirmation. The freedom-to-operate prescreen was conducted against a database of over 300,000 materials patents, and the result is that no third-party composition-of-matter claim has been identified that would block device-use or method-of-screening claims on NdNiSi or related Nd-arm members. This is consistent with the general picture for rare-earth silicides: the academic literature knows these compounds well, but patent coverage of the specific device integration and computational screening workflows remains sparse. The Lattice Graph claim set has been designed to operate entirely within the whitespace identified by that prescreen. No direct first-party DFT data (lattice constants, electronic density of states, computed superconducting Tc, or DFPT dielectric tensors) has yet been generated for this specific composition; those calculations are explicitly identified as the next open validation gate. The simulations completed to date are the four-engine phonon consensus described above and the patent-whitespace prescreen. These are sufficient to support the current claim strategy — which does not assert that NdNiSi is a confirmed superconductor, but rather that it is a computationally pre-qualified candidate identified by a novel screening workflow — but they fall short of what would be needed to license this asset on the strength of its predicted properties alone. Targeted DFPT calculations for the dielectric tensor, full DFT band-structure and density-of-states resolution of the f-electron manifold, and ideally migration-barrier or interface molecular dynamics simulations for device-relevant configurations are the natural next steps. Measured characterization — even powder XRD confirmation of phase purity and a resistivity-versus-temperature measurement — would substantially strengthen any device-use licensing conversation.

The superconducting electronics and cryogenic sensing market is the primary target. Cryogenic device manufacturers, quantum computing hardware companies, and advanced sensing platform developers all face material selection decisions at cryogenic temperatures where the choice of superconducting or low-temperature-compatible materials directly affects device performance, fabrication compatibility, and ultimately system cost. The addressable portion of this market relevant to a method-of-use and device-integration claim on a class of rare-earth silicide candidates is estimated in the range of $1–5 billion, though this estimate should be treated as a rough sizing rather than a precise forecast. The upper bound reflects a future in which rare-earth silicides achieve broad adoption across quantum processor interconnects and cryogenic sensor arrays; the lower bound reflects a scenario where adoption is limited to niche superconducting detector applications. The licensing logic for this asset type is primarily cross-license and design-freedom rather than royalty-on-product. A cryogenic device manufacturer that is independently investigating RE-silicide materials for a sensor or qubit-coupler application may want to ensure that their internal computational screening workflow — which may functionally overlap with what is claimed here — does not create IP exposure. A method-of-use license in that context is a relatively low-cost insurance purchase. The same logic applies to foundries or materials suppliers who synthesize these compounds for device customers. The more valuable licensing scenario, however, is a strategic acquisition or exclusive license by a party building a proprietary materials database for superconducting electronics, where the full rare-earth silicide portfolio (all three arms) becomes part of a defensible IP stack rather than a single-composition wager.

Nd/related arm with device-use whitespace

method-of-screening + device-use only; no composition-of-matter to literature-known Nd/related members

The most natural acquirers or licensees for this asset are parties already licensing or considering the full rare-earth silicide superconductor candidate portfolio, for whom this Nd-arm dossier provides completeness of coverage across the genus. Cryogenic hardware companies — particularly those developing superconducting detector arrays, quantum processor interconnects, or low-temperature SQUID-based sensing platforms — represent the primary end-user customer class, and any of those organizations building an internal materials IP stack would have strategic reasons to hold method-of-use coverage on a pre-screened class of candidates. Materials informatics companies seeking to license a validated screening workflow rather than individual compositions are a secondary target; the method claims here are compatible with a workflow-licensing model. Defensive acquirers — companies in the superconducting electronics supply chain who want to ensure freedom to operate their own computational materials selection programs — are also a plausible buyer class, even if they have no present intention to commercialize NdNiSi specifically. The value in that scenario is insurance: holding the method claims prevents a third party from asserting a blocking position against a broadly similar internal R&D workflow. Given the early stage of this specific asset and the bounded nature of the claims, transaction structures most likely to be viable are inclusion in a full-portfolio license of the rare-earth silicide superconductor candidate portfolio, a non-exclusive method-of-use license at a modest annual fee, or a joint-development agreement in which the buyer funds the open validation experiments (DFPT, measured characterization) in exchange for preferred licensing terms on the resulting strengthened asset.

The principal risk for this asset is the gap between majority MLIP consensus and the full validation that would be needed to support an aggressive licensing position. The absence of first-party DFT data and measured characterization means that the asset is currently positioned on computational pre-qualification alone, and a sophisticated counterparty in a licensing negotiation will identify that gap immediately. If the DFPT calculations, when completed, reveal imaginary phonon modes not captured by the majority of the MLIP ensemble, the asset's value as a device-use claim would be substantially diminished — though the method claim, which covers the screening process itself, would remain intact. A secondary risk is the intentional absence of a composition-of-matter claim: because the compositions are literature-known, the asset cannot be used to prevent a party from making, using, or selling NdNiSi, only from using it in specific device contexts identified through the covered screening methodology, which limits the leverage available in any enforcement scenario. The roadmap to de-risking is clear and relatively low-cost. Commissioning DFPT phonon calculations from a DFT group with experience in rare-earth intermetallics would take weeks and would either confirm or refute the MLIP majority-stable result. Commissioning synthesis and basic transport characterization — arc-melting or solid-state synthesis followed by powder XRD and resistivity measurement — would take months and would substantially strengthen the commercial narrative. If both gates return positive results, this asset transitions from a supporting defensive dossier to a validated member of the portfolio with measured data behind it, which is a meaningful step up in licensing value. The investment required to reach that gate is modest relative to the potential upside in a full-portfolio transaction.