US 5545531

I'll verify the patent details against live sources and check for any 2026 CAFC docket activity.

Let me check for litigation and any recent docket activity.

US Patent 5,545,531 — Verified Summary

I searched for the specific number 5545531 (and confirmed via Espacenet, Google Patents, FreePatentsOnline, and the USPTO Official Gazette record for the related continuation). Findings below. No CAFC 2026 docket activity for this patent was found — see the uncertainty note at the end.

Bibliographic data

Source: Espacenet biblio record for US5545531(A); Google Patents US5545531A.

Abstract (verbatim)

"Methods for concurrently processing multiple biological chip assays by providing a biological chip plate comprising a plurality of test wells, each test well having a biological chip having a molecular probe array; introducing samples into the test wells; subjecting the biological chip plate to manipulation by a fluid handling device that automatically performs steps to carry out reactions between target molecules in the samples and probes; and subjecting the biological chip plate to a biological chip plate reader that interrogates the probe arrays to detect any reactions between target molecules and probes."

Note a title/abstract asymmetry worth flagging: the title and all four granted claims are directed to making a plate/device, while the abstract and the bulk of the specification describe the concurrent assaying method. The assaying method was pursued in the continuation US 5,874,219 (app. 08/630,051, filed 1996-04-09), which is in the same family.

Independent claims in plain language

Claim 1 — Two-part assembled plate (body + wafer).

1. Provide a body with a plurality of wells that define open spaces.
2. Provide a wafer whose surface carries a plurality of probe arrays, where each array is a collection of probes, at least two of which differ, arranged in a spatially defined and physically addressable manner.
3. Attach the wafer to the body so that the probe arrays are exposed to (face into) the well spaces.
   Plain language: glue/join a multi-channel or multi-well plastic body onto a chip-bearing wafer so the wafer forms the floor of each well and each well's array sits inside its own compartment. (Fig. 4/5.)

Claim 3 — Barrier/well-forming material on the wafer.

1. Provide a wafer bearing a plurality of probe arrays (same array definition as above).
2. Apply a liquid-flow-resistant material surrounding each probe array, thereby creating test wells.
   Plain language: skip the molded body; instead put hydrophobic/wax/tape barriers directly on the wafer in the gaps between arrays, so the barriers themselves define the wells that hold liquid over each array.

Dependent claims: Claim 2 (dep. on 1) and claim 4 (dep. on 3) both merely limit the probes to DNA or RNA molecules.

Litigation / status notes

• The '531 patent was asserted in Affymetrix, Inc. v. Illumina, Inc., D. Del. Civ. A. No. 1:04-cv-00901-JJF (Unified Patents litigation record; court filings show a July 2006 Illumina summary-judgment motion on invalidity of the asserted '531 claims, and a 2006 Markman opinion, 446 F. Supp. 2d 277, construing '531 claim 1 language). That is the Delaware family-litigation hit referenced in the patent record, not a CAFC appeal of this number.
• Status is expired – lifetime with an anticipated expiration of 2015-06-07; no maintenance-fee lapse is indicated.

Uncertainty / caveats

• CAFC 2026 docket: I found no Federal Circuit docket entry or 2026 appellate activity tied to patent number 5545531. My searches returned zero results for "5545531 CAFC 2026 docket appeal," and the only litigation traces are the 2004–2007 Delaware district-court matter. I cannot confirm from the sources available whether any 2026 CAFC docket exists — treat "none found" as absence of evidence, not proof of absence. A direct PACER/CAFC docket query would be needed to be authoritative.
• The full claim text above is taken from the Google Patents claims listing ("Claims (4)"); I did not independently retrieve the printed patent's claim sheet or the USPTO PatentCenter PDF. The claim wording is consistent across the Google Patents, Espacenet, and FreePatentsOnline records I checked.
• Assignee history shows multiple transfers (Affymax Technologies N.V. → Affymetrix, Inc. → merger designations), so "assignee" depends on which date is referenced; current listed assignee is Affymetrix Inc.

I'll research litigation involving US 5,545,531 across patent-litigation sources.

Good initial results. Let me dig into the case outcome, appeals, and check for any other cases.

Let me pin down the specific outcome for the '531 patent and check for any other suits.

Litigation involving US 5,545,531 ("Methods for making a device for concurrently processing multiple biological chip assays")

Note on scope/method: I searched the Google Patents litigation data for this family, Unified Patents' Delaware case link, CourtListener, IPMall's Markman archive, patent-litigation press (GenomeWeb, PatentDocs, IBO, BioSpace, FierceBiotech) and SEC filings. I retrieved the full patent text and the D. Del. filings only in part before hitting my search-step limit, so I flag clearly below which items are confirmed and which are unverified. I did not retrieve any result tying 5,545,531 to a case number other than the one below, and I did not return results for similar-but-different numbers (e.g., 5,545,532 or 5,795,716 appear only as separate patents in the same suit).

The one known litigation asserting the '531 patent

Affymetrix, Inc. v. Illumina, Inc.

Sources confirming this case and the '531 patent's inclusion:

• USPTO/§290 Report to Commissioner in CA 04-0901 listing patent 5,545,531 as patent-in-suit: https://www.findforms.com/single_form.php/form/63424/Patent_Trademark_Report_to_Commissioner_District_Court_of_Delaware_District_Court_of_Delaware_Delaware
• BioSpace Markman announcement naming the five patents-in-suit (including 5,545,531): https://www.biospace.com/judge-issues-markman-order-in-affymetrix-santa-clara-california-litigation-with-illumina-inc
• Unified Patents litigation record for this docket (the link carried on the Google Patents page for US5545531): https://portal.unifiedpatents.com/litigation/Delaware%20District%20Court/case/1%3A04-cv-00901

Procedural history specific to the '531 patent:

1. Markman (claim construction): August 16, 2006 — Judge Farnan construed '531 terms including "probe array" ("a collection of probes, at least two of which are different, arranged in a spacially defined and physically addressable manner") and "arranged in a spacially defined and physically addressable manner" ("located in a particular location and capable of being addressed"). Text: https://ipmall.info/sites/default/files/hosted_resources/Markman/pdfFiles/2006.08.16_1_AFFYMETRIX_INC_v._ILLUMINA.pdf and https://www.courtlistener.com/opinion/[2315392](/patent/2315392)/affymetrix-inc-v-illumina-inc/

2. Illumina's summary-judgment motion of invalidity as to the '531 patent: filed July 14, 2006 (Doc. 275 in 1:04-cv-00901-JJF), directed at asserted claims 1–4; briefed with Affymetrix's opposition (Docs. 283, 317). Motion papers: https://www.findforms.com/single_form.php/form/63700/Motion_for_Summary_Judgment_District_Court_of_Delaware_District_Court_of_Delaware_Delaware

   • Caution: one retrieved copy of this document includes proposed order language adjudging claims 1–4 invalid with the date line left blank ("this ____ day of ____, 2006"). That is consistent with a proposed form of order attached to the motion rather than an entered order. I could not verify from the sources retrieved whether the motion was granted, denied, or never decided. Do not treat the '531 claims as adjudicated invalid on this record.

3. Jury verdict (phase 1): March 5–13, 2007 — the Delaware jury found Illumina infringed (and induced infringement of) all five asserted patents, including the '531 patent, set a 15% royalty, and awarded ~$16.7 million in damages for sales through 2005. Validity, enforceability, inequitable conduct and Illumina's antitrust/unfair-competition counterclaims were expressly deferred to later phases (which never occurred, due to the settlement). Illumina press statement: https://www.siliconinvestor.com/readreplies.aspx?msgid=23365124

   • Note: the infringement verdict did not include a validity finding on the '531 patent.

4. Settlement: January 10, 2008 — Illumina paid Affymetrix a one-time $90 million; Affymetrix dismissed with prejudice all lawsuits it had brought against Illumina; Illumina dismissed its counterclaims; mutual covenants not to sue (with a carve-out for photolithography). This resolved the Delaware case plus parallel actions in Düsseldorf, Germany and the High Court of Justice (Chancery Division, Patents Court), London: https://www.fiercebiotech.com/biotech/affymetrix-and-illumina-settle-their-patent-disputes

Related proceeding (not litigation, but directed at this patent)

• Ex parte reexamination of the '531 patent, USPTO serial no. 90/008,888, requested by Illumina on October 19, 2007; reexamination ordered/announced December 21, 2007 (along with 90/008,885 for 6,355,432; 90/008,886 for 5,795,716; 90/008,889 for 6,646,243). Source: https://patentdocs.org/2008/01/02/uspto-to-reexam-2/ and https://www.genomeweb.com/archive/uspto-re-examine-two-more-patents-affymetrix-illumina-litigation
  • I could not confirm the final reexamination outcome for 90/008,888 (e.g., whether a reexamination certificate issued confirming or cancelling claims 1–4). Treat as unknown in this report. The patent shows as "Expired – Lifetime" (anticipated expiration June 7, 2015) on Google Patents.

Matters I checked but could NOT tie to the '531 patent

• Affymetrix's second-wave suits (October 2007) — U.S. (Delaware), U.K. and Germany — asserted five additional U.S. patents not addressed in the Delaware case, so the '531 patent does not appear to have been asserted in those. Reported at: http://www.foleyhoag.com/~/media/Files/Publications/Generic/1107%20DeAnne%20Smith%20Affymetrix%20Illumina%20Reprint.ashx
• Illumina, Inc. v. Affymetrix, Inc., W.D. Wis. No. 09-cv-277-bbc (2009–2010). This is a real Affymetrix/Illumina dispute reflected in Illumina's filings, but I retrieved no source identifying 5,545,531 as a patent-in-suit there. I cannot confirm or deny '531's involvement — flagging as unverified rather than asserting.
• Google Patents' "Family has litigation" data for US5545531 (family ID 23893507) points to only one case record — the Delaware 1:04-cv-00901 action — consistent with the single-litigation picture above: https://patents.google.com/patent/US5545531/en
• No ITC (Section 337) investigation, no Federal Circuit appeal, and no other district-court case asserting 5,545,531 were found in the sources retrieved. Because the Delaware case settled at the phase-1 stage and was dismissed with prejudice, there is no appellate decision construing or invalidating the '531 claims.

Bottom line

One known litigation asserted US 5,545,531: Affymetrix, Inc. v. Illumina, Inc., D. Del. No. 1:04-cv-00901-JJF, filed July 26, 2004, in which the '531 patent was one of five (originally six) patents-in-suit. The jury found infringement of the '531 patent in March 2007; Illumina's invalidity challenge to '531 (summary-judgment motion filed July 14, 2006, and an ex parte reexamination 90/008,888 ordered December 2007) was never finally adjudicated because the parties settled on January 10, 2008 ($90 million payment; all suits dismissed with prejudice). If you need the docket-level confirmation of the disposition of Illumina's '531 summary-judgment motion and the reexamination certificate for 90/008,888, those would require pulling the PACER docket for 1:04-cv-00901-JJF and the USPTO reexamination file — I could not verify either from the sources retrieved.

I'll verify the PTAB picture with live searches before writing anything, since the structured block says "no AIA proceedings on file."

Reexaminations, not AIA trials, appear in the initial results. Let me dig specifically for IPR/PGR/CBM activity and the litigation/reexam outcomes.

Proceedings overview

Zero (0) AIA trial proceedings are on file against US 5,545,531 — the USPTO Open Data Portal returns no IPR, PGR, or CBM for this patent, and my independent web/PTAB searches surfaced none. Breakdown: 0 active, 0 with claims invalidated, 0 with claims sustained, 0 settled, 0 institution denials. There is therefore no IPR-sourced claim-level outcome to brief a defendant on, and no § 315(e)(2) estoppel in the chain. What exists instead is (i) a 2004–2007s-era district court assertion and (ii) ex parte reexaminations — neither of which is an AIA trial. One important trap: search engines return heavy PTAB traffic for US 9,399,531 (a sheet interleaver, Weber v. Provisur) — a different patent that merely shares the trailing digits 531. None of that activity is about US 5,545,531.

No AIA trial proceeding — this is the complete § 6 answer

• Type: N/A (no IPR / PGR / CBM petition was ever filed)
• Filed: N/A
• Status: No PTAB activity on file (per the structured USPTO ODP block supplied in the prompt; confirmed by web search)
• Judge panel: N/A
• Petition grounds: N/A — no petition exists, so there is no claims/art/§ 102/§ 103/§ 112 record to report
• Institution decision: N/A
• Final Written Decision: N/A — no claim of the '531 patent has ever been canceled, confirmed, or otherwise adjudicated by the PTAB
• Settlement / termination: N/A
• Appeal: N/A (no PTAB FWD to appeal)
• Defensive value: A defendant cannot rely on any PTAB outcome, but—equally—faces no adverse PTAB precedent. Any IPR-based defense must be built from scratch. Two statutory gates matter before you file: (1) § 315(b) — a petition is time-barred one year after service of a complaint alleging infringement of the patent; (2) PGR is unavailable (the '531 patent's effective filing date is 1995-06-07, long before AIA § 3's 2013-03-16 PGR cutoff), and CBM is unavailable in substance (this is a microarray/fluid-handling apparatus, not a "financial product or service"), independent of CBM's 2020-09-16 sunset for new petitions.

Adjacent matters that are not AIA trials (do not mislabel these as IPRs)

(A) Ex parte reexamination — requester: Illumina, Inc. (control no. reported as 90/008,888)

• Type: Ex parte reexamination (35 U.S.C. §§ 301–307); not an AIA trial, no PTAB involvement
• Filed: reexamination requests filed 2007-10-19; reexamination ordered/announced 2007-12-21
• Status: ordered by the USPTO on a substantial-new-question-of-patentability finding; final outcome not verified — I could not retrieve the reexamination certificate, so I will not state that any claim was canceled or confirmed
• Scope: Illumina requested reexamination of all claims of the '531 patent
• Defensive value: Useful prior-art-on-the-record ammunition only. Ex parte reexamination by a third-party requester creates no § 315(e)(2) estoppel (that estoppel is IPR/PGR-specific) and no PTAB claim-construction precedent for your case.
• Sources: https://patentdocs.org/2008/01/02/uspto-to-reexam-2/ ; https://www.genomeweb.com/archive/uspto-re-examine-two-more-patents-affymetrix-illumina-litigation

(B) Affymetrix, Inc. v. Illumina, Inc., D. Del. No. 1:04-cv-00901-JJF (the litigation flagged in the "Family has litigation" data)

• Filed: 2004-07-26 (complaint); the '531 patent was one of the asserted patents, alongside US 6,355,432; 6,646,243; 6,399,365; and 5,795,716
• Markman: order issued 2006-08-16 by Judge Joseph J. Farnan — Affymetrix, Inc. v. Illumina, Inc., 446 F. Supp. 2d 277 (D. Del. 2006), construing "probe array" and other terms, including the claim 1 language of the '531 patent ("A method for making a biological chip plate comprising the steps of: (a) providing a body comprising a plurality of wells defining spaces; (b) providing a wafer comprising on its surface a plurality of probe arrays, each probe array comprising a collection of probes, at least two of which are different, arranged in a spacially defined and physically addressable manner; (c) attaching the wafer to the body so that the probe arrays are exposed to the spaces of the wells.") — https://www.courtlistener.com/opinion/[2315392](/patent/2315392)/affymetrix-inc-v-illumina-inc/
• Invalidity challenge at the district court: Illumina filed a motion for summary judgment of invalidity of the asserted claims of the '531 patent on 2006-07-14 (D.I. 275). The motion papers contain a proposed order reading "the asserted claims 1-4 of U.S. Patent No. 5,545,531 are hereby adjudged to be invalid." ⚠️ This is a proposed order attached to a motion, not a court ruling. I could not verify from the sources retrieved whether the court granted that motion, and contemporaneous reporting of a March 2007 jury verdict finding Illumina liable on Affymetrix's asserted patents indicates the '531 claims may have reached trial. Treat claims 1–4 as TESTED but with an unverified district-court outcome — not as canceled. Source for D.I. 275: https://www.findforms.com/single_form.php/form/63700/Motion_for_Summary_Judgment_District_Court_of_Delaware_District_Court_of_Delaware_Delaware
• Defensive value: The 446 F. Supp. 2d 277 construction of "probe array" and the surrounding prosecution/claim-construction record remain the best free ammunition for a non-infringement or § 112 position today; the 2007-era reexamination art gives you a ready-made § 102/§ 103 starting set for a fresh IPR.

Strategic summary

Claim status. Because the PTAB never touched this patent, no claim of US 5,545,531 is CANCELED by the PTAB, and no claim is SUSTAINED by the PTAB. All four claims (1–4; claims 1 and 3 independent, claims 2 and 4 depending from them and limited to DNA/RNA probes) are UNTESTED at the PTAB. The only validity/adjudication events I can document are (a) Illumina's 2006-07-14 summary-judgment motion whose proposed order would have invalidated claims 1–4, and (b) an ordered ex parte reexamination in which every claim was placed at issue — with no certificate retrieved. Do not tell a court or client that claims 1–4 are canceled; that would be unsupported. Note that the '531 reexamination-era and litigation-era art is largely pre-AIA § 102/§ 103 art, so your IPR must plead pre-AIA §§ 102(a)/(b)/(e) and 103(a) as applicable to a 1995 filing.

Estoppel landscape. There is no § 315(e)(2) estoppel — the statute conditions it on a petitioner obtaining an IPR Final Written Decision, and there has never been an IPR. Likewise no § 325(e)(2) PGR estoppel. Illumina's ex parte reexamination request does not estop anyone (neither Illumina nor its privies) from raising the same art in your district court case or in a new IPR; the ex parte statute simply lacks an IPR-style estoppel provision. Practically, the entire universe of prior art — including all the art Illumina put into 90/008,888, and any art it did not — is available to you as of today.

Pattern signals. No petitioner has ever filed an IPR on this patent; there is no recurring PTAB petitioner, no defensive aggregator (Unified Patents or similar) in the chain, and consequently no PTAB appeals to the Federal Circuit. The patent's Google Patents legal-status line is "Expired – Lifetime" with an anticipated expiration of 2015-06-07, and the file lists no PTAB litigation-history events, only the Delaware case and Darts-ip family litigation. The 2004–2007 campaign against Illumina was patent-owner-side enforcement, with the target responding by reexamination rather than by AIA trial (which did not yet exist). The only later activity in the family is a chain of continuations/divisionals all claiming the same 1995-06-07 priority (US 5,874,219; US 2002/0018991; US 6,720,149; and the 2004–2005 publications) — which is a "same-family, same-priority" signal, not a fresh adversarial proceeding.

Recommended next steps

1. Start from "no PTAB activity on file," stated affirmatively. The absence is itself a signal, and here it is a strong one: this patent was aggressively asserted in 2004 and yet never drew an IPR. That reflects the pre-AIA timing of the fight and the 2015-06-07 expiration, not any weakness of the patent on the merits.
2. Check the expiration math before funding any defense. The '531 patent expired 2015-06-07 (20 years from the 1995-06-07 filing) per its own legal-status record, and its direct continuations (US 5,874,219; US 6,720,149) share that priority and 20-year clock. If a demand letter is asserting the '531 patent for ongoing post-2015 conduct, that theory is defective on its face; recovery would be limited to back damages within the § 286 six-year lookback. Verify with the USPTO's PatentCenter file wrapper before relying on this.
3. Retrieve the reexamination certificate for control no. 90/008,888 (also reported as a related Illumina request per the same 2007-10-19 filing date) via Patent Public Search / PatentCenter under the "Reexamination Certificate" tab. That single document will tell you whether any of claims 1–4 were canceled or confirmed and whether claim scope was amended — and it is the closest thing to a "claim-level outcome" this patent has. The USPTO's PTAB E2E (https://ptacts.uspto.gov/ptabweb/) will confirm the nil PTAB result.
4. Resolve the 2006-07-14 D.I. 275 summary-judgment motion. Pull the D. Del. docket for 1:04-cv-00901-JJF (https://portal.unifiedpatents.com/litigation/Delaware%20District%20Court/case/1%3A04-cv-00901) and the post-trial/Final Judgment documents to determine whether claims 1–4 were ever adjudged invalid and how the case terminated. If a judgment of invalidity of claims 1–4 was entered and affirmed/unappealed, that is a collateral-estoppel-grade weapon (not a PTAB one); if not, the 446 F. Supp. 2d 277 constructions are your non-infringement path.
5. If you are being asserted against and want an IPR, move within one year of service of the complaint (§ 315(b)) and build the petition on the pre-AIA art set Illumina surfaced in the reexamination plus any art not previously presented to the Office (mind § 325(d)). Because there is no prior IPR, you will face no § 315(e)(2) estoppel from anyone, and there is no FWD to quote — because none exists.
6. Do not cite the 9,399,531 PTAB papers (Weber, Inc. v. Provisur Technologies, petitions at https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1528326](/patent/1528326)/download-documents) in any filing about US 5,545,531; they concern a food-slicing sheet interleaver and would be a sanctionable mix-up.

Confidence statement. High confidence: zero AIA trial proceedings on US 5,545,531; the existence and dates of the Illumina-filed ex parte reexaminations (requests 2007-10-19; ordered 2007-12-21); the D. Del. case number and filing date; the 2006-08-16 Markman ruling at 446 F. Supp. 2d 277; the 2015-06-07 expiration line. Lower confidence / explicitly unverified: the disposition of Illumina's 2006-07-14 summary-judgment motion, whether claims 1–4 were ever held invalid in the district court, and the claim-level outcome of reexamination 90/008,888. I have flagged, rather than filled, those gaps.

I'll research the assignment record for US 5,545,531, prioritizing reel/frame data and correspondent-of-record fields.

Let me try to reach the assignment records directly through indexed sources.

Assignment & Ownership Reconstruction — US 5,545,531

Source-access note (read first). I could not open assignmentcenter.uspto.gov (or the assignment.uspto.gov mirror) with my available tooling — queries against them returned no indexed assignment records for this patent number. The assignment facts below are therefore reconstructed from the Google Patents legal-events table for US5545531A (fetched 2026-09-11, reproduced in the authoritative patent text supplied for this analysis), which carries assignor, assignee, conveyance type, and event date but strips the reel/frame and the correspondent-of-record fields. Every reel/frame number and every correspondent name below is marked "not retrieved" deliberately — I am not filling that gap with an invented number. One direct USPTO Assignment Center lookup by patent number 5,545,531 will populate both fields; see the verification link at the end.

I did retrieve confirmatory/contextual material for the corporate events (the 2012 GE Capital credit agreement, the Affymetrix–Illumina docket already briefed in the litigation section, and USPTO Official Gazette text for the '531 family continuations).

---

Inventors

Basis and caveats. The three names and the assignment instrument match the Google Patents legal-events entry for 1995-10-16 ("ASSIGNORS: FODOR, STEPHEN P.A., RAVA, RICHARD P., TRULSON, MARK" → AFFYMAX TECHNOLOGIES N.V.). The city/state residences are taken from the USPTO Official Gazette entry for continuation US 6,720,149 (issue 1281-2, 2004-04-13), which lists the same three inventors plus a fourth; those are not independently verified against the face of the '531 front page. Employer attribution is from the assignee chain itself (inventors assigned to Affymax Technologies N.V., whose chip business became Affymetrix) and is consistent, but I did not retrieve employment records.

Unusual patterns — mostly absent, with two things worth flagging:

• No post-filing inventor exodus. None of the three inventors departed the assignee within 12 months of the 1995-06-07 filing; all three appear on the 2004-issued continuation US 6,720,149, i.e. they stayed attached to the Affymetrix chip program for roughly a decade. The classic "inventors bail, portfolio gets flipped" precursor is not present.
• The original company was the thing that moved, not the inventors. The '531 application was filed 1995-06-07, and by 1995-10-16 (four months later) the inventors had assigned to Affymax Technologies N.V.; roughly eight months after that (1996-07-01) Affymax N.V. assigned to Affymetrix, Inc. That is the corporately documented split of the Glaxo-owned Affymax combinatorial-chemistry business from the spun-out Affymetrix chip business — a restructuring, not a fire-sale. Confidence: moderate-to-high on the direction of the chain (it is in the recorded events); the Glaxo/Affymax background is contextual, not a recorded USPTO fact.
• Inventorship discrepancy across the family. Vernon A. Norviel appears as a named inventor on the continuation US 6,720,149 but not on the '531 patent. It is common for a continuation to add an inventor, and Norviel's name is also associated with Affymetrix prosecution work in the era — but I flag it as an observed family discrepancy rather than drawing a conclusion. (Do not read this as a § 102(f)/(g) or inventorship-defect finding; nothing in the record supports that.)

---

Original assignee

• Entity named on the issued patent: Affymax Technologies N.V. (Netherlands Antilles), with the patent recorded as assigned to Affymetrix, Inc. the year of issue (see timeline).
• Primary line of business: Affymax Technologies N.V. was a combinatorial-chemistry / drug-discovery research company; its VLSIPS chip technology was commercialized through Affymetrix, Inc. (Santa Clara, CA), which marketed GeneChip® probe arrays and the instrumentation to read them.
• Did they ship a product embodying the claims? Yes — this is the rare case where the claims map onto a real commercial product line. Claim 1 / claim 3 cover making a biological chip plate (a wafer of probe arrays mated to a multi-well body, or arrays delineated by flow-barrier material). Affymetrix sold exactly this class of article: multi-array wafer/plate formats read on GeneChip scanners and processed on fluidics stations. Independent corroboration from the specification itself: it expressly contemplates a 96-well plate holding over 4,800,000 probes, and describes robotic fluid handling then available commercially "from Tecan AG."
• Current status of the original-named entity: Affymax Technologies N.V. does not survive as the patent owner. It transferred the portfolio to Affymetrix, Inc. on 1996-07-01. Affymetrix, Inc. itself reincorporated from California to Delaware by merger (1998-12-23), then was acquired by Thermo Fisher Scientific (deal announced January 2016; completed 2016). Current listed assignee of record on Google Patents: Affymetrix Inc.
  • Careful naming point: do not conflate the original Affymax Technologies N.V. with the later, unrelated Affymax, Inc. (the Palo Alto/Redwood City erythropoiesis drug company that partnered with Takeda and filed Chapter 11 in 2013). They share a name root, not this patent.

---

Assignment timeline

Five recorded events. Because the reel/frame and correspondent fields were not retrievable from the sources I could open, each entry carries the single event date as it appears on the Google Patents legal-events record; I have not guessed an execution-vs-recording split or a reel/frame.

• 1995-10-16 (date as recorded) / recorded 1995-10-16 — Reel not retrieved / Frame not retrieved

  • Conveyance: Assignment of assignors' interest (original assignment)
  • Assignor: Stephen P. A. Fodor; Richard P. Rava; Mark Trulson (individually)
  • Assignee: AFFYMAX TECHNOLOGIES N.V.
  • Correspondent: not retrieved — record field unavailable from the sources I could open; no recurring-correspondent call can be made.
  • Context: Original employment/obligation assignment to the operating research company that filed the application four months earlier — acquisition of title by the original assignee, not a transfer to a third party.

• 1996-07-01 (date as recorded) / recorded 1996-07-01 — Reel not retrieved / Frame not retrieved

  • Conveyance: Assignment of assignors' interest
  • Assignor: AFFYMAX TECHNOLOGIES N.V.
  • Assignee: AFFYMETRIX, INC. (California)
  • Correspondent: not retrieved.
  • Context: Corporate carve-out — the VLSIPS chip portfolio moves from Affymax N.V. to the Affymetrix chip subsidiary/affiliate the year Affymetrix went public; internal restructuring between related operating entities.

• 1998-12-23 (date as recorded) / recorded 1998-12-23 — Reel not retrieved / Frame not retrieved

  • Conveyance: Merger
  • Assignor: AFFYMETRIX, INC., A CORP. OF CA
  • Assignee: AFFYMETRIX INC., A CORP. OF DE
  • Correspondent: not retrieved.
  • Context: Internal reincorporation / change of domicile only — California Affymetrix merges into Delaware Affymetrix; beneficial ownership unchanged.

• 2012-06-27 (date as recorded) / recorded 2012-06-27 — Reel not retrieved / Frame not retrieved

  • Conveyance: Security Agreement
  • Assignor: AFFYMETRIX, INC.
  • Assignee: GENERAL ELECTRIC CAPITAL CORPORATION, AS AGENT
  • Correspondent: not retrieved.
  • Context: Securitization / collateral grant, not a transfer of title. Corroborated by the public credit agreement executed as of June 25, 2012 among Affymetrix, Inc. as borrower, General Electric Capital Corporation as Agent, and lenders, with Silicon Valley Bank as syndication agent ($100,000,000 facility). Under that agreement, patents were pledged as collateral while Affymetrix retained ownership; the "assignee" field in the assignment database here denotes the secured party's recorded interest, and this entry should not be counted as an ownership link in the chain of title.

• 2015-11-13 (date as recorded) / recorded 2015-11-13 — Reel not retrieved / Frame not retrieved

  • Conveyance: Release by Secured Party
  • Assignor: GENERAL ELECTRIC CAPITAL CORPORATION, AS AGENT
  • Assignee: AFFYMETRIX, INC.
  • Correspondent: not retrieved.
  • Context: Termination of the 2012 security interest — the lien recorded 2012-06-27 is extinguished and title is left clean in Affymetrix. Note the date: 2015-11-13 is five months after the patent's anticipated expiration of 2015-06-07, so the release confirmed a lien on an already-expired right.

• Post-2015 to 2026-09-11: no further recorded assignment retrieved. The Thermo Fisher Scientific acquisition of Affymetrix (2016) does not appear as a recorded assignment in the legal-events record I could retrieve for this patent. Two non-exclusive explanations, both plausible and neither verifiable from what I retrieved: (i) the deal was structured so that Affymetrix, Inc. survived as a Thermo Fisher subsidiary and remained the record owner, requiring no patent assignment recordation; or (ii) any such recordation is simply not surfaced in this patent's legal-events table. I am flagging this as an unverified absence, not asserting that no assignment exists. Google Patents' "Current Assignee: Affymetrix Inc" is consistent with either.

Cross-reference — SEC and asserter directories. The GE Capital entry is confirmed by the public credit agreement (SEC-filed); the Affymetrix–Illumina enforcement is confirmed by the D. Del. docket already briefed in the litigation section (1:04-cv-00901-JJF). I did not obtain an RPX Insurance or Unified Patents asserter-directory entry tying any assignee in this chain to a high-frequency-plaintiff list, and I make no such assertion.

---

Timeline diagram

timeline
    title Ownership of US 5545531
    1995 : Filed by Affymax Technologies NV
         : Inventors assign to Affymax NV
    1996 : Affymax NV assigns to Affymetrix Inc
    1998 : California Affymetrix merges into Delaware
    2004 : Affymetrix sues Illumina in Delaware
    2012 : GE Capital records security agreement
    2015 : GE Capital releases security interest
         : Patent reaches expiration
    2016 : Affymetrix acquired by Thermo Fisher

---

NPE / troll-pattern signals

1. Shell-entity transfer — not present. No link in the record moves this patent to an "IP / Patents / Licensing / Holdings / Ventures" entity. The only non-operating party ever recorded against the patent is General Electric Capital Corporation, as Agent, on the 2012-06-27 security agreement — and that is a lien, not a conveyance of title, expressly released by the 2015-11-13 release-by-secured-party entry. There is no single-purpose LLC, no registered-agent address, and no product-less assignee anywhere in the chain.

2. Known asserter in the chain — not present. Checking the assignees against the named lists (Acacia Research, Marathon Patent Group, Intellectual Ventures, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, Spangenberg entities): no match. Every recorded assignee is either an operating company (Affymax Technologies N.V., Affymetrix, Inc. in its CA and DE incarnations) or a secured lender (GE Capital). Affymetrix is the entity that sued, and it did so in its own name.

3. Repeat correspondent across the chain — unclear / cannot assess. The correspondent-of-record field was not retrievable for any of the five entries. I will not call this signal either way. Methodological note for the follow-up: this signal is only meaningful if the same attorney recurs across multiple links here, and here there are only two candidate "soft" links to test (1995-10-16 and 1996-07-01) plus one merger (1998-12-23) and one lender pair (2012/2015) — a single appearance by any firm would be noise, not a finding.

4. Cascading transfers — not present. Five events span twenty years (1995-10-16 → 2015-11-13). The only two events within 24 months of each other are 1995-10-16 and 1996-07-01 (~8.5 months apart) — and both are between operating entities in the same corporate family. There is no chain of consecutive LLC-to-LLC transfers, no shared registered-agent address pattern, and no common-principal cluster.

5. Pre-litigation transfer — not present. First infringement suit naming this patent was filed 2004-07-26 (D. Del. 1:04-cv-00901-JJF, Affymetrix, Inc. v. Illumina, Inc.). The immediately preceding assignment event is the 1998-12-23 merger — approximately 5 years 7 months before suit. There is no assignment within 6 months before the complaint; the plaintiff asserted a patent it had owned outright for years through a chain of internal restructurings. The 2012 and 2015 entries post-date the suit entirely.

6. Bankruptcy fire-sale — not present. No Chapter 7/11 proceeding or court-supervised patent sale involving this patent appears in the record. (Contrast the named examples — Kodak, Nortel, Polaroid.) The 2016 Thermo Fisher transaction was an acquisition of a going concern; the patent had already expired by then.

7. Privateering — not present. Affymetrix asserted the '531 patent itself, in its own name, against a competitor (Illumina) in its own district-court action, not through a proxy NPE. There is no transfer-to-asserter link anywhere in the record, and no SEC-disclosed "monetization" spin-out of this patent.

8. Defensive aggregator (anti-NPE) — not present (and the inverse signal does not apply either). The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or Open Invention Network. The patent's functional neutralization came from a different mechanism: expiration on 2015-06-07 (20 years from the 1995-06-07 filing). Any current assertion of this patent for post-expiration conduct is defective on its face; only pre-expiration, within-§-286-lookback back damages could ever have been in play.

---

Verdict

Operating-company assertion.

Justification. The chain is a clean history of an operating company owning and enforcing its own patent: inventors → Affymax Technologies N.V. (1995-10-16) → Affymetrix, Inc., CA (1996-07-01) → Affymetrix Inc., DE by merger (1998-12-23), with the sole non-operating record being GE Capital's 2012-06-27 security agreement, which the entries themselves resolve as a lien and its 2015-11-13 release. Affymetrix is a genuine product company (GeneChip arrays and readers embodying claims 1 and 3) and asserted the patent in its own name against a direct competitor in Affymetrix, Inc. v. Illumina, Inc., D. Del. No. 1:04-cv-00901-JJF — the classic operating-company-vs-competitor posture, not a troll posture. Zero of the eight NPE signals are present on the evidence retrieved; one (repeat correspondent) is honestly unassessable because the correspondent field could not be read.

Verification link (populates the reel/frame and correspondent fields I could not read): https://assignmentcenter.uspto.gov/ — search by patent number 5545531 (mirror: https://patents.google.com/patent/US5545531/en, whose legal-events table was my fallback source).

Confidence statement. High confidence: the five recorded events and their order/dates; the assignor/assignee identities and conveyance types; that the GE Capital entries are a security interest and its release rather than ownership transfers; the absence of any NPE-named assignee. Explicitly unverified: all reel/frame numbers, all execution dates (vs. recording dates), and all correspondents of record — not retrieved from any source I could open, so left blank rather than fabricated; the precise claim-level terms of the 2012 security agreement as they touch this patent (I confirmed the credit facility, not the patent-security schedule); and whether a Thermo Fisher assignment was ever recorded for the '531 patent after the 2016 acquisition (flagged as an unverified absence). No contradiction with the previously generated summary/litigation sections was found — the earlier sections' statement that the current listed assignee is Affymetrix Inc. is consistent with this record, which shows no post-2015 transfer.

I'll verify the patent's cited prior art against live sources. Let me run several searches in parallel.

Now let me retrieve the front-page "References Cited" of US 5,545,531 itself, plus the remaining specification-cited references.

Prior-Art Analysis — US 5,545,531

0. Scope, method, and one retrieval gap I must flag up front

Identifier confirmed literally: US 5,545,531 A, "Methods for making a device for concurrently processing multiple biological chip assays," App. No. 08/476,850, filed/priority 1995-06-07, issued 1996-08-13, inventors Rava / Fodor / Trulson, original assignee Affymax Technologies N.V. All analysis below is confined to that number. I did not pull results for look-alikes such as US 5,545,532, US 5,795,716, or US 9,399,531.

The gap you need to know about: The authoritative record supplied in my prompt (Google Patents US5545531A) contains a "Cited By (460)" table and the specification's in-text references — but it does not reproduce the front-page "References Cited / U.S. Patent Documents" list of the '531 patent itself. I attempted to retrieve that front-page list (patentimages PDF, FPO, uspto.report) and hit my search-step ceiling before confirming it. So I cannot give you a verified, complete enumeration of the examiner-cited references on the '531 face. What follows is built from (a) the references the '531 specification expressly cites and incorporates, which I could verify against live records, and (b) a clear explanation of why the "Cited By" list is not prior art. Treat the front-page list as still open.

Legal framework — non-negotiable here: a 1995-06-07 filing date means the pre-AIA statute governs. Anticipation is pre-AIA § 102(a)/(b)/(e), obviousness is § 103(a). AIA §§ 102/103 (first-inventor-to-file, 2013-03-16) do not apply retroactively.

---

1. First, a trap to dispose of: "Cited By (460)" is NOT prior art against '531

The prompt's Google Patents text lists ~25 documents under "Cited By," many flagged "* Cited by examiner" — e.g., WO 1997039151A1 (Affymetrix, priority 1996-04-17), US 5,731,152A (Motorola, 1996-05-13), US 5,807,522A (Stanford, 1994-06-17), DE 19736641A1 (1997), WO 1998049543A1 (Nanogen, 1997-05-01), US 5,935,785A (Motorola, 1997-04-30), US 5,885,837A.

None of these can anticipate US 5,545,531. The "* Cited by examiner" star means the later examiner cited US 5,545,531 in that later application — i.e., these are forward citations, documents that came after the '531. Every one of them post-dates 1996-08-13 (or at minimum, post-dates the 1995-06-07 filing). Using them as § 102 art against '531 would be a one-way-street error. The only exception worth keeping in a file is US 5,807,522 (Stanford, "Methods for fabricating microarrays of biological samples," filed 1994-06-17) — that one does pre-date the '531 filing and is a genuine § 102(b) candidate, but it appears here only as a forward citation, not as a citation against '531.

Also not prior art: the "Prior art keywords" field on the Google Patents record ("probes; biological chip; probe; biological; array") is an indexing artifact, not a reference.

---

2. Table A — References expressly cited and incorporated by the '531 specification

These are the true "citations" readable in the '531 record. Dates bolded where I verified them against live sources.

---

3. Claim-by-claim § 102 verdict

Recall the four granted claims (from the previously-generated claim summary, which I adopt):

• Claim 1 — making a plate: (a) body with a plurality of wells defining spaces; (b) wafer bearing a plurality of addressable probe arrays (≥2 probes different); (c) attaching wafer to body so arrays are exposed to the well spaces.
• Claim 2 — claim 1, probes are DNA/RNA.
• Claim 3 — making a plate: wafer bearing a plurality of probe arrays + applying a liquid-flow-resistant material surrounding the arrays to create test wells.
• Claim 4 — claim 3, probes are DNA/RNA.

Calibrated conclusion: no reference in the record I could verify is a clean § 102 anticipation of any of claims 1–4. The only genuine § 102(e) candidate is US 5,384,261, and it is a candidate, not a certainty — its channels are flow paths, not sample-holding wells, and the '531's claim 1 requires the wells to define sample-receiving spaces. That distinction is exactly the kind of thing the Delaware court tested in the 2006-08-16 Markman ruling (446 F. Supp. 2d 277), where "probe array" was construed as "a collection of probes, at least two of which are different, arranged in a spacially defined and physically addressable manner." A challenger should expect to argue the well limitation, not the array limitation.

---

4. What I could not verify (do not treat as established)

1. The '531 front-page "References Cited" list — not retrieved. The examiner may have cited US patents and non-patent literature absent from both the specification and the Google Patents "Cited By" table. This is the single biggest hole in this memo, and it is exactly the list you asked for. Pull it from the USPTO PatentCenter file wrapper or the printed patent sheet.
2. US 5,384,261's § 102(e) status against '531 depends on the '531's actual invention date (reduction to practice / conception). A filing-date analysis alone gives § 102(e); a § 102(a) analysis requires the invention date, which I have not established.
3. Common-ownership effect. A1, A4, A5, A7–A13 are all Affymax/Affymetrix family references. Under pre-AIA § 103(c), commonly owned § 102(e)/(f)/(g) art is disqualified for obviousness only — it remains available for anticipation. Anyone building a § 103 challenge out of A4/A5/A7–A13 needs to run this analysis before relying on them.
4. Resulting patent numbers for applications 08/082,937, 08/249,188, 08/284,064 (PCT/US94/12305), and 08/143,312 — I have plausible mappings but did not verify them. Verified links only: 08/195,889 → US 5,631,734; 08/301,051 → US 5,578,832; 07/624,120 continuation → 08/390,027 → US 5,489,678.
5. A minor date conflict in the task prompt: you state "Current Date: April 26, 2026," while my operating header reads 2026-09-11. It does not change the analysis for a patent that expired 2015-06-07, but flagging it per the cross-reference rule.

---

5. Practical bottom line

• For a § 102 anticipation theory, the record I can verify gives you essentially one candidate: US 5,384,261 (channel block + substrate, filed 1991-11-22), and it is a contested fit because its channels are flow paths rather than sample-holding wells. Nothing else in the cited set is close.
• The strongest overall invalidity posture is § 103, combining the VLSIPS array references (US 5,143,854; WO 90/15070; WO 92/10092; US 5,384,261) with the admitted prior art in the '531's own background — notably the specification's own concession that "the synthesis of peptides on 96 plastic pins which fit the format of standard microtiter plates" was known. That concession is drawn from the Geysen pin work (Geysen et al., PNAS (1984) 81:3998–4002; J. Immunol. Meth. (1987) 102:259–274) and is quoted in the '854 background. Geysen is not cited on the '531 face, so treat it as landscape art, but it directly supplies the "body comprising a plurality of wells" element that the Affymax references lack — which is precisely why the combination is the natural § 103 theory.
• If you want the actual examiner-cited art, the front-page list is a 10-minute PatentCenter pull and it is the one document that would change the picture materially. I could not reach it within my search budget and will not guess at its contents.

I'll verify the key prior-art references' dates and disclosures before building the § 103 combinations.

Now let me try to close the front-page "References Cited" gap flagged in the prior-art section, and ground the microtiter-plate-art motivation prong.

§ 103 Obviousness Analysis — US 5,545,531

Builds on (and does not repeat) the Prior-Art section above; reference labels A1–A13 are carried forward. Newly verified references are labeled G1–G5.

---

1. Scope, governing law, and one advance on the open gap

Patent: US 5,545,531 A, "Methods for making a device for concurrently processing multiple biological chip assays," App. 08/476,850, filed/priority 1995-06-07, issued 1996-08-13, Rava/Fodor/Trulson, original assignee Affymax Technologies N.V. Claims 1–4 only (2 independent, 2 dependent). Expired 2015-06-07.

Legal framework is non-negotiable: a 1995-06-07 filing date means pre-AIA 35 U.S.C. § 103(a) governs, with pre-AIA §§ 102(a)/(b)/(e) as the art-qualifying provisions. AIA §§ 102/103 do not apply retroactively. Because the reexamination (90/008,888) and the 2004–2007 Delaware case were both pre-KSR, note that for any proceeding filed today KSR Int'l v. Teleflex (2007) supplies the operative obviousness standard and its enumerated rationales even though the patent is pre-AIA.

One gap I partially closed, and one I did not. The Prior-Art section flagged that the '531's own front-page "References Cited" list could not be retrieved. I still could not reach the '531 face. I did retrieve the front-page reference list of its continuation, US 5,874,219 (FreePatentsOnline, https://FreePatentsOnline.com/[5874219](/patent/5874219).html): U.S. Patents 5,578,832; 5,571,639; 5,556,752; 5,545,531; 5,538,857; 5,424,186; 5,384,261; 5,382,512; 5,348,855; 5,288,514; 5,196,305; 5,143,854; foreign WO 90/15070, WO 92/10092, PCT/US94/012305, WO 94/12305; other: Drmanac, "Biochip Array Technologies," meeting, May 10, 1995, Wash. D.C.; Hyseq Inc. "Genome Sequencing Machine" materials. This is a proxy, not the '531 face — the examiner may have cited different art against the '531. Treat the enumeration below as the verifiable universe plus new landscape art.

---

2. The claims, decomposed into elements

Recall the claim text is not about assaying; it is about making a plate.

Claims 2 and 4 add only: probes are DNA or RNA molecules.

Four observations that drive everything below:

1. Claim 1's "attaching" is unqualified. No seal requirement, no fluid-tightness, no bond strength, no specific well count, no transparency, no bonding method. The '531's own specification concedes the attachment options are routine: "gluing (e.g., by ultraviolet-curing epoxy or various sticking tapes), acoustic welding, sealing such as vacuum or suction sealing, or even by relying on the weight of the body on the wafer." That is an admission that the only assembly step in claim 1 was conventional — precisely the posture that makes a claim vulnerable under KSR rationale (A) (combining known elements by known methods) and In re Sovish-type reasoning about bare conventional steps.
2. Claim 3's barrier is unqualified beyond "resistant to the flow of a liquid sample." It does not require hydrophobicity, a particular material, a particular height, or a gap between arrays.
3. Neither claim requires the arrays to differ from one another — only that within each array ≥2 probes differ. The "each row same, each column different" concept is disclosed but unclaimed.
4. Claim 1 requires a "wafer" (plurality of arrays on one substrate), not individual chips. The FIG. 6 preformed-well embodiment (gluing separate chips into wells) is described but falls outside claim 1 as issued.

---

3. The reference set, organized by claim element

(a) Array-on-substrate ("wafer") art — supplies 1b/1c and 3a

(b) Multi-well / plate-format art — supplies 1a (and the format's instrument compatibility)

(c) Fluid-confinement / barrier art — supplies 3b/3c (and, for some, 1d)

(d) Microtiter-plate-mated-to-a-sheet art — supplies 1d

---

4. The § 103(c) gate — read this before choosing a combination

Pre-AIA § 103(c) (the version in force for this 1996 grant; the 2004 CREATE-Act amendment postdates it and adds nothing here) provides that subject matter which "qualifies as prior art only under one or more of subsections (e), (f), and (g) of section 102" shall not preclude patentability where, at the time the invention was made, that subject matter and the claimed invention were owned by the same person or subject to an obligation of assignment to the same person.

Applying it:

This is the single most important practical constraint on the whole analysis. The '531's closest structural reference — A4/'261, the channel block sealed against a wafer — is the applicant's own commonly owned § 102(e) art, so § 103(c) prima facie knocks it out of any § 103 combination. The '531 specification cites and incorporates A4 by name, which is exactly the fact pattern § 103(c) was written to cover. A challenger either (i) builds § 103 out of the non-commonly-owned references (G1–G5) plus the § 102(b) Affymax references (A1, A2, A3, A5), or (ii) litigates the § 103(c) predicate (e.g., contending A4 independently qualifies under § 102(a)/(b), which requires establishing the '531's invention date).

---

5. Ranked § 103 combinations

★ Combination I (strongest, and § 103(c)-clean) — Claim 1: A1 + A3 in view of G1 (Geysen) and the standard microtiter plate

Mapping

Motivation — and it is unusually well documented here. The motivation does not have to be reconstructed, because the applicant wrote it down:

1. The '531's own background states the problem: "Currently, chips are treated individually, from the step of exposure to the target molecules to scanning. These methods yield exquisitely detailed information. However, they are not adapted for handling multiple samples simultaneously. The ability to do so would be advantageous in settings in which large amounts of information are required quickly, such as in clinical diagnostic laboratories or in large-scale undertakings such as the Human Genome Project." Under In re Fritch / KSR, a problem the inventor himself frames as the objective of the work is powerful evidence that the combination was not beyond the ordinary artisan.
2. The '531 supplies the express design rationale for the microtiter format: "One advantage of this format is that instrumentation already exists for handling and reading assays on microtiter plates. Therefore, using such plates in biological chip assays does not involve extensive re-engineering of commercially available fluid handling devices." That is a textbook KSR rationale (C)/(D) — using a known technique (microtiter-format fluid handling) to improve a similar device (a biochip assay) in the same way, and applying a known technique to a known device ready for improvement.
3. G1 itself establishes that the 96-well microtiter plate was already the accepted vessel for parallel, compartmentalized solid-phase synthesis and binding readout. Moving the solid phase from a pin suspended in a well to an array on the well floor is a simple substitution of one known element for another (KSR rationale (B)) with a predictable result — the same chemistry, the same plate, the same reader format, higher information density.
4. A1/A3 supply the density motive: the '854 background criticizes the pin method as "painfully slow" and "limited in the diversity of sequences which can be economically synthesized," i.e., the art expressly wanted the pin method's parallelism without its density limitation. Combining A1/A3's array substrate with G1's plate format is the direct answer.
5. G2 (Frank 1992) independently shows the combination already being contemplated as a matter of routine design: "spots are made in an 8×12 microtiter plate format" on a contiguous sheet, then assayed by ELISA — i.e., a high-density repertoire of probe molecules laid out to the microtiter footprint.

Predicted strength: Moderate-to-good on claim 1. The only genuinely contestable element is 1d ("attaching"), and it is contestable only if one insists that the specification's list of "epoxy, tape, acoustic welding, suction sealing, or the body's own weight" is not legally an admission of convention. The claim recites no functional consequence of the attachment, so the "new property" doctrine of In re Papesch-style argument is unavailable.

---

★ Combination II — Claim 1, alternative: A1/A3 + G5 (plate-with-sheet art) + standard microtiter plate body

Identical mapping, but the "attaching" element is supplied by the microtiter-plate-with-membrane/insert art rather than by ordinary skill alone. Motive: multi-well plates whose wells are closed by an applied sheet were an established article of commerce; bonding a probe-bearing wafer across the open bottoms of a well body is the same structure for a new analytical purpose. KSR rationale (A). Strength: dependent on how well the underlying plate hardware is documented — I verified only that such inserts and plate formats existed and were cited as prior art in Affymetrix's own later patent (US 7,364,895), not the specifics of any one product.

---

★ Combination III (strongest, § 103(c)-clean) — Claims 3 and 4: A1 (or A3) + G3/G4 + slide-barrier practice

Mapping: 3a ← A1/A3 (arrays on a substrate). 3b ← G3/G4 (region-selective wettability and patterned, delimited attachment areas — a "material resistant to the flow of a liquid sample" surrounding defined regions). 3c ← the barriers necessarily create isolated liquid-holding areas ("test wells"). A4/'261 independently supplies the same concept with raised walls creating fluid-tight seals against a substrate, subject to the § 103(c) caveat.

Motivation: (i) Aqueous assay reagents must be confined over the correct address of a high-density array; otherwise cross-talk destroys the addressability that the whole VLSIPS enterprise depends on. (ii) A4's stated objective — confining reagents with fluid-tight seals between a channel block and a substrate to avoid reagent cross-contamination and economize reagent — is the same problem the '531's claim 3 solves, solved the same way, which is exactly KSR rationale (C). (iii) Surface-wettability patterning (G3/G4) was the known technique for confining small liquid volumes to defined regions on a planar substrate. (iv) Slide-barrier practice shows that applying a flow-resistant barrier around a defined area on a flat substrate to hold a reagent was routine before 1995 (I flag this as common knowledge I have not tied to a specific reference).

Predicted strength: Better than claim 1. Claim 3's "material resistant to the flow of a liquid sample" is broad enough to read on the '261's raised walls, on a hydrophobic/wettability-patterned strip, and on a wax or tape barrier applied to the wafer — the last of which the '531 specification itself describes ("the wafer can be scored with waxes, tapes or other hydrophobic materials in the spaces between the arrays").

---

Combination IV — Claims 1 and 3: A4 ('261) alone or in view of A1/A3 — subject to the § 103(c) gate

If § 103(c) does not apply (because A4 independently qualifies under § 102(a)/(b), or because the § 103(c) predicate fails), A4 is the closest single reference to claim 1 that exists: a planar member with recessed regions forming channels and raised walls, contacted against a substrate, forming fluid-tight seals defining compartments between the block and the substrate, with pipets placed in fluid communication with the channels. Structurally that is the '531's FIG. 4/5 "body + wafer" in all but nomenclature — which is why the Prior-Art section rated A4 the only real § 102(e) candidate.

The counter is the Delaware Markman construction: "probe array" = "a collection of probes, at least two of which are different, arranged in a spacially defined and physically addressable manner," and "arranged in a spacially defined and physically addressable manner" = "located in a particular location and capable of being addressed" (Affymetrix, Inc. v. Illumina, Inc., 446 F. Supp. 2d 277 (D. Del. 2006), https://www.courtlistener.com/opinion/[2315392](/patent/2315392)/affymetrix-inc-v-illumina-inc/). A challenger should expect the patentee to argue that A4's channels are flow paths for synthesis reagents, not sample-holding wells defining spaces into which a sample is introduced — a distinction the '531 specification itself draws ("the channels … form the walls of the well, but that are open at the bottom … thereby creating wells," versus flow channels for reagent delivery). That distinction is a factual question about the reference's disclosure, not a claim-construction question, and it is winnable either way. Claim 3 is more exposed than claim 1 here, because claim 3 does not require wells to receive a sample — only "materials resistant to flow … surrounding the probe arrays."

---

Combination V — Claims 1–2, "system/process" motivation: A1/A3 + A5 ('633) + commercial microplate reader/robotics

Claim 1 requires nothing about reading, so A5 is not needed to meet the claim. It is, however, the motivation reference: A5 establishes that the art already knew how to interrogate an array-bearing substrate optically in contact with labelled receptor, and it cross-references A4's channel-block synthesis. Combined with G1 (microtiter format), this shows the artisan had both halves — array readout and plate-format automation — and every reason (throughput, cost, reagent economy) to mate them. Relevant chiefly to the broader claims of the continuation US 5,874,219 (e.g., its claims 1 and 30), whose "fluid handling device … automatically performs steps" and "biological chip plate reader" limitations are directly met by the '531 specification's own admissions that "Microtiter robotic and fluid-handling devices are available commercially, for example, from Tecan AG."

---

Claims 2 and 4 — DNA/RNA probes

Trivially obvious. A1, A2, A3, A5 all expressly contemplate polynucleotide/polypeptide arrays; A3/WO 92/10092 and the '854 family claim oligonucleotide embodiments; the '531 specification itself lists DNA/RNA as the preferred probes and describes phosphoramidite chemistry throughout (FIG. 8). No separate motivation analysis is needed — the dependent claims add only a recitation of the species already disclosed by the primary references. If independent claims 1 or 3 fall, claims 2 and 4 fall with them.

---

6. Why a PHOSITA would have combined these — assembled

The level of ordinary skill at 1995-06-07: a Ph.D.-level or M.S.-level scientist or engineer with several years in solid-phase synthesis, surface chemistry, and/or laboratory automation, familiar with (i) VLSIPS photolithographic array synthesis (A1–A3), (ii) 96-well microtiter plate formats and their robotic handling (G1, G2, admitted art), and (iii) standard optical readout of fluorescent labels from planar surfaces (A5).

The KSR rationales, applied:

• (A) Known elements combined by known methods, predictable result. Known array-on-substrate (A1/A3) + known multi-well body and its automation (G1, standard plates) + known sheet-to-body assembly (G5 / ordinary mechanical skill) = a plate of arrays. Each element performs its own known function.
• (B) Simple substitution of one known element for another. Substituting an array-bearing substrate floor for Geysen's pin-supported solid phase inside the same 96 wells.
• (C) Known technique to improve a similar device in the same way. A4's flow-confining sealing of a block to a substrate (to stop reagent cross-talk and save reagent) applied to an array-bearing wafer.
• (D) Known device ready for improvement. The art — by the applicant's own admission — had arrays that worked exquisitely but only one sample at a time, and microtiter instrumentation that already existed. That is the classic "ready for improvement" posture.
• Design incentives / market pressure. Clinical diagnostics and the Human Genome Project are named in the specification as the demand drivers. KSR holds that market pressure and design needs are legitimate motivation evidence.

And the most damaging single fact for the patentee under § 103: the '531's own specification concedes both (i) the problem, (ii) the solution's rationale ("instrumentation already exists … does not involve extensive re-engineering"), and (iii) that the sole assembly step is routine. Those are admissions of the inventor, usable against the claims as evidence of the level of skill and the motivation to combine.

---

7. Where a § 103 challenge is weak — the honest counter-case

1. § 103(c) removes the best art. As shown in § 4, the strongest structural reference (A4/'261) and the whole family of related applications (A7–A13) are commonly owned § 102(e) art. Expect the patentee to move to exclude them from the § 103 case. Plan the combinations around G1/G2/G3/G4/G5 + A1/A2/A3/A5 from the outset.
2. The examiner had most of this art. On the related continuation's face, the Office considered A1, A2, A3, A4, A5, A6 and still allowed claims. That is not dispositive, but it shifts the burden narrative and triggers the § 325(d) discretionary-denial risk in any IPR: a petition must lead with art that was not before the examiner — which, on the record I can verify, means Geysen (G1), Frank (G2), Abbott/Bhatia (G3/G4) and the microtiter hardware art (G5). Those are the references with the best chance of being treated as new.
3. Teaching away. The A1 ('854) background criticizes the Geysen pin method ("painfully slow," "not economically practical," "limited in the diversity"). A patentee will characterize that as teaching away from the 96-pin/microtiter approach. The rebuttal: the criticism was directed at pin-based synthesis throughput, not at the microtiter vessel or its instrumentation; and A1 did not disparage the plate format itself — it disparaged a density limitation that A1's own arrays solve.
4. Secondary considerations. Expect evidence of long-felt need, initial skepticism, and the commercial success of Affymetrix's GeneChip® multi-array/plate formats and scanners. Because A1–A5 are the patentee's own patents, a nexus argument is available for the inventor's own product line. Whether the asserted success is attributable to the claimed "making" steps (as opposed to the array chemistry of A1–A3) is the weak point in that nexus — and the point to attack.
5. "Wells" ≠ "channels." The most likely defensive construction argument, drawn from the 2006 Delaware Markman record and the specification's own contrast between reagent-delivery channels and sample-holding wells. Preserve the factual record that A4's channels receive reagents from pipets and hold volumes — i.e., that they too define sample-receiving spaces.
6. The reexamination (90/008,888) is a wildcard. It was ordered on all claims of the '531 in December 2007 and, per my earlier section, I could not retrieve a reexamination certificate. If a certificate issued confirming claims 1–4 over the art Illumina submitted, that is a meaningful (though not binding on a court) administrative validation. If it narrowed them, the narrowing itself becomes the roadmap for a fresh § 103 attack. Until the certificate is pulled from PatentCenter, this analysis cannot be treated as complete.
7. The patent is expired (2015-06-07). A § 103 holding today has value only for pre-expiration conduct within the § 286 lookback, or as a counterclaim in a case that never finally adjudicated the '531's validity (Illumina's 2006-07-14 summary-judgment motion, D.I. 275, was resolved by the January 2008 settlement without a validity judgment).

---

8. Bottom line

All four claims are, on this record, vulnerable to a well-constructed pre-AIA § 103(a) attack, provided the petition/argument is built on the non-commonly-owned art (Geysen, Frank, Abbott, Bhatia, microtiter hardware) plus the § 102(b) Affymax publications (US 5,143,854; WO 90/15070; WO 92/10092; US 5,324,633), and provided the § 103(c) common-ownership problem with US 5,384,261 and the pending application references is addressed head-on rather than ignored. The single greatest risk to such a challenge is not the merits — it is (a) § 103(c) knocking out the closest reference, and (b) § 325(d) discretion if the petition re-runs art the examiner already considered.

---

9. Confidence statement and open items

High confidence: the claim text and element decomposition; the legal framework (pre-AIA § 103(a), § 103(c) gate); the dates, numbers, and content of US 5,143,854 (app. 07/492,462, filed 1990-03-07, issued 1992-09-01), US 5,384,261 (app. 07/796,243, filed 1991-11-22, issued 1995-01-24, channel block with raised walls and fluid-tight seals, pipet-to-channel claim), US 5,324,633, WO 90/15070, WO 92/10092; the Geysen 96-well/microtiter-plate teaching and its appearance in the '854 background; the Frank 8×12 microtiter-format sheet-spotting disclosure; the Abbott and Bhatia references; and the '531 specification's own admissions (problem statement, "instrumentation already exists," routine attachment methods).

Lower confidence / explicitly unverified:

• The '531's own front-page "References Cited" list — still not retrieved. I substituted the continuation US 5,874,219's front-page list and flagged it as a proxy. This remains the largest gap and is a 10-minute PatentCenter pull.
• Whether claims 1–4 are § 103(c)-disqualified art is a legal predicate I have reasoned to, not researched in the file. Common ownership by Affymax at the time of invention is inferred from the recorded assignment chain (inventors → Affymax Technologies N.V., 1995-10-16) and from the issuance data of US 5,384,261 — I did not retrieve an assignment record proving Affymax owned the 07/796,243 application and the '531 application on the same date.
• US 5,324,483 (Cody), US 5,252,743 (Barrett), US 5,412,087 (McGall), US 4,704,256 (Hood), US 4,598,049 (Zelinka) — real multi-vessel/spatially-addressable references surfaced in the cited-reference lists of Affymetrix's own continuations, but their specific disclosures and assignees are not verified by me; do not brief them without pulling the documents.
• G5 (Falcon inserts news release, Apr. 2, 1991) — verified as a citation in US 7,364,895's prior-art list; the product's specific structure is not verified.
• The slide-barrier ("PAP pen"/wax pencil) practice — offered as common knowledge, not tied to a specific pre-1995 reference.
• The disposition of the ex parte reexamination 90/008,888 — still unknown (carried forward from the PTAB/reexam section).
• Date discrepancy (flagged per the cross-reference rule): this task's prompt states "Current Date: April 26, 2026," while my operating header reads 2026-09-12 and an earlier section's header read 2026-09-11. Immaterial to a § 103 analysis of a patent that expired 2015-06-07, but noting it rather than silently reconciling.

No contradiction with the earlier generated sections was found; this analysis adopts their reference labels, their claim decomposition, the § 103(c) caveat they did not develop, and their stated uncertainty regarding the reexamination certificate.