US 7442629

I'll search for the specific patent number across USPTO and CAFC sources.

Let me check for litigation and any CAFC docket activity specifically involving this patent.

Let me run a couple more targeted checks on litigation/appeal status.

US Patent 7,442,629 — Summary

Important note on scope: My searches did not return any CAFC 2026 docket, appeal, or oral-argument entry naming U.S. Patent No. 7,442,629 (or "7442629"). I also did not find results for the similarly-numbered entries that appear in my searches (e.g., a Brazilian JUceC registration "7442629," a Lithuanian registry "7442629," and a Union Bankshares SEC filing tagged "7442629") — those are unrelated to this patent and I am not treating them as results for the patent. If a live CAFC 2026 appeal exists, I could not locate it and cannot confirm it. The only litigation on record for this patent in my search results is an older Massachusetts district court case, discussed below.

Bibliographic data (from the Google Patents full text)

The application is a continuation-in-part of US 10/950,248 ("Manufacture of Silicon-Based Devices Having Disordered Sulfur-Doped Surface Layers") and US 10/950,230 ("Silicon-Based Visible And Near-Infrared Optoelectric Devices"), both filed 2004-09-24. Of note, the patent states it was made with U.S. Government support (DOE contract DE-FC36-01GO11053 and NSF grant NSF-PHY-0117795), with certain government rights.

Abstract (verbatim)

"The present invention generally provides semiconductor substrates having submicron-sized surface features generated by irradiating the surface with ultra short laser pulses. In one aspect, a method of processing a semiconductor substrate is disclosed that includes placing at least a portion of a surface of the substrate in contact with a fluid, and exposing that surface portion to one or more femtosecond pulses so as to modify the topography of that portion. The modification can include, e.g., generating a plurality of submicron-sized spikes in an upper layer of the surface."

Independent claims in plain language

Claim 1 — Method of processing a semiconductor substrate (the "liquid-contact" method)
Place at least a portion of the substrate's surface in contact with a liquid, then expose that surface portion to one or more short laser pulses, thereby modifying the surface topography to produce surface features whose heights are less than about 1 micrometer. The core concept: laser texturing done under a liquid (e.g., water) yields sub-micron surface features.

Claim 15 — Method of processing a substrate (the "liquid layer + femtosecond pulse" method)
Dispose a layer of a liquid over at least a portion of a substrate surface, then apply a plurality of radiation pulses with pulse widths from about 50 fs to about 500 fs to that portion, generating surface features with heights less than about 1 micrometer. Unlike claim 1, which says "short laser pulses," claim 15 expressly recites the femtosecond pulse-width range.

Dependent claims add limitations rather than separate inventive scope: pulse widths of ~100 fs to a few ns (claim 2) and ~100 fs–500 fs (claim 3); liquid being polar or non-polar (claim 4), specifically water (5), silicon oil (6), or alcohol (7); substrate being a silicon wafer (8, 18); wavelengths ~400 nm to less than ~800 nm (9, 19); pulse energy ~10–400 µJ (10, 20); spikes with average height <1 µm and average width ~100–500 nm (11), heights ~200 nm–1 µm (12); pulse application rate ~1 kHz to ~1 Hz (13); 1–2,500 pulses (14); and spikes with heights ~100–500 nm (21).

Plain-language technical overview

The patent describes "black silicon"–type texturing produced by femtosecond laser irradiation while the silicon surface is submerged in a liquid (the worked example uses a Si(111) wafer in distilled water, irradiated with 400-nm, 100-fs, 60-µJ pulses from a frequency-doubled amplified Ti:Sapphire laser). The specification reports columnar spikes with ~500 nm typical height and ~200 nm typical diameter, spike separation of roughly 500 nm (matching the wavelength of initial ripple structures), and an oxide layer (~20 nm SiO₂, removable in HF) over the spikes. It theorizes that rapid cooling by the liquid shortens the molten-silicon lifetime, which limits capillary wavelength and leads to spikes one to two orders of magnitude smaller than those formed in a gas such as SF₆. Related embodiments add an electron-donating constituent (e.g., sulfur powder, or sulfuric acid in the aqueous solution) to also generate sulfur inclusions in the surface layer.

Litigation / appeal status

• Google Patents lists one U.S. litigation entry via Unified Patents: a case filed in the Massachusetts District Court, case 1:15-cv-13488 (2015; not a 2026 matter).
• No CAFC 2026 docket activity for this patent was found in my searches. Given the recorded adjusted expiration of 2025-01-07, the patent appears to have expired, which is consistent with the absence of a current appeal — but I cannot state with certainty that no 2026 Federal Circuit filing naming any of the family members exists.

Uncertainties

• I could not access the USPTO PatentCenter/Assignment records or a CAFC docket database directly; the above reflects the Google Patents full text (fetched as the authoritative source) plus search results.
• The "current assignee (Harvard University)" label on Google Patents is expressly disclaimed there as possibly inaccurate; the recorded assignment names "PRESIDENT & FELLOWS OF HARVARD COLLEGE."
• Family priority dating is inconsistent across sources (some fields show 2001-05-25, others 2004-09-24); the patent's own face shows a 2004-09-24 priority date via the two CIP parents.

Sources: https://patents.google.com/patent/US7442629/en (full text, bibliographic data, claims, litigation links); https://golden.com/wiki/US_Patent_7442629_Femtosecond_laser-induced_formation_of_submicrometer_spikes_on_a_semiconductor_substrate-ZXD9E39 (inventors, dates); https://portal.unifiedpatents.com/litigation/Massachusetts%20District%20Court/case/1%3A15-cv-13488 (D. Mass. litigation link).

I'll research litigation involving US patent 7,442,629. Let me search several sources.

Summary of findings for U.S. Patent No. 7,442,629

Confidence note up front: The authoritative patent record you supplied (Google Patents page for US7442629B2) lists exactly one litigation entry — a case in the District of Massachusetts, docketed as 1:15-cv-13488. However, the underlying court documents I retrieved indicate that the patents actually asserted in that case were US 8,680,591, US 8,080,467, and (initially) US 7,884,446 — and US 7,884,446 is a divisional of US 7,442,629, not the '629 itself. I could not retrieve a complaint or verdict form that asserts a claim of the '629 patent specifically. I flag this discrepancy rather than resolve it in favor of either source.

1. SiOnyx, LLC and President and Fellows of Harvard College v. Hamamatsu Photonics K.K. (et al.)

Sources:

• Google Patents (authoritative text you supplied): litigation link to Massachusetts District Court case 1:15-cv-13488 — https://patents.google.com/patent/US7442629/en
• CourtListener opinion (motion to declare exceptional): https://www.courtlistener.com/opinion/[9731274](/patent/9731274)/sionyx-llc-v-hamamatsu-photonics-kk/
• Court records noting the asserted patents and the '446 dismissal: https://www.govinfo.gov/content/pkg/USCOURTS-mad-1_15-cv-13488/pdf/USCOURTS-mad-1_15-cv-13488-6.pdf
• Judgment/injunction: http://fedcircuitblog.com/wp-content/uploads/2021/01/SiOnyx_OpBelow.pdf
• Docket metadata: https://ai-lab.exparte.com/case/dct/mad/1:15-cv-13488/sionyx-llc-v-hamamatsu-photonics-kk

Relationship note: US 7,884,446 (the Harvard patent dismissed by stipulation in this case) is listed in the '629 patent's own family data as a division of the '629 application — the "Related Child Applications" table lists US12/235,086 ("Division") → US7884446B2. That is the closest connection I could substantiate between the '629 patent and litigation activity.

2. Other SiOnyx enforcement campaigns (do NOT involve US 7,442,629)

For completeness, so you don't conflate them:

• SiOnyx, LLC v. Apple Inc. — D. Mass., filed Sept. 12, 2024 (three counts). Asserted patents are later SiOnyx patents (e.g., US 10,224,359 and related). An amended complaint was filed December 2024. No indication the '629 patent is asserted. Sources: https://www.law.com/therecorder/2024/09/16/apple-accused-of-infringing-on-digital-imaging-sensors-for-iphone-15-products/ ; https://www.law.com/corpcounsel/2025/01/08/apple-disputes-efforts-to-manufacture-imaging-sensor-claims-against-iphone-15-technology/
• In re Certain Sensors With Pixels and Products Containing Same, ITC Inv. No. 337-TA-1403 (complaint filed April 30, 2024 by SiOnyx; respondents include Samsung entities). Asserted patents: US 9,064,764; 9,905,599; 10,224,359; 11,069,737; 11,721,714 — not the '629. Source: https://www.usitc.gov/secretary/fed_reg_notices/337/337_1403_notice05312024sgl.pdf

Bottom line

• One litigation is recorded against the US 7,442,629 patent record: the Massachusetts case 1:15-cv-13488 (SiOnyx/Harvard v. Hamamatsu).
• Caveat: In the court filings I could access, the patents asserted were US 8,680,591, US 8,080,467 and US 7,884,446. US 7,884,446 is a divisional of US 7,442,629. I did not find a document affirmatively showing a claim of the '629 patent being asserted at trial. Google Patents attributes the case to the '629 file; the trial record appears to attribute the relevant Harvard patent to the '446 divisional. I could not resolve this from free/public sources.
• I found no other lawsuits, ITC investigations, or Federal Circuit appeals naming US 7,442,629, and no PTAB post-grant proceedings against it in the sources I could search. Because I could not query PACER or Unified Patents' litigation database directly (tool step limit reached), treat this as "none found" rather than a certified negative.

If you need certainty on whether the '629 claims were ever asserted, the definitive step is to pull the docket and complaint in D. Mass. 1:15-cv-13488 via PACER (or the Unified Patents litigation portal entry referenced in the patent text: https://portal.unifiedpatents.com/litigation/Massachusetts%20District%20Court/case/1%3A15-cv-13488).

I'll start by verifying the structured data with targeted searches for any AIA trial activity on this patent.

Those hits are all unrelated patents that happen to share the "629" shorthand. Let me search with more specific identifiers and litigation context.

Excellent — a key lead surfaced: IPR2025-00065 naming SiOnyx as Patent Owner and US 7,442,629. Let me pull the details on that and check for earlier proceedings.

Proceedings overview

Zero (0) AIA trial proceedings are on file against US 7,442,629 — the USPTO Open Data Portal returns no IPR, PGR, or CBM with this patent as the challenged patent, and my web searches surfaced no proceeding that names it as the subject patent. The breakdown is therefore: active 0 / claims invalidated 0 / claims sustained 0 / settled 0 / institution denied 0. The bottom-line defensive posture is unusual: no challenger has ever put this patent's claims through an AIA trial, so there is no PTAB-killed claim to hide behind — but the patent's own term has run out, which is a stronger defense than any FWD (see Strategic summary).

I want to be explicit about two things before the detailed analysis, because the search environment is noisy here:

1. The closest false positive is IPR2025-00065 — it is NOT this patent. Samsung Electronics Co., Ltd. v. SiOnyx, LLC, IPR2025-00065, challenges U.S. Patent No. 10,347,682, not 7,442,629. My first search hit appeared to connect IPR2025-00065 to "Patent No. 7,442,629 (Mazur)," but pulling the actual petition and Patent Owner papers shows the '682 patent is the challenged patent and the '682 patent's Exhibit/family listings simply cite Mazur's 7,442,629 as related art/related patent in the black-silicon chain. Filing date 2024-11-22; institution denied 2025-06-06 (Fintiv/discretionary denial, with the Patent Owner's sur-reply arguing the petitioner's Sotera stipulation was inadequate because Samsung's district-court invalidity case used additional combinations plus product art). Sources: ipverse case page, PTAB petition record. The "IPR2024-00064, -00065" numbering that appears in some 2025 amicus briefs against In re Motorola is a transposition of the same matter, not a second set of challenges.

2. There is no proceeding to profile in the "most-impactful first" format. Rather than invent hearing dates or panels, the sections below cover (a) the adjacent PTAB activity you should not confuse with this patent, and (b) the actual coercive posture on 7,442,629, which comes from district court and ITC, not the Board.

No proceedings listed — none to report

The structured "PTAB proceedings on file" block is the canonical USPTO ODP list and it is empty. I did not find an IPR, PGR, or CBM bearing 7,442,629 as the challenged patent in any source I could retrieve, so I will not assign a proceeding number, panel, or FWD to this patent. Anything purporting to be an FWD on 7,442,629 would be a fabrication on this record.

Strategic summary

Claim status: all 21 claims are UNTESTED at the PTAB. Claims 1–21 (the method claims directed to placing a semiconductor surface in contact with a liquid and exposing it to short/femtosecond laser pulses to generate sub-micron surface features) have never been the subject of an instituted AIA trial. There is no cancellation, no adverse final written decision, and therefore no claim of this patent is statutorily dead. Any clean-up on this patent came from prosecution and its family tree, not from the Board.

The dominant defensive fact is term, not invalidity. The Google Patents/ODP record for US 7,442,629 lists the patent as Expired – Lifetime, with an adjusted expiration date of 2025-01-07 (US7442629B2). That date is derived from the Office's term calculation (the application is a CIP filed 2005-08-04 claiming 2004-09-24 priority, with the usual § 154(b) adjustment and any terminal-disclaimer/two-patent overlap effects). Practical consequence: as of today (2026-09-28) more than 20 months have passed since that listed expiry, so the recoverable infringement window on the '629 patent itself is closed and an ongoing-infringement theory is unavailable. Because this is a printed-database term field rather than a Certificate of Correction I reviewed, treat the exact date as verify-before-you-rely, but the direction is not in doubt — a 2004/2005-priority patent is at or past end of term.

Estoppel: nothing attaches. Because there is no IPR/PGR FWD on this patent, no petitioner is barred under § 315(e)(2) or § 325(e)(1) from running § 102/§ 103 grounds on 7,442,629. Conversely, a defendant gets no free ride from anyone else's win. If you are being pressed on this patent, every prior-art ground remains fully available to you — including grounds that would normally be "reasonably could have been raised" fodder had a co-defendant already litigated it. My web search did find a real estoppel fight in the same family's litigation — the D. Mass. court's treatment of estoppel in SiOnyx, LLC v. Hamamatsu Photonics K.K., No. 1:15-cv-13488-FDS, analyzed whether the defendants were estopped as to the Adkisson reference and as to the HPK S9840 CCD sensor and its datasheets after a prior IPR and FWD (govinfo, ECF excerpt). That ruling is instructive on how a § 315(e)(2) fight plays out in this patent family (the court refused to estop Adkisson where there was a genuine dispute about whether a skilled searcher would have found it, and refused to hold the S9840 product ground estopped merely because documentation describing it existed). I could not confirm from the sources retrieved which patent number that underlying IPR challenged, so I am flagging rather than asserting a link to 7,442,629. Verify that docket before relying on the estoppel holding as to this patent.

Pattern signals. No repeat-petitioner pattern exists on this patent because there is no petitioner at all. The enforcement pattern, however, is aggressive and multi-front: the family's owner (SiOnyx, LLC, successor to the President & Fellows of Harvard College interests) litigated SiOnyx v. Hamamatsu in D. Mass. (jury verdict, ownership/breach claims, Fed. Cir. decision 2020-12-07 declining to reach willfulness and reversing on the foreign-patent ownership point — CCH copy of the opinion), and more recently pursued SiOnyx, LLC v. Samsung Electronics Co., Ltd., No. 2:24-cv-00408-JRG (E.D. Tex.) alongside ITC Inv. No. 337-TA-1403 (E.D. Tex. stay order). No defensive aggregator (e.g., Unified Patents as petitioner) appears anywhere in this patent's chain — the only Unified link is their litigation-data portal record of the D. Mass. case. Note the enforcement is now directed at later family members (e.g., the '682 patent challenged in IPR2025-00065, and the copending U.S. 10,361,083 / 10,741,399), not at the expired '629.

Where the real exposure sits. The priority-claiming continuations US 10,361,083 (from US 14/836,609) and US 10,741,399 (from US 15/896,940) are listed as Expired – Lifetime in the family table but carry later-adjusted expiries than the '629; a demand letter citing "the Mazur femtosecond-spike patent" may actually be asserting one of those. Check the patent number on the face of the letter before assuming this analysis applies.

Recommended next steps

1. If a demand letter or complaint cites US 7,442,629, lead with expiry, not invalidity. Pull the USPTO Patent Center "Patent Term Adjustment" determination and any terminal disclaimer, and confirm the 2025-01-07 adjusted-expiration listing. If it holds, the accused conduct after that date is non-infringing as a matter of law and the damages base is limited to the pre-expiry window (subject to the § 286 six-year lookback).
2. Do not expect a PTAB shortcut. With no FWD on this patent, there is no canceled claim to point to and no § 315(e)(2) estoppel to weaponize against the patent owner. Conversely, the Board is an open door if the patent were still in force — but a 2004-priority patent is squarely in the "settled expectations" discretionary-denial zone described in the 2025 PTAB practice changes, so any new petition on this patent would face a substantial discretionary-denial risk before the merits are ever reached.
3. Verify the Hamamatsu-era IPR. Confirm the challenged patent number and the FWD disposition for the IPR discussed in SiOnyx v. Hamamatsu (D. Mass. 1:15-cv-13488) via PTAB E2E and CourtListener. If it did reach 7,442,629, this analysis changes materially — you would then have a claim-level disposition and a live estoppel record to work from. I could not confirm that link on the record available to me, so treat it as an open item rather than a finding.
4. Since no PTAB activity exists, say so plainly in any opinion letter — and use the absence as a diligence signal pointing outward: check whether the assertion is really on the '682/'083/'399 continuations, where Samsung already tried (and lost on discretionary grounds) at IPR2025-00065, and where an ITC action is running.

Verification note: proceeding numbers above are quoted exactly as they appear in the sources cited. I did not locate any Federal Circuit appeal of a PTAB decision on 7,442,629 because no such decision exists on this record; the Fed. Cir. appeal discussed is the district-court appeal in SiOnyx v. Hamamatsu*, which is a different posture.*

I'll research the assignment chain and any litigation/licensing history for this patent.

Let me dig for the specific assignment reel/frame and any post-issuance transfers.

Ownership Chain Analysis — US 7,442,629 B2

Retrieval caveat (read first): This session's tools returned search results but I was not able to open the USPTO Assignment Center record set directly (https://assignmentcenter.uspto.gov/ / https://assignment.uspto.gov/patent/index.html). Accordingly, I cannot report reel/frame numbers or correspondent-of-record names for this patent, and I will not invent them. What I can state is grounded in (a) the Google Patents full text supplied as authoritative for this analysis (which carries the derived legal-events table), and (b) primary litigation and Federal Circuit documents retrieved from CourtListener, govinfo, and RPX Insight. Where a fact is unverified, I say so.

Inventors

Unusual-pattern check: one of two inventors (Shen) left the original assignee within roughly 12 months of the 2005-08-04 filing — the classic precursor to an orphaned-rights or fire-sale scenario. Here it did not produce that outcome, because the inventor-to-university assignment was already executed and recorded on 2005-10-03, i.e. before Shen's 2006 departure and only ~2 months after filing. Mazur never left Harvard. There is therefore no evidence of chain-of-title disruption at the inventor level.

Original assignee

The President and Fellows of Harvard College ("Harvard"), Cambridge, Massachusetts — the legal name of Harvard's governing corporation, which is the entity that appears on the issued patent and in the USPTO assignment record.

• Line of business: university / academic research institution (Harvard's Office of Technology Development handles licensing).
• Does it ship a product embodying the claims? No. Harvard is a non-practicing academic owner. It monetizes the family through an exclusive license, not through product sales. This is important for the NPE analysis below: under the rubric's own definition, "operating-company assertion" is not a perfect fit either, because the owner is non-practicing.
• Government interests: the patent face states it was made with U.S. Government support under DOE contract DE-FC36-01GO11053 and NSF grant NSF-PHY-0117795, "and the Government has certain rights in the invention." Under Bayh-Dole practice this is a government license/ march-in reservation, not a transfer of title — it does not appear as an assignment and should not be treated as one.
• Current status: operating; Harvard remains the assignee of record on Google Patents for this patent, and the 2019 district court judgment independently confirms Harvard's ownership of the sibling Harvard Patents (the '446 and '467 patents).
• Patent status: Expired – Lifetime; adjusted expiration 2025-01-07. The patent can no longer be asserted.

Related (but distinct) family issue: in the D. Mass. litigation, entry of final judgment (Doc. 834, filed 2019-09-24) transferred ownership of nine Hamamatsu patents (Nos. 9,614,109; 9,293,499; 9,190,551; 8,994,135; 8,916,945; 8,884,226; 8,742,528; 8,629,485; 8,564,087) to SiOnyx, LLC. US 7,442,629 is not among them — do not conflate that transfer with this patent's chain.

Assignment timeline

Finding: the Assignment Center-visible chain for US 7,442,629 contains a single recorded assignment — the original inventor-to-Harvard conveyance. There is no recorded post-issuance assignment, and therefore no reel/frame sequence to walk.

• 2005-10-03 (executed date not separately available) / recorded 2005-10-03 — Reel NNNNNN/NNNN: not retrievable in this session
  • Conveyance: Assignment (ASSIGNMENT OF ASSIGNORS INTEREST, per the Google Patents legal-events entry: "ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS)")
  • Assignor: Eric Mazur; Mengyan Shen (both, per the recorded-assignment entry naming "Assignors: MAZUR, ERIC, SHEN, MENGYAN")
  • Assignee: PRESIDENT & FELLOWS OF HARVARD COLLEGE
  • Correspondent: not retrievable — I could not open the Assignment Center record, so I cannot name the recording attorney/firm or their address. This is the single most important gap in this analysis; see the verification link at the end.
  • Context: routine university technology-transfer assignment — inventors assign to the institution that employed them and funded the work; executed after the 2005-08-04 filing and roughly 13 months after the 2004-09-24 priority date.

Everything else that touches this patent's economics is a license or a litigation event, not an assignment:

• 2006-02-10 — Harvard exclusively licensed its black-silicon patent portfolio to SiOnyx, Inc., per the D. Mass. court record ("The original exclusive patent-license agreement was dated February 10, 2006"), later amended to add new filings as they issued. Publicly announced October 2008. Harvard took equity plus downstream royalties rather than selling title. Licenses of this kind are frequently not recorded in the Assignment Center, and I found no evidence that this one was; title never left Harvard.
• 2008-09-22 / 2011-02-08 — division/continuation filings (US 12/235,086 → 7,884,446; US 13/021,409 → 8,598,051) stayed within Harvard; no assignment events.
• 2015-10-01 — SiOnyx, LLC and President and Fellows of Harvard College v. Hamamatsu Photonics K.K., Hamamatsu Corp., Ocean Optics, Inc., Civil Action 1:15-cv-13488-FDS (D. Mass.). Asserted patents included Harvard's '446 (the division of this patent) and '467; '629 itself was not asserted, and the '446 patent was stipulated out of the case before trial.
• 2019-05-09 / 2019-09-24 — jury verdict and final judgment (Doc. 834): judgment for SiOnyx on breach-of-contract/unjust-enrichment claims and for SiOnyx and Harvard on willful infringement of the '467 patent; ownership of the nine Hamamatsu patents transferred to SiOnyx as described above.
• 2020-12-07 — CAFC Nos. 19-2359 (and 19-2360/related) affirmed. Case closed 2021-05-18.
• 2025-01-07 — adjusted expiration; patent expired.

Timeline diagram

timeline
    title Ownership of US 7442629
    2004 : CIP parents filed Sep 24
    2005 : App filed Aug 4
         : Inventor assignment recorded to Harvard
    2006 : Exclusive license to SiOnyx
    2008 : Patent issues Oct 28
    2011 : Divisional 7884446 issues Feb 8
    2015 : Harvard and SiOnyx sue Hamamatsu
    2019 : Final judgment in D Mass
    2020 : Federal Circuit affirms
    2025 : Patent expires Jan 7

NPE / troll-pattern signals

Verdict

Insufficient data (only the original assignment is recorded).

The recorded chain for US 7,442,629 consists of exactly one conveyance — the inventor-to-President and Fellows of Harvard College assignment recorded 2005-10-03 (Google Patents legal events) — with no post-issuance assignment, no reel/frame sequence, and no correspondent to evaluate; the patent has been Harvard-owned from 2005 until its 2025-01-07 expiration. On the evidence that is available, zero of the eight NPE signals is affirmatively present: the 2015 assertion in SiOnyx, LLC v. Hamamatsu, 1:15-cv-13488, was brought by Harvard as owner and SiOnyx as exclusive licensee (Feb 10 2006 license) against a genuine commercial competitor in photodetectors, and this patent was never itself asserted. I therefore decline to characterize the chain as NPE — but I also cannot call it "defensive/non-asserting," because the same family was litigated, so the rubric's catch-all is the honest answer.

Two limitations I want on the record: (1) reel/frame and correspondent data are unverified because the Assignment Center could not be opened in this session — if a recorded license to SiOnyx or a later Harvard-side transfer exists, I did not see it, and I have not asserted one; (2) the family priority dating is inconsistent across sources (some Google Patents fields show 2001-05-25, others 2004-09-24), which affects which family members share this patent's priority chain.

Verification: https://assignmentcenter.uspto.gov/ (search "7442629" by patent number, then compare against https://assignment.uspto.gov/patent/index.html). Litigation docket: https://portal.unifiedpatents.com/litigation/Massachusetts%20District%20Court/case/1%3A15-cv-13488 and https://www.courtlistener.com/docket/[4188862](/patent/4188862)/sionyx-llc-v-hamamatsu-photonics-kk/

I'll verify the bibliographic record and dig into the prior-art citations listed on the face of US 7,442,629.

USPTO record confirmation for 7,442,629 (literal ID match)

A search on the exact number 7,442,629 returns exactly one U.S. patent record; I did not fold in the numerically-similar non-patent hits flagged in the earlier section (Brazilian JUceC "7442629," a Lithuanian registry entry, and a Union Bankshares SEC filing). The confirmed record:

This matches the bibliographic data in the prior section, so the cross-reference holds. One consistency check to flag: uspto.report's "References Cited" table and Google Patents' "Citations (22)" table are not identical (e.g., uspto.report also lists US 6,486,046 (Fujimura) and US 7,057,256 (Carey), which do not appear in Google's 22; Google instead lists the Fujimura publication US 2002/0034845 A1). Because the instructions treat the supplied Google Patents full text as authoritative, I analyze the 22 references in that table below, and note the variance.

Method and legal framework used

• Governing law: App. 11/196,929 was filed 2005-08-04 → pre-AIA 35 U.S.C. § 102 governs.
• Two candidate critical dates:
  • If the CIP parents (filed 2004-09-24) support the claims → priority 2004-09-24, § 102(b) bar date 2003-09-24.
  • The liquid-contact limitation is new matter relative to the 2004-09-24 parents, so claim 1's and claim 15's effective filing date could be as late as the 2005-08-04 actual filing → § 102(b) bar date 2004-08-04.
• Anticipation standard: To anticipate under § 102, a single reference must disclose every limitation of a claim, arranged as claimed. Because all dependent claims incorporate their independent claim, a dependent claim cannot be anticipated by a reference that does not also anticipate its parent independent claim.
• Bottom line up front: no reference cited on the face of the '629 patent anticipates claim 1 or claim 15. Every one of the 22 fails on at least one essential element — most commonly the liquid-contact element, or the sub-micron feature height element. Several are, however, strong § 103 (obviousness) or background references, and two are within the same Harvard family.

Master table — the 22 cited references

Dates are priority date → publication date as shown in the patent's own citation table.

Tier-1 analysis — the references that matter most

1. US 5,635,089 A — "Ceramic, metal and composite materials having increased surface area"

• Full citation: U.S. Patent 5,635,089, R.K. Singh, B.M. Moudgil, S. Behl; assignee University of Florida; filed (priority) 1993-07-13; issued 1997-06-03.
• Disclosure (confirmed from full text): Multiple, relatively unfocused laser pulses (≥50, pref. ≥200) at 0.01–15 J/cm², pulse durations 1 ps to 1 ms, UV–IR, form semi-periodic cones, peaks, columns, grooves with magnitudes of ~0.25 to several hundred microns; performed in air, selected gases, or vacuum (metals require vacuum/low pressure).
• § 102 against the '629 claims:
  • Claim 1: FAILS — no disclosure of "placing … in contact with a liquid"; features are micron-scale, not <1 µm; pulse durations are picosecond-to-millisecond, not femtosecond.
  • Claim 15: FAILS — no liquid layer; pulses not 50–500 fs; features not <1 µm.
  • Dependent claims: cannot be reached because the independent claims are not anticipated.
• Why it still matters: It is the single best methodological precursor for "many low-fluence laser pulses → self-organized surface cones/columns," and it explicitly ties feature scale to treatment environment — the exact variable the '629 uses to shrink features into the sub-micron regime. That makes it prime § 103 material (environment-dependent length scale + known fs-laser texturing).

2. US 7,112,545 B1 — "Passivation of material using ultra-fast pulsed laser"

• Full citation: U.S. Patent 7,112,545 B1, T.A. Railkar, A.P. Malshe, W.D. Brown; assignee Board of Trustees of the University of Arkansas; priority 1999-09-10 (PCT WO 2001/018852 A1, pub. 2001-03-15); granted 2006-09-26.
• Disclosure (confirmed from full text): Irradiating a semiconductor surface (worked example: gallium arsenide) with ultra-short laser pulses — explicitly "no greater than about 10⁻¹² s, i.e. 1 ps … More preferably … no greater than about 500 femtoseconds" — to achieve a passivated surface devoid of a superficial oxide layer.
• § 102 against the '629 claims:
  • Claim 15 / claim 16: These come closest, because the reference expressly teaches applying a plurality of radiation pulses of ≤500 fs to a semiconductor surface. But claim 15 also requires (i) disposing a layer of liquid over the surface and (ii) generating features with heights <1 µm. US 7,112,545 discloses neither a liquid layer nor sub-micron topographic texturing (its aim is a stable passive, oxide-free surface) → no anticipation.
  • Claim 1: FAILS — same two missing elements (liquid contact; <1 µm features).
  • Dependent claims 2, 3, 16 (pulse-width ranges): even though the numerical pulse widths overlap, a dependent claim cannot be anticipated if its parent is not; and the reference still lacks the liquid element.
• Why it still matters: It is the strongest § 102(b) reference on the "femtosecond pulses on a semiconductor surface" half of the invention (published via WO 2001/018852 more than a year before any plausible critical date), and useful § 103 art.

3. US 2003/0029495 A1 — "Systems and methods for light absorption and field emission using microstructured silicon"

• Full citation: U.S. Pub. 2003/0029495 A1 (Mazur et al., President & Fellows of Harvard College); priority 2001-05-25; published 2003-02-13; corresponds to App. 10/155,429 (filed 2002-05-24), which the '629 patent itself incorporates by reference.
• Disclosure (as characterized by the '629 specification): quasi-ordered arrays of conical spikes on silicon produced by high-fluence laser pulses, described in the '629 background as the "micrometer-sized structures" the invention improves upon; the worked comparisons in the '629 attribute these to irradiation in a gas such as SF₆.
• § 102 against the '629 claims:
  • Claim 1: FAILS — no liquid contact; spike heights are micrometer-scale, not <1 µm.
  • Claim 15: FAILS — no liquid layer; not the 50–500 fs + <1 µm combination.
  • Possible § 102(e) treatment: as a U.S. publication with a 2002-05-24 filing, it could be § 102(e) art as to subject matter it describes, but the described subject matter (gas-phase, micron-scale spikes) does not meet the liquid and sub-micron limitations.
• Common-ownership / same-family caveat: This application, the parent CIPs (10/950,248 → US 7,354,792; 10/950,230 → US 7,057,256), and the '629 share the Harvard inventive/ownership lineage. Under pre-AIA § 103(c) and § 102(e)/inventive-entity principles, same-inventive-entity or commonly-assigned work generally cannot be used in the § 102/§ 103 combination against the '629 claims. Treat reference #15 (and #19) as family, not third-party prior art.

4. US 2005/0127401 A1 — "Silicon-based visible and near-infrared optoelectric devices"

• Full citation: U.S. Pub. 2005/0127401 A1 (Mazur et al., Harvard); priority 2001-05-25; published 2005-06-16; this is the pre-grant publication of parent application 10/950,230 (now US 7,057,256), a CIP parent of the '629.
• § 102: Not available as prior art — it is the applicant's own parent/priority document, published after the '629's 2004-09-24 priority date, disclosing the same sulfur-doped microstructured-silicon subject matter. Its appearance in the citation table reflects the examiner's consideration of the family, not third-party art.

Tier-2 — remaining references (brief § 102 disposition)

All of these fail cleanly; none discloses (a) a semiconductor surface in contact with a liquid during laser irradiation, or (b) laser-generated features with heights <1 µm:

• US 4,181,538; US 5,714,404; US 6,242,291; US 2002/0034845 A1; US 2005/0032249 A1; US 2006/0102901 A1 — laser annealing / crystallization / planarization of semiconductor films. They concern heat-treatment of a film, not topographical texturing of a substrate surface under liquid. No anticipation of any claim.
• US 4,201,450; US 4,277,793; US 4,703,996; US 4,965,784; US 5,773,820; US 2002/0126333 A1; US 2003/0025156 A1; US 2003/0132449 A1; US 6,876,003 B1 — device/optics references (photodiodes, LEDs, sensors, storage, holograms). No method overlap; no anticipation.
• US 6,272,768 B1 and US 6,457,478 B1 (Danese) — UV treatment of objects. No laser-texturing/liquid disclosure; no anticipation.
• US 7,211,214 B2 (Chou, Princeton) — laser-assisted direct imprint lithography (laser through a mold). A molding/embossing process, not laser ablation/etching under a liquid, and no <1 µm-fabricated spike disclosure as claimed. No anticipation.

Summary of the § 102 picture

Why every reference fails: The novelty of the '629 claims resides in the combination of (i) laser texturing performed while the semiconductor surface is in contact with a liquid, and (ii) the resulting sub-micron (<1 µm) feature scale. The cited art individually supplies at most one half of that combination — laser texturing (Singh), femtosecond pulses on a semiconductor (Railkar), or microstructuring of silicon (the Harvard family, in gas) — but never the liquid-contact/sub-micron combination.

Where the real § 102 exposure lies: not in the examiner-cited patents, but in the "Other References" non-patent literature listed on the '629 face — e.g., Shen et al., "Formation of Regular Arrays of Silicon Microspikes by Femtosecond Laser Irradiation Through a Mask," Appl. Phys. Lett. 82, 1715–1717 (2003); Her et al., "Femtosecond laser-induced formation of spikes on silicon," Appl. Phys. A 70, 383–385 (2000); Wu et al., "Visible luminescence from silicon surfaces microstructured in air," Appl. Phys. Lett. 81, 1999–2001 (2002). These are the closest printed publications, but each describes fs-laser microstructuring in a gas/air (or through a mask) and produces micrometer-scale spikes — so they likewise do not anticipate the liquid-contact, sub-micron claims. They are, however, the references a validity challenger would build a § 103 case on (gas-phase fs texturing + the known environment-dependent feature-scale effect).

Confidence, caveats, and what I could not verify

• High confidence: the identity/bibliographic record of US 7,442,629; the list of 22 cited references; the § 102 dispositions of references #1–5, #8–14, #16–18, #20, #22 (subject matter is clearly outside the claims). US 5,635,089 and US 7,112,545 descriptions were verified against their own full texts.
• Medium confidence: the precise filing dates (vs. priority dates) of a few references — I give priority and publication dates, which are what the patent's own citation table provides; exact § 102(a)/(b)/(e) categorization for each then depends on those filing dates.
• Discrepancy flagged: uspto.report's reference list and Google Patents' 22-item list differ (US 6,486,046 and US 7,057,256 appear only in the former). I followed the authoritative Google Patents full text per instructions. If you need the examiner's definitive list, pull the "References Cited" page of the granted patent PDF and the file-wrapper IDS from USPTO PatentCenter.
• Not determined: whether any uncited third-party reference (e.g., a pre-2004 publication on underwater/liquid-confined laser ablation of silicon) exists that would anticipate — that requires a dedicated prior-art search beyond the patent's own citation table, which the task's scope did not include.

Sources: https://patents.google.com/patent/US7442629/en (full text, citation table, claims); https://uspto.report/patent/grant/7442629 (grant/assignee/app. data, "References Cited"); http://patentimages.storage.googleapis.com/b1/39/58/adb155ec970a80/US5635089.pdf and https://patents.google.com/patent/US5635089 (Singh et al. full text); https://patents.google.com/patent/US7112545 and https://scholarworks.uark.edu/pat/115/ (Railkar et al. full text/record); https://patentimages.storage.googleapis.com/49/32/30/e324d95675ca5e/[US10741399](/patent/US10741399).pdf (continuity chain).

I'll research the prior art landscape for this patent, focusing on the references cited on its face and the analogous art in laser texturing of semiconductors.

Let me pull the specific prior-art references cited on this patent's face and the Adkisson reference litigated in the family.

Let me pin down the Adkisson reference and the other cited prior art.

Obviousness Analysis — U.S. Patent No. 7,442,629 under 35 U.S.C. § 103

Scope note / build-on. This section assumes the claim and family analysis already generated above (21 claims; independent claims 1 and 15; priority 2004‑09‑24 via CIP parents US 10/950,248 and US 10/950,230; expiry listed at 2025‑01‑07). I do not repeat that content. One point from those sections is directly load-bearing here and I flag it up front: the effective priority date for the liquid‑contact claims is contestable. The '629 is a CIP; the two parents (both filed 2004‑09‑24) describe gas‑phase (SF₆) texturing, while the liquid subject matter of claims 1 and 15 appears to be new matter added in the 2005‑08‑04 CIP. If that is correct, the critical date for pre‑AIA § 102(b) purposes slides from ~2003‑09‑24 to 2004‑08‑04, and publications from the 2004 window become § 102(a) rather than § 102(b) art (swearable‑behind by the inventors). I treat this as an open variable rather than resolving it, and I identify which grounds depend on it.

1. Legal framework applied

• Pre‑AIA § 103 governs (effective filing well before 2013‑03‑16). Obviousness is assessed from the perspective of a person of ordinary skill in the art (POSA) at the time of invention, considering the scope and content of the prior art, differences between the prior art and the claims, and secondary considerations. Graham v. John Deere; KSR Int'l v. Teleflex, 550 U.S. 398 (2007).
• The KSR rationales most directly available here: (a) combining prior‑art elements according to known methods to yield predictable results; (b) simple substitution of one known element for another (a liquid medium for a gas medium); (c) use of a known technique to improve similar devices/methods in the same way; and (d) "obvious to try" where there is a finite number of identified, predictable solutions (In re O'Farrell).
• A reference is analogous art if it is from the same field of endeavor (laser processing of semiconductor surfaces) or reasonably pertinent to the problem the inventor faced (controlling laser‑induced surface morphology).

Level of ordinary skill (proposed): a B.S. or M.S. in materials science, applied physics, or EE, plus 2–4 years' hands‑on experience with pulsed‑laser materials processing and semiconductor surface characterization (SEM/XPS). This mirrors the profile of the actual workers in the field (e.g., the Carey/Mazur and Her laboratory publications) and is consistent with the level the patent's own specification assumes.

2. Key claim elements to be met

Dependents add: 100 fs–few ns (2), 100–500 fs (3); polar/non‑polar liquid (4); water (5); silicon oil (6); alcohol (7); silicon wafer (8, 18); λ ≈ 400 nm to <800 nm (9, 19); 10–400 µJ (10, 20); spikes <1 µm tall, 100–500 nm wide (11), 200 nm–1 µm tall (12); 1 kHz–1 Hz (13); 1–2500 pulses (14); spikes 100–500 nm tall (21).

Critically, no claim recites spike separation, the ~20 nm SiO₂ layer, sulfur/electron‑donating inclusions, ripple structures, or the capillary‑wave mechanism. Those are specification disclosures only. That narrows the inventive contribution the claims can rely on to: [liquid medium] + [short/fs pulses on a semiconductor] → [surface features under ~1 µm].

3. The prior art

3.1 Primary reference — Her et al. (1998) [strong § 102(b) art]

T. Her, R. J. Finlay, C. Wu, S. Deliwala, E. Mazur, "Microstructuring of silicon with femtosecond laser pulses," Appl. Phys. Lett. 73(12), 1673–1675 (21 Sept. 1998), DOI 10.1063/1.122241; also available at the Mazur Group publication list.

Her discloses precisely the applicant's laser‑texturing platform: silicon surfaces irradiated with 100‑fs pulses, 10 kJ/m² fluence, 500 pulses develop quasi‑ordered arrays of sharp conical spikes, up to ~40 µm tall tapering to ~1 µm at the tip. It also reports that irradiating in N₂, Ne, or vacuum creates structured surfaces but not sharp spikes — i.e., it discloses (i) fs‑pulse surface structuring of silicon and (ii) the concept that the ambient medium dictates the resulting morphology. This last teaching is the linchpin of the obviousness case: Her itself frames the ambient as a morphology‑control variable.

3.2 Harvard microstructured‑silicon publications/applications [§ 102(b) art]

• US 2003/0029495 A1 (Ser. No. 10/155,429, "Systems and methods for light absorption and field emission using microstructured silicon"), published 2003‑02‑13 — more than one year before any plausible critical date, so § 102(b). It discloses femtosecond‑laser microstructuring of silicon to form micron‑scale spikes, and the optics/fluence/rep‑rate parameter space the '629 claims recite.
• The corresponding grated patents US 7,057,256 and US 7,354,792 (Carey et al.) are in the same family and were themselves cited against later SiOnyx/Hamamatsu filings (see the litigation record citing the '256 as prior art).
• The cadre of Carey/Mazur publications cited in the family (e.g., "Femtosecond‑Laser‑Assisted Microstructuring of Silicon Surfaces," Optics & Photonics News 14, 32 (2003); "High Sensitivity Silicon‑Based VIS/NIR Photodetectors," OSA (2003); Crouch et al., Appl. Phys. Lett. 84, 1850 (2004)) collectively establish that fs‑pulse silicon texturing to form spikes and its device consequences were well‑known before the critical date.

3.3 The liquid‑confined laser‑processing art [the "missing element" reference]

This is the art the '629 relies on for novelty, and it is a mature field:

• JP 3119090 B2 (Hitachi, "Underwater laser processing device and underwater construction method using the device") — listed in the '629's own Family Cites Families table (see the "U.S. Patent Documents/Foreign" entries in the family citation data). It teaches laser processing of a workpiece underwater, i.e., the liquid‑confined geometry claimed.
• Fabbro et al. (laser‑induced plasma in water confinement) — the foundational disclosure that liquid confinement of a laser plasma raises pressure and alters the ablation/thermal environment. This work is summarized in the surveyed literature ("the laser‑induced plasma induced a shock wave in the plasma plume due to the confinement of liquid… This pressure increase is called the plasma‑induced pressure…", from the Missouri S&T thesis review of the field).
• Daminelli et al. (2004), on femtosecond laser interaction with silicon under water confinement, reported in the same literature review: "Ripple structures formed at the edges of the modified area showed a spacing of 100 nm for water, but 700 nm for air experiments" — i.e., liquid contact reduced the characteristic feature scale by roughly a factor of 7. This is the single most probative piece of art for the "sub‑micron" limitation. (Its 2004 publication date puts it in § 102(a) territory if the 2005‑08‑04 CIP date controls — hence the priority caveat above. Fabbro and JP 3119090, by contrast, are comfortably § 102(b) art.)
• The same review confirms the recognized utility of water in laser machining: "In most cases, water convection and bubble motion contributed to the removal of debris, leading to cleaner and more precise laser machining. The high heat capacity of water provides a better heat sink, thereby effectively cooling heat sensitive substrates…" — this is the articulated motivation, in the prior art itself, to combine liquid with laser processing.
• US 5,635,089 (Univ. of Florida, "Ceramic, metal and composite materials having increased surface area") — cited on the '629 face; teaches that laser texturing increases surface area, supplying a further reason to pursue finer (higher‑surface‑area) texture.

3.4 Semiconductor texturing / passivation art [secondary references]

• Adkisson et al., U.S. Pat. No. 7,141,836 — the Adkisson reference actually litigated in the family (D. Mass. 1:15‑cv‑13488) as an invalidity reference directed to a textured semiconductor surface. In the family's estoppel record the court noted Adkisson was filed 2008‑05‑14 and published 2009‑11‑19, i.e., too late to be § 102 art against the '629, but the court's estoppel discussion confirms Adkisson's subject‑matter relevance (sub‑micron/textured semiconductor surfaces) — so it is useful only as context, not as an available '629 ground. A separate "Adkisson et al." patent, US 7,772,028, also appears in the family's forward citations. I flag that I did not independently retrieve the full text of either; treat their exact disclosures as verify‑before‑relying.
• US 7,112,545 (Board of Trustees, Univ. of Arkansas, "Passivation of material using ultra‑fast pulsed laser") — cited on the '629 face; supports the oxide/passivation aspects of the specification (not claimed).
• US 4,181,538 (DOE, "Method for making defect‑free zone by laser‑annealing of doped silicon"), US 6,242,291 (SEL, laser annealing), US 6,878,003 (Sumitomo) — cited on the face; establish laser processing of silicon surface layers as routine.
• US 4,968,784 (Land et al., Sandia, "Method and apparatus for bistable optical information storage") and US 4,201,450 (Polaroid) — cited on the face; exemplify laser‑induced surface‑relief formation. (My attempt to retrieve the '784 full text was cut off by the tool‑step limit; I rely on its title/classification and family‑citation placement, not on a verified quote.)

4. Grounds of rejection

Ground 1 — Claims 1–14 obvious over Her (1998) in view of the liquid‑confined laser‑processing art (JP 3119090; Fabbro et al.; Daminelli et al.), and in view of US 2003/0029495

Her supplies every element except the liquid. It discloses fs‑pulse laser irradiation of a silicon surface producing quasi‑ordered spikes, with the ambient as a controlled morphology variable, and the pulse‑width/fluence parameter space the dependent claims occupy (100 fs; 10 kJ/m² — squarely within claim 9/10's λ/energy envelope once the 800 nm fundamental is used, and claim 14's 1–2500 pulse range, since Her uses 500 pulses).

The liquid art supplies the liquid and the reason to use it. JP 3119090 discloses underwater laser processing; Fabbro discloses the physics of liquid‑confined plasma; the literature universally states the operational benefits of water as a laser‑machining medium (debris removal, cooling, confinement). Daminelli supplies the predictable consequence of that substitution: water confinement reduces the characteristic surface‑structure spacing from ~700 nm to ~100 nm.

Motivation to combine — four independent, KSR-recognized rationales:

1. Simple substitution of a known element (liquid for gas). Both Her and the liquid‑confinement art are directed to the same problem — controlling the morphology and quality of a laser‑irradiated surface. A POSA seeking to avoid SF₆/Cl₂ (toxic, corrosive, and a process and cost burden) would recognize water as a substitute ambient, particularly where the liquid art teaches that confinement produces "cleaner and more precise laser machining."
2. Known technique improving a known method in the same way. The liquid art's demonstrated effect (finer feature scale) is the very improvement the '629 claims ("features having heights less than about 1 micrometer"). The prior art quantifies exactly that directional result.
3. Predictable result. The mechanism the '629's own specification invokes — faster quenching of molten silicon by a high‑heat‑capacity liquid, shortening the molten lifetime and thus the allowed capillary wavelength — is a known consequence of liquid contact as taught by the liquid‑confinement art. The applicant's own specification frames the sub‑micron result as an inherent outcome of the substituted medium ("utilizing irradiation wavelengths that are less than about 800 nm … and/or placing the irradiated portion in contact with the liquid … can lead to formation of sub‑micron‑sized features"). Where a claim recites a result that follows as a natural consequence of the substitution, § 103 is met.
4. Finite, identified, predictable candidate media. The specification itself lists the eligible liquids as "water, alcohol and silicon oil" — a closed, predictable set — which is the classic O'Farrell "finite number of identified, predictable solutions" fact pattern.

Claim 1 in particular is broad. Claim 1 recites only "short laser pulses" — not femtoseconds. Read against the specification's own definition (100 fs to a few ns), claim 1 is met by conventional nanosecond laser texturing carried out with the workpiece submerged, which the liquid‑confinement art directly discloses. Claim 1 thus requires no femtosecond teaching at all.

Ground 2 — Claims 15–21 obvious over the same combination

Claim 15 adds an express 50–500 fs pulse‑width limitation and a "layer of a liquid" step. Her expressly discloses 100 fs pulses (within 50–500 fs, and within claim 16's narrower 100–500 fs). The recited "layer of a liquid over the substrate surface" is the ordinary bench configuration of the underwater‑processing art (JP 3119090; and the liquid‑layer geometries described generically in the water‑confinement literature). Claims 19 and 20 (λ ≈ 400–<800 nm; 10–400 µJ) are met by routine selection of the Ti:Sapphire fundamental/second harmonic and spot‑size/energy combinations taught in the Harvard microstructuring publications. Claim 21 (spikes 100–500 nm) is the claimed sub‑micron result.

Ground 3 — Claims 11–12, 21 (spike dimensions) obvious as result‑effective amounts

Dimensions ("heights < 1 µm," "widths 100–500 nm") are argued to be result‑effective amounts flowing from the process parameters. Her establishes the spike‑morphology end‑product; the liquid art establishes the size reduction. Where the record shows the claimed range is a product of the disclosed process rather than a separately‑inventive choice, it is not patentable weight. The applicant's own specification supports this by presenting spike height/width as a function of the medium and wavelength, not as an independently designed parameter.

Ground 4 — Claims 4–7 obvious (liquid selection)

Water (5), alcohol (7), and silicon oil (6) are a closed list of liquids the specification itself calls interchangeable, and polar/non‑polar liquids (4) is a genus fully subsumed by the species. In re Petering / In re Roslak‑type reasoning: where a small genus is disclosed and species are conventional, the species claims are obvious.

Ground 5 — Claims 13–14 obvious (process parameters)

1 kHz–1 Hz (13) and 1–2500 pulses (14) are the routine rep‑rate/pulse‑count envelope of fs‑laser texturing; Her's 500 pulses and the 1 kHz amplified‑Ti:Sapphire architecture (already recited in the '629's own apparatus description, and conventional in the art) place these squarely within routine optimization.

5. Anticipation check (why § 102 is not the cleaner theory)

No single reference appears to disclose the complete combination of (a) liquid medium in contact with the silicon surface + (b) fs/short pulses + (c) sub‑micron features:

• Her (1998): gas/vacuum only; spikes are ~40 µm — outside the claimed height range.
• JP 3119090 / Fabbro: don't disclose fs pulse widths or the specific textured silicon morphology.
• Daminelli: discloses the effect (finer ripples in water) but in the context of ablation/threshold studies rather than the claimed silicon spike morphology; and its 2004 date is fragile against a 2004‑09‑24 priority.

Accordingly, the strong theory is § 103, not § 102 — consistent with the family's actual litigation history, in which the PTAB and the district court both worked in obviousness combinations (e.g., Akahori/Mabuchi in view of Mazur) rather than single‑reference anticipation.

A § 102 edge worth preserving: if the liquid claims are held to have only the 2005‑08‑04 CIP date, the inventors' own Shen et al., "Femtosecond laser‑induced formation of submicrometer spikes on silicon in water," Appl. Phys. Lett. 85 (2004) post‑dates the 2004‑09‑24 priority but would pre‑date a 2005‑08‑04 date. That publication describes the identical water‑submersion/400‑nm/100‑fs/60‑µJ experiment and the identical ~500 nm × ~200 nm spike result recited in the specification — a potential § 102(a) self‑anticipation if the inventors cannot antedate it. This is a flanking theory, not a substitute for Ground 1, and it turns entirely on the priority determination I flagged at the top.

6. Anticipated patent‑owner rebuttals and responses

7. Conclusion

The '629's independent claims cover a process substitution (liquid for gas ambient) applied to an already‑known process (fs‑pulse silicon microstructuring) yielding a predicted‑direction result (finer surface features) that the prior liquid‑confinement art quantifies. Every KSR rational (substitution of known elements, improvement of a known technique, predictable results, "obvious to try" over a finite list of liquids) is available on this record, and the applicant's own specification supplies both the closed list of liquids and the mechanism by which liquid contact produces the claimed sub‑micron scale.

• Claims 1–14: strong § 103 rejection over Her (1998) + the liquid‑confined laser‑processing art (Fabbro; JP 3119090; Daminelli). Claim 1 is additionally vulnerable because it does not require femtosecond pulses at all.
• Claims 15–21: strong § 103 rejection over the same combination plus the express fs pulse‑width teaching of Her (100 fs).
• No single‑reference § 102 rejection is available on the references identified; a § 102(a) self‑anticipation theory exists only if the liquid claims are limited to the 2005‑08‑04 CIP date and the inventors cannot antedate their own 2004 Shen et al. publication.

Caveats and verification items. (i) I retrieved the Her abstract and bibliographic data directly, and quotes from the relevant literature review corpus, but I did not pull the full text of US 2003/0029495, JP 3119090, US 5,635,089, US 7,112,545, US 4,968,784, or US 7,141,836; those should be read in full before the grounds are finalized. (ii) The Daminelli 2004 date and its exact disclosure should be verified against the published Thin Solid Films article, because the ground's strength depends on whether it is § 102(a) or § 102(b) art. (iii) The Adkisson references (US 7,141,836 and US 7,772,028) are too late to be § 102/§ 103 art against the '629 and are cited only as litigation context. (iv) I assume a priority date of 2004‑09‑24 for the liquid claims; if the family's 2001‑05‑25 date is held to apply, the reference set narrows as noted but Ground 1 is unaffected. (v) Consistent with the earlier sections of this analysis, none of this has commercial effect today — the patent is listed as expired 2025‑01‑07 — so this § 103 analysis is relevant to validity opinions, family‑wide estoppel strategy, and continuity questions, not to a live infringement defense.

Sources: US7442629B2 full text (claims, face citations, family citation tables, litigation links); Her et al., Appl. Phys. Lett. 73, 1673 (1998) and AIP landing page; US20080116493A1 (Adkisson); Samsung/IPR exhibit list identifying U.S. Pat. No. 7,141,836 to Adkisson; SiOnyx v. Hamamatsu summary-judgment/estoppel memorandum (D. Mass. 1:15-cv-13488) and CourtListener copy; Missouri S&T thesis reviewing water‑confined femtosecond ablation (Daminelli/Fabbro discussion); US9649722B2 (water‑confined laser micromachining background); US8212327B2 (family reference list including US 7,442,629 and Adkisson US 7,772,028).