Single-phase Li5AlO4 aluminate anode-side interlayer for garnet solid-state batteries

The deep technical challenge in garnet-based solid-state batteries is not the bulk garnet electrolyte itself — it is the lithium-metal/garnet interface. Garnet ceramics (typified by Li7La3Zr2O12, LLZO) are inherently poor wetting surfaces for lithium metal: the ceramic is stiff, the contact is discontinuous, and local current focusing at voids and grain boundaries drives lithium dendrite nucleation and eventual short-circuit. The standard industry response has been to add lithium aluminate as a sintering aid — during co-sintering it distributes as a glassy or fine-crystalline residue throughout the garnet grain boundaries, helping densification but providing no spatially-controlled protection at the anode interface. This patent family covers something categorically different: a deliberate, post-densification, single-phase Li5AlO4-rich interlayer — concentrated asymmetrically on the anode-facing surface of the garnet pellet, not distributed throughout it. The inventive step is the combination of three precise elements: (1) a phase-pure or near-phase-pure Li5AlO4 layer (at least 60 mol% in the designated orthorhombic crystal structures) rather than a mixed or amorphous aluminate residue, (2) an engineered positional asymmetry — the interlayer is at least 1.5 times thicker or denser on the anode side than on the cathode side, with preferred ratios of 3–5×, and (3) a physical dimension that falls within the electrochemically-relevant range (5 nm to 2 µm) rather than the multi-micron sintering-aid residues that distribute through bulk ceramic. Together these features make the anode-side protection a measurable, engineered artifact — one that can be verified by cross-section electron microscopy and that provides a distinct, continuous interfacial buffer rather than a statistical scatter of aluminate grains throughout the pellet. The commercial argument is straightforward: manufacturers who already handle lithium aluminate precursors for sintering can apply a modified post-densification coating or in-situ conditioning step, making adoption compatible with existing infrastructure while producing cells that meet high-current cycling targets (critical current density at or above 0.8 mA/cm² at 25 °C, interfacial resistance at or below 30 Ω·cm²). The timing of this filing is strategically important. The global pivot toward lithium-metal solid-state batteries is accelerating, with automotive OEMs and consumer-electronics integrators publicly committing to ceramic electrolyte roadmaps for the late 2020s. No broadly-held patent yet claims a positionally-defined, single-phase aluminate interlayer with an explicit anode-side asymmetry ratio — the field has been dominated by sintering-aid process patents and by cathode-coating aluminate patents, both of which the claims expressly exclude. This creates a defensible whitespace claim that is broad enough to be strategically valuable to a garnet cell manufacturer while remaining technically precise enough to withstand scrutiny.

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.

Li5AlO4 is a lithium-rich aluminate crystallizing in the orthorhombic Pmmn (space group No. 59) structure as its primary polymorph, with Pbca (No. 61) as an alternative structural form — both are covered by the claimed scope. The material is electronically insulating (computed bandgap approximately 4.9 eV), which is a necessary attribute for any electrolyte-side interlayer: it must not create a mixed electronic-ionic conductor region that would propagate dendrites electronically. Its bulk ionic conductivity is modest — NEB-computed migration barriers run approximately 0.73 eV — making it unsuitable as a primary bulk solid electrolyte, and this is not its claimed function. The role of Li5AlO4 here is as a compliant, lithium-chemically-compatible buffer that presents a more wettable surface chemistry to lithium metal than bare garnet oxide while also offering chemical stability against reduction by metallic lithium at the anode potential. The aluminate structure provides this stability because Li5AlO4 is already in a highly lithiated state; it does not readily accept additional lithium reduction at the relevant electrochemical window. The computational validation of this material proceeded through two entirely independent machine-learning interatomic potential (MLIP) frameworks operating on both crystal structure variants. The Pmmn and Pbca polymorphs were each evaluated under MACE and CHGNet — two distinct MLIP architectures trained on separate DFT datasets — and both potentials agree that both structures are dynamically stable: no imaginary (negative-frequency) phonon modes appear in the phonon dispersion, and the lowest-frequency optical modes are real-valued (MACE computed approximately 1.41 THz; CHGNet computed approximately 1.69 THz for the relevant zone-center modes). Consensus across two independent potentials is the firm threshold the computational workflow requires before advancing a candidate; the agreement here provides genuine confidence rather than single-model artifact. This phonon consensus is supported by three independent DFT reference calculations drawn from the published literature and databases. The combination — consensus MLIP phonon stability plus multi-source DFT grounding — constitutes the current computational proof floor. Beyond static structural stability, the simulation program includes a nudged-elastic-band (NEB) calculation of the bulk lithium migration barrier, which returned approximately 0.73 eV. This number is deliberately included in the analysis not to promote Li5AlO4 as a fast-ion conductor — it is not — but to characterize the material honestly for its role: a thin interlayer with high barrier is tolerable if the layer is sufficiently thin (5 nm to 2 µm) that the areal resistance contribution remains within the 30 Ω·cm² target. Critically, at nanometer-scale thickness, even a material with a 0.73 eV bulk barrier can present acceptable total interfacial resistance while providing the chemical and mechanical buffering function. A gallium-substituted variant of Li5AlO4 has been explored in a simulation arm; partial Ga substitution can modify lattice parameters and potentially reduce the migration barrier, providing a development pathway if the unsubstituted composition proves insufficient at scale. Amorphous LixAlyOz compositions in the 4:1 to 6:1 Li:Al ratio range are also included in the claimed scope, reflecting that practical deposition processes (sputtering, ALD, solution coating) may produce partially or fully amorphous films that retain the interface-buffering function of the crystalline phase. Two key open validation gates remain. First, a direct experimental measurement of critical current density and the anode-side asymmetry ratio on a physical coupon — this is the P-type experimental proof that would convert the computational case into demonstrated device performance. Second, explicit-layer dielectric-function and phonon calculations (DFPT) targeting the specific Li5AlO4/garnet interface geometry rather than the bulk material alone; interface phonon calculations capture soft modes or instabilities that bulk calculations can miss. These are acknowledged gaps, not buried risks; any licensing or acquisition discussion should treat them as near-term de-risking milestones rather than disqualifying uncertainties.

The addressable market for solid-state battery electrolytes and interfaces is not uniformly realizable — the technology is in a staged commercialization phase, and the near-term accessible segment is the garnet-specific manufacturing supply chain rather than the entire solid-state battery market. The total addressable market for garnet-based solid-state battery components is estimated at roughly $3–5 billion over the relevant commercialization window, reflecting the anticipated ramp in automotive and consumer applications that have publicly committed to ceramic-electrolyte architectures. This is a rough estimate based on the projected scale of garnet cell manufacturing investment, not a bottom-up build, and should be treated as a directional planning figure. The actual buyer of this technology is the garnet cell manufacturer — companies that are building or licensing garnet electrolyte production lines for lithium-metal cells targeting automotive (range-extended EV packs) and premium consumer applications (medical devices, aerospace). These manufacturers face the same interface problem regardless of their specific garnet formulation: the lithium-metal/oxide ceramic interface is the dominant performance and lifetime limiter. A supplier of interlayer materials or a licensor of an interlayer process that demonstrably addresses this interface at commercial scale could extract royalty economics tied to per-Ah cell output or per-pellet processing volume. Licensing logic is straightforward: the interlayer process is a per-cell input with measurable yield impact, creating a natural per-unit royalty basis. An exclusivity premium is plausible for a Tier-1 automotive cell supplier seeking to lock out competitors from a proven interface solution. The enforceability of value capture depends critically on the specificity of the claim: the anode-side asymmetry ratio (at least 1.5×, preferred 3–5×) and the single-phase purity threshold (at least 60 mol%) together create a claim that is narrow enough to enforce — it requires measurement, not just presence of aluminate — while capturing any commercially viable implementation, because a manufacturer that achieves the performance targets almost certainly meets the structural thresholds. This makes the intellectual property position practically valuable, not just technically interesting.

simple manufacturable single-phase interlayer with measurable asymmetry endpoint

engineered positionally-defined interlayer + anode-side asymmetry distinguishes sintering-aid residue and cathode-coating use

The most natural acquirers or licensees are garnet solid electrolyte manufacturers and their supply-chain partners who face the lithium-metal wetting problem as an active production challenge. This includes companies building LLZO-based cell manufacturing capacity for automotive OEMs in North America, Europe, and Asia, as well as the OEMs themselves who are backward-integrating into electrolyte production. A Tier-1 automotive customer with a garnet electrolyte roadmap would have strong incentive to license this family exclusively, both to gain the interface performance benefit and to block competitors from accessing the same solution. Equipment and materials suppliers to the garnet manufacturing ecosystem — precursor chemical companies, thin-film deposition system vendors, and contract coating service providers — represent a second acquirer category: the interlayer process is a value-added service step that could be sold as a turnkey solution bundled with this IP. A third category is strategic acquirers from adjacent materials sectors — alumina and specialty ceramics manufacturers who already produce lithium aluminate at scale for other applications (refractory, phosphor, optical) and who might see forward integration into battery interface materials as a growth path. For any of these acquirer types, the key value proposition is the combination of a clean FTO position, a computationally validated material foundation, a precisely scoped claim that distinguishes from prior art, and a manufacturable process pathway that does not require exotic deposition tools — sputtering, ALD, or solution coating steps all fall within the claimed scope, giving a manufacturing partner flexibility in process selection.

The most significant technical risk is the migration-barrier magnitude: 0.73 eV is a substantial bulk barrier, and if the interlayer must carry appreciable ionic current (rather than functioning purely as a passivation and wetting layer), total cell impedance could be penalized. This risk is manageable at the lower end of the thickness specification (5–50 nm) but requires explicit experimental validation of the resistance-performance tradeoff across the full 5 nm to 2 µm thickness range. The gallium-substituted variant is a direct hedge against this risk — Ga substitution in aluminate structures typically modifies the local Li environment and can lower migration barriers — but the Ga-substituted arm has been explored only computationally and also requires experimental validation. The manufacturing uniformity of the asymmetric deposition is a second open question: achieving a controlled 3–5× anode-to-cathode thickness ratio at commercial scale on a porous or rough sintered ceramic surface requires deposition process development that has not yet been demonstrated. The strategic de-risking roadmap is clear. Near term, a small-batch coupon experiment (full-cell or symmetric-cell configuration) measuring critical current density and interfacial resistance for sputtered or ALD Li5AlO4 films on LLZO pellets would simultaneously address the performance gate and the manufacturability question. Concurrent DFPT interface calculations would provide the computational complement. Medium term, optimization of the asymmetry deposition process (directional sputtering, selective-face coating, or shadow-mask ALD) with cross-section microscopy verification of the asymmetry ratio would confirm the manufacturing pathway. These are well-scoped milestones that a well-resourced partner could complete within 12–18 months, converting a computationally grounded but experimentally open filing into a licensed product-ready technology.