Copper electroplating bath and waveform for void-free fill of high-aspect-ratio glass-core vias

The transition from silicon to glass cores in advanced packaging substrates is no longer speculative — it is happening in production lines at the leading OSAT and substrate houses, driven by the electrical performance advantages of glass (ultra-low loss tangent, tight CTE control, and lithographic planarity) combined with escalating bandwidth demands for AI accelerators and next-generation mobile chipsets. The indispensable unit operation enabling that transition is through-glass-via (TGV) copper metallization: filling blind and through vias with aspect ratios from 3:1 up to 20:1 without voids or seams. Electroplating baths developed for silicon TSVs are not adequate substitutes. Silicon-via chemistries are tuned to electrochemically distinct sidewall surfaces and narrower aspect-ratio windows; glass sidewalls, modified by adhesion coatings and seed-layer processes, present different adsorption kinetics and a far wider target aspect-ratio range. This invention addresses that gap directly with a purpose-built acid copper bath and pulse-reverse waveform architecture. The core insight is structural: rather than a single monolithic additive formulation, the invention organizes the plating chemistry into three independently deployable additive arms, each capable of achieving void- and seam-free bottom-up fill on its own, with different performance envelopes and cost profiles. This architecture matters commercially because it means a chemistry vendor can license one arm, two arms, or all three — broadening the deal structure — and because it provides IP coverage across the additive chemistry space that glass-core packaging suppliers are likely to explore. The glass-core advanced-packaging substrates portfolio of which this is the lead process asset reflects a deliberate strategy of covering the fill-chemistry design space at a level of specificity that maps directly onto what a next-generation substrate supplier must buy or design around. The asset is classified as a lead within that portfolio, which means it anchors the commercial licensing conversation for TGV fill process chemistry. It is not a defensive filing or a speculative backup; it represents the primary claimable process around which the rest of the portfolio's chemistry and method filings are organized.

The invention covers an acid copper sulfate electroplating bath combined with pulse-reverse waveforms, configured specifically for bottom-up superfilling of glass-core vias. The bath architecture comprises three independently functional additive arms. The first arm pairs DMSPAPS (a bis-sulfopropyl disulfide-type accelerator precursor) with PVI (a polyvinylimidazole-type carrier) and PEG (polyethylene glycol suppressor). DMSPAPS-class accelerators are well understood in damascene copper: they adsorb preferentially at via bottoms due to curvature-driven surface concentration, establishing the bottom-up growth gradient. PVI carriers modulate accelerator delivery kinetics. The PEG suppressor suppresses sidewall and field deposition to redirect current density downward. In a glass-core geometry — where the sidewall composition, roughness, and seed-layer chemistry differ from thermally oxidized silicon — the relative concentrations and molecular weights of these three components require re-optimization, and the claim scope reflects that retuning. The second arm uses BMIM-Cl, an imidazolium-based ionic liquid, as a leveling agent. Ionic-liquid levelers are a relatively recent entrant to via-fill chemistry; they function by differential adsorption at high-current-density protrusions, suppressing field deposition and leveling surface topography after the via is filled. The use of a room-temperature ionic liquid in this role is mechanistically distinct from conventional polyamine or Janus-green-type levelers, and that distinction is part of the claim space. The third arm combines CP-DMBI (a reducing-agent class organic, used here as a leveler in a secondary role) with BSAED (a bis-sulfoethyl-amino ethylene-type accelerator). This pairing represents an alternative accelerator-leveler couple with a different adsorption-energy profile at the copper (100) surface — the dominant surface exposed during bottom-up fill — which was characterized computationally (see proof narrative). The separation of arms allows a formulator to choose the leveler chemistry independently of the accelerator chemistry, providing performance tuning latitude that a single locked formulation does not. The pulse-reverse waveform is co-claimed with the bath chemistries. Pulse-reverse electroplating periodically reverses current direction, which serves to re-equilibrate additive surface coverage, dissolve locally deposited nodules at high-aspect-ratio via shoulders, and allow mass transport of fresh additive to the via bottom. For aspect ratios above roughly 10:1 — which are expected in the thicker glass-core substrates being developed for AI package interposers — direct-current plating produces shoulder pinch-off and void entrapment at unacceptable rates. The waveform parameters (forward pulse duration, reverse pulse duration, duty cycle, peak current densities) interact with additive concentrations in a coupled optimization space; the claims cover that coupled space as well as the individual arms.

The addressable market for TGV copper fill chemistry is best understood as a slice of the broader advanced-packaging materials market, which is itself expanding rapidly on the back of AI infrastructure build-out, high-bandwidth memory stacking, and the shift toward chiplet-based heterogeneous integration. Glass-core substrates are being pursued by a set of substrate manufacturers — primarily in Asia and increasingly in North America under CHIPS Act-sponsored capacity — as an alternative to conventional organic (BT resin) cores for the next generation of package substrates. Volume-production timing estimates range from 2026 to 2029 depending on the supply chain participant, with the earliest adopters targeting server-class AI accelerator packages where the electrical performance premium of glass justifies the process cost. The estimated addressable market for the specialty chemistry serving TGV fill in production-scale glass-core substrates is in the range of $500 million to $2 billion annually once glass-core substrate production reaches meaningful scale — these are estimates, and the upper end assumes broad adoption across AI, mobile, and automotive advanced-packaging markets rather than a server-only scenario. The customer for this IP is not the substrate manufacturer directly; it is the plating chemistry vendor who formulates and sells the acid copper bath to the substrate line. That intermediary — a company such as a specialty electrochemical materials supplier — takes the additive package, qualifies it on customer process equipment, and sells it as a consumable with ongoing supply. Licensing or acquisition of the bath chemistry claims thus targets a vendor rather than an OEM, which means the commercialization pathway is a royalty-bearing license on chemistry sales or an outright acquisition of the chemistry IP to control the vendor's freedom to operate in this space. Royalty logic would typically attach to the volume of additive package sold per unit of substrate area processed, making the economics track directly to glass-core substrate production volumes.

bottom-up fill tuned to glass-core sidewall vs Si-TSV baths

three independent arms not subject to watch-list; dependent leveler arms need per-arm clearance

The natural acquirers or licensees for this asset are specialty electrochemical materials companies that supply semiconductor process chemistry — the tier of vendors who develop and qualify copper plating baths for logic foundries, memory manufacturers, and substrate houses. Companies in this category are actively investing in TGV process chemistry to position themselves ahead of production ramp at glass-core substrate manufacturers, and a filing with independently usable claim arms covering three distinct additive approaches provides meaningful IP leverage in their product roadmap. Acquisition of the asset would allow such a company to block or monetize competitors entering the glass-core via-fill market with similar formulations, while a license would provide freedom to operate with a defined royalty structure. Either path is commercially logical for a vendor seeking to differentiate its product line in this emerging segment. Substrate manufacturers with in-house plating chemistry development — a smaller but real set of vertically integrated players — are a secondary buyer class. For them, the interest is less in licensing the chemistry externally and more in acquiring the IP to secure freedom to operate for their own process development without risk of infringement. The asset's lead status within the glass-core advanced-packaging substrates portfolio means a full-portfolio acquisition, which would include the related chemistry, process, and materials filings, is also a credible deal structure for a buyer who wants comprehensive coverage of the glass-core substrate supply chain.

The primary technical risk is the open experimental validation gate: the void/seam cross-section coupon program has not yet been completed. The computational work — Cu(100) adsorption energies, Cu(II)-additive cluster calculations, xTB frontier orbitals — provides mechanistic support for the additive design choices but does not substitute for physical fill-quality data. Until coupons are plated and cross-sectioned at representative aspect ratios (at minimum the low, mid, and high ends of the 3-20 range), the claim that the bath achieves void- and seam-free fill is a computational prediction with strong theoretical grounding, not a demonstrated experimental result. This is a standard and addressable risk — coupon fabrication and SEM cross-section characterization are routine semiconductor process steps — but it is the single most important item on the de-risking roadmap before prosecution arguments or licensee-facing performance claims can be made with full confidence. The secondary risk is the narrow FTO position on dependent claims, discussed in the FTO section. Completion of the per-arm clearance workstream is the de-risking step. A third risk is timing: the window for IP-based leverage in glass-core via chemistry will narrow once the leading additive chemistry suppliers complete their own TGV-specific development programs and begin filing. The absence of a disclosed race window in the commercial data reflects the early stage of the market rather than absence of competitive urgency; a buyer should treat the three-to-five-year production ramp timeline for glass-core substrates as the practical clock on establishing IP position.