US 6368198

Search-scope note

I ran targeted web searches against Google Patents/Justia (USPTO full-text mirrors), the ITC and district-court documents that surfaced, and general CAFC 2026 opinion coverage. I found no CAFC 2026 docket, appeal, or PTAB proceeding in which US 6,368,198 is the asserted patent. The only 2026 CAFC matters surfacing under a "63 68198" query were unrelated (e.g., Ironburg v. Valve, No. 2024-2088, June 18, 2026; the Arlton/Lite Machines '763 patent petition) — different patents entirely, and I am not auto-substituting them. I cannot access a live USPTO PatentCenter/PAIR record or a certified litigation docket from here, so the litigation statement below reflects "nothing found," not a certified negative.

---

US 6,368,198 B1 — bibliographic data

Source: https://patents.google.com/patent/US6368198/en (also mirrored at https://patents.justia.com/patent/6368198).

Uncertainty flags: (a) Google Patents' own "family applications" table inconsistently prints the priority date as 1997-04-04 while the priority/continuation-in-part field says 1999-11-22; the specification text supports an 09/447,620 (1999-11-22) CIP parent, and separately lists US 08/835,117 (filed 1997-04-04, issued as US 6,039,641) as a related parent. Treat the true earliest effective date with care. (b) I have not independently verified the assignment/chain-of-title beyond the recorded assignment notice.

Abstract (verbatim)

"A CMP pad dresser which has a plurality of uniformly spaced abrasive particles protruding therefrom. The abrasive particles are super hard materials... brazed to a substrate which may be then coated with an additional anti-corrosive layer... usually a diamond or diamond-like carbon... to prevent erosion of the brazing alloy by the chemical slurry... This immunity to chemical attack allows the CMP pad dresser to dress the pad while it is polishing a workpiece... the abrasive particles extend for a uniform distance away from the substrate, allowing for even grooming... A method of producing such a CMP pad dresser is also disclosed."

Independent claims — plain-language overview

Claim 1 — core fabrication method (template/grid braze-sheet route). A method of making a CMP pad dresser: (a) provide a substrate; (b) lay a template with a predetermined aperture pattern onto a sheet of braze alloy; (c) fill the apertures with abrasive particles; (d) remove stray particles not sitting in an aperture; (e) press the particles in the apertures down into the braze sheet so they become partially embedded; (f) lift off the template, leaving the particles in place on the sheet; and (g) attach the particle-bearing braze sheet to the substrate. In plain terms: the grit pattern is set before brazing, using a stencil on a flexible braze preform, so the particles don't cluster the way they do when sprinkled onto molten braze.

Claim 19 — uniform spacing + near-level protrusion. A method of making a CMP pad dresser: (a) provide a substrate; (b) uniformly space multiple abrasive particles on a substrate surface; and (c) affix them so each particle extends to within about 50 micrometers of a predetermined uniform height above the substrate. In plain terms: an outcome-oriented claim — evenly spaced grits whose tips are leveled to within ~50 µm, which is what produces even pad dressing depth.

Claim 20 — uniform spacing + coating with an anti-corrosive layer. A method of making a CMP pad dresser: (a) provide a substrate; (b) uniformly space multiple abrasive particles on the surface; (c) affix them so each extends to a predetermined height; and (d) coat the abrasive particles and the braze alloy with an anti-corrosive layer. In plain terms: the leveled, evenly spaced dresser of claim 19 plus a sealing coating over both grit and braze so the acid slurry can't eat the bond.

Dependent claims (2–18, 21–24) in brief

• 2: grit is single-crystal or polycrystalline diamond or cBN.
• 3–4: apertures sized to hold exactly one particle, at a predetermined size.
• 5–6: particle size ~100–350 µm; all particles within 50 µm of each other.
• 7–9: aperture spacing set to give a predetermined inter-particle distance; that distance ~1.5–10× particle size; pattern is a grid.
• 10–12: particles euhedral, of predetermined shape, with a sharp point facing away from the substrate.
• 13–14: metallic substrate, specifically stainless steel.
• 15–16: braze sheet made by binding braze powder with an organic binder, formed by rolling, extruding, or tape casting.
• 17: braze is a Ni alloy with ≥ ~2 wt% chromium.
• 18: post-braze thickness leaves ~10–90% of each particle exposed.
• 21–24: anti-corrosive layer is diamond or diamond-like carbon; thickness < ~3 µm; DLC atomic carbon content ≥ ~95%; deposited by cathodic arc.

Note: claim 24 reads "A method of making... of 20" — the printed transitional wording is irregular (likely intended "of claim 20"); I am reporting it as it appears, not correcting it.

What the disclosure adds beyond the claims

The specification frames three prior-art failures: electroplated nickel holds grit only mechanically; conventional brazing forms weak braze that acid attacks; and molten-braze surface tension causes clustering, gaps, and "mounds" of uneven tip height. The claimed template-on-braze-sheet route is asserted to give grid spacing, near-monodisperse grit selection, and level tips (template thickness sets protrusion), with concave braze fillets supporting each particle. Examples 1–2 report: 135 µm and 225 µm grids (900 µm pitch) on 100 mm × 6.5 mm disks brazed at 1000 °C, outperforming a randomly loaded disk (the 135 µm disk ~2× it); and DLC-coated (1 µm, cathodic arc) 225/275 µm disks surviving >30 h of in-situ dressing in pH 3.0 acid slurry versus ~1 h for a conventional disk.

Relationship to later family and enforcement color

US 6,368,198 is a parent of US 6,884,155 (10/109,531) and grandparent-lineage to US 7,201,645 (contoured CMP pad dresser), all Chien-Min Sung/Kinik-family. Kinik's "DiaGrid" CMP conditioners were the accused products in ITC Inv. No. 337-TA-449 (3M v. Kinik), where the '198-adjacent brazed-diamond DiaGrid process was examined in detail (see the Commission opinion findings at usitc.gov pub3530.pdf, FF 136–184). Separately, Chien-Min Sung v. Kinik Co., C.A. 1:20-cv-00247 (D. Del.), concerns the '802 patent and ownership/license issues — not the '198 patent. I found no assertion of the '198 patent itself in any 2026 CAFC matter.

I'll search for litigation involving US patent 6368198 across patent litigation databases.

Let me search more specifically for cases asserting this patent.

Let me look specifically for cases asserting the '198 patent.

Litigation Search: US Patent No. 6,368,198

Bottom line: I could not identify any litigation in which US 6,368,198 ("Diamond grid CMP pad dresser") was asserted as a patent‑in‑suit. The number 6,368,198 appears in the public record primarily as (a) a cited prior‑art reference/patent‑family member in other patents, and (b) the patent Kinik identifies as the basis for its commercial DiaShield® DLC coating. I found no docket entry or case summary naming it as an asserted patent. Below I separate that finding from the several related-but-distinct disputes involving the same parties and technology, and I flag where my search was inconclusive.

The patent itself (for identification)

• US 6,368,198 B1; application 09/558,582; inventors Chien‑Min Sung and Frank Lin; assignee Kinik Company; filed 2000‑04‑26; granted 2002‑04‑09; priority 1999‑11‑22 as a continuation‑in‑part of 09/447,620. Status: Expired – Lifetime (anticipated expiration 2019‑11‑22). Source: https://patents.google.com/patent/US6368198/en

Because the '198 patent expired on 2019‑11‑22, any U.S. infringement action asserting it would have had to be filed before that date, and damages would be limited to pre‑expiration conduct.

Where the '198 number does appear

• Cited/listed as a reference in other patent documents, e.g., the family listing in WO2012006281A3 and the specification citations in CN1867428A ("美国专利……6368198B1"), i.e., prior‑art/family citations, not litigation.
• Cited in Kinik technical literature as the patent covering Kinik's "DiaShield®" DLC barrier coating (Sung & Lin, US Patent 6,368,198), e.g., https://www.scientific.net/MSF.534-536.1133 and the Mendeley record of the same paper. This is a commercial‑product identification, not a lawsuit.

Related litigation in the same technology/party space — different patents

These are frequently confused with the '198, but each involves a different patent number. I am not auto-correcting these numbers; I am reporting them literally and noting they are distinct from 6,368,198.

Sources for the IPR/related‑matters list: Kinik's IPR2021‑00638 petition (real party‑in‑interest and related‑matters sections), https://www.docketalarm.com/cases/PTAB/IPR2021-00638/Kinik_Company_v._Sung_Chien-Min/docs/03-10-2021-Petitioner/Petition-1-... ; IP Verse Kinik PTAB case list, https://ipverse.greyb.com/competitive-analysis/company/kinik

Important caveats and open items

1. No confirmed '198 assertion. Despite targeted searches, I found no complaint, ITC filing, PTAB proceeding, or docket in which claim(s) of US 6,368,198 were asserted. If litigation exists, it is not surfacing in the public aggregators I could reach.
2. The "3M Co. v. Kinik Co." (D. Minn.) summary is ambiguous. The Studicata summary refers to the "'198 and '498 Patents," which could correspond to 6,368,198 — but the summary is a secondary/derived source and may be abbreviating different numbers. I could not verify this against the underlying complaint, and I am not asserting that 6,368,198 was one of those three patents. Treat this as unconfirmed.
3. The Sung v. Kinik D. Del. 2014 case (three patents) is the most likely candidate if the '198 was ever asserted, given the ownership dispute over JV‑generated patents, but the dockets I retrieved do not list the asserted patent numbers.
4. Search limitations: I was unable to complete exhaustive queries against paid litigation databases (PACER/Unified Patents/RPX Insight) within this session. I hit a tool‑call limit before retrieving a full PACER docket or an RPX Insight litigation page for 6,368,198.

Recommended verification steps

• Run the patent number directly on RPX Insight (insight.rpxcorp.com) and Unified Patents Portal (https://portal.unifiedpatents.com/litigation) — both index asserted US patent numbers in district court and ITC complaints.
• Pull the D. Del. 1:14‑cv‑01027‑RGA complaint via PACER to read the three asserted patent numbers.
• Check the ITC EDIS record for any § 337 complaint (beyond 337‑TA‑449) listing 6,368,198.

If no litigation is known: Based on the searches performed, the accurate statement is that no litigation asserting US Patent No. 6,368,198 is known to me, and the patent is now expired (2019‑11‑22), which materially reduces the likelihood of any assertion.

I'll start by verifying the canonical "no proceedings" default against the public record.

Let me run a couple more targeted checks before settling on the "no proceedings" default.

One final check for any Kinik-initiated or Kinik-targeted AIA trials tied to this patent.

Proceedings overview

Total AIA trial proceedings on US 6,368,198: 0. Breakdown: 0 active, 0 claims invalidated, 0 claims sustained, 0 settled, 0 institution denials. This is the canonical list from the USPTO Open Data Portal ("The USPTO ODP API returns no AIA trial proceedings for this patent as of the most recent ingest"), and my independent searches surfaced nothing that the ODP missed — no IPR, PGR, or CBM on this patent exists on the public record. The defensive posture this gives a defendant is therefore not an IPR-estoppel story; it is an expiration story: the record shows the patent as "Expired - Lifetime" with an anticipated expiration of 2019-11-22, so the practical defense is that there is no live patent to infringe — not that some panel already killed the claims.

No proceedings to itemize — and what that does (and does not) mean

Because the count is zero, the per-proceeding template has no entries. Two corroborating data points and one caveat:

• Google Patents' PTAB tab for this patent (the structured source used here) lists no PTAB proceedings: https://patents.google.com/patent/US6368198/en
• The PTAB E2E / Patent Trial and Appeal Board search portal returns no trial for patent number 6,368,198: https://ptab.uspto.gov/
• Near-misses that are not this patent (do not conflate them — petitioners win cases by getting this wrong):
  • IPR2014-01523, Kinik Company v. Chien-Min Sung — directed to US 8,777,699 (a sibling Sung CMP-dresser patent), not the '198. Petitioner's demonstratives are dated 2015-09-16: https://www.docketalarm.com/cases/PTAB/IPR2014-01523/Inter_Partes_Review_of_U.S._Pat._8777699/
  • IPR2021-00638, Kinik Company v. Chien-Min Sung — directed to US 9,724,802; filed 2021-03-10, status Terminated-Settled (decision date 2021-08-27); panel included APJs Jason W. Melvin and Phillip J. Kauffman: https://www.law360.com/ptab_cases/6049f68dd63686005b4a0e4d
  • IPR2024-00535, Samsung Austin Semiconductor LLC — appears to concern the same '802 patent (the docket includes Dr. Sung's declaration on the '802): https://www.docketalarm.com/cases/PTAB/IPR2024-00535/Samsung_Austin_Semiconductor_LLC/
  • EP counterpart, not PTAB: EP1151825B1 ("A diamond grid CMP pad dresser"), granted 2006-06-28, with DE60029089D1 and ATE331590T1 in the same family.
  • Foreign invalidity challenge, not PTAB (verify before relying): Korean Patent Court decision 2005허179 (등록무효 / registration invalidation), decided 2007-03-28, lists Kinik Company as plaintiff against Sae Sol Diamond, Ehwa Diamond, and Shinhan Diamond, and dismisses the plaintiff's claim. The searching tool returned only a partial text; the party roles and the identity of the Korean patent are ambiguous in that snippet, so treat this as a lead, not a finding.

Strategic summary

Claim status. All 24 claims of US 6,368,198 are UNTESTED before the PTAB: nothing canceled, nothing sustained, nothing disclaimed. Claim 1 (the template/braze-sheet method), claim 19 (uniform height within 50 µm), and claim 20 (anti-corrosive coating over particles and braze) are all intact as issued — but "intact" is doing less work than usual, because the patent's term has run. The structured record lists: application 09/558,582 filed 2000-04-26; priority 1999-11-22 (CIP of 09/447,620); granted 2002-04-09; "Expired - Lifetime," anticipated expiration 2019-11-22. One record inconsistency worth flagging literally: the family table in the same source lists a 1997-04-04 priority (via US 08/835,117 / US 6,039,641), which would put expiration even earlier. Either way, the patent is expired.

Estoppel landscape. There is no § 315(e)(2) estoppel attached to this patent, because no petitioner ever filed an IPR or PGR against it. That cuts both ways for a defendant: (a) you are not blocked from any prior-art ground, and (b) you also get no free ride from an earlier petitioner's work product. The numerous IPRs in the Sung/Kinik family (IPR2014-01523, IPR2021-00638) estop only their petitioners, and only as to the sibling patents they challenged ('699 and '802). Also note the mechanics if you were contemplating an AIA filing on the '198 itself: this is a pre-AIA patent, so PGR is unavailable (§ 321 applies to first-inventor-to-file patents), and the CBM program has sunset (no new CBM petitions after 2020-09-16). An IPR against an expired patent is technically permissible, but there is no § 315(b) one-year clock running because no complaint asserting the '198 can create new infringement.

Pattern signals. The pattern is Sung-family, not '198-specific. Dr. Chien-Min Sung and Kinik were joint-venture partners from roughly 1996–2012, then became adverse: Kinik filed IPRs against Sung's '699 and '802 patents, IPR2021-00638 terminated in settlement, and Samsung later entered as a petitioner on the '802. The '198 is a Kinik-owned patent (original assignee Kinik Co.; current assignee Kinik Company per the record), with Sung and Frank Lin as inventors — so it sits on the opposite side of that family feud from the patents that actually drew IPRs. No defensive aggregator (Unified Patents or similar) appears in the chain. I found no evidence that the '198 was itself asserted in a U.S. district court or ITC action; the ITC and Federal Circuit activity I located (Inv. No. 337-TA-449, Kinik Co. v. ITC, 362 F.3d 1359 (Fed. Cir. 2004)) concerned 3M/UAS's US 5,620,489 — a different patent, asserted against Kinik.

Recommended next steps

1. If you are facing a demand or suit on the '198, lead with expiration, not invalidity. The record shows expiration 2019-11-22. A complaint filed today (2026-09-29) can only reach infringement occurring after 2020-09-29 under the six-year lookback of 35 U.S.C. § 286 — a window in which, on the record's dates, no infringement is possible. Pull the front-page term calculation and any § 154(b) adjustment from the patent's own file wrapper before you brief it, and confirm the expiration date directly with Patent Center rather than relying on a third-party database.
2. If you want an invalidity adjudication anyway (e.g., a declaratory-judgment posture or a defensive publication strategy), the only Office route left is ex parte reexamination under 35 U.S.C. §§ 301–307, which remains available for expired patents and permits an anonymous requester. There is no PGR (pre-AIA) and no new CBM.
3. Document the absence of PTAB activity before representing it to a court or an adversary: screenshot the Google Patents PTAB tab at https://patents.google.com/patent/US6368198/en and the PTAB E2E search at https://ptab.uspto.gov/, and note the search date. The absence is itself a data point — well-asserted U.S. patents in this art field did attract IPRs (the sibling '699 and '802 patents), and this one did not.
4. Watch the sibling patents, not this one. If a counterparty is asserting CMP-dresser patents against you, the live risk is in the Sung/Kinik family (e.g., US 9,724,802, US 8,777,699) and in the continuation US 6,884,155 ("Diamond grid CMP pad dresser") and US 7,204,645 ("Contoured CMP pad dresser") — all expired or near-expired, but each with its own claim set. Verify each one's expiration and PTAB history individually rather than assuming the '198's clean record carries over.
5. If you need the Korean lead resolved (case 2005허179), obtain the full Korean-language opinion and confirm which patent was at issue and Kinik's party role before citing it anywhere.

Confidence note: I am highly confident there are zero AIA trial proceedings on US 6,368,198 — this rests on the canonical ODP block in the prompt plus the corroborating searches above. I am not asserting any finding about claim outcomes, judge panels, or FWDs, because none exist to report. The Korean proceeding and the sibling-patent outcomes are flagged as unverified leads.

I'll research the assignment record for US 6368198. Let me pull from multiple sources.

Let me query the USPTO assignment API and related sources directly.

Let me try the USPTO assignment API endpoint and check for any later transfers off Kinik.

Assignment Analysis — US 6,368,198 B1 "Diamond grid CMP pad dresser"

Analyst caveat up front: I was able to retrieve the recorded-assignment event data for this patent, but not the reel/frame numbers or the correspondent-of-record fields. Those fields are exposed only in the USPTO Assignment Center detail view / assignment API, which I could not reach within this session. I have left them explicitly marked as unverified rather than reconstructed. The verification page for the fields I could not pull is https://assignment.uspto.gov/patent/index.html (search by patent number 6368198).

---

Inventors

Unusual-pattern check:

• No evidence of inventor flight. Both inventors are on the single recorded assignment to Kinik (recorded 2000-07-24), and Sung remained Kinik's diamond-technology principal for at least ~a decade after filing (Kinik Diamond Technology Center papers through 2009). The "all inventors depart within 12 months" fire-sale precursor is not present.
• Prolific-inventor flag on Sung. Sung is a very high-volume inventor whose name later appears as inventor/assignor on a large family of abrasives patents (the ~18 priority-linked applications listed on the Google Patents page, e.g. US 6,884,155; US 7,124,753; US 7,201,645). One sibling in that family line — US 8,777,699 B2 ("Superabrasive tools having substantially leveled particle tips") — is assigned to Ritedia Corporation, a Sung-linked holding entity. That matters for portfolio-level NPE screening of the family, but Ritedia does not appear anywhere in the chain of US 6,368,198 and I will not import it into this chain.
• CIP structure worth noting for chain tracing. This patent is a continuation-in-part of US 09/447,620 (filed 1999-11-22), which is itself a CIP of US 08/835,117 (filed 1997-04-04, issued as US 6,039,641, "Brazed diamond tools by infiltration," Sung). The 1997-04-04 priority date propagates into ~18 family members. Chain analysis of this patent must not be confused with chain analysis of the 1997-04-04 family.

---

Original assignee

Kinik Company (Chinese: 中國砂輪企業股份有限公司; "China Grinding Wheel"), Ying-Kuo, Taipei Hsien, Taiwan. Listed on the issued patent as assignee (Google Patents shows current assignee as "KINIK / Kinik Co").

• Primary line of business: manufacturer of bonded abrasives, diamond tools, and semiconductor consumables — CMP pad conditioners ("diamond disks"), dicing blades, backside grinding wheels, and bonding tools. Founded in Yingge, Taiwan; over 70 years in grinding-wheel manufacturing; publicly listed on the Taiwan Stock Exchange (TWSE: 1560).
• Did it ship a product embodying the claims? Yes — strongly documented. Kinik's own product literature describes the diamond-grid CMP dresser line (DiaGrid®, later I-DiaGrid®, S-DiaGrid®, equaDia®, Pyradia®, O-Pyradia®) as commercial product. Kinik's CMP Diamond Disk page describes exactly the claimed advantage: "A dresser that contains a diamond grid is much more efficient to dress the pad than conventional disks with randomly distributed grits… both the dresser life and pad life can be doubled." More directly, a 2005 Kinik paper states: "Kinik Company pioneered diamond pad conditioners protected by DLC barrier (DiaShield® Coating) back in 1999 (Sung & Lin, US Patent 6,368,198)" — an express corporate admission that claims 20–24 (the anti-corrosive / DLC coating claims) were commercialized as DiaShield®.
• Current status: Operating. No bankruptcy, no acquisition of Kinik by a third party, no dissolution surfaced. Corporate disclosures are filed with Taiwan's Market Observation Post System (MOPS), not the SEC (so there is no 10-K/8-K to cross-reference; the Taiwan MOPS filings serve the same function).
• Patent status: Expired — Lifetime. Anticipated expiration 2019-11-22. As of today (2026-09-29) this patent has been unenforceable for nearly seven years. This is dispositive for any NPE-assertion analysis: no entity can currently sue on it.

---

Assignment timeline

The retrieved sources expose exactly one recorded conveyance. Google Patents' "Legal Events" for this file shows a single reassignment entry dated 2000-07-24 (Google's label: "reassignment"), reading "KINIK — ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: LIN, FRANK, SUNG, CHIEN-MIN." No subsequent assignment, security agreement, merger, change of name, or license is recorded.

• Executed: not exposed in retrieved data (recorded 2000-07-24; the natural execution date is on or about the 2000-04-26 filing date, but I will not assert that as fact) / recorded 2000-07-24 — Reel NOT RETRIEVED / Frame NOT RETRIEVED
  • Conveyance: Assignment of Assignors' Interest (see document for details)
  • Assignor: Chien-Min Sung; Frank Lin (individually)
  • Assignee: KINIK (Kinik Company), 64 Chung-San Rd., Ying-Kuo, Taipei Hsien, Taiwan
  • Correspondent: NOT RETRIEVED. I could not reach the Assignment Center detail view, and the correspondent-of-record is not mirrored in Google Patents' legal-events feed. Because I cannot cite a name, I make no repeat-correspondent finding for this chain. (Flag for follow-up: this is the single highest-value field still outstanding, since it would let us tie this recording to Kinik's other filings and to Sung-family recordings.)
  • Context: Standard founder/inventor-to-employer assignment at filing — a new-employee/inventor rights transfer, not an acquisition, fire-sale, or reorg.

Post-issuance assignment activity: none recorded (2002–2026). The patent was granted 2002-04-09 and no transfer out of Kinik was ever recorded — consistent with Kinik's stated current-assignee status. In particular, there is no assignment of this patent to Ritedia Corporation, despite Ritedia holding other Sung-family patents.

If the Assignment Center detail view shows any additional entry (e.g. an internal Kinik Co → Kinik Company change of name, or a security agreement against the Kinik portfolio that names this property), that would be an incremental finding I have not been able to confirm and should not be assumed.

---

Timeline diagram

timeline
    title Ownership of US 6368198
    1999 : Inventors file parent application
    2000 : Application filed by Kinik Co
         : Sung and Lin assign rights to Kinik
    2002 : Patent issued to Kinik Co
    2019 : Patent expires
         : Kinik Co remains owner of record

---

NPE / troll-pattern signals

Net signal count: zero strong, zero weak. Four categories non-assessable/negative, one unassessable for lack of a retrieved field.

---

Verdict

Insufficient data — in the taxonomy's literal sense that only the original assignment is of record.

Justification: The Assignment Center-derived record for US 6,368,198 contains a single conveyance — the 2000-07-24 recording of the inventors' interest assignment from Chien-Min Sung and Frank Lin to KINIK (Kinik Company) — with no post-issuance assignment, security agreement, or change of name recorded between issuance on 2002-04-09 and expiry on 2019-11-22. Every substantive NPE indicator is affirmatively negative: the sole assignee is a TWSE-listed operating abrasives manufacturer that shipped the claimed product (Kinik's own 2005 paper expressly credits "Sung & Lin, US Patent 6,368,198" as the basis of its DiaShield® DLC-coated conditioner line), so the shell-entity, known-asserter, cascading-transfer, fire-sale, privateering, and pre-litigation-transfer signals are all absent, and the patent is now expired and unassertable. The verdict is "insufficient data" only because the record is trivial, not because it is missing a link — the two genuinely outstanding data points are the reel/frame and the correspondent of record for the 2000-07-24 recording.

Verify at: https://assignment.uspto.gov/patent/index.html — search patent number 6368198 (recorded 2000-07-24; assignee KINIK; assignors Sung, Chien-Min and Lin, Frank). Re-run at https://assignmentcenter.uspto.gov/ to capture the reel/frame and correspondent fields this session could not retrieve.

Supporting sources

• Google Patents legal events / bibliographic data for US 6,368,198: https://patents.google.com/patent/US6368198/en
• Kinik CMP Diamond Disk product page (DiaGrid diamond-grid dresser claims and benefits): https://www.kinik.com.tw/en-us/Products/CMPDiamondDisk.html
• Kinik tech paper expressly crediting Sung & Lin, US 6,368,198 for the DiaShield® DLC barrier (NSTI-Nanotech 2005): https://briefs.techconnect.org/wp-content/volumes/Nanotech2005v2/pdf/511.pdf
• Kinik investor/corporate disclosures (Taiwan MOPS, substituting for SEC filings): https://mopsov.twse.com.tw/nas/STR/156020260811M001.pdf
• EP family member (same chain, Kinik): EP 1 151 825 B1

Residual uncertainty I will not paper over: (i) the reel/frame is unknown to me; (ii) the correspondent is unknown to me; (iii) the execution date of the 2000-07-24 recording is unknown to me; (iv) I did not run a docket-level litigation search, so "no litigation" rests on the absence of any suit in the sources reviewed rather than on a docket query. None of these gaps, on the evidence available, would change the verdict — a chain that never leaves a product-shipping operating manufacturer and that expired in 2019 is not an NPE chain.

I'll verify the patent record and the cited references against live sources before analyzing anticipation.

Let me verify the remaining cited references.

Prior-Art Analysis — US 6,368,198 B1

Verification note. A targeted search confirms the record: US 6,368,198 B1, "Diamond grid CMP pad dresser," app. 09/558,582, Kinik Co. (inventors Chien-Min Sung; Frank Lin), priority 1999-11-22, filed 2000-04-26, issued 2002-04-09 (https://patents.google.com/patent/US6368198/en; Justia mirror https://patents.justia.com/patent/[6368198](/patent/6368198)). Per the strict-identifier rule, I confirm the only USPTO grant matching "6368198" is US 6,368,198. The 7-digit string 6368198 is also a JP engine-start-controller grant — unrelated, not substituted.

Search limitation (stated up front). I could not open a live USPTO PatentCenter/PAIR prosecution record from this environment, and I exhausted my search budget before independently verifying every Japanese-language citation. The reference list below is taken from the patent's own face ("Patent Citations (16)") and from Google Patents' citation data. Where I retrieved the reference text, I say so; where I did not, I flag it as unverified rather than invent a description.

Legal framework used. The '198 patent has a pre-AIA effective date (priority 1999-11-22; filed 2000-04-26). The relevant categories are §102(a)/(b) (knowledge/use/printed publication, and the one-year bar keyed to ~1999-04-26) and §102(e) (US patents/applications as of their filing dates). All 16 cited references predate the 1999-11-22 priority date, so each is presumptively available as prior art. None of them, on the record I can see, discloses every element of any independent claim (1, 19, 20) — so I characterize them by strength, not as clean anticipations.

The independent claims to test against are: claim 1 (template-on-braze-sheet fabrication: lay apertured template on a sheet of braze alloy → fill → remove excess → press in → lift template → attach sheet to substrate); claim 19 (uniform spacing + tip height leveled to within ~50 µm); claim 20 (uniform spacing + uniform height + anti-corrosive coating over particles and braze).

---

The 16 examiner-cited references, in citation order

---

Family/relatives citations (context, generally not §102 art against the '198)

Google Patents lists eight "Family Cites Families" — references cited in the other family members, some of which are same-inventor family (excluded from §103 by pre-AIA §103(c), and not §102 art against the same inventor's earlier disclosure):

• US 4,078,906 (Elgin Diamond, 1976) — abrading tool with discontinuous diamond surfaces.
• US 4,925,457 / US 5,049,165 (Ultimate Abrasive Systems) — method of making an abrasive tool; composite material.
• GB 9309972 D0 (De Beers, 1993) — tool insert.
• JP 2896657 B2 (Asahi Diamond, 1996-06-28) — "Dresser and manufacturing method thereof" (content unverified; like JP H11-77536, a method-of-making-a-dresser candidate worth pulling).
• US 6,039,641 (Sung, 1997-04-04) — "Brazed diamond tools by infiltration" — the '198's family parent, not hostile art.
• TW 394723 B (Sung, 1997-04-04) — "Abrasive tools with patterned grit distribution and method of manufacture" — same inventor/family; describes the patterned-grit concept underlying the '198 but is not third-party art.
• US 6,051,495 (AMD, 1997-10-31) — seasoning a wafer polishing pad to polish tungsten.

---

Bottom line

1. No cited reference appears to anticipate any independent claim outright. Claims 1, 19, and 20 each require a combination (template-on-flexible-braze-sheet preform; ~50 µm tip leveling; or a coated brazed dresser) that the single references supply only in fragments.
2. The two strongest examiner references are:

• US 6,190,240 (Nippon Steel) — uniform, spaced, partially embedded grains in a braze/joining alloy on stainless steel, with spacing ≈0.1–10× grain size. It is the best §102 candidate against the substance of claim 19 and against several dependent claims (13–14, 17), but it lacks the template/braze-sheet route of claim 1.
• US 5,921,856 (Sp3/Zimmer) — uniformly distributed diamond grit plus a diamond overcoat sealing grit and bond; best candidate against the coating aspect of claim 20/21, though the process (HFCVD) and absence of braze distinguish it.

1. Two unverified references deserve manual retrieval before reliance: JP H11-77536 (Asahi Diamond) and JP 2896657 B2 (Asahi Diamond) — both are method-of-manufacturing-a-dresser disclosures and are the only cited items whose titles plausibly map onto claim 1's fabrication sequence. I will not characterize them as anticipatory on title alone (per the "say so explicitly rather than fabricate" rule). US 6,106,382 (3M) is a third unverified item.
2. Everything else (items 1–4, 6, 8–12, 14, 16) is background: process/apparatus art or non-abrasive preconditioners that do not disclose the claimed dresser structure or fabrication method. Item 2 in particular teaches away from abrasive particles.

If you want, I can do a focused follow-up pull on the three unverified references (Asahi Diamond JP H11-77536 and JP 2896657 B2; 3M US 6,106,382) plus the examiner's actual §102/§103 rejections from the '198 file history, which would pin down whether any rejection was ever maintained against these claims.

I'll ground the analysis in the actual cited prior art. Let me verify what the load-bearing references disclose.

Let me verify the remaining load-bearing citations before drawing conclusions.

Caveats before the analysis

Date inconsistency (flagging, not correcting): the task header sets "Current Date: April 26, 2026," while my operating date is 2026-09-29. The prior section also cites a CAFC opinion dated June 18, 2026 — a date that is future as of April 26, 2026. I cannot resolve which clock is authoritative and will not silently harmonize them. Nothing in this §103 analysis turns on the date.

Scope: I use only the references the page itself lists as prior art — the 16 items under Patent Citations and the 8 under Family Cites Families. I exclude everything under Cited By (those post-date the '198 priority/filing and are not citable against it) and I specifically flag one such entry below because it looks tempting. I could not retrieve full text for US 6,106,382, US 6,213,856, JP 2896657 B2, or JP H1177536 A; statements about those four are limited to their titles/dates as printed on the page.

---

1. Bottom line

The '198 is in serious §103 trouble, and claim 1 in particular sits close to the line between obviousness and outright anticipation. The reason is not any single CMP-dresser reference; it is that the abrasive-tool art listed on this very page (US 4,925,457 / US 5,049,165 / US 4,078,906) already taught template/mesh/screen placement of uniformly spaced diamond that is then fixed in a matrix with the grit protruding, and that art is plainly analogous to a CMP pad dresser. The CMP-dresser art on the page (US 5,921,856, US 6,213,856, US 6,190,240, US 6,106,382, JP 2896657, JP H1177536) supplies the motivation and the coating limitation. A PHOSITA combining them would arrive at every limitation of claims 1–24.

The strongest non-obviousness hooks are narrow: the specific flexible braze-sheet preform with grit pre-embedded before brazing, and the template thickness as the height-control mechanism. Both are themselves suggested by the cited art, so they are weak hooks.

---

2. The critical-date gate — it determines which references exist

The '198 is pre-AIA (filed 2000-04-26). Its printed priority is 1999-11-22 (CIP of US 09/447,620), but the page's own tables also tie it to 1997-04-04 through US 08/835,117 / US 6,039,641. The prior section already flagged this inconsistency as unresolved; I will not resolve it either, but the two scenarios produce different art sets:

Important: even under the applicant's most favorable date, the Ultimate Abrasive and Elgin art remains available. The obviousness case does not depend on Scenario A. Under Scenario A it becomes overwhelming.

One trap: TW 394723 B is the reference printed on the page, but its grant date is 2000-06-21 — after the '198's 2000-04-26 filing. It is therefore not citable as a printed publication by its own grant date under Scenario A. The citable document is its PCT sibling, WO 98/45092 A1, "Abrasive tools with patterned grit distribution and method of manufacture," published 1998-10-15. Any office action or petition that cites "TW 394723" as the primary reference should be re-papered to WO 98/45092. I flag this because it is the kind of error that loses an IPR on the pleadings.

---

3. Prior-art inventory (as printed on the page)

3.1 Abrasive-tool art — the core of the §103 case

3.2 CMP-dresser/conditioner art — supplies motivation and the coating

Reference to exclude: US 6,517,424 B2, "Protective coatings for CMP conditioning disk" (Abrasive Technology), appears in the Cited By list with priority 2000-03-10 — after the '198's 1999-11-22 priority. It is not prior art to the '198 and must not be used. It is nonetheless a useful signal: a competitor patented essentially the "protective coating on a conditioning disk" concept shortly after the '198, which bears on §102/§103 (it is not anticipatory of the '198) but also on secondary-considerations nexus arguments.

---

4. Claim 1 — the strongest §103 combinations

Claim 1 steps: (a) substrate; (b) template with predetermined aperture pattern on a sheet of braze alloy; (c) fill apertures with abrasive particles; (d) remove excess; (e) press particles into the sheet so they are partially embedded; (f) remove template, particles remain in place; (g) attach the grit-bearing sheet to the substrate.

Combination 1 — US 4,925,457 (primary) + US 6,039,641 (secondary)

• (a) substrate: US 6,039,641's matrix support material; US 4,925,457's "carrier" that can be shaped to conform to substrates; US 4,078,906's metal disc body.
• (b) template/mesh with predetermined apertures: US 4,925,457 literally — "placing them in the openings of a mesh and the mesh is then removed," where the spacing is "uniform." US 4,078,906 independently confirms screens/sieves used "to hold particles of diamonds in position."
• (c)–(d) fill apertures, remove excess: US 4,925,457's mesh-and-openings placement; US 4,078,906's screen with pattern openings plus removal of non-adhered particles.
• (e) press into a matrix and partially embed: US 4,925,457's "embedding the particles in the carrier, and fixing the particles to the carrier with the particles protruding from the carrier." The Ultimate Abrasive family elsewhere teaches applying "physical force" to orient/seat the particles.
• (f) remove template, particles remain: US 4,925,457 — "the mesh is then removed."
• (g) attach the carrier to a substrate: US 4,925,457 — flexible carrier "shaped to conform to substrates of complex shapes"; US 6,039,641 — braze bonded to a matrix support material.
• Braze chemistry: US 6,039,641 (Cr/Mn/Si/Al carbide-former braze) makes the "braze alloy" limitation a routine selection.

Motivation (KSR / MPEP 2143): both references are in the same field of endeavor (superabrasive tools) and address the same deficiency — non-uniform grit distribution causing redundant particles, debris blockage, and premature failure. The Ultimate Abrasive family states the benefit explicitly: "spaces between the abrasive areas allow for the removal of debris and the inflow of coolant," and uniform spacing "optimiz[es] the stock removal rate and the quality of the machined surface." That is precisely the '198's stated problem (pad glazing; uneven grooming; over/under-dressed zones). A PHOSITA adapting a known uniform-pattern placement method to a known brazed CMP conditioner does so with a reasonable expectation of success.

Combination 2 — WO 98/45092 (Sung) + US 6,039,641 (Sung)

Same inventor's own prior publications disclose (i) abrasive grains in regular arrangement "planted by a template tool, which has holes in specified pattern" (WO 98/45092) and (ii) diamond "disposed in the matrix support material in a predetermined pattern" bonded by a 2–50% Cr/Mn/Si/Al braze (US 6,039,641, claim 1). The only meaningful gap between that pair and claim 1 is the physical form of the braze as a rolled flexible sheet rather than loose powder. That is a design choice, and the '198's own claim 15/16 recites rolling/extruding/tape-casting with an organic binder — a conventional braze-preform practice ("brazing tape"). Under Scenario A, this combination is the most dangerous one, and it is self-inflicted: it is the applicant's own §102(b) publication.

Combination 3 — US 4,078,906 (silk-screen) + US 6,039,641 (braze)

US 4,078,906 teaches pattern application through a screen onto a disc, nonselective sprinkling of abrasive, and selective fixing; US 6,039,641 supplies a braze that chemically bonds diamond. The stated advantage in US 4,078,906 is "an economy of valuable diamond particles and increased cutting action by provision of a discontinuous pattern" — the same economy argument the '198 makes for a grid.

Anticipation caution on claim 1

US 4,925,457's abstract is close to a limitation-by-limitation reading of claim 1 if its "carrier" is read as the braze sheet (the family describes a "preform consisting of unsintered, partially sintered or completely sintered metal powder"). If the carrier is a braze/metal-powder preform, claim 1 is arguably anticipated, not merely obvious. I would not rely on that reading without the full US 4,925,457 specification, which I have not reviewed.

---

5. Claim 19 — uniform spacing + tips within ~50 µm

Elements: (a) substrate; (b) uniformly space particles on the surface; (c) affix so each extends within about 50 µm of a predetermined uniform height.

• (b): US 4,925,457 ("uniformly spacing"); US 5,921,856 ("mono-layer of diamond grit substantially uniformly distributed," grain size within ±20% on a given disk).
• (c): The flat-plate press of the '198 is a mechanism for leveling tips; the prior art teaches the equivalent height-control concepts. US 5,921,856 constrains grit density and states grain size is matched to about ±20%; the Ultimate Abrasive family teaches applying physical force to orient particles uniformly before fixing. The stencil-brazing art that surfaced in the search results (an AU publication in the stencil/brazing-paste family; assignee not verified) states that "the thickness of the stencil sheet will determine the height of the abrasive parcels" — i.e., the template-thickness-as-height-control idea the '198 relies on.
• Numeric limit: "within about 50 micrometers" is a result-effective range with no criticality demonstrated. Routine optimization of a disclosed working range is obvious absent unexpected results — In re Aller, 220 F.2d 454 (CCPA 1955); In re Boesch.

Motivation: the CMP-dresser references identify the exact harm the leveled-tip feature addresses. US 5,921,856's discussion of the ideal conditioner — remove only the outer cell layer, maximize pad life, avoid wafer scratching — supplies the incentive to control tip height; a PHOSITA would level the grit to avoid the deep-penetrating "killer asperities" that over-dress the pad.

---

6. Claims 20–24 — uniform spacing + uniform height + anti-corrosive coating

This is the §103 combination the page supports most cleanly.

• Claim 20 (coating over grit and braze): US 5,921,856 discloses a conditioning head with uniformly distributed grit encased in and bonded by an outer CVD diamond layer grown over the grit-covered substrate. The '198's step (d) is the same step with a braze substrate underneath. The motivation is not merely generic — US 5,921,856 states it verbatim: the nickel plating "react[s] with the [metal] slurry… dissolve[s]… and cause[s] a major loss of the diamond grit from the plate, potentially scratching the wafers." That is the '198's stated corrosion/dislodgement problem. Both references therefore address the same defect, and the combination yields the predictable benefit the '198 claims (in-situ dressing in acid slurry).
• Claim 21 (diamond or DLC): US 5,921,856 teaches diamond; US 5,246,884 (IBM) teaches diamond or diamond-like carbon films in a CMP environment. Selection between the two is a nominal choice of known CMP-facing materials — KSR, 550 U.S. at 417 (predictable variation).
• Claim 22 (< ~3 µm): thickness optimization; the '198 itself presents the range (0.5–5 µm, preferably 1–3 µm) without criticality. Routine optimization.
• Claim 23 (DLC atomic carbon ≥ ~95%): composition optimization; no evidence of criticality. The '198's own specification states 90–95%, so the claimed 95% floor is a routine narrowing.
• Claim 24 (cathodic arc): the '198's own specification admits "PVD methods such as the use of a cathodic arc with a graphite cathode, are known in the art." An applicant's admission that a technique is known is itself strong evidence of obviousness.

Substitution argument to watch: US 5,921,856 uses CVD diamond, thick (10–250 µm; at least 10% of grit size); the '198 uses PVD DLC, thin (0.5–5 µm). A PHOSITA recognizing why CVD diamond protects the grit (chemical barrier, encasement) would be motivated to substitute a thinner, cheaper, lower-temperature PVD/DLC coating to preserve tip geometry while retaining corrosion protection — especially since the '198's own examples show 1 µm DLC by cathodic arc surviving >30 h in pH 3.0 slurry. Because claim 21 expressly claims DLC as an alternative to diamond, the substitution is claimed, not avoided.

---

7. Dependent claims 2–18

---

8. Secondary considerations — and the nexus problem

The '198's only objective evidence is internal: Table 1 (grid disks outperform a random-placement disk; the 135 µm disk ≈ 2×) and Table 2 (>30 h vs. ~1 h in pH 3.0 acid slurry, in-situ dressing).

• The 2× result is attributed to uniform spacing. But uniform spacing is the very thing US 4,925,457 and US 4,078,906 already taught, and the inventor's own published work (WO 98/45092; the 2006 powder-metallurgy paper) states that patterned distribution "may double the tool life and the cutting speed." Objective evidence that is coextensive with the prior-art teaching does not rebut obviousness.
• The 30 h vs. 1 h result is the more impressive datum, and it is the coating's doing. But US 5,921,856 already identified the identical failure mode (nickel/metal-bond dissolution in metal CMP slurry → grit loss → wafer scratching) and prescribed a diamond coating to solve it. The nexus therefore runs to a known solution, not to the claimed invention as a whole (In re Kao; Orme nexus analysis).

If I were defending the '198, I would press (i) that the combination of pre-embedded particle-bearing braze sheet + template-thickness-controlled tip leveling + dense PVD DLC seal produces a synergy not shown in any single reference, and (ii) unexpected results in high-acid in-situ dressing. That is a genuinely arguable position for claims that require the braze-sheet preform (claims 1, 15–18), but it is weak for claims 19–24, which are essentially outcome claims.

---

9. Where the §103 case is weakest — and what would defeat it

1. The "sheet of braze alloy" as the carrier. No reference on the page expressly discloses a free-standing, handleable braze-alloy sheet into which grit is pressed before brazing. US 4,925,457's "carrier" and US 4,078,906's disc are the ultimate substrates, not consumable preforms. Expect a defense to argue that this substitution solves a problem the references did not face (transporting pre-placed grit to the furnace without dislodging it) — an argument that itself echoes US 4,925,457's own observation that loosely held particles are "difficult to transport before the metal is deposited."
2. Template thickness as the height-setting mechanism. The '198 ties uniform tip height to template thickness. If the accused prior-art combination sets height by grit size or by a flat plate rather than by template thickness, the "predetermined uniform height" limitation is not met in the way claim 19 describes. The stencil-brazing art I located (AU publication, assignee unverified) undercuts this, but I could not confirm its publication date against the '198 priority, so I would not build a chart on it without verification.
3. The four references I could not read (US 6,106,382; US 6,213,856; JP 2896657; JP H1177536) are all CMP-dresser/dresser-manufacture references. Any of them could contain a disclosure that is directly on point — or could, conversely, contain a teaching away. The analysis above is deliberately conservative about them.
4. The GB 9309972 D0 citation is defective on its face. It is an application number, not a publication. It cannot be used as a §102(b) reference without the corresponding published GB specification.

---

10. Confidence and uncertainty flags

• High confidence: the reference identities, dates, and the quoted disclosures for US 4,925,457 (abstract), US 4,078,906 (object statement and background), US 5,921,856 (summary and nickel-degradation rationale), US 6,039,641 (claim 1), and the TW 394723 abstract. These are quoted from the retrieved text.
• Moderate confidence (secondary-source only): the WO 98/45092 publication date of 1998-10-15 and its relationship to TW 394723 — the page and PubChem corroborate each other, but I did not open the WO document itself.
• Not verified: the content of US 6,106,382, US 6,213,856, JP 2896657 B2, JP H1177536 A; the assignee of the AU stencil-brazing document that appeared in the search results; whether US 4,925,457's "carrier" is a metal-powder preform capable of supporting an anticipation (rather than obviousness) reading of claim 1.
• Excluded and why: US 6,517,424 (Abrasive Technology) and all other Cited By entries post-date the '198 priority and cannot be prior art against it. Their appearance in the "protective coatings for CMP conditioning disk" space is evidentiary color only.
• No representation: I did not find, and do not assert, any actual USPTO rejection, IPR, or court holding on the '198's validity. This is an independent Graham/KSR analysis built from the page's prior-art listings.