Trirutile niobate, polar borate, and brownmillerite ceramics for mm-wave filters and solid oxide fuel cells

The millimeter-wave filter market is entering a period of forced substitution. As 5G sub-6 GHz deployments mature and 6G planning accelerates toward the 28, 39, and 60 GHz bands, the incumbent dielectric resonator platforms — predominantly composite scheelite ceramics — face fundamental loss-tangent ceilings and thermal-expansion management problems that degrade filter selectivity at mm-wave frequencies. The trirutile-structure MgTa2O6, polar borate SrB4O7, and ring-borate YAl3B4O12 represent a structurally diverse yet compositionally coherent set of monolithic ceramic candidates that are computationally validated for dynamic stability and whose intrinsic crystal chemistries are favorable for the combination of high Q-factor, low temperature coefficient of resonant frequency (tau_f near zero), and controlled permittivity that mm-wave filter designers require. The portfolio's value is not confined to RF hardware: the same filing also secures the brownmillerite Ba2In2O5 as a proton- and oxide-ion conducting solid oxide fuel cell electrolyte, and extends into polar-borate piezo/nonlinear-optical device space, making this a genuinely cross-vertical ceramic platform rather than a single-application bet. The timing logic is driven by two converging pressures. On the RF side, base-station and fixed-wireless-access hardware developers face a component shelf that has not materially innovated at the crystal-chemistry level in over a decade; a computationally pre-screened, patent-clean monolithic material with validated dynamic stability substantially de-risks materials selection for next-generation module designs. On the energy side, the global push for intermediate-temperature solid oxide fuel cells (600–800°C operating range) has created sustained demand for proton-conducting electrolytes with high ionic conductivity and processing compatibility; Ba2In2O5 brownmillerite is a known fast-ion candidate and the inclusion of an independent stability assessment adds credibility that purely literature-sourced claims cannot match. Together, these two application domains — telecommunications hardware and electrochemical energy conversion — sit in markets large enough to justify licensing across separate vertical channels from a single asset. The portfolio to which this asset belongs, the "dielectric, ferroelectric and wide-bandgap oxides" portfolio, covers a range of structurally related oxide classes. Within that portfolio, this family functions as the lead asset for monolithic microwave-dielectric and polar-borate applications, and the computational work completed to date represents meaningful prior investment that a licensee or acquirer can build on directly rather than replicating from scratch.

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.

The three lead compositions span distinct crystal structure types that converge on complementary dielectric performance profiles. MgTa2O6 adopts a trirutile superstructure (space group P4₂/mnm) in which Mg²⁺ and Ta⁵⁺ order alternately on the octahedral sites of the parent rutile framework. This cation ordering is precisely what produces a near-zero tau_f and a Q×f product competitive with established microwave resonator ceramics: the ordered arrangement eliminates the anharmonic phonon scattering pathways that in disordered solid solutions drive up dielectric loss. SrB4O7 crystallizes in the polar Pmn2₁ space group — a structure that lacks an inversion center — making it simultaneously a microwave-frequency low-loss dielectric and a piezoelectric/nonlinear-optical material; its relatively low permittivity (εr typically in the 8–12 range for borates of this class) is favorable for applications where impedance matching and spurious mode suppression matter more than miniaturization. YAl3B4O12, adopting the huntite-related R32 structure, is a ring borate where [B3O9] and [BO4] groups share connectivity through the AlO6 framework; the rigid, covalently cross-linked boron-oxygen network gives the material an exceptionally stiff lattice and correspondingly low intrinsic dielectric loss at millimeter-wave frequencies. The computational validation methodology applied here goes beyond single-point energy minimization. Two independent machine-learning interatomic potentials — drawn from the ensemble that includes MACE, CHGNet, MatterSim, and ORB — were run on the MgTa2O6, SrB4O7, YAl3B4O12, and BaHfN2 structures, and both potentials independently confirm dynamic stability: the computed phonon dispersion curves show no imaginary modes across the full Brillouin zone, confirming these structures sit at genuine energy minima rather than saddle points. This two-potential consensus, achieved as of May 2026, is the threshold the validation pipeline requires before advancing a material; the agreement is meaningful because MACE and its siblings have different training corpora and different functional forms, so disagreement between them flags structures that may be artifactually stable in one potential's energy landscape but not in the physical one. For Ba2In2O5 and Sm2Zr2O7, three of four potentials find dynamic stability (one dissents), which is a majority-stable result — encouraging but one confidence level below the lead group, and honestly reflected as such. The scheelite-class members BaWO4 and SrWO4 fared poorly: BaWO4 is majority-unstable and SrWO4 shows instability across both checked potentials. This is disclosed candidly; the scheelite branch of the family is not pursued as a primary composition. Two open validation gates remain before the lead compositions can be positioned for licensing with full experimental backing. First, density-functional perturbation theory (DFPT) calculations of the piezoelectric and electro-optic tensors for SrB4O7 and YAl3B4O12 have not yet been completed; these would quantify whether the polar-borate branch carries commercially relevant piezo or NLO coefficients, or whether that application is speculative at this stage. Second, actual Q×f and tau_f resonator measurements — the standard figures of merit for microwave dielectric ceramics, typically obtained at 1–10 GHz by the Hakki-Coleman or cavity methods and extrapolated to mm-wave bands — have not been reported for the specific monolithic forms claimed here. Both gaps are addressable through targeted simulation campaigns (DFPT) and ceramic processing plus resonator metrology (Q×f measurement), representing defined near-term de-risking work rather than open-ended research. The full family extends to additional members including Sr2MgB2O6, BaHfN2 (an oxide-to-nitride bridge composition bridging borate and nitride dielectric chemistries), Ba2SnO4, and Sr2TiO4 Ruddlesden-Popper phases. The Ruddlesden-Popper members introduce layered perovskite chemistry that is structurally related to Ba2In2O5 but with different cation arrangements, broadening the claim scope while maintaining chemical coherence. BaHfN2 was included in the two-potential stable consensus run, suggesting it is a credible member of the validated lead group despite being a nitride rather than an oxide.

The addressable market for millimeter-wave dielectric ceramics is anchored by the hardware supply chain for 5G and emerging 6G base stations, fixed-wireless-access customer-premises equipment, satellite ground terminals, and automotive radar. Filter and duplexer components alone represent a substantial fraction of radio-frequency module bill-of-materials cost, and dielectric resonator filters are preferred over cavity filters at mm-wave frequencies because of their smaller footprint and lower insertion loss. Market estimates for the microwave dielectric ceramics segment broadly cluster in the $1–3 billion range on a global addressable basis, with the mm-wave sub-segment growing at a meaningfully faster rate than the legacy 4G infrastructure market that dominated the prior decade. These figures should be treated as estimates given the heterogeneous sourcing; the point is directional rather than authoritative. The buyer profile for the RF dielectric application is a ceramic components manufacturer or a vertically integrated module house seeking to license a patent-protected composition for internal production. Royalty structures in ceramics IP typically run on a per-unit or per-kilogram basis tied to verified performance specifications; the natural licensing trigger for this asset would be a licensee demonstrating that a monolithic MgTa2O6 or borate ceramic meets its filter insertion-loss and temperature-stability spec, at which point the composition and device-use claims become the relevant IP boundary. The proton-conducting electrolyte application in solid oxide fuel cells addresses a separate buyer pool — SOFC and proton-conducting fuel cell (PCFC) stack manufacturers pursuing intermediate-temperature designs — and represents a distinct licensing channel that could be structured independently, covering Ba2In2O5 brownmillerite formulations under the same family filing. A third, smaller commercial thread runs through the polar-borate members: SrB4O7 and YAl3B4O12 have documented application interest in ultraviolet nonlinear optics and in rad-hard packaging dielectrics, where the borate framework's resistance to radiation damage has been noted in the literature. Defense and aerospace customers in these segments tend to accept higher unit prices and longer development timelines, making even a niche licensing arrangement commercially meaningful. The cross-vertical span — RF telecommunications, electrochemical energy, and specialty photonics/defense — is the primary commercial argument for the breadth of this family's composition claims.

monolithic microwave-dielectric + polar-borate piezo/NLO + brownmillerite electrolyte across one phase family

monolithic single-crystal/ceramic/thick-film only; composite scheelite+Bi2O3/CeO2 genus excluded; CaWO4 not led-on (102-disclosed)

The most likely near-term licensees are ceramic component manufacturers supplying the RF telecommunications infrastructure market, particularly those with existing mm-wave filter product lines who are actively searching for next-generation dielectric materials to replace or complement their current scheelite composite portfolios. Companies in Japan, South Korea, Taiwan, and Germany have historically dominated the microwave ceramic components market and maintain active R&D programs in advanced dielectric compositions; any of these players would have the ceramic processing infrastructure to rapidly evaluate monolithic MgTa2O6 and borate resonator bodies and the commercial motivation to secure IP-protected materials ahead of 6G hardware development cycles. Module-level integrators building front-end RF modules for fixed-wireless and satellite applications are a second tier of potential licensees, particularly if they are vertically integrating ceramic processing. For the SOFC and PCFC electrolyte application, the relevant buyers are fuel cell stack developers and systems integrators operating in the intermediate-temperature solid oxide space, including established players in Europe, the United States, and Japan who are developing next-generation stacks for distributed power generation and hydrogen applications. Defense and aerospace customers seeking rad-hard package dielectrics or UV-transparent borate optics represent a smaller but potentially high-value niche. A strategic acquirer assembling a broad ceramic IP position — whether a materials conglomerate, a specialty chemicals company with a ceramics division, or a deep-tech investment vehicle — would find this asset's cross-vertical span attractive as a portfolio addition to the "dielectric, ferroelectric and wide-bandgap oxides" portfolio, providing coverage across RF, electrochemical, and photonic application domains from a single prosecution family.

The primary technical risk is the absence of experimental resonator data. Dynamic stability by machine-learning potential consensus is a meaningful positive signal, but it does not substitute for measured Q×f and tau_f values in sintered ceramic bodies; the actual microwave performance depends critically on grain boundary chemistry, sintering atmosphere, and dopant-level control that computational methods cannot fully capture. If resonator measurements on sintered monolithic MgTa2O6 or borate ceramics reveal Q×f values below the performance thresholds required by filter designers (typically >10,000 GHz for demanding applications), the commercial case for the RF application weakens substantially. The open DFPT gate on piezoelectric/electro-optic tensors similarly means that the polar-borate application claims are currently supported by structural symmetry arguments rather than calculated or measured coefficients. The FTO posture is narrow by design, and buyers should conduct independent FTO analysis rather than relying on the carve-out characterization provided here, particularly for the brownmillerite electrolyte application where prior experimental literature is extensive. The de-risking roadmap is clear: commission DFPT calculations on the polar borates, process and characterize monolithic ceramic resonator bodies for MgTa2O6 and the borates, and obtain independent FTO analysis for the Ba2In2O5 electrolyte space — three defined workstreams that a well-resourced buyer or licensee could pursue concurrently within a 12–18 month timeframe.