Scandium manganese nickel double-perovskite magnetodielectric oxide

Sc2MnNiO6 is an ordered A2BB'O6 double perovskite disclosed within the dielectric, ferroelectric and wide-bandgap oxides portfolio as a development-stage dependent composition. The strategic rationale is straightforward: the magnetodielectric oxide segment sits at the intersection of oxide-dielectric thin-film engineering and spintronic-adjacent device architectures, and the scarcity of computationally characterized, patent-clean ordered double perovskites with both a useful bandgap and meaningful magnetodielectric coupling makes even a near-hull candidate with clean freedom-to-operate worth staking a dependent claim against. The filing is honest about its stage — it is explicitly not advanced as the lead composition in this family, but rather as a parallel backup disclosure that preserves optionality for licensees or acquirers who want to extend coverage into Mn-Ni B-site ordered systems. The timing logic for this class of materials is driven by forced substitution dynamics in oxide dielectric layers: as device architectures demand tighter control of leakage, magnetic response, and thermal stability simultaneously, single-transition-metal perovskites are increasingly inadequate, and ordered double perovskites — where two distinct B-site cations arrange in a rock-salt superlattice — offer a structural handle on coupling dielectric and magnetic order parameters that single-cation systems cannot match. A disclosed, computationally validated, FTO-clean candidate in this space, even at the dependent-filing tier, carries real option value for parties building a freedom-to-operate position in next-generation oxide-stack components. The dossier below is candid about the open proof gates: ordered-phase synthesis has not been demonstrated in the lab, DFPT dielectric-tensor confirmation is pending, and the HSE-corrected bandgap has not been computed. These are real gaps, and any acquirer or licensee should price them in. What the computational evidence does establish — three out of four independent machine-learning interatomic potentials agreeing on dynamic stability, and a DFT energy-above-hull placing the compound close enough to the convex hull to be practically accessible — represents a meaningful first filter that most candidate oxides never pass.

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.

Sc2MnNiO6 adopts the ordered A2BB'O6 double perovskite archetype, in which scandium occupies the A-site and manganese and nickel alternate on the crystallographically distinct B and B' sites in a rock-salt ordering pattern. This B-site ordering is the structural feature that distinguishes a true double perovskite from a disordered or phase-separated mixture of binary perovskites, and it is the source of virtually all the technologically interesting properties: magnetodielectric coupling, anomalous permittivity enhancement near magnetic ordering temperatures, and the ability to tune optical and electronic properties by controlling B-site charge states independently. The ordered double perovskite space group has not been resolved from the current computational data, which is itself an honest flag — resolving the ground-state symmetry (likely monoclinic P21/n or trigonal R-3 depending on distortion pattern) is a prerequisite for quantitative DFPT calculations and should be addressed in any follow-on synthesis campaign. The bandgap computed at the standard DFT level is 1.82 eV. This value places Sc2MnNiO6 in a useful window for oxide-dielectric applications: wide enough to suppress direct tunneling leakage at typical oxide layer thicknesses, yet not so wide as to make optical characterization difficult. The caveat is that standard DFT (GGA or GGA+U) systematically underestimates bandgaps in transition-metal oxides containing Mn and Ni, which carry strong correlation effects. The HSE06 hybrid-functional correction is an open proof gate; the true quasiparticle gap is expected to be higher, perhaps in the 2.2–2.8 eV range, which would only improve the dielectric application case but must be measured or computed before confident device-level projections can be made. The thermodynamic stability assessment uses the DFT energy-above-hull (EAH) as the primary metric. At approximately 34.6 meV per atom, Sc2MnNiO6 sits near — but not on — the convex hull of the Sc-Mn-Ni-O quaternary system. In practical terms, this means the compound is a metastable phase: it does not spontaneously decompose at 0 K under the DFT ground state, but there exist competing phase assemblages (likely combinations of ScNiO3, MnO, and Sc2O3 or related binaries/ternaries) that are marginally more stable. An EAH of roughly 35 meV/atom is within the range that many known and synthesizable oxide phases occupy — it is not a barrier to synthesis, but it does mean that synthesis conditions (temperature, oxygen partial pressure, substrate epitaxial strain) will require careful optimization to suppress phase separation. Epitaxial stabilization on an appropriate substrate is a plausible route, and thin-film deposition under non-equilibrium conditions routinely accesses metastable phases with EAH values considerably higher than this. Dynamic stability — the question of whether the structure sits at a true energy minimum with respect to atomic displacements, rather than a saddle point that would lead to spontaneous distortion or decomposition — was assessed using four independent machine-learning interatomic potentials from distinct training lineages: MACE, CHGNet, MatterSim, and ORB. Three of the four potentials return a stable verdict (no imaginary phonon modes across the Brillouin zone), with one dissenting. This three-of-four majority-stable result is a meaningful positive signal — it means the instability flagged by one potential is likely an artifact of that potential's training distribution rather than a true physical instability — but it is honestly below the consensus threshold (all four potentials agreeing) that the portfolio reserves for its lead compositions. DFPT phonon calculations from first principles remain an open gate and would either confirm the majority verdict or identify the specific soft mode responsible for the minority dissent.

The addressable market for magnetodielectric oxide materials is best framed as a sub-segment of the broader oxide thin-film dielectric market, which spans applications in DRAM capacitor stacks, ferroelectric memories, microwave resonators, and spintronic tunnel junctions. The high-permittivity oxide dielectric market for semiconductor memory alone is measured in the low billions of dollars annually, but magnetodielectric materials specifically — those in which dielectric permittivity is tuneable by an applied magnetic field — occupy a smaller, more specialized segment estimated in the $0.5–1 billion range. This addressable figure is best understood as an estimate reflecting early-stage commercial traction: magnetodielectric oxide layers are not yet in high-volume silicon fab processes, and the dominant current customers are research institutions, defense contractors pursuing phase-shifter and tunable-filter components, and early commercial actors in oxide-based neuromorphic and spintronic architectures. The economics of licensing or acquiring IP in this space reflect that developmental stage. A patent position covering a novel, computationally characterized, FTO-clean magnetodielectric composition is primarily valuable as a component of a broader oxide-dielectric stack strategy: it enables a licensee to publish, develop, and commercialize without clearing third-party IP, and it provides a priority date for a composition the market is moving toward. Royalty logic for materials IP in specialty oxide stacks typically attaches to device-level licensing — a per-wafer or per-unit fee negotiated at the point of integration into a manufacturable process node — rather than to composition sales directly. Given the development stage, early licensing arrangements would most likely take the form of sponsored research agreements or option-to-license structures tied to synthesis milestones, with royalty-bearing licenses triggered upon demonstration of the ordered phase in thin-film form.

magnetodielectric oxide option with clean FTO; weaker proof tier (near-hull, 3-of-4)

near-hull candor-flagged dependent; not claimed as a 4-of-4 lead

The most natural acquirers or licensees for this asset are parties already building IP portfolios in oxide-dielectric or magnetodielectric stacks who want to extend coverage into the ordered double perovskite compositional space at low incremental cost. Advanced semiconductor materials suppliers (particularly those active in high-k dielectric precursor development), ferroelectric memory device companies exploring oxide alternatives to HfO2-based stacks, and defense-adjacent component manufacturers pursuing magnetically tunable microwave materials are all plausible strategic fits. Because the asset is a dependent filing at a development stage, the most likely near-term transaction structure is an option-to-license bundled with a sponsored research agreement that funds the DFPT and synthesis milestones — converting the option to a full license upon synthesis confirmation of the ordered phase. Academic spinouts and national-lab-affiliated ventures working on magnetodielectric thin films represent a secondary acquirer class: they may have the synthesis infrastructure to close the open proof gates and the publication incentive to generate the experimental validation that would strengthen the claim. For a larger incumbent oxide materials company, the asset's value is primarily defensive — acquiring it prevents a competitor from staking an adjacent position in Sc-based double perovskites while the parent family's lead compositions are commercialized. In either case, the clean FTO status is the headline attribute, and any diligence process should center on the synthesis feasibility question, for which consulting engagement with groups experienced in Sc-based perovskite thin-film deposition (pulsed laser deposition or molecular beam epitaxy) would be the appropriate first step.

The principal risk is synthesis: ordered B-site arrangement in A2BB'O6 double perovskites is kinetically demanding, and the Sc-Mn-Ni system has not been reported in the experimental literature in ordered form. If synthesis attempts yield only phase-separated ScNiO3 and MnO-containing assemblages rather than an ordered double perovskite, the composition-plus-device-use claim loses its material basis for enablement, and the asset's value is substantially diminished. This is an honest assessment — the synthesis proof gate is not a formality. The near-hull EAH of 34.6 meV/atom and the three-of-four stability majority are encouraging indicators, but neither guarantees that an experimentalist can produce phase-pure ordered material under accessible conditions. The roadmap to de-risk is sequential: first, resolve the space group and full atomic positions computationally (DFT structural relaxation with multiple starting configurations to rule out lower-energy distorted variants); second, run DFPT to close the phonon stability gate and compute the static dielectric tensor, which would provide the first direct computational evidence of the magnetodielectric application target; third, commission a targeted synthesis experiment — pulsed laser deposition on a SrTiO3 or LaAlO3 substrate with appropriate lattice parameter match is the most likely route to epitaxial stabilization of the metastable ordered phase. Completing these three steps would transform this asset from a candor-flagged dependent with open gates into a computationally and experimentally validated composition, at which point renegotiating the licensing terms or pursuing an independent claim filing would be appropriate. The FTO position is clean today and is not a risk factor, but patent prosecution timelines mean that the synthesis validation should be completed before any continuation filing deadlines expire.