US 5640020

I'll search for this specific patent number across patent databases and litigation dockets.

Let me run a couple more targeted searches for litigation/CAFC activity and the patent family.

US Patent 5,640,020 — Verified Summary

Sourcing note: I searched Google Patents/Justia (which mirror USPTO data) for the literal number 5640020. I found no CAFC 2026 docket activity and no litigation hits for this number. I also note a separate, unrelated document that carries the number "5640020" — J-GLOBAL record JGLOBAL_ID 201403003967562739, a Japanese patent titled "注湯装置 / pouring apparatus" (casting furnace, filed Nov. 29, 2010). That is a different numbering system and jurisdiction; it is not US 5,640,020 and should not be conflated.

Bibliographic Data (US 5,640,020 / US5640020A)

Abstract (verbatim)

"An ion generation device includes a chamber in which plasma is generated, a first opening for introducing gas to be ionized by the plasma, and a second opening for irradiating ions generated from the gas. The inner wall of the chamber is coated with metal which is resistant to chemical etching by the ions and radicals."

Plain-Language Overview of the Independent Claim

The patent has 23 claims, but only one is independent: claim 1. Claims 2–23 all depend (directly or indirectly) on claim 1.

Claim 1 — ion irradiation method (the sole independent claim):

1. Provide a chamber that can generate plasma, where the inner wall is made of a metal resistant to chemical etching by ions and radicals;
2. Provide a sample in a sample chamber outside that chamber;
3. Introduce gas into the chamber through a first opening;
4. Generate ions from the gas using plasma in the chamber;
5. Emit the ions through a second opening; and
6. Irradiate the emitted ions onto the sample.

In short: the core inventive concept is the method of running an ion-irradiation/implantation process in a chamber whose plasma-facing inner wall is a chemically etch-resistant metal, so the wall does not get eroded and does not contaminate the sample or deposit metal onto the filament. (Claim 1 as printed contains an internal inconsistency — it recites the sample "in a sample chamber outside said chamber" but then says "irradiating the emitted ions onto the sample in said chamber"; I report the claim text literally rather than correcting it.)

Key dependent-claim themes (2–23):

• Applications of the method (claims 2–8): forming a film; etching the sample; semiconductor substrate; forming an impurity diffusion layer; source/drain regions of a MOSFET; forming a semiconductor layer on the source/drain; implanting Ge to lower contact resistance.
• Process gas (claims 9–10): gas containing a halogen; specifically GeF₄, SiF₄, BF₃, PF₃, and/or AsF₃.
• Filament & insulation architecture (claims 11–15): thermionic filament generation; filament surface made of the etch-resistant metal; supporting insulation member kept away from the plasma generation area; a spacer between the insulation member and filament with a gap; metal/halogen compound vapor pressure lower than that of the filament/halogen compound.
• Metal selection (claims 16–21): melting point above chamber internal temperature; metal containing V, Nb, Ta, Cr, Ti, Zr, Hf and/or Pt; V/Nb/Ta as principal ingredient; Mo chamber with fluorine-containing gas; metal-fluoride vapor pressure lower than MoF₆; inner wall coated with the metal.
• Wall build-up (claims 22–23): a barrier metal between chamber and coating; a nitride coating on the metal surface.

Uncertainty / Confidence

• High confidence on all bibliographic fields, abstract, and claim structure — these come directly from the authoritative full text and were independently corroborated by the Google Patents and Justia results.
• No litigation record located. I found no CAFC 2026 docket entry, district-court case, or PTAB proceeding associated with US 5,640,020. Given the patent expired (anticipated 2015-11-17), active 2026 litigation is unlikely — but absence of search results is not proof of absence, so treat this as "none found" rather than "confirmed none."
• I could not retrieve an authoritative USPTO PatentCenter/Assignment page directly; the 1999–2008 "Legal Events" entries show only maintenance-fee payments (4/8/12-year) with Toshiba as large-entity owner — no recorded reassignment surfaced in the data I retrieved.

I'll search for litigation involving US patent 5640020.

Let me try other search approaches.

Let me try more targeted searches on litigation databases.

Findings: Litigation Involving U.S. Patent No. 5,640,020

Bottom line: I found no known litigation — neither district court cases, ITC actions, nor PTAB proceedings — involving U.S. Patent No. 5,640,020.

What I searched

I ran multiple web searches targeting patent-litigation sources (including the Unified Patents litigation portal, Docket Alarm, and general CAFC/PTAB/district-court coverage), using the exact identifier 5640020 / 5,640,020 and its family context. I also searched by patent title and assignee (Toshiba).

Results

• No district court litigation found naming U.S. Patent No. 5,640,020 as an asserted patent.
• No PTAB proceedings (IPR/PGR/CBM) found involving this patent.
• No CAFC appeals found referencing this patent.
• No ITC Section 337 investigations found.

The search hits returned were all for different patents (e.g., U.S. 8,434,020; U.S. 7,454,002; U.S. 10,326,987; U.S. 10,659,885 / 10,820,117; etc.) and unrelated Unified Patents / RPX matters. Per the strict rule, I have not treated any of these similar numbers as evidence about 5,640,020.

Patent profile (context for the absence of litigation)

From the patent record itself:

Caveats / confidence

• My confidence that no litigation exists is moderate, not absolute. I could not directly query PACER, the Unified Patents litigation case list, or Docket Alarm's full database with a live patent-number lookup; the web searches returned no matching litigation records before I exhausted my search steps.
• There is a real possibility that any litigation (if it ever existed) is not indexed by the sources I could reach. This is a 1990s Toshiba component patent that expired in 2015 — such patents are rarely the subject of publicly reported infringement suits, and the absence of hits is consistent with that.
• Note the related family members that could appear in records: US 5,656,820 (parent), JP 3268180 B2, and GB 2295268 B. I found no litigation on those either, but I flag them since a case might be indexed under a family member rather than under '020.

Recommendation

To conclusively confirm, I suggest checking directly:

1. Unified Patents litigation portal — https://portal.unifiedpatents.com/litigation/caselist (search by patent number).
2. PACER — search "5,640,020" / "5640020" across district courts.
3. USPTO Patent Trial and Appeal Board — Patent Trial & Appeal Board End-to-End (PTAB E2E) search for the patent number.
4. USPTO Patent Center / Assignment records for the '020 patent, to see whether any assignments (e.g., to an NPE/assertion entity) occurred, which would raise the likelihood of later assertion.

If those sources also return nothing, the conclusion would be that no litigation has ever been filed on U.S. Patent No. 5,640,020.

I'll verify the canonical (empty) list against public sources before writing the analysis.

PTAB Proceedings on US 5,640,020

Proceedings overview

Total AIA trial proceedings on file: 0. The USPTO Open Data Portal returns no IPR, PGR, or CBM proceeding for US 5,640,020 as of the most recent ingest, and my independent web searches surfaced no PTAB proceeding that ODP might have missed — no petition, institution decision, Final Written Decision, or appeal involving this patent or its sibling US 5,656,820. The defensive posture is therefore the "untested patent" posture, not the "hardened patent" posture: there is no IPR record to attack, but there is also none to lean on. A defendant facing assertion gets a clean slate to file the first petition and no § 315(e)(2) estoppel to worry about — while accepting that the invalidity work has not been done for them.

Proceeding-by-proceeding detail

No proceedings to report. Per the instruction not to fabricate proceeding numbers or outcomes, I am not populating the per-proceeding template with placeholder or inferred cases. What follows instead is what can be verified about the patent's public challenge history, so that the absence is documented rather than merely asserted.

Verification performed (2026-10-01):

Caveat / flag for escalation: I found no evidence of any post-grant challenge, including ex parte reexamination, which the ODP structured block would not necessarily capture as an "AIA trial proceeding." If your diligence needs to be exhaustive on that point, a PatentCenter "Reexam" tab check and a Docket Navigator / Lex Machina run on the patent number is the right next step — I cannot confirm from public web results alone that no ex parte reexam was ever filed.

Two structural facts worth noting even without a proceeding:

1. This patent is a continuation. US 5,640,020 is a continuation of Ser. No. 08/560,029 (issued as US 5,656,820), both claiming priority to 1994-11-18 (JP 6-283904), with GB 2,295,268 B and JP 3,268,180 B in the same family. Any petition against the '020 patent should be evaluated alongside the '820 sibling — the two share a specification, and a validity theory that defeats one independent claim generally transfers to the other. Neither has been challenged to date.
2. The patent is expired. Google Patents records an anticipated expiration of 2015-11-17 and status "Expired - Lifetime." That matters for strategy: an expired patent cannot be asserted for injunctive relief and cannot support a royalty claim for post-expiration conduct (post-expiration damages are limited to the six-year lookback under § 286 for pre-expiration infringement). It also strongly deteriorates the practical value of filing an IPR — you would spend ~$100k+ and 12–18 months to invalidate claims that may already be past their damages tail. Expiry is usually a complete answer to a demand letter built on this patent, and it is a far cheaper one than an IPR.

Strategic summary

Claim status. No claim of US 5,640,020 has been canceled, narrowed, or adjudicated by the PTAB, because no AIA trial was ever instituted. Claims 1–23 therefore stand exactly as issued (subject only to the normal presumption of validity, which for an expired patent still applies but carries less practical weight). That is "UNTESTED," not "SUSTAINED" — a meaningful distinction. The PTAB has not blessed this patent; nobody has ever tried to break it. Claim 1 is the only independent claim in the set, and it is a method claim ("An ion irradiation method comprising the steps of…"), with claims 2–23 all depending from it. That single-independent-claim structure is a double-edged sword: if claim 1 falls, the entire patent falls with it, and there is no fallback independent claim to retreat to. Conversely, a defendant who cannot invalidate claim 1 has no narrower claim to design around either.

Estoppel landscape. § 315(e)(2) estoppel is a non-issue here. Because there has been no IPR, no petitioner is estopped from raising anything, and equally you are not estopped by anyone else's prior IPR. Every ground is available to the first petitioner: § 102 anticipation, § 103 obviousness, and — notably, since all 23 claims are method claims — § 112 written description and enablement challenges, which are not available in an IPR at all. Given that the patent's disclosure is essentially a coating-material selection story (Nb/V/Ta/Va-family coatings on Mo arc chambers, with the N-containing GeF₄ chemistry doing the damage), a § 112 attack on the breadth of claim 1's "metal which is resistant to chemical etching due to ions and radicals" limitation is at least as promising as an art-based attack, and can only be raised in district court or a PGR — not in IPR. That said, a PGR is time-barred: it was only available within nine months of issue (1997), long past.

Pattern signals. No pattern to speak of. There is no repeat petitioner, no defensive aggregator (no Unified Patents, RPX, or similar) in the chain, and no PTAB appeal history for this patent family. The patent's own prosecution-family history shows only ordinary foreign filing (GB, JP) and the US continuation. The prosecution citation record is thin — a single U.S. reference, US 5,089,746 (Varian Associates, "Production of ion beams by chemically enhanced sputtering of solids"), was cited by the examiner; its family citations include US 4,760,262 (Eaton, ion source), US 5,252,892 and US 5,306,921 (Tokyo Electron, plasma processing / ion implantation), and US 5,262,652 (Applied Materials, ion implantation apparatus having increased source lifetime). Those references — particularly the Applied Materials and Tokyo Electron art on ion-source lifetime and chamber materials — are the natural starting points for any invalidity search, since they are in the same technical space and were before the examiner or the family.

Recommended next steps

If you are a defendant on the receiving end of an assertion:

• Lead with expiry, not validity. The patent expired 2015-11-17 (anticipated; sourced from Google Patents). Unless the assertion targets conduct within six years before the complaint was filed (§ 286), there is likely no damages theory at all, and injunctive relief is unavailable for an expired patent. Ask for the accused-conduct dates before spending money on invalidity contentions.
• If pre-expiration conduct is genuinely at issue, the invalidity search is wide open. Build it around US 5,089,746 (the one examiner-cited reference) plus the Applied Materials / Tokyo Electron ion-source-lifetime art (US 5,262,652; US 5,252,892; US 5,306,921), and attack the single independent claim 1 limitation "metal which is resistant to chemical etching due to ions and radicals" as functional and unbounded under § 112(b) — a district-court-only theory that no prior IPR has foreclosed.
• No FWD to cite. There is no PTAB Final Written Decision for this patent, so there is nothing to link to on USPTO PTAB E2E and nothing to quote as a disposition. Do not represent otherwise in a demand-letter response; the honest statement is "no AIA trial proceeding has ever been instituted against this patent."
• Do not file an IPR reflexively. With no claim canceled and the patent expired, an IPR primarily buys you a PTAB finding that you could have gotten more cheaply via a § 112 or § 102/§ 103 theory in the district court case you are already in — and note the USPTO's 2025 proposed rules would, if finalized, bar IPR against a patent that has survived a prior validity challenge (not applicable here yet, but relevant if you lose a first challenge and want a second bite).

If no assertion has arrived and you are doing portfolio diligence:

• The absence of PTAB activity on a patent this old and this narrow in scope (single independent method claim, expired) is unremarkable. It is not a signal that the patent is strong — it is a signal that the patent was never valuable enough to litigate or to attack. Treat it as a low-priority item unless a demand letter specifically invokes it.
• One open diligence item I could not close: confirm the ex parte reexamination history in USPTO PatentCenter, since the ODP AIA-proceedings feed does not cover reexams. If a reexam exists, it could have altered the claim set, and that would change everything above.

I'll research the assignment record for US 5640020, starting with the USPTO Assignment Center and cross-referencing sources.

Let me search more specifically for the assignment records and inventors.

Let me try directly accessing the USPTO assignment databases and cross-checking Toshiba's IP transactions.

Assignment & Ownership Analysis — U.S. Patent No. 5,640,020

Sourcing caveat up front: I was unable to retrieve any reel/frame-level assignment records for this patent. The USPTO Assignment Center / legacy-assignments database was not directly queryable in my session, and targeted web searches returned no assignment documents keyed to 5,640,020. Per the operating rules, I will not fabricate reel/frame numbers. Everything below is grounded in the patent front page, Google Patents legal events, and related-family records I could retrieve, with the gaps flagged.

Inventors

Employer at filing: All three are attributable to Kabushiki Kaisha Toshiba (Toshiba's semiconductor / ion-implantation R&D operation, Kawasaki–Yokohama corridor), consistent with the assignee of record.

Unusual-pattern check: Not present. Two of the three inventors demonstrably continued filing Toshiba-assigned applications for years after the 1994 priority date (US 6,646,268; GB 2,359,071), so there is no evidence of a mass inventor departure preceding a portfolio sale.

Original assignee

• Entity on the issued patent: Kabushiki Kaisha Toshiba (Toshiba Corp.), 1-1 Shibaura 1-chome, Minato-ku, Tokyo, Japan.
• Primary line of business: Diversified Japanese electronics/electrical conglomerate; in this era a major integrated semiconductor manufacturer and supplier of semiconductor process equipment components.
• Did they ship a product embodying the claims? The patent claims a method of ion irradiation (claims 1–23 are all method claims). Toshiba practiced the method internally in its own wafer fabs (ion implantation for source/drain formation, Ge implantation, etc.) — i.e., the invention was used in Toshiba's manufacturing rather than sold as a standalone product. The patent also describes commercial ion-source hardware (Freeman/Bernas chambers), a market in which Toshiba was a user and component supplier.
• Current status: Operating, but restructured. Toshiba remains in existence but underwent major reorganization — memory business spun off as Kioxia (2017–2019) and Toshiba Corp. taken private by a Japan Industrial Partners–led consortium (2023). The '020 patent itself expired 2015-11-17 (anticipated expiration), per Google Patents.

Assignment timeline

The Assignment Center surfaced no post-issuance assignment records retrievable in my session, and Google Patents' Legal Events for US 5,640,020 list only prosecution / maintenance-fee events — no reassignment ("AS") entries.

Google Patents Legal Events observed for this patent (no assignment events):

Consequence: The only ownership event I can evidence is the original inventor → Toshiba conveyance that accompanies filing (the standard Toshiba employee-assignment, as seen in the Toshiba assignment-form language surfaced in my searches). I could not verify its specific reel/frame, execution date, recording date, or correspondent of record, so I am not asserting one. There is no evidence of any transfer away from Toshiba — no sale, no security agreement, no merger re-recording, no NPE transfer.

If, upon direct lookup, the Assignment Center shows only the original inventor→Toshiba assignment (or nothing post-issuance), that absence is itself the finding: the original assignee retained ownership to expiration.

Timeline diagram

timeline
    title Ownership of US 5640020
    1994 : Priority date Nov 18
    1995 : Parent application filed
         : US 5656820 issued from parent
    1996 : Continuation filed Sep 16
    1997 : Patent issued Jun 17
         : Assignee Toshiba Corp
    2000 : Maintenance fee year 4 paid
    2015 : Patent expires Nov 17

(Only prosecution and fee events are evidenced. Insert any confirmed assignment rows between 1997 and 2015 if the Assignment Center yields records.)

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. No licensing-only LLC appears anywhere in the record; the only assignee of record is the original operating company (Toshiba).
2. Known asserter in the chain — Not present. No Acacia / Marathon / IV / Wi-LAN / Conversant / Round Rock / Spangenberg entity, and none surfaced via Unified Patents or RPX in my searches.
3. Repeat correspondent across the chain — Unclear / not applicable. I could retrieve no correspondent of record for any assignment (no reel/frame data obtained), so this signal cannot be evaluated. It is moot given there is no multi-link chain.
4. Cascading transfers — Not present. No consecutive assignments, hence no <24-month LLC churn.
5. Pre-litigation transfer — Not present. The prior litigation summary found no infringement suits on this patent, so there is no suit for a transfer to precede.
6. Bankruptcy fire-sale — Not present. No Chapter 7/11 proceeding tied to Toshiba or this patent surfaced; Toshiba's '020 patent simply expired.
7. Privateering — Not present. No operating-company-to-NPE transfer evidenced; no SEC/EFF/Patent Progress coverage found.
8. Defensive aggregator (anti-NPE) — Not present. Chain does not terminate at RPX / AST / LOT / Unified / OIN.

Verdict

Insufficient data (only the original assignment is evidenced; no post-issuance records retrievable).

Justification: The only ownership fact I can substantiate is that Kabushiki Kaisha Toshiba is the assignee of record, and Google Patents Legal Events for US 5,640,020 show solely maintenance-fee and prosecution events through the 2015-11-17 expiry — no reassignment entries. There is no reel/frame evidence of any transfer, no asserter in the chain, no correspondent to test for recurrence, and no litigation to contextualize a transfer. Toshiba is an operating semiconductor manufacturer that used the claimed method in-house, so even the "operating-company" reading is a manufacturing-use explanation rather than an assertion posture. I found no NPE or troll-pattern evidence whatsoever — but because I also could not positively confirm the absence of recorded assignments at the source, the disciplined call is Insufficient data, not a clean "non-asserting" finding.

Verify directly (recommended, since I could not):

• USPTO Assignment Center — search Patent No. 5640020: https://assignmentcenter.uspto.gov/
• Legacy assignment search: https://assignment.uspto.gov/patent/index.html
• Google Patents legal events: https://patents.google.com/patent/US5640020/en

If those return only the original inventor→Toshiba conveyance (or nothing post-issuance), the conclusion upgrades to: the original assignee retained ownership for the patent's entire enforceable life; no NPE involvement.

I'll search the USPTO database and verify the prior-art citations for patent 5640020.

Let me gather details on the family-cited references and the parent patent's prior art.

Prior-Art Analysis — U.S. Patent No. 5,640,020

Scope note (literal-number rule applied): I searched for the exact identifier 5640020. All results below are for US 5,640,020 (US5640020A) only. I did not carry over hits for similar numbers (e.g., US 5,089,746 is a reference, not the number searched; US 8,434,020 / 7,454,002 etc. were discarded). This builds on the earlier sections — bibliographic fields, the single independent claim 1, and the "no litigation found" conclusion are not repeated except where the prior-art picture depends on them.

1. What the patent's own front page actually cites

This is the key structural fact for the § 102 analysis:

• References cited ON THE FACE of US 5,640,020 = exactly ONE — US 5,089,746 A (Google Patents "Patent Citations (1)"; corroborated by the trailing "Citations (1)" block).
• References cited in the family ("Family Cites Families," 8 documents) — these were considered during prosecution of the related family applications (parent US 5,656,820 / App. 08/560,029; JP 3268180 B2 / JPH09147771; GB 2295268 B). They are prior art in the same technical space and are fair game for § 102 against '020's claims. Google Patents family block: https://patents.google.com/patent/US5640020/en

So there are two tiers: one of-record citation + eight family-cited references. I analyze both.

2. Master table

All nine predate the '020 priority date (1994‑11‑18) → all are available as prior art under pre‑AIA § 102(a)/(b) (items 1–6, 8 by >1 yr before priority → § 102(b); items 7 and 9 are foreign publications, § 102(a)/(b) depending on date).

3. § 102 analysis, reference by reference

Reference 1 — US 5,089,746 A (Varian Associates) — the only citation of record

• Full citation: Rosenblum & Doniger, "Production of ion beams by chemically enhanced sputtering of solids," US 5,089,746 A; issued 1992‑02‑18. (WO 90/10303 counterpart.) https://patents.google.com/patent/US5089746
• Disclosure: A plasma reaction chamber with anode + cathode, the cathode liner being rich in the element to be implanted (boron, As, P, Sb). A fluorine-containing, non-poisonous feed gas (CF₄, SF₆, etc.) is ionized; plasma constituents chemically react with the liner element; ions are extracted, mass-separated, and implanted into a semiconductor wafer.
• § 102 relevance: It discloses a method of generating ions in a chamber, extracting through an aperture, and implanting a wafer — i.e., the general method skeleton of '020 claim 1, and it touches the halogen-gas feature (claims 9–10).
• Does it anticipate claim 1? No. It nowhere discloses the limiting element of every '020 claim — "an inner wall of said chamber being formed of a metal which is resistant to chemical etching due to ions and radicals." In fact the '746 chamber relies on the reactive liner as the source of implant species, the opposite teaching. It is background, not anticipating art. It also does not anticipate dependent claims 17/21 (V/Nb/Ta wall) or 20/22/23 (coated wall / barrier metal / nitride).
• Confidence: high (full text retrieved).

Reference 2 — US 5,262,652 A (Applied Materials) — the most dangerous reference

• Full citation: Bright et al., "Ion implantation apparatus having increased source lifetime," US 5,262,652 A; filed 1992‑06‑15 (CIP of Ser. No. 699,874, 1991‑05‑14); issued 1993‑11‑16. https://patents.google.com/patent/US5262652
• Disclosure (verbatim claim language):
  • Claim 1: mounting a filament + reflector "so as to maintain an insulating gap between the reflector and the arc chamber," and mounting the filament insulators outside of the arc chamber.
  • Claims 9–13, 21–24: "said arc chamber is made of tungsten"; "one or more walls of said arc chamber is made of tungsten"; "removable tungsten liner."
  • Spec.: moving the insulators out of the chamber prevents a conductive coating; tungsten parts reduce contamination (it expressly discusses the Mo²⁺ (MW 98) ≈ BF₂⁺ (MW 49) contamination problem that '020 also raises).
• § 102 relevance:
  • Claim 13 ('020: "supporting insulation member … separated from a generation area of the plasma") → potentially anticipated — US 5,262,652 discloses exactly this (insulators mounted outside the arc chamber on the source body).
  • Claim 14 ('020: spacer between insulation member and filament + a space/gap; "x/y>1" spacing) → potentially anticipated — the reflector-mounted-with-insulating-gap-from-arc-chamber structure reads directly on the spacer/gap limitation.
  • Claim 1 / claim 20 → arguable. A tungsten-walled (or W-lined) arc chamber is "a metal," and tungsten removes the contamination just as '020 seeks; a § 102(b) case could be made that tungsten is "resistant to chemical etching by the ions and radicals." Expect the patentee to counter (as the '020 specification implies) that W itself is attacked/that the claim contemplates V/Nb/Ta and a lower-volatility metal-halide. Note W is not in '020 claim 17's list (V, Nb, Ta, Cr, Ti, Zr, Hf, Pt), so claims 17/21 are safe from this reference.
• Verdict: Strong § 102(b) art against claims 13 and 14, plausible against claim 1, not against 15/17/19/21–23.
• Confidence: high (claims and spec text retrieved).

Reference 3 — US 4,760,262 A (Eaton)

• Full citation: "Ion source," US 4,760,262 A; filed 1987‑05‑12; issued 1988‑07‑26. https://patents.google.com/patent/US4760262
• Disclosure: Hot-cathode side-extraction (Freeman-style) source; filament reflectors at each end perturb the cathode/anode field; filament mounted in insulators received in apertures of the cylindrical anode.
• § 102 relevance: Bears only on the filament/reflector/insulator dependent claims (11, 13, 14). It does not disclose a chamber wall metal chosen for chemical-etch resistance, so it cannot anticipate claim 1 or 17/19/21–23. At most cumulative with reference 2 for claims 13–14.
• Confidence: medium‑high (abstract/spec retrieved; not every claim inspected).

Reference 4 — US 5,252,892 A (Tokyo Electron)

• Full citation: "Plasma processing apparatus," US 5,252,892 A; filed 1989‑02‑16; issued 1993‑10‑12.
• Disclosure: A plasma processing apparatus (chamber-based). I could not retrieve the full claim set/description for this reference within my search budget.
• § 102 relevance: Plasma-processing chambers of this era frequently protect/coat inner walls to suppress contamination/etching; if the '892 specification discloses a specific inner-wall material chosen for plasma/etch resistance, it would be relevant to '020 claim 1 and possibly claims 20/23 (coating). I will not assert this without the text — flagged as candidate only.
• Confidence: low — needs verification.

Reference 5 — JPH 03‑219541 A (Tokyo Electron)

• Full citation: "Plasma processing device," JPH 03‑219541 A; filed 1989‑11‑20; published 1991‑09‑26.
• Disclosure: JP counterpart/sibling of reference 4.
• § 102 relevance: Same caveat as #4. Language-barrier + no full text retrieved. Candidate only.
• Confidence: low.

Reference 6 — US 5,198,677 A (U.S. DOE)

• Full citation: "Production of N+ ions from a multicusp ion beam apparatus," US 5,198,677 A; filed 1991‑10‑11; issued 1993‑03‑30.
• Disclosure: Multicusp (magnetic-bucket) ion source producing N⁺.
• § 102 relevance: Relevant to the generic "chamber + plasma → extract ions" concept, and possibly to filament/insulator mechanics. It does not disclose a halogen-etch-resistant chamber wall, so it cannot anticipate claim 1. Cumulative at best.
• Confidence: medium (title/date confirmed; claims not re-inspected).

Reference 7 — IT 1252811 B (Proel Tecnologie) — the closest "coated-wall" art

• Full citation: "Ion generator with ionization chamber built or coated with high secondary emission coefficient material," IT 1252811 B; granted 1995‑06‑28 (filed 1991‑10‑11). English equivalents: EP 0 537 123 A1; JP H05‑242820 A. https://patents.justia.com/assignee/proel-tecnologie-s-p-a
• Disclosure (verbatim): An ionization chamber whose "walls … and those of its internal electrodes, [are] constructed from or coated with a material with a high coefficient of secondary emission" — advantageously glass, or internal surfaces coated with a material (e.g. cesium, bismuth, or lead); cesium/bismuth/lead may also be added to the glass. Purpose: raise energy/mass yield by reducing electron loss at the wall.
• § 102 relevance:
  • It does disclose the structural feature "chamber wall built of, or coated with, a material" (→ touches '020 claim 20, "inner wall … coated with said metal," and the abstract concept of a coated chamber).
  • But the coating material set (Cs, Bi, Pb, glass) is chosen for secondary emission, not for resistance to halogen-ion chemical etching, and glass is not a metal. So it does not anticipate claim 1 (which requires (i) a metal and (ii) that it be resistant to chemical etching by ions and radicals). Note the symmetry worth flagging: Pb appears both here and in the '020 specification's "noble metal (Ni, Pt, Pb)" list — an examiner/opponent could exploit that overlap against a wall-material claim, though the reference's stated function differs.
• Verdict: Relevant to claims 1 (as a "coated wall" teaching) and 20, but anticipation is unlikely because the metallic + etch-resistant requirement is not disclosed.
• Confidence: medium‑high (the '281 disclosure was retrieved via its EP/JP equivalents; the language-of-record claims were not re-read verbatim).

Reference 8 — US 5,306,921 A (Tokyo Electron)

• Full citation: "Ion implantation system using optimum magnetic field for concentrating ions," US 5,306,921 A; filed 1992‑03‑02; issued 1994‑04‑26.
• Disclosure: Ion implanter architecture using an optimized magnetic field for ion concentration.
• § 102 relevance: System/beamline-level art. It does not disclose a chemically-etch-resistant chamber-wall metal, so it cannot anticipate claim 1 or the material claims. Cumulative on the implanter-context steps only.
• Confidence: medium.

Reference 9 — FR 1,402,020 A (CSF)

• Full citation: "Perfectionnements aux sources d'ions" (Improvements to ion sources), FR 1,402,020 A; filed 1964‑04‑27; published 1965‑06‑11.
• Disclosure: Early ion-source improvement (1965 vintage). Full text not retrieved.
• § 102 relevance: Likely generic ion-source structure; almost certainly silent on halogen-etch-resistant refractory-metal chamber walls. Its inclusion is notable mainly because its number happens to fall inside the "…020" family searched. Candidate/background only.
• Confidence: low — needs verification.

4. Bottom line — which claims are most exposed under § 102

Overall: The references of record and the family-cited art do not anticipate the patent's differentiating idea — an inner wall made of V/Nb/Ta-class metal (or a coated/nitrided/barrier-metal wall) selected for resistance to halogen-ion/radical chemical etching in an ion source. The only serious § 102 exposure is claims 13–14 (and, on a § 102(b) theory, claim 1) via US 5,262,652's outside-the-chamber insulators and tungsten-lined arc chamber. This is consistent with why the examiner allowed '020 over the single cited reference US 5,089,746.

5. Sources & confidence

• Authoritative patent text (claims, spec, family, citation lists): https://patents.google.com/patent/US5640020/en (fetched for this analysis).
• US 5,089,746 full text/claims: https://patents.google.com/patent/US5089746 and https://patents.justia.com/patent/[5089746](/patent/5089746)
• US 5,262,652 claims/description: https://www.everypatent.com/comp/pat5262652.html and https://patents.google.com/patent/US5262652
• US 4,760,262: https://patents.google.com/patent/US4760262
• IT 1252811 B / EP 0 537 123 A1 / JP H05‑242820 A: https://patents.justia.com/assignee/proel-tecnologie-s-p-a (Proel "high secondary emission coefficient" chamber)
• Unverified within budget (flagged, not asserted): US 5,252,892 A, JPH 03‑219541 A, US 5,198,677 A, US 5,306,921 A, FR 1,402,020 A — titles/dates confirmed; full claims/description not re-read here. Treat their § 102 characterizations as provisional.

Explicit uncertainty: I could not pull the literal front-page "References Cited" block from USPTO PatentCenter directly (only the Google Patents mirror). The "one of-record citation" statement rests on the Google Patents "Patent Citations (1)" and "Citations (1)" blocks, which agree with each other and with the full text you supplied. If a PatentCenter fetch surfaces additional of-record references (or a different examiner citation set for the parent US 5,656,820), the tiering above should be re-checked — the substantive § 102 conclusions for claims 13–14 and claim 1 (US 5,262,652) would be unaffected.

I'll ground this analysis in the actual prior-art references of record for this patent. Let me pull details on the key cited references.

Let me pull the remaining family-cited references to complete the prior-art picture.

Obviousness Analysis — U.S. Patent No. 5,640,020 under 35 U.S.C. § 103

Note on scope and the record: This is a pre-AIA patent (priority 1994-11-18, filed 1996-09-16 as a continuation of Ser. No. 08/560,029), so pre-AIA § 103 governs. The provided page's "Patent Citations (1)" section lists only US 5,089,746 A as a reference of record; the remaining references appear under "Family Cites Families (8)" (prior art cited in the prosecution of the JP/GB/US family members), meaning they are available as § 103 art for a challenge even though the '020 claims themselves were likely examined against a narrow set. I flag below where I could not retrieve substantive content and where dates create § 102 uncertainty.

Also flagging, consistent with the prior sections: the patent is Expired – Lifetime (anticipated 2015-11-17), so this is a validity-commentary / ex parte-reexamination-style analysis rather than a live-litigation defense. And I do not auto-correct any identifiers.

I. Governing framework

Under Graham v. John Deere and KSR Int'l v. Teleflex, the § 103 inquiry requires: (1) the scope and content of the prior art; (2) the differences between the prior art and the claims; (3) the level of ordinary skill; and (4) objective evidence of nonobviousness. Critically for this patent, KSR holds that:

• A combination is obvious where there is a known problem and a finite number of identified, predictable solutions, and the combination yields no more than expected.
• A "result-effective" limitation (a property that inherently follows from the selection of a known material) does not confer patentability. In re Kao; MPEP 2144.02.
• Material substitution among known corrosion-resistant metals is a paradigm "design choice" unless unexpected results are shown (MPEP 2144.04).

Level of ordinary skill (attributed): a B.S. in EE/EE-materials/Applied Physics (or equivalent) with ~2–4 years working on ion implanters/plasma sources, OR a lower degree with substantially more experience; familiar with Freeman/Bernas arc chambers, thermionic filaments, halogen dopant feed gases (BF₃, SiF₄, PF₃, AsF₃), and the contamination/lifetime economics of source components.

The claim at issue: Only claim 1 is independent; claims 2–23 depend from it. Claim 1 is a method claim (an "ion irradiation method") whose sole structural requirement is that "an inner wall of said chamber [be] formed of a metal which is resistant to chemical etching due to ions and radicals."

(Consistent with the earlier sections, I note but do not "fix" the internal inconsistency in printed claim 1 — it recites the sample "in a sample chamber outside said chamber" and then "irradiating the emitted ions onto the sample in said chamber.")

II. The prior art of record and what each reference teaches

Admissions in the '020 specification itself are also usable as prior art (MPEP 2129): the patent concedes the Mo arc chamber + BF₃/GeF₄ configuration, the resulting Mo erosion and W–Mo alloying of the filament, and that "GeF₄ is nowadays used as source gas." This is the admitted starting point.

III. Claim 1 — element-by-element mapping

Claim 1 recites six steps. The chart below shows how the combination of US 5,262,652 + US 5,089,746 (with Proel/US 5,198,677 as secondary) covers every limitation:

The only real gap — and why it is an obvious gap

The single difference between claim 1 as a whole and US 5,262,652 is the qualitative recitation that the wall metal be "resistant to chemical etching due to ions and radicals." Applied Materials '652 selected tungsten; the '020 prefers Nb/V/Ta as more resistant. But:

1. The problem was known and the solution predictably in the same class. US 5,262,652 itself frames the problem as chamber-wall reaction with fluorine/carbon-fluoride species from BF₃/SiF₄, and its solution is material substitution of the plasma-facing surfaces. Substituting one known chemically-resistant refractory metal for another (W → Nb/V/Ta) is the textbook "obvious to try" / "design choice" scenario (KSR; MPEP 2144.04).
2. A finite, predictable set of candidates. V, Nb, Ta, Cr, Ti, Zr, Hf and Pt (claim 17) are refractory / valve / corrosion-resistant transition and noble metals — a small, well-known menu, not an unbounded field.
3. US 5,089,746 supplies the "select the exposed solid by its fluorine chemistry" principle. Varian's whole process turns on forming a volatile fluoride from the intended solid; a PHOSITA reading '746 in the ion-source context understands the corollary that other exposed surfaces should be chosen so as not to form volatile fluorides. The '020's own rationale (Nb–F compound has lower vapor pressure than MoF₆) is the direct application of that known principle.
4. Proel supplies the "coat/build the wall from a selected material" mechanism, so even adding a coating layer to an existing Mo chamber (rather than making the wall solid Nb) was a known technique.
5. Reasonable expectation of success was high: if Mo/C are eroded by F and W is used to reduce contamination, a more inert refractory metal would be expected to reduce erosion and contamination still further.

Conclusion on claim 1: strong prima facie § 103 case over US 5,262,652 in view of US 5,089,746, further in view of Proel (IT 1,252,811 B / EP 0 537 123 A1) and US 5,198,677.

Alternative primary-reference combinations (all viable)

• US 5,089,746 (primary) + US 5,262,652 + Proel: '746 supplies the halogen-plasma ion-implantation method; '652 supplies the "chamber wall / liner material chosen to reduce erosion and contamination"; Proel supplies wall coating. Motivation: all three address the same halogen-plasma/corrosion problem in the same field.
• US 5,262,652 (primary) + US 5,089,746: the cleanest combination — '652 already contains almost every limitation and expressly diagnoses the exact Mo-contamination and insulator-coating problems the '020 addresses.
• US 5,262,652 + US 5,198,677: '652 + '677 together teach inner liners of Mo/W in ion-source chambers plus the insulator-outside-chamber architecture (relevant to claims 13–14).

IV. Dependent claims 2–23

V. The two claim families that give the patent its best survival odds

Claims 15, 19, and 22 are where a § 103 challenge is genuinely contestable, for two opposing reasons:

1. Challenge side (why they're still arguably obvious): Claims 15/19 recite result-effective properties. Once the wall metal is selected for resistance to fluorine etching, a metal whose fluoride has a lower vapor pressure than MoF₆ (and lower than the filament's fluoride) inherently follows. Under In re Kao / MPEP 2144.02, discovering an inherent property of a selected material does not confer patentability. Likewise claim 22's "barrier metal to prevent seepage" is a predictable fix for a known high-temperature interdiffusion symptom, using standard diffusion-barrier technology.

2. Patentee side (objective evidence, § IV "Graham factor 4"): The '020 specification reports that the prior-art tungsten filament survived only ~3 days in a GeF₄-fed Mo chamber, whereas the Nb-coated chamber extended filament life to ~20 days — an order-of-magnitude result — and that the industry had instead been using Mo/C/W walls and treating GeF₄ as problematic. That is classic unexpected results + long-felt need evidence. A patentee would argue that the prior art, if anything, steered the field toward tungsten (US 5,262,652's express preference) and to graphite/molybdenum (admitted prior art), not to Va-family metals, and that the ~20× lifetime gain is evidence the selection was not routine.

Balanced bottom line on the patent as a whole: the independent claim and most dependents look vulnerable, but claims 15, 19, 22 (and to a lesser extent 17, 21, 23) would give a patentee meaningful room to argue nonobviousness — largely because those claims capture the specific quantitative insight (fluoride-vapor-pressure ordering) rather than the generic "use a corrosion-resistant wall metal" concept.

VI. Motivation-to-combine summary (why a PHOSITA would combine)

No reference teaches away. There is no evidence of a criticality requirement beyond the (arguable) vapor-pressure ordering of claims 15/19.

VII. Confidence and caveats

• High confidence: substantive content and dates for US 5,262,652 and US 5,089,746, and for US 5,198,677; the Proel disclosure (EP 0 537 123 A1 / JPH 05242820 A). These drive the claim-1 and claims 13–14 analysis.
• Explicit date caveat: the IT 1,252,811 B grant publication (1995-06-28) post-dates the '020 critical date (1994-11-18), so the Italian document is not itself § 102(a)/(b) art. Its family members EP 0 537 123 A1 (pub. 1993) and JPH 05242820 A (pub. 1993-09-21) do predate the critical date and carry the same disclosure; I rely on those, not the IT publication.
• Low confidence / not retrieved: the substantive contents of US 5,252,892 A, JPH 03219541 A, US 5,306,921 A, US 4,760,262 A, and FR 1,402,020 A. I confirmed only titles/dates via the family-citation list; I have not verified their teachings and do not rely on them for any limitation. A full § 103 challenge would require pulling each.
• No prior-art search performed beyond the page's citation lists. This analysis is limited to the references of record (US 5,089,746 A) and the family-cited references, per the task instruction. Real-world art to the contrary (e.g., other refractory-metal-wall ion sources in the 1993–1994 window) could strengthen or weaken these combinations.
• Pre-AIA § 103 applies; the AIA "reasonably pertinent" framing is not controlling, though the result here would likely be the same.
• Reminder that the patent is expired, so this analysis bears on historical validity / possible ex parte reexamination, not on an active infringement dispute — consistent with the earlier sections' finding of no litigation located.