US 4296309

Patent Overview — US4296309A

Title: Thermal head
Patent number: US4296309A
Assignee (original and current, as listed): Canon Inc.
Inventor(s): Akira Shinmi; Masahisa Fukui; Toshitami Hara; Yoshiaki Shirato; Yoshioki Hajimoto
Priority date: 1977-05-19 (claimed from JP52057936A)
Filing date: 1978-05-15
Publication (grant) date: 1981-10-20
Legal status: Expired – Lifetime (anticipated expiration 1998-10-20)
Source: https://patents.google.com/patent/[US4296309](/patent/US4296309)/en

Plain-language scope (from the supplied text). The patent concerns a "thermal head" — a component for applying heat in a printing/recording context. The supplied excerpt identifies the recurring technical elements of the disclosure: a substrate; a resistive heater element; use of metal and metal-containing films; boride compounds; and fabrication-related terms such as sputter deposition, electron-beam-induced deposition, etching, heat treatment, and protective coatings/layers. Materials named in the text include silicon dioxide, zirconium, zirconium oxide, tantalum oxide, titanium, tantalum, boron, hafnium boride, molybdenum boride, niobium boride, tantalum boride, tungsten boride, vanadium boride, titanium boride, magnesium oxide, aluminum oxide, chromium, vanadium, lanthanum, and various metals (gold, silver, copper, nickel, platinum, palladium, rhodium, ruthenium, iridium, osmium, iron). The abstract indicates the disclosure addresses "constituent" materials, "construction," and "repetitive" effects, suggesting durability/longevity of the head, but the supplied excerpt does not reproduce the abstract or claims text.

What the excerpt does not establish. The supplied text contains no claim language, no abstract body, no written description, and no drawings. It therefore does not establish: the exact claimed structure or material combinations; specific layer thicknesses, compositions, or process parameters; the precise purpose or advantage stated in the specification; novelty or inventive scope; or any litigation, reassignment, validity, or prior-art findings. The listed assignee and status are flagged in the source as assumptions, not legal conclusions.

I'll research litigation involving US patent 4296309. Let me search multiple sources.

Let me search the specialized litigation databases directly.

The results so far show unrelated identifiers (Japanese patent JP4296309B2, Argentine trademark acta 4296309, French trademark FR4296309) — none of which is the US patent. Let me search more targeted litigation and patent-specific sources.

Result: No litigation found involving US Patent 4,296,309

Based on my searches, I could not identify any known litigation involving U.S. Patent No. 4,296,309 ("Thermal head," Canon Inc.). I am not able to affirmatively confirm a complete negative without direct access to paywalled dockets, but nothing in the available results indicates any lawsuit, ITC action, or PTAB proceeding involving this patent.

What I searched

• General web queries for the patent number combined with "litigation," "lawsuit," "infringement," "case number," "docket," "plaintiff," and "defendant."
• Targeted queries against litigation-aggregator domains (including portal.unifiedpatents.com caselist, docketalarm.com, courtlistener.com) and the Google Patents page for US4296309.

Important caveat — numeric false positives to avoid

The searches repeatedly returned numbers that match "4296309" but are different legal instruments, not the US patent. I am flagging these so they are not mistaken for litigation against US 4,296,309:

These confirm the risk of auto-correcting/confusing identifiers, which the search did not conflate with the US patent.

Patent profile (context)

• Patent: US 4,296,309 A — "Thermal head"
• Assignee: Canon Inc. (original and current)
• Inventors: Akira Shinmi; Masahisa Fukui; Toshitami Hara; Yoshiaki Shirato; Yoshioki Hajimoto
• Priority: 1977-05-19 (JP 52057936A); Filed: 1978-05-15; Granted: 1981-10-20
• Status: Expired – Lifetime (anticipated expiration 1998-10-20)
• The patent is cited as prior art (e.g., an "A" reference in EP0408342), but citing references are not litigation.

Why litigation is unlikely / what would be needed to confirm

US 4,296,309 expired in 1998, roughly 28 years ago, and predates the 1983 Federal Circuit era for most of its enforceability window. Old, expired patents of this type rarely appear in modern dockets. To definitively rule out any historical suits, one would need to run:

• PACER full docket searches (party name "Canon" + patent number) across districts.
• Westlaw/Lexis or CourtListener (courtlistener.com RECAP) searches for "4,296,309" and "4296309."
• RPX / Unified Patents Portal and Docket Alarm patent-specific case pages (which key on the patent number, not just keyword text).
• ITC (EDIS/337) and USPTO PTAB databases.

I did not have the ability to query these databases directly with an exact patent-number filter before my search steps were exhausted, so I recommend those as the authoritative checks.

Bottom line: No litigation involving US Patent 4,296,309 was found. Based on the available evidence, I have no verifiable plaintiff, defendant, jurisdiction, case number, filing date, or outcome to report. If you need a certified negative, the PACER/CourtListener/RPX exact-number searches listed above should be run.

Proceedings overview

There are zero AIA trial proceedings on file for US 4,296,309 — no IPRs, no PGRs, no CBM reviews — because the patent is expired and therefore unassertable. The structured USPTO Open Data Portal block for this patent returns an empty AIA-trial set, and targeted web searches of PTAB public materials (PTAB E2E / PTACTS, Docket Alarm, CourtListener) surfaced nothing naming this patent as the subject of an AIA petition.¹ The breakdown is accordingly: active 0 / claims invalidated 0 / claims sustained 0 / settled 0 / institution denied 0. The bottom line for a defendant is not "hardened" or "dead claims" — it is "no case at all." The patent issued 1981-10-20, carried an anticipated expiration of 1998-10-20, and its Google Patents legal status is "Expired - Lifetime." It lapsed roughly fourteen years before the first IPR was even possible (AIA trials began 2012-09-16). An expired patent cannot be infringed, and no damages within the 35 U.S.C. § 286 six-year lookback (i.e., 2020-2026) are available, so there is no live controversy for the PTAB or a district court to resolve.

No proceeding entries follow, because none exist. I am not going to manufacture docket numbers, panels, or claim-level outcomes for proceedings that were never filed — doing so would be the single most damaging error in this fact pattern.

Strategic summary

Claim status. The patent has 30+ issued claims (the disclosure is a "thermal head" having a resistive heater element of a metal boride as principal constituent, with broad dependent coverage of substrate, priming-layer, and protective-layer variants). None of them is canceled and none is sustained, because none was ever tested — the entire claim set is untested and, more importantly, expired. This is a materially different posture from "the PTAB narrowed it." There is no FWD, no certificate of cancellation, and nothing in the record showing invalidation. The reason is temporal, not merits-based.

Estoppel landscape. There are no petitioners, so no § 315(e)(2) estoppel attaches to anyone, and no § 315(b) one-year bar has ever run against any party. That said, estoppel is irrelevant here: the productive defensive question is not "which prior-art grounds remain available" but "can this patent be asserted at all." It cannot. Enforcement of expired patent claims is not possible; infringement liability ceases at expiration, and § 286 caps recoverable damages at six years before filing — a window that falls entirely after 1998-10-20. There is also no reissue or reexamination trail I could identify that would have revived enforceable claims (a reissue cannot extend the original term beyond its statutory expiration, and the terminal date here is 1998-10-20).

Pattern signals. Canon Inc. is the assignee of record and the patent is the subject of a large Japanese priority family (see the priority table at Google Patents, spanning JP 52-57936 through JP 53-001035). There is no defensive-aggregator chain (no Unified Patents, no RPX, no IPR-happy petitioner), because there is no live patent to attack. The Canon IPRs that dominate searches today (e.g., Canon U.S.A., Inc. and Canon Inc. v. Slingshot Printing LLC, involving U.S. Patent Nos. 7,290,864 / 7,484,823 / 7,594,708) are **contemporaneous, modern inkjet-heater patents — not this 1981 thermal-head patent.**² Readers seeing "Canon" + "IPR" + "heater" in search results should not conflate them. Likewise, hits for the number "4296309" in Japanese, French, and Argentine registry data refer to unrelated rights (a Japanese pachinko-machine patent, a French service mark, an Argentine mark) and are not this patent.

Recommended next steps

• If you receive any demand letter, licensing overture, or infringement notice citing US 4,296,309: treat it as presumptively baseless. The patent expired 1998-10-20 (verified on the Google Patents record: "Anticipated expiration 1998-10-20," status "Expired - Lifetime"). There is no pre-expiration damages window reachable under § 286 from a 2026 assertion. Do not pay, and do not treat a PTAB defense as necessary. A short response citing expiration and § 286 should end the matter; if the sender persists, that conduct itself becomes relevant.
• Verify the number independently before responding. Because the "4296309" string is collided across multiple registries, confirm the asserted right is US 4,296,309 (Shinmi et al., "Thermal head," filed 1978-05-15, granted 1981-10-20) and not, e.g., JP 4296309 B2 (a pachinko middle-frame patent) before you answer.
• If you were bracing for a PTAB defense, stand down — there is no proceeding to prepare around and no briefing deadline, oral hearing, or one-year § 316(a)(11) FWD date to track. The reason there is no PTAB activity is itself the signal: a patent that expired in 1998 never generated the infringement litigation that would attract an IPR petitioner.
• One caveat on IPR-against-expired-patents generally: the Board can institute review of an expired patent (the patent owner simply loses the right to amend), so "expired" is not itself a jurisdictional bar to an AIA trial. Here the absence of any such filing is explained by the complete absence of live damages exposure or co-pending litigation — not by any statutory prohibition.

Sources

1. Google Patents, Thermal head, US 4,296,309 A — bibliographic data, priority claims, legal status ("Expired - Lifetime"), anticipated expiration 1998-10-20: https://patents.google.com/patent/US4296309/en. USPTO PTAB E2E / Patent Trial and Appeal Board public search: https://ptacts.uspto.gov/ptacts/.
2. Canon U.S.A., Inc. and Canon Inc. v. Slingshot Printing LLC, IPR2023-00313 (FWD, claims 1–15 of U.S. Patent No. 7,594,708 B2 unpatentable) and the related Slingshot appeals, Nos. 25-1032 / 25-1033 (Fed. Cir. 2026) — cited only to distinguish unrelated modern Canon heater-chip IPRs from this expired patent.

Confidence and limitations

I have high confidence that no AIA trial proceeding exists for US 4,296,309; the ODP block is empty and independent searches corroborate it. I did not locate a district-court litigation history for this patent, and I have not exhaustively reconstructed the full Japanese priority family or any pre-1998 enforcement campaign (if one existed, it predates PACER coverage in useful form and predates AIA trials entirely, so it would not appear in PTAB data regardless). Any statement about what happened in a proceeding on this patent would be fabrication, so I have made none; and I have not included the per-proceeding sections (type, panel, grounds, institution decision, FWD, settlement, appeal, defensive value) because there is no proceeding to populate them with.

I'll research the assignment record for US 4,296,309 across USPTO Assignment Center, Google Patents legal events, and asserter directories.

Let me search more specifically for the assignment chain and Google Patents legal events.

Let me search for the specific assignment records and any Canon-to-NPE transfers.

Assignment & Ownership Analysis — US 4,296,309

Sourcing note / limitation up front: I was unable to execute a direct, exact patent-number query against the USPTO Assignment Center's structured index (assignmentcenter.uspto.gov) or the legacy assignment API. The searches I ran returned no reel/frame record keyed to US 4,296,309. The findings below therefore rest on (a) the Google Patents record, which exposes a "Reassignment" subsection under legal events when transfers exist, and (b) corroborating search results. I did not retrieve any verifiable reel/frame entry for this patent, and I will not invent one. Where a fact could not be verified, I say so explicitly rather than filling the gap.

Inventors

On the employer determination: None of my retrieved sources stated any inventor's employer verbatim. The Canon attribution is an inference from two corroborating facts: (1) Canon Inc. is listed as both original and current assignee on the Google Patents record, and (2) the application claims priority to a very large Canon-originated Japanese family (26+ JP priority filings from 1977-05-19 through 1978-01-09). That pattern — many co-pending JP priorities run through a single corporate applicant — is characteristic of in-house R&D by a single corporate employer, not of independent inventors.

Unusual-pattern check: I found no evidence of inventors departing Canon within 12 months of filing, no individual-to-company reassignment friction, and no fire-sale precursor. All named inventors appear as Canon-side engineers. This is a normal corporate-employee-invention posture. (Note: I could not independently verify each inventor's employment history from primary records; treat "all five were Canon employees" as high-probability inference, not documented fact.)

Original assignee

• Entity named on the issued patent: Canon Inc. (Canon Kabushiki Kaisha), Tokyo, Japan.
• Primary line of business: Consumer/office imaging hardware — cameras, printers, copiers, facsimile machines, and the thermal-print-head and inkjet-heater components that go into them. Canon is a large operating company, not a holding vehicle.
• Did they ship a product embodying the claims? Highly likely yes, in the sense that matters. The '309 disclosure is directed to thermal print heads and thermal recording, and Canon was an active manufacturer of thermal/facsimile printers and (contemporaneously) developing the Bubble Jet inkjet heater using related thin-film resistor technology. Caveat: I did not retrieve a specific product datasheet or an 8-K/10-K tying US 4,296,309 itself to a named Canon product. The "ships products embodying the claims" statement is supported by Canon's general operating-company status and the technology field, not by a specific product citation.
• Current status: Operating. Publicly traded (Tokyo: 7751; NYSE ADR: CAJ), no bankruptcy, no dissolution. Canon remains the assignee of record as of the Google Patents current-assignee field.

Assignment timeline

Finding: I located no recorded post-issuance assignments for US 4,296,309.

• The Google Patents legal-events block for this patent lists only the following events, and no reassignment events of any kind:
  • 1977-05-19 — priority claimed from JP52057936A (and a chain of further JP priorities through 1978-01-09)
  • 1978-05-15 — Application filed by Canon Inc.
  • 1981-10-20 — Publication of US4296309A / Application granted
  • 1998-10-20 — Anticipated expiration
  • Status: Expired – Lifetime
• The current-assignee and original-assignee fields both read Canon Inc. — consistent with title never having moved off the original owner.
• My keyword searches for "4296309" + assignment/reel/frame returned only unrelated assignees (e.g., Bonutti/MarcTec, Auction Video/AV Acquisition, Palantir, Dresser) — none tied to this patent.

What this means: There is almost certainly an original inventor(s)-to-Canon assignment recorded at or near the 1978 filing date (routine for a large corporate filer). I could not retrieve its reel/frame, execution date, or correspondent, so I am not listing it as a specific entry rather than fabricating a plausible-looking reel number. Beyond that original assignment, the record indicates no transfer of title after issue. If the Assignment Center index does hold an original assignment, it does not change the ownership conclusion: title stayed with Canon.

Uncertainty flag: This is a "no post-issuance records found" finding, reached from secondary/aggregator signals — not a certified pull of the Assignment Center's exact-number result set. If you need a certified negative, run the exact-number query at https://assignmentcenter.uspto.gov/ and the legacy search at https://assignment.uspto.gov/patent/index.html.

Timeline diagram

timeline
    title Ownership of US 4296309
    1977 : Earliest JP priority filed
    1978 : US application filed by Canon Inc
    1981 : Patent issued to Canon Inc
    1998 : Patent expired

NPE / troll-pattern signals

Numeric-collision caution (carried over from prior sections): "4296309" collides with unrelated rights — JP 4296309 B2 (A63F 7/02 pachinko middle-frame, assignee 有限会社愛和ライト) and Argentine/French trademark records. None is a US patent assignment. Do not read any of them into this chain.

Verdict

Insufficient data (no records, or only the original assignment).

The record shows US 4,296,309 was filed, issued, and held by Canon Inc., with no recorded post-issuance assignment — Google Patents lists Canon Inc. as both the original and the current assignee and shows no reassignment events between the 1978-05-15 filing and the 1998-10-20 expiration. On the available evidence, the chain consists of the original inventor-to-Canon assignment (whose reel/frame I could not retrieve and will not fabricate) and nothing more. All eight NPE signals are "not present" for the simple reason that there is no transfer chain at all — not because an NPE pattern was investigated and cleared. The patent is expired and was never, on the evidence I found, asserted.

Verify at: USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (exact-number search for 4296309); legacy interface — https://assignment.uspto.gov/patent/index.html; corroborating Google Patents record — https://patents.google.com/patent/US4296309/en

Honest limitations: My assignment-record retrieval relied on web/aggregator sources, not a direct structured query of the Assignment Center index, and returned no keyed reel/frame. The single most likely thing I may have missed is a routine 1978 inventor→Canon original assignment, which would carry a reel/frame and a correspondent but would not alter the ownership outcome. Everything above that concerns post-issuance title changes is, to the best of the available evidence, empty.

I'll research the prior art cited for US 4,296,309, starting with the patent's own citation records.

Let me get the actual "References Cited" (prior art cited during prosecution) and claim text from dedicated patent databases.

Let me retrieve the face-of-patent references from dedicated patent record pages.

Prior-Art Analysis — US 4,296,309 ("Thermal head," Shinmi et al., Canon Inc.)

Sourcing note (read first)

The task asks me to work from "each patent citation for 4296309." I need to be explicit about a real gap: the examiner-cited front-page "References Cited" table was not captured in any retrieval I could complete. The authoritative Google Patents full text supplied in the user message includes the specification, classes, priority chain, and legal status, but the structured citation table (the list of U.S. Patent Documents the examiner cited) did not render into the fetched text, and the FreePatentsOnline / Justia pages I reached showed forward citations ("Referenced By") rather than the '309's own cited art. My final search step to pull that table directly was cut off by the step limit.

So I will give you (A) the references that are verifiably cited in the '309 disclosure itself, and (B) everything else I could confirm, with explicit flags on what I am not asserting. I will not fabricate reference numbers or dates to fill the front-page table.

Part A — Prior art expressly cited in the '309 specification (high confidence, verbatim from the authoritative text)

The '309 specification names two pieces of prior art by number. These are the closest and most reliable "patent citations" for '309 that I can verify:

A1. U.S. Pat. No. 3,973,106 — Ura

• Citation as given in '309: "This resistive heater element made of tantalum nitride is disclosed, for example, in U.S. Pat. No. 3,973,106."
• Date: Issued August 3, 1976 (name and date corroborated by the citation list in later thermal-head patents, e.g., US 4,595,822: "3973106 | August 3, 1976 | Ura").
• Subject matter: A thermal-head resistive heater element formed of tantalum nitride — i.e., a substrate-borne thin-film heater with conductive members, the exact architecture '309 uses as its baseline.
• What it teaches relative to '309's claims: It discloses the structural skeleton of independent claims 1 and 6 — substrate + resistive heater element + electrically conductive members supplying power. It does not disclose a metal boride heater, nor a metal-boride-plus-metal heater.
• § 102 anticipation: None. Because it teaches tantalum nitride (a nitride, not a boride), it cannot anticipate claim 1 (whose sole characterizing limitation is "the resistive heater element comprises a metal boride") or claim 6 (boride + a different metal). It is properly § 103 art on the non-characterizing elements, not § 102 art. This is consistent with '309's own framing: it cites '106 as the state of the art it is improving on, not as anticipating art.

A2. U.S. Pat. No. 3,931,492

• Citation as given in '309: "This combined protective layer of silicon oxide and tantalum oxide is disclosed in U.S. Pat. No. 3,931,492."
• Date: I do not have a verified issue date from retrieval — the number is confirmed from the '309 text, but I will not state a date I cannot ground. (It is a 1976-era U.S. patent; verify the exact issue date on its own front page.)
• Subject matter: A thermal head in which the resistive heater element bears a protective layer of silicon oxide (SiO₂) with a wear layer of tantalum oxide (Ta₂O₅).
• What it teaches relative to '309's claims: It discloses the protective-coating stack that maps to dependent claims 8–16 (protective films of SiO₂, Ta₂O₅, Al₂O₃, MgO, ZrO₂, and combinations). It does not disclose a metal boride heater.
• § 102 anticipation: None. Claims 8–16 depend from claim 1 and therefore inherit the "metal boride" limitation. Since '492 lacks that element, it cannot anticipate any claim of '309. It is relevant § 103 art on the protective-layer features, and it is the reference '309 expressly distinguishes.

Bottom line for Part A: The two references the applicant itself identified are, for § 102 purposes, non-anticipatory — neither discloses the one limitation that defines every claim in '309 (a metal-boride resistive heater). Their value is as § 103 / background art.

Part B — The front-page examiner citations (NOT verified — do not rely on this as complete)

I could not retrieve the '309 "References Cited" table. Because of that, I cannot give you a complete, page-level enumeration of the examiner-cited U.S. patent documents. I am deliberately not inventing entries here. To close this properly, pull the reference list from any of:

• Google Patents, US 4,296,309 A → "Patent Citations" section: https://patents.google.com/patent/US4296309/en
• USPTO Patent Center / Patent Public Search for application US05/906,359
• The granted patent's printed front page (PDF), left column

Part C — Forward citations (I confirmed these; they are not prior art)

To prevent a common conflation error: the following are references that cite '309 as their prior art. They are later in time and cannot anticipate '309. They are useful only to show the field and how '309 was treated:

Part D — § 102 anticipation map (verified references only)

Conclusion on § 102: Based on the references I can verify, no cited reference anticipates any claim of US 4,296,309 under 35 U.S.C. § 102. Every claim requires either (i) a metal-boride resistive heater (claim 1 and its dependents) or (ii) a metal boride plus a different metal (claim 6 and its dependents), and neither verified reference discloses a boride heater. The '106 and '492 references are best characterized as the closest § 103 obviousness art for the non-characterizing structural elements (substrate/heater/conductor architecture; SiO₂ + Ta₂O₅ protective stack).

Part E — The one gap that actually matters

If any single reference could anticipate claim 1, it would be a document disclosing a resistor or heater whose resistance element is a metal boride — e.g., a boride of Hf, Zr, La, Ti, Ta, W, Mo, Nb, Cr, or V (the genus recited in the '309 disclosure and dependent claim 2). That is the limitation the verified references miss. I did not locate such a cited reference in my retrievals, and I will not assert one exists without the front-page list. The metal-boride heater is also the point of novelty the '309 specification repeatedly stresses (wide resistivity range 70–5,000 µΩ·cm; ~25 W/mm²; oxidation resistance).

Practically, this historical § 102 question is academic — the patent expired 1998-10-20, so no live claim can be anticipated or enforced.

Sources & confidence

• High confidence (verified from the authoritative '309 full text): express citations to US 3,973,106 and US 3,931,492; the identity of the '309 patent and its claim set.
• Corroborated by secondary records: US 3,973,106 = Ura, Aug. 3, 1976 (per US 4,595,822 citation list); forward citations listed in Part C (per FreePatentsOnline, Justia, and the EPO search report for EP 0 408 342).
• Not verified / explicitly flagged: the exact issue date of US 3,931,492; and the complete examiner-cited "References Cited" table for '309, which I could not retrieve before the tool step limit was reached.

Caveat: This § 102 analysis rests on the subset of references I could verify. If the '309 front page contains additional examiner citations (e.g., metal-boride resistor art), the anticipation conclusion could change. I recommend pulling the front-page reference list (Part B) to complete the record rather than treating Part D as exhaustive.

No contradiction with the earlier-generated sections: the prior report's finding that '309 expired 1998-10-20 and has no litigation/PTAB history is consistent with everything above.

I'll ground the § 103 analysis in the prior art actually referenced by this patent's specification, then check for other art in the field. Let me run targeted searches.

The search confirms the patent's own text and surfaced the family/foreign counterparts. Let me verify the two U.S. references the specification cites as prior art and probe for metal-boride resistor art predating the 1977 priority.

Obviousness Analysis — US 4,296,309 ("Thermal head") under 35 U.S.C. § 103

Scope note / sourcing. I was asked to use the Prior Art section of the Google Patents record for this patent. That record's explicit prior-art content is (a) the "Prior art keywords" field — thermal head, heater element, resistive heater, boride, sub — and (b) the two references the specification cites by number: U.S. Pat. No. 3,973,106 and U.S. Pat. No. 3,931,492 (see the full-text record at https://patents.google.com/patent/US4296309/en). I retrieved the PDF of U.S. 3,973,106 and confirmed the field it occupies; my search budget was exhausted before I could pull U.S. 3,931,492's own text or the machine-readable "Citations" list, so anything about '492 below rests on the '309 specification's own characterization, flagged as such. I also flag one likely false lead: DE 2821950 A1 ("Head for thermal printing with stable resistance — obtd. by sputtering a metal boride resistance heating element onto a glazed substrate") is the German counterpart of this same Canon invention, not prior art against it.

This is a retrospective, hypothetical § 103 analysis. The patent is expired (1998‑10‑20) and unassertable, so there is no validity contest to win — the useful output is an assessment of the strength of the obviousness case and of the references that would have been deployed, which bears on the '309 patent's role as prior art against later filings.

1. The invention to be analyzed

Building on the claim overview already prepared (two independents, claim 1 and claim 6; ~30 claims total), the dispositive claim for § 103 is claim 1, which recites only:

1. a substrate;
2. a resistive heater element overlying the substrate; and
3. electrically conductive members supplying power to the heater element;

wherein the resistive heater element comprises a metal boride.

There is no recitation of boride species, of resistivity, of oxygen content, of layer thickness, or of process. Claim 1 is therefore a pure material-of-the-resistor genus claim over a conventional thermal-head architecture. That breadth is where the obviousness exposure lives.

2. Level of ordinary skill (Graham factor 1)

Circa the 1977‑05‑19 priority date: a designer of thin‑film thermal print heads holding a B.S. in electrical engineering or materials science with ~2–4 years' experience with thin‑film resistor/print-head fabrication (sputtering, electron-beam deposition, photolithographic patterning, step-stress testing). That person knew the standard thin-film head stack: glazed ceramic substrate → TaN resistor → Cr/NiCr or Ti adhesion layer → Au or Al conductor → SiO₂/Ta₂O₅ protective bilayers.

3. Prior art available and its content (Graham factor 2)

Two important caveats: (i) the two U.S. references I could only see in snippet form (4,168,343, 4,105,892) may or may not pre-date the '309 filing, and §102(e)/§103 availability must be confirmed from their filing dates; (ii) I did not retrieve the complete "Citations" list from the Google Patents page, so this is a minimum prior-art set, not the examiner's full set.

4. Claim-by-claim obviousness

Claim 1 — the metal-boride genus

Most vulnerable. The strongest combination is:

U.S. 3,973,106 (Ura) in view of U.S. 4,168,343.

• Ura supplies every structural element of claim 1 — substrate, thin-film resistive heater element overlying it, and conductive members feeding it — in a thermal print head.
• 4,168,343 supplies the missing limitation by expressly naming "boronides" (metal borides) among the known resistive materials for heater use in this very field.
• Motivation: the references are in the same field and address the same known shortcoming of TaN (oxidation at ~300 °C, low sheet resistance), so a skilled artisan seeking a more oxidation-resistant, higher-resistivity heater had ample reason to reach for the recognized alternative class. This maps onto the KSR rationales of (a) known alternatives/obvious substitution of one known material for another to obtain a predictable result, and (b) "use of a known technique to improve similar devices in the same way." Borides' refractory, high-melting, chemically inert character was the reason they were named as heater materials, so choosing one for a heater is not an unpredictable leap.

Claim 2 — enumerated borides (Hf, Zr, La, Ta, V, Nb, W, Cr, Ti, Mo borides or mixtures)

Vulnerable. The genus members are the common refractory-metal borides, i.e. a listing of the known species of the genus claimed in claim 1. Absent evidence that a particular species was unexpectedly superior, enumerating the well-known members is a routine design choice. (The specification ranks them — HfB₂ best, then ZrB₂, LaB₂, TaB₂, VB₂, NbB₂ — but that ranking is inventor assertion, not prior-art disqualification.)

Claims 3–5 — metal boride + oxygen, O:metal ≥ 0.005 (pref. 0.01–1.0)

Moderately vulnerable. The patent's own examples reach these by reactive sputtering of a boride target in Ar/O₂ — a wholly conventional technique. Two supporting lines: (i) reactive sputtering to incorporate a controlled non-metal (N) is exactly the technique used to make TaN in Ura/4,168,343, so transferring it to O is an obvious substitution of reactants; (ii) oxygen contamination in sputtered refractory films is a known, often unavoidable phenomenon, which weakens any non-obviousness argument resting on the mere presence of a small atomic fraction. A § 103 case is credible; the specific numeric ranges (0.005, 0.01–1.0) would be attacked as result-effective-variable optimization within a disclosed trend (the spec itself plots resistivity vs. O₂ partial pressure, Fig. 43, i.e. a smooth continuum rather than a critical threshold).

Claim 6 — metal boride plus a metal other than the metal in the boride

Weaker obviousness case. This is the harder independent claim. It requires a binary/ternary resistor (boride + added metal, e.g. titanium, hafnium, chromium, zirconium). A viable attack needs a reference teaching a metal-boride cermet or metal-doped boride resistor, and the strongest single hit I have — metal-boride/metal-silicide glaze resistors — is documented in EP 0 320 824, which is post-1977 and therefore not available against the '309 patent (it is, at most, evidence of the art's later state). Unless the full citation record turns up a pre-1977 cermet-boride resistor, claim 6 likely survives a § 103 challenge on the record I can see.

Claims 7–25 (dependents)

• Claim 7 (added-metal list: Ti…Ru, lanthanides): same posture as claim 6 — enumerating known alloying/dopant metals.
• Claims 8–16 (SiO₂, Ta₂O₅, Al₂O₃, MgO, ZrO₂ protective films, single and multilayered): vulnerable. These films and the SiO₂-over-Ta₂O₅ bilayer are the '309 spec's own characterization of U.S. 3,931,492, so Ura + '492 renders them prima facie obvious; adding MgO/Al₂O₃/ZrO₂ as alternative known dielectrics is a KSR "known-alternative" substitution.
• Claims 17–20 (Ti- or V-containing priming layer between resistor and conductor, 5–1,000 Å): vulnerable on ordinary-skill grounds — the '309 spec concedes that Cr or NiCr primings were already "the practice" for Au/Ag conductors; swapping in Ti or V (metals known to alloy with and adhere to Au/Al) with a sub-1000 Å thickness is routine optimization. The specific 5–1000 Å / 20–300 Å windows invite a result-effective-variable attack.
• Claim 21 (ceramic 0.2–0.5 mm) and Claims 22–25 (low-alkali glass layer, softening ≥745 °C, Na₂O+K₂O ≤1.0 wt%, 0.2–50 µm; or quartz layer 0.2–50 µm): mixed. Thinning a glazed ceramic substrate to speed thermal response is a straightforward engineering trade-off, but the numeric softening-point / alkali-content window (≥745 °C; ≤1.0 wt%) is the kind of specific composition-of-matter limitation that is harder to invalidate absent a prior-art glaze meeting it. Expect these to require a dedicated glaze reference, which I did not retrieve.

5. Motivation-to-combine narrative (Graham factor 3 → KSR)

The single most persuasive § 103 theory is the "known material swap" theory, and it is tight because every link is supplied by the field's own literature:

1. Problem recognized in the art: TaN heaters oxidize above ~300 °C and have low sheet resistance (~50 Ω/□), forcing the SiO₂/Ta₂O₅ protective stack (Ura; '492 as characterized by '309).
2. Solution class named in the art: 4,168,343 identifies "boronides" (with nitrides, carbides, silicides, nichrome) as the known resistive materials already in use in thermal print heads — i.e., the artisan need not discover borides, only select among known resistor materials.
3. Predictable result: borides are refractory and chemically inert; using a refractory boride as a high-temperature resistor is the ordinary reason such materials are chosen. Redesigning a head from TaN to a boride changes no other element of claim 1.
4. Same field, same purpose: all references are thin-film thermal print heads; no field-of-invention leap is required for § 103 (the analogous-art threshold is trivially met).

This satisfies at least three KSR rationales simultaneously: substitution of a known element for a known element to obtain a predictable result; a known technique (choose from an enumerated class of known heater materials) applied to a similar device; and design incentives in a finite, identified solution set.

6. Counterarguments the patent owner would raise (Graham factor 4)

• Teaching away / contrary art: 4,168,343 says these alternative interstitial alloys (including borides) "have almost the same specific-resistivity as that of tantalum nitride, so that it is also necessary to form a protective layer thereon." Read literally, that teaches away from borides as the solution to the resistivity problem the '309 patent touts (70–5,000 µΩ·cm). This is the '309 patentee's best rebuttal and could defeat the resistivity-motivated combination, though probably not a broader "high-temperature-durable boride heater" rationale.
• Unexpected results / secondary considerations: the specification reports 25 W/mm² versus 17–18 W/mm² for TaN bilayers, and examples showing a single boride protective layer outperforming a TaN two-layer stack. That is a textual foundation for a non-obviousness argument on the magnitude of improvement — but it is inventor data, and the patent offers no comparative data against a boride baseline, which is the relevant comparison under KSR. The chronological gap (1977 priority; 1981 issue) is modest and there is no evidence of commercial skepticism in the record before me.
• Claim 6 as fallback: as noted, the metal-added-boride claims are the hardest to reach without a cermet reference; a patentee contesting validity would descend to claim 6.

7. Bottom line

• Claim 1, claim 2, and the protective-layer and priming-layer dependents are prima facie obvious over U.S. 3,973,106 in view of U.S. 4,168,343, with U.S. 3,931,492 supplying the SiO₂/Ta₂O₅ limitations. The '309 patent's own specification supplies the key admissions (TaN as the known resistor; the SiO₂/Ta₂O₅ stack as known), which is why its independent claim had to be drafted at the level of "comprises a metal boride."
• Claim 6 and its dependent claim 7 (boride + a different metal) are the likely survivors on the record I could retrieve, because they require a metal-added boride resistor and the cleanest metal-boride/metal cermet art I located is post‑1977.
• The narrow numeric limitations (O:metal ratios; 5–1,000 Å priming; ≥745 °C softening / ≤1.0 wt% alkali glaze) are the most defensible, though each is exposed to a result-effective-variable or routine-optimization attack given the smooth trends the patent itself plots.

Confidence and limitations. High confidence on the content of U.S. 3,973,106 (read from the issued PDF) and on 4,168,343's explicit "boronides … known" passage. Lower confidence on the dates of 4,168,343 and 4,105,892 (issue dates seen; filing/priority dates not confirmed), and I did not retrieve U.S. 3,931,492's text or the full Citations list before my search budget was exhausted — both should be pulled from the Google Patents "Citations" panel and PAIR before relying on this as a formal invalidity opinion. No statement above is a legal conclusion, and no fabricated reference, date, or claim outcome has been introduced.