Rare-earth disilicate dielectric platform for glass-core and redistribution-layer applications

The rare-earth disilicate dielectric platform claims Y2Si2O7, Lu2Si2O7, and a family of related rare-earth silicates as packaging dielectrics for glass-core substrates and redistribution layers. The core materials bet is that these crystalline silicates can simultaneously deliver low millimeter-wave permittivity and thermal stability above 600°C — a performance combination that polymer RDL dielectrics and silicon dioxide cannot jointly provide. Organic dielectrics fail the thermal budget of glass-core processing; SiO2 does not offer the tunable permittivity envelope that rare-earth substitution enables across this silicate chemistry. The timing is governed by the industry transition to glass-core packaging, which is creating an active search for dielectric materials that meet the electrical and thermal requirements of next-generation substrate build-up layers. Two disilicate members have been elevated to high-confidence status through density-functional perturbation theory calculations of the full dielectric tensor, providing a measured computational foundation that distinguishes this platform from literature-analog estimates. The IP strategy ties those two well-characterized compositions to a broader composition claim that spans monosilicates and an aluminoborate analog, giving a licensee compositional optionality within a single agreement. Within the broader critical-mineral recovery and recycling separations portfolio, this platform attacks the high-frequency, high-Tg corner of the advanced-packaging dielectric market, complementing positions in borate dielectrics and other oxide chemistries. The freedom-to-operate position is clean, with prior rare-earth silicate art confined to thermal-barrier and semiconductor-chamber coatings — a different application field with distinct device context and claim language.

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.

From density-functional perturbation theory applied to the Materials Project reference structures, Y2Si2O7 (beta-disilicate, mp-5652) has a static permittivity of 9.25, decomposed as an ionic contribution of 6.01 and an electronic contribution of 3.24, at a computed bandgap of 4.80 eV. Lu2Si2O7 (mp-18385) shows a static permittivity of 10.16 — ionic 6.77, electronic 3.39 — at a 5.20 eV gap. Both values are attested by two independent DFT source calculations, which is what justifies elevating these two compounds from literature-analog to high-confidence status. The permittivity decomposition is the critical materials-science point: at millimeter-wave frequencies the high-frequency (electronic) contribution dominates the loss-relevant dielectric response, so the 3.2–3.4 electronic permittivity range is favorable for signal integrity even as the ionic term brings the static value to 9–10. A low optical permittivity in an inorganic crystal is unusual — most high-static-permittivity ceramics carry a correspondingly high electronic term — and the rare-earth disilicate framework achieves the combination through the stiff Si-O-Si bonding network that limits lattice polarizability while rare-earth substitution tunes the ionic response. Dynamic stability has been assessed via phonon calculations using the MACE machine-learning interatomic potential, returning a minimum phonon frequency of 0.143 THz for Y2Si2O7 with no imaginary modes — confirming the structure is dynamically stable. A rare-earth-silicate survey covering phonon spectra and energy-above-hull across the composition family (including monosilicate and disilicate members) provides the broader thermodynamic context. The rare-earth disilicate framework also supports thermal stability well above 600°C, consistent with the melting points and phase stabilities known for this compound class. Deposition routes compatible with semiconductor toolsets — sol-gel, physical vapor deposition, atomic layer deposition, chemical vapor deposition, and frit-bonding — all exist in the literature for rare-earth silicates, giving integration flexibility without requiring custom equipment development.

The addressable market for advanced semiconductor packaging dielectrics — specifically redistribution-layer and glass-core build-up dielectric films — is estimated at one to five billion dollars. The buyers are glass-core substrate vendors and outsourced semiconductor assembly and test providers. Value accrues at the dielectric layer level, where pricing logic follows dielectric area per substrate multiplied by glass-core unit volume, with a premium for nodes where the combined low-loss and high-Tg specification excludes commodity alternatives. These are estimates reflecting the stated addressable range; no revenue commitment or market data beyond the provided range is implied. The premium portion of the market sits in millimeter-wave RDL and glass-core inter-layer dielectrics, where signal integrity at frequencies above 30 GHz demands both low permittivity and low loss tangent, while glass-core processing temperatures eliminate most organic dielectric options. That joint constraint is the economic gate that creates the opening: when neither polymer nor oxide commodity materials can be qualified, a materials licensor with a clean IP position and computational validation can negotiate platform licensing terms rather than per-unit supply agreements. Royalty and licensing logic supports a platform structure because the composition claim spans multiple silicate members. A glass-core vendor or OSAT that licenses the platform gains the freedom to optimize composition for a specific node — permittivity target, CTE match, deposition process — without renegotiating intellectual property. That optionality has value independent of any single member compound's performance, supporting a platform-license fee structure above what a single-compound license would command.

low mm-wave eps + high Tg where organic dielectrics and silica cannot jointly meet spec

RDL/glass-core packaging-dielectric use distinguishes thermal-barrier / chamber-coating prior art

Glass-core substrate vendors are the most probable primary licensees. Low-loss, high-Tg dielectric is a core technical requirement for a competitive glass-core product, and the multiple deposition routes — PVD, ALD, CVD, sol-gel — fit the range of integration approaches different vendors are pursuing. A field-of-use license scoped to glass-core and redistribution-layer applications gives a vendor the composition optionality to tune permittivity and CTE match at each node without reopening IP negotiations. OSATs are secondary buyers, more likely to take non-exclusive field-of-use licenses scoped to advanced packaging processes rather than substrate fabrication. Given the clean freedom-to-operate position and the complementary relationship with the borate dielectric platform in the same portfolio, a materials or substrate strategic that wants to control the dielectric IP space for glass-core packaging would find both platforms worth evaluating together. The same strategics reviewing the borate platform are the natural audience for this silicate position, and a bundled portfolio conversation covering multiple dielectric corners is more likely to produce a platform-level transaction than two separate single-asset negotiations. Milestone structures tied to the densified-film qualification gate are the expected deal shape, aligning licensor and licensee incentives on the validation outcome.

The primary risk is that device-relevant properties are computed or stated as design targets, not measured. Permittivity, loss tangent, Tg, and CTE all await densified-film experimental data; radiation hardness is a design target only. Claims and value arguments that lean on the monosilicate members or YAl3B4O12 carry additional risk because those members are supported by literature-analog reasoning rather than the direct DFPT attestation available for the two disilicates. Film densification at packaging scale — producing low-defect, low-loss films via PVD, ALD, or sol-gel at commercially relevant thicknesses — is unproven at even coupon level, and CTE compatibility with specific glass-core substrates must be confirmed experimentally before any node qualification can begin. The clear de-risking path is straightforward: fabricate a densified-film coupon and measure permittivity, loss tangent, Tg, and CTE. That single experimental campaign closes the most consequential open gate, converts the computational foundation into qualification-relevant data, and produces the dataset that strategic buyers require before committing to a licensing milestone. Extending the validation to a second rare-earth disilicate member — Lu2Si2O7 — and to one monosilicate member would further strengthen the genus claim and demonstrate composition-tuning across the permittivity range. Radiation-hardness testing, relevant for aerospace and satellite applications, can follow once the baseline dielectric and thermal properties are in hand.