Thulium orthophosphate (TmPO4) phonon-stable member of the rare-earth-phosphate separation platform

Thulium orthophosphate (TmPO4) is a dependent member of Lattice Graph's rare-earth phosphate separation platform, contributing structural and phonon evidence that anchors the platform's claimed claim coverage across the lanthanide series. The asset's role is strategic rather than performance-primary: by demonstrating that TmPO4, like GdPO4 and SmPO4, is dynamically stable in its xenotime structure at realistic temperatures, it helps establish that phonon instability within this orthophosphate family is localized to mid-shell occupancies — specifically the 4f8 (Tb) and 4f9 (Dy) configurations — rather than being a diffuse or unpredictable phenomenon across the series. This kind of systematic, species-by-species structural mapping is what transforms a collection of individual computations into a defensible genus claim. The practical significance sits at the intersection of two large, underserved challenges in critical-mineral supply chains. First, rare-earth elements are essential feedstocks for permanent magnets, catalysts, phosphors, and defense systems, but primary mining supply is geographically concentrated and increasingly subject to export controls. Recycling and secondary-recovery routes — particularly hydrometallurgical leaching of end-of-life magnets and lamp phosphors — are gaining regulatory and commercial urgency. Selective crystallization and precipitation using tailored phosphate ligand systems is one credible separation route. Second, patent protection for such separation chemistry requires breadth: a narrow claim covering only one or two members of a chemically similar family is readily designed around. TmPO4's inclusion in the claimed family expands claim coverage to the heavy lanthanide end of the xenotime series, closing a gap that a competitor or generic manufacturer might otherwise exploit. The asset is candid in its scope: no separation or recovery performance data is yet attributed to TmPO4 itself, and the primary validation employed one machine-learning interatomic potential rather than the full multi-potential consensus that the portfolio's lead members carry. The honest characterization is that TmPO4 is a well-computed, structurally sound supporting member whose contribution is evidentiary — it completes a pattern across the rare-earth series that the lead filings (EF11 and EF14) require to maintain broad coverage.

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.

TmPO4 crystallizes in the xenotime structure type (tetragonal, space group I4₁/amd), which is the thermodynamically preferred polymorph for the heavier rare-earth orthophosphates. Thulium carries a 4f12 electron configuration, placing it near the heavy end of the lanthanide series, well past the mid-shell where crystal-field and magnetic-exchange effects are most complex. The xenotime structure accommodates rare-earth ions in an eight-coordinate environment with alternating phosphate tetrahedra, and its rigidity makes it an attractive host for ions across a wide radius range. The material referenced by Materials Project entry mp-5884 is well-characterized in the experimental literature as a stable, refractory ceramic with no known thermodynamically competing polymorph under ambient conditions. Phonon stability was assessed using the MACE-MP-0 universal machine-learning interatomic potential on a 2×2×2 supercell with a 6×6×6 q-point mesh — a mesh density appropriate for resolving soft modes at zone-boundary wavevectors, which is precisely where rare-earth phosphates carrying 4f7–4f9 occupancies have been observed to show incipient instability in related families. The calculation yields a minimum phonon frequency of +0.137 THz with zero imaginary modes anywhere in the Brillouin zone. A positive minimum frequency with no imaginary branches is the standard criterion for dynamic (harmonic) stability: it indicates that every normal mode of the crystal represents a restoring force toward equilibrium rather than a runaway displacement. Complementing the harmonic calculation, a finite-temperature ab-initio molecular dynamics trajectory (MACE-MP-0 potential, 350 K, 2,000 steps) showed no bond dissociation or structural decomposition, confirming that anharmonic effects at a practically relevant temperature do not destabilize the structure. This combination — zero imaginary phonon modes plus intact structure through a thermal trajectory — constitutes a meaningful two-tier stability proof for purposes of claim support. An important honest qualification is warranted: the phonon computation used one machine-learning potential (MACE-MP-0), whereas the portfolio's lead members and highest-confidence assets are validated against multiple independent potentials (including CHGNet, MatterSim, and ORB) requiring consensus before advancement. For TmPO4, CHGNet validation is not yet reported and DFT phonon or energy calculations are absent. This means the stability verdict is credible — MACE-MP-0 is a well-benchmarked universal potential with demonstrated transferability across oxide chemistry — but it has not yet been cross-confirmed at the multi-potential consensus level that the portfolio reserves for its primary candidates. The asset is therefore appropriately categorized as a supporting member rather than an independently validated lead. The significance of the +0.137 THz minimum frequency is also worth contextualizing: it is a small but clearly positive value, meaning the structure sits in a shallow but genuine potential energy minimum. It is not the robust +0.5 THz or higher seen in chemically simpler oxides, but it is unambiguously stable by standard computational criteria. The pattern across GdPO4 (4f7, half-filled shell), SmPO4 (4f5), and TmPO4 (4f12) — all stable — versus Tb (4f8) and Dy (4f9) — the locus of soft-mode behavior — is chemically coherent. Lanthanide contraction and the specific crystal-field splitting patterns near the 4f7/4f8 crossover point create conditions for enhanced electron-phonon coupling and potential lattice softening. TmPO4's stability, sitting well past that problematic zone, is consistent with expectation and adds a data point that makes the soft-mode localization hypothesis more rigorously bounded. This is precisely the kind of negative-space evidence that strengthens genus claim coverage: a patent covering rare-earth orthophosphate separations is on firmer ground if it can demonstrate, member by member, which compositions are structurally viable within its claimed scope.

The commercial context for TmPO4 is defined entirely by the market for rare-earth separation and recycling technology, not by any applications of thulium phosphate as a material in its own right. Rare-earth elements — particularly the magnet metals neodymium, praseodymium, dysprosium, and terbium — are facing demand growth driven by electric vehicles, wind turbines, and defense hardware, while primary supply remains concentrated in China, which has periodically restricted exports. Secondary supply through recycling of end-of-life permanent magnets and fluorescent lamp phosphors is widely recognized as both a national-security priority and an emerging commercial opportunity. Hydrometallurgical routes that use selective precipitation with phosphate ligands to isolate individual rare-earth fractions from mixed leachates are under active development by multiple academic groups and at least several commercial entities. Within that context, TmPO4 itself is a minor constituent in commercial rare-earth streams. Thulium is one of the least abundant rare-earth elements in end-of-life magnet scrap and phosphor waste; it does not carry the volume or pricing of dysprosium, neodymium, or terbium. The asset therefore does not drive standalone revenue but instead contributes to the breadth of intellectual property coverage across the orthophosphate family — coverage that protects the EF11/EF14 lead assets from design-around attempts using adjacent heavy lanthanide compositions. No specific total addressable market figure is assigned to this member. The relevant market sizing belongs to the platform as a whole, where the separation and recycling of rare-earth elements from spent magnets and lamp phosphors represents a multi-billion-dollar global opportunity as recycling mandates in the European Union and the United States accelerate adoption of secondary-supply infrastructure. The monetization logic for this supporting member runs through the lead platform: a license or acquisition of the rare-earth phosphate separation portfolio gains value from the completeness of its claimed coverage, and TmPO4's inclusion is part of the evidence base that demonstrates the claimed family is scientifically well-founded across the xenotime heavy-lanthanide series. A licensee operating a commercial rare-earth hydrometallurgical facility would benefit from the certainty that comes with a patent portfolio covering a wide compositional range, not just the two or three members for which separation performance has been directly demonstrated.

genus completeness supporting the EF11/EF14 phosphate breadth

RE-recovery / magnet-recycling use per EF11/EF14 leads

The natural acquirers or licensees for this asset are buyers of the EF11/EF14 rare-earth phosphate separation platform as a whole rather than purchasers of TmPO4 in isolation. Hydrometallurgical companies building commercial rare-earth recycling capacity — particularly those processing end-of-life neodymium-iron-boron magnets or spent lamp phosphors — would benefit from holding broad claimed coverage over the rare-earth orthophosphate family, as it protects their process chemistry from generic or competitor filings on adjacent compositions. Strategic acquirers could include specialty chemicals companies with rare-earth processing divisions, rare-earth recycling startups seeking IP foundations prior to scale-up, and battery or magnet manufacturers seeking to internalize critical-mineral recovery technology to reduce supply-chain exposure. Defense-sector primes and critical-mineral funds operating under government mandates to develop domestic rare-earth supply chains represent a second buyer category, particularly given that thulium and neighboring heavy lanthanides appear in defense-relevant alloy and device specifications. In all cases, TmPO4 contributes incremental value to a platform transaction rather than anchoring a standalone deal. Its value is clearest in a due-diligence context where a buyer auditing the claimed family asks whether each named member is computationally enabled and free of blocking third-party claims — and for TmPO4, the answer is yes on both counts, with the honest caveat that multi-potential consensus and DFT validation remain outstanding.

The principal risk for this asset is that its validation depth is shallower than the portfolio's lead members. With only one machine-learning potential applied and no DFT phonon benchmark, a sophisticated technical reviewer during acquisition due diligence might flag TmPO4 as insufficiently corroborated compared to members that have achieved multi-potential consensus. This is not a fatal weakness — MACE-MP-0 is a well-validated foundation model for oxide chemistry, and zero imaginary modes plus a stable 350 K AIMD trajectory is a meaningful result — but it means the asset occupies a supporting rather than anchor role. The roadmap to de-risk this is straightforward: run CHGNet and at least one additional potential on the same structure to achieve the portfolio's standard consensus threshold, followed by a single-point DFT energy and optionally a DFT phonon calculation using the mp-5884 structure as the starting geometry. A secondary risk is functional: no separation or recovery performance has been asserted for TmPO4, meaning its inclusion in the claimed family as an enabled member could be challenged in prosecution if an examiner requires that each named member be shown to function in the claimed application. The mitigation here lies with the lead members (EF11/EF14): if GdPO4, SmPO4, or other family members generate separation performance data, TmPO4 benefits from genus-level enablement under the same structural family argument. The market risk — that thulium's low commercial abundance in rare-earth scrap makes TmPO4-specific separation economically marginal — is real but largely irrelevant to the asset's role, which is claimed family breadth and defensive coverage rather than direct commercial exploitation of thulium recovery.