US 9615453

I'll search for authoritative information on this specific patent number.

Let me search specifically for litigation and post-grant activity, and verify the claims text.

Let me check specifically for any litigation, PTAB, or Federal Circuit activity tied to this patent or its owner.

US 9,615,453 B2 — Verified Summary

Note on searching: I searched for this exact number (9,615,453). Google Patents, FreePatentsOnline, and secondary databases all return the same single patent; the only confusable hits were unrelated art (e.g., WO 96/15453, FR 9615453 — neither is this patent). I found no Federal Circuit 2026 docket for this patent or its owner. I did find active district-court and PTAB activity, detailed below.

Bibliographic data (verified)

Source: https://patents.google.com/patent/US9615453/en ; https://www.freepatentsonline.com/y2014/0085842.html

Abstract (verbatim)

"A substrate comprising a solid glass core having a first surface and a second surface opposed to the first surface; multiple conductors extending through the solid glass core beginning at the first surface and ending at the second surface, wherein one of the conductors has a third surface and a fourth surface, wherein the third surface and the first surface are substantially coplanar, wherein the second surface and the fourth surface are substantially coplanar, wherein one of the conductors comprise a copper-tungsten alloy material, wherein the solid glass core is directly contact with the conductor; and a first dielectric layer and a first metal layer formed at the first surface, wherein the first metal layer at the first surface is electrically coupled with one of the conductors."

Plain-language overview of the claims

Claim 1 (independent — apparatus/substrate claim): A substrate, not a method. It has four required elements:

1. A solid glass core with two opposed surfaces (first/top, second/bottom).
2. Multiple conductors running all the way through the glass core, from the first surface to the second surface — i.e., through-glass vias.
3. The conductors are flush with the glass: the conductor's top face is substantially coplanar with the glass's top surface, and the conductor's bottom face is substantially coplanar with the glass's bottom surface. At least one conductor is made of a copper-tungsten alloy.
4. The glass is directly in contact with that conductor (no intervening adhesion/liner layer required by the claim), and a first dielectric layer plus a first metal layer are formed at the first surface, with the metal layer electrically coupled to a conductor.

Note: FreePatentsOnline renders this final element as "first dielectric layer and a first metal layer… first metal layer… electrically coupled."

Second independent claim (identity uncertain): The specification's second summary embodiment recites a substrate/composition variant in which the conductor is instead a multi-layer conductor — "a first metal layer and a second metal layer coated [on] the first metal layer" — with the glass again directly contacting the conductor, and "a first dielectric layer and a third metal layer formed at the first surface" electrically coupled to a conductor. The claim tail retrieved matches this text exactly. I could not authoritatively retrieve this claim's number from the sources available; based on the 20-claim set and staged claim structure I believe it is a later independent claim (likely claim 11), but I am flagging that as an inference, not a verified fact.

Dependent claims (12–20 range): Not retrieved in full. Based on the disclosure, they would narrow features such as the covering layers on the conductor (first covering layer 6a, e.g., Ni/Zn/Ti/Ta/Ag/Cr or an anti-oxidation oxide; second covering layer 6b as an adhesion layer), conductor cross-sectional shape (circular, square, oblong, rectangular, flat), and dimensions/thicknesses. Treat this paragraph as description-derived, not claim-verified.

Important drafting observation: The title and specification describe a "method for fabricating" a glass substrate package, but the independent claim coverage I verified is directed to a substrate apparatus, not a method. If method claims exist, I did not retrieve them. Also note the specification contains apparent unit inconsistencies (e.g., "thermal resistance layer 8 has a thickness between 0.05 and 1 meter"; "glass layer 16 has a thickness greater than 0.5 meters") — these are likely intended as micrometers. Flagging as an on-face anomaly, not making a correction.

Related family (continuations claiming priority from 14/036,256)

US 15/261,956 → US 10,622,310 B2; US 15/476,999 → US 10,096,565 B2; US 16/139,040 → US 10,453,819 B2; US 16/752,650 → US 11,107,768 B2; US 17/372,459 → US 11,538,763 B2; US 17/985,827 → US 11,894,306 B2; US 18/524,118 → US 12,062,618 B2; US 18/778,983 → US 2024/0379566 A1; US 19/001,689 → US 2025/0132260 A1.

Litigation / PTAB activity found (not CAFC)

• MYW Semitech, LLC v. [Apple Inc.](/litigations/by-plaintiff/Apple%20Inc.), No. 1:25-cv-00504 (D. Del.) — patent infringement action by the current assignee. Docket filings dated through early 2026 (scheduling order Sept. 5, 2025; a Feb. 12, 2026 order and hearing transcript appear as IPR exhibits).
• Taiwan Semiconductor Manufacturing Company Ltd. v. MYW Semitech, LLC, IPR2026-00067 (P.T.A.B.) — a 2026 inter partes review petition naming the same patent owner; its exhibit list references the Apple D. Del. case, including a stay-related dispute.

Uncertainty flag: I confirmed the parties and case numbers, but I did not confirm from the search results that US 9,615,453 specifically is the patent-in-suit in the Apple case or the patent challenged in IPR2026-00067 — those dockets involve multiple MYW Semitech patents. Do not treat the '453 number as established for those proceedings without checking the complaints/petition directly.

CAFC 2026: No Federal Circuit appeal or docket involving US 9,615,453 was found. As of this search I have no evidence of CAFC activity. I state that as absence of evidence, not as proof none exists.

I'll search for litigation involving US patent 9615453 across patent litigation databases.

Let me search more targeted litigation databases.

Found a lead — MYW Semitech, LLC v. [Apple Inc.](/litigations/by-plaintiff/Apple%20Inc.) Let me verify which patents are asserted and check for other cases.

Summary

I found no litigation that asserts U.S. Patent No. 9,615,453. I searched patent‑litigation trackers (Unified Patents portal, RPX Empower/litigation, Docket Alarm, CourtListener/Justia dockets, E.D. Tex. blog) and general web sources for the literal string "9615453" in a litigation context. Every hit for that number was a false positive unrelated to this patent (e.g., WO 96/15453 to Spectral Diagnostics; FR 9615453 (cosmetics); VA Citation NR 9615453; various Brazilian/Chinese registry numbers). None was a U.S. patent case, ITC investigation, or PTAB proceeding involving this patent.

I want to be explicit about the confidence level: this is a negative finding based on the searches I was able to run, not a certification that no case exists. I could not exhaustively query PACER or the Unified Patents caselist by patent number before exhausting my search steps, so a recent or obscure filing could have been missed. What I can state with confidence is that no assertion of the '453 patent appeared in any source I retrieved.

Important context: litigation against sibling patents in the same family

The '453 patent is the earliest‑issued member and the priority anchor of a family owned by MYW Semitech, LLC (assignee recorded Dec. 27, 2024; inventor Ping‑Jung Yang). Per the Google Patents family data, nearly every later member claims priority back through the '453 application (US 14/036,256): US 10,622,310, US 10,096,565, US 10,453,819, US 11,107,768, US 11,538,763, US 11,894,306, US 12,062,618, US 2024/0379566, and US 2025/0132260.

MYW Semitech is actively litigating, but on sibling patents in that chain — not on the '453:

Related PTAB proceedings (filed Nov. 21, 2025) by Taiwan Semiconductor Manufacturing Company Ltd. et al. v. MYW Semitech LLC, none involving the '453:

• IPR2026‑00065 — IPR of US 11,107,768
• IPR2026‑00067 — IPR of US 11,894,306
• (the corresponding third petition — the '763 — falls between these numbers)

Sources:

• https://litigation.rpxcorp.com/litigation/dedce-88910-myw-semitech-v-apple
• https://dockets.justia.com/browse/court-dedce/noscat-10/nos-830?page=6
• https://ai-lab.exparte.com/case/dct/ded/1:25-cv-00504/doc/analysis/18
• https://sandbox.docketalarm.com/cases/PTAB/IPR2026-00067/… (exhibit list citing MYW Semitech, LLC v. Apple Inc., No. 1:25‑cv‑00504)
• https://litigation.rpxcorp.com/ptab/19953-… (IPR2026‑00065, filed 11/21/2025)
• https://edtexweblog.com/page/3/ (report that the stay was granted after institution on all three asserted patents)

Collateral matter worth flagging

There is a third‑party litigation funding dispute in the Delaware action. A document styled with the E.D. Tex. cause number "25‑cv‑o" (a suggestion to the court filed by the International Legal Finance Association) describes an Apple subpoena seeking discovery into MYW Semitech's funding, identifying Davidson Kempner Capital Management LP and Piccadilly Patent Funding LLC, and naming inventor Ping‑Jung Yang. The docket also lists Davidson Kempner and Piccadilly Patent Funding as participants. This is consistent with the USPTO assignment record, which shows a patent security agreement in favor of Piccadilly Patent Funding LLC as security holder, recorded March 25, 2025, shortly before the complaint was filed.

• https://www.uscourts.gov/sites/default/files/document/25-cv-o_suggestion_from_international_legal_finance_association_-_tplf.pdf

Bottom line

• US 9,615,453: no known litigation, no known IPR/PGR, no ITC §337 investigation. The patent's legal status in the record is "Expired – Fee Related," with an adjusted expiration of 2034‑09‑09.
• The only active assertion campaign by its owner — MYW Semitech, LLC v. Apple Inc., 1:25‑cv‑00504 (D. Del., filed Apr. 24, 2025) — pleads different patents in the same family ('768, '763, and apparently '306), not the '453.

Caveat: I interpret "US9615453" literally and applied no auto-correction. Given my limited number of search steps, I recommend a direct PACER docket search on "9,615,453" and a query of the Unified Patents litigation caselist to confirm this negative finding before relying on it.

I'll start by verifying the structured data against live sources, then search for any PTAB or Federal Circuit activity.

Let me run targeted searches on the patent owner and possible assertion activity.

Let me dig into the related TSMC IPRs and confirm whether the '453 patent itself was ever challenged.

PTAB Proceedings — US 9,615,453 ("Method for fabricating glass substrate package")

Verification note up front: The structured "PTAB proceedings on file" block reports that the USPTO ODP API returns no AIA trial proceedings for US 9,615,453. My independent web sweep — searching for 9615453 + IPR/PTAB/FWD, the assignee MYW Semitech, and inventor Ping-Jung Yang — surfaced no petition, institution decision, or Final Written Decision naming the '453 patent. I hit a tool-call cap mid-sweep, so this is a strong-but-not-exhaustive negative. Everything below distinguishes what is confirmed from what is inference.

Proceedings overview

Total AIA trials on US 9,615,453: 0. There are no IPRs, PGRs, or CBM reviews — no claims invalidated, none sustained, none settled, no institution denials, and no Federal Circuit appeals from any '453 FWD, because no FWD exists. The defensive bottom line is therefore not "the patent has survived two IPRs and is hardened" and not "claims 1–5 are canceled." It is: the '453 has never been validity-tested at the PTAB at all — but it appears to have lapsed for failure to pay maintenance fees ("Expired – Fee Related" per Google Patents, adjusted expiration 2034-09-09), and the patent family it anchors is under active, multi-front attack on sibling continuations. For a defendant, the '453 is a paper threat, not a litigated one; the live fight is over its descendants.

No proceedings to report on the '453 itself

There are no per-proceeding entries to write. Rather than manufacture one, here is the adjacent PTAB activity — proceedings on patents in the same family, which matter enormously for anyone evaluating the '453 — clearly labeled as not challenges to the '453.

IPR2026-00065 — Taiwan Semiconductor Manufacturing Company Limited et al. v. MYW Semitech LLC

• Type: Inter Partes Review
• Patent challenged: US 11,107,768 (a continuation in the '453 family — priority traced through the same chain)
• Filed: 2025-11-21
• Status: Third-party database (RPX Empower) lists an institution date of 2026-05-06. I could not independently confirm the institution decision text or the panel. Treat as provisional.
• Judge panel: Not confirmed in the sources retrieved.
• Petition grounds: Not retrieved — I did not locate the petition PDF or its Ground table.
• Institution decision: Listed by RPX as 2026-05-06; the panel's reasoning is unknown to me.
• FWD: None confirmed; if institution issued 2026-05-06, the statutory 12-month deadline puts an FWD around 2027-05-06.
• Settlement / termination: None reported.
• Appeal: None confirmed.
• Defensive value: Zero direct estoppel effect on the '453, but it demonstrates that TSMC is willing to fund PTAB attacks on this family — a signal that a '453 petition would be institution-viable in TSMC's hands.

IPR2026-00066 — Taiwan Semiconductor Manufacturing Company Limited et al. v. MYW Semitech LLC

• Type: Inter Partes Review
• Patent challenged: US 11,538,763 (same family)
• Filed: 2025-11-21
• Status: RPX Empower lists an institution date of 2026-05-15 (again, provisional — not verified against the PTAB docket).
• Judge panel / grounds / FWD / settlement / appeal: Not retrieved.
• Defensive value: Same as above; reinforces the pattern of coordinated family-wide petitions, not a one-off.

IPR2026-00067 — Taiwan Semiconductor Manufacturing Company Limited et al. v. MYW Semitech LLC

• Type: Inter Partes Review (caption per Docket Alarm / RPX)
• Patent challenged: Not confirmed. Given the three patents asserted in the parallel district court case ('768, '763, '306), the third petition most likely targets US 11,894,306 — but I could not verify this, and you should not rely on it.
• Filed: 2025-11-21
• Status / panel / grounds / FWD / settlement / appeal: Not retrieved.
• Defensive value: Unknown pending docket confirmation.

Why this matters for the '453 even though it isn't the '453: these three patents are direct continuations of the '453 (the '453's Priority to entries on the Google Patents page run to US 15/261,956 → US 10,622,310; US 15/476,999 → US 10,096,565; US 16/139,040 → US 10,453,819; US 16/752,650 → US 11,107,768; US 17/372,459 → US 11,538,763; US 17/985,827 → US 11,894,306; US 18/524,118 → US 12,062,618). Anyone litigating over the family should expect the '453's specification and disclosure to be the common § 112 and claim-construction battleground.

Strategic summary

Claim status of the '453: every claim is UNTESTED. Nothing has been canceled, nothing has been confirmed patentable, and no claim has been construed by the PTAB. The only validity posture information available is the patent's own prosecution and the Google Patents legal-status field, which reads "Expired – Fee Related," with an adjusted expiration listed as 2034-09-09. That field indicates the patent lapsed early for non-payment of a maintenance fee rather than running to its adjusted term. I was unable to pull the underlying USPTO maintenance-fee record in this session, so verify this directly on Patent Center before relying on it — but if accurate, it is dispositive in a way no IPR could be: an expired patent cannot be infringed prospectively, which guts any injunctive or ongoing-royalty theory and leaves at most a six-year lookback on past damages. Timing is suggestive: the '453 was assigned to MYW Semitech LLC on 2024-12-27, and a patent-security agreement to Piccadilly Patent Funding LLC was recorded 2025-03-25 — a window that overlaps the 7.5-year maintenance-fee grace period for a patent granted 2017-04-04. That overlap is a hypothesis, not a finding.

Estoppel landscape. Because no AIA trial has been instituted on the '453, no § 315(e)(2) estoppel exists on the '453 for anyone. Every prior-art ground — § 102, § 103, and (for a PGR-eligible later-filed continuation, § 112) — remains fully available to any petitioner. Conversely, the § 315(b) one-year bar is the defendant's clock, not the patent owner's: if you are served with a complaint asserting the '453, you have one year to petition, and the family IPRs suggest the art is out there and the budgets are committed. Note the asymmetry the family shows — MYW Semitech did not assert the '453 in MYW Semitech, LLC v. [Apple Inc.](/litigations/by-plaintiff/Apple%20Inc.), No. 1:25-cv-00504 (D. Del., filed 2025-04-24), which pleads only the '768, '763, and '306 patents. If you are a defendant sitting on a '453 demand letter, your one-year clock may not have started at all.

Pattern signals. (1) Same petitioner, multiple petitions: TSMC (with at least one co-petitioner — the caption reads "et al," which I could not resolve) filed three coordinated IPRs on 2025-11-21 against sibling patents, which is the classic family-wide squeeze. (2) No defensive aggregator visible: I found no Unified Patents, RPX, or other aggregator challenge on the '453; the funding trail instead runs to Davidson Kempner Capital Management LP and Piccadilly Patent Funding LLC, both of which Apple has subpoenaed in third-party discovery aimed at litigation-funding and real-party-in-interest issues. (3) Patent owner posture: no Federal Circuit appeals by the patent owner on any '453 proceeding, because there is none; the patent owner's enforcement energy is going into the continuation patents, not the '453.

Recommended next steps

1. Confirm the maintenance-fee status of the '453 today. Pull the Patent Center fee history and the maintenance-events tab. If the patent indeed lapsed, that fact — not any PTAB outcome — is the cleanest defense, and it costs you one afternoon. Cite Google Patents only as a lead, never as proof.
2. Confirm the sibling IPR dockets. IPR2026-00065/‑00066/‑00067 were filed 2025-11-21; if they instituted in May 2026, Final Written Decisions are due roughly one year later (≈2027-05-06 and ≈2027-05-15) under the § 316(a)(11) statutory deadline, with oral hearing typically 2–3 months prior. Verify the third petition's target patent and the joinder/co-petitioner identity — co-petitioners and privity matter for who gets estopped in the district court case.
3. If you hold a '453 demand letter, treat it as low-urgency but not zero-urgency. An expired patent supports no prospective relief. A pre-suit investigation into the fee lapse, plus a preservation letter, is proportionate. Do not build a licensing position on the assumption the patent is live.
4. If the '453 is ever asserted, petition immediately and preserve the § 112 record. With no prior institution, an IPR is available within one year of service, and given that the family's asserted claims are already under TSMC attack, the prior art and expert infrastructure for a '453 petition likely already exists.
5. Flag the funding/RPI angle. The Piccadilly Patent Funding security interest (recorded 2025-03-25) and the Davidson Kempner discovery fight are the same tools that defeated or complicated other funder-backed monetization campaigns. RPI and privity discovery can affect both standing-adjacent arguments and estoppel scope.

Sourcing note: I could not retrieve a petition PDF, institution decision, or FWD for any family IPR within this session, and I was rate-limited before reaching the PTAB E2E docket or CourtListener. The RPX and Docket Alarm data above should be treated as leads; the authoritative record is PTAB E2E and the Google Patents page for US 9,615,453.

I'll research the assignment record for this patent. Let me run several targeted searches.

Let me search for more specific assignment and ownership details.

I was able to retrieve the conveyance events, parties, and dates, but not the reel/frame numbers or correspondent-of-record fields — the Assignment Center's underlying records were not exposed in the results I could reach. I flag every place where that gap matters rather than filling it in.

Inventors

Pattern note: This is a single-inventor, individually-owned portfolio, not a corporate-derived one. That inverts the usual NPE narrative: there is no operating company whose product line was sold off. Instead, one inventor built a large continuation family (at least 9 later filings claiming priority from 14/036,256, per the earlier section) and then conveyed it wholesale to a licensing LLC in 2024. Also note that Yang is not anonymous in the assertion campaign — Apple's third-party subpoena expressly names "Ping-Jung Yang (the named inventor of the Patents-in-Suit)" as a party whose communications with the funder are sought (uscourts.gov, D. Del. 25-cv-504, subpoena dated ~2026). That is a stronger-than-normal inventor-to-asserter continuity link.

I found no evidence that Yang was ever employed by a semiconductor manufacturer or that any employer held rights. Absence of evidence, not proof.

Original assignee

• Entity on the issued patent (2017-04-04): Individual — Ping-Jung Yang.
• Primary line of business: Not determinable from the record. There is no evidence of any commercial product embodying the claims. The specification is written as a fabrication-process disclosure (glass-core substrate with through-glass copper-tungsten conductors, damascene and embossing metallization, COG bonding to OLED/MEMS/LCD panels), and the granted coverage is apparatus-directed — consistent with an inventor-held portfolio, not a product company's patent.
• Current status of the original assignee: Yang remains an individual; he is not the current owner. He executed an assignment of assignor's interest to MYW Semitech, LLC, recorded 2024-12-27 (Google Patents legal events; reel/frame not retrieved).
• Successor owner today: MYW Semitech, LLC, subject to a recorded security interest.

On "did they ship a product": I have no evidence of any shipped product from Yang or from MYW Semitech. MYW Semitech's observable activity is litigation, not manufacturing. I state that as a finding about the record, not about Yang's private activities.

Assignment timeline

Retrieval limitation, stated plainly: The two post-issuance conveyances below are attested by Google Patents' legal-events table for US9615453B2. I could not retrieve the reel/frame numbers, execution dates, or correspondent-of-record names from USPTO Assignment Center within this session. I am therefore reporting the recorded/effective dates and conveyance types that are documented, and leaving reel/frame and correspondent fields explicitly blank rather than inventing them. Verify at https://assignmentcenter.uspto.gov/ .

2024-12-27 (recorded) — Reel/Frame: not retrieved (field exists; I could not pull the number)

• Conveyance: Assignment of Assignors' Interest ("SEE DOCUMENT FOR DETAILS")
• Assignor: YANG, PING-JUNG, MR. (individual, sole inventor)
• Assignee: MYW SEMITECH, LLC
• Correspondent: not retrieved. I could not confirm the recording attorney/firm. Flag: this is precisely the field where an NPE chain usually reveals a repeat-player filer, and it is the single most valuable item still missing from this reconstruction.
• Context: transfer-to-asserter — inventor-owned assets conveyed to a licensing/assertion LLC roughly four months before the first infringement complaint in the family (see next entry).

2025-03-25 (recorded) — Reel/Frame: not retrieved

• Conveyance: Patent Security Agreement (grant of security interest; assignee recorded as "PICCADILLY PATENT FUNDING LLC, AS SECURITY HOLDER")
• Assignor: MYW SEMITECH LLC
• Assignee: PICCADILLY PATENT FUNDING LLC, as security holder
• Correspondent: not retrieved.
• Context: securitization — collateral pledge of the patent estate to a litigation-funder-affiliated entity, recorded ~30 days before MYW Semitech filed suit against Apple (complaint dated 2025-04-24).

No other recorded assignments found. No merger, change-of-name, release, or correction filings surfaced. No Chapter 7/11 sale, no defensive-aggregator transfer.

Execution dates vs. recording dates: I could only confirm recording dates. Given that the '453 assignment was recorded roughly four months before suit and the security agreement roughly one month before suit, the execution dates were almost certainly on or before those recordings — but I have not verified them and will not assert them.

Timeline diagram

timeline
    title Ownership of US 9615453
    2012 : Provisional application filed
    2013 : Nonprovisional application filed
    2017 : Patent issued to inventor Yang
    2024 : Assigned to MYW Semitech LLC
    2025 : Security interest to Piccadilly
         : MYW Semitech sues Apple

NPE / troll-pattern signals

Important scoping caveat first: In the D. Del. action, the patents actually asserted are US 11,107,768, US 11,538,763, and US 11,894,306 — not US 9,615,453. So for '453 specifically I have no direct assertion evidence. What I do have is that Apple's subpoena defines "Semitech Patents" as "any U.S. or foreign patent assigned to MYW Semitech LLC, now or in the future (including the Patents-in-Suit)," and '453 is in that assigned family (all four asserted/accompanying patents descend from application 14/036,256). Treat the assertion-linked signals below as family-level, not '453-level.

Additional signal the checklist doesn't enumerate but the record supports — funder-controlled assertion: The 2025-03-25 Patent Security Agreement to Piccadilly Patent Funding LLC, as security holder, paired with (a) Davidson Kempner's appearance as a docket participant in 25-cv-504 and (b) Piccadilly being the target of Apple's third-party discovery on funding and valuation, is direct record evidence that a third-party funder holds a secured financial position in this estate. A security interest in the patents themselves — not merely a funding contract — is the most aggressive form of this arrangement.

Verdict

NPE — high confidence.

The chain is: sole inventor Ping-Jung Yang → MYW Semitech, LLC by assignment recorded 2024-12-27 (100% of assignor's interest) → secured to Piccadilly Patent Funding LLC as security holder by Patent Security Agreement recorded 2025-03-25 — with the first family complaint (MYW Semitech v. Apple, 1:25-cv-00504-RGA, D. Del.) dated 2025-04-24. That is two strong signals (shell-entity transfer to a licensing-only LLC, and a pre-litigation transfer within ~4–5 months of suit) plus a documented third-party funder with a security interest in the patents themselves. The only thing materially missing is the reel/frame and correspondent-of-record for both entries, which I could not retrieve — and which is the field most likely to convert signal #3 from "unclear" to "present." Caveat carried forward from the prior section and still unresolved: US 9,615,453 is not among the three patents actually asserted against Apple; its NPE posture is established at the family level via the "Semitech Patents" definition, not by direct assertion.

Verify the reel/frame entries and correspondents at: https://assignmentcenter.uspto.gov/ (search patent number 9615453); cross-check legal events at https://patents.google.com/patent/US9615453/en.

I'll search for the specific patent and its cited prior art references.

Let me try to retrieve the actual citation lists from the patent record.

Let me try to access the front-page "References Cited" directly.

Prior-Art Analysis — US 9,615,453 B2

Important correction to the task framing / date conflict: the system header for this session gives today's date as 2026-09-26, while the task instruction states April 26, 2026. I am flagging the discrepancy rather than silently picking one; nothing in the substance below turns on it.

Number-integrity note (per the strict rule): I searched the literal string 9615453 / 9,615,453. All recurring hits that are not this patent were excluded and are named here so they are not mistaken for it: WO 96/15453 A1 (Spectral Diagnostics, protein assay membranes) and FR 9615453 — neither is US 9,615,453. No auto-correction was applied.

1. Bottom line on the specific request

I could not retrieve the authoritative examiner-cited "(56) References Cited" list for US 9,615,453. I want to be explicit about this rather than dress up field knowledge as a citation list:

What I can state with confidence from the supplied authoritative text: the specification body contains no in-text patent citations at all. I read the full description as provided — there is exactly one cross-reference (provisional US 61/705,649, filed 2012-09-26) and no "U.S. Pat. No. ___" passages anywhere. So the entire prior-art list, if any, is examiner-added on the front page, not applicant-supplied. That is a real finding, and it means any §102 argument runs through the examiner's search, whose scope is visible in the classification: H01L23/15 (substrate characterised by material — ceramic/glass), H01L51/0096, H01G4/129 (glass-ceramic), H05K1/09, and H10W/H10K77 glass-substrate groups. Prior-art keywords listed on the record: layer, substrate, chip, metal, conductors.

Therefore Section 3 below is labelled by evidence class. Do not treat any of it as the patent's own citation list.

2. §102 framework I applied (this matters for every mapping below)

US 14/036,256 was filed 2013-09-25, i.e. after 2013-03-16, so the AIA first-inventor-to-file provisions of 35 U.S.C. §102(a)(1)/(a)(2) govern — unless the 2012-09-26 provisional supports the subject matter, in which case the effective filing date is 2012-09-26 for supported claims, and pre-AIA §102(a)/(b)/(e) becomes the operative test for that subject matter. Both were checked. Practical consequence:

• Art published before 2012-09-26 → §102(a)(1)/pre-AIA §102(a)/(b) — anticipatory if it discloses every element.
• Art filed before 2012-09-26 but published after → §102(a)(2)/pre-AIA §102(e) — requires "effectively filed" analysis.
• Art published 2012-09-26 → 2013-09-25 is §102(a)(1) only as to the AIA filing date, not the provisional date; it cannot be §102(a)(2) prior art unless its own filing predates the effective date.

Claim elements that any §102 reference must meet (claim 1 as verified in the prior section):

The second independent claim (identified in the prior section as the multi-layer-conductor variant, likely claim 11 but that number remains an inference, not verified) swaps E5 for a first metal layer + second metal layer coated on the first metal layer, and recites a third metal layer at the first surface.

Analytical read: E5 (Cu-W alloy) and E6 (direct glass-to-metal contact) are the narrowest and most likely novelty anchors. E1–E4 and E7–E8 describe a generically planarised glass interposer and are the elements most likely to have been met by cited art.

3. Prior art identified, graded by evidence class

Class A — Surfaced on the record as citation-linked to '453 (moderate confidence; source table unconfirmed)

My "US9615453B2 cited-by" searches returned Google Patents pages whose tables contain the exact row US9615453B2 (en) | 2012-09-26 | 2017-04-04 | Ping-Jung Yang | Method for fabricating glass substrate package. Because those host patents (e.g. US 7,749,900, issued 2010) predate '453, a row naming '453 in such a table can only be a forward/citation link from '453, i.e. it is evidence '453 cites them. I could not confirm which table each row sat in (Google uses similar layouts for "Patent Citations," "Cited By," and "Similar Documents"), so I grade this moderate confidence.

A1. US 7,749,900 B2

• Full citation: US 7,749,900 B2, "Glass core substrate for integrated circuit devices and method of making same" (Intel Corporation). Related sibling: US 8,207,453 B2, same title (Intel), issued 2012-06-26; and US 2011/0147059 A1 (Qing Ma), "Substrate for integrated circuit devices including multi-layer glass core and methods of making the same," published 2011-06-23.
• Dates: US 7,749,900 issued 2010-07-06 (pre-dates both the '453 provisional and filing).
• Description: Glass core for an IC package substrate; through-holes/vias formed in the glass and metallised, with build-up dielectric and metal layers laminated on the glass faces and connected to the vias. This is the closest structural analogue to claim 1 in the record.
• §102 mapping: Potentially meets E1, E2, E3, E4, E6 (if glass is metallised directly), E7, E8 of claim 1. It likely does not disclose E5 (copper-tungsten alloy) — Intel's glass-core work uses Cu metallisation/conductive pastes. Net effect: a §102 weak, §103 strong reference for claim 1; a primary reference for the second independent claim (the multi-metal-layer variant) if its build-up stack is read onto it.

A2. US 7,337,540 B2

• Full citation: US 7,337,540 B2, "Method of manufacturing a structure body bonding with a glass substrate and semiconductor substrate." (Surfaced via its Google Patents page listing the '453 row.)
• Dates: Issued 2008-05-27 (pre-AIA; well before the effective date).
• Description: Bonding/joining of a semiconductor structure body to a glass substrate, with associated via/interconnect formation and planarisation — i.e. the glass-substrate-plus-semiconductor-structure assembly concept.
• §102 mapping: Directed at the method aspects the '453 title implies but its verified claims do not cover. As to claim 1, it is unlikely to meet E5; could bear on E1, E7, E8 and on dependent claims drawn to glass composition/thickness. Low anticipation value; possible §103 value for the dielectric/metal-layer-on-glass element.

A3. DE 10034507 C1

• Full citation: DE 10034507 C1, "Verfahren zum Erzeugen von Mikrostrukturen auf Glas- oder Kunststoffsubstraten nach der Heißformtechnologie…" (process for producing microstructures on glass or plastics substrates by hot-embossing, plus the associated moulding tool). Surfaced via its Google Patents page listing the '453 row.
• Dates: German patent, priority filed 2000-07-14; granted ~2002 (I did not verify the grant date in these results — flagging as unverified).
• Description: Hot-embossing/hot-forming of microstructures into a glass (or polymer) substrate using a shaping tool. This maps directly onto the '453 specification's FIGS. 20a–20i embossing process for forming the metal layer on the glass, and onto the general "shape the glass while soft" teaching.
• §102 mapping: On its face an embossing-of-glass reference; it would not meet E2/E3/E4/E5. Its realistic role is §103 combined with A1 for the embossing embodiment, or as §102 art against any dependent claim directed purely to an embossed/patterned glass surface.

Class B — Surfaced as technically on-point in the same field (not citation-linked to '453 in any result)

B1. US 2013/0050155 A1 — Qualcomm MEMS Technologies

• Full citation: Petersen, K.E.; Shenoy, R.V.; Black, J.P.; Burns, D.W.; Ganapathi, S.K.; Stephanou, P.J.; Buchan, N.I., "Glass as a substrate material and a final package for MEMS and IC devices," US 2013/0050155 A1, App. 13/221,744, Assignee Qualcomm MEMS Technologies Inc.
• Dates: Filed 2011-08-30; published 2013-02-28.
• Description: Glass substrate + cover glass encapsulating MEMS/IC devices, with bond pads in electrical communication with the encapsulated device and configured for a flexible connector. URL: https://patentimages.storage.googleapis.com/4a/cc/6b/8089deb5348a71/US20130050155A1.pdf
• §102 mapping: Published 2013-02-28 — after the 2012-09-26 provisional but before the 2013-09-25 filing. Its own filing (2011-08-30) predates the '453 effective date, so under §102(a)(2) it is available prior art only if the '453 claims are denied the provisional date; under the AIA analysis with the provisional date it is not §102(a)(2) art. Relevance: E1, E7 (glass substrate carrying electrical connections) — no through-glass conductor of Cu-W, no coplanar through-via teaching. §103 value only.

B2. WO 2013/033124 A1 — through-package-via glass interposer

• Full citation: WO 2013/033124 A1, PCT/US2012/052719, "Through-package-via (TPV) structures on inorganic interposer and methods for fabricating same." Search report completed 2013-01-31.
• Dates: International filing in 2012 (PCT/US2012/052719 — I did not verify the exact filing day); published 2013-03-07.
• Description (from the search report text): explicit process flow "Provide Glass Substrate → Form Through-Glass Via Holes → Metallize Sidewalls of Through-Glass Via Holes → Dice Glass Substrate to Form Multiple Packages," with sandblasting + wet etch (HF) via formation, conformal metal seed layer, plating, and joining the plated glass substrate to a device substrate. This is the glass-interposer/TGV art closest in subject matter to the '453 core concept. (Assignee not confirmed in my results; the technical content matches the Georgia Tech glass-interposer programme — flagging that attribution as inferred.)
• §102 mapping: Its international filing predates 2012-09-26, so as an effectively-filed-earlier disclosure it is a serious §102(a)(2) candidate against E1, E2, E3, E4, E7, E8 — if a US national-stage publication/patent exists with the requisites and if the claims are denied the provisional date. Its vias are lined/plated (seed layer + plating), which cuts against E6 (direct glass-to-conductor contact) and it does not disclose E5. Its strongest bite is on the method subject matter the '453 title claims but the verified claims do not.

B3. US 7,084,073 B2 — via hole through a glass wafer

• Full citation: "Method of forming a via hole through a glass wafer" — US 7,084,073 B2 (Korean priority, Appl. No. 2002-70121, filed 2002-11-12). Source: https://uspto.report/patent/grant/7,084,073
• Dates: Korean priority 2002-11-12; granted 2006-08-01 (I did not independently confirm the grant date — flagged).
• Description: Two-step via formation through a glass wafer — rough "preliminary via" by sandblasting (with a film-resist mask), then wet etch (HF) to yield a smooth via free of undercut and micro-cracks, for MEMS packaging.
• §102 mapping: E2 (a through-glass hole) only — a hole, not a conductor; no coplanarity of a conductor face, no Cu-W, no dielectric/metal layer. §102 weak/anticipation of nothing in claim 1; §103 value as the TGV-formation background.

Class C — Field-identified as the most likely actual novelty-defeating art (analysis, not retrieval)

This is my technical judgement, clearly labelled as inference: because E5 (copper-tungsten alloy) is the narrow element, the art the examiner most plausibly relied on is hermetic glass-to-metal seal / feedthrough art and co-fired ceramic (HTCC/LTCC) metallisation art — both of which use Cu-W and W pins/vias precisely for CTE-matching to glass (the same CTE rationale the '453 specification gives for glass generally). Cu-W with 10–30 wt% Cu is a commodity material for glass-sealed pins. I did not retrieve a specific such reference from the '453 record and do not name a number for it, because I cannot verify one. If a §102 attack on claim 1 is the goal, this is the class to pull first.

4. Consolidated §102 mapping

No reference identified in this analysis discloses E5 (copper-tungsten alloy conductor) together with E6 (direct glass-to-conductor contact) and E3/E4 (coplanar through-conductor faces). On the evidence I have, claim 1 is not anticipated by any reference I surfaced; the exposure is §103-type and depends on whether an unretrieved Cu-W-feedthrough or co-fired-tungsten-via reference was cited. That is exactly the gap created by not having the (56) list.

5. What remains unverified — and how to close it

1. The (56) reference list itself — the single most important missing datum. Get it from the front page of the granted patent image PDF (the (56) field sits in the left column of page 1), or via USPTO Patent Public Search / PatentCenter "References Cited" for US 9,615,453, or Global Dossier.
2. Claim 11 identity — the prior section's inference that the second independent claim is claim 11 is still an inference; nothing in these results changed it. The (56) list should be mapped per-claim only once the claim numbers are confirmed.
3. Whether '453 is actually the patent-in-suit / IPR patent — unchanged from the prior section: the MYW Semitech v. Apple, No. 1:25-cv-00504 (D. Del.) and TSMC v. MYW Semitech, IPR2026-00067 dockets involve multiple MYW patents; I have no confirmation that '453 is the asserted/challenged patent. If it is, the IPR petition's ground section is itself a curated prior-art list and would be the single fastest way to obtain the strongest art.
4. Grade-A rows (A1–A3) should be confirmed against the actual front page before being cited as "the references cited in '453." I will not upgrade that grade without the (56) field.

Standing caveat: no CAFC activity for '453 or MYW Semitech was found (absence of evidence, not proof of none), and the earlier-flagged on-face unit anomalies in the specification (e.g. "thermal resistance layer 8 has a thickness between 0.05 and 1 meter") remain uncorrected and are not relied on above.

I'll ground this analysis in actual retrievable prior-art records before opining, since the "Prior Art section" referenced in your instruction was not included in the material passed to me.

Let me pull the front-page cited art and a few more anchor references.

Analyst Note on Inputs — Read First

The "Prior Art section of this page" you instructed me to use was not included in the material passed to me. The message contains (a) the full text of US 9,615,453 from Google Patents, and (b) the previously generated summary sections. There is no "Prior Art" heading, no IDS/cited-references list, and no examiner citation table.

Rather than fabricate a prior-art set to fit the instruction, I ran targeted searches and assembled a set of references I could actually verify (title, assignee, date, URL). Everything below is grounded in those results. Where I could only see a title/date and not the specification, I say so. If you can paste the actual Prior Art section, I will re-run the mapping against those specific references instead.

Also flagging one contradiction-style issue with the prior summary: the earlier section states the second independent claim's number is "likely claim 11" but unverified. That uncertainty persists and materially affects the grounds below, because the second independent claim has a different conductor limitation (multilayer metal rather than CuW alloy) and therefore needs a different primary-art mapping. Two of my grounds are split accordingly.

1. Governing Framework

• Statutory basis: 35 U.S.C. § 103. The application was filed 2013-09-25, i.e., after the AIA's 2013-03-16 cutoff, and the earliest claim of priority is provisional US 61/705,649 (2012-09-26). Because both the AIA and pre-AIA § 102/103 windows are arguable here, I analyze under AIA § 103 with a 2012-09-26 critical date, and I flag date-sensitive references separately.
• Standard: KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007) — a combination is obvious where the elements are known, the combination is a predictable variation of prior-art elements, and there is an articulable reason (design need, market pressure, known problem) to combine. MPEP § 2143 (motivation) and § 2144.01 (obvious to try, finite number of predictable solutions) govern the motivation analysis.
• Claim construction posture: the two limitations that will decide this case are (i) conductor "substantially coplanar" with both glass surfaces, and (ii) "the solid glass core is directly contact with the conductor" (i.e., a linerless / no-intervening-layer limitation). Everything else in claim 1 is conventional. The CuW alloy is the third distinct hook.

2. Verified Reference Set

The R5 date problem, stated plainly: if US 9,615,453's claims are fully supported by the 2012-09-26 provisional, then R5 (effective filing 2012-11-13) is not § 102(a)(2) art and cannot be used in an IPR petition. R3, however, is dated 2012-01-01 and is available either way. A petitioner should lead with R3, not R5.

3. Element-by-Element Mapping — Claim 1

Conclusion on claim 1: every limitation except the CuW selection and the linerless contact is squarely in the glass-interposer art as of 2012. The § 103 question reduces to whether selecting CuW over W or Cu as a via fill, and omitting a liner, would have been obvious to a PHOSITA.

4. Grounds of Rejection

Ground 1 — Claim 1: R1 (or R3) in view of R4, further in view of R6/R7

Primary: R1 (Intel glass core substrate) or, more safely on dates, R3 (Tummala 2012 IMAPS paper disclosing an ultra-thin glass interposer with through-package-vias, double-sided RDL metallization, and 0.5–5 µm wiring — see the retrieved abstract and the Georgia Tech companion paper).
Secondary: R4 (tungsten-filled through-glass via, 2011).
Tertiary: R6 and/or R7 (CuW is a glass-/ceramic-CTE-matched metallization).

Motivation (articulable, specific, and traceable to the art itself):

1. A recognized problem in the same field. Copper vias in glass are a known thermal-expansion mismatch problem. The '453 specification itself frames the field this way: "pre-testability, post-bonding visual inspection, and TCE (Temperature Coefficient of Expansion) matching to avoid solder bump fatigue are still challenges." A reference may be combined with the patentee's own admitted problem statement.
2. A finite, predictable solution set. R4 had already put tungsten inside a through-glass via. R6 teaches CuW as a material whose "rate of thermal expansion [is] substantially the same as some glass and ceramic materials." R7 teaches CuW specifically as the metal joined to ceramics using a glass bonding agent, expressly because it is "relatively wettable with glass" and because CTE differentials must stay ≤5×10⁻⁶/°C. Adding Cu to W to tune CTE downward while raising conductivity is the textbook two-way-option optimization of a composite; the '453 spec's own list of ratios (Cu:W from 30:70 to 90:10) is precisely the conventional CuW composition range.
3. Reasonable expectation of success. R4 demonstrates W can be filled into glass vias; R6/R7 establish that CuW is a glass-compatible, CTE-tailorable metallization. No new chemistry or new process class is required.
4. Linerless contact is a known, permitted design choice. The glass-interposer literature (e.g., the background discussion in US 2022/0059436, citing demonstrated "glass wafers with metalized tungsten-plugged vias, which provide electrical connection and hermetic sealing") shows linerless/plugged glass vias. Omitting a process step that the specification elsewhere lists as optional is a classic KSR "predictable variation."

Ground 2 — Claim 1: R3/R5 in view of R4, R6, R7 and R2

If the claim's RDL element ("first dielectric layer and a first metal layer") is argued to be a point of novelty, R2 (Intel multi-layer glass core substrate) and R3 (Georgia Tech double-side RDL on glass) independently disclose building a dielectric-plus-metal stack on a glass core surface and routing it to the through-vias. The motivation is the entire purpose of an interposer: fan-out from fine-pitch vias to package-level I/O.

Ground 3 — Claim 1 (alternative): co-fired ceramic / LTCC tungsten-via art as the primary

Pre-existing co-fired-glass/ceramic substrate art (tungsten metallization co-fired in glass-ceramic green tape, with CTE-matched refractory fills) plus the glass-core-substrate teachings of R1/R2 renders the claimed combination obvious from the other direction. The WITEC reference (R6) is a powder-metallurgy CuW product with glass-matching expansion, and the '453 specification's own fabrication flow — liquid glass poured around pre-positioned metal traces (FIGS. 6–15), then CMP — is structurally analogous to green-tape co-firing with via fills.

Ground 4 — Second independent claim (multilayer conductor): Grounds 1–3 + R8

The second independent claim recites the conductor as "a first metal layer and a second metal layer coated the first metal layer" (with a "third metal layer" at the surface). That is the standard barrier/seed + fill architecture of via metallization, which R8 (nucleation/barrier layer before tungsten fill) and the PVD/CVD barrier-layer teachings recited in the '453's own spec (Ti, TiN, TiW, Ta, TaN, Ni, NiV seed layers) disclose. Motivation: adhesion and diffusion-barrier performance — the reason such bilayers exist in every via process.

5. Dependent Claims (probable scope, description-derived)

The previously generated section flagged claims 12–20 as unretrieved; I did not retrieve them either. If they narrow to covering layers 6a (Ni/Zn/Ti/Ta/Ag/Cr or an oxide anti-oxidation layer) and 6b (adhesion layer), those are individually taught by R8 and by ordinary surface-finish practice (the literature review of US 2022/0059436 recites the identical stack: "adhesion layer is usually tungsten, titanium, chromium… metal layer is copper, gold, or silver… diffusion barrier is nickel, tungsten, titanium, or chromium… passivation layer is nickel or gold"). Cross-sectional shape (circular/square/oblong/flat) and dimensional ranges are result-effective-variable selections with no unexpected result asserted. I will not assert what these claims say until the claim text is produced.

6. Where the Patent Has Genuine Defensive Traction

An honest § 103 analysis must state the patentee's best arguments, because a well-drafted petition has to pre-empt them:

1. Teaching away from linerless glass-via contact. Much of the glass-interposer literature (notably the Georgia Tech TPV work, R5) deliberately teaches a polymer liner or stress-relief layer between glass and conductor to manage CTE mismatch. R5's abstract expressly recites "a stress relief barrier" and "a metallization seed layer." Under In re Gurley-type reasoning, a patentee may argue the art taught that an intervening layer was necessary, which would weigh against "directly contact."
2. CuW as a powder-metallurgy composite, not a via fill. R6/R7 describe CuW as a sintered/infiltrated bulk product and a braze-joined member — not as a damascene or plug via fill. A patentee will argue non-analogous-use or inoperability. The counter is R4 (W actually filled into glass vias) plus the general recognition that Cu lowers W's resistivity and processing temperature.
3. Date vulnerability of the best references. As noted, R5 (Georgia Tech '934/'555) is the closest structural reference but is dated after the '453 priority date. If the petitioner relies on it and the patentee perfects the provisional priority, that ground collapses.
4. Priority-date attack as a two-edged sword. A petitioner may argue the '453's claims are not entitled to the 2012-09-26 provisional (which would pull R5 into the prior art under § 102(a)(1)/(a)(2)), but that same argument risks conceding a 2013-09-25 effective date and inviting the patentee to rely on the provisional's disclosure for other purposes.

7. Bottom Line

• On the record I could verify, claim 1 is vulnerable under § 103 as a combination of (i) a glass core/interposer with through-vias and build-up dielectric+metal layers on its surfaces (R1, R2, R3), (ii) a tungsten-filled through-glass via (R4), and (iii) copper-tungsten as a glass-/ceramic-CTE-matched metallization (R6, R7) — with the "substantially coplanar" and "directly contact" limitations met by standard CMP planarization and by omitting an optional liner.
• The motivation is unusually strong here because the patent's own Background names CTE matching as an outstanding problem, and because the prior art supplies the exact CTE-matching rationale for CuW-to-glass.
• The weakest link in the petitioner's case is not motivation but dates and the depth of my reference set: my two structurally closest references (R5, and possibly R8) require date verification, and I read R1/R2 at the title level only.

8. Explicit Uncertainty Flags

1. The Prior Art section you referenced was not supplied. This analysis substitutes independently verified references. Re-run it if you produce that section.
2. I did not read the full specifications of R1 (US 8,207,453) or R2 (US 2011/0147059). I verified title, assignee, and dates only. The coplanarity and dielectric-layer mappings to those two are inferred from their titles/assignees and must be confirmed against the text before being asserted in a petition.
3. R5 (US 9,275,934 / US 2013/0119555) carries a Nov. 13, 2012 date that is after the '453 priority date and is therefore date-contingent, not a safe primary reference.
4. The second independent claim's number remains unverified (previously advanced as "likely claim 11"). Ground 4 is keyed to that claim's language, not to a confirmed number.
5. Claims 12–20 (or the full 20-claim set) were not retrieved. Section 5 is description-derived, not claim-verified.
6. Litigation/PTAB context (MYW Semitech, LLC v. [Apple Inc.](/litigations/by-plaintiff/Apple%20Inc.), No. 1:25-cv-00504 (D. Del.); TSMC v. MYW Semitech, IPR2026-00067) is noted only as context; it remains unconfirmed that US 9,615,453 is the patent-in-suit or the challenged patent in those proceedings.

Sources: US9615453B2 — Google Patents · US2014/0085842 — FreePatentsOnline · Tummala, IMAPS 2012DPC · Sun et al., IEEE EPTC 2011 · US 9,275,934 B2 · DE 3881030 T2 — WITEC · US 2003/0221854 A1 — Sumitomo · US 8,778,797 B2 · US 2020/0098551 A1 · US 2022/0059436 A1