US 6420169

I'll verify the patent details and check for litigation/docket information.

I have the bibliographic data and litigation info. Now let me try to retrieve the actual claim text, since the provided full-text excerpt did not include the claims.

Let me try to retrieve the full claim set (especially independent claims 1, 7, 8, 10, 11, 15).

US Patent 6,420,169 — Analyst Summary

Retrieval limitations up front: I searched the open web for USPTO/Google Patents/FPO records and for litigation dockets. I did not query USPTO PAIR/PatentCenter or the CAFC docket system directly, and the authoritative full text you supplied does not include the claim body (it stops at the classification section). Claim information below is therefore assembled from third-party indexes (FreePatentsOnline, uspto.report) and is flagged where incomplete. Where I could not verify something, I say so rather than reconstruct it.

1. Bibliographic data (verified, multiple sources)

Minor indexing discrepancy worth noting: one FreePatentsOnline search-result title rendered the patent as "Affymetrix, Inc. – Olson et al," which conflicts with the inventor list on Google Patents and in your authoritative text (Read/Fodor/Stryer/Pirrung/Hoeprich). I did not auto-correct it; I'm flagging it as an index artifact rather than accepting it, since two independent sources plus the patent's own front page give the five-inventor list.

2. Abstract

The indexed abstract (FreePatentsOnline) reads:

"A method for synthesizing oligonucleotides on a solid substrate. The method provides for the irradiation of a first predefined region of the substrate without irradiation of a second predefined region of the substrate. The irradiation step removes a protecting group therefrom. The substrate is contacted with a first nucleotide to couple the nucleotide to the substrate in the first predefined region. By repeating these steps, an array of diverse oligonucleotides is formed on the substrate."

Caveat: the abstract describes a method, while the title and the claim fragments I could retrieve are directed to an apparatus (a flow cell / reactor). This mismatch is common in this Affymetrix family (the specification carries both), but I am not asserting the abstract text is complete — Google Patents' abstract body was not captured in the retrieved text.

3. Independent claims — plain-language overview

The retrieved claim fragments show apparatus claims built around a shallow, sealed reaction cavity mated to a derivatized substrate surface — the "flow cell" that carries activated monomers across a light-directed synthesis surface. Based on repeated "An apparatus … said apparatus comprising:" preambles, claims 7, 8, 10, 11 and 15 appear to be independent apparatus claims; claims 9, 12, 13, 14 appear dependent. I could not retrieve the text of claims 1–6, so I cannot state with confidence what claim 1 covers (it may be the method claim matching the abstract).

• Claim 7 (polynucleotides, independent — fragments): An apparatus with a glass substrate whose surface carries linker molecules bearing functional groups reactive with nucleotides or nucleosides; that surface is mated to a body having a cavity and seals the cavity so the linker molecules are in fluid communication with it; the cavity is less than 1000 μm (1 mm) deep; an inlet port and an outlet port communicate with the cavity; and a fluid-flowing means is coupled to the inlet.
• Claim 8 (polypeptides, independent — fragments): Same architecture, but the surface functional groups are reactive with amino acids, and the substrate surface contacts and seals the cavity.
• Claim 10 (polypeptides, independent — fragments): A body with a sealed cavity disposed therein, less than 1000 μm deep, containing a substrate whose surface has amino-acid-reactive functional groups.
• Claim 11 (polypeptides, independent — fragments): A body having a sealed cavity less than 500 μm (0.5 mm) deep. (This is the narrowest of the depth-based claims.)
• Claim 15 (polypeptides, independent — fragments): Glass substrate + linker molecules reactive with amino acids, mated to a cavity-sealing body, cavity <1000 μm deep, inlet/outlet ports, a fluid-flowing means positioned to deliver fluid into the cavity, and a means for heating the cavity.

Dependent claims retrieved:

• Claim 9: adds a pump positioned to flow fluid through the cavity (inlet → outlet).
• Claim 12: the functional groups are protected with a protecting group.
• Claim 13: the protecting group comprises MeNPOC.
• Claim 14: the protecting group is nitroveratryloxycarbonyl (NVOC).

This is consistent with the Google Patents "claims" keyword cloud, which contains substrate, protecting group, nucleotide, heat treatment, sealing — i.e., photolabile-protecting-group chemistry plus a sealed, optionally heated, thin reaction cavity.

4. Litigation / docket check

• US district court (as linked by Google Patents): Affymetrix, Inc. v. Illumina, Inc., D. Del. 1:07-cv-00670, filed 2007-10-24, patent infringement (35 U.S.C. § 271). Confirmed on the Docket Alarm/Justia docket page. Google Patents labels the patent's family as having litigation.
• Context (GenomeWeb, 2007): Affymetrix's late-2007 complaints in the US, UK and Germany asserted, among others, US 6,420,169 ("Apparatus for Forming Polynucleotides or Polypeptides") against Illumina's BeadArray and Solexa-acquired products. This is a different cohort from the five patents tried in March 2007 (which produced the $16.7M jury verdict and 15% royalty in the earlier 2004-filed Delaware case).
• CAFC 2026 dockets: I found no CAFC 2026 appeal activity referencing 6,420,169. Given the recorded status (Expired – Fee Related; anticipated expiration 2019-07-16, ~17 years of family history), 2026 Federal Circuit activity concerning this patent is implausible. Stated uncertainty: absence of hits in my web searches is not proof of absence from the CAFC docket; I did not query the Federal Circuit's CM/ECF or PACER directly.

5. Bottom line

US 6,420,169 is an Affymetrix patent from the original light-directed (VLSIPS) synthesis family, claiming the shallow sealed flow-cell/cavity reactor — a substrate surface with photolabile-protected, monomer-reactive linker groups mated to a body defining a sub-1 mm (and sub-0.5 mm) deep cavity with inlet/outlet ports, fluid-flow means, and optionally heating. It was asserted against Illumina in the 2007 Delaware action (1:07-cv-00670) and is now expired. The precise scope of claim 1 and the total claim count remain unverified in this session.

I'll search for litigation involving US patent 6420169.

Let me dig into the litigation reference found on Google Patents and search litigation databases.

I'll verify the structured data against live sources before drafting, since the prompt instructs me to flag any proceedings the ODP ingest may have missed.

The patent was asserted in district court and was subject to reexamination requests (a pre-AIA mechanism, not an AIA trial), so I need to distinguish those. Let me search specifically for AIA trial numbers.

No IPR/PGR numbers are surfacing. Let me run two final targeted checks — reexamination (the pre-AIA mechanism Illumina actually used) and any PTAB docket entry.

PTAB Proceedings — US 6,420,169 ("Apparatus for forming polynucleotides or polypeptides")

Verification note before the findings. I treated the structured PTAB proceedings on file block (USPTO ODP) as canonical — it returns no AIA trial proceedings. I then ran independent searches (patent number paired with IPR2012-/IPR2013-/IPR2014- prefixes, "PTAB"/"Patent Trial and Appeal Board," and petitioner-side queries) and surfaced no IPR, PGR, or CBM number involving this patent. My search coverage is not a substitute for the PTAB API/E2E docket; the authoritative record remains the ODP block and PTAB E2E. Everything below that is not an AIA trial is labelled as such, and inferences are flagged.

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 defensive posture this gives a defendant is not the usual "the patent survived an IPR and is hardened." It is stronger and simpler than that: there is no PTAB record to attack or to rely on, because the patent expired on 2019-07-16 (listed "Anticipated expiration" on Google Patents, legal status "Expired – Fee Related"). A defendant facing assertion today should lead with expiration and the 35 U.S.C. § 286 damages window, not with PTAB estoppel or an FWD.

AIA trial proceedings

None. There are no proceedings to list, because no IPR, PGR, or CBM was ever filed against US 6,420,169. No FWD exists, no institution decision exists, no panel exists to name, and no § 315(e)(2) or § 325(e)(2) estoppel attaches to anyone.

Per the constraints in this task, I am not populating the per-proceeding template with invented numbers, panels, or dispositions. (Note also that this patent is not a CBM candidate — it is a microarray synthesis apparatus/method patent, not a "financial product or service" patent under AIA § 18.)

Adjacent proceedings that are not AIA trials (flagged, since a defendant will encounter them in a diligence file)

(a) Ex parte reexamination request — Illumina, Inc. (pre-AIA mechanism)

• Filed: 2007-10-19 (Illumina's own statement, reported 2007-10-25, News-Medical)
• Scope: Illumina requested reexamination of "these five patents" — the five U.S. patents at issue in the Delaware litigation, which reporting identifies as US 5,902,723; 6,403,320; 6,420,169; 6,576,424; and 7,056,666 (GenomeWeb).
• Outcome — NOT CONFIRMED for '169. A December 2007 headline reports the USPTO "Orders the Re-Examination of Two Patents Included in the Patent Litigation between Illumina and Affymetrix." I could not confirm from available sources which two patents were ordered, or whether any reexamination certificate ever issued for '169. Consistent with this, the patent is indexed as US6420169B1 with no C1/C2 reexamination-certificate kind code — which may indicate the request was denied for '169, but it may equally be an indexing gap. This must be verified in PatentCenter before relying on it.
• Defensive value: none adverse — reexamination creates no IPR-style estoppel. But if a certificate exists and narrowed the claims, the amended claim text governs any infringement analysis. Pull the file history.

(b) District court litigation (the source of the Google Patents "Family has litigation" flag)

• Affymetrix, Inc. v. Illumina, Inc., D. Del. (case flagged as 1:07-cv-00670 via Unified Patents litigation data), announced 2007-10-23/24. US 6,420,169 was one of the patents asserted against Illumina's Solexa sequencing technology (GenomeWeb; IBO).
• Earlier Delaware action (filed 2004-07-26): a March 2007 jury found Illumina infringed "one or more claims" of all five asserted Affymetrix patents and awarded >$16.7M at a 15% royalty; validity was expressly reserved for a later phase (Illumina 10-K excerpt). Reporting indicates the 2007 U.S. suit re-asserted the same five patents, which — if correct — means '169 was among the patents a jury found infringed. I am flagging this as an inference from matching patent lists, not a confirmed docket fact.
• Critical limit: no court is reported to have adjudicated the validity of '169, and I could not confirm the disposition of 1:07-cv-00670. Check PACER/CourtListener before representing otherwise to a court or adversary.

Strategic summary

Canceled vs. sustained vs. untested. There is nothing canceled and nothing sustained — all claims of US 6,420,169 are UNTESTED at the PTAB. (Third-party search reports cite the document's claims as "1-23"; confirm the exact independent/dependent structure against the printed claims, which were not in my source set.) The patent's practical status is driven by term, not by trial: priority traces to 1989-06-07 / 1990-03-07 (US 07/492,462), it issued 2002-07-16 from an application filed 1994-11-30, and it expired 2019-07-16. Its entire family (US 5,143,854; 5,424,186; 5,445,934; 5,744,305; 5,800,992; 6,403,320; 6,416,952; 6,506,558; 6,576,424, etc.) shares that 1989/1990 priority and is likewise expired. If a demand letter cites "the '169 family," the whole family is time-barred from prospective relief.

Estoppel landscape. § 315(e)(2) / § 325(e)(2) estoppel requires an instituted IPR/PGR that reaches FWD. Here there is no IPR, therefore no estoppel, therefore every ground remains available to a defendant — § 102, § 103, and § 112 are all untouched by any Board proceeding. The ex parte reexamination (if one concluded) creates no statutory estoppel against third parties; only prosecution-history effects flow from it. Conversely, the § 315(b) one-year clock is likewise un-triggered, but an IPR in 2026 over a patent that expired in 2019 has only narrow salvage value (knocking out pre-expiration damages exposure in a pending case) — and no amendment path of practical use to the owner.

Pattern signals. No serial petitioner (no petitioner at all). No Patent Owner appeal to the Federal Circuit from any PTAB decision, because there is no PTAB decision. No defensive aggregator (e.g., Unified Patents) IPR in the chain — the Unified Patents link in the structured data is litigation tracking, not a filing by Unified. The invalidity pressure historically came from a competitor (Illumina) via reexamination + district court, which is a pre-AIA fact pattern consistent with the 2007–2010 vintage. Current assignee of record is Affymetrix, Inc. (now within Thermo Fisher Scientific — background only; verify current ownership in the USPTO assignment records before directing a license or covenant-not-to-sue inquiry).

Recommended next steps

1. Lead with expiration. Any complaint filed today carries a § 286 damages window looking back six years (i.e., 2020-01-01 → 2026), which falls entirely after the 2019-07-16 expiration — leaving no recoverable pre-expiration damages. Injunctive relief is unavailable. This is a dispositive posture in most scenarios and does not depend on any PTAB record.
2. Confirm the reexamination history before filing anything. Search PatentCenter for the '169 application number 08/348,471 and any reexamination proceeding (90/… prefix) and for a reexamination certificate. I could not confirm which of the five 2007 reexamination requests were granted, nor whether '169's claims were confirmed or amended. Do not assert "no certificate exists" in a brief without checking.
3. Confirm the district court record. Pull 1:07-cv-00670 (D. Del.) and the 2004 Delaware action on PACER or CourtListener; specifically determine (i) whether '169 was in the 2004 case at the March 2007 infringement trial, (ii) what happened to the validity phase, and (iii) whether the 2007 action was resolved, dismissed, or settled. Note that a finding of infringement of "one or more claims" in 2007 is not a validity adjudication and has no collateral effect here.
4. Because there is no IPR estoppel, keep all invalidity theories on the table — § 102/§ 103 on the 1989–1990 priority date (the field is dense with pre-1990 solid-phase synthesis art, and the European counterpart EP 0 619 321 was revoked by the opposition division and the appeal was decided in T 0384/08, per EPO Boards of Appeal — useful comparative ammunition), plus § 112 for any claims narrowed in reexamination.
5. If a complaint is served imminently and pre-expiration damages are somehow still in play, the § 315(b) clock starts on service — file the IPR within one year. But treat this as a fallback: the expiration argument should resolve the matter without PTAB fees. Link the operative PTAB record (or its absence) via PTAB E2E and, for any decision cited, USPTO PTAB Decisions.

Bottom line for a defendant: there is no PTAB activity to worry about and none to hide behind — but you do not need it. Claims 1-23 of US 6,420,169 were never tested at the Board, and the patent expired on 2019-07-16. An assertion premised on ongoing or future infringement is unsupportable; verify the reexamination and district court files to close the last gaps.

I'll research the assignment record and litigation history for US 6420169.

Let me dig deeper into the assignment records, Affymetrix corporate history, and the litigation posture.

Assignment & Ownership Analysis — US 6,420,169 B1

Data provenance note (read first). The authoritative sources for this section are the USPTO Patent Assignment Center (https://assignmentcenter.uspto.gov/) and its index at https://assignment.uspto.gov/patent/index.html. The event data below is taken from the Google Patents "Legal Events / reassignment" record for US 6,420,169, which mirrors the USPTO assignment database. The Google Patents legal-events feed reproduces the execution/recording events, parties, and conveyance type, but it does not expose reel/frame numbers or the correspondent of record, and I was unable to retrieve those specific fields for this patent within the research budget for this task. I therefore report the chain with the fields I can verify and flag the reel/frame and correspondent columns as a genuine data gap rather than supplying placeholder numbers. No assignment is invented or inferred.

Inventors

Unusual-pattern check. There is no "all inventors departed within 12 months" fire-sale precursor here. The inventor group split cleanly into company employees (Read, Fodor, Hoeprich) and Stanford academic collaborators (Stryer, Pirrung) — the classic Affymax/Affymetrix model of in-house scientists plus consulting professors. Fodor remained with the entity family through the Affymetrix spin-out and IPO (June 1996), so the continuity between inventor and eventual owner is intact. This is a normal corporate/organic-chemistry research-team signature, not a precursor to a portfolio dump.

Original assignee

Affymetrix, Inc. (Santa Clara, CA) — the entity named on the issued patent (per the D. Del. AO-120 report of record, which lists holder "Affymetrix Inc." for the '169 patent).

• Product embodying the claims: Yes. Affymetrix commercialized the GeneChip® microarray system (first DNA microarray on the market, 1994), built on the photolithographic in-situ synthesis technology described in this family. Its microarray/genotyping/genetic-analysis products are the direct commercial embodiment.
• Primary line of business: Life-sciences tools — DNA microarrays, genetic analysis, genotyping, cytogenetics, gene expression.
• Current status: Acquired, still operating. Affymetrix (Nasdaq: AFFX) was acquired by Thermo Fisher Scientific Inc. (NYSE: TMO) for ~$1.3B ($14.00/share, announced 2016-01-12; closed 2016-03-31, integrated into Thermo Fisher's Life Sciences Solutions segment). Affymetrix survived as a subsidiary entity. Not dissolved, not in bankruptcy. (A competing ~$1.6B bid by Origin Technologies / Centrillion, formed by former Affymetrix employees, was withdrawn.)

Assignment timeline

Recorded events, from the USPTO-derived Google Patents reassignment record. Reel/frame and correspondent-of-record could not be confirmed from the sources reachable in this task (see provenance note) and are deliberately left unstated rather than guessed.

• Executed/recorded 1995-03-23 — Reel NNNNNN/NNNN (not retrieved)

  • Conveyance: Assignment
  • Assignor: J. Leighton Read; Stephen P.A. Fodor; Lubert Stryer; Michael C. Pirrung; Paul D. Hoeprich, Jr. (all named inventors)
  • Assignee: Affymax Technologies N.V. (Amsterdam, Netherlands)
  • Correspondent: not retrieved — cannot assess recurrence; flagged as a gap.
  • Context: Original inventor→company assignment of rights to the applicant/employer entity (Affymax), the routine first link in the chain.

• Executed/recorded 1996-06-10 — Reel NNNNNN/NNNN (not retrieved)

  • Conveyance: Assignment
  • Assignor: Affymax Technologies N.V.
  • Assignee: Affymetrix, Inc.
  • Correspondent: not retrieved.
  • Context: Internal spin-out / corporate reorganization — the microarray (GeneChip) IP was carved out of Affymax and moved to Affymetrix as the subsidiary became the operating microarray business.

• Executed/recorded 1999-09-07 — Reel NNNNNN/NNNN (not retrieved)

  • Conveyance: Merger (reorganization; California corp → Delaware corp)
  • Assignor: Affymetrix, Inc., a California corporation
  • Assignee: Affymetrix, Inc., a Delaware corporation
  • Correspondent: not retrieved.
  • Context: Change of domicile only — internal reincorporation merger; no change in beneficial ownership.

Post-1999: the Google Patents legal-events feed for this patent shows no further recorded assignments. The two later corporate events that touched Affymetrix are not assignments of title to this patent:

• 1993/1996 spin-out and June 1996 IPO (AFFX) — the 1996 assignment above captures the relevant transfer.
• 2016-03-31 Thermo Fisher acquisition of Affymetrix — structured as a stock/acquisition transaction in which the Affymetrix, Inc. entity survived as a subsidiary; title to the patent therefore remained with Affymetrix, Inc. (now a Thermo Fisher Scientific subsidiary) and no USPTO assignment record was required. The patent itself expired 2019-07-16 (Google Patents "anticipated expiration"; status Expired – Fee Related), consistent with the 17-years-from-grant term that applies to this pre-URAA, pre-June-8-1995-filing family.

Timeline diagram

timeline
    title Ownership of US 6420169
    1989 : Priority date
    1994 : Application filed by Affymetrix
    1995 : Inventors assign to Affymax Technologies NV
    1996 : Affymax transfers to Affymetrix Inc
    1999 : Merger into Affymetrix Delaware corp
    2002 : Patent issued to Affymetrix
    2007 : Affymetrix sues Illumina in Delaware
    2008 : Illumina pays 90M and case settles
    2016 : Thermo Fisher acquires Affymetrix
    2019 : Patent expires

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. The chain terminates at Affymetrix, Inc., a long-lived operating company with shipped products (GeneChip). No LLC with an "IP / Holdings / Licensing / Ventures" suffix, no registered-agent-only address, no single-member shell appears anywhere in the recorded chain.

2. Known asserter in the chain — Not present. No assignee matches any public NPE list (Acacia, Marathon, IV, IPNav, Wi-LAN, Conversant/Mosaid, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, Spangenberg entities, etc.). Every assignee in the chain is an Affymax/Affymetrix family entity or Thermo Fisher.

3. Repeat correspondent across the chain — Unclear / not assessable. The correspondent of record is not exposed in the Google Patents event feed, and I could not retrieve the reel/frame entries from the Assignment Center within this task. I will not infer recurrence without the underlying records; this signal is logged as a data gap, not a finding. (If pulled directly, expect a single Affymetrix/Affymax corporate-IP attorney across the three 1995–1999 recordings — that would be an internal-reorg signature, not an NPE one.)

4. Cascading transfers — Not present. Three recorded transfers span 1995-03-23 → 1999-09-07 (≈4.5 years), all internal (inventors→Affymax, Affymax→Affymetrix, CA→DE merger). No <24-month chained-LLC sequence, no shared shell addresses.

5. Pre-litigation transfer — Not present. The 2007 suit was brought by the entity that had held title since 1996/1999. There is no assignment within 6 months (or at all) before the 2007-10-24 filing — no chain arranged for the assertion, and no venue-setup transfer.

6. Bankruptcy fire-sale — Not present. Affymetrix was solvent throughout; it was sold in a $1.3B all-cash acquisition (Thermo Fisher, 2016), not a Chapter 7/11 asset sale. No Kodak/Nortel-style patent auction.

7. Privateering — Not present. Affymetrix asserted the patent in its own name, in its own interest, against a direct commercial competitor — the opposite of transferring to an NPE to assert on its behalf.

8. Defensive aggregator (anti-NPE) — Not present. The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. The patent is held by the operating family (now Thermo Fisher) and has expired.

Verdict

Operating-company assertion.

Justification: the recorded chain (inventors → Affymax Technologies N.V., 1995-03-23; Affymax → Affymetrix, Inc., 1996-06-10; Affymetrix CA → Affymetrix DE merger, 1999-09-07) is an unbroken internal-corporate lineage ending at a company that shipped the GeneChip microarrays embodying the claims — not a shell or asserter. The patent was asserted by that same operating owner against a genuine competitor, Illumina, in Affymetrix, Inc. v. Illumina, Inc., D. Del. 1:07-cv-00670 (filed 2007-10-24), where US 6,420,169 is listed on the AO-120 report alongside US 5,902,723, 6,403,320, 6,576,424 and 7,056,666; the dispute ended in the January 2008 $90M Illumina payment plus a covenant not to sue. No shell transfer, known-asserter, cascading-LLC, pre-litigation-transfer, bankruptcy, privateering, or defensive-aggregator signal is present. Two caveats: (a) reel/frame and correspondent fields could not be verified and are flagged as a data gap; (b) the patent expired 2019-07-16 (Expired – Fee Related), so it is no longer an assertable asset regardless of ownership.

Verification link (USPTO Assignment Center, search by patent number):
https://assignmentcenter.uspto.gov/ — query "6420169".

I'll research patent US6420169 in the USPTO database and identify its cited prior art references.

Let me retrieve the full "References Cited" section from the patent's front page.

Let me get the complete cited references list from additional sources.

Prior-Art Analysis — US Patent 6,420,169 B1

0. Two contradictions/limitations flagged up front

1. Date contradiction. The task framing says "Current Date: April 26, 2026," while the tool environment and the source page I fetched are stamped 2026-09-27. I use 2026 for currency statements but note the mismatch.
2. Your previously-generated "Litigation summary" is essentially empty. It contains only process narration ("I'll search for litigation involving US patent 6420169… Let me dig into the litigation reference…") with no actual findings. I therefore treat it as non-authoritative and cannot cross-reference substantive content from it. Separately, the authoritative Google Patents record does contain a litigation hook (Delaware District Court case 1:07-cv-00670, "First worldwide family litigation filed" per Darts-ip, family 27502959), which the earlier section did not capture.

1. Verification of the patent record (USPTO / Google Patents / FPO)

I confirmed this number specifically; I did not return results for near-numbers (6,420,169 B1 vs. e.g. 6,416,952 B1, 6,420,169-family members 6,379,895 / 6,403,320 / 6,506,558 / 6,919,211, all of which are different patents in the same family).

Important textual limitation: neither of the two full-text mirrors I could retrieve (Google Patents, FreePatentsOnline) returned the verbatim claim set of 6,420,169 in the fetched excerpts. The abstract confirms the claims are directed to "a method for synthesizing oligonucleotides on a solid substrate… irradiation of a first predefined region… removal of a protecting group… contacting with a first nucleotide… repeating… to form an array of diverse oligonucleotides." I therefore map prior art to claim subject matter, not to verified claim numbers, and say so explicitly below.

2. The critical § 102 timing problem (this governs everything else)

US 6,420,169 is a continuation (via 08/348,471, filed 1994-11-30) claiming benefit back to 1989-06-07. That creates two very different prior-art windows, and conflating them is the most common error in analyzing this patent:

• If the claims are entitled to 1989-06-07: only references published or patented before 1989-06-07 can anticipate under § 102(a)/(b). Everything dated 1990–1994 on the face of the patent is not § 102 art and was cited only as background/context or under a different theory.
• If the claims are NOT entitled to 1989-06-07 (e.g., new matter in the 1990 or 1994 applications, or a claim not supported by the 1989 disclosure): the 1990–1994 references become live § 102(a) / § 102(e) art.

This matters enormously here because a large block of the cited patents is dated 1990–1992 — after the 1989 date but before the 1994 filing. Those are the references with the greatest practical § 102 exposure, and they include the family's own parent (US 5,143,854) plus US 5,079,600 (Schnur), US 5,153,319 (Caruthers), US 5,047,524 (Andrus), US 5,141,813 (Nelson), US 5,112,962 (Letsinger), US 4,981,985 (Kaplan), US 4,973,493/4,979,959 (Guire), US 4,965,188 (Mullis) — all listed as "US Patent References" on the 6,420,169 face.

Family-members-are-not-prior-art point: US 5,143,854 (Pirrung), US 5,424,186 (Fodor), US 5,445,934 (Fodor), US 5,510,270 (Fodor), US 5,770,456, WO 90/15070 and WO 92/10092 share the same inventors/applicant and/or are in the same priority chain. They appear throughout the citations of other patents, but they are not § 102 prior art against 6,420,169 — they are the parent/sibling invention.

3. Cited references (as retrieved — partial, explicitly incomplete)

I retrieved the printed "References Cited" list in two chunks. The full list on the face of the patent numbers roughly 200+ U.S. patents plus foreign documents and other publications; I was able to verify the following, but I could not retrieve the complete list, the "Foreign Patent Documents" block, or the "Other Publications" block. I will not fabricate entries.

3a. Pre-1989 U.S. patents (candidate § 102(a)/(b) art if the 1989 date holds)

3b. 1989–1992 U.S. patents (NOT § 102 art against a 1989-06-07 date; live only if priority fails)

(The list continues into additional U.S. patents, foreign patent documents, and other publications that I could not retrieve. Treat § 3 as representative, not exhaustive.)

4. Most relevant prior art — ranked

Tier 1 — Closest art, and the only references legally capable of anticipating a 1989-priority claim

1. U.S. 4,689,405 — Frank et al., "Simultaneous synthesis of a plurality of different oligonucleotides on a solid phase," granted 1987-08-25.
The single most on-point pre-1989 reference. It discloses synthesizing many different oligo sequences in parallel on discrete solid supports (cellulose discs arrayed in columns) — i.e., the "plurality of diverse sequences at known locations" concept.

• Anticipation potential: would reach claims directed to parallel/plural-site synthesis of diverse polynucleotides on a solid support. It does not disclose light-directed masking of a single contiguous substrate, so it does not anticipate claims reciting irradiation-through-a-mask deprotection. If the 6,420,169 claim is a method of forming an array on a single substrate by selective irradiation, Frank is the best § 103 combinable reference, not an anticipatory one.

2. U.S. 4,631,211 — Houghton, granted 1986-12-23 (plural reaction vessels / packaged solid supports for multiple peptide synthesis).

• Anticipation potential: apparatus/plural-site aspect. Again not light-directed.

3. U.S. 4,458,066 and U.S. 4,500,707 — Caruthers et al. (1984-06 and 1985-02).

• Only reach the chemistry layer: protected nucleotide coupling to a solid support. They do not disclose spatially selective photodeprotection. Relevant to any claim reciting "contacting with a nucleotide to couple to the substrate."

4. U.S. 4,728,502 — Hamill (1988-03).

• Solid-phase synthesis on a surface/array format; cited in the family's own background. Best reaches substrate-surface synthesis claims, not photolithography.

5. U.S. 4,683,202 — Mullis (1987-07).

• § 102 exposure only if a claim recites amplification. Peripheral.

Tier 2 — Art that anticipates only on a "priority date fails" theory (1990–1992)

These are the references with real § 102(a)/(e) bite if the 6,420,169 claims are not entitled to 1989-06-07 (they pre-date the 1994-11-30 filing but post-date 1989-06-07):

• U.S. 5,079,600 (Schnur et al., 1992-01-07) — "High resolution patterning on solid substrates." Closest to the mask/pattern element.
• U.S. 5,091,652 (Mathies et al., 1992-02-25) — laser/confocal fluorescence scanning (relevant to any detection-apparatus claim).
• U.S. 5,112,962 (Letsinger et al., 1992-05-12) and U.S. 5,141,813 (Nelson, 1992-08-25) — solid-phase oligo anchor/support chemistry.
• U.S. 5,153,319 (Caruthers et al., 1992-10-06) — polynucleotide preparation process.
• U.S. 5,047,524 (Andrus et al., 1991-09-10) and U.S. 5,026,773 (Steel, 1991-06-25) — automated synthesis apparatus (relevant to the "apparatus" title of 6,420,169).
• U.S. 4,981,985 (Kaplan et al., 1991-01-01) — photolabile-group synthesis.

Caveat: even here, none of these alone disclose the full combination (photoremovable protecting group + patterned irradiation of one region while not irradiating another region of the same substrate + iterative coupling to build a spatially addressed array). No verified reference in Tier 2 is a clean § 102 anticipation of that combination.

Tier 3 — Not prior art (same family / same inventors)

US 5,143,854; US 5,424,186; US 5,445,934; US 5,510,270; US 5,770,456; US 6,379,895; US 6,403,320; US 6,506,558; US 6,919,211; WO 90/15070; WO 92/10092. Same inventive entity/priority chain — § 102 cannot be met by them.

Tier 4 — Non-patent literature that should be in the "Other Publications" block (could not verify verbatim)

I could not retrieve the "Other Publications" block of 6,420,169, so I list these as expected/commonly cited in this family rather than as verified citations: Merrifield, J. Am. Chem. Soc. 85:2149-2154 (1963); Geysen et al., J. Immunol. Meth. 102:259-274 (1987); Frank & Döring, Tetrahedron 44:6031-6040 (1988); Caruthers, Science 230:281-285 (1985); Fodor et al., Science 251:767-777 (1991). Verify against the patent's own "Other Publications" listing before relying on any of these.

5. Bottom-line answer on § 102

• The most relevant prior art cited on the face of 6,420,169 is U.S. 4,689,405 (Frank et al., 1987-08-25), followed by U.S. 4,631,211 (Houghten, 1986-12-23) and the Caruthers solid-phase synthesis patents (4,458,066; 4,500,707).
• No reference I verified is a clean, single-reference § 102 anticipation of the full light-directed, masked, iterative array-synthesis combination claimed. Frank/Houghten reach the "plurality of diverse sequences at known sites" element but lack patterned photodeprotection; Caruthers reaches the coupling chemistry but not spatial addressing.
• The strongest § 102 exposure is conditional: it depends entirely on whether the 6,420,169 claims actually retain the 1989-06-07 priority date. If they do, the large 1990–1992 block (Schnur 5,079,600, Letsinger 5,112,962, Nelson 5,141,813, Caruthers 5,153,319, Mathies 5,091,652, Andrus 5,047,524, Steel 5,026,773) is simply not prior art and the anticipation case collapses to the 1976–1988 references. If the claims do not retain 1989, that 1990–1992 block becomes the real battleground — and several of those references (particularly Schnur '600) then become strong § 102/§ 103 candidates.

6. Explicit gaps (stated rather than fabricated)

1. Claim numbers not verified. I could not retrieve the verbatim text of claims 1–N of 6,420,169; the "which claim(s)" mapping above is by claim subject matter, not by numbered claim.
2. Citation list incomplete. I verified roughly 45–50 U.S. references; the face of the patent contains substantially more, plus a foreign-documents block and an other-publications block, none of which I could retrieve. Do not treat § 3 as the complete 6,420,169 bibliography.
3. Reference substance. For a number of low-relevance entries I could confirm citation + date but not substantive content; those are marked "—".
4. No "Foreign Patent Documents" entries verified (e.g., the family's EP 0 476 014 / EP 0 619 321 / GB 2 248 840 and WO 90/15070 are relevant context but I did not confirm they appear on the 6,420,169 front page).

If you want, I can attempt targeted retrieval of (a) the full "References Cited" front page, (b) the verbatim claim set, and (c) the Foreign Patent Documents / Other Publications blocks to close gaps 1–4.

I'll ground this analysis in the actual prior-art record for US 6,420,169. Let me pull the cited references and the key art.

Let me pull the full cited-reference list and the non-patent literature for this patent.

Now I need the actual claim language of US 6,420,169 to analyze claim-by-claim. Let me retrieve it.

Obviousness Analysis — US 6,420,169 under 35 U.S.C. § 103

0. Scope, sources, and two explicit evidentiary caveats

Caveat 1 — the "Prior Art section" supplied to me is empty of citations. The full text of US 6,420,169 that I was given contains the bibliographic data, the legal-events timeline, the CPC/IPC classification block, and the machine-generated chemistry annotations (the "prior art keywords: substrate; cavity; synthesis; mask; group" list, and the compound-frequency table). It does not contain the "(56) References Cited" table or the "Referenced By"/"Citations" tables. I therefore retrieved the cited-reference lists independently:

• FreePatentsOnline "US Patent References" list for 6420169 — https://www.freepatentsonline.com/6420169.html
• uspto.report reference list for 6420169 — https://uspto.report/patent/grant/6420169
• Front-page references of the parent family member US 5,143,854 (Pirrung, Read, Fodor, Stryer) — http://www.everypatent.com/comp/pat5143854.html
• Reexamination certificate front page for US 5,445,934 (family member, Ex Parte Reexam 90/008,675) — https://patentimages.storage.googleapis.com/d9/17/63/91a6c7748f0925/US5445934.pdf
• Sibling specification text for the same "Very Large Scale Immobilized Polymer Synthesis" disclosure — https://patentimages.storage.googleapis.com/b8/9c/3d/57e780114b03c7/US20050148027A1.pdf

Caveat 2 — I do not have verbatim claim text for US 6,420,169. The claims were not in the supplied page text, and my searches did not return them. Everything below about claim scope is inferred from (a) the title "Apparatus for forming polynucleotides or polypeptides," (b) the abstract/keyword set on the page ("substrate," "cavity," "synthesis," "mask," "group," "fluid," "sealing," "heat treatment"), (c) the family specification's reactor disclosure (substrate mount engaging the substrate around its periphery, reactor space/cavity, mask placed on or focused on the substrate, pumped reaction fluids), and (d) the apparatus claims of related family members (e.g., US 5,405,783, US 5,445,934, US 5,744,305). Any element-by-element mapping I offer should be re-verified against the actual claim set before being relied on in a filing or opinion. I flag this rather than paper over it.

Cross-reference to the earlier section: the previously generated "Litigation summary" section identified the Delaware District Court case recorded for this patent (https://portal.unifiedpatents.com/litigation/Delaware%20District%20Court/case/1%3A07-cv-00670) and the Darts-IP worldwide-family litigation flag. I do not repeat that content here; note only that an obviousness defense under § 103 is precisely the kind of challenge validity litigation of this patent would have generated.

1. Two possible critical dates — the single most important variable

The page states, consistently with the family record:

This produces two constructively different § 103 analyses, and the difference is decisive:

• Route A — effective filing date 1989-06-07 (or 1990-03-07). Pre-AIA § 102(b) one-year bar art = everything published before 1988-06-07. Art published in the June 1988–June 1989 window is only § 102(a) art and is rebuttable by a pre-publication invention date. Critically, Fodor et al., Science 251:767–773 (15 Feb. 1991), the 1991 "light-directed, spatially addressable parallel chemical synthesis" paper, is not prior art at all on this route — nor is Khrapko et al., FEBS Lett. 256:118 (1989), nor Southern's WO 89/10977 (Nov. 1989), nor Pease et al., PNAS 91:5022 (1994). An obviousness rejection built on the 1991 Science paper would be legally defective.

• Route B — effective filing date 1994-11-30 for claims lacking support in the 1989/1990 disclosure. Because 6,420,169 is captioned "Apparatus for forming polynucleotides or polypeptides," while the 1989/1990 disclosure lineage is polypeptide-centric (the polynucleotide/oligonucleotide array work entered the family through later-filed applications in the '934/'639 chain), any claim that is not supported under § 112 ¶ 1 by the 1989/1990 parent gets only the 1994 date. On that route the entire 1989–1994 literature and patent corpus becomes available, including the inventors' own 1991 Science paper, US 5,143,854 (issued 1992-09-01), Pirrung et al., US 5,405,783 (1995), Southern WO 89/10977, Khrapko 1989, Pease 1994, and Caruthers/Letsinger-type phosphoramidite art.

Realistically, the highest-probability invalidity theory against 6,420,169 is therefore not a clean two-reference obviousness attack on the 1989 priority disclosure; it is an effective-date attack that converts the inventors' own 1991/1992 publications into § 102(b) art, followed by a § 103 attack. A corollary worth stating plainly: the apparatus claims most likely have the strongest 1989 support (the reactor/cavity/mask disclosure is squarely in the original specification as reproduced in the sibling publication at https://patentimages.storage.googleapis.com/b8/9c/3d/57e780114b03c7/US20050148027A1.pdf, ¶¶ [0013]–[0019]), while the polynucleotide-side claims have the weakest — the opposite of what one might assume.

A disqualification most analyses miss: pre-AIA § 103(c)

Several references on the FPO/uspto.report lists (e.g., US 5,143,854 — Pirrung et al., 1992-09-01; US 5,151,319 — Caruthers, 1992-10-06; and the rest of the Affymax/Affymetrix family) issued after the 1989/1990 priority date. As US patents they can only be prior art under § 102(e) if their filing dates precede the applicant's invention date. But under pre-AIA 35 U.S.C. § 103(c)(1), subject matter that qualifies as prior art only under § 102(e), (f), or (g) and that was commonly owned at the time the invention was made cannot be used in a § 103 rejection. Both US 5,143,854 and US 6,420,169 descend from Affymax Technologies N.V. / Affymetrix (per the assignment events on the Google Patents page and the '934 reexam certificate). Accordingly:

• The mere fact that US 5,143,854 is listed on the face of US 6,420,169's reference list is not evidence that it was § 103 prior art. IDS citations are not rejections, and a same-chain reference is usually § 103(c)-disqualified.
• Likewise, the 1991 Science paper is the inventors' own work and, if the 1989 date governs, is not prior art at all.

This is why the analysis below draws on non-family, third-party art — the only art that can do § 103 work on Route A.

2. The prior-art landscape, sorted by claim element

Stack 1 — Spatially addressed parallel synthesis at known locations (the "where is which sequence" problem)

Stack 2 — Photoremovable ("photolabile") protecting groups for stepwise synthesis

Stack 3 — Masked photolithographic patterning of surfaces (outside biology)

Stack 4 — Polynucleotide arrays at known locations for hybridization

Stack 5 — Fluid-handling / reaction-chamber apparatus

Automated peptide and DNA synthesizer flow columns (e.g., Andrus et al., US 5,047,524 "Automated system for polynucleotide synthesis and purification"), HPLC/flow-cell hardware, and standard 96-well microtiter plate formats established, before 1989, that a sealed reaction cavity having fluid inlets/outlets, through which reagents are pumped over a support, and removable by sealing means, was conventional laboratory engineering.

3. The obviousness combinations, with motivations

The controlling framework is Graham v. John Deere Co., 383 U.S. 1 (1966) as refined by KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007): scope and content of the prior art, differences between the prior art and the claims, level of ordinary skill, and secondary considerations; and "the combination of familiar elements according to known methods is likely to be obvious when it does no more than yield predictable results." A teaching, suggestion, or motivation need not be explicit in the references — a "design incentive" or "market pressure" plus a finite number of identified, predictable solutions suffices.

Combination 1 — The core "spatially addressed synthesis" claim set

Geysen US 4,833,092 (or Geysen 1984/1987) + Houghten US 4,631,211 + Frank US 4,689,405 + Zehavi & Patchornik 1972 / Amit 1974 (photolabile groups) + Barzynski US 3,849,137 / Levy IBM TDB 1967 (masked illumination).

Motivation. All five references are in the same field (solid-phase synthesis of diverse sequence libraries at known locations) and address the same express problem the '169 specification identifies: existing methods — Merrifield, multiple tubular reactors, foraminous containers, and 96-pin plates — "cannot practically be used to synthesize a sufficient variety of polypeptides for effective screening." Geysen, Houghten and Frank each explicitly confront the density/addressability limit and each solve it mechanically (pins, bags, discs). The predictable next increment, given that (i) photolabile protecting groups for α-amino acids were known since 1972 and expressly proposed for peptide synthesis, and (ii) the semiconductor industry had for two decades patterned surfaces with a mask plus light, was to make the deprotection step spatially selective by light rather than by physical separation of the synthesis sites. A PHOSITA facing the Geysen/Houghten throughput ceiling would have had every reason to substitute light for pipetting: it removes the mechanical manipulation that dominated cycle time and it uses the one tool (projection/contact lithography) already capable of defining sub-millimeter regions.

Strength. This is the strongest combination for a claim whose novelty resides in "different sequences at known locations, made by repeated light-directed deprotection + coupling." Its principal weakness is the absence of an explicit statement in any single reference that masking should be applied to a synthesis surface, and the real chemical obstacles (NVOC/MeNPOC stability, photolysis side-products, coupling efficiency) revealed by Pease et al. 1994.

Combination 2 — The masking/patterning claims specifically

Barzynski US 3,849,137 + Gazard 1980 + Morita 1983 + Schnur US 5,079,600 + Geysen US 4,833,092.

Motivation. If a claim recites illuminating the substrate through a mask to deprotect selected regions, each element is disclosed: the mask-plus-light patterning is Barzynski/Levy/Gazard/Morita; the substrate surface bearing reactive functionality and monomer units is Barzynski/Gazard/Schnur; the "selected regions define which sequence grows where" insight is Geysen. KSR teaches that combining elements disclosed in the prior art because each does what it is known to do is obvious. This combination is analytically clean on elements but requires an assumption about the claim's precise language that I could not verify.

Combination 3 — The polynucleotide-side claims

Southern EP 226,970 (1987) + Caruthers phosphoramidite chemistry (US 4,415,732 / 4,458,066 / 4,500,707) + Zehavi/Amit photolabile groups + Barzynski/Levy masked patterning + Bains 1988 / Lysov 1988 / Drmanac 1989.

Motivation. Southern establishes the objective (arrayed polynucleotides on a support, read out by hybridization). Bains/Lysov/Drmanac establish that a complete set of defined sequences should be arrayed at known locations. Caruthers establishes the iterative coupling/deprotection cycle and the protected phosphoramidite monomers. Barzynski/Levy establish that a mask plus light defines where chemistry happens. A PHOSITA seeking the arrays of Southern/Bains/Drmanac would predictably substitute a photolabile 5′ (or 3′) hydroxyl protecting group for the acid-labile dimethoxytrityl group, because photolabile groups on hydroxyls were known and because spatial addressing was the recognized problem. Note the instructive detail that appears in the later Duke patent US 5,908,926 ("5′ to 3′ nucleic acid synthesis using 3′-photoremovable protecting group," https://www.freepatentsonline.com/[5908926](/patent/5908926).html), whose background states as of the mid-1990s that light-directed synthesis "requires probe immobilization" and that the array count scales as 4 l rather than 4^l — the kind of uncontested background admission that supports the pre-1994 availability of the concept.

Date caution. On Route A (1989 priority), Khrapko 1989 and Pease 1994 drop out; on Route B (1994 date for unsupported claims) they come in as § 102(b) art.

Combination 4 — The cavity/seal/fluid-handling apparatus claims

Steel US 4,794,150 or US 5,026,773 (reaction vessel for solid-phase synthesis) + Andrus US 5,047,524 (automated synthesizer with fluid delivery) + a standard 96-well/flow-cell arrangement + Schnur US 5,079,600 (patterned substrate) + a mask-alignment stage (Levy 1967 / Barzynski 1974).

Motivation. If — as the family disclosure indicates — the apparatus claims are directed to a body that engages the substrate around its periphery to define a reactor space (cavity), with reaction fluids pumped through it and a mask placed on or focused on the substrate for illumination, then the claim is a combination of known, individually unpatentable mechanical elements: (i) a sealed reaction chamber for solid-phase synthesis (Steel; ordinary synthesizer columns); (ii) fluid inlet/outlet and pumping (Andrus; chromatography); (iii) an O-ring/gasket seal (ubiquitous); and (iv) a mask held in registration with the substrate (standard contact/proximity lithography, Levy; Barzynski). Under KSR, "the combination of familiar elements according to known methods is likely to be obvious when it does no more than yield predictable results," and improving a known apparatus by adding a known sealing element falls within the "predictable variation" branch. The patentability, if any, must come from the cooperating relationship between the cavity and the light path — i.e., that the reaction chamber must remain both fluid-tight and optically transmissive/irradiation-accessible while the mask is registered to the substrate. That functional synergy is the only place where a non-obviousness argument has purchase, and it is exactly the kind of argument that a claim term like "cavity" alone does not capture.

Combination 5 — The "monomer bearing a photoremovable group, repeat" claim

Houghten US 4,631,211 + Frank US 4,689,405 + Zehavi & Patchornik 1972 + Amit 1974 + Barzynski US 3,849,137.

Motivation. Stepwise coupling of a protected monomer, followed by selective deprotection and a further coupling, was the universal practice since Merrifield; the only substitution is the nature of the protecting group (photo- rather than acid/Base-labile) and the means of selectivity (masked light rather than physical compartmentalization). Both substitutions were individually known. This combination is strongest against any claim drafted generically as "providing a monomer having a photoremovable protecting group to a substrate having selected activated regions, and repeating."

Summary table

4. Why a PHOSITA would have combined — the affirmative motivation story

1. Identity of field and problem. All of Stacks 1 and 4 are "solid-phase synthesis of many different sequence-defined polymers at known locations, then screen them"; all of Stack 2 addresses "how to remove a protecting group selectively"; all of Stack 3 addresses "how to define very small regions of a surface spatially." KSR permits motivation to be found in "the interrelated teachings of multiple patents[;] the effects of demands known to the design community or present in the marketplace; and the background knowledge possessed by a person having ordinary skill in the art."
2. Recognized, express design incentive to miniaturize. Geysen's pin format, Houghten's bags, and Frank's discs each cap out at hundreds to low thousands of sequences because the format is defined by physical manipulation. Reducing feature size while increasing count is the canonical design incentive; the only available technology for sub-millimeter feature definition circa 1988–89 was photolithography, which necessarily puts a mask between the light source and the substrate.
3. The photochemistry was not speculative. Zehavi & Patchornik (1972) and Amit et al. (1974) made light a known deprotection trigger for α-amino functions, and Haridasan et al. (Chem. Abstr., Feb. 1989) applied 2-nitrobenzyloxycarbonyl photolysis specifically to peptide synthesis — all before 1989-06-07.
4. Predictability of result. Combining a known solid-phase cycle, a known photolabile group and known masked illumination yields only the expected result: sequences built where light struck. There is no new principle of operation — which is the KSR touchstone for "predictable variation."
5. The apparatus additions are routine. Seals, inlets, outlets, and heat treatment (recited in the keyword/claim-vocabulary set for this patent) are ordinary mechanical expedients; a patent cannot rest on the mere discovery that a reaction chamber must be closed while reagents flow through it.

5. Counterweights — why a § 103 attack on the 1989/1990 claims would probably fail

A rigorous analysis must state the other side, because the record here is unusually favorable to the patentee on Route A:

• No express TSM. No pre-June-1989 reference states that light should be used to address synthesis sites on a common substrate. The art used physical partitioning; the invention replaced it with optical partitioning. That substitution is not merely "a predictable variation" if the art contains no suggestion of it.
• Teaching away / contrary art. Geysen's own methodology teaches away from sharing a single continuous surface — the whole point of pins is that each sequence is on its own pin. Melding synthesis onto one surface risks cross-contamination during coupling, a concern the art treats as a reason to keep sites physically separate.
• Real, documented technical obstacles. Pease et al. (PNAS 91:5022, 1994), summarizing the inventors' own reduction to practice, shows that substantial additional work (MeNPOC photochemistry, photolysis side-reactions, coupling efficiency, light-scattering, mask alignment) was required after the concept. Post-KSR case law still requires that the combination have been "reasonably likely to succeed" — not merely "obvious to try" where the art identifies no finite set of solutions.
• Secondary considerations. The strongest evidence here — long-felt need for high-density parallel synthesis, failure of the mechanical approaches to scale, recognition by the field (the 1991 Science paper received the Newcomb Cleveland Prize for the best research article in Science; see the historical account at https://www.scienceopen.com/document_file/ad6ec2d2-6898-4359-9423-2fd2ebcfcf96/ScienceOpenPreprint/MicrochipEssay.pdf), commercial adoption of the resulting VLSIPS™/GeneChip arrays, and affirmative corroboration that at least one highly-experienced photochemist considered the NVOC/MeNPOC choice a genuine contribution — all cut against obviousness under the fourth Graham factor. (Note the licensing/royalty record is likely covered in the earlier "Litigation summary" section and should be folded in there rather than restated.)
• The family survived reexamination. Ex Parte Reexamination 90/008,675 (filed 2007-06-25) of family member US 5,445,934 concluded with the issuance of a reexamination certificate on 2009-04-21 (https://patentimages.storage.googleapis.com/d9/17/63/91a6c7748f0925/US5445934.pdf). That is not dispositive for 6,420,169 — different claims, and I found no reexamination certificate for 6,420,169 itself — but it shows the core disclosure withstood a district-of-art challenge.
• The § 103(c) shield. As discussed in § 1, the most tempting "prior art" (US 5,143,854, US 5,151,319, the rest of the Affymax/Affymetrix family) is either not prior art at all or is disqualified from § 103 if it qualifies only under § 102(e) and was commonly owned. Much of the art listed on the '169 front page was cited by the applicant in IDSs and was never applied by the examiner.

6. Bottom line

1. On a 1989-06-07 (or 1990-03-07) effective date, the best § 103 case is Combination 1 (Geysen + Houghten + Frank + Zehavi/Amit + Barzynski/Levy), and secondarily Combination 2 and Combination 5. It is a legally credible case — same field, same express problem, every element independently known — but it is genuinely vulnerable to the absence of any pre-1989 suggestion to replace mechanical with optical partitioning, to Geysen's implicit teaching away from a shared surface, and to strong secondary-consideration evidence. The 1991 Science paper cannot be used on this route.
2. On a 1994-11-30 effective date (claims unsupported by the 1989/1990 disclosure — most plausibly the polynucleotide-oriented claims), the picture changes materially: the inventors' own 1991 and 1992 publications become § 102(b) art, Pease 1994 and Khrapko 1989 come in, and Combination 3 (Southern + Caruthers + photolabile groups + masked patterning + Bains/Lysov/Drmanac) becomes a strong, arguably clean, § 103 case against the broad polynucleotide claims.
3. The apparatus/″cavity″ claims (Combination 4) carry the greatest § 103 risk of all, if they recite no more than a sealed reaction chamber for a masked, light-directed synthesis support, because KSR's "familiar elements according to known methods" doctrine is directly on point. Their survival depends entirely on whether the claims capture the functional cooperation between fluid-tightness and optical addressability rather than merely reciting the chamber.

Recommended next step for a definitive analysis: obtain the granted claim set of US 6,420,169 verbatim (USPTO PatentCenter / Global Dossier), the ex parte reexamination and prosecution histories for the '169 and its siblings, and the filing dates of the § 102(e)-potentially-qualifying references (Schnur '600, Andrus '524, Nelson '813, Letsinger '962, Mathies '652, Kaplan '985, Steel '773, Macevicz '007, Church '625) so each can be tested against the correct Graham/KSR/pre-AIA § 103(c) framework. Without the verbatim claims, the element-by-element mapping in § 3 above is an analytical scaffold, not a conclusion.

Unverified/residual uncertainty, stated explicitly: (i) I could not retrieve the granted claims of US 6,420,169, so all claim-scope statements are inferred; (ii) I did not verify the filing dates of the post-1990-issued US references, which is what determines whether they qualify under § 102(e); (iii) I did not verify the parties or outcome of Delaware 1:07-cv-00670, which should be reconciled with the earlier Litigation section; (iv) the EP 0 226 970 publication date (1987-07-01) and the Amit/Patchornik/Sheradsky 1974 citation are from memory and should be confirmed against the primary sources before being relied on. Where the supplied page and my retrieval agree — the 1989-06-07 prior-art date, the 1990-03-07 parent, the 1994-11-30 filing, and the expired status — there is no contradiction; the only internal oddity is that the page labels 1989-06-07 the "prior art date" while also showing a 1990-03-07 priority claim, which is explained by the 07/362,901 → 07/492,462 continuation-in-part chain documented on the '934 reexamination certificate.