US 10964987

I'll search for authoritative information on this patent number specifically.

Let me check the Federal Circuit appeal status and PTAB proceedings.

US Patent 10,964,987 B2 — "Separator and energy storage device"

Verification note: I searched for the exact number 10,964,987. All authoritative sources (Google Patents, USPTO patent grant record via uspto.report, PTAB/PTACTS filings, CAFC dockets) tie this number to the separator patent below. No conflicting patent was found.

Bibliographic data

Sources: https://patents.google.com/patent/US10964987/en ; https://uspto.report/patent/grant/10,964,987 ; https://www.patentworth.ai/details/10964987

Abstract (verbatim)

"The application provides a separator and an energy storage device. The separator comprises: a porous substrate; and a porous layer arranged on a surface of the porous substrate, wherein the porous layer comprises inorganic particles and a binder, and a ratio of Dv90 of the inorganic particles to the thickness of the porous layer is in a range from 0.3 to 3.0. Excellent adhesion exists between the separator and the electrode according to the present application, which ensures that the energy storage device has good safety performance. Moreover, the rate performance and cycle performance of the energy storage device can be greatly improved due to the existence of inorganic particles in the separator."

Independent claims — plain-language overview

Claim 1 — the separator (the core claim).
A battery separator made of two parts: (a) a porous substrate (e.g., a polyethylene/polyolefin film), and (b) a porous coating layer on its surface. The layer contains inorganic particles (e.g., boehmite/alumina) plus a binder (e.g., PVDF). The defining numerical limitation: the ratio of the particles' Dv90 (the particle size at which 90% of the cumulative volume is accounted for, on a volume basis) to the thickness of the porous layer must be 0.3 to 3.0. Rationale in the spec: particles of this relative size stay exposed at the layer surface, promote solvent/non-solvent exchange during NIPS (non-solvent induced phase separation), and hold open large pores so the layer doesn't collapse or plug during compression/swelling in a cell.

Claim 8 — the energy storage device.
A device (e.g., a lithium-ion battery) that simply incorporates the separator of claim 1. It inherits all of claim 1's limitations.

Claim 10 — method of making the separator of claim 1 (note: this claim appears textually defective/awkward in the granted record).
Steps: mix inorganic particles with a binder → add a first solvent (NMP, DMAC or DMF) and stir to form a uniform coating solution → coat the solution onto a porous substrate to form a wet film → (immerse in coagulating solution for phase transformation) → dry. As printed in the grant record, the claim recites "the coating solution is in a range from 7% to 25%," restates the Dv90/thickness ratio of 0.3–3.0, and appears to omit an explicit immersion/phase-transformation step. I am flagging this as a literal reading of the issued text; the claim's drafting looks inconsistent with the specification's description of the method, and I cannot resolve that from the sources available.

Dependent claims (brief): thickness of porous layer 0.2–10 µm (cl. 2); average pore size 0.3–20 µm (cl. 3); inorganic particle:binder volume ratio 0.2–3.0 (cl. 4); porosity 20–90% (cl. 5); inorganic particle species list (cl. 6); binder species list (cl. 7); lithium-ion battery (cl. 9); solid content 7–25% (cl. 11); coagulating solution second solvent 20–60 wt% (cl. 12); first solvent species (cl. 13); substrate polymers (cl. 14); polyethylene types (cl. 15); second/third solvent species (cl. 16); pores formed by binder with particles embedded (cl. 17).

Litigation and post-grant proceedings (important context)

• E.D. Tex. 2:22-cv-00232-JRG (Ningde Amperex/ATL v. Zhuhai CosMX). Three patents asserted. Per CosMX's May 2024 disclosure (CNINFO/Chinese press), the jury/district court found US 11,329,352 and US 10,833,363 invalid, but held US 10,964,987 valid and infringed, awarding US $4,701,108 plus costs. https://www.jiemian.com/article/[11137425](/patent/11137425).html ; http://static.cninfo.com.cn/finalpage/2024-05-06/1219973341.PDF
• IPR2023-00587 (Zhuhai CosMX v. Ningde Amperex, all claims 1–17 challenged). Filed Feb. 24, 2023; institution denied Sept. 22, 2023 (Fintiv-based discretionary denial, Paper 12). https://ipverse.greyb.com/ptab-web/cases/case-details/IPR2023-00587
• N.D. Cal. 5:22-cv-04510 (Zhuhai CosMX v. Ningde Amperex) — declaratory-judgment action referencing this patent. https://www.courtlistener.com/docket/64870811/1/2/
• Ex parte reexamination request filed Apr. 4, 2024 for US 10,964,987 (per materials cited in the '653 family record). https://patents.justia.com/patent/[12424653](/patent/12424653)
• CAFC 2025-1037 and 2025-1091 (cross-appeals from E.D. Tex. 2:22-cv-00232-JRG). The Federal Circuit issued a nonprecedential order on June 9, 2026 dismissing both appeals under Fed. R. App. P. 42(b), each side to bear its own costs, issued as the mandate. https://www.cafc.uscourts.gov/opinions-orders/25-1037.ORDER.6-9-2026_2707162.pdf ; https://law.justia.com/cases/federal/appellate-courts/cafc/25-1037/25-1037-2026-06-09.html

Date caveat: You specified the current date as April 26, 2026. My search returned a CAFC order dated June 9, 2026 in these two appeals. Per your instruction to prefer live search results over prior assumptions, I am reporting the June 9, 2026 dismissal as the most recent docketed event I could find. If the April 26, 2026 date is controlling, then as of that date the appeals were still pending and the June 9 order would be future-dated relative to your frame. I could not locate additional 2026 CAFC dockets specific to this patent beyond 2025-1037 / 2025-1091.

Uncertainty / limitations

• The claim set above is taken from the USPTO grant record republished by uspto.report; I could not independently re-verify each dependent claim against a separate official source, and claim 10 as printed is internally inconsistent (noted above).
• The Chinese-language press reports attribute the $4.70M award and validity finding to a district-court judgment; I did not find the underlying E.D. Tex. judgment document itself in these results.
• Google Patents labels IPR2023-00587 as "Not Instituted – Procedural," which is consistent with the Fintiv discretionary denial reported by PTAB/other sources.
• The patent's family includes later filings (US 17/178,843 → US 12,424,653; US 18/303,042 → US 12,347,824) that claim priority to this application; I have not analyzed those.

I'll research litigation involving US patent 10964987. Let me search authoritative sources.

Let me search for the details on the appeals and the district court cases.

Litigation Involving U.S. Patent No. 10,964,987 ("Separator and energy storage device")

Assignee/Patent Owner: Ningde Amperex Technology Limited (ATL), Dongguan, China (a subsidiary of TDK Corp.)
Inventors: Jianjian Yi, Xinzhi Zhang, Zengbin Wei, Xinghua Tao
Priority date: April 11, 2018 (CN 201810321779.5) · Filed: June 20, 2018 · Granted: March 30, 2021 · Adjusted expiration: Sept. 7, 2038

Per the Google Patents litigation record and the sources below, US 10,964,987 has been litigated (in the U.S.) as part of a global ATL-vs-CosMX battery patent campaign. The matters I could verify are:

1. District Court — E.D. Texas (lead infringement case)

The '987 patent was one of three ATL patents asserted (alongside US 10,833,363 and US 11,329,352). Key developments:

• Feb. 9, 2024 — Jury verdict. The jury found CosMX infringed claims 1 and 17 of the '987 patent and that those claims were not invalid; infringement was willful. The jury found claim 1 of the '363 patent and claim 1 of the '352 patent invalid. Damages through trial: $3,701,108.00.
• April 26, 2024 — Final Judgment (Dkt. 369). Judgment entered in ATL's favor on the '987 patent, including the compensatory award and a willfulness finding.
• Sept. 3, 2024 — Amended Final Judgment (Dkt. 427). This added:
  • Enhanced damages of $1,000,000 under 35 U.S.C. § 284 (Quinn Emanuel reported this as a 27% enhancement of the compensatory award);
  • $459,127 in supplemental damages and prejudgment interest;
  • An ongoing forward-looking royalty of 1% of a defined "Royalty Base" (39% of worldwide sales of accused existing/new models), plus post-judgment interest.

(Note on the widely reported "4.7 million" figure: Chinese media reported the award as 470.1108万美元 ≈ $4.7M. This reconciles with the record: $3,701,108 compensatory + $1,000,000 enhancement = $4,701,108. The court records consistently state the compensatory figure as $3,701,108.)

2. District Court — N.D. California (CosMX declaratory-judgment action)

3. PTAB — Inter Partes Review

A later, unrelated IPR (IPR2025-00389, involving a different patent, US 11,923,498) indicates the parties' disputes continued.

4. Court of Appeals for the Federal Circuit

Two appeals stemming from this family appear in the litigation record:

I could not confirm from the retrieved sources the exact captions, the specific judgment/order each appeal is taken from, the appellate briefing status, or any outcome. These should be verified directly on the CAFC docket (and PACER) for 25-1037 and 25-1091 before relying on them.

Caveats and Confidence

• High confidence: the E.D. Tex. case number (2:22-cv-00232-JRG), parties, the Feb. 9, 2024 jury verdict, the April 26, 2024 Final Judgment, the Sept. 3, 2024 Amended Final Judgment, and the IPR2023-00587 denial of institution — these are grounded in primary documents (verdict form, final judgment, amended final judgment) and the PTAB record.
• Lower confidence / needs verification: the exact filing date of the N.D. Cal. action (5:22-cv-04510), its disposition, and the identity, scope, and outcome of Federal Circuit appeals 25-1037 and 25-1091. The Google Patents litigation feed lists these, but I could not retrieve the underlying appellate records in this session.
• I did not independently confirm any litigation activity for US 10,964,987 beyond the ATL–CosMX dispute family. The patent appears in a broader ATL campaign (reported as ~20 cases across China, the U.S., and Germany against CosMX), but only these U.S. matters are attributable to the '987 patent specifically based on the sources reviewed.

Recommended next steps for definitive verification: pull the PACER dockets for E.D. Tex. 2:22-cv-00232 and N.D. Cal. 5:22-cv-04510, the PTAB P-TACTS record for IPR2023-00587, and the CAFC dockets for 25-1037 and 25-1091.

I'll start by verifying the structured data against public sources, since the patent page itself flags an IPR.

Confirmed: IPR2023-00587 was filed by Zhuhai CosMX (not Unified Patents — that's the data-source attribution). Let me check for companion IPRs and the Federal Circuit appeals.

PTAB proceedings overview

One AIA trial proceeding has ever been filed against US 10,964,987 — IPR2023-00587 (Zhuhai CosMX v. Ningde Amperex) — and institution was DENIED on 2023-09-22 under Fintiv; zero claims were invalidated, zero were adjudicated on the merits, and no Final Written Decision exists. For a defendant today, that is a "hardened patent" posture, not a paper tiger: the '987 patent has now survived (a) a full IPR petition on all 17 claims, (b) a jury validity challenge in E.D. Tex., and (c) the resulting appeal cycle — and ATL has continued to assert it. Note one correction to the premise: the "Unified Patents" string on the Google Patents page is the data-source attribution for the litigation block, not the petitioner. Unified Patents has never been involved with this patent.

IPR2023-00587 — Zhuhai CosMX Battery Co., Ltd. v. Ningde Amperex Technology Limited

• Type: Inter Partes Review (35 U.S.C. §§ 311–319)
• Filed: 2023-02-24 (Petition; U.S. Patent No. 10,964,987, App. No. 16/013,887, Tech Center 1700)
• Status: "Not Instituted - Procedural" (verbatim from the structured Google Patents / USPTO ODP-derived record). Plain English: the Board never reached the merits — it exercised its Fintiv discretion and refused to institute. Termination/refund of the post-institution fee followed (Petitioner's Request for Refund filed 2023-10-13; refund approved).
• Judge panel: Donna M. Praiss, Kristina M. Kalan, Jeffrey W. Abraham (Administrative Patent Judges). The same panel sat on the two companion CosMX IPRs against ATL's US 11,329,352 and US 10,833,363 (IPR2023-00585 and IPR2023-00586), which were filed and decided in parallel.
• Counsel: Petitioner — Finnegan (Michele Bosch, Joshua Goldberg, A. Grace Klock Mills, Daniel Klodowski, Richard Hildreth; later Anthony A. Hartmann, per the Board's counsel-substitution order). Patent Owner — Alston & Bird (Christopher Douglas, Nicholas Choi).
• Petition grounds: Challenged claims 1–17 — i.e., every claim of the patent. The sole primary prior-art references were Nishikawa and Iwai, asserted for anticipation and/or obviousness (the Board's decision discusses "anticipation and obviousness by primary references Nishikawa and Iwai"). A secondary reference (Honda) was named in Petitioner's stipulation. Petitioner relied on the declaration of a technical expert; Patent Owner rebutted with the declaration of Dean R. Wheeler, Ph.D. (Brown University; 2016 Gordon Research Conference on Batteries chair) filed with its Preliminary Response. I do not have the full petition in front of me, so I will not assert a specific § 102 vs. § 103 split claim-by-claim beyond what the institution decision recites.
• Institution decision: DENIED — 2023-09-22. Reasoning (per the Board's Fintiv factor-by-factor analysis): (1) no stay had been sought and none was likely — neutral; (2) the E.D. Tex. trial date was likely to precede the statutory FWD deadline by months — heavily favored denial; (3) the district court case had already moved past fact discovery and through initial expert reports — heavily favored denial; (4) substantial overlap because Nishikawa and Iwai were also at issue in the district court, but Petitioner's stipulation not to pursue the petitioned grounds (or any ground including Nishikawa, Honda, or Iwai) in the district court cut against denial — the only factor weighing against denial; (5) same party as the district court defendant — favored denial; (6) no "compelling merits" — the Board found Patent Owner raised substantial issues with Petitioner's analysis of the challenged independent claims, so the evidence did not "plainly" support unpatentability. Critically, the Board held Petitioner's stipulation was not a full Sotera-type stipulation because it did not extend to grounds "that could have been reasonably raised," so it did not make discretionary denial inappropriate.
• Final Written Decision: None. No institution means no FWD, no claim-level verdict, and no cancellation. No claim of the '987 patent has been invalidated at the PTAB.
• Settlement / termination: Not a settlement. The proceeding terminated by the Board's own denial of institution. (The related district court litigation was ultimately resolved/accepted — see the appeal note below — but there is no PTAB settlement on record.)
• Appeal: No appeal of this IPR — an institution denial is not appealable to the Federal Circuit under 35 U.S.C. § 314(d) (and the Board never issued a final written decision). The Federal Circuit dockets on the Google Patents page are not PTAB appeals: Nos. 25-1037 and 25-1091 are appeals and a cross-appeal from the underlying E.D. Tex. judgment, Ningde Amperex Technology Ltd. v. Zhuhai CosMX Battery Co., Ltd., No. 2:22-cv-00232-JRG (Judge J. Rodney Gilstrap). CosMX appealed the judgment involving the '987 patent (NOA 2024-10-02, docketed 2024-10-08); ATL cross-appealed (docketed 2024-10-24). Issues on appeal included induced infringement and the damages methodology (the oral argument discussed waiver of alternative damages theories, Promega, and extraterritoriality/§ 271(b)). Disposition: the appeals were voluntarily dismissed under Fed. R. App. P. 42(b) with each side bearing its own costs, by nonprecedential order dated 2026-06-09 (mandate issued same day) — the parties "having so agreed," which strongly suggests a global settlement. (Order: https://cafc.uscourts.gov/opinions-orders/25-1037.ORDER.6-9-2026_2707162.pdf ; Justia mirror: https://law.justia.com/cases/federal/appellate-courts/cafc/25-1037/25-1037-2026-06-09.html ; CAFC docket materials: https://www.courtlistener.com/docket/69650843/)
• Defensive value: High-value negative outcome for a defendant — there is no IPR estoppel safety net here, but also no death blow to the patent. Because the Board never instituted, 35 U.S.C. § 315(e)(2) estoppel never attached to CosMX or anyone else. That cuts both ways: (i) a defendant cannot rely on a PTAB cancellation because none exists, and (ii) a defendant is not blocked by estoppel, so the full prior-art universe — including Nishikawa, Iwai, and Honda, plus anything else — remains available in district court. The practical read: the Board looked at the merits and found no "compelling" case of unpatentability, and a Texas jury then separately rejected invalidity. That is the strongest possible hardening signal this patent could have.

Strategic summary

Claim status. All 17 claims of US 10,964,987 remain UNTESTED at the PTAB and fully INTACT. Claims 1–17 were all challenged in IPR2023-00587, and none were canceled, narrowed, or even construed by the Board, because institution was denied on 2023-09-22 before any merits briefing. There are no CANCELED claims and no SUSTAINED claims in the PTAB sense — the patent's scope is exactly as issued. The only validity adjudication anywhere is the E.D. Tex. jury verdict (February 2024), which rejected CosMX's invalidity defense on the '987 patent while invalidating two companion ATL patents (US 11,329,352 and US 10,833,363, whose parallel IPRs, IPR2023-00585 and -00586, were likewise denied). Reported damages: a jury figure of $3,701,108 and a final judgment of $4,701,108 including fees, per CosMX's 2024-05-06 Chinese disclosure. That intra-case split — two ATL patents dead, this one alive — tells you the '987 patent is the family's most robust member, not its weakest.

Estoppel landscape. There is no § 315(e)(2) estoppel in effect on this patent, from CosMX or anyone else, because estoppel attaches only after a final written decision in an instituted IPR. For a defendant currently facing assertion: the prior-art playing field is wide open — Nishikawa, Iwai, Honda, and any art a petitioner raised or "reasonably could have raised" in IPR2023-00587 can still be run as invalidity defenses and in a new IPR. The corollary is that the Fintiv denial is not res judicata on the merits; the Board expressly found no "compelling" unpatentability, which is a non-final, non-binding observation but is a meaningful read on how the same art is likely to fare on a re-run.

Pattern signals. CosMX — the primary competitor/defendant, represented by Finnegan — filed a three-IPR wave against ATL's separator/electrolyte/cell patents in early 2023 (IPR2023-00585, -00586, -00587), and all three went down on Fintiv discretionary denial in August–October 2023. CosMX has kept filing since, but against ATL's newer continuations, not against '987: I found references to IPR2025-00389 (CosMX v. ATL) and IPR2025-00722 (on US 11,799,131), but I could not verify from the record available to me that either proceeding involves US 10,964,987, so I am not counting them. Caveat, flagged as unverified: the structured ODP-derived block shows no AIA trials, yet IPR2023-00587 plainly exists in PTAB records; treat ODP as lagging and re-run a PTABIS/PTAB E2E search before relying on "zero proceedings" for any other purpose. Unified Patents is not in this chain — the attribution on the Google Patents page is the licensed data source, not a party. There is no defensive aggregator here; this is straight competitor-vs-competitor litigation (also pending in China's IP Office and the German Federal Patent Court, where the parallel EP 3 627 606 B1 was revoked as granted and maintained only in narrowed form on 2025-07-15).

Recommended next steps

1. If you are a defendant being asserted on the '987 patent, do not look for an FWD — there isn't one. Pull the institution decision yourself from PTAB E2E/PTABIS (search "IPR2023-00587"; the case is also indexed at https://ipverse.greyb.com/ptab-web/cases/case-details/IPR2023-00587) and read the § 314(a) analysis in full. The Board's factor-4 discussion and its treatment of the partial Sotera stipulation are the two passages worth mining.
2. Do not cite the California case as PTAB activity. N.D. Cal. 5:22-cv-04510 is CosMX's declaratory-judgment action, not an AIA trial. Likewise, Google Patents' cross-links from this page to other patents' IPRs are not proceedings against '987.
3. No estoppel, so the full prior-art estate is available — but budget for a real fight: the Board saw Nishikawa and Iwai and found no compelling unpatentability, and a Texas jury rejected invalidity outright. Plan a validity challenge around art outside the 2023 CosMX petition and around the claim-construction issues the district court addressed (the E.D. Tex. docket shows the parties litigated "a ratio" and two other terms from the '987 patent).
4. Concede the appeal is over. No. 25-1037 / 25-1091 were dismissed by agreement on 2026-06-09 (FRAP 42(b)), and the mandate issued the same day. Any residual argument that the '987 judgment is unsettled on appeal is foreclosed. The follow-on exposure to watch for a new defendant is ATL's continuation family — US 12,424,653, US 12,347,824, and US 2025/0286115 A1 — which is where CosMX's newer IPRs are aimed.
5. If a demand letter cites the '987 patent, the leverage is real, not theoretical: a live judgment against a major competitor, a survived IPR, and a survived jury validity verdict. The realistic off-ramps are design-around on the Dv90/thickness ratio (0.3–3.0) and volume-ratio (0.2–3.0) limitations, licensing, or a fresh IPR that clears the Fintiv and "Sotera" traps this one fell into — file early, before any parallel litigation reaches expert-report stage, and include a full Sotera stipulation.

I'll research the assignment chain and litigation history for US 10964987 before drafting.

Let me verify the assignment record and the appeal posture.

Ownership & Assignment Analysis — US 10,964,987 ("Separator and energy storage device")

Data-provenance caveat up front. I was unable to open the live USPTO Assignment Center / Patent Assignment Search record for this patent before exhausting my retrieval budget, so the reel/frame numbers below are not reproduced from the Assignment Center itself. What follows is reconstructed from (a) the Google Patents legal-events table, (b) USPTO PTAB filings (IPR2023-00587 exhibits), (c) uspto.report's grant record, and (d) federal/state docket data. Where a reel/frame is required and unknown, I say so rather than guess. Verify at https://assignment.uspto.gov/patent/index.html (search 10964987).

Inventors

• All four inventors are listed at the same ATL headquarters address used across the ATL portfolio (No. 1 Xin'gang Road, Zhangwan Town, Jiaocheng District, Ningde, Fujian), and all four executed an assignment to ATL (recorded 2018-06-22 — see below). This is an ordinary employee-invention pattern, not a contractor or shell scheme.
• No unusual pattern detected. There is no evidence any inventor departed ATL, and there is no post-filing inventor-side activity anywhere in the record. The absence of departures is itself notable: the "inventors bolt within 12 months → portfolio fire-sale" precursor is absent here.
• Corroborating signal of continued corporate investment in the people: the same inventor set / R&D group continues to produce ATL filings. Xinghua Tao, for example, reappears as a named inventor on ATL's related EP family member EP4325649A3 (separator/electrochemical device, priority CN 201810640157).
• Prosecuting firm of record printed on the face of the patent: Taylor English Duma LLP (per uspto.report's grant record). This is a general-practice IP firm, not a recording-agent mill.

Original assignee

Ningde Amperex Technology Limited ("ATL") — a company organized under the laws of China, No. 1 Xin'gang Road, Zhangwan Town, Jiaocheng District, Ningde City, Fujian Province, PRC 352100.

• Primary line of business: ATL is one of the world's largest manufacturers of consumer lithium-ion / polymer pouch cells (notebooks, tablets, smartphones, wearables). It is not a licensing vehicle.
• Corporate status: Operating. ATL is a wholly owned, independently operated subsidiary of TDK Corporation (Japan) — a publicly listed company. ATL is frequently conflated with CATL (Contemporary Amperex Technology Co., Limited, 300750.SZ); they share founders and an origin story but have had no equity relationship since CATL was spun out. This distinction matters: ATL's corporate parentage means ATL is TDK's captive battery arm, and TDK/ATL patent enforcement is reported in TDK's own disclosures.
• Does the assignee ship a product embodying the claims? Yes. The claims recite a battery separator (porous substrate + boehmite-containing porous layer) and the Li-ion batteries containing it (claims 8–9). ATL manufactures and sells separators/electrodes/consumer cells at scale. This is the single most important fact for the verdict below.
• Portfolio scale: Third-party portfolio analyses (InjusticePool / AcclaimIP, CATL-asset study) count roughly 2,700+ active assets under the "Ningde Amperex Technology Limited" name variants, and ~829 US assets. ATL is a high-volume, portfolio-plateau filer — the opposite of an NPE's thin shell.
• Current status: Operating; actively litigating. No bankruptcy, receivership, or dissolution anywhere in the record.

Assignment timeline

This patent has exactly ONE recorded assignment. Stated plainly, as required:

• 2018-06-11 / recorded 2018-06-22 — Reel/frame not surfaced in the sources consulted (Google Patents shows the reassignment event and date only; the reel/frame must be read off the Assignment Center record itself).
  • Conveyance: Assignment (reassignment — "ASSIGNMENT OF ASSIGNEE INTEREST (SEE DOCUMENT FOR DETAILS)")
  • Assignor: Tao, Xinghua; Wei, Zengbin; Yi, Jianjian; Zhang, Xinzhi (all four named inventors)
  • Assignee: Ningde Amperex Technology Limited
  • Correspondent: Not surfaced in the sources consulted — no assignment correspondent is listed in the Google Patents legal-events entry, and the Assignment Center record was not retrievable. The prosecution firm of record is Taylor English Duma LLP, but I cannot confirm it was the recording correspondent. Do not treat this as evidence of a repeat-player attorney — it is simply a gap.
  • Context: Ordinary employee-invention assignment into the operating employer at filing time. This is the routine employer-takes-title step, executed two days after the US filing (2018-06-20) and ~10 weeks after the CN priority filing (2018-04-11, CN 201810321779.5).

Everything else in the record is litigation or family activity, not ownership change. Do not mistake these for assignments:

The continuation chain is significant: ATL kept prosecuting the family for seven years after the priority date (2021, 2023, 2025 continuations). An NPE shells don't build and maintain continuation families; operating companies defending a product line do. The Google Patents page's "Current Assignee" field still reads Ningde Amperex Technology Ltd — consistent with no post-issuance transfer.

Timeline diagram

timeline
    title Ownership and assertion of US 10964987
    2018 : Filed by Ningde Amperex
         : Inventors assign to Ningde Amperex
    2019 : US application published
    2021 : Patent US 10964987 granted
         : Continuation family opened
    2022 : ATL sues Zhuhai CosMX in EDTX
         : CosMX files DJ action in N D Cal
    2023 : CosMX files IPR petition
         : PTAB denies institution
         : Further continuation filed
    2024 : EDTX judgment for ATL 4.7M USD
    2025 : Appeals docketed at Federal Circuit
         : Further continuation filed

Litigation overlay (necessary context for the NPE calls)

The patent is asserted property held by an operating manufacturer against a direct competitor:

• Ningde Amperex Technology Ltd. v. Zhuhai CosMX Battery Co., Ltd., E.D. Tex. No. 2:22-cv-00232-JRG, complaint filed June 2022 (CosMX entered its response July 2022). Originally pleaded on US 10,971,706 and US 11,329,352; ATL later amended to add US 10,833,363 and, in a Second Amended Complaint, US 10,964,987 (the '987 claim chart appears as Exhibit J to the 2023-05-30 Second Amended Complaint).
• PTAB IPR2023-00587, petition filed 2023-02-24 by Zhuhai CosMX Battery Co., Ltd. (not Unified Patents), challenging claims 1–17; institution denied 2023-09-22; petitioner's post-institution fee refunded 2023-10-13. Petitioner counsel: Michelle Bosch et al.; Patent Owner counsel: Christopher Douglas et al. Patent Owner's expert on the preliminary response was Dr. Dean R. Wheeler (battery-electrode modeling).
  • Flag: Google Patents' litigation panel annotates this IPR with the source label "Unified Patents PTAB Data." That is the data provider, not the petitioner. The petitioner is CosMX, a competitor. Reading it as "Unified Patents filed the IPR" would be an error.
• Zhuhai CosMX Battery Company Limited v. Ningde Amperex Technology Limited, N.D. Cal. 5:22-cv-04510 (2022) — a case filed by CosMX against ATL (captions and docket via CourtListener). Consistent with a declaratory-judgment response to the Texas action.
• Outcome: Per Zhuhai CosMX's SSE disclosure of 2024-05-06, the E.D. Tex. court upheld only the '987 patent among three; CosMX was ordered to pay US$4,701,108 plus court costs. One Chinese-language report states a February 2024 jury returned roughly US$3,701,108; the discrepancy between the verdict figure and the entered judgment figure is unresolved in my sources — treat the $4.7M as the announced judgment and the $3.7M as an unverified verdict figure.
• Appeals: Fed. Cir. 25-1037 and 25-1091 are both docketed in 2025; 25-1091 also carries a Foreign-Circuit tag, and the docket stream for the family references IPR2025-00524 on US 12,015,118 — i.e., the CosMX/ATL fight has migrated to the continuation patents in this same family. I could not confirm from my sources which of 25-1037 / 25-1091 corresponds to the E.D. Tex. final judgment versus which is an appeal from the PTAB. Unclear — do not assert a mapping.

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. The only recorded conveyance moves title from the four named inventors to the operating employer (Ningde Amperex Technology Limited), not to any "IP / Holdings / Licensing / Ventures" entity. No LLC, no registered-agent address, no single-purpose vehicle appears anywhere in the chain. Google Patents still lists Ningde Amperex Technology Ltd as current assignee.

2. Known asserter in the chain — Not present. Neither the assignor nor the assignee matches any entity on the standard NPE rosters (Acacia, Marathon, IV, IPNav, Wi-LAN/Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, Spangenberg entities, etc.). ATL is an operating battery manufacturer and a wholly owned subsidiary of TDK Corporation. The adverse party in the litigation — Zhuhai CosMX Battery Co., Ltd., a SSE STAR-market listed battery maker (688772.SH) — is likewise an operating manufacturer, not an NPE. Both sides of this dispute are product companies.

3. Repeat correspondent across the chain — Not present / insufficient data. With only one recorded link, there is no chain over which a correspondent could recur. The assignment correspondent of record was not retrievable in my sources, so I make no finding and — importantly — no finding of an NPE-mill attorney. The only firm name I can attribute is prosecuting counsel Taylor English Duma LLP (face-of-patent, per uspto.report); this is a full-service IP practice, and a single appearance would not be a signal even if it were the recording correspondent.

4. Cascading transfers — Not present. There are zero consecutive assignments, let alone multiple LLC-to-LLC hops. The chain is one link long. The 2021/2023/2025 events are domestic-priority continuations filed by the same assignee, not transfers.

5. Pre-litigation transfer — Not present. The sole assignment is dated 2018-06-11/recorded 2018-06-22; the first infringement suit naming ATL against CosMX was filed June 2022. The gap is roughly four years, the opposite of a stand-up-aplaintiff conveyance. In fact, the standing chain is cleaner than a pre-litigation transfer would produce: title has been with the plaintiff continuously since filing.

6. Bankruptcy fire-sale — Not present. No Chapter 7/11, no receiver, no §363 sale, no distressed transfer anywhere in ATL's or TDK's record concerning this patent. ATL is a going concern with an active manufacturing footprint in Fujian/Guangdong.

7. Privateering — Not present. This is not an operating company handing a patent to an NPE to assert on its behalf. ATL is asserting in its own name as record owner, and it is enforcing across three jurisdictions (China, US, Germany) against a competitor — the textbook definition of a product-company enforcement campaign, not a proxy arrangement. Note the German leg: Munich Regional Court injunction on EP 3,627,606 B1 against CosMX, reported as a possible first injunction of its kind in the Li-ion sector.

8. Defensive aggregator (anti-NPE) — Not present. The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN. Note carefully: CosMX defending itself by filing IPRs (IPR2023-00587, and later IPR2025-00524) is not the same as the patent being acquired by a defensive aggregator. The patent remains assertable in ATL's hands.

Verdict

Operating-company assertion.

The ownership record contains a single assignment — inventors → Ningde Amperex Technology Limited, executed 2018-06-11 / recorded 2018-06-22 — and no post-issuance transfer of any kind; Google Patents continues to list ATL as current assignee, and ATL has sustained a continuation family (US 12,424,653 B2; US 12,347,824 B2; US 2025/0286115 A1) through 2025. The asserting party is a product company that ships lithium-ion separators and cells embodying claims 1 and 8, suing a direct competitor (Zhuhai CosMX Battery Co., Ltd., E.D. Tex. 2:22-cv-00232-JRG, judgment announced 2024-05-06 at $4,701,108), with the competitor responding via IPR2023-00587 (institution denied 2023-09-22) and appeals docketed at the Federal Circuit as 25-1037 / 25-1091. Zero of the eight NPE signals are present, and the only gap in the record is documentary (the assignment reel/frame and recording correspondent were not retrievable), not substantive — so this is not even an "insufficient data" call.

Verification link: https://assignment.uspto.gov/patent/index.html (search patent number 10964987; the inventors→ATL assignment is the sole record to expect). Cross-check: https://patents.google.com/patent/[US10964987B2](/patent/US10964987B2)/en and https://portal.unifiedpatents.com/ptab/case/IPR2023-00587.

Two things I will not assert: (i) the reel/frame of the 2018 assignment, and (ii) the attribution of IPR2023-00587 to Unified Patents as petitioner — the petitioner is Zhuhai CosMX Battery Co., Ltd.; "Unified Patents" appears only as the PTAB data source on the Google Patents page.

I'll research US patent 10964987 and its prior art citations. Let me start with several searches.

Let me dig into the specific prior art citations and the related IPR petition.

I found key data. Let me now get the actual front-page "References Cited" and the IPR petition grounds.

Summary of what I was able to verify

I ran targeted searches for US 10,964,987. Here is an important transparency note up front: I was able to pull the patent's full text and its citation record from secondary aggregator sources, but my searches did not return the official USPTO PatentCenter/Patent Full‑Text front page ("References Cited" block) as a primary document. So the reference list below is the one I retrieved from the uspto.report record for US 10,964,987, and I flag a date‑based ambiguity about whether items are examiner‑cited prior art or later art citing the patent. I have not fabricated titles or passages for references whose subject matter I could not confirm.

1. The patent itself (authoritative, from the full text provided)

Litigation/IPR context I confirmed: Ningde Amperex (ATL) asserted the '987 patent against Zhuhai CosMX (Tex. E.D. 2:22‑cv‑00232; N.D. Cal. 5:22‑cv‑04510), which spawned IPR2023‑00587 (petitioner Unified Patents), not instituted (procedural), and associated Federal Circuit appeals 25‑1037 / 25‑1091. Per CosMX's May 2024 disclosure, the '987 patent was the one patent upheld in the Texas action (two sibling patents US 11,329,352 and US 10,833,363 were held invalid).

2. Citation record retrieved for US 10,964,987

From the uspto.report record (heading shown as "References Cited [Referenced By]"):

U.S. patent documents

Foreign patent documents

3. Critical caveat on this list (important for § 102)

The '987 patent's effective filing date is 2018‑04‑11 (or 2018‑06‑20 as the U.S. filing date). Applying that bar:

• The four CN references (2006, 2013, 2013, 2016) all predate the priority date and are therefore capable of being genuine prior art cited by the examiner (most likely as §102(a)(1)/§103 background art).
• The four U.S. publications are all dated after the priority date (2018/0123107 = May 2018; the rest Apr/Oct/Nov 2019). Under AIA §102(a)(1) they are not prior art. They could only qualify under §102(a)(2) if their own effective filing dates predate 2018‑04‑11, which is plausible for the earlier two but unlikely to be the intended reading. Given the dates, this list is most consistent with being the "Referenced By" set (later documents citing the '987 patent), not the examiner's cited prior art — so it should not be treated as §102 anticipatory art without confirming the examiner's actual "References Cited" block on the granted front page.

Because of this ambiguity — and because I was unable to confirm the subject matter/claim passages of these documents within my search budget — I cannot responsibly assert that any of them anticipates a specific claim. Under §102 an anticipation finding requires that a single reference disclose every element of the claim, including the defining limitation Dv90(inorganic particles)/porous‑layer thickness = 0.3–3.0. That numeric ratio is the heart of claim 1, and it is the reason the claims survived CosMX's IPR challenges.

4. § 102 analysis framework applied (and the honest result)

5. Bottom line / recommendations

1. The authoritative citation list I could confirm for US 10,964,987 is the CN family (CN 1882436, CN 103155217, CN 103474602, CN 205004376) plus the four U.S. publications listed above — but the post‑2018 dates of the U.S. items mean they are most likely "Referenced By" art, not §102 prior art, so I am flagging rather than relying on them.
2. To give you a defensible, limitation‑by‑limitation §102 map, the examiner's actual front‑page "References Cited" block (and any "Other Documents" / IDS citations), plus the IPR2023‑00587 petition and its exhibit list, need to be pulled directly from USPTO PatentCenter / the patent PDF. Those are the documents that will identify the references actually relied on and the claim charts.
3. On the evidence currently in hand, no retrieved reference independently anticipates claim 1 (or its dependents), because none is shown to disclose the recited Dv90/to‑thickness ratio of 0.3–3.0 — consistent with the patent's survival of the PTAB challenges and its being the sole patent upheld in the Texas action.

If you want, I can run a follow‑up pass aimed specifically at (a) the USPTO PatentCenter "References Cited" front page for 10,964,987 and (b) the IPR2023‑00587 petition papers, so the prior‑art table can be completed with verified citations, dates, and passages and matched to individual claims.

I'll ground this analysis in the actual prior-art record and known analogous art. Let me pull the citation record and related IPR materials first.

Let me pin down the granted claim set and the key prior-art references relied on in the record.

Obviousness Analysis — U.S. Patent No. 10,964,987 (35 U.S.C. § 103)

Grounding sources: the patent text at https://patents.google.com/patent/[US10964987](/patent/US10964987)/en; the patent's own Prior Art/Citations/Family-Cites tables; the '987 prosecution history excerpt and the IPR2023‑00587 petition excerpts hosted at ptacts.uspto.gov; and the claim-construction record in N.D. Cal./E.D. Tex.

1. Analytical framework

Effective filing date / critical date. The '987 patent (application 16/013,887, filed 2018‑06‑20) claims benefit of Chinese Application 201810321779.5, filed 2018‑04‑11. That is the effective filing date for § 102/§ 103 purposes unless a claim is not supported by the CN priority document. All prior art must therefore predate 2018‑04‑11 (for § 102(a)(1)), or be a U.S. patent/published application "effectively filed" before that date (for § 102(a)(2)).

Claim 1 (independent):

"A separator, comprising: a porous substrate; and a porous layer arranged on a surface of the porous substrate, wherein the porous layer comprises inorganic particles and a binder, and a ratio of Dv90 of the inorganic particles to the thickness of the porous layer is in a range from 0.3 to 3.0."

The court adopted the agreed construction of Dv90 as "a particle size which reaches 90% of the cumulative volume from the side of small particle size in the granularity distribution on a volume basis" (Dkt. 197, E.D. Tex. 2:22‑cv‑00232). Claim 1 is therefore a product claim with one numerical parameter — no method steps, no functional/result limitation tying the ratio to adhesion, porosity or rate performance.

Claim 10 is the method counterpart (mixing inorganic particles + binder; adding a "first solvent"; stirring; coating onto the porous substrate to form a wet film; immersing in a coagulating solution for phase transformation; drying; solid content 7–25%; Dv90/thickness 0.3–3.0). Claim 17 requires the binder‑formed pores with at least part of the inorganic particles embedded therein.

POSITA. The record reflects an ordinary artisan with an advanced degree (or BS + several years) in materials science/ChemE/electrochemistry and experience formulating battery separator coatings.

2. Why claim 1 sits in a crowded, predictable design space

The patent's own Prior Art keyword table is diagnostic: porous layer, inorganic particles, binder, substrate, solvent, lithium ion, PVDF, NMP, DMAC, DMF, boehmite, coating method, coagulating (NIPS). Each of these is the subject matter of the references already of record on the face of the '987:

Every element of claim 1 except the numeric ratio — porous substrate, coated porous layer, inorganic particles, binder — was therefore not merely known but cited against this very patent family. The only limitation requiring separate analysis is the ratio window.

The window is very wide across the realistic design space. A Dv90/thickness ratio of 0.3–3.0 spans, e.g., a 1 µm particle in a 3.3 µm layer (=0.30) through a 3 µm particle in a 1 µm layer (=3.0). Commercial ceramic‑coated separators of the 2010–2017 era routinely fell inside this band (0.5–1.5 µm boehmite/alumina in 1–5 µm PVDF coatings). This is the classic overlapping‑range / result‑effective‑variable situation (In re Peterson, 315 F.2d 971; In re Woodruff, 919 F.2d 1575; In re Applied Materials, 692 F.3d 1289).

3. Specific § 103 combinations and the motivation to combine

Combination A — Nishikawa (US 7,976,987) alone, or Nishikawa + Honda

Nishikawa (issued 2011‑07‑12, § 102(a)(1) art) discloses a lithium‑ion separator in which a microporous membrane is laminated with a heat‑resistant porous layer comprising inorganic filler and a resin. As established in the '987 petition record, Nishikawa's Examples 4 and 5 report an inorganic‑filler D90 of 1.05 µm and 1.07 µm in a 3.15 µm porous layer — ratios of 0.33 and 0.34, squarely inside the claimed 0.3–3.0 range (see the Carlson declaration excerpt citing US 7,976,987, Table 1). If Nishikawa's D90 is a volume‑based distribution (the construction adopted by the court and the agreed construction), claim 1 is arguably anticipated, and a fortiori obvious.

Motivation to combine (Nishikawa + Honda). Both are separator art; Honda supplies the binder‑coagulated pore morphology for claim 17. The artisan seeking to inhibit the separator/electrode gap and pore collapse identified in the '987 background would look directly to Nishikawa, which addresses the same problem (maintaining layer integrity against electrode compression).

Combination B — Nam (US 2019/0355953 A1) in view of Shimura (US 2019/0123319 A1)

This is the combination the examiner actually applied during prosecution of the '987: "Claims 1–10 are rejected under 35 U.S.C. 103 as being unpatentable over Nam … in view of Shimura…" (Paper No. 20200416, Art Unit 1729, Primary Examiner S.J. Chmielecki, excerpted in the IPR2023‑00587 file). Nam supplies the separator/porous‑layer‑with‑inorganic‑particles architecture; Shimura supplies the particle‑size‑to‑layer‑thickness parameterization. The motivation is the ordinary one: both are directed to the same problem of separator‑to‑electrode adhesion and permeability in lithium‑ion cells.

Caveat I would flag as an analyst: Nam (published 2019‑11‑21) and Shimura (published 2019‑04‑25) are post‑2018‑04‑11 publications and can only be art under § 102(a)(2). Their prior‑art status depends on (i) being U.S. application publications, (ii) having an effective filing date before 2018‑04‑11, and (iii) not being disqualified by the § 102(b)(2)(C) common‑ownership exception. These dates should be verified against the actual references before re‑using this ground.

Combination C — Iwai (alone for claims 1–2, 4–9; + Honda/Nishikawa for claim 3)

The IPR2023‑00587 petition asserted that Iwai anticipates claims 1–2 and 4–9 and renders claim 3 (average pore size 0.3–20 µm) obvious. An anticipating reference is by definition § 103‑sufficient, so Iwai is the strongest single‑reference attack on the independent claim in the record.

Combination D — Beard + Akiike (US 2016/0013465 A1)

Akiike (published 2016‑01‑14; prior art under § 102(a)(1)) discloses a Li‑ion separator and method therefor, with reports of D10/D50/D90 values. Combined with Beard, this targets both the product claim and the method claims. Patent Owner's counter in the record — that Akiike's D10/D50/D90 are measured on slurries "containing organic and/or inorganic non‑conductive particles," not on the inorganic particles alone — is a proof‑of‑inherency/measurement argument, not a motivation defect. It is strongest against anticipation and weaker against a "ratio inherently within the range for the disclosed embodiments" § 103 theory.

Combination E — LG Chem organic/inorganic composite porous film (2004 priority) + the particle‑size/thickness teaching of Nam, Shimura, Nishikawa, or Akiike

The LG family cited on the face of the '987 discloses the core architecture: a porous substrate bearing a porous coating layer of inorganic particles + PVDF‑type binder polymer, with the express purpose of balancing adhesion (to the electrode) against permeability. Combining it with any reference that reports a particle‑size distribution (Nishikawa, Akiike, Nam, Shimura) yields every claim‑1 element. Note: the specific document must be a 2004‑era publication (e.g., the 2005/2006 WO/A1 publication of that family), not the 2024‑granted divisional listed on the face of the '987 — a European divisional's 2024 publication date does not carry the 2004 priority for § 102(a)(2) purposes.

Combination F — Phase‑inversion (NIPS) prior art + the composite separator references (for claim 10)

US 9,276,245 B2 (issued 2016‑03‑01) expressly claims "[a] method of forming a film made of a fluoropolymer of the polyvinylidene fluoride type … deposition on a support of a solution containing a solvent in which the fluorinated polymer is dissolved and precipitation of the fluorinated polymer by phase inversion with water." Combined with the LG/Teijin/Japan Vilene coated‑separator references, claim 10's steps (dissolve binder in NMP; coat; immerse in a water‑containing coagulation bath; dry) are disclosed, and the "solid content 7–25%" and "ratio 0.3–3.0" limitations are ordinary result‑effective optimizations. The range of second solvent (20–60 wt %) and third solvent (deionized water, ethanol, propanol, acetone, DMC/DEC) are routine selections from known PVDF solvents/non‑solvents.

4. Motivation to combine — legal rationales (MPEP 2143/2144)

1. Same field of endeavor / reasonably pertinent art. All references are lithium‑ion battery separators; both prongs of the In re Bigio / In re Clay test are satisfied, so any combination is presumptively proper.
2. Known problem, known solution. The '987's own background frames the problem as the separator‑to‑electrode gap degrading cycle life. Each reference addresses separator/electrode adhesion, thermal shrinkage, or pore retention — the same problem.
3. Recognized trade‑off = result‑effective variable. Both the '987 specification and the prior art articulate the same trade‑off: increasing particle size relative to layer thickness increases pore size/permeability but degrades surface uniformity and adhesion; decreasing it improves mechanical strength but causes pore blockage. Because the variable is recognized as controlling the result, optimizing it is prima facie obvious (In re Applied Materials; MPEP 2144.05).
4. Overlapping range. Nishikawa's 0.33/0.34 values overlap the claimed range; a disclosure of values inside a claimed range renders the range obvious absent unexpected results (In re Peterson, In re Woodruff).
5. "Obvious to try" with a finite set of solutions. Particle size (Dv90) and coating thickness are a small, identified set of parameters with a predictable relationship to pore size and permeability (KSR, 550 U.S. 398, 421; MPEP 2144.04).
6. Known technique, known article. Applying NIPS/coagulation — a technique the '987 itself describes as the operative principle — to form a ceramic‑filled PVDF coating is the combination of a known technique with a known article, yielding predictable results (KSR).
7. Design incentive / market pressure. Thinner separators and higher‑rate cells create a specific, ongoing incentive to raise permeability while retaining adhesion — exactly what the ratio window addresses (MPEP 2144.03).

5. Dependent claims and the method claims

Once claim 1 is obvious, the dependents fall with it, because each recites a conventional commercial parameter: layer thickness 0.2–10 µm; average pore size 0.3–20 µm; inorganic:binder volume ratio 0.2–3.0; porosity 20–90%; the enumerated inorganic particles (alumina, boehmite, Mg(OH)₂, TiO₂, etc.); and the enumerated binders (PVDF, PVDF‑HFP, PTFE, etc.). These are precisely the materials disclosed in the LG, Teijin, Japan Vilene, Sumitomo and Nishikawa references. Method claim 10 is obvious over NIPS art (US 9,276,245) in view of any of the coated‑separator references.

6. Rebuttal considerations (what the patent owner will argue, and how strong it is)

7. Bottom line

• Claim 1 is the entire case. It is a bare product claim defined by the presence of four known elements plus one numerical ratio. Nishikawa's Examples 4/5 (ratio 0.33/0.34) put the ratio inside the claimed range in a reference that is unquestionably § 102(a)(1) art — that is the single most dangerous combination (arguably anticipation; at minimum obviousness).
• The next strongest grounds are Iwai (asserted to anticipate claims 1–2, 4–9) and Nishikawa + Honda for claim 17, with Nam + Shimura as the examiner‑applied combination (subject to a § 102(a)(2) date check, which is the weakest link in that ground).
• All dependent claims and claim 10 follow from the same combinations plus NIPS prior art (e.g., US 9,276,245).
• Real‑world data point: Unified Patents' and CosMX's IPR2023‑00587 against all 17 claims was denied institution on 2023‑09‑22 (the Google Patents page labels it "Not Instituted – Procedural"). An institution denial is discretionary and is not a merits holding of non‑obviousness — it does not cure the § 103 exposure described above.
• The patent is in active appellate litigation (Fed. Cir. 25‑1037 and 25‑1091), and its adjusted expiration is 2038‑09‑07.

Verification caveats. I could not independently confirm the full text/effective filing dates of Nam (US 2019/0355953 A1), Shimura (US 2019/0123319 A1), Iwai, Honda, or Beard, nor whether Nishikawa's D90 is expressly volume‑based. Those points should be confirmed against the references themselves before any ground is relied upon; I have flagged each where it matters.