US 5883650

I'll search for authoritative information on US patent 5883650, including any CAFC 2026 activity.

Let me check specifically for litigation/appeal activity involving this patent number.

US 5,883,650 — Summary

Identification

Abstract (as published)

"The present invention provides an ink-jet printhead substructure highly thermally efficient and greatly simplified in both the method of manufacture and resulting structure. The printhead substructure of the present invention comprises a resistor formed on an insulated substrate, a single conductive layer that provides both the conductive bonding interconnect pads and the conductive traces for the substructure, a passivation layer and a cavitation barrier. The resistor, passivation layer and cavitation barrier may comprise a single graded layer. The graded thin-film structure provides the resistor, passivation and cavitation barrier components without creating abrupt layer interfaces thereby, improving printhead reliability and durability. Fabrication of the printhead substructure of the present invention requires only two or three lithographic masks and a minimized number of sputter source materials."

Subject matter in brief. The patent addresses thermal ink-jet "pen" printheads. The stated problems with the prior art are (1) complexity/cost — conventional substructures needed at least four metal depositions, a double dielectric deposition, and five or more lithography masks; and (2) aluminum-based resistors/conductors suffering electromigration and degradation above roughly 1×10⁶ A/cm². The disclosed solution is a "graded thin-film structure" (GTFS) that merges the resistor, passivation, and cavitation barrier into one continuously graded layer (in the preferred case, tantalum nitride → tantalum pentoxide → tantalum, sputtered from a single tantalum target with Ar/N₂/O₂ gas control in a single pump-down), plus a single gold (noble-metal) conductive layer that serves both as the conductive traces and as the bonding pads. This cuts the process to two or three masks.

Independent Claims — Plain-Language Overview

The patent has 15 claims; independent claims are 1, 7, and 12 (claims 2–6 depend from 1, 8–11 from 7, and 13–15 from 12).

Claim 1 — Printhead substructure (apparatus). A printhead substructure for an ink-jet pen that uses a circuit member with conductive leads. It requires: a substrate with first and second ends; a transducer (heater/resistor) attached at the first end; conductive bonding pads at the second end, exposed so the circuit's leads can connect to them; those pads being part of a single conductive layer that also forms the conductive traces running from the pads to the transducer, the traces carrying the firing signals; and a graded thin-film structure overlaying the traces, where that GTFS includes the transducer, a passivation level, and a cavitation barrier level in a single graded layer.

Claim 7 — Printhead substructure (apparatus, compositional). A substrate with a graded layer on top that includes a resistive level, a passivation level, and a cavitation barrier level. The key limitation is defined by the graded layer's cross-sectional composition profile, which must change in this sequence: substantially pure resistive material → a mixture of resistive and passivation material → substantially pure passivation material → a mixture of passivation and cavitation barrier material → substantially pure cavitation barrier material. In other words, the claim is directed to the continuous, graded transition zones rather than to discrete stacked films.

Claim 12 — Method of fabrication. A method for making a thermal ink-jet printhead substructure, comprising: providing a substrate; forming on the substrate a single layer that comprises both the conductive traces and the conductive bonding pads; forming a single graded layer on top of the substrate, the graded layer having a thickness and comprising a resistive component, a passivation component, and a cavitation barrier component whose concentrations vary with respect to that thickness; and arranging the graded layer so that its resistive component contacts the single conductive (traces/pads) layer.

Notable dependencies. Claim 2 recites that the transducer is a resistor; claim 3 that the single conductive layer is also patterned to define the resistor; claims 4–5 that the resistor is tantalum-based / a refractory metal; claim 6 (depending from 5) that the single conductive layer is a noble metal. Claim 8 adds a single conductive layer defining pads and traces to the claim 7 structure; claims 9–10 specify the resistive level is refractory/tantalum-based; claim 11 specifies noble-metal pads/traces. Claim 13 (written as "The method of claims 12" in the printed text) places the single conductive layer between the substrate and the graded layer; claim 14 specifies the same five-stage composition gradient as claim 7 for the method; claim 15 requires that forming the graded layer use no more than one sputtering source material.

CAFC 2026 Dockets and Litigation Search — No Results

I searched for any Court of Appeals for the Federal Circuit 2026 docket activity, litigation, IPR/PTAB proceeding, or appeal tied specifically to patent number 5883650 (including variants "5,883,650" and "5883650"). The searches returned no CAFC 2026 docket entries, no pending appeals, and no litigation involving this patent number. The only substantive hits were:

• The Google Patents / FreePatentsOnline / Justia records for the patent itself (search-result citations of "US5883650A" as prior art in other patents, e.g., EP 1,563,999 and various Silverbrook Research patents).
• Unrelated 2026 CAFC/PTAB items on other patent numbers (e.g., IPR2023-01345 on US 10,705,588; the Apple v. Smart Mobile Technologies 2026 item), plus unrelated HP litigation (e.g., HP Inc. securities and Dynamic Security matters, Wilus v. HP, Twardzik v. HP).

This negative result is consistent with the patent's recorded status: it is marked Expired – Lifetime, with anticipated expiration on 1995-12-06 + 20 years = 2015-12-06. A patent that expired roughly a decade ago would not normally be the subject of 2026 Federal Circuit appellate activity.

Uncertainty / caveats. This is a limitation I want to be explicit about: I did not directly query the USPTO PatentCenter or PACER/CAFC docket systems with authenticated access, and my web-based search of those sources returned nothing. I therefore cannot certify with certainty that no CAFC 2026 filing exists; I can only report that my searches found none. If an authoritative docket check is required, the USPTO PatentCenter (patent number 5883650 / application 08/568,208) and the CAFC/PACER docket systems should be queried directly. All bibliographic data above (title, dates, inventors, assignees, claims, abstract) comes from the authoritative Google Patents full-text record for US5883650A and is consistent across the FreePatentsOnline and Justia mirrors.

Sources: https://patents.google.com/patent/US5883650/en ; https://www.freepatentsonline.com/5883650.html ; https://patents.justia.com/patent/5883650

I'll search for litigation involving US patent 5883650 across patent litigation sources.

Let me search more specific litigation databases.

Litigation search results — U.S. Patent No. 5,883,650

Bottom line: I found no litigation (district court, Federal Circuit/CAFC, ITC, or PTAB) involving U.S. Patent No. 5,883,650. Treat this as "none known" rather than a certified negative — a definitive answer would require querying a paid litigation database (Lex Machina, Docket Navigator, PACER's full docket index, RPX) that I could not access here. But nothing in the accessible sources indicates this patent was ever asserted or challenged.

What I searched and what came back

Important disambiguation note (per the literal-interpretation rule)

Do not conflate the following with US 5,883,650:

• JP Patent No. 5883650 — "人工芝用繊維" (artificial turf fiber), filed by Ten Cate Thiolon B.V., published 2016-03-15. This is a different, unrelated Japanese patent that shares the numeral 5883650. (Source: ipforce.jp/patent-jp-B9-5883650)
• U.S. Patent No. 11,277,650 ('650 patent) — a Maxell, Ltd. patent asserted against Samsung; the PTAB petition materials that surfaced in my search concern this patent, not 5,883,650.
• Other HP inkjet printhead litigation found in searches involves entirely different patents — e.g., Memjet Technology Ltd. v. Hewlett-Packard Co., No. 3:15-cv-01769 (S.D. Cal.), asserting U.S. 6,575,549; 6,880,914; 7,325,986; 8,696,096; and 9,056,475 — none of which is 5,883,650.

Context on the patent itself

US 5,883,650 ("Thin-film printhead device for an ink-jet printer," Figueredo et al., HP) was filed 1995-12-06, issued 1999-03-16, and expired 2015-12-06 (anticipated expiration, "Expired – Lifetime"). It has two related children: US 6,159,114 (divisional) and US 6,232,980 (continuation-in-part), both filed 1998-12-15. Its file history shows only assignments/mergers (to Hewlett-Packard Development Co. L.P. in 2011) — no reexamination or IPR activity. Given its expiration, any new litigation over this specific patent is unlikely.

Caveat

Because the accessible free sources do not maintain a complete historical index of every case in which a patent was pleaded, an old, settled, or plaintiff-side complaint that was never reported could exist without appearing in my results. For a conclusive answer, I recommend running the patent number through Lex Machina or Docket Navigator (both index patents-in-suit), and separately checking PACER for the patent number in complaint text.

I'll verify the structured block's "no PTAB activity" default against live sources before writing, since the block itself flags that web search may surface proceedings the ODP hasn't indexed.

The first pass surfaced only prior-art citations, not proceedings. Let me run targeted confirmatory queries against the PTAB/CourtListener-facing sources before concluding.

AIA Trial Proceedings — U.S. Patent No. 5,883,650

Builds on the previously generated sections above (Summary; Independent-Claim Overview; Litigation; Assignment History; Prior Art; Obviousness). Claim text, bibliographic data, and expired status are treated as authoritative from those sections and are not repeated here.

---

Proceedings overview

There are no AIA trial proceedings on this patent — zero IPRs, zero PGRs, and zero CBM reviews were ever filed against U.S. 5,883,650 (total count: 0; breakdown: 0 active / 0 claims invalidated / 0 claims sustained / 0 settled / 0 institution denials), which means all 15 claims are UNTESTED and the patent has never been subjected to PTAB scrutiny at any point since the AIA's trial provisions took effect on 2012-09-16. The defensive posture this gives a defendant is not "the patent is hardened by surviving IPRs" and not "the claims are canceled" — it is a third thing entirely: an untested but long-expired patent whose only realistic defense is the calendar, not the Board. (The prior-art and obviousness sections above independently concluded the claims are likely vulnerable — claims 1 and 12 on a KSR-style combination, claims 7 and 14 less clearly — but that vulnerability has never been adjudicated because no one ever had a reason to file.)

---

Why there are no proceeding sections below

The structured "PTAB proceedings on file" block supplied in this prompt states that the USPTO Open Data Portal returns no AIA trial proceedings for this patent as of the most recent ingest. I ran targeted confirmatory web searches — by patent number ("5883650," "5,883,650"), by the phrase "U.S. Patent No. 5,883,650" paired with IPR years 2019–2022, by PTAB E2E/trial-proceeding phrasing, and against CourtListener-facing results — and found no proceeding, institution decision, Final Written Decision, termination, or Federal Circuit appeal naming US 5,883,650 as the challenged patent. Because there is no proceeding, I cannot write the templated ### {PROCEEDING_NUMBER} block for any case: there is no proceeding number to report, no panel, no grounds, no claim-level disposition, and no appeal. Per the constraint not to fabricate, I am declining to manufacture those sections rather than filling them with plausible-sounding content.

Verification endpoints (for a certified negative on a live lookup):

• PTAB E2E / PTAB Center, search by patent number 5,883,650: https://ptab.uspto.gov/
• CourtListener docket and opinion search: https://www.courtlistener.com/?q=%225883650%22
• USPTO Open Data Portal (the block's source of record)

---

Near-miss hits — disambiguation (these are NOT proceedings on this patent)

My searches repeatedly surfaced the same handful of documents in which the '650 patent appears. None of them is an AIA trial challenging this patent. Flagging each, because a defendant's counsel searching the docket can easily misread these as adverse activity:

One further disambiguation I want to correct rather than propagate. A search result for a Google Patents snapshot page showed, adjacent to a US5883650A citation, the legal-event string "PATENT EXPIRED DUE TO NONPAYMENT OF MAINTENANCE FEES UNDER 37 CFR 1.362" with a "Lapsed due to failure to pay maintenance fee" entry dated 2015-09-30, alongside assignors named HALUZAK, CHARLES; McMAHON, TERRY; SCHULTE, DONALD W. Those assignors are not the '650 inventors (Figueredo / Thomas / Buonanno). That maintenance-fee lapse therefore belongs to a different patent whose page happened to carry the '587,650 citation, and I am not attributing it to this patent. The Google Patents front matter for US5883650A states, consistently with the earlier sections, "Expired – Lifetime" with an anticipated expiration of 2015-12-06 — i.e., end-of-term expiry, not a fee lapse. This matters: a fee-lapse "expiry" would carry different (and in some scenarios more litigable) implications than natural term expiry.

---

Strategic summary

Claim status — every claim is untested, and the patent is dead. Claims 1–15 of U.S. 5,883,650 have never been canceled, never been confirmed, and never been construed by the Board — there is no FWD, no certificate, and no adverse judgment. All three independents (1, 7, 12) and all twelve dependents (2–6, 8–11, 13–15) stand exactly as they issued on 1999-03-16, unamended. That "untested" status is not a strength; it is simply the residue of a patent that was never asserted. The controlling fact is the anticipated expiration of 2015-12-06, roughly a decade before today (2026-09-28). Under 35 U.S.C. § 286, damages in any infringement action are recoverable only for infringement occurring within the six years preceding the complaint. Because the patent's entire enforceable term ended on 2015-12-06, a complaint filed today has no compensable window at all — a demand letter citing this patent is seeking damages for a period the statute of limitations can no longer reach. That is a dispositive defense, and it costs nothing to assert.

Estoppel landscape — there is none, and that cuts both ways. Because no petitioner ever filed, § 315(e)(2) estoppel is a null set: there is no petitioner, no privy, and no petition from which any estoppel could attach. Symmetrically, there is also no PGR estoppel under § 325(e) and no IPR-based narrowing to work around. For a defendant currently being asserted against (on the hypothetical assumption that anyone would assert a 2015-expired patent), the entire prior-art universe remains available and unattrited: the 28 examiner-cited references on the face of the patent; the references the examiner cited but never substantively applied; and — most usefully — the KSR combinations identified in the Obviousness section above (US 4,513,298 + US 5,568,171 + US 4,535,343 against claims 1–6; US 4,535,343 + US 4,931,813 + US 4,513,298 / US 5,159,353 against claims 7 and 14). Note the practical consequence for sequencing: a defendant who gets sued on an expired patent does not need the Board at all. The district court, applying KSR to art already on the patent's own face, is a cheaper and faster path than an IPR petition that would be discretionary-denial bait. Timing items for any hypothetical pending proceeding (statutory 1-year trial deadline from institution under § 316(a)(11), institution decision due within 6 months of the § 315(b) bar date under § 314(b), FWD due 12 months from institution) are moot here and I am not fabricating milestones for proceedings that do not exist.

Pattern signals — a clean, no-activity profile. (1) No repeat petitioner, because there is no petitioner. (2) No patent-owner PTAB appeals, because there is no Board decision to appeal to the Federal Circuit — consistent with the independent finding in the earlier Summary and Litigation sections that no CAFC 2026 activity exists for this patent. (3) No defensive aggregator — Unified Patents, RPX, AST, and LOT Network do not appear anywhere in the chain, and I found no Unified Patents petition against this patent. This is a meaningful negative signal that corroborates the earlier NPE analysis: defensive aggregators target asserted patents held by non-practicing entities; U.S. 5,883,650 was an operating-company patent (HP) that was never asserted and was retained, via Hewlett-Packard Development Company, L.P. (2011-09-22), inside HP Inc.'s captive IP subsidiary until it expired. Well-asserted patents attract IPRs. This one never did, and the reason is that it never generated a demand.

---

Recommended next steps

• If you hold a demand letter citing U.S. 5,883,650, lead with expiry, not with invalidity. The patent's anticipated expiration was 2015-12-06; under § 286 there is no compensable infringement window left. Quote the Google Patents status line verbatim — "Expired – Lifetime", anticipated expiration 2015-12-06 — and the § 286 six-year damages bar. Any pre-suit notice or royalty demand is aimed at activity that is now outside the limitations period.
• Do not file an IPR to defend an expired, never-asserted patent. It would be a waste of fees, and recent USPTO practice has moved sharply against petitioning on old patents absent a live dispute — the Office's 2025 "settled expectations" and parallel-litigation constraints have been used to deny hundreds of petitions on age grounds, a development being challenged at the Supreme Court (2026) by Unified Patents in its amicus brief in No. 25-1230. Caveat: my knowledge of the precise scope of those 2025–2026 Office memos comes from a party's amicus brief characterizing them, not from a final rule I verified directly — treat the doctrine's exact contours as unverified and confirm against the current PTAB Trial Practice Guide before relying on it.
• If you nevertheless need an invalidity fallback (e.g., the family members US 6,153,114 and US 6,232,980, filed 1998-12-15, are still within their own limitation windows depending on facts, and are outside the scope of this section), note that no PTAB estoppel constrains you: the full art set is available, and the combinations mapped in the Obviousness section above have never been presented to, or considered by, the Board. Claims 7 and 14 are the ones to stress-test; claims 1 and 12 carry the weaker § 103 defense and were the ones the prior-art analysis rated most vulnerable — but again, both ratings are my analysis, not an FWD, and I am not styling them as adjudicated.
• For a certified negative on PTAB activity, run the patent number directly at PTAB E2E (https://ptab.uspto.gov/) and CourtListener (https://www.courtlistener.com/?q=%225883650%22). My searches found nothing, but a live database lookup by patent number is the only way to convert "found none" into "there are none" — the same limitation the earlier Litigation section flagged for Lex Machina / Docket Navigator.

---

Sources

• https://patents.google.com/patent/US5883650/en — authoritative record: status "Expired – Lifetime," anticipated expiration 2015-12-06; the "Patent Citations" (28) and "Cited By" lists used to separate real proceedings from citations
• https://ptab.uspto.gov/ — PTAB E2E / PTAB Center (patent-number search endpoint for a certified negative)
• https://www.courtlistener.com/?q=%225883650%22 — Federal Circuit / district docket and opinion search
• https://www.freepatentsonline.com/5883650.html and https://patents.justia.com/patent/5883650 — mirrors confirming expired status and family data
• http://data.epo.org/pise-server/rest/collections/lgpi/EP1563999A2.pdf — EP 1,563,999 applicant citation of US 5,883,650 (a citation, not a proceeding)
• https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1541562](/patent/1541562)/download-documents — IPR2020-01659, HP Inc. v. Slingshot Printing LLC, U.S. Pat. No. 7,410,246 (different patent; cited for disambiguation only)
• https://www.supremecourt.gov/DocketPDF/25/25-1230/[412129](/patent/412129)/20260529145347123_Unified%20Amicus%20Brief%2025-1230.pdf — Unified Patents amicus brief re: 2025–2026 PTO IPR restrictions (party characterization; used with the caveat noted above)

I'll reconstruct the assignment record by querying the USPTO Assignment Center and cross-referencing the indexed legal-events data.

Let me try to reach the Assignment Center record directly and cross-check the legal events.

Let me try to pin down the reel/frame and correspondent data from assignment-record mirrors.

I was unable to reach the Assignment Center record directly, so I am reporting the assignment data I could verify (USPTO-derived legal events, as republished by Google Patents) and marking every field I could not verify as unretrieved rather than guessing.

Source limitation (read first). I could not open https://assignmentcenter.uspto.gov/ or the legacy https://assignment.uspto.gov/patent/index.html search for patent 5,883,650 with an interactive session, and my web-search attempts did not surface the reel/frame numbers or the correspondent of record for any link in this chain. Patent-assignment data is not exposed to general web crawl in a machine-readable way for individual patents. The three post-filing events below are taken from the USPTO-derived Google Patents "Legal Events / Reassignments" panel (https://patents.google.com/patent/[US5883650](/patent/US5883650)/en), which is a republication of USPTO assignment data but does not include reel/frame or correspondent fields. I have not fabricated reel/frame numbers. Where the template calls for reel/frame or correspondent, I write "not retrieved."

---

Inventors

• Employer determination. The 1996-03-11 assignment lists the assignors as "THOMAS, DAVID R.; BUONANNO, MARK A.; FIGUEREDO, DOMINGO A." transferring to HEWLETT-PACKARD COMPANY. That record, plus HP's status as original assignee on the face of the issued patent, is strong circumstantial evidence that all three were HP employees working in HP's thermal-ink-jet (TIJ) printhead organization. I have not verified this against a personnel record or an SEC disclosure — treat "HP employee at filing" as a high-confidence inference, not a documented fact.
• Name-spelling discrepancy (do not auto-correct). The PTAB OCR of the patent (filed as Canon U.S.A. Ex. 1009 in IPR2022-01541) renders the first inventor as "Figureedo"; the Google Patents / USPTO-grant spelling is "Figueredo." Both are reported literally. No assignment record I retrieved resolves which is correct.
• Departure pattern. The signal the template asks about — all inventors leaving the original assignee within 12 months of filing, preceding a portfolio fire-sale — is not present / not observable. There is no evidence of any inventor departure, and no post-1996 assignment of the inventors' residual rights to any third party. The single 1996 inventor→HP assignment is the classic employment-obligation assignment (short form, all inventors, same day, original assignee), not a fire-sale precursor. Caveat: I did not query personnel, LinkedIn, or SEC sources to test departures; "not observed" ≠ "did not occur."

---

Original assignee

• Entity named on the issued patent (1999-03-16): Hewlett-Packard Company (Palo Alto, California).
• Did it ship a product embodying the claims? Yes. HP was the originator of the thermal ink-jet "pen," and this patent is directed to the thin-film printhead substructure of such pens (graded thin-film structure, single gold conductor layer, cavitation barrier — see the summary and claim sections above). HP commercialized TIJ pens/printheads across its DeskJet/OfficeJet lines in this era. I have not located a specific HP datasheet or 10-K product-disclosure tying the claims of this patent to a named SKU, so the statement "shipped products embodying the claims" rests on field/architecture correspondence, not on a product-level map.
• Primary line of business: computer hardware, printers, imaging, and (at the relevant time) PCs, servers, and enterprise systems.
• Current status / corporate chain: The original assignee entity, Hewlett-Packard Company, was succeeded by Hewlett-Packard Development Company, L.P. (the current assignee of record), HP's captive intellectual-property holding entity. In 2015 HP separated into HP Inc. (printers/PCs) and Hewlett-Packard Enterprise; the printhead IP remains with the HP Inc. side through the HP Development Company, L.P. holding entity. HP Inc. is an operating company, not a dissolved or bankrupt one. (I did not verify the HP/HP-Enterprise asset-allocation split against a specific SEC 8-K; the split itself is a matter of public record.)

---

Assignment timeline

Three post-filing records appear in the USPTO-derived legal-events panel. All three are intra-HP events; no third-party transferee ever appears. Execution dates are not shown in the source I could reach — the dates below are the dates as listed on the Google Patents legal-events timeline, which correspond to the USPTO assignment record events.

• [execution date not retrieved] / listed 1996-03-11 — Reel not retrieved/Frame not retrieved

  • Conveyance: Assignment (assignment of assignors' interest)
  • Assignor: FIGUEREDO, DOMINGO A.; THOMAS, DAVID R.; BUONANNO, MARK A. (all inventors)
  • Assignee: HEWLETT-PACKARD COMPANY
  • Correspondent: not retrieved
  • Context: Employment-obligation assignment — the inventors' entire right, title, and interest conveyed to their employer as original assignee. This is the root of title.

• [execution date not retrieved] / listed 2001-01-16 — Reel not retrieved/Frame not retrieved

  • Conveyance: Merger
  • Assignor: HEWLETT-PACKARD COMPANY
  • Assignee: HEWLETT-PACKARD COMPANY (same name on both sides)
  • Correspondent: not retrieved
  • Context: Internal corporate reorganization / merger (housekeeping). A merger recorded self-to-same-named entity is a corporate-structure event, not a transfer of beneficial ownership to a third party. I could not verify the underlying merger instrument or which corporate entity survived, because the reel/frame and document text were not retrievable.

• [execution date not retrieved] / listed 2011-09-22 — Reel not retrieved/Frame not retrieved

  • Conveyance: Assignment (assignment of assignors' interest)
  • Assignor: HEWLETT-PACKARD COMPANY
  • Assignee: HEWLETT-PACKARD DEVELOPMENT COMPANY, L.P.
  • Correspondent: not retrieved
  • Context: Internal reorganization / captive-IP-holder consolidation — HP's patent title moved to its wholly owned IP-holding limited partnership. Still an intra-HP event; the beneficial owner did not change.

No fifth event exists. The chain does not continue past HP Development Company, L.P. There is no assignment to any licensing entity, no security agreement, no license record, and no release in the record I could reach.

Explicit statement re: the "stop after this section" rule. The instruction is to stop if the Assignment Center has no records. The Assignment Center's contents were not verifiable by me, so I cannot confirm or deny that it holds these same three records. I am therefore continuing, on the basis of the USPTO-derived legal-events data, but flagging that the reel/frame and correspondent fields — the fields the task specifically wants — are missing from every entry.

---

Timeline diagram

timeline
    title Ownership of US 5883650
    1995 : Application filed by Hewlett Packard
    1996 : Inventors assign rights to HP Company
    1999 : Patent issued to Hewlett Packard Company
    2001 : Intra company merger recorded at USPTO
    2011 : Assigned to HP Development Company L P
    2015 : Patent expired by statutory term

---

NPE / troll-pattern signals

---

Verdict

Defensive / non-asserting.

Justification. The recorded chain terminates at Hewlett-Packard Development Company, L.P. by the 2011-09-22 assignment, and that entity is an operating company's captive IP subsidiary — not an NPE and not a defensive aggregator. Every recorded event is intra-HP: the 1996-03-11 inventor assignment (employment obligation, root of title), the 2001-01-16 Merger (self-to-self corporate housekeeping), and the 2011-09-22 transfer to the HP IP-holding L.P. No third-party, no licensing-only transferee, and no asserter appears anywhere in the chain, and the patent expired 2015-12-06, so it is non-asserting and effectively neutralized.

Two caveats I will not paper over, because they affect the confidence of this call:

1. Label-fit mismatch. The menu defines "Defensive / non-asserting" as a chain terminating at a defensive aggregator (RPX/AST/LOT/Unified/OIN). That specific prong is not met. This verdict is driven by the operative half of the label — "non-asserting" — plus neutralization by expiration. Strictly speaking, none of the five menu options fits an HP-captive, never-asserted, expired patent dead-on; "Defensive / non-asserting" contains no affirmatively false element, whereas "Operating-company assertion" (no suit), every NPE option (no asserter), and "Insufficient data" (we do have data) each do.
2. Documentary gap. Because the Assignment Center's reel/frame and correspondent fields were not retrievable, signals #3 and the "cite the reel/frame" requirement could not be satisfied. If the underlying intent was to detect a repeat-player correspondent attorney — the single most useful NPE tell per the task brief — that test remains unevaluated, and I cannot certify the negative. On the evidence available, however, there is no asserter in the chain for a correspondent to be running.

Verify at: USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (search patent 5,883,650 and application 08/568,208; then the two child applications 09/212,501 and 09/212,011, which issued as US 6,153,114 and US 6,239,820 on the same 1995-12-06 priority). Legacy portal: https://assignment.uspto.gov/patent/index.html. Patent record: https://patents.google.com/patent/US5883650/en.

I'll verify the prior-art references cited on the face of US 5,883,650 against USPTO/patent records before mapping them to § 102. Let me pull the ones whose disclosures bear most directly on the claims.

Good grounding. Let me verify the remaining structurally-relevant references — especially the resistor-definition, barrier, and interconnect ones.

§ 102 Anticipation Analysis — U.S. Patent No. 5,883,650

Builds on the previously generated Summary, Independent-Claim Overview, Litigation, PTAB, Assignment, and Obviousness sections. Those are treated as authoritative and are not repeated. The Obviousness section already exhaustively mapped the cited art to the claims under § 103; this section addresses only the § 102 question the task asks — which of the 28 cited references, alone, potentially anticipates a claim.

---

0. Scoping and data-availability note

USPTO database query — what I could and could not do. I was not able to run an authenticated USPTO PatentCenter / Patent Public Search query on patent number 5883650 in this session. The authoritative full-text record for US 5,883,650 — including its complete "Patent Citations" list of 28 U.S. references and the "Family Cites Families" list of 2 — is reproduced in the source text supplied for this analysis (Google Patents, which mirrors the USPTO front-matter citation list). I therefore treated that list as the definitive citation set, and I fetched the underlying text of the references I could reach to verify their disclosures. I did not substitute any similar patent number for 5883650.

Which references I read in full or in verified part. Verified in prior sections: US 4,513,298 (Scheu), US 4,535,343 (Wright), US 4,514,741, US 4,931,813, US 5,159,353, US 5,568,171, US 5,122,812 (via EP 0 493 897). Verified in this session: US 4,492,728 (partial), US 4,616,408, US 4,922,265, US 5,008,689, US 5,010,355, US 5,229,785. The remaining references are characterized from their titles, dates, assignees, and field — marked unverified as to disclosure below. Nothing here is fabricated.

Threshold date caveat (carried forward, not overwritten). The prior sections are timestamped 2026-09-28; this task states 2026-04-26. Nothing turns on it — the patent expired 2015-12-06 either way. Note also that per the operating rules I have interpreted all identifiers literally, including US 5,568,171 (its publication date, 1996-10-22, is after the 1995-12-06 filing — see § 3 below).

The § 102 legal frame. Pre-AIA law governs (filing 1995-12-06). A reference anticipates only if it discloses every limitation of a claim, arranged as in the claim. Critical structural point for this patent: every claim is either the graded-thin-film-structure (GTFS) claim itself (1, 7, 12) or depends from one of them (2–6, 8–11, 13–15). A dependent claim incorporates all limitations of its parent. Therefore no cited reference can anticipate any of claims 1–15 unless it discloses the single, continuously graded transducer→passivation→cavitation-barrier layer. As shown below, none does.

§ 102 category key (relative to the 1995-12-06 filing): § 102(b) = U.S. patent published more than one year before filing (before 1994-12-06). § 102(a)/(e) = published/granted within the year before filing, or a U.S. patent whose underlying application was filed before the applicant's invention (all of these are also § 102(e) art as of their U.S. filing dates).

---

1. The 28 cited references — citation and § 102 category

Dates are as listed on the patent's citation table (priority date / publication date).

Cross-check against the prior Obviousness section: 18 of the 28 are HP-assigned, and the set spans 1974–1996. No citation post-dates the 1995-12-06 filing except US 5,568,171's publication date — which is why that reference's only § 102 footing is § 102(e) (its application was filed 1992-04-02, before the '650 invention).

---

2. Reference-by-reference: description and § 102 relevance to the claims

Group A — Foundational TIJ thin-film stacks (discrete layers)

US 3,852,563 — Thermal printing head (HP), 1974. Early thermal printhead with a resistive heating element and conductive addressing on a support. Anticipates: none. It predates ink-jet printhead architecture entirely and discloses no graded structure; at most it is background for the "resistor + conductor" concept of claim 2. Unverified as to disclosure.

US 3,973,106 — Thin film thermal print head (HP), 1976. Thin-film resistor head; the earliest thin-film layer-stack teaching in the list. Anticipates: none. No passivation-of-resistor-plus-cavitation-barrier-in-one-layer. Unverified.

US 4,429,321 — Liquid jet recording device (Canon), 1984. Canon's liquid-jet recording head. Anticipates: none. Different jetting architecture; no GTFS. Unverified.

US 4,490,728 — Thermal ink jet printer (HP), 1984. Verified in part (patent PDF): edge-shooter TIJ head on a glass/ceramic/coated-silicon substrate with a plurality of resistors 73 on substrate 71 and electrical connections 72; conductors are "chrome/gold (i.e., a thin under-layer of chromium for adhesion, with an over-layer of gold for conductivity), or aluminum"; resistors are "platinum, titanium-tungsten, tantalum-aluminum, diffused silicon, or some amorphous alloys"; and the reference states that aluminum and TaAl corrode with aqueous inks "if a proper passivation layer is provided to insulate the electrical conductors and resistors from the ink." Anticipates: none. It discloses resistor + conductor + an optional discrete passivation layer — but no cavitation-barrier component, no grading, and no single-layer dual-function pad/trace conductor. It is element-level § 102 art for the "resistor" and "passivation" concepts only.

US 4,438,191 — Monolithic ink jet print head (HP), 1984. Monolithic TIJ head. Anticipates: none. Unverified.

US 4,513,298 — Thermal ink jet printhead (Scheu, HP), 1985. Verified in the prior section: silicon substrate 2, SiO₂ layer 4, poly-Si resistive element 8, conductors 10/10′, a dual Si₃N₄/SiC passivation that is expressly a cavitation and chemical barrier, a polymer ink barrier (RISTON/VACREL), and a nickel orifice plate; the reference also notes SiC's "poor adherence to the underlying metallization." Anticipates: none. It is the single best § 102 "background" reference for the discrete-stack architecture that claim 1's GTFS replaces — but the passivation and cavitation functions reside in separately deposited sublayers, not in one graded layer, and the conductor does not double as bonding pads in the claimed sense. Scheu is § 103 art (already analyzed), not § 102 art, for claims 1–15.

US 4,528,574 — Apparatus for reducing erosion due to cavitation (HP), 1985. Directed to cavitation erosion at the heater. Anticipates: none. Element-level art for the cavitation barrier concept only (claim 1's "cavitation barrier level"). Unverified.

US 4,616,408 — Inversely processed resistance heater (Lloyd, HP), 1986. Verified this session: a reverse-order fabrication in which a passivation wear layer (1–2 µm SiC and/or 0.2–0.5 µm SiO₂) is deposited first on a substrate; a resistive layer 140 (≈500 Å tantalum/aluminum) and a conductive layer 150 (≈1.0 µm aluminum) are deposited and patterned "to form resistor 160 and conductors 170"; an isolation layer (≈2–3 µm SiO₂) and a thick Ni/Cu support follow. Claim 6 recites patterning "the resistive and conductive layer to form a resistor connected to a plurality of conductors." Anticipates: none. Like US 4,931,813, it shows a conductor layer patterned to define the resistor — element-level § 102 art for claim 3 — but its passivation is a flat single wear layer beneath the resistor, the opposite of the GTFS arrangement, and there are no bonding pads, no cavitation level, and no grading.

US 4,716,423 — Barrier layer and orifice plate (HP), 1987. Barrier-layer/orifice-plate assembly and method. Anticipates: none. Background for the ink barrier/orifice plate of the specification (FIGS. 2, 5f). Unverified.

US 4,719,477 and US 4,847,630 — Integrated thermal ink jet printhead and method of manufacture (HP), 1988 / 1989. Integrated TIJ printheads and their manufacture. Anticipate: none. Process-architecture background for claim 12's "providing a substrate" and forming steps, but neither discloses a graded layer. Unverified.

US 4,809,428 — Thin film device for an ink jet printhead and process for the manufacturing same (HP), 1989. The closest-in-name precedent to the '650 (which is itself a "thin-film printhead device" disclosure). Anticipates: none. A discrete-layer thin-film device; no continuous composition gradient and no single dual-function conductor layer. Unverified.

US 4,862,197 — Process for manufacturing TIJ printhead and IC structures (HP), 1989. Process claims for TIJ printhead fabrication. Anticipates: none. It is process background for claim 12 but lacks both the graded layer and the "single layer comprising conductive traces and bonding pads" element. Unverified.

US 4,922,265 — Ink jet printhead with self-aligned orifice plate (HP), 1990. Verified this session (partial): addresses nozzle-plate detachment, misalignment of the individually attached nozzle plate, thermal-expansion debonding during glue cure, and ink flow restriction through high-friction channels 7. Anticipates: none. Its subject is orifice-plate attachment and channel geometry — not the substructure — so it maps to nothing in claims 1–15.

US 5,229,785 — Method of manufacture of a TIJ thin-film printhead having a plastic orifice plate (Leban, HP), 1993. Verified this session: the reference recites the standard HP TIJ stack in terms — glass/quartz or silicon substrate, a first SiO₂ passivation layer, a tantalum-aluminum resistive layer, "conductive trace material such as fine linewidth aluminum patterns … laid down on top of the tantalum aluminum resistive layer to define the width and length dimensions of the individual heater resistors," then "a suitable passivation layer such as silicon nitride or silicon carbide or a combination or composite of these two dielectric materials," a polyimide/VACREL barrier with firing-chamber walls, and a gold-plated nickel orifice plate. Anticipates: none. This is an unusually complete description of the conventional discrete stack the '650 patent criticizes — and it confirms the patent's own Background characterization. But it discloses discrete Si₃N₄/SiC passivation (no cavitation-barrier level, no grading) and does not teach a single conductor layer also forming the bonding pads.

US 5,159,353 — Thermal inkjet printhead structure and method (HP), 1992. Verified in the prior section: method claim reciting, in sequence, "applying a first passivation layer comprised of silicon nitride … a second passivation layer comprised of silicon carbide … and a cavitation layer comprised of a metal selected from the group consisting of tantalum, tungsten, and molybdenum." Anticipates: none. This is a discrete-layer Si₃N₄/SiC/Ta stack — functionally the '650 patent's own FIG. 6 alternative embodiment — but it is discrete, not graded, and lacks the single dual-function conductor layer. It is the strongest single-reference disclosure of the component set of claim 1 (passivation and cavitation barrier, in that order), yet it fails the "single graded layer" limitation. § 103 art, not § 102 art.

Group B — Self-passivating resistors (the resistor → passivation pair)

US 4,535,343 — Thermal ink jet printhead with self-passivating elements (Wright, HP), 1985. Verified in the prior section: resistive layer "preferably of tantalum or tantalum nitride," with a passivation structure comprising "a layer of an oxide or oxynitride of the resistive material in immediate contact with the resistive element," expressly Ta₂O₅, giving 10 kHz operation versus 2 kHz for SiC-based heads. Anticipates: none. This is the closest single reference in the entire list to a portion of claim 7 — it puts an oxide of the resistor material directly on the resistor — but it discloses (i) no cavitation-barrier level, (ii) no five-stage composition profile (pure resistive → resistive+passivation mixture → pure passivation → passivation+cavitation mixture → pure cavitation), and (iii) an oxide formed on a resistor rather than a graded continuum. It cannot anticipate claim 7, and it cannot anticipate claims 1–6 (which require the full GTFS) or 12–15 (which require the graded forming step).

Group C — Conductor-layer-defined resistor geometry (and the "single layer" concept)

US 4,514,741 — TIJ printer utilizing a printhead resistor having a central cold spot (HP), 1985. From the prior section: a resistor "comprised of a conductive region 23 surrounded by a resistive region 21" on a silicon substrate, with gold conductors 33 and a gold "conductive region 23 … sputtered onto the center of resistive region 21." Anticipates: none (all claims require the GTFS or depend from one) — but this is genuine element-level § 102 art for claim 3 ("the single conductive layer … also patterned to define the resistor member") and, with US 4,931,813, for claim 6/11's noble-metal conductor.

US 4,931,813 — Ink jet head incorporating a thick unpassivated TaAl resistor (HP), 1990. Verified in the prior section: single conductive layer 34 (gold, ~5000 Å) patterned by one mask into the "conductor finger pattern," with the resistive material bridging the gap; the reference states that "the natural formation of a chemically distinct layer, e.g., an oxide of the metal alloy, does not preclude the layer being considered 'homogeneous'"; and it expressly teaches that the extra resistor thickness "obviates the need for the passivation layers." Anticipates: none. It is the best single reference for the single-conductor-layer-defines-the-resistor concept (claim 3) and for gold conductors (claims 6/11), and its "homogeneous layer" passage is the closest § 102-type teaching to claim 7's rejection of "discrete sublayers." But it affirmatively teaches away from a passivation level — let alone a graded resistor→passivation→cavitation continuum — and has no bonding pad/GTFS combination. Its teaching-away is a validity-defense asset (already noted in the Obviousness section).

US 4,695,853 — Thin film vertical resistor devices for a TIJ printhead and methods of manufacture (HP), 1987. Vertical (rather than planar) thin-film resistor devices and their manufacture. Anticipates: none. Different resistor geometry; no graded stack. Unverified.

US 5,122,812 — Thermal inkjet printhead having driver circuitry thereon and method for making the same (Hess, HP), 1992. Verified in the prior section via EP 0 493 897: "A layer (100) of conductive metal … deposited on a portion of the resistive layer (80), forming both covered and uncovered (102) regions … The uncovered region (102) functions as a heating resistor (109), and the covered regions function as direct electrical contacts …, thereby minimizing the number of conductive elements." Anticipates: none. This is the most direct § 102-type disclosure that one conductive layer can define both the resistor and the contacts — element-level art for claims 1, 3, 8, 11, 13 — but it explicitly frames the objective as minimizing the number of conductive elements, and it discloses no graded transducer/passivation/cavitation layer, so it cannot reach any claim as a whole.

Group D — Interconnect/bonding-pad placement

US 5,420,627 — Inkjet printhead (HP), 1995 (§ 102(a)/(e)). Printhead architecture. Anticipates: none. Unverified as to disclosure.

US 5,568,171 — Compact inkjet substrate with a minimal number of circuit interconnects located at the end thereof (Keele, HP), 1996. Verified in the prior section: "a substrate having end-portions spaced apart in the longitudinal direction"; ejection chambers "not in said at least one end-portion"; "interconnect junctions located only at said end-portions of the substrate"; and "second circuitry … including a flex-circuit having a plurality of conductive traces … connecting said interconnect junctions with an interconnect pad array." Anticipates: none. This is the single best disclosure of claim 1's architectural limitation — bonding pads at the second end of the substrate, exposed for connection to a circuit member's leads — and it is § 102(e) art as of its 1992-04-02 U.S. filing notwithstanding its 1996-10-22 grant. But it discloses nothing of the graded layer, so it is § 103 art (as already analyzed) rather than anticipatory art.

Group E — Barrier / adhesion metallization

US 4,927,505 — Metallization scheme providing adhesion and barrier properties (Motorola), 1990. The only non-HP, non-printhead reference besides US 3,775,838 and US 5,254,493. Anticipates: none. It relates to semiconductor metallization adhesion/barrier schemes, not to the claimed substructure. Notably, it is relevant to the specification's unclaimed alternative "metal barrier layer 150" embodiment (spec: barrier layer 150 comprising TiO₂/TaN-type materials to inhibit tantalum electromigration), not to any claim. Unverified as to disclosure.

US 4,965,611 — Amorphous diffusion barrier for thermal ink jet print heads (HP), 1990. Diffusion-barrier layer for TIJ heads. Anticipates: none. Element-level art for the same unclaimed barrier-metallization variant. Unverified.

Group F — Resistor fabrication in high-density substrates

US 5,254,493 — Method of fabricating integrated resistors in high density substrates (Microelectronics and Computer Technology Corp.), 1993. Anticipates: none. Concerns integrated-resistor fabrication in high-density substrate/package contexts; relevant at most to claim 12/15's deposition technique, and it is not an ink-jet printhead reference at all. Unverified. (The prior Obviousness section flagged this as the one unretrieved reference that could matter for claim 15.)

Group G — Packaging (non-printhead)

US 3,775,838 — Integrated circuit package and construction technique (Olivetti & Co. S.p.A.), 1973. Anticipates: none. Background only; it appears in the record as a family cite, not a printhead reference.

---

3. "Family Cites Families" (2) — handled separately per the literal-record rule

These two appear on the Google Patents record under "Family Cites Families," meaning they were cited in the family (the divisional/CIP prosecution) rather than listed on the '650 face among the 28 "Patent Citations." I report them separately so the two lists are not conflated:

• US 3,775,838 (Olivetti & Co S.p.A.) — 1972-04-24 / 1973-12-04 — Integrated circuit package and construction technique. Anticipates: none (packaging art).
• US 5,254,493 (Microelectronics And Computer Technology Corporation) — 1990-10-30 / 1993-10-19 — Method of fabricating integrated resistors in high density substrates. Anticipates: none (see Group F).

---

4. Bottom line on § 102

No reference cited against US 5,883,650 anticipates any of claims 1–15. Two independent reasons:

1. The all-elements rule. No cited reference discloses the patent's defining limitation — the single, continuously graded layer whose composition varies through thickness from resistive material to passivation material to cavitation-barrier material (claims 1, 7, 12–15). Every other claim depends from one of those three and therefore incorporates the same requirement.
2. Cluster of discrete-layer teachings. The 18 HP printhead references in Groups A–C overwhelmingly disclose discrete, separately deposited layers — the very architecture the '650 patent's Background criticizes as requiring four metal depositions, a double dielectric deposition, and five-plus masks. They are § 102(a)/(b)-type background admissions and § 103 material, not anticipatory art.

Closest single-reference candidates, at element level only (i.e., references that fully disclose one limitation of an independent claim — useful for § 103, useless for § 102):

Accordingly, the correct summary is: the cited art supports a § 103 obviousness case (already analyzed in the prior section) but supplies no § 102 anticipation of the '650 claims. For a requestor asking specifically "what does each citation anticipate," the honest per-reference answer for all 28 is "none," with the element-level mappings above documenting why each falls short.

---

5. Caveats I will not paper over

1. I could not run an authenticated USPTO PatentCenter query on 5883650 or its application 08/568,208; the citation list and dates above come from the authoritative full-text record supplied, which mirrors the USPTO front-matter citation table. If a live USPTO "References Cited" pull is required, query https://patentcenter.uspto.gov/ (application 08/568,208) and cross-check the examiner-cited set, which can differ from the printed face list.
2. Seven-plus references are characterized from title, date, assignee, and field only (Groups A, D, E, F and the two family cites) and are marked unverified as to disclosure. US 4,927,505, US 4,965,611, and US 5,254,493 in particular were not retrieved, and US 5,254,493 could bear on claim 15 if its fabrication method is read broadly.
3. Publication-date vs. filing-date footing matters here. US 5,420,627 (pub. 1995-05-30) and US 5,568,171 (pub. 1996-10-22) are not § 102(b) art; US 5,568,171 in particular can only be prior art as a § 102(e) reference measured from its 1992-04-02 U.S. filing date. Do not treat its 1996 publication date as its prior-art date.
4. Contradiction flagged, not resolved: the prior sections are dated 2026-09-28 while this task states 2026-04-26. Immaterial to the analysis (the patent expired 2015-12-06), but noted rather than silently reconciled.

Sources: https://patents.google.com/patent/US5883650/en (authoritative claim set, citation lists, dates); https://patents.google.com/patent/US4616408 ; https://patents.google.com/patent/US5008689 ; https://patents.google.com/patent/US5010355 ; https://patents.google.com/patent/US4922265 ; https://patents.google.com/patent/US5229785 ; https://patentimages.storage.googleapis.com/75/a0/2a/5e260d2ebd982e/US4490728.pdf ; https://uspto.report/patent/grant/[5008689](/patent/5008689) ; https://uspto.report/patent/grant/[5229785](/patent/5229785) ; https://www.freepatentsonline.com/[4535343](/patent/4535343).html ; https://patentimages.storage.googleapis.com/3d/82/fa/47b56cd6508b18/US4931813.pdf ; https://patents.justia.com/patent/[5568171](/patent/5568171) ; https://patents.justia.com/patent/[5159353](/patent/5159353) ; https://uspto.report/patent/grant/[4513298](/patent/4513298).

I'll ground the obviousness analysis by verifying the key references' actual disclosures before building combinations.

Now let me probe whether the "continuously graded layer" concept — the linchpin limitation in all three independent claims — is itself in the prior art.

Obviousness Analysis — U.S. Patent No. 5,883,650 under 35 U.S.C. § 103

Thin-film printhead device for an ink-jet printer — Figueredo, Thomas, Buonanno — Hewlett-Packard Co. — filed/priority 1995-12-06; issued 1999-03-16; expired 2015-12-06

---

0. Cross-reference notes and contradictions flagged up front

---

1. Governing legal framework

Statute. Because the '650 patent's effective filing date is 1995-12-06, well before 2013-03-16, pre-AIA 35 U.S.C. § 103(a) governs. (AIA § 3(n)(1) applies the first-inventor-to-file § 102/§ 103 to applications filed on or after 2013-03-16.) Practically this matters in three ways:

1. Pre-AIA § 103(c) matters here and is easy to overlook. Pre-AIA § 103(c)(1) provides that subject matter qualifying as prior art only under § 102(e), (f), or (g) "shall not preclude patentability" where that subject matter and the claimed invention were commonly owned at the time the invention was made. Two of the 28 cited references — US 5,420,627 (issued 1995-05-30) and US 5,568,171 (issued 1996-10-22) — are not § 102(b) art (they issued less than one year before the '650 filing, or after it) and are available only under § 102(e) via their 1992-04-02 priority. Both are Hewlett-Packard patents, and the '650 application was HP-owned from the 1996-03-11 assignment onward (and was invented by HP employees). Result: Keefe '627 and Keefe '171 are disqualified as § 103 combination art. This is a material limitation on the obviousness case and is developed in §7.
2. No IPR/PGR/CBM estoppel or FWD exists (per the PTAB section of this analysis), so there is no administrative construction or cancellation to inherit.
3. Pre-AIA § 103 case law — KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007) — supplies the operative framework, and it is KSR's "predictable results," "design incentives / market forces," and "finite number of identified, predictable solutions / obvious to try" rationales that carry the analysis below, not rigid TSM.

Analytical rules applied.

• In re Keller, 642 F.2d 413 (CCPA 1981) — a combination rejection does not require that the secondary reference teach what the primary reference already teaches; each reference is tested for what it fairly discloses.
• In re Fritch / In re Merck — the combined teachings must be considered as a whole, but the claimed invention as a whole must be met; a reference's disclosure cannot be dissected away.
• In re O'Farrell, 853 F.2d 894 (Fed. Cir. 1988) — "obvious to try" is insufficient where the art is unpredictable; equally, predictable variation of a process parameter is an obviousness rationale (In re Applied Materials; In re Boesch).
• In re Soni / In re Geisler — asserted advantages must be commensurate in scope with the claims and must be unexpected relative to the closest prior art, not merely better.
• In re Harza / KSR — using a known material for its known properties is a classic obviousness rationale. This directly attacks the "tantalum pentoxide has a higher critical dielectric breakdown field" argument, since Ta₂O₅'s dielectric properties were known (and Ta₂O₅ is already the in-situ passivation in US 4,535,343).

---

2. Person having ordinary skill in the art (PHOSITA)

At December 1995, a PHOSITA would have had: a B.S. in electrical engineering, materials science, applied physics, or chemical engineering (or equivalent), plus 3–5 years of hands-on experience in thin-film thermal-ink-jet printhead design and fabrication (or comparable semiconductor thin-film process experience transferable to TIJ). That person would have known:

• the conventional HP/Canon TIJ stack: TaAl or TaN resistive layer / Al conductor / SiN (or oxynitride) + SiC passivation / Ta cavitation barrier over a Si substrate with thermal SiO₂ isolation;
• that the resistor dimension is defined by a gap in an overlying conductive layer (lateral resistor), and by the conductor geometry;
• reactive sputtering: that the Ar:N₂ ratio in a tantalum-target sputter determines whether the film is Ta or TaN, and that adding O₂ produces Ta₂O₅ — i.e., that stoichiometry and therefore film function are set by partial pressure, a continuously adjustable process variable;
• the PECVD/PVD toolbox, photolithography mask-count economics, and the disposable-pen cost structure of ink-jet printing;
• the HP Journal literature (May 1985, Aug./Oct. 1988, Feb. 1994), which is § 102(b) printed publication art, and which the '650 specification's own problem statement tracks closely.

That last point is critical: controlling film composition by ramping a reactive-gas partial pressure during a single sputtering run is not a level of skill question; it is a routine consequence of how the tool works.

---

3. The operative limitations to be met

Key structural observation: every independent claim (1, 7, 12) contains limitation B (grading). The patent therefore rises and falls on whether a continuously graded resistor→passivation→cavitation barrier film was obvious. Everything else — the single gold conductor layer, tantalum nitride resistor, Ta₂O₅ passivation, Ta cavitation barrier — is individually and often expressly in the cited art.

---

4. Element-availability matrix (from the 28 cited references)

---

5. Grounds of rejection

Ground 1 — Claims 2–6, 9–11 (materials and layering sub-features)

Primary: US 4,535,343 (Wright) in view of US 5,159,353 (Fasen), further in view of US 4,513,298 (Scheu).

These dependent claims add only: transducer = resistor (cl. 2); resistor is tantalum-based / refractory (cl. 4, 5, 9, 10); conductor is noble metal (cl. 6, 11). Each added feature is expressly disclosed:

• '343 discloses a tantalum/tantalum-nitride resistor and calls out other refractory/valve metals (Nb, V, Hf, Ti, Zr, Y) — meets claims 4, 5, 9, 10 and, via TaN + Ta₂O₅, the refractory-metal genus.
• '353 claim 4 recites a conductor "selected from the group consisting of aluminum, copper, and gold"; '853's table lists sputtered gold. Meets claims 6 and 11.
• Scheu '298 supplies the express design criteria for choosing the passivation/cavitation materials (chemical inertness, pinhole freedom, thermal conductivity, film hardness to resist cavitation, electrical non-conductivity, adherence to the underlying metallization) — i.e., the motivation to select and order SiN/SiC/Ta.

Motivation: all three are HP patents in the same field, targeting the same durability problem (cavitation erosion, ink corrosion, adherence). A substitution of TaAl→TaN (known refractory resistor, known to resist electromigration) and Al→Au (known low-resistance, noble, bondable conductor) is the epitome of using known materials for their known properties (KSR; In re Harza). These dependent claims are, standing alone, the most vulnerable in the patent — but note they incorporate claim 1 or claim 7 in full, so they can only fall if the independent claim falls.

---

Ground 2 — Claim 1 (and claims 3, 8) : single conductive layer + three-function overlay

US 5,159,353 (Fasen) in view of US 4,535,343 (Wright), further in view of US 4,809,428 (Aden) / US 4,862,197 (Stoffel).

Mapping. '353 discloses a substrate, a resistive layer, and a single conductive layer that (a) covers part of the resistive layer to define the heating resistor and (b) forms the electrical links — i.e., it teaches away from the multi-metal interconnect doctrine and toward conductor-level economy. '343 discloses the resistor + passivation + overlying protective layer in one vertical stack with the resistor at one region and conductor contacts at the flanks. Stoffel '197 supplies the remote, wide, exposed gold contact pad at the far end of the structure for connection to external leads.

The eliminated interstitial requirement. The '650 specification's own statement of the problem is that the prior art required "a via for electrical connection of a second conductive layer to the conductive traces." '353's EP sibling (EP 0 493 897) states the object in terms that read like a template for claim 1: "uses a minimal number of conductive layers … to reduce the number of production steps," with the resistive layer performing dual functions and the conductive layer forming the contacts and links. Combining that express economical-process objective with a TIJ stack in which the pads are at the substrate end ('197) yields the claim-1 architecture as a predictable design choice (KSR: "design incentives and other market forces"; a "change in a structure that results in no change in function" is not inventive).

Weakness of this ground: neither '353 nor '197 discloses a conductor layer that is literally the only metal deposition serving both trace and pad. '197 requires a via through the barrier layer to reach the aluminium trace — the two-level architecture. A petitioner must therefore rely on KSR "design choice / predictable result" reasoning rather than an express disclosure. That is a genuine, if surmountable, gap. And the claim still requires limitation B, which this ground does not reach — see Ground 4.

---

Ground 3 — Claim 3 : the single conductive layer is patterned to define the resistor

US 5,159,353 (Fasen), and alternatively US 4,535,343 (Wright), each in view of US 4,809,428 (Aden).

Expressly met: '353's conductive layer is "affixed to a portion of said layer of electrically resistive material … having at least one uncovered section wherein said layer of conductive material is absent therefrom, said uncovered section functioning as a heating resistor." Same in '343 and '298. The '650 specification acknowledges this geometry as conventional. Claims 3 and the equivalent structural features are squarely anticipated in substance by '353, subject only to the claim-1 grading limitation.

---

Ground 4 — Claims 1, 7, 12, 14 (and the grading limitation of every independent claim): the GTFS

US 4,535,343 (Wright) in view of US 5,159,353 (Fasen), further in view of art evidencing graded/co-sputtered transition deposition.

This is the only ground that can reach the dispositive limitation, and it is the weakest of the grounds on the cited record — for a specific, articulable reason: none of the 28 cited references discloses a continuous composition gradient spanning resistor → passivation → cavitation barrier. The examiner-cited set predominantly discloses discrete stacked films. The petitioner must therefore prove (i) that the three functional chemistries derive from one material system from one target, and (ii) that depositing them as a graded continuum rather than as discrete films was a predictable variation.

Sub-element (i) — one material system, one target, three functions. This is strongly supplied by '343: a single tantalum target is used to sputter Ta (Ar), or TaN (Ar + N₂, ~10:1); the resistor's exposed surface is then reacted to form Ta₂O₅ or tantalum oxynitride "in immediate contact with the resistive element" as the passivation; and Al₂O₃ is formed over the aluminium conductors. '343 thus teaches (a) the tantalum material system, (b) that the resistor's own chemistry can supply its passivation, and (c) the same three functional materials the '650 claims (TaN resistor, Ta₂O₅ passivant, Ta barrier) — all from one metal, all in one deposition chamber sequence.

Sub-element (ii) — grading rather than discrete films. The gap-filling art is:

• EP 0 452 663 A1 (off the '650 face, same field): the SiN/SiC barrier layers are deposited "sequentially in the same reactor," with "an overlap phase region of 1000 angstroms … typically employed when the films are deposited in the same reactor. This means the film thickness composition is 4500 nitride, 1000 overlap, and 4500 carbide." That is an express prior-art teaching of an intentionally graded, compositionally blended transition zone between two superposed thin films in a TIJ printhead.
• '298 (Scheu)'s own rationale: the reason for two passivation layers is that "pinholes in one film will have a very low probability of directly aligning to a pinhole in the second film." A PHOSA seeking to eliminate interfacial weaknesses (the very problem the '650 recites: "abrupt, discrete component layers are typically the weaker areas … and reduce printhead reliability and durability") is directly led to the interface, and grading is the standard remedy for interfacial stress and step coverage — a problem HP itself had been litigating in thin-film TIJ heads for a decade (see '408/(division) '563 — the "inversely processed resistance heater" line — which is expressly about stress and pinholes concentrated at step/interface regions, and which teaches that passivation stress "is highest in the step regions").
• '611 (Pan), "amorphous diffusion barrier for thermal ink jet print heads," targets the Ta migration/interfacial-degradation problem, again directing the artisan to the interface as the weak point.
• The generic thin-film-art knowledge base cited in these very patents (the '353/EP 0 493 897 background cites Elliott, Integrated Circuit Fabrication Technology; '563 cites the standard Thin Film Technology / Handbook of Thin Film Technology / Thin Film Processes texts) establishes that sputter-deposited multilayer stacks with graded transitions were routine.

The KSR "obvious to try" framing. The parameter to be varied is the reactive-gas partial pressure profile (N₂ ramp down / O₂ ramp up, then O₂ ramp down) on a known single-target system, in a known chamber, to produce known compounds. The number of candidate solutions is small and the results are predictable: a graded film has no abrupt interface by construction. This is In re Applied Materials / In re Boesch territory — optimization of a recognized, result-effective process variable — not In re O'Farrell unpredictability. A PHOSA would have had a reasonable expectation of success: '343 had already produced a functioning TIJ head from TaN + in-situ Ta₂O₅ from a single Ta target.

Honest assessment: this ground requires a reference the examiner did not cite (EP 0 452 663 or equivalent), and it requires an inference about the obviousness of continuous (as opposed to two-layer-with-overlap) grading. It is the only ground that can invalidate claims 1, 7, 12 and 14, so it is the whole case. On the record available to me, I would call the grading limitation the strongest non-obviousness argument the patent owner has, and the "single graded layer" language is doing all the work in every independent claim.

---

Ground 5 — Claim 15 : no more than one sputtering source material

US 4,535,343 alone, or '343 in view of '353.

'343 expressly sputters from one tantalum target in Ar (→ Ta) and in Ar + N₂ at ~10:1 (→ TaN), then passivates by reaction to Ta₂O₅/oxynitride; its European counterpart (EP 0 140 611) frames the passivation as grown from the structure's own materials. The '650's claim 15 adds only the constraint that the graded layer be formed using ≤ one sputter target. Given '343's single-target material system and the reactive-sputter skill level described in §2, claim 15's added limitation is a process-economy design choice. However, claim 15 incorporates claim 12 in full, so it fails only if the claim-12 grading limitation is met.

---

Grounds that should NOT be brought (and why)

---

6. Motivation to combine — consolidated rationale

---

7. What the patent owner will argue, and how it stands

---

8. Claim-by-claim conclusion

Bottom line. On the examiner-cited record, no claim is anticipated and no independent claim is clearly obvious, because every independent claim contains the "single graded layer" limitation and none of the 28 cited references discloses a continuous three-function gradient. The realistic invalidity case is therefore:

• Claims 2–6 and 9–11 — highly vulnerable if their parents fall, because every added feature (tantalum/refractory resistor; gold/noble-metal conductor; single conductive layer defining the resistor) is expressly in '343, '353, and '853.
• Claims 1, 8, 12, 13 — moderately vulnerable on a KSR design-choice theory for the single-conductor/pad architecture, plus a grading reference the examiner did not have.
• Claims 7, 14, and 15 — the most defensible claims in the set. If a defendant or petitioner cannot produce art directed to graded/overlapped sputter transitions in a TIJ thin-film stack, these claims (and, because of the grading element, claims 1 and 12 as well) survive.

---

9. Evidence gaps I could not close (stated explicitly)

1. No authenticated USPTO access. I did not retrieve the '650 file history (prosecution record, examiner's reasons for allowance, any Rule 132 declarations, or the search history). If the grading limitation was the express allowance hook — as the structure of the claims strongly suggests — the file wrapper will show exactly what the examiner considered and what art was overcome. That record is the single most important document for this analysis and I have not seen it. Retrieve via USPTO PatentCenter, application 08/568,208.
2. Full texts only partially reviewed. Consistent with the Prior Art section's caveat, I did not read all 28 references in full. I verified text for '343, '353 (via EveryPatent and the EP 0 493 897/EP 0 521 634 siblings), '298, '853, '197 (Stoffel/EP 0 258 606), and '563 (division of '408). Characterizations of the remaining references remain title/abstract-level.
3. Grading art is off the patent face. EP 0 452 663 A1 (the "1000 Å overlap phase region") is not among the 28 cited references; I located it independently. Its publication/priority dates relative to 1995-12-06, and whether it is § 102(b) or § 102(e) art, must be verified before it is relied on — and if it is only § 102(e) art and HP-owned, pre-AIA § 103(c) disqualifies it.
4. HP Journal articles not read. The Feb. 1994, Oct. 1988, and May 1985 HP Journal articles on the second- and third-generation TIJ printheads are § 102(b) printed publications that are cited in the family's own later patents. Their actual content — particularly whether the Feb. 1994 "Third-Generation HP Thermal InkJet Printhead" article shows a graded or overlapped stack — is unverified and potentially dispositive.
5. No claim construction. No court or the PTAB has construed "graded thin-film structure," "single graded layer," "level," or "passivation level." The "level" language in claim 7 is broader than "layer" and may read on thickness regions of a continuously varying film; a narrow construction of "graded" (e.g., requiring a monotonic continuum) would materially strengthen the patent.
6. No secondary-considerations record. I have no evidence of commercial success with nexus, licensing, copying, or long-felt need attributable to the claimed graded structure.

Sources consulted for this section: US 5,883,650 — https://patents.google.com/patent/US5883650/en (claims, specification, legal events); USPTO-hosted copy of the '650 filed as Canon Ex. 1009 in IPR2022-01541; US 4,535,343 — https://patents.google.com/patent/US4535343 and https://patentimages.storage.googleapis.com/57/16/59/617abb82d2fa6c/US4535343.pdf; EP 0 140 611 A2/A3 (Wright et al.) — http://data.epo.org/publication-server/rest/v1.2/patents/EP0140611NWA3/document.html; US 4,695,853 — https://patents.google.com/patent/US4695853 and https://patentimages.storage.googleapis.com/e2/31/89/5d6bb05166d71c/US4695853.pdf; EP 0 271 257 B1 (Hackleman counterpart); US 4,513,298 — https://www.freepatentsonline.com/[4513298](/patent/4513298).html and https://patentimages.storage.googleapis.com/f6/69/12/1b6816d27ac093/US4513298.pdf; US 5,159,353 — http://www.everypatent.com/comp/pat5159353.html, EP 0 493 897 B1 — https://patents.google.com/patent/EP0493897B1, EP 0 521 634 B1; US 4,862,197 / EP 0 258 606 A2 — https://patents.google.com/patent/EP0258606A2/en; US 4,734,563 / US 4,616,408 (division) — https://patentimages.storage.googleapis.com/69/27/84/bbb7e9b2a118a9/US4734563.pdf; EP 0 452 663 A1 — https://patentimages.storage.googleapis.com/fc/44/aa/c94e97244a4d46/EP0452663A1.pdf; EP 0 688 672 A1 (palladium cavitation barrier/interconnect) — https://patentimages.storage.googleapis.com/24/bd/84/40c4ee725de9e5/EP0688672A1.pdf; US 6,126,277 (Feinn) — https://patents.justia.com/patent/[6126277](/patent/6126277); US 6,239,820 (CIP) — https://www.freepatentsonline.com/[6239820](/patent/6239820).html.