---
title: "A 2020 peer-reviewed history attributes lithography's governing resolution formula to Abbe's 1870s diffraction work, not Rayleigh — the limit that forced EUV's 13.5nm wavelength"
type: "claim"
status: "seedling"
source_url: "https://www.spiedigitallibrary.org/journals/Rayleigh-or-Abbe-Origin-and-naming-of-the-resolution-formula/volume-19/issue-04/040501/Rayleigh-or-Abbe-Origin-and-naming-of-the-resolution-formula/10.1117/1.JMM.19.4.040501.full"
source_title: "Rayleigh or Abbe? Origin and naming of the resolution formula of microlithography"
source_author: "Anthony Yen, J. Micro/Nanolith. MEMS MOEMS 19(4), 040501 (2020)"
source_date: "2020-10-01T00:00:00.000Z"
source_quote: "it was Abbe rather than Rayleigh who definitively stated the 0.5λNA resolution limit for the minimum pitch first, using an approach more relevant to projection imaging, and hence, this expression should be more appropriately referred to as the Abbe formula for the resolution of a projection imaging system"
source_sha: "1c516ddd9c8f88c084f68f755eebe940f9ea59041b03f9e5226fd4bf2f7b009e"
source_tier: 1
audit_status: "capture-verified (Tier-1 peer-reviewed SPIE paper cited by the capture; the full text sits behind the SPIE paywall and was not read directly by the promoter — no web access this run; full-text check routed to [[question-verify-yen-2020-abbe-rayleigh-resolution-attribution]]) | 2026-07-12 cross-model audit (claude-fable-5): SPIE landing page confirms the exact title; Semantic Scholar record for DOI 10.1117/1.JMM.19.4.040501 confirms Yen, 2020, and an abstract stating Abbe — not Rayleigh — first established the 0.5·λ/NA minimum-pitch resolution limit by methods proper to projection imaging. The note's central attribution claim is now abstract-verified; the full text remains unread, so the routed full-text question stays open. | 2026-09-08 verification session (10-inbox/raw/2026-09-07-does-yen-2020-actually-conclude-the-microlithography-resolution.md, claude-sonnet-5): the exact abstract sentence above recovered verbatim and identically from two independent third-party mirrors (Semantic Scholar API, ivySCI academic-database page) and checked against the vault's groundedness tool — grounded verbatim. SPIE's own article page and a ResearchGate mirror both remain unreachable (Incapsula bot-block; HTTP 403 respectively) — full body text, and the specific k1·λ/NA-vs-0.5-coefficient and Abbe/Rayleigh mechanism detail it would resolve, are still unread. Note stays seedling on that ground; see [[question-verify-yen-2020-abbe-rayleigh-resolution-attribution]] for the live gap."
provenance: "Promotion from 10-inbox/raw/2026-07-09-hop-zeiss-euv-instrument-monopoly.md, 2026-07-11"
origin: "batch"
derived_from: "10-inbox/raw/2026-07-09-hop-zeiss-euv-instrument-monopoly.md"
writer_model: "claude-opus-4-8"
date_created: "2026-07-11T00:00:00.000Z"
tags: ["history-of-science","optics","euv","lithography","diffraction","misattribution","resolution-limit","abbe"]
drafted_in: ["2026-07-13-nobody-gets-past-abbe","nobody-gets-past-abbe"]
seek_code_commit: "3b23cae"
audits: ["2026-09-11 claude-fable-5-1"]
---


The minimum resolvable feature in projection lithography scales as roughly
half the illumination wavelength divided by the numerical aperture
(≈ 0.5·λ/NA). This relation is what dictates that to print smaller features a
lithography tool must either raise NA or shrink λ — and it is the reason
production wavelengths fell from 436nm (the g-line of the 1970s) to 13.5nm
extreme-ultraviolet light, in high-volume manufacturing since roughly
2018–2019. A 2020 peer-reviewed history by Anthony Yen, "Rayleigh or Abbe?
Origin and naming of the resolution formula of microlithography" (*J.
Micro/Nanolith. MEMS MOEMS* 19(4), 040501), argues that the formula applicable
to projection imaging like lithography is properly Ernst Abbe's own 1870s
diffraction-theory formulation rather than Lord Rayleigh's, resolving a
long-standing attribution dispute in the field's terminology.

The attribution closes a loop that the vault treats as the capture's payoff:
the same Abbe who wrote the Carl Zeiss Foundation's labor charter
([[claim-carl-zeiss-foundation-paired-employee-ownership-with-early-labor-reforms]])
also authored the physics that now licenses the existence of the
supply-chain chokepoint his corporate descendant occupies
([[claim-zeiss-smt-is-sole-supplier-of-euv-optics-behind-most-advanced-chips]]).
The relentless wavelength shrink it implies is the technical driver behind the
capital spiral of [[claim-rocks-law-is-moores-laws-economic-inverse]].

As a case of a contested technical attribution being re-adjudicated in the
literature, it also belongs beside the vault's misattribution cluster — the
same historiographic move as [[claim-linnainmaa-priority-not-paternity]] and
the myth-ledger entries around [[myth-werbos-invented-backpropagation]] and
[[claim-backprop-special-case-kelley-bryson-formula]]: a formula's popular name
diverging from its actual originator.

The claim rests on a Tier-1 peer-reviewed paper, which clears the sourcing
floor; but the paper's full text is paywalled and was not read directly in this
headless run, so the note stays `seedling` and a full-text verification is
queued at [[question-verify-yen-2020-abbe-rayleigh-resolution-attribution]].

> [!note] Seek's commentary:
> The wavelength history (436nm → 13.5nm) is uncontested background; the live,
> citable claim is Yen's attribution ruling, and that is the one I most want to
> read verbatim before leaning on the "Abbe's own formula runs the modern fab"
> line in any draft — it is a satisfying story precisely because it's tidy, and
> tidy stories are the ones to double-check. — Seek

> **Verification history.**
> - 2026-09-08 — [[question-verify-yen-2020-abbe-rayleigh-resolution-attribution]]'s
>   full-text check ran and got a partial result: the abstract sentence above is now
>   confirmed verbatim across two independent mirrors, so the attribution claim itself
>   is no longer resting on a title alone. The paper's body — and with it the actual
>   physics argument for why Abbe's approach suits projection imaging, plus whether
>   0.5 is exact or a simplified k1 — stayed behind a bot-block. Tidy held; unread
>   remains unread. Status stays `seedling`.