US 11402343

I'll search for authoritative information on US patent 11402343, including any litigation or CAFC 2026 activity.

Let me run a few more targeted searches for litigation and USPTO records specific to this patent number.

US Patent 11,402,343 B2 — Verification Summary

Note on litigation/CAFC search: I searched for this patent number in connection with USPTO records and Federal Circuit 2026 dockets. I found no CAFC 2026 docket, appeal, IPR, or district-court litigation involving patent 11,402,343. Searches combining "11402343," "11,402,343," and "litigation/docket/CAFC" returned either unrelated patents (e.g., 11,582,343 / 11,597,402) or generic results. Google Patents lists no "Cited by examiner" litigation, and the legal-events log shows only routine prosecution, assignment, and maintenance-fee entries. I state this as a negative finding based on searches — not as a certified clearance.

---

Bibliographic Data

Data-quality flag: A third-party aggregator (patentleaderboard.com) labels these inventors under "IBM" and "Hitachi." That conflicts with the authoritative record (Rigaku Corporation, Tokyo) and appears to be an unreliable name-matching artifact. I do not treat it as accurate.

---

Abstract (verbatim)

"An X-ray fluorescence spectrometer of the present invention includes a counting time calculation unit (13) configured to: by a predetermined quantitative calculation method, determine each of quantitative values by using reference intensities of one standard sample and repeatedly perform a procedure of determining each of the quantitative values in a case where only a measured intensity of one of measurement lines is changed by a predetermined value, to calculate a ratio of a change in each of the quantitative values to the predetermined value as a quantitative-value-to-intensity change ratio, the one of the measurement lines having the measured intensity to be changed being different on each repetition of the procedure; and use quantitative-value-to-intensity change ratios calculated thereby for all the measurement lines to calculate a counting time for each of the measurement lines from a quantification precision specified for each of the quantitative values."

---

Plain-Language Overview of the Independent Claims

Background problem: Existing XRF spectrometers picked counting times assuming a measurement line's intensity precision maps one-to-one onto a component's quantitative precision. That assumption breaks down for high-content elements (e.g., Cr in stainless steel, where the calibration curve is convex) and for thin-film thickness/composition analysis by the fundamental parameter (FP) method, because one line affects several quantitative values and vice versa.

Claim 1 — Counting time calculation unit (the "forward" direction: precision in → counting times out)

Given a specified quantification precision for each quantitative value, the spectrometer computes the counting time needed for each measurement line:

1. Measure standard samples in a provisional counting time to derive calibration-curve constants/correction coefficients or FP instrument-sensitivity constants.
2. Using one standard sample, set each measurement line's measured intensity as a reference intensity.
3. First procedure (perturbation/noise-sensitivity analysis): one at a time, perturb only one measurement line's intensity by a fixed amount, recompute all quantitative values, and record the ratio of the change in each quantitative value to the perturbation — the quantitative-value-to-intensity change ratio (F_ij). This creates a full cross-sensitivity matrix capturing coupling between lines.
4. For each perturbed line, divide the specified precision by the corresponding F_ij; take the tightest (smallest absolute value) result as the provisional necessary intensity precision.
5. Convert that to a counting time per line using the reference intensity (e.g., T = 1/(σ²·I·1000)).
6. Compute an estimated quantification precision for each value (σ²_Wj = Σ F²_ij σ²_reli) and compare it to the specification (allowance e.g. ≤103%).
7. If all pass → output final counting times. If not, second procedure: for each failing quantitative value, hypothetically add a fixed time (e.g., 1 s) to one line at a time, recompute the estimated precision, treat the precision improvement as the "expected improved quantification precision," and derive the necessary additional time by scaling the residual precision gap by that improvement.
8. Add time only to the line with the shortest necessary additional time, by a multiple of that time, and update the estimate.
9. Iterate until the updated precision meets the specification; if other quantitative values still fail, repeat for the worst offender (highest ratio of estimated to specified precision).
10. Round the final counting time of each line to set digits/units and output.

Claim 2 — Quantification precision calculation unit (the "reverse" direction: counting times in → precision out)

Structurally mirror-image: given a specified counting time per measurement line, the spectrometer (i) determines calibration constants or FP sensitivity constants from standard samples, (ii) sets reference intensities for one standard sample, (iii) performs the same perturbation procedure to obtain the F_ij cross-sensitivity ratios, then (iv) computes each line's intensity precision from the specified counting time and reference intensity, converts that through the F_ij ratios into a quantification precision for each quantitative value, and outputs it.

Core inventive concept common to both claims: the F_ij cross-sensitivity matrix, empirically derived by digital perturbation through the chosen quantification algorithm, corrects the false assumption that a given quantitative value's precision depends only on one line's intensity precision. This makes optimal counting-time allocation (or precision prediction) work across sample types and quantification methods (calibration-curve or FP), including thin films where thickness and composition are coupled.

---

Confidence and Caveats

• High confidence on bibliographic data, abstract, and claim scope — these are taken directly from the full patent text and Google Patents record.
• Negative finding, moderate confidence: No CAFC 2026 docket or litigation identified. I could not directly query PACER, CourtListener, or USPTO Patent Center for this specific number within the search budget, so absence of litigation results should be read as "not found in searched sources," not as definitive proof none exists.
• The patent's two claims are both independent; there are no dependent claims.

I'll search for litigation involving US patent 11402343 across patent litigation databases.

The initial searches returned irrelevant results. Let me refine my search to patent litigation databases specifically.

None of these results relate to the patent. Let me try more targeted searches on specific litigation databases.

Search Results: Litigation Involving U.S. Patent No. 11,402,343

No litigation involving U.S. Patent No. 11,402,343 ("X-ray fluorescence spectrometer," Rigaku Corp.) was found.

I ran multiple targeted searches across general web indexes and litigation-focused sources (including queries pointed at the Unified Patents litigation portal, PACER-derived dockets, and CAFC coverage). None returned any district court case, CAFC appeal, ITC investigation, or PTAB proceeding naming this specific patent.

Case-by-case table

There are no cases to list.

What the searches did surface (and why it is not the patent)

• Patent #11402343 appeared once — on Patent Leaderboard (patentleaderboard.com) in the inventor profile for Takao Moriyama, listing "11402343 – X-ray fluorescence spectrometer – Yoshiyuki Kataoka – 2022-08-02." This confirms the patent's bibliographic data but is not a litigation record.
• Numerous hits for Unified Patents litigation/IPR activity and other patent suits (e.g., IPRs involving U.S. Patents 10,311,073; 10,621,228; 8,971,914; 9,253,239; 9,848,295; 8,605,794) are false positives matching on unrelated case numbers, party names, or the "Unified Patents portal" text — none reference U.S. 11,402,343.
• A docket-snippet hit ("Hangzhou Yingsen Technology... (1:26-cv-02343)") matched only on the partial string "02343" in a case number and is unrelated to this patent.

Context and caveats (important, per your "do not fabricate" instruction)

• Likelihood: U.S. 11,402,343 is an analytical-instrument patent owned by Rigaku Corporation (a spectrometer maker), issued 2022-08-02 from PCT/JP2019/047933. Such patents are typically practiced internally by the manufacturer and its competitors rather than asserted by non-practicing entities, making litigation less likely — but this is an inference, not a confirmed fact.
• Search limitations I must disclose: I did not get direct, positive confirmation from PACER, the Unified Patents litigation caselist, Docket Navigator, or Lex Machina for this specific patent number. My tool access returned indexed web results only, and some of those sources (e.g., PACER, paywalled docket databases) are not fully queryable through them. A definitive "no litigation" conclusion would require a direct patent-number search in PACER/Docket Navigator/Lex Machina.
• I did not find any similar number's results that I am presenting as this patent's — everything unrelated was excluded.

Recommended verification steps

1. Run the exact number 11,402,343 (or "11402343") as a party/patent filter directly in Lex Machina, Docket Navigator, or Unified Patents Portal → Litigation.
2. Search PACER Case Locator for "11402343" as a patent-in-suit term.
3. Check the CAFC docket and USPTO PTAB (IPR/PGR) for the patent number — none appeared in my searches, but direct database queries are more reliable than indexed web search.

I'll verify the structured "no proceedings on file" finding against live sources before drafting.

Let me run a couple more confirmatory searches specifically on the PTAB/PTAB E2E for this patent number and on Rigaku XRF PTAB activity.

PTAB / AIA Trial Analysis — U.S. Patent No. 11,402,343 B2

Cross-reference to prior sections: This is consistent with the earlier-generated summary and litigation sections, which likewise found no PTAB, district-court, ITC, or CAFC activity for this patent number. No contradiction to flag.

---

Proceedings overview

Total AIA trial proceedings on file: 0 (0 active, 0 with claims invalidated, 0 with claims sustained, 0 settled, 0 institution-denied). The USPTO Open Data Portal "PTAB proceedings on file" block returns no IPR, PGR, or CBM for U.S. 11,402,343 — and my independent web searches surfaced no proceeding either. The defensive posture this gives a defendant is the opposite of hardened-by-IPR: the patent is entirely untested at the Board, with both of its independent claims fully intact and no petitioner estoppel available to borrow. A defendant cannot point to any PTAB cancellation, any surviving-claims roadmap, or any adverse FWD. Everything — claim construction, prior-art mapping, and § 101/§ 112 theories — remains open and must be won from scratch.

---

Proceedings

None. There is no proceeding number to list. Per the constraints of this task, I will not invent one. The following is the evidentiary basis for the null result:

False positives rejected (do not confuse these with our patent): every near-number hit in the search results belongs to a different patent — U.S. 11,582,343 (Red Hat v. Competitive Access Systems, multipath communications); U.S. 11,006,343 (T-Mobile v. Dali Wireless); U.S. 11,677,798 (a third-party IPR); U.S. 6,603,343 (Apple/LG/Samsung v. Arigna); and U.S. 6,212,343 (Ricoh ITC matter). None involve Rigaku or X-ray fluorescence. A search snippet matching "IPR2024-00043" was a docket string coincidence, not a proceeding on U.S. 11,402,343.

Search-limitation disclosure (per the "do not fabricate" instruction): my confirmation is index-based — I did not execute a direct patent-number query inside PTAB E2E / PTAB Center, Docket Navigator, Lex Machina, or Unified Patents' PTAB portal. A definitive negative requires that direct query. I state "no proceedings" as a well-supported finding, not as a certified clearance.

---

Strategic summary

Claim status: both claims UNTESTED. U.S. 11,402,343 issued 2022-08-02 with exactly two claims, both independent: claim 1 (counting-time calculation unit — specified precision in, per-line counting times out) and claim 2 (quantification-precision calculation unit — specified per-line counting times in, precision out). Neither has ever been construed by the Board, and neither has been canceled, disclaimed, or amended in any AIA trial. Contrast this with the prior sections' finding of no litigation: the patent has simply never been challenged anywhere.

Estoppel landscape: none exists. Because no IPR/PGR was instituted, 35 U.S.C. § 315(e)(2) estoppel is not triggered for anyone — there is no petitioner, no privy, and no RPI bound by any ground. Every § 102/§ 103 ground a defendant can find is available both at the Board and in district court, subject only to the ordinary statutory bars: § 315(b)'s one-year clock from service of a complaint alleging infringement, § 325(d) (art already before the examiner), and General Plastic / Fintiv-type discretionary considerations if parallel litigation exists. There is no prior FWD whose findings could be adopted, and equally no prior FWD whose adverse claim construction a defendant would have to litigate around.

Pattern signals: absent. No serial petitioner, no defensive aggregator (Unified Patents, RPX, etc.) in the chain, and no PTAB appeal by Rigaku — because there has been no proceeding to appeal. The prior sections' inference holds and is reinforced here: this is an internally practiced analytical-instrument patent from a Japanese manufacturer (Rigaku Corporation), not an NPE-asserted portfolio patent, which is the ordinary explanation for why a four-year-old patent of this kind has drawn no IPR. Method-claim subject matter also matters strategically: IPR reaches only § 102/§ 103 grounds built on patents and printed publications — so a § 101 eligibility attack on these "spectrometer comprising a calculation unit" claims, or a § 112 indefiniteness attack, is not available at the PTAB and would have to be litigated in district court or the ITC.

---

Recommended next steps

• If you are a defendant: there is no FWD to link or quote — there is no proceeding. Your invalidity work is unconstrained. Start from the prior art already of record: the patent's own IDS cites, among others, JP 2000-65765 (Rigaku Industrial — the admitted-conventional counting-time approach described in the Background), US 2003/0118148 A1 (Rigaku Industrial), JP H05-264480 A, JP 2003-270177 A (Horiba — measurement-time setting), WO 2005/106440 A1 / US 7,382,855 B2 (Matsushita), and US 2017/0322165 A1 (Rigaku), plus the non-patent Ekinci et al. (1999) thickness-by-intensity-ratio paper. The core inventive hook — the empirically derived quantitative-value-to-intensity change ratio matrix F_ij and the iterative "necessary additional time" allocation of claim 1 — is the place to focus a § 103 obviousness theory.
• Timing if you file an IPR: the statutory one-year FWD deadline runs from institution, and the § 315(b) one-year bar runs from service of an infringement complaint. With no prior IPR, you are free to be the first petitioner — but weigh General Plastic/Fintiv discretion if a parallel district-court case is already scheduled for trial.
• If you want a certified negative on PTAB activity: query the patent number 11,402,343 directly in PTAB E2E / PTAB Center, and cross-check Docket Navigator, Lex Machina, or the Unified Patents PTAB portal. Indexed web search cannot substitute for those direct database queries.
• Absence as a signal: no IPR four-plus years post-issue on a currently-maintained patent (4th-year fee paid 2026-01-21, per the earlier bibliographic section) is consistent with low assertion pressure. Treat that as context, not as immunity — the absence of PTAB activity says nothing about the claims' actual validity.

Sources for verification: PTAB E2E (https://ptacts.uspto.gov/), USPTO Patent Trial and Appeal Board Decisions (https://www.uspto.gov/patents/ptab/decisions), and CourtListener (https://www.courtlistener.com/) — note that no opinion exists to cite for this patent, and I have deliberately not fabricated a proceeding number, panel, or disposition.

I'll verify the assignment record for US 11,402,343 through the USPTO Assignment Center and cross-reference sources.

Ownership Reconstruction — US 11,402,343 B2 ("X-ray fluorescence spectrometer," Rigaku)

Sourcing note (stated plainly): I could not open a live session of the USPTO Assignment Center (assignmentcenter.uspto.gov) — my tooling returns indexed web results only, and that portal is not queryable through it. The reel/frame and event data below are taken from the Google Patents legal-events record for US 11,402,343 (which mirrors the Assignment Center's "Patent Assignment Abstract of Title" fields), cross-checked against the patent front matter. Where a field was not exposed in the indexed record — specifically the correspondent of record — I say so rather than reconstruct it.

---

Inventors

Pattern assessment: No unusual inventor pattern. Both are long-tenured Rigaku XRF instrument-development staff named across a portfolio of Rigaku analytical-instrument filings. There is no evidence of either inventor departing the original assignee within 12 months of filing, and no evidence of inventor-held title that was later sold. Both are employee-inventors whose rights were confirmatorily assigned to the corporate applicant (see timeline).

Data-quality flag (carried forward, reconfirmed): Third-party aggregators mislabel these inventors. patentleaderboard.com files Takao Moriyama under "IBM" (24 patents) and Yoshiyuki Kataoka under "Hitachi" (31 patents). This is a same-name matching artifact: there is a different Takao Moriyama who is an IBM inventor (e.g., US 10,448,202 B2 "Dynamic boundary setting," IBM), and the aggregator has merged the two people. The patent-at-issue inventors are the Rigaku Moriyama and Kataoka (both lists do correctly show Rigaku as a secondary employer). Do not treat the IBM/Hitachi attributions as accurate.

---

Original assignee

Rigaku Corporation — 3-9-12 Matsubara-cho, Akishima-shi, Tokyo 196-8666, Japan (XRF engineering historically at 14-8 Akaoji, Takatsuki, Osaka).

• Business: Specialty manufacturer of scientific/analytical instrumentation built around X-ray technology — X-ray diffraction (XRD), X-ray fluorescence (XRF), X-ray transmission, and related analytical systems. Founded 1951; sells into semiconductors, electronics, batteries, pharma, steel, and cement markets in 90+ countries.
• Product embodying the claims: Yes. Rigaku is the manufacturer of the wavelength-dispersive sequential XRF spectrometer lines (e.g., the ZSX family) described in Fig. 3 of the patent — i.e., the goniometer/spectroscopic-device/detector architecture claimed. The claimed counting-time optimization is instrument firmware/software behavior, which Rigaku ships with its XRF platforms. The assignee is therefore a practicing operating company, not a holding vehicle.
• Current status: Operating, publicly listed — but under a new parent. Carlyle acquired Rigaku in 2021; on 2024-10-25 the group listed on the Tokyo Stock Exchange Prime Market as Rigaku Holdings Corporation (ticker 268A), headquartered at Rigaku's Akishima address, with Carlyle's holding vehicle Atom Investment, L.P. retaining roughly 42% post-listing. Importantly, that 2024 IPO/2021 LBO moved shares in the parent, not patent title — plain title to this patent was never re-recorded, which is why the assignment chain below terminates in 2021. Rigaku is not in bankruptcy and has not been dissolved.

---

Assignment timeline

Total recorded assignments for this patent: one (1). Every other entry in the legal-events log is a prosecution/publication/status event or a maintenance-fee payment, not a conveyance.

• 2021-07-15 (executed) / recorded 2021-10-01 — Reel 057672 / Frame 0952
  • Conveyance: Assignment of assignors' interest (see document for details) — i.e., a plain title assignment, not a security agreement, license, merger, or change of name.
  • Assignors: KATAOKA, YOSHIYUKI; MORIYAMA, TAKAO
  • Assignee: RIGAKU CORPORATION, Japan
  • Correspondent: Not retrieved / not exposed in the indexed record. I did not locate the attorney or firm of record for this reel/frame and will not guess one. (Flag: if a correspondent is recoverable from the Assignment Center abstract for 057672/0952, it should be recorded here — see verification steps. On the face of the record this is a single-link chain, so the "repeat correspondent" heuristic has nothing to recur against within this patent.)
  • Context: Internal confirmatory assignment — PCT applicant Rigaku Corporation took formal title from its two employee-inventors as part of US national-stage entry (application 17/599,228; national-stage entry events dated just before recording). Routine; not an acquisition, fire-sale, or transfer to an asserter.
  • Related non-conveyance events for context: 2021-09-28 FEPP entry setting owner entity status to large entity/undiscounted; 2022-05-26 Notice of Allowance; 2022-08-02 patent granted; 2026-01-21 4th-year maintenance fee paid (large entity, code M1551) — the fee payment is direct evidence the patent is still being maintained in force by the original corporate owner.

No other assignments exist. There is no change-of-name record, no merger record, no security agreement, and no license record for this patent — even though the assignee's corporate parent changed (Carlyle 2021, IPO 2024). That absence is itself informative: no entity ever took a title interest in this patent other than Rigaku Corporation and its employee-inventors.

---

Timeline diagram

timeline
    title Ownership of US 11402343
    2019 : PCT filed by Rigaku Corporation
    2020 : JP 6732347 B1 granted
    2021 : Inventors assign title to Rigaku Corp
         : Carlyle acquires Rigaku
    2022 : US 11402343 B2 issued
    2024 : Rigaku Holdings lists on Tokyo exchange
    2026 : Fourth year maintenance fee paid

---

NPE / troll-pattern signals

---

Verdict

Insufficient data (per the rubric's own definition: "no records, or only the original assignment") — but read it as benign, not unknown. The complete recorded chain is a single confirmatory employee→employer assignment, executed 2021-07-15 and recorded 2021-10-01 at Reel 057672 / Frame 0952, moving title from inventors Kataoka and Moriyama to Rigaku Corporation, which remains the owner today (4th-year maintenance fee paid 2026-01-21, large-entity code M1551). None of the eight NPE signals is present, no asserter appears in the chain, and no post-issuance conveyance of any kind exists — so there is no NPE chain to score, even though the assignee is a healthy operating company rather than an unknown party.

Why not "Operating-company assertion": that category requires the current assignee to be suing actual competitors over the claims. Rigaku is an operating company that ships the claimed spectrometer architecture, but there is no identified litigation against competitors naming this patent (see the litigation section's negative finding). Ownership posture = practicing operating company, non-asserting on this record.

Why not "NPE — high/moderate confidence": zero signals present. Zero strong signals, zero weak signals. Nothing to support it.

---

Verification steps (to close the one gap)

1. Open the Assignment Center abstract for application 17/599,228 / patent 11,402,343 and read Reel 057672 / Frame 0952 directly to capture the correspondent of record — the only field I could not retrieve, and the one field that would let signal #3 be scored definitively: https://assignmentcenter.uspto.gov/ (patent-number search; legacy interface at https://assignment.uspto.gov/patent/index.html).
2. Confirm the abstract shows Total Assignments: 1 and cross-check the legal-events log for any 2024–2026 change-of-name or merger record that Google Patents may not have mirrored.
3. If asserting a "no-NPE" conclusion for a portfolio analysis, also check the JP 6732347 B1, EP 3951370 B1, CN 113692533 B, and WO 2020/202644 A1 siblings — foreign registers would show a parallel outbound transfer that the US record alone would miss.

I'll verify the patent record and the cited references before drafting the prior-art analysis.

Let me verify the remaining cited references.

Verification of the patent record

I confirmed the target patent before analyzing its prior art:

• US 11,402,343 B2 — "X-ray fluorescence spectrometer" — Rigaku Corporation; inventors Yoshiyuki Kataoka and Takao Moriyama; App. No. 17/599,228 (US national stage of PCT/JP2019/047933); priority JP 2019-065572 (2019-03-29); issued 2022-08-02. Record is consistent across the authoritative full text and Justia.

Methodology / limitation disclosure: The USPTO patent-number search and partial web verification I ran returned the correct patent (no similar-number substitution). My reference-by-reference analysis below is grounded primarily in the examiner citation list embedded in the authoritative patent record ("Citations (11)" and "Patent Citations (15)"), which is the definitive list of references of record for this patent. I verified several key references directly (JP2000065765A via J-GLOBAL; US6118844A via Google Patents; US20030118148A1 via Google Patents; JP2003270177A via Google Patents/J-GLOBAL; WO2005106440A1). I hit the search-step limit before independently verifying every one of the eleven, so descriptions of the least-verified references (JPH05264480A, JP2001133419A, JP2009097957A, JP2014035334A) are derived from their titles/assignees/dates as recorded in the patent, and are flagged as such.

Anticipation framing (important): Under 35 U.S.C. § 102, anticipation requires a single reference disclosing every element of a claim, arranged as claimed. As shown below, no cited reference fully anticipates either claim; the cited art is best understood as § 103 material. Both claims require the two-part inventive core — (a) the empirical perturbation-derived quantitative-value-to-intensity change ratio matrix F_ij generated across all measurement lines by the chosen quantification algorithm, and (b) its use either to allocate per-line counting time (claim 1, including the iterative "necessary additional time" loop) or to predict per-value precision (claim 2). I have mapped each reference to the claim element(s) it touches, and stated explicitly where anticipation fails. All listed references predate the 2019-03-29 priority date and are therefore available as § 102(a)(1)/102(a)(2) and § 103 art.

---

Master table of cited prior art

---

Reference-by-reference analysis

1. JP H05-264480 A — Sumitomo Metal Industries (pub. 1993-10-12)

Citation: Japanese unexamined patent application H05-264480, filed 1992-03-17, published 1993-10-12, "Fluorescent X-ray analysis method and apparatus used for the method."
Description: An early fluorescent-X-ray analysis method/apparatus from Sumitomo Metal. It appears in the record as general background art for XRF quantitative analysis and method steps; I did not independently retrieve its full text.
§ 102 mapping: No anticipation of claim 1 or claim 2. It predates the perturbation/precision-allocation architecture by decades and does not disclose the F_ij change-ratio matrix, reference-intensity setting, or the counting-time/precision algorithms. Useful only as general-evidence background.

2. JP 2000-065765 A — Rigaku Industrial (pub. 2000-03-03) — the spec's "Patent Document 1"

Citation: JP 2000-065765 A, filed 1998-08-19, published 2000-03-03, Rigaku Industrial Co., "X-ray fluorescence analyzer" (verified via J-GLOBAL, JGLOBAL_ID 200903056331029115).
Description (verified): Measures peak intensity and background intensity of a representative sample (means 12); a first calculation means (14) calculates the measurement time that yields a specified analytical precision, and a display means (19) displays it. It computes a peak measurement time (and, when peak+background measurement is instructed, both peak and background measurement times) for the representative sample based on the stored peak/background intensities.
§ 102 mapping: This is the closest single reference to the preamble and several elements of claim 1, because it expressly discloses:

• measuring a representative sample to obtain peak/background intensities;
• specifying an analytical precision and calculating a counting time to achieve it (claim 1's "calculate a counting time … on the basis of the reference intensity" and "specified quantification precision" concepts).

Why it does not fully anticipate claim 1: It relies on the admitted one-line/one-value assumption (intensity ∝ content) and does not disclose (i) determining calibration-curve constants/correction coefficients or FP instrument-sensitivity constants from plural standard samples in a provisional counting time; (ii) the first procedure across all lines producing the F_ij quantitative-value-to-intensity change ratio; (iii) the provisional necessary intensity precision with smallest absolute value; (iv) the estimated-quantification-precision comparison; or (v) the iterative "necessary additional time" update loop. It likewise does not touch claim 2's reverse computation. Anticipation status: none for either claim; primary § 103 reference (the specification itself admits this as conventional art).

3. JP 2000-074857 A — Rigaku Industrial (pub. 2000-03-14)

Citation: JP 2000-074857 A, filed 1998-08-27, published 2000-03-14, Rigaku Industrial Co., "X-ray fluorescence analyzer."
Description: Same-assignee, same-era XRF analyzer filed eight days after JP 2000-065765; consistent with the family of counting-time/precision-based XRF analyzers.
§ 102 mapping: No anticipation. Same § 103 relevance as reference 2 — general evidence that specifying precision and computing counting time was known; discloses nothing of the F_ij perturbation matrix or the iterative allocation algorithm.

4. US 6,118,844 A — Helmut Fischer GmbH (granted 2000-09-12)

Citation: US 6,118,844 A, priority 1997-09-09 (DE 197 39 321.1), granted 2000-09-12, Helmut Fischer GmbH & Co., "Method and device for the determination of measurement uncertainties in X-ray fluorescence layer thickness" (verified).
Description (verified): Determines measurement uncertainty for an XRF layer-thickness gauge. For a given thickness d and measurement time t, it generates a simulated spectrum S(K) and injects a Poisson-distributed random contribution per channel via a random generator; the standard deviation σ(d) of the thicknesses obtained over repeated simulations is taken as the measurement uncertainty (reciprocal of repetition precision). The device thus predicts achievable precision from measurement time without repeated physical measurement.
§ 102 mapping: This is the closest cited reference to the direction of claim 2 (measurement-time in → uncertainty/precision out) and shares claim 2's subject matter (thin-layer thickness quantification precision), touching the "calculate … a quantification precision … on the basis of the specified counting time" concept.
Why it does not fully anticipate claim 2: Its mechanism is Monte-Carlo/Poisson simulation of a spectrum, not (i) determination of calibration constants or FP instrument-sensitivity constants from plural standard samples; (ii) setting one standard sample's measured intensities as reference intensities; (iii) the perturbation procedure producing the F_ij change ratios across all lines; or (iv) combining per-line intensity precision with F_ij via σ²_Wj = Σ F²_ij σ²_reli. Those are the distinguishing elements of claim 2. Anticipation status: none; strong § 103 reference for the "predict precision from specified counting time" concept.

5. JP 2001-133419 A — JEOL Ltd. (pub. 2001-05-18)

Citation: JP 2001-133419 A, filed 1999-11-02, published 2001-05-18, JEOL Ltd., "X-ray elemental analyzer."
Description: X-ray elemental analyzer background art (title/assignee/date taken from the record; full text not independently retrieved).
§ 102 mapping: No anticipation of either claim. General background; discloses no counting-time-allocation or precision-prediction algorithm of the claimed type.

6. US 2003/0118148 A1 — Rigaku Industrial Corp. (pub. 2003-06-26; granted US 6,668,038 B2)

Citation: US 2003/0118148 A1, filed 2001-12-06 (DE priority) / 2001-12-10 (US), published 2003-06-26, Rigaku Industrial Corp., "X-ray fluorescence spectrometer"; granted as US 6,668,038 B2 (2003-12-23) (verified).
Description (verified): XRF spectrometer for thin-film samples with a "measuring-line evaluating means" that, for each candidate secondary-X-ray line, calculates a first theoretical intensity at a specified thickness and composition of each layer, and a second theoretical intensity at that thickness/composition when the thickness or concentration has been changed by a predetermined quantity, then calculates a precision of the thickness or concentration based on the first and second theoretical intensities and determines applicability/inapplicability of using that line (with a display). It also contemplates selecting lines based on the calculated thickness/concentration precision.
§ 102 mapping: This is the single most relevant cited reference to the core computational idea shared by both claims — it discloses the conceptual germ of the F_ij procedure: determine a quantitative value at a reference state, perturb the variable by a "predetermined quantity," recompute, and derive a precision from the change. It reads directly on the claims' "determine each of the quantitative values … in a case where only the measured intensity … is changed by a predetermined value … calculate a ratio of a change … as a quantitative-value-to-intensity change ratio" in substance (though it frames the perturbation as a change in thickness/concentration, and the "precision" as a line-selection criterion).
Why it does not fully anticipate: It does not disclose: the perturbation being applied to each measurement line's intensity to build a full cross-sensitivity matrix F_ij over all lines; the use of those ratios for counting-time allocation from a specified precision (claim 1) or precision prediction from specified counting times (claim 2); the iterative necessary-additional-time loop; or the rounding/output step. Anticipation status: none in full; the leading § 103 reference for the perturbation/change-ratio element of both claims — a defendant's obviousness theory should center here combined with reference 2.

7. JP 2003-270177 A — Horiba Ltd. (pub. 2003-09-25)

Citation: JP 2003-270177 A, filed 2002-03-13, published 2003-09-25, Horiba Ltd., "Method for setting measurement time for fluorescent X-ray analysis under a plurality of conditions" (verified; inventors Sakuma & Sato).
Description (verified): Sets measurement times in an energy-dispersive XRF analysis performed under two tube-voltage conditions (low voltage for light elements, high voltage for heavy elements); it makes the low-voltage measurement time and high-voltage measurement time different lengths within a preset total measurement time, to obtain good-efficiency results quickly and reduce the standard deviation of quantified values.
§ 102 mapping: No anticipation. It touches the general concept of allocating measurement time across multiple measurement conditions (superficially akin to claim 1's per-line time allocation), but it does not disclose the F_ij change ratios, the specified-quantification-precision-to-counting-time inversion, or any precision-propagation formula. § 103 background only.

8. WO 2005/106440 A1 — Matsushita Electric (pub. 2005-11-10) / US 7,382,855 B2

Citation: WO 2005/106440 A1, filed 2004-04-28, published 2005-11-10, Matsushita Electric Industrial Co.; also US 2007/0248211 A1 (pub. 2007-10-25) and US 7,382,855 B2 (granted 2008-06-03), "Fluorescent X-ray analysis method and fluorescent X-ray analysis device" (verified in part).
Description (verified from the WO publication): Sets measuring conditions and starts measurement, then computes the measured concentration and the measurement precision of elements in the sample during measurement, and terminates the measurement when the measurement precision satisfies a predetermined precision condition, outputting the concentration at that point. The example shows an input precision condition ("is the precision ≤ 5% of the concentration?") and a setting of measurement time t with a maximum t_max = 200 s. Its ISR cites JP H08-043329 A / US 5,570,406 A (Horiba) as category "Y."
§ 102 mapping: Relevant to claim 2 (computing a quantification/measurement precision) and to claim 1's specified-precision objective (obtaining a measurement that meets a specified precision). It does not, however, disclose the F_ij change-ratio matrix, the per-line necessary-additional-time allocation loop, or the reference-intensity/standard-sample provisional-counting-time framework. Anticipation status: none for either claim; relevant § 103 secondary reference.

9. JP 2009-097957 A — JEOL Ltd. (pub. 2009-05-07) / US 7,742,565 B2

Citation: JP 2009-097957 A, filed 2007-10-16, published 2009-05-07, JEOL Ltd., "Analysis method using X-ray spectrum"; also US 2009/0310748 A1 (pub. 2009-12-17) and US 7,742,565 B2 (granted 2010-06-22).
Description: Analysis method using X-ray spectra (title/assignee/date from the record and Google Patents family listing; full text not independently retrieved).
§ 102 mapping: No anticipation of either claim; general spectral-analysis background with no counting-time-allocation or F_ij-precision-propagation content.

10. JP 2014-035334 A — Hitachi High-Tech Science (pub. 2014-02-24)

Citation: JP 2014-035334 A, filed 2012-08-10, published 2014-02-24, Hitachi High-Tech Science Corp., "Fluorescent X-ray analysis method and fluorescent X-ray analyzer."
Description: Fluorescent X-ray analysis method/apparatus (title/assignee/date from the record; full text not independently retrieved).
§ 102 mapping: No anticipation. Background art contemporaneous with the general field; discloses no F_ij perturbation matrix or precision/counting-time inversion as claimed.

11. US 2017/0322165 A1 — Rigaku Corp. (pub. 2017-11-09) / WO 2017/026200 A1

Citation: US 2017/0322165 A1, filed 2015-08-10, published 2017-11-09, Rigaku Corporation, "X-ray fluorescence spectrometer"; counterpart WO 2017/026200 A1 (verified to exist as a later-cited/related Rigaku XRF document).
Description: Rigaku XRF spectrometer of the same general wavelength-dispersive architecture; appears in the record as XRF-spectrometer background art and is the most recent cited reference (2017).
§ 102 mapping: No anticipation of either claim. As a Rigaku XRF spectrometer disclosure it supports the apparatus context (sample stage, X-ray source, spectroscopic device, detector, control unit) recited in the claims' preambles, but discloses none of the distinguishing counting-time-calculation or precision-calculation units/algorithm. § 103 background only.

Non-patent literature — Ekinci et al. (1999)

Citation: Ekinci et al., "Determination of the Sample Thicknesses By Intensity Ratio Measurement By Energy Dispersive X-Ray Fluorescence Spectrometry," Instrumentation Science & Technology, vol. 27, No. 3, pp. 181-189 (1999).
Description: Determines thin-sample thicknesses by intensity-ratio measurement using ED-XRF.
§ 102 mapping: No anticipation. It concerns thickness determination via intensity ratios (a calibration/quantification technique), not counting-time allocation from a specified precision or precision prediction from specified counting times, and discloses no F_ij cross-sensitivity matrix.

---

Synthesis

Best anticipatory candidates (and why none is a clean § 102 hit):

1. JP 2000-065765 A (Rigaku Industrial; "Patent Document 1") — best single-reference § 102 candidate against claim 1: it discloses measuring a representative sample and calculating a counting time to achieve a specified analytical precision. It fails to disclose the F_ij perturbation matrix and the iterative necessary-additional-time algorithm, so claim 1 is not anticipated by it alone.
2. US 2003/0118148 A1 (Rigaku Industrial) — best single-reference § 102 candidate for the perturbation/change-ratio element common to claims 1 and 2 (compute a value at a reference state, perturb by a "predetermined quantity," recompute, derive precision). It lacks the all-line F_ij matrix and both claims' specialized uses of it, so neither claim is anticipated.
3. US 6,118,844 A (Helmut Fischer) — best single-reference candidate against the "precision-out" concept of claim 2 (uncertainty of thin-layer thickness as a function of measurement time). Its random/Poisson simulation mechanism is materially different from the claimed F_ij-based propagation, so claim 2 is not anticipated.
4. WO 2005/106440 A1 / US 7,382,855 B2 (Matsushita) — secondary relevance to both claims (measurement-time setting tied to a precision condition; in-run precision computation), but missing the F_ij machinery.

Likely § 103 combination to focus a challenge: the primary disclosure of count-time-from-specified-precision (JP 2000-065765 A) + the perturbation/change-ratio line-evaluation technique (US 2003/0118148 A1) + precision-from-time prediction (US 6,118,844 A and/or WO 2005/106440 A1), with the Horiba multi-condition time-allocation (JP 2003-270177 A) as evidence that allocating time among multiple measurement conditions was known. That combination, if it renders the iterative necessary-additional-time loop and the cross-line F_ij propagation obvious, is the realistic route to invalidity — because no single reference supplies those elements, and (as established in the earlier PTAB section) there is no FWD, no IPR, and no litigation to date, so nothing has yet tested the claims.

Caveats:

• Descriptions for references 1, 5, 9, and 10 rely on the patent record's citation data plus titles/assignees/dates, not full-text retrieval; independent full-text review (JPO/J-PlatPat for the JP references, USPTO Patent Center for the family members) is recommended before relying on them in a validity opinion.
• "Potentially anticipates" is used strictly as a § 102 screening label. On the present record no cited reference anticipates either of claims 1 or 2 as a whole; each entry above identifies the claim element(s) it touches and the missing elements that defeat anticipation. I have not fabricated any disclosure not supported by the sources cited.

I'll verify the content of the key prior-art references from the "Prior Art"/Citations section before building the § 103 analysis.

Obviousness Analysis Under 35 U.S.C. § 103 — U.S. Patent No. 11,402,343 B2

Patent: "X-ray fluorescence spectrometer" · Rigaku Corporation · issued 2022-08-02 · priority 2019-03-29 (JP 2019-065572)

Cross-reference: This builds on the previously generated patent summary, litigation summary, and PTAB sections. Those sections established (a) the two-claim, both-independent structure and the F_ij core concept, and (b) the absence of any PTAB/litigation record. Nothing here contradicts them; the invalidity posture is analyzed entirely from the patent's own "Prior Art"/Citations record.

---

0. Method, and what this analysis is not

This section maps the record's cited art against the claim elements and identifies KSR-type rationales for combination. Two caveats stated up front:

1. I did not retrieve full texts of every cited reference. I verified the substance of the most load-bearing references (JP 2000-65765, US 2017/0322165 A1, US 2003/0118148 A1, US 6,118,844, JP 2003-270177, JP 2009-097957/US 7,742,565). Where I rely on a reference's title/assignee/date only, I say so.
2. Date discipline. Only art published before the 2019-03-29 priority date is § 102/§ 103 art. Several documents on the Google Patents page under "Similar Documents" and "Families Citing this family" (e.g., WO 2021/059597 A1, JP 7190749 B2, JP 7249666 B2, JP 7178725 B2, JP 7233756 B2, US 11,832,981 B2, CN 114641687 A) appear to post-date the priority date and are therefore flagged as not available as prior art without proof of earlier priority. See § 9.

Date note for the record: the task header states April 26, 2026; my operating context states 2026-09-29. This discrepancy does not affect the § 103 analysis, which is keyed to the 2019-03-29 priority date.

---

1. Claim element decomposition

Claim 1 (apparatus: XRF spectrometer comprising a counting time calculation unit), steps:

Claim 2 (apparatus: XRF spectrometer comprising a quantification precision calculation unit) is the structural mirror: elements 1.1, 1.3, 1.4, 1.5 are identical; then given a specified counting time per line, compute each line's intensity precision from the specified time and reference intensity, convert through F_ij to a quantification precision per value, and output it (single pass; no iterative allocation).

The only substantive delta over the admitted conventional art is: (i) the F_ij cross-sensitivity matrix obtained by digital perturbation through the chosen quantification algorithm, and (ii) in claim 1, a greedy iterative time-allocation loop built on that matrix.

---

2. Person of ordinary skill in the art (POSITA)

A POSITA here would be an engineer/physicist with a bachelor's degree in physics, analytical chemistry, or instrumentation plus 2–5 years' XRF experience — or equivalent — familiar with: wavelength-dispersive sequential XRF hardware (goniometer, analyzing crystal, detector), Poisson counting statistics and error propagation, the calibration-curve and fundamental-parameter quantification methods, and routine numerical methods including finite-difference sensitivity analysis. Notably, the record's own "Prior art keywords" — precision, intensity, quantitative, quantification, value — confirm the Examiner treated this as a crowded, well-explored field of precision/intensity bookkeeping.

---

3. Prior-art landscape (from the record)

Sources: https://patents.google.com/patent/JP2000065765A/en · https://patents.google.com/patent/US20170322165A1/en · https://patents.google.com/patent/US20030118148A1/en · https://patents.google.com/patent/JP2003270177A/en · https://patents.google.com/patent/[US6118844A](/patent/US6118844A)/en · https://patents.google.com/patent/[US7742565B2](/patent/US7742565B2)/en

---

4. Claim 1 — mapping and the two principal combinations

Combination A (primary): JP 2000-65765 + US 2017/0322165 A1 + JP 2003-270177 A

Combination B (alternative): JP 2000-65765 + US 2003/0118148 A1 + JP 2003-270177 A

US 2003/0118148 A1 supplies the same sensitivity idea from the other direction — it perturbs the quantitative value ("thickness or concentration changed by a predetermined quantity") and reads the theoretical intensity change to compute a precision. A POSITA inverting a Jacobian is doing nothing more than linear algebra; the patent's own claim treats F_ij as a ratio, and inverting a ratio is not inventive. This combination is weaker because US '148 is the transposed matrix, but it is a viable secondary theory.

---

5. Claim 2 — mapping

Claim 2 is claim 1 with the loop deleted and the input/output swapped:

Claim 2 therefore stands or falls with Claim 1 on the same F_ij + error-propagation core; if Claim 1 is obvious, Claim 2 adds only the deletion of an iterative loop — broadening, not narrowing, and hence a fortiori obvious once the F_ij matrix is in the art.

---

6. Motivation to combine (KSR rationales)

1. Same field, same problem, same assignee lineage. JP 2000-65765 (Rigaku Industrial) and US 2017/0322165 A1 (Rigaku) are the same corporate family as the patent's assignee. A POSITA at Rigaku would start with its own prior filings; combining a Rigaku counting-time method with a Rigaku sensitivity-matrix method is the paradigm of "combination of familiar elements according to known methods." KSR, 550 U.S. 398, 417 (2007).
2. Identified, finite set of known options. The record shows a small, well-known menu for hitting a precision target in XRF: (a) increase the counting time (JP '765, JEOL, Horiba); (b) model intensity↔precision sensitivity (US '148, US '165); (c) propagate counting statistics (Fischer). Selecting and combining these is not inventive.
3. Reasonable expectation of success. Digital perturbation of a deterministic quantification algorithm returns exact, reproducible sensitivity coefficients; first-order error propagation over a linearized model is standard. There is no unpredictability to defeat the combination.
4. Motivation supplied by the patent's own admissions. The Background concedes the conventional approach fails for convex calibration curves (Cr in stainless steel) and thin films, and states the object as "measurement in a suitable counting time and with a suitable quantification precision." Horiba's stated object — "短時間で…測定結果の定量値の標準偏差を小さくできる適正な測定時間を設定" (set an appropriate measurement time that shortens the run while reducing the standard deviation of quantitative values) — is the same object. That is an express teaching-direction, satisfying the "articulated reasoning" requirement.
5. Design incentive / market pressure. XRF throughput is a selling point (RoHS-style high-volume QC). Trimming overset counting times is a recognized efficiency driver — a classic "known technique to improve a similar device" rationale.

---

7. Why the residual differences do not confer nonobviousness

• "Only one measurement line changed at a time." This is finite-difference (one-at-a-time) sensitivity analysis — the default numerical Jacobian method when closed-form derivatives of the FP iteration are unavailable. Nothing more.
• "Smallest absolute value → provisional necessary intensity precision." A trivial "tightest constraint governs" selection.
• "Necessary additional time" linear scaling. First-order Newton/secant step on a monotone-in-time function; the patent's own worked example (Table 7, Repetitions 1–3) shows plain iteration to a tolerance.
• "Rounding up to whole seconds." Instrument-resolution design choice.
• "Re-measure standards if final time > provisional time." Conventional calibration hygiene, self-evident from the statistics.
• No new hardware. Both claims are pure software/firmware functional language operating on the admitted spectrometer of FIG. 3/FIG. 4.

---

8. Objective indicia (secondary considerations)

The specification contains no evidence of unexpected results, no comparative criticality beyond the admitted baseline, and no data showing that the F_ij approach solves something a POSITA could not have predicted. Table 8's achieved precisions (thickness 0.082 nm; Fe 0.039 mass%) are the consequence of applying the method, not evidence of nonobviousness. There is also no patentability nexus for any alleged commercial success — the accused/owner product line (Rigaku WDX sequential spectrometers) embodies the admitted JP '765 hardness too. As previously found, the patent has never been challenged at the PTAB or in court, so there is no adjudicated secondary-considerations record either way.

---

9. Anticipation check (§ 102) and date-flagged near misses

• No single reference appears to anticipate. JP 2000-65765 lacks the F_ij perturbation matrix and the iterative allocation; US 2017/0322165 A1 lacks the counting-time calculation and the iterative second procedure. The patent's novelty sits in the union, which is why this is a § 103, not § 102, analysis.
• Date-flagged items to verify before use: CN 114641687 A (a Rigaku-related Chinese publication) describes, on its face, both directions — "计算计数时间" (compute counting time from a specified precision) using σ_I = I·(σ_w/W), and the reverse ("指定了计数时间的情况下") computing analysis precision σ_C = W/(T·I·1000)^½ — i.e., the substance of both claims. I could not confirm its filing/priority date; CN 114641687 published around 2022 and appears to be a later Rigaku filing, in which case it is not § 102/§ 103 art against a 2019-03-29 priority date. Verify its priority date, because if it (or an earlier-parent WO/JP) predates 2019-03-29, it becomes the single most dangerous reference in the field — potentially anticipatory under § 102(b) and certainly combinable. The same date-check applies to WO 2021/059597 A1 and the JP 71xxxxx/72xxxxx family members listed as "citing" documents.
• Non-analogous-art defense is unavailable. Every reference is XRF or XRF layer-thickness metrology.

---

10. Weak points and expected counterarguments

1. Owner's best rebuttal: that no reference motivates feeding a sensitivity matrix into a counting-time optimizer, and that JP '765's one-to-one assumption positively teaches away from needing F_ij. Rebuttal: teaching-away requires a disparagement or a criticality the art lacks; JP '765's assumption is an admitted approximation, and US 2017/0322165 (same assignee) expressly recognizes the multi-line coupling ("the intensities of the measurement lines greatly depend on both the composition and the thickness"), which supplies the very recognition of the problem.
2. Owner's second rebuttal: the 103% acceptance band and the 1-second/0.5× multipliers are "specific" — but these are unclaimed-or-design-choice parameters; claim 1 recites only "predetermined" values.
3. My evidentiary exposures: (a) I have not quoted US 2017/0322165 A1's full specification; the "predetermined amount" perturbation and the FP recomputation are confirmed from its own published summary, but an F_ij ratio is my characterization of that procedure. (b) I have not verified the full content of the Matsushita, Sumitomo, JEOL '419, or Hitachi references; they are used only as background. (c) The Horiba reference's multi-condition setting is conditions (tube voltage), not measurement lines; a POSITA would still read it as teaching unequal time allocation to control quantitative standard deviation.

---

11. Conclusion

Both claims are, on the record cited by the patent itself, strong § 103 candidates for invalidation over:

• Claim 1: JP 2000-65765 + US 2017/0322165 A1 + JP 2003-270177 A (optionally + US 6,118,844 for the thin-film uncertainty-propagation point and JP 2009-097957 for the variable-time control point);
• Claim 2: the same set, minus the iterative loop, which claim 2 omits anyway.

The single most important reference is US 2017/0322165 A1 (Rigaku, publ. 2017-11-09) — it discloses, almost element-for-element, the claim-1 "first procedure" (perturb one measurement line's intensity by a predetermined amount, recompute via the FP method, iterate across lines) that the patent presents as its point of novelty. Combined with JP 2000-65765's admitted precision↔counting-time conversion and Horiba's unequal time allocation, the remaining steps are arithmetic and design choice, supported by express KSR rationales (same field, same problem, same-assignee lineage, finite known options, predictable results).

Recommended next steps: (1) obtain the full text of US 2017/0322165 A1 / EP 3239702 B1 and US 2003/0118148 A1 to lock down the perturbation and precision-calculation disclosures; (2) confirm the priority date of CN 114641687 A and WO 2021/059597 A1 — a pre-2019-03-29 date converts this from an obviousness case into a potential § 102 case; (3) commission an F_ij-perturbation § 112/§ 101 workup separately, since § 101 eligibility is unavailable at the PTAB (per the earlier PTAB section).

Confidence: High on the identification of US 2017/0322165 A1 as the closest art to the F_ij "first procedure" and on JP 2000-65765 as the admitted base reference; moderate on the weighting of Horiba and JEOL as supplying the iterative-allocation motivation, because my verification of those two is summary-level, not full-text; low/none on the date status of the "similar document" set, which must be independently checked. This is a technical analysis, not a legal opinion on validity.