Layered palladium and platinum sulfide chalcogenides for broadband infrared windows and nonlinear optics

The infrared optics and nonlinear photonics industries are built on a handful of chalcogenide crystals that have not fundamentally changed in decades. Silver gallium sulfide (AgGaS2) and its selenium analog (AgGaSe2) remain the workhorses for second-harmonic generation and mid-infrared transmission, but they carry real liabilities: silver is a regulated and strategically sensitive metal, the compounds are mechanically soft, and the phase-matching windows are constrained. More practically, silver-containing materials face tightening scrutiny in defense procurement and precision-instrument supply chains that demand long-term source security and environmental compliance. That pressure has opened a genuine substitution window — one that requires not merely a different material, but a material with the right combination of transparency range, phase-matching geometry, and non-centrosymmetric structure to host second-order nonlinear interactions. This patent family covers the class of A2MQ2 layered d8 chalcogenides — alkali or alkaline-earth palladium and platinum sulfides and selenides — as broadband infrared window materials and second-harmonic-generation hosts. The central insight is structural: palladium(II) and platinum(II) adopt square-planar d8 configurations that are inherently closed-shell, meaning no unpaired electrons, no paramagnetic loss, and clean optical properties across a transparency window computed to span roughly 0.5 to 25 microns. Four representative compositions have been computationally validated as dynamically stable, the broader chemical family has been enumerated across a 13-member composition space, and the family's freedom-to-operate landscape has been cleared with comparative SHG claims carefully scoped to avoid patent conflict. This is a lead asset in the integrated packaging, storage, and PFAS-treatment systems portfolio, filed as a composition-and-device-use claim covering both the materials themselves and their application as IR windows and nonlinear-optical hosts. The timing argument is straightforward: IR photonics is entering a growth phase driven by hyperspectral imaging, quantum cascade laser systems, atmospheric sensing, and defense-grade thermal imaging — all of which require high-quality mid-wave and long-wave IR transmitting optics. The incumbent crystal growers are supply-chain constrained, and no non-toxic, closed-shell layered chalcogenide has been staked in the patent record for these applications. The combination of demonstrated phonon stability, a clean freedom-to-operate position, and a defined composition family makes this an acquirable foundation for a next-generation IR crystal platform.

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 A2MQ2 system exploits a structural motif that is well-known in solid-state chemistry but underexplored for optical applications. Palladium(II) and platinum(II) are classic d8 metals: their ground-state electron configurations yield square-planar coordination geometry, fully occupied d-orbitals, and no unpaired spins. When these metals are embedded in a sulfide or selenide lattice with alkali or alkaline-earth charge-balancing cations — yielding compounds such as K2PdS2, Na2PdS2, K2PtS2, and Rb2PtS2 — the result is a layered structure with tetragonal D4h symmetry. The closed-shell nature of the d8 center is the optical key: paramagnetic centers in crystal lattices scatter and absorb light in ways that degrade both linear transmission and nonlinear conversion efficiency. Eliminating paramagnetic loss by design, rather than by purity control, is a meaningful engineering distinction. The transparency window claimed — approximately 0.5 to 25 microns — spans the near-infrared through the long-wave infrared, covering the atmospheric transmission bands (3-5 micron mid-wave, 8-12 micron long-wave) that are most commercially valuable for thermal imaging, quantum cascade laser pumping, and molecular fingerprinting spectroscopy. For comparison, AgGaS2 transmits from roughly 0.5 to 13 microns, and AgGaSe2 extends to about 18 microns but introduces phase-matching constraints. The proposed d8 chalcogenides, if the computed transparency window holds through synthesis and characterization, would represent a meaningful spectral extension into the long-wave band in a non-silver composition. The layered nature of the structure also raises the prospect of piezoelectric and potentially non-centrosymmetric polymorphs within the same compositional family, which is why the claim family is written to cover the layered piezo mode of use alongside the optical applications. Four compositions — K2PdS2, Na2PdS2, K2PtS2, and Rb2PtS2 — have each been subjected to phonon-stability calculations using two independent machine-learning interatomic potentials. Both potentials agree that all four structures are dynamically stable: the minimum phonon frequencies are positive across the full Brillouin zone, with values of +0.085 THz (K2PdS2), +0.935 THz (Na2PdS2), +2.231 THz (K2PtS2), and +1.516 THz (Rb2PtS2). Positive minimum phonon frequencies with no imaginary modes indicate that the crystal structures sit in true local energy minima, not saddle points — the structure would not spontaneously distort or decompose under small perturbations at low temperature. At least one of these compositions has independent DFT-level confirmation of this stability assessment. The energy hull data (convex hull positions reported in the description for the compositional variants) further supports thermodynamic accessibility of these phases. Importantly, the K2PtS2 and Rb2PtS2 members show the largest phonon frequency margins, suggesting greater dynamic robustness and potentially easier synthesis conditions. Two additional compositions — Cs2Pd3S4 and BaPt2S3 — are adopted in the family as prophetic backup members. Both sit on the convex hull thermodynamically, indicating they are not disfavored by thermodynamics, but neither has been confirmed in the literature (and for BaPt2S3, literature existence has not been confirmed at all). These are held as synthesis targets reserved for experimental verification; they are not presented as computationally validated leads in the same sense as the four primary members. The broader 13-member enumeration of the composition family — including selenide analogs (Na2PdSe2, K2PdSe2), heterometallic variants (K4Cr2PdS8, V2PdS4), and phospho-sulfide members (K4P2PtS8, NaPPtS4, GePtS3) — defines the scope of the composition claim and provides the family breadth needed to block obvious work-arounds. Critically, the one remaining computational gate is an in-house DFT or time-dependent DFT calculation of the second-order nonlinear optical (SHG) tensor for the primary members. This is needed both to confirm the non-centrosymmetric symmetry of the relevant polymorph and to quantify the d-coefficient for head-to-head comparison with AgGaS2 benchmarks.

The total addressable market for infrared optical materials and components sits in the $1-5 billion range, spanning crystal growth and characterization services, finished optical elements (windows, lenses, prisms, phase-matching crystals), and integrated photonic devices. The principal demand segments are defense and aerospace thermal imaging, quantum cascade laser systems for spectroscopy and countermeasures, atmospheric and environmental sensing, and the growing field of mid-infrared fiber optics. Within nonlinear optics specifically, the second-harmonic generation crystal market is smaller but high-margin: crystal growers for AgGaS2 and AgGaSe2 serve a relatively concentrated customer base of laser OEMs, defense primes, and national laboratory procurement offices, where the value per kilogram of polished crystal is substantial and switching costs (once qualified into a laser system) are high. The royalty and licensing logic follows two distinct tracks. In the crystal-growth-and-supply track, a materials company or specialty crystal grower would license the composition IP to manufacture and sell IR window blanks and NLO crystal boules — royalties in this model are typically assessed on finished-crystal revenue, with rates in specialty optical materials commonly running 3-8% of net sales for a foundational composition patent. In the device-integration track, a laser OEM or photonic systems integrator would license the device-use claims to incorporate the material into qualified laser or imaging products, with royalties assessed on system revenue at lower headline rates but on a larger revenue base. A third path — outright acquisition by a crystal grower or photonics materials company that wants to own the IP rather than license it — would be valued on a multiple of the downstream royalty stream or strategic exclusivity premium. The absence of any competing patent position in this chemical class (confirmed by freedom-to-operate screening across 300,000+ materials patents) means the acquirer would hold a clear blocking position rather than a crowded-field license.

non-toxic broadband IR-window SHG host with paramagnetic-loss suppression

comparative SHG removed from Clause CC-1; SHG comparison recited prophetically only

The most likely strategic acquirers for this family are specialty IR optical crystal growers and photonics materials companies that currently derive revenue from AgGaS2 or AgGaSe2 crystal sales and want to own the next-generation composition before a competitor does. Companies in this category include vertically integrated crystal growers serving defense and laser OEM customers, where long-term supply-chain control is a competitive differentiator. A second buyer class is laser OEM companies and defense primes with internal photonics R&D programs — particularly those developing quantum cascade laser systems, mid-IR LIDAR, or hyperspectral imaging payloads — who would acquire the composition IP to lock in materials exclusivity for their product roadmaps. Photonics holding companies that aggregate specialty optics IP for licensing-back programs represent a third path. Licensing without outright acquisition is also commercially viable given the clean FTO position and well-defined claim scope. A crystal grower could take an exclusive field-of-use license for IR windows, while a separate NLO laser integrator holds an exclusive SHG license, without either party conflicting with the other — the claim structure supports this kind of field-splitting. The portfolio context (integrated packaging, storage, and PFAS-treatment systems) is broader than photonics alone, so a photonics-focused buyer acquiring this single family can do so cleanly without needing to engage with the rest of the portfolio.

The primary technical risk is the absence of an experimentally synthesized sample for any member of the A2MQ2 family described here. Computed phonon stability is a strong predictor of synthesizability but not a guarantee — kinetic barriers, competing phases, and laboratory-specific challenges can prevent isolation of a predicted structure. For the prophetic backup members Cs2Pd3S4 and BaPt2S3 in particular, synthesis has not been attempted, and BaPt2S3 has not been confirmed in the literature at all. The second open risk is the SHG tensor: the transparency window and closed-shell advantages are computed or inferred, but the actual nonlinear coefficient has not been quantified. It is possible that the relevant polymorph is centrosymmetric, in which case the SHG application would require engineering a non-centrosymmetric form (e.g., through strain, templating, or compositional modification). These risks are addressable through a defined experimental roadmap: (1) solid-state synthesis of K2PdS2 and K2PtS2 from alkali sulfide and palladium/platinum sulfide precursors, followed by powder X-ray diffraction to confirm phase purity; (2) transmission spectroscopy from near-IR through long-wave IR to confirm the transparency window; (3) DFT/TDDFT calculation of the SHG tensor to quantify nonlinear coefficients and confirm non-centrosymmetric symmetry; and (4) single-crystal growth for phase-matching angle measurement. Steps (1) and (3) are independently executable and could proceed in parallel, with experimental synthesis de-risking the composition claims while the tensor calculation de-risks the NLO performance claims.