US 11309444

Patent Overview: US11309444B1

Title: Microstructure enhanced absorption photosensitive devices

Patent number: US11309444B1 (application US17/532,831; also published as US20220102563A1)

Inventors: Shih-Yuan Wang and Shih-Ping Wang

Assignee: W&W Sens Devices Inc (original and current assignee per the listing)

Key dates:

• Priority date: 2015-11-20 (claimed from US14/947,718)
• Filing date: 2021-11-22
• Publication date: 2022-04-19
• Grant date: 2022-04-19
• Anticipated expiration: 2035-11-20
• Legal status: Active

Purpose and plain-language scope:
The patent relates to photosensitive devices designed to absorb light more effectively through microstructure features. The recurring prior-art keywords — "pillar," "region," "pillars," "doped," "absorption" — indicate the disclosed structures involve pillars and doped regions arranged to enhance absorption.

Classifications place the subject matter in optics and semiconductor light-detecting devices: photonic crystals and metamaterials (G02B1/002, G02B1/005), diffraction gratings with pitch less than or comparable to the wavelength (G02B5/1809), optical waveguides and integrated optical circuits (G02B6/122, G02B6/136), and a broad set of photodetector classes including PN heterojunction photovoltaic cells (H10F10/16, H10F10/164), avalanche and PIN photodiodes (H01L31/105, H01L31/107), and related semiconductor body shapes and doping configurations. This suggests the devices may serve light sensing and/or photovoltaic applications, potentially across infrared, visible, or shorter wavelengths given the H10F subclass scope.

The patent also lists priority claims to several related applications (US15/797,821; US16/528,958; US17/182,954), indicating a family of continuations or divisionals.

Source: https://patents.google.com/patent/[US11309444](/patent/US11309444)/en

What the excerpt does not establish:

• The specific claims, claim language, or claim count.
• Detailed structure, dimensions, materials, or fabrication steps beyond keyword topics.
• The full specification, drawings, or embodiments.
• Any independent verification of inventorship, priority chain, or assignee accuracy — Google Patents explicitly disclaims legal conclusions for these.
• Improvements over any particular prior-art reference beyond keyword labels.

(No litigation, validity, ownership-transfer, or prior-art analysis is included, per scope.)

I'll search for litigation involving US patent 11309444.

Initial searches didn't surface litigation for this specific patent. Let me run more targeted searches.

Found a significant case. Let me verify whether US 11,309,444 specifically was asserted and check for other cases.

Bottom line

I found no litigation asserting U.S. Patent No. 11,309,444 ("the '444 patent") specifically. Searches for the number 11,309,444 returned no complaint, docket, CAFC appeal, ITC action, or PTAB proceeding that is about that patent. As far as I can determine, the '444 patent has never been asserted in a lawsuit or challenged in an AIA trial.

I want to be transparent about method and limits: I searched generally (not by directly browsing the Unified Patents caselist or PACER), so I cannot rule out a very recent or obscure filing. However, the '444 patent's owner is highly litigious and its enforcement campaign is well documented, and in none of the materials surfaced does the '444 patent appear as an asserted patent.

Why it's worth explaining: the '444 patent sits inside a heavily litigated family

US 11,309,444 B1 ("Microstructure enhanced absorption photosensitive devices," W&W Sens Devices, Inc., inventors Shih-Yuan Wang and Shih-Ping Wang, granted 2022-04-19, filed 2021-11-22 as Ser. No. 17/532,831) is part of the same continuation family as the patents that are being litigated. Its priority chain runs back to US 14/947,718 (US 10,622,498), and it in turn is a parent of US 17/707,429 → US 11,621,360 B2. So although the '444 patent itself is not on any public asserted-patent list I found, sibling members of the family are.

The known litigation (does NOT include the '444 patent)

Case: W&Wsens Devices Inc. v. Samsung Electronics Co., Ltd.; [[Samsung Electronics America](/litigations/by-defendant/Samsung%20Electronics%20America), Inc.](/litigations/by-plaintiff/Samsung%20Electronics%20America%2C%20Inc.); Samsung Semiconductor, Inc.; and Samsung Austin Semiconductor LLC

• Court / jurisdiction: U.S. District Court for the Eastern District of Texas, Marshall Division (Judge Rodney Gilstrap)
• Case number: 2:24-cv-00854-JRG
• Filing date: 2024 (docket number reflects a 2024 filing; my sources do not give the exact day)
• Asserted patents: originally U.S. Pat. Nos. 9,525,084; 10,446,700; 10,468,543; 11,621,360; 12,087,871; and later 12,243,948. 11,309,444 is not among them. (W&Wsens's Rule 3-1/3-2 Disclosure of Asserted Claims, Ex. 1074 in IPR2025-00994, lists the Asserted Patents explicitly and the '444 patent is absent.)
• Accused products: Samsung ISOCELL CMOS image sensors (e.g., GN3, HP2; Dual Pixel / Tetracell / Nonacell architectures) and ISOCELL Vizion time-of-flight sensors (33D, 63D, 931), plus Galaxy smartphones incorporating them.
• Status / outcome (as of now): Pending. Key developments:
  • Samsung's § 1404(a) motion to transfer to the N.D. Cal. was denied (W&Wsens's venue choice of E.D. Tex. was upheld).
  • The court issued a claim construction order (Dkt. No. 261) construing "hole" as "a void whose primary purpose is to increase the absorption of light entering the void by the light absorbing material abutting the void" — a construction adverse to W&Wsens's broader "mere absence of material" reading.
  • By a joint partial-dismissal order (Dkt. 82-1, filed 10/07/2025), claims on the '084 patent were dismissed with prejudice as to Samsung's existing products; Samsung Semiconductor, Inc. and Samsung Austin Semiconductor LLC were dismissed without prejudice; and the case continues only against Samsung Electronics Co., Ltd. and Samsung Electronics America, Inc.
  • Trial had been set for roughly August 2026 per the docket control order exhibits (Pet. Exs. 1079–1083).

Related PTAB proceedings (same owner, sibling patents — again, none on the '444 patent)

All are Samsung Electronics Co., Ltd. v. W&Wsens Devices Inc., with the Board applying Fintiv-style discretionary denial in view of the co-pending E.D. Tex. trial.

Other records I noticed (not litigation)

• IP LITFIN US 2024 LLC appears in the assignment chain (reassignment recorded 2024-10-23) under a "PATENT SECURITY AGREEMENT" from W&WSENS Devices, Inc. That is a financing/collateral arrangement, not a suit.
• No CAFC appeal involving the '444 patent was found. The CAFC results surfaced (e.g., the 2024-1684 IdeaHub v. Unified Patents Rule 36 judgment, and Bosch v. Snap-On) are unrelated.

Caveats

• I could not directly query the Unified Patents litigation caselist, PACER, or the ITC EDIS interface in this session; findings rest on indexed web results (docketalarm, CourtListener, Justia, govinfo, PTAB PTACTS documents).
• Because the '444 patent is a live family member in an active campaign, it could be added to a complaint by amendment at any time. If you need a definitive negative, it should be confirmed against Docket Navigator / PACER (E.D. Tex. and W.D. Tex.) and the Unified Patents caselist filtered on 11309444, as well as the USPTO PTAB API for any petition naming that patent.

Correctness note: I have not "corrected" any identifier. The patents discussed — 9,525,084; 10,446,700; 10,468,543; 11,621,360; 12,087,871; 12,243,948; and 11,309,444 — are reproduced exactly as they appear in the source materials.

I'll start by verifying the structured data against live sources.

Let me search for the patent owner and related family members.

Significant PTAB activity is emerging on the family, but I need to confirm whether any proceeding targets US11309444 itself. Let me dig.

Proceedings overview

Total AIA trial proceedings against US11309444: 0. The structured "PTAB proceedings on file" block from the USPTO Open Data Portal returns no IPR, PGR, or CBM for this patent, and two rounds of targeted web searching (patent number, patent owner, petitioner, and family-member queries) surfaced no petition, institution decision, or Final Written Decision naming US11309444. Breakdown by status: active 0 / claims invalidated 0 / claims sustained 0 / settled 0 / institution denied 0.

Bottom line for a defendant: this is not "hardened" — it is unbloodied. US11309444 has never been through an AIA trial, so every claim is untested, no § 315(e)(2)/§ 325(e)(2) estoppel attaches to it, and no PTAB record exists to leverage. That is the opposite of the troll-with-dead-claims scenario: if you are asserted on this patent, the entire IPR toolbox is still open. The counterweight is that the patent's siblings — same specification, same inventors, same owner — are under sustained multi-front attack by Samsung, so there is a large, publicly available body of work (prior art, expert declarations, claim constructions) that you could borrow almost verbatim.

None — no proceeding on file for US11309444

No per-proceeding sections can honestly be written, because no proceeding exists. Rather than fabricate docket entries, the relevant intelligence is the family-level activity below. None of these proceedings challenges US11309444, and none of them has any estoppel or preclusive effect on it.

Verified caveat: the ODP ingest can lag filings by weeks-to-months. A very recently filed petition against '444 could exist that neither ODP nor my searches captured. I found none, and I will not invent one.

---

Family-level AIA activity (context — NOT proceedings on US11309444)

US11309444 sits in the W&Wsens Devices / Wang family (inventors Shih‑Yuan Wang and Shih‑Ping "Bob" Wang) that also includes U.S. Patents 9,525,084; 10,446,700; 10,468,543; 11,621,360; 12,087,871; and 12,243,948. Samsung filed a coordinated set of six AIA petitions against five of those siblings in 2025, in parallel with W&Wsens Devices Inc. v. Samsung Electronics Co., Ltd., No. 2:24‑cv‑00854‑JRG (E.D. Tex.) (Chief Judge Rodney Gilstrap).

Supporting sources: Docket Alarm PTAB search and IPR2025‑00996 docket (https://www.docketalarm.com/cases/PTAB/IPR2025-00996/Samsung_Electronics_Co._Ltd._v._W%26Wsens_Devices_Inc/); IPVerse case page for IPR2025‑00996 (https://ipverse.greyb.com/ptab-web/cases/case-details/IPR2025-00996); PTAB filings mirrored in the E.D. Tex. record; Patent Owner's Authorized Response to Director Review Request, IPR2025‑00995 (https://www.docketalarm.com/cases/PTAB/IPR2025-00995/Samsung_Electronics_Co._Ltd._v._W%26Wsens_Devices_Inc/docs/11-25-2025-Patent_Owner/Other__other-21-Authorized_Response_to_Director_Review_Request.pdf?download=true).

Notes and uncertainties (stated rather than glossed):

• The PGR2025‑00082 panel is reported as including APJs Jon M. Jurgovan, Christopher L. Ogden, Peter D. Shapiro, and Kalyan K. Deshpande. That is four names for a three‑judge panel, so the mirror likely conflated the panel with the Acting Chief APJ; treat the panel composition as unconfirmed. The IPR2025‑00996 docket mirror similarly lists "Kalyan K. Deshpande and Gilstrap" — "Gilstrap" is the district judge, not an APJ, so that field is corrupted.
• The institution reasoning in the ‑993/‑994/‑995/‑996 denials was heavily discretionary (§ 314(a), Fintiv-derived factors plus the March 26, 2025 Director memo on settled expectations), not merits-based. Samsung argued the Office committed a material examiner error — that the '543 claims were substantively identical to claims rejected in the parent '821 application but omitted the amendment that secured allowance. See Samsung's referral request: https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1557882](/patent/1557882)/download-documents. A discretionary denial is non-merits and non-precedential; it says nothing about the validity of the underlying claims.
• The one institution that did happen — PGR2025‑00082 on the '948 patent — is the family's real merits test. A PGR FWD would be due within 12 months of institution (statutory deadline from § 326(a)(11)); with an August 2025 filing, expect institution in early 2026 and an FWD around early-to-mid 2027. I could not confirm the institution date or the grounds from an official PTAB source in this session.
• No Federal Circuit appeals were found in this family. No appeal is possible from a denied institution.
• At the district court, the parties narrowed rather than settled: the '084 patent claims were dismissed with prejudice as to Samsung's existing products and colorable variations (Joint Motion for Partial Dismissal, Dkt. 82‑1, 2025-10-07), and IPR2025‑00997 was to be jointly dismissed. That is a settlement-flavored outcome but is not a PTAB settlement on US11309444.
• A patent security agreement with IP LITFIN US 2024 LLC was recorded 2024-10-23. This is a financing/collateral arrangement, not a defensive aggregator such as Unified Patents. The patent is listed as Active with anticipated expiration 2035-11-20.

---

Strategic summary

Claim status. For US11309444 specifically: no claim is canceled, no claim is sustained, and no claim is adjudicated — every claim is untested. There is no IPR certificate, no FWD, no adverse judgment, no disclaimer. If someone demands a license under '444, the patent as issued is fully intact. (I am deliberately not describing what '444 claims, because the authoritative text supplied here does not include its claim set.) By contrast, in the family, PGR2025‑00082 on the '948 patent has been instituted and is the only live PTAB merits proceeding; the other five petitions are dead — four by discretionary denial and one by dismissal.

Estoppel landscape. No estoppel attaches to US11309444 under either § 315(e)(2) (IPR) or § 325(e)(2) (PGR). Estoppel runs claim-by-claim and proceeding-by-proceeding; since no trial was instituted on '444, there is nothing to estop. Samsung's discretionary denials on the siblings produce no estoppel either. For a defendant now facing an assertion of '444, the § 102/§ 103 prior-art-plus-printed-publication space is completely open. The only practical constraints are § 315(b) (one-year bar running from service of a complaint alleging infringement of that patent — '444 has never been asserted, so no bar has accrued) and § 325(d) (the Board may decline grounds already presented to the Office during prosecution). Note also that § 315(a)(1) can bar a petitioner that first filed a DJ action of invalidity; check whether any co-defendant has already filed a counterclaim specifically as to '444.

Pattern signals. One petitioner — the Samsung entity cluster (Samsung Electronics Co., Ltd.; Samsung Electronics America, Inc.; Samsung Semiconductor, Inc.; Samsung Austin Semiconductor LLC) — has filed six AIA petitions across this family, but zero against US11309444. The patent owner (W&Wsens) is fighting hard on the discretionary-denial front and, as of the November 2025 filings, was winning: three of Samsung's five IPRs were denied institution and the Director upheld the denial on the '996 petition in January 2026. W&Wsens has not pursued any CAFC appeal from a PTAB decision, because it has not needed to — no adverse FWD has issued against it. There is no defensive aggregator (no Unified Patents, no RPX IPR) in the chain. The '444 patent is not among the six patents asserted in the E.D. Tex. action (the amended complaint asserts the '084, '700, '543, '360, '871, and '948 patents), so '444 is currently an unasserted family member — a classic reserve patent. That is a meaningful signal: unasserted family members are often held back precisely to be asserted later, at which point a fresh § 315(b) clock starts.

One important defendant-side asset. In the parallel litigation the court issued a Markman order (Dkt. No. 261) construing "hole" narrowly as "a void whose primary purpose is to increase the absorption of light entering the void by the light absorbing material abutting the void," and rejecting W&Wsens's "mere absence of material" position — expressly finding that "holes" are not "isolation trenches." See https://law.justia.com/cases/federal/district-courts/texas/txedce/2:2024cv00854/[233673/261](https://assignmentcenter.uspto.gov/search/patent/reelFrameDetail?reelFrame=233673-0261)/ and https://www.courtlistener.com/opinion/[10940060](/patent/10940060)/wwsens-devices-inc-v-samsung-electronics-co-ltd-and-samsung/. Because the family shares a common specification, that construction is directly transplantable to any '444 dispute involving the term "hole."

---

Recommended next steps

1. Treat "no PTAB activity on '444" as the operative fact, and verify it once more before you rely on it. Re-run the ODP/PTAB search immediately before filing or before any IPR deadline, since the ingest lags. The official source is PTAB E2E / PatentCenter: https://ptacts.uspto.gov/ptacts/. If a petition has quietly been filed, the § 315(b) analysis and the estoppel analysis change completely.

2. If you are a defendant facing assertion of '444, do not assume you are barred. Because '444 was never asserted in 2:24‑cv‑00854, no § 315(b) one-year clock has started for it, and no estoppel exists. If W&Wsens amends its complaint to add '444, calendar the § 315(b) deadline from the date of service of that complaint.

3. If a proceeding on '444 is filed while you read this, track the statutory milestones. Under § 314(b) the institution decision is due within 6 months of the petition's filing; the trial then runs on a statutory 1-year deadline from institution to FWD under § 316(a)(11), extendable up to 6 months for good cause. Note also the current discretionary-denial environment — three of Samsung's five IPRs in this family were denied on discretionary grounds, and the Acting Director's March 26, 2025 memo (plus the "material Office error" exception Samsung invoked) is doing real work here.

4. Mine the family proceedings rather than reinventing them. Even though the petitions were denied institution, Samsung's petitions and exhibits are public and contain a mapped prior-art record for essentially the same specification, including an extensive prior-art table (see, e.g., Patent Owner's Exhibit 2017 in IPR2025‑00995, https://www.docketalarm.com/cases/PTAB/IPR2025-00995/). A fresh petition against '444 can adopt that art, but must distinguish Samsung's § 112/claim-construction positions in the district court, and must overcome § 325(d) if any of the art was before the examiner.

5. If you need an invalidity theory in district court now, the Markman order is your starting point, not the PTAB. The court's construction of "hole" as a light-absorption void that excludes isolation trenches is the single most useful intrinsic-record holding in this family, and it is adverse to the patent owner on the term that drives most of the asserted claims. Any '444 theory that depends on reading trenches as "holes" should be re-evaluated immediately.

6. Do not treat the discretionary denials as a merits win for W&Wsens. They are not. Samsung's repeated "material examiner error" argument — that the family's claims mirror claims rejected in the parent '821 application minus the amendment that won allowance — remains untested on the merits for '444 and is the strongest structural attack available if the claims of '444 track those parents.

I'll research the assignment chain, the entities involved, and any litigation or NPE signals before writing the analysis.

Strong leads: an active Samsung litigation, PTAB IPRs, and a 2024 security agreement with a "LitFin" entity. Let me dig into each.

Assignment & Ownership Analysis — US 11,309,444 B1

Verification limit (read first). I could not open the USPTO Assignment Center / Patent Assignment Search UI from this session, so reel/frame numbers and the recorded correspondent for either assignment are not captured below. Google Patents' legal-events feed (the authoritative text supplied to me) confirms two post-issuance recordation events but does not expose reel/frame or correspondent. I have not invented those values. The analyst should pull them directly at the Assignment Center search page for patent 11309444 before this memo is relied on for standing or chain-of-title purposes.

---

Inventors

Unusual-pattern check — the classic warning flag is absent, and inverted. The usual tell (all inventors leaving the original assignee within 12 months of filing, presaging a fire-sale) does not apply here: the two inventors are the assignors and the principals of the assignee. The 2021-11-22 assignment is an inventor→own-company transfer recorded on the same day the application was filed (application 17/532,831 filed 2021-11-22), i.e. routine employment/assignment paperwork, not a departure.

One related-family artifact: M. Saif Islam (UC Davis) is a named co-inventor on sibling family members — the '700 and '543 patents — but is not a named inventor on the '444 patent. Do not carry him into this chain.

---

Original assignee

W&WSens Devices, Inc. (also styled "W&W Sens Devices, Inc." on the face of the '444 grant), 4546 El Camino, Suite 215, Los Altos, CA 94022.

• Business: fabless/asset-light photodetector R&D and IP development — silicon and Ge-on-Si photodiodes/APDs/SPADs using light-trapping microstructures, positioned for datacenter interconnect, LiDAR/ToF and 3D imaging.
• Did it ship a product embodying the claims? No evidence of a commercial product. Its own 2019 outreach material (quoted in PTAB Exhibit 2003, IPR2025-00995) markets the company as holding "an extensive global IP portfolio for light trapping technology in thin silicon photosensors," states it is "currently engaging in M&A discussions and… being represented by Bank of America Merrill Lynch," and solicits meetings with an OEM. That is a licensing / sell-the-portfolio posture, not a product-shipping posture. I found no evidence of a W&WSens catalogue part, foundry tape-out, or revenue product.
• Current status: operating as a patent-assertion plaintiff, not dissolved. DoD data through the mid-2020s shows it contracting with Dr. M. Saif Islam at UC Davis for proof-of-concept fabrication runs (Plaintiff's First Supplemental Interrogatory Responses, W&WSens v. Samsung). It is the sole real party-in-interest in its litigation and in its PTAB proceedings ("The real party-in-interest for Patent Owner is W&Wsens Devices Inc.", IPR2025-00995 Mandatory Notices, 2025-06-10).
• Not a distressed seller: no Chapter 7/11, no assignment-for-creditors, no UCC Article 9 foreclosure appears anywhere in the record I could reach.

---

Assignment timeline

Google Patents legal events for US 11,309,444 B1 record two post-issuance entries. Reel/frame not captured (see verification limit) — flagged rather than fabricated.

1. 2021-11-22 (executed) / recorded 2021-11-22 — Reel not captured

• Conveyance: Assignment (Assignment of Assignors' Interest)
• Assignor: WANG, Shih-Ping; WANG, Shih-Yuan
• Assignee: W&W SENS DEVICES, INC. (Los Altos, CA)
• Correspondent: not captured — Assignment Center UI not reachable this session.
• Context: Formation/employment assignment — inventors to their own company, executed on the application filing date (17/532,831 filed 2021-11-22). Routine.

2. 2024-10-23 (executed) / recorded 2024-10-23 — Reel not captured

• Conveyance: Patent Security Agreement (a collateral assignment / grant of security interest — not an outright transfer of title)
• Assignor: W&WSENS DEVICES, INC.
• Assignee: IP LITFIN US 2024 LLC
• Correspondent: not captured — this is the single most valuable missing field; the recording attorney on this entry is the key repeat-player check.
• Context: Litigation financing / securitization. The assignee name ("IP LitFin" = IP Litigation Finance) plus the 2024 formation year and the recording date landing in the middle of the company's active E.D. Tex. campaign against Samsung is the profile of a funder taking a security interest in the asserted portfolio. Ownership of record stays with W&WSens absent default.

Two entries only. No cascading LLC chain, no intermediate holding company, no change-of-name, no release recorded on this patent as of the last Google Patents sync (2026-09-24).

---

Timeline diagram

timeline
    title Ownership of US 11309444
    2015 : Priority date 20 Nov 2015
    2021 : Application 17 532 831 filed
         : Inventors assign rights to W&W Sens Devices
    2022 : US 11309444 B1 granted 19 Apr 2022
    2024 : W&W Sens sues Samsung in E D Texas
         : Security Agreement to IP LITFIN US 2024 LLC
    2025 : Samsung files IPRs on family patents

---

NPE / troll-pattern signals

Context on the seller's financial trajectory (supports the financing reading): W&WSens marketed itself for M&A in 2019 (PTAB Ex. 2003, email dated 2019-08-14), retained Bank of America Merrill Lynch, elevated "Lihan Wang, VP Strategy & Business Development," and by 2024 had pivoted to enforcement financed by a 2024-formed funder. That arc — no product, no exit, then assert — is the classic funded-monetization sequence.

Scope note on this patent: US 11,309,444 is not among the patents currently asserted in 2:24-cv-00854. The asserted set is 12,087,871; 11,621,360; 10,446,700; 9,525,084 (and, per the First Amended Complaint, 12,243,948); the '084 claims were dismissed with prejudice as to currently existing products by the joint partial-dismissal order filed 2025-10-07, with IPR2025-00997 to be jointly dismissed. The '444 is a sibling in the same priority family (inter alia through 16/528,958 → US 11,121,271, and 15/797,821 → US 10,446,700) and is being held in reserve.

---

Verdict

NPE — moderate confidence.

The owner of record is the original, inventor-founded R&D company W&WSens Devices, Inc., which shows no evidence of a commercial product embodying the claims, publicly marketed itself for licensing/M&A rather than product sales (PTAB Ex. 2003, 2019-08-14), and is now asserting the family against Samsung in E.D. Tex. 2:24-cv-00854 — a textbook non-practicing profile. The supporting monetization signal is the 2024-10-23 Patent Security Agreement from W&WSENS DEVICES, INC. to IP LITFIN US 2024 LLC, a 2024-formed entity whose name denotes litigation finance, recorded during the active campaign. It is not high confidence, and I decline to call it high, because the strong structural tell is missing: the 2024-10-23 conveyance is a collateral security interest, not a title transfer (title stayed with W&WSens), there is no shell-LLC ownership hop, no known asserter appears, there is no cascading transfer chain, and the patentee is the original assignee suing an actual competitor rather than an aggregator that bought claims. Signals 3 (repeat correspondent) and 5 (pre-litigation timing) are unresolved pending reel/frame and correspondent retrieval — obtaining those could move this to high confidence, since a single recording attorney appearing on both the 2021-11-22 and 2024-10-23 entries would satisfy signal 3 outright.

Verification link: https://assignmentcenter.uspto.gov/ (search patent 11309444; corroborated at https://assignment.uspto.gov/patent/index.html and in the legal-events feed at https://patents.google.com/patent/US11309444/en).

I'll research US patent 11309444 and its prior art citations.

Let me get the specific citation data from the patent's Google Patents page.

Prior-Art Analysis — US 11,309,444 B1

⚠️ Scope and confidence disclosure (please read first)

I must be explicit about a limitation before giving the analysis, per my operating rules:

1. I could NOT retrieve the patent's own face-page "(56) References Cited" table. The authoritative full text you supplied (Google Patents rendering) contains the classification lists, the definitions/summary section, and the drawing descriptions — but it omits the front-page citation tables (U.S. Patent Documents, Foreign Patent Documents, and Other Publications). My search attempts to pull the Google Patents "Citations" panel for US11309444B1 returned only citing documents (forward citations) and family members, not the cited (backward) prior art.
2. Therefore, the references below are drawn from (a) the specification's own citations, (b) the search report of the closest family member (EP 3221895 A4 / WO 2016/081476), which shares the same priority chain, and (c) the related U.S. family. I flag confidence for each. You should verify against USPTO Patent Center "References Cited" for application 17/532,831 before relying on this for any §102 position.
3. Full claim text was not in the supplied document; the "Definitions" section reads as a summary of embodiments that correspond to claim features. Claim mapping below is therefore provisional.

---

1. Patent identification (verified from the authoritative text)

Critical date for §102 purposes: the effective filing date is the 2015-11-20 priority date (assuming the chain via 14/947,718 → 15/797,821 → 16/528,958 → 17/182,954 is intact). Any reference must therefore predate 2015-11-20 to qualify under §102(a)(1)/(a)(2). Note the terminal-disclaimer/continuation relationships flagged on the family members — relevant to §102(b)(2)/(b)(3) exceptions and to obviousness-type double patenting, not §102 art.

---

2. Related U.S. family members — NOT prior art (but critical to the §102 "same inventor / common ownership" analysis)

These appear in the priority chain and thus are not §102 prior art against 11,309,444, but they are the references a third party would try to use under §102(a)(2)/§102(e) if priority is challenged:

Because all of these share inventors and assignee, they fall within §102(b)(2)(C) common-ownership and §102(b)(2)(A) "same inventor" exceptions. They should not defeat 11,309,444 on §102 grounds.

---

3. Non-patent literature cited in / on the face of the specification (highest-confidence §102 candidates)

These are expressly named in the 11,309,444 disclosure (see the "Nanowires" and "nanoholes" sections of the supplied text). All predate 2015-11-20, so they are §102(a)(1) printed publications:

3.1 Garnett et al.

• Full citation: E. Garnett and P. Yang, "Light trapping in silicon nanowire solar cells," Nano Letters, 2010, 10, 1082–1087.
• Publication date: 2010 (published 2010; ~March 2010).
• Description: An ordered array of silicon nanowires is shown to increase the optical path length in silicon by ~73× versus bulk, i.e., an effective absorption coefficient far above bulk Si.
• Potential §102 relevance: The specification itself adopts the Garnett number ("the effective absorption coefficient of the ordered array of Si nanowires is 73× that of bulk silicon; bulk α for Si is 760/cm and nanowire Si α is 5.548×10⁴/cm at 850 nm"). Garnett therefore is a strong §102(a)(1) candidate against any claim reciting a silicon microstructure array that increases effective absorption, and against claims reciting the numerical performance attributed to microstructures. It does not disclose the reverse-biased PIN/PIPIN diode architecture, the cathode/anode "reverse biasing circuitry," or the >5–10 Gb/s bandwidth limitations.

3.2 Kelzenberg et al.

• Full citation: M. D. Kelzenberg et al., "Enhanced absorption and carrier collection in Si wire arrays for photovoltaic applications," Nature Materials, vol. 9, March 2010, pp. 239–244.
• Publication date: March 2010.
• Description: Si microwire/wire arrays for PV — enhanced absorption plus radial carrier collection.
• Potential §102 relevance: §102(a)(1) against claims to a microstructured absorbing semiconductor region for photon conversion and against PV/zero-bias embodiments. Weak-to-moderate; does not show the photodiode/APD reverse-bias limitations.

3.3 Li et al.

• Full citation: X. Li et al., "Optical absorption enhancement in silicon nanowire and nanohole arrays for photovoltaic applications," Proc. SPIE Vol. 7772, 77721G-1 (2010) (Next Generation (Nano) Photonic and Cell Technologies for Solar Energy Conversion).
• Publication date: 2010.
• Description: Absorption enhancement in both Si nanowire and nanohole arrays.
• Potential §102 relevance: Directly relevant to claims reciting "holes" as the microstructure (not just pillars) — the specification cites Li for the proposition that "Microstructure hole arrays have similar properties as microstructured pillars." Strong §102(a)(1) candidate for the "pillars and/or holes" claim concept.

3.4 Chang-Hasnain (HCG)

• The specification repeatedly cites "Chang-Hasnain" for high-contrast-grating (HCG) theory — the near-wavelength regime, low-Q absorption modes, and Q values up to 10⁷ with minimal absorption loss.
• Potential §102 relevance: Relevant to the "absorbing mode high contrast grating" claim concept. However, HCG art teaches high-Q, non-absorbing gratings; it would more naturally support a §103 obviousness theory (modifying an HCG to a lossy/absorbing mode) than a §102 anticipation. Caveat: I do not have the exact Chang-Hasnain citation string from this patent's front page and should not fabricate the year/venue — verify.

---

4. References cited in the sibling family search report (EP 3221895 A4 / WO 2016/081476)

The PubChem/Espacenet record for EP-3221895-A4 — same inventors, same priority line (priority 2014-11-18, filing 2015-11-17) — lists the following cited documents. These are the references a parallel examiner considered against the same disclosure and are the most likely entries in the US 11,309,444 "(56)" table as well:

⚠️ I am deliberately not providing descriptions of the technical content of US 4,814,847, US 7,977,637, US 8,357,960, US 2007/0215968 or US 2010/0059822 because I could not verify them from a primary source in this session, and my operating rules prohibit fabricating references. The "(Y)" designations on US 2010/0059822 and US 8,357,960 indicate the examiner used them in an obviousness combination, not as standalone anticipatory art — so they are more likely §103 material than §102 material.

---

5. Provisional claim mapping under 35 U.S.C. § 102

Because the granted claim set was not available, I map to the claim concepts evident from the specification/definitions section:

---

6. What I recommend you do to close the gaps

1. Pull the official "(56) References Cited" list for application 17/532,831 from USPTO Patent Center or Patent Public Search — this is the definitive citation set and the one the examiner actually applied.
2. Pull the file wrapper / office actions for 17/532,831 (and its parent 14/947,718) to identify which references were applied as §102 vs §103 rejections and what was overcome.
3. Note the third-party activity I observed in the search results: a PTAB proceeding referencing claim language about a "photosensitive imager device … textured region …" (docket string IPR2025-00811) surfaced alongside this query. That claim language resembles the SiOnyx-type textured-photodetector patents, which would be strong §102/§103 art (laser-textured absorbing regions, micron/nano surface features, black silicon). I could not confirm the parties or patent number, so treat this lead as unverified — but the SiOnyx "textured region" family is worth a dedicated search as the closest commercial prior art to the "microstructure enhanced absorption" concept.

---

Bottom line: On the evidence I could verify, the most relevant §102(a)(1) prior art is the Garnett 2010 Nano Letters paper (Si nanowire absorption enhancement, ~73×) and Li 2010 SPIE (Si nanohole arrays), both of which are expressly cited in — and used as quantitative support by — the 11,309,444 specification itself. Neither is believed to disclose the reverse-biased cathode/anode photodiode architecture, the APD multiplication-layer arrangements, or the >5–10 Gb/s at 850 nm QE ≥60% performance limits, which remain the strongest novelty positions. The patent's own face-page citation list and Office Action history should be retrieved from USPTO Patent Center to complete this analysis rigorously.

I'll build this analysis carefully. First, let me try to retrieve the actual granted claims and the cited prior-art references, since the earlier section flagged that claim text could not be verified.

Let me pull the actual claim set and the cited references for this patent.

I have claim 1 of the '444 patent. Let me pull the full claim set and the "References Cited" (Prior Art) section.

Obviousness Analysis — U.S. Patent No. 11,309,444 (35 U.S.C. § 103)

Analyst note on the date: The task header states the current date is April 26, 2026, while the session metadata says September 29, 2026. The earlier-generated section of this analysis also cites events dated September 26, 2026. Those cannot all be correct. I flag the inconsistency rather than resolve it, and I do not rely on any post-April-2026 development below.

---

0. Evidence base and an important caveat about the "Prior Art section"

Caveat, stated up front. I could not retrieve the "Prior Art" / "References Cited" tab of the US 11,309,444 Google Patents page (or the PTO-1449 / examiner citation list from its file wrapper) within this session. The full text you supplied to me is also truncated — it ends mid-specification at "Ge thin film can be grown epitaxially on Si," and it contains no "References Cited" list, no claims column, and no drawings text.

I therefore do not have the authoritative prior-art list for this patent. Rather than fabricate citations, this analysis uses only art I could actually verify from the record I retrieved, and I label the confidence of each reference explicitly. Any combination below must be re-run against the actual PTO-1449 before being relied on.

What I did verify, and it matters:

1. The actual granted claim 1 of US 11,309,444 (retrieved from the Justia claim transcription). This is a significant correction to the earlier-generated section of this analysis, which reconstructed the claims from summary-of-invention language and described them as covering a generic "cathode region / anode region / reverse-bias circuitry / microstructure-enhanced absorbing region." The real claim 1 is far narrower and structurally specific. See §3 below.
2. Corroboration: the Google Patents "prior art keywords" field for US 11,309,444 reads "pillar, region, pillars, doped, absorption" — consistent with the pillar-and-doped-region claim I retrieved, and inconsistent with the broader cathode/anode framing in the earlier section.
3. References discussed in the specification itself (applicant admissions).
4. Two citation lists from family members.

---

1. Framework

Under Graham v. John Deere and KSR Int'l v. Teleflex, I address: (a) scope and content of the prior art; (b) differences between the prior art and the claims; (c) level of ordinary skill; and (d) secondary considerations. Under § 103 the question is whether the claimed subject matter as a whole would have been obvious — not whether each element was separately known. KSR controls the motivation analysis: a combination is obvious where the improvement is the "predictable use of prior art elements according to their established functions," and where "a technique has been used to improve one device, and a person of ordinary skill in the art would recognize that it would improve similar devices in the same way."

---

2. Level of ordinary skill in the art (PHOSITA)

A person of ordinary skill in the art as of the relevant date would have a Ph.D. or M.S. in electrical engineering, applied physics, or materials science, plus 2–5 years of experience in semiconductor optoelectronics — specifically Si photonics, photodetector/APD design, and CMOS-compatible process integration — or a B.S. with 5–10 years of such experience. This person would be familiar with: (i) light-trapping by sub-wavelength and near-wavelength Si pillar/hole arrays; (ii) high-contrast grating (HCG) physics; (iii) SOI/BOX processing; (iv) anisotropic wet etching of Si (KOH/TMAH) producing pyramidal facets; (v) dielectric planarization/fill; and (vi) the RCA/CMOS requirement that a photonic device be fabricated in a commercial CMOS foundry. The patent itself is directed to exactly the intersection of these skills and could be practiced by such a person using a textbook-level combination of known process steps.

---

3. The claimed subject matter

3.1 Claim 1 (verified, from the Justia transcription)

Claim 1 is directed to "A device converting light photons to electrons" comprising:

3.2 Dependent claims (partially retrieved)

• 2 — pillar-separating regions extend only partway into the Si absorption region.
• 3 — in-pillar holes have bottom ends with non-zero lateral cross-sectional area.
• 4 — further including an antireflection coating.
• 5 / 6 — each pillar has a plurality of in-pillar holes / only a single in-pillar hole (mutually exclusive alternatives).
• 7 — the top electrical contact extends over the in-pillar holes.
• 8 — the top electrical contact comprises a transparent, electrically conducting region.
• 9 — pillar-separating regions extend down at least to the lower doped Si region.
• 10 — at least one contact is electrically coupled to a plurality of the pillars.
• 11 — doping: one of the upper/lower regions > 5×10¹⁷/cm³, the other > 5×10¹⁸/cm³, absorption region < 5×10¹⁶/cm³.
• 12 — a second independent claim (light-photon-to-electron device) — text truncated in my source; not fully analyzed.
• 13 — a plurality of the in-pillar holes are pyramid-shaped.
• 14 — a plurality of the in-pillar holes are round in lateral cross-section.
• 15 / 16 / 17 — mirror claims 5 / 7 / 9.
• 18 — text truncated; not analyzed.
• 19 — a third independent claim; at least one in-pillar hole elongated in lateral cross-section.
• 20 — a plurality of in-pillar holes overlap.
• 21 — text truncated; not analyzed as an independent claim; appears to depend from claim 1.

Flagged contradiction with the earlier-generated section. The earlier section's "plain-language overview of the independent claims" (cathode/anode regions, reverse-bias circuitry, generic pillars/holes/voids in Si, Ge, Si/Ge or a III-V family, plus microwave-transmission-line / waveguide / heat-exchanger claims) describes the specification's summary-of-invention and the EP/PCT family, not the granted US claim set. If one tries to invalidate the '444 using that reconstructed framing, the analysis will be wrong. The controlling claim 1 is silicon-only, pillar-only, and requires in-pillar holes. Treat the earlier overview as superseded for § 103 purposes.

---

4. Prior-art inventory (with confidence labels)

4.1 Applicant-admitted prior art (HIGH confidence — quoted in the '444 specification itself)

4.2 References cited on the face of the EP family member (MEDIUM confidence)

The PubChem/Espacenet record for EP 3 221 895 A4 (the EP counterpart in this family, priority 2014-11-18) lists these citations:

• US 4,814,847 A
• US 2007/0215968 A1
• US 2010/0059822 A1 (marked "Y")
• US 7,977,637 B1
• US 8,357,960 B1 (marked "Y")

The "Y" markings mean the EPO considered these relevant in combination for inventive step. I could not verify their disclosures in this session — I am listing them as candidates precisely because an EPO "Y" citation often maps onto a US § 103 combination, and because they may be the same references the US examiner cited. Do not treat the characterizations below as verified.

4.3 Background art (HIGH confidence as general knowledge, not a specific citation)

• Anisotropic wet etching of (100) Si with KOH/TMAH producing pyramidal and inverted-pyramidal facets — the textbook "random pyramid texturing" of solar cells, in use since the 1970s–80s and standard in CMOS-adjacent process flows.
• SOI/BOX substrates providing a buried dielectric layer beneath a device layer, with the BOX acting as an etch stop for backside vias.
• Dielectric gap-fill / planarization (spin-on glass, polyimide, PECVD oxide, nitride) between high-aspect-ratio features.
• Transparent conducting oxides (ITO) as top contacts to pillar arrays.
• Vertical PIN / mesa micropillar photodiodes with reverse bias.

---

5. Element-by-element mapping and combinations

Ground A — Garnett in view of Lin, further in view of KOH-pyramid texturing

Why the combination would have been made. The '444 specification itself supplies the motivation, which is powerful § 103 evidence because it is an applicant admission:

"The microstructures … effectively increase the effective absorption length resulting in a greater absorption of the photons." … "the enhanced absorption allows the APD/PD to: (1) extend the sensitivity … closer to the material band gap …; (2) operate at a high data rate bandwidth (e.g. >10 Gb/s), due to shorter absorption length …; (3) increase quantum efficiency …; (4) be manufactured of silicon …; (5) be compatible with Si Photonics and CMOS processes."

The problem to be solved — "Si is weakly absorbing at 850 nm and longer, so a 30 µm absorption region is needed, which kills bandwidth" — was a known, articulated problem in the field, and Garnett, Lin, and the HCG literature each addressed it in the same way (light trapping by near-wavelength Si microstructure). Combining pillar trapping (Garnett), hole trapping (Lin), and a pyramidal facet geometry (KOH) is the application of known techniques to a known device ready for improvement, yielding a predictable result.

KSR/MPEP § 2143 rationales engaged: (A) known elements combined to yield predictable results; (C) known technique to improve a similar device in the same way; (D) known technique applied to a known device ready for improvement; (F) design incentive/market force (850 nm datacom bandwidth demand, explicitly recited in the spec).

---

Ground B — Garnett in view of the Chang-Hasnain HCG art, further in view of SOI/BOX

This alternative framing avoids needing Lin/KOH for the pyramidal-hole element by attacking the claim through the functional limitation (1i) and the geometry limitations (1h):

• HCG art teaches that a periodic array of high-index structures with a period between λ/n₁ and λ/n₂ produces resonance and in-plane propagation, i.e., a drastically lengthened effective absorption path. The '444 spec calls this "the absorption mode of high contrast grating (HCG)" and expressly states the period can be "approximately 100-2000 nm" — which is numerically identical to claim 1's recited 100–2,000 nm center-to-center spacing. The reference range and the claim range coincide, and the specification offers no evidence that 100–2,000 nm is critical or unexpected.
• Changing pillar to hole (or adding holes to pillars) in an HCG array is a known design variant: the '444 spec itself says structures "such as voids, air gaps, and/or holes" all lower the effective index and support the same resonance physics. Lin confirms holes behave like pillars. Substituting a hole for a pillar is an art-recognized equivalent in this context — KSR rationale (B).
• Pyramidal hole walls from KOH etching provide graded effective index — a classic antireflection/texturing approach; predictable.
• SOI/BOX supplies element 1g and also the low-index cladding that makes the HCG resonance possible — and the spec admits BOX/SOI is used "to lower the effective refractive index."

---

Ground C — the claim's own "1.1×" and "CMOS" limitations are weak

Two claim elements deserve separate treatment because they are the patentee's likely fallback positions and both are vulnerable:

1. "increase photon absorption … by a factor of 1.1 or more." This is a result-effective limitation recited as a 10% improvement. In the '444 spec, the comparable admitted figure for a pillar array alone is 73× (Garnett). A 10% increment attributable to adding a hole is squarely within the range of predictable, routine optimization. Under In re Aller / In re Woodruff reasoning, where a recited numerical result is a result-effective variable and the prior art suggests the variable, optimization is routine and the burden shifts to the patentee to show criticality.
2. "CMOS compatible for mass production using CMOS foundries." This is a statement of intended use / capability, not a structural limitation. In re Schreiber — reciting that a known device is used in its known way for its known purpose does not confer patentability. Moreover, the reference to a foundry is a manufacturing-location/process statement, which the specification confirms is achieved by using Si and standard deposition/etch steps. It adds nothing structural over Garnett/Lin.

---

Ground D — Anticipation (§ 102) consideration

I did not identify a single reference I can verify that discloses all of: a vertical Si PIN pillar array, dielectric-filled pyramidal in-pillar holes, dielectric-filled pillar-separating regions, and a buried dielectric. Lin (holes) lacks pillars; Garnett (pillars + polyimide) lacks in-pillar holes. So on the record I can verify, claim 1 is not anticipated; the attack is a § 103 attack. (I cannot rule out anticipation by a reference I was unable to retrieve — see § 0.)

---

6. Dependent claims — additional art needed

Dependent claims 2–21 add nothing that would independently salvage the claim set if claim 1 falls, because each limitation is either a routine process/geometry choice, an inherent crystallographic result, or a result-effective numerical optimization.

---

7. The priority-date issue — why the prior-art universe may be much larger

This is the single most important strategic point and it connects directly to the earlier-generated section's finding about the W&Wsens v. Samsung priority dispute.

• US 11,309,444 issued from application 17/532,831, filed 2021-11-22, and claims priority back to 2015-11-20 (via 17/182,954 / 16/528,958 / 15/797,821 / 14/947,718, with CIP links to PCT/US16/67977, PCT/US18/57963, PCT/US18/43289, and Ser. Nos. 16/296,985 / 16/042,535).
• Under § 120, the '444 is entitled to the 2015 date for a given claim only if that claim's subject matter is supported under § 112 by the earlier application(s). Claim 1's "dielectric-filled, pyramid-shaped, in-pillar holes," "rectangular laterally extending cross-section," and the specific 100–2,000 nm / 100–5,000 nm ranges must each find written-description and enablement support in 14/947,718 (filed 2015-11-20) and its ancestors.
• My limitation: the '444 full text provided to me is truncated and I could not confirm that "pyramid-shaped, in-pillar holes" appears in the 2015 material at all. I flag this as unresolved. If that language was added in a later CIP, the effective filing date for claim 1 could be later than 2015-11-20 — potentially as late as 2018, 2019, or 2021.
• The earlier-generated section reports that in W&Wsens Devices Inc. v. Samsung Electronics Co., Ltd., No. 2:24-cv-00854-JRG (E.D. Tex.), Samsung is contesting priority, arguing a November 22, 2021 date for related family members (the '948, '360 and '871 patents), relying in part on W&Wsens' September 26, 2026 filing in App. No. 19/036,506 concerning the "holes extend 100 to 10000 nm in the light direction" limitation.

Why this matters for § 103. Every reference published between the contested priority date and the 2021-11-22 filing becomes § 102(a)(1)/(a)(2) prior art if priority is lost. On the record I retrieved, that would sweep in Zang et al. (2017) — expressly cited on the face of US 10,446,700 — and a very large body of 2016–2021 Si-photonics/light-trapping literature and patents. A priority loss would convert this from a hard obviousness case into an easy one. Conversely, if priority holds at 2015-11-20, the petitioner is confined to pre-2015 art.

I cannot resolve this without the file wrappers, and I will not guess.

---

8. Secondary considerations (§ 103 rebuttal — patentee's side)

The patentee's best arguments, and my assessment:

1. Unexpected results. The '444 spec asserts >10 Gb/s at 850 nm with ~90% QE in Si, and ~54 Gb/s / 98% QE in FIG. 4A–4B. Weakness: the specification attributes the effect to known mechanisms — HCG resonance, scattering, near-field, interference — and expressly relies on Garnett's 73× measured enhancement. A result that the prior art already predicted and measured is difficult to call unexpected. Worse, claim 1 recites only 1.1×, so the patentee cannot tie the impressive 54 Gb/s/98% data to a limitation that distinguishes the claim from the art.
2. Long-felt need / failure of others. There is a genuine story here — Si's weak absorption at 850 nm forced 30 µm absorption regions and capped bandwidth at ~1.5 GHz. That is a legitimate secondary consideration. But the problem was known and being actively solved in the 2010–2015 window (Garnett 2010, Lin 2010, HCG work), which undercuts "failure of others."
3. Industry praise / licensing. Not evidenced on the record I have.
4. Copying / commercial success. The W&Wsens v. Samsung case (accused Samsung ISOCELL products) might support a copying/nexus theory, but the asserted patents there are different family members (the '543, '084, '948, '360, '871), and US 11,309,444 is not asserted. Commercial success of a product not shown to be coextensive with this claim carries little weight.

---

9. Weaknesses in the attack (what could defeat it)

Being candid about the other side:

1. The "compared to a like device lacking said one or more in-pillar holes" limitation. This is a comparative/negative limitation tied to a structural difference. A petitioner must either (a) show the prior art structure differs from the claim only by the absence of in-pillar holes, or (b) establish that the recited 1.1× enhancement is inherent. Ground A above is designed to do (a) using Garnett's pillars-plus-polyimide as the "like device" baseline, but this requires careful claim-charting and likely a declaration with optical modeling.
2. "Pyramid-shaped" is a specific shape. Absent a reference that teaches or suggests pyramidal holes in a pillar, the patentee will argue that the KOH-step combination is hindsight. A petitioner would be well served by a reference showing pyramidal/textured holes in Si.
3. I do not have the actual cited-art list. If the examiner already considered and allowed over Garnett/Lin/Chang-Hasnain, the petitioner must present new art or a materially different combination, and must reckon with the statutory-estoppel-adjacent and § 325(d) arguments in any IPR.
4. The claim requires a full vertical PIN within each pillar. Garnett's radial junction must be converted, and the petitioner should point to a reference teaching axial PIN in a micropillar (common in the micropillar-photodiode art) rather than resting on "routine substitution" alone.

---

10. Bottom line

1. The real claim 1 is much narrower than the earlier-generated section assumed. It is Si-only, pillar-only, and requires dielectric-filled, pyramid-shaped, in-pillar holes, a vertical PIN within each pillar, dielectric pillar separators, a buried dielectric, 100–2,000 nm pitch, and 100–5,000 nm absorption lengths. The earlier "cathode/anode/reverse-bias/microstructure-enhanced region" overview describes the specification's summary of invention, not the granted claims, and should not be used for § 103 purposes.
2. On the verified record, the strongest § 103 case is a three-way combination: Garnett (Si pillar array, polyimide dielectric fill between pillars, 73× path-length enhancement, 530 nm pitch) as the primary reference, Lin (Si nanohole arrays as equivalent light traps) to supply holes, and routine KOH/TMAH anisotropic etching of (100) Si to supply the pyramidal geometry — optionally with SOI/BOX (admitted in the spec) for element 1g and Chang-Hasnain HCG art for the 100–2,000 nm near-wavelength period.
3. The motivation is supplied by the applicant's own specification, which articulates the 850 nm weak-absorption/high-bandwidth problem and identifies resonance, scattering, near-field, and HCG effects as the mechanism. Under KSR, applying a known light-trapping technique to a known Si photodiode "ready for improvement" is obvious.
4. The claim's 1.1× threshold and "CMOS compatible" recitation are its soft underbelly — a 10% incremental enhancement and a statement of intended use add essentially nothing over the art.
5. The priority date is the pivot. If claim 1 loses § 120 benefit (a live issue in the W&Wsens v. Samsung litigation over the related '948/'360/'871 patents, per the earlier section), the 2016–2021 art — including Zang et al. 2017, already on the face of a family member — becomes available and the obviousness case becomes substantially easier. I could not verify priority support for "pyramid-shaped in-pillar holes" in the 2015 material, and I flag that as the single highest-value item to check.

---

11. What I would need to make this authoritative

1. The PTO-1449 / "References Cited" list from the 17/532,831 file wrapper, and the examiner's reasons for allowance — to know what was already considered.
2. The complete claim set (claims 12, 18, 19, 21 were truncated in my source).
3. Written-description support for "pyramid-shaped, in-pillar holes" in 14/947,718 (filed 2015-11-20), to resolve the priority question.
4. The contents of US 4,814,847; US 2007/0215968; US 2010/0059822; US 7,977,637; US 8,357,960 — the "Y"-marked EP citations, which are the most likely candidates for the US combination.
5. The IPR2025-00995 petition against US 10,468,543, to see the petitioner's established art set for this family.

Caveat repeated: I could not open the "Prior Art" section referenced in the instruction. Nothing above should be treated as a verified claim chart until items 1–4 are obtained.