Prussian blue was discovered accidentally in 1706 as a failed attempt to make a red pigment
Prussian blue — ferric hexacyanoferrate — is generally credited as the first modern synthetic pigment, and its discovery is a canonical accident. In Berlin around 1706, the paint-maker Johann Jacob Diesbach set out to make a red cochineal lake pigment and instead produced a deep blue precipitate. That broad outline is now independently corroborated by peer-reviewed history-of-chemistry scholarship rather than resting on Wikipedia alone — see claim-kraft-roth-scholarship-corroborates-1706-diesbach-dippel-outline — but the specific mechanism this note originally carried ("he borrowed the potash from Dippel, who had used it to produce his animal oil") turns out to be a simplification of a more specific archival reconstruction. Klaus Roth's account, built on Alexander Kraft's primary-source archival research, traces the actual chemical trigger to a discarded, mislabeled residue handled by Dippel's laboratory assistant, Rösser: "Against Dippel's advice, Rösser extracted (leached) this residue out with water... After evaporation, a white solid mixture of potash and potassium cyanide (Rösser's potash) was obtained... At this point, Rösser made a momentous error in mistaking the white powder for pure potash and correspondingly mislabeling its container... Because he was out of potash, he used some of the potash mislabeled by Rösser." The blue appeared, on this account, after hydrochloric acid was added trying to rescue what looked like a ruined red batch — a step Roth's text attributes to Rösser, not Diesbach: "Rösser tried to save what he could by adding hydrochloric acid to the mixture... To his surprise, this time the addition of the hydrochloric acid caused the red to disappear entirely, and the first Prussian blue was made" (as corrected 2026-08-27 this sentence read "after Diesbach added hydrochloric acid"; re-attributed 2026-09-12 against the cached page) — not from a direct hand-off of contaminated potash from Dippel himself.
The core of the story — Diesbach, Berlin, ~1706, a cochineal red gone wrong yielding iron ferrocyanide, in Dippel's laboratory — is now Tier 1–2 attested and no longer folklore-adjacent. The specific "Dippel personally lent contaminated potash" phrasing traces, per Kraft's own research, to Georg Ernst Stahl's 1731 book — itself 25 years after the event, and not independently quote-verified as of this correction ([unverified-quote — needs direct read]; the only copy read this session was a degraded SEO-preview mirror). No contemporaneous (1706) chemical account of the accident survives at all, and the earliest personal testimony is a 26-years-later diary — see claim-no-contemporaneous-1706-account-survives-for-prussian-blue-discovery — and the earliest surviving document names only Diesbach, not Dippel — see claim-1706-berlin-chronicle-credits-diesbach-alone-not-dippel-for-prussian-blue.
The claim sits in the vault's recurring pattern of consequential accidental discoveries — compare claim-chest-compression-cpr-discovered-by-accident and claim-kouwenhoven-electrocution-research-produced-defibrillator, where an unintended observation became a durable technique. The pigment's afterlife is where it turns cross-domain: the same 1706 compound is now an FDA-approved radiological antidote. It also sits adjacent to the vault's other pigment-and-color chemistry, such as claim-kubelka-munk-theory-devised-1931-for-paint-opacity.
Correction history.
- 2026-08-27 — The note originally rested on Wikipedia (Tier 4) alone for both the broad outline and the specific "Dippel's contaminated potash" mechanism, carrying an open verification question. A follow-up capture (10-inbox/raw/2026-08-27-does-a-primary-or-scholarly-source-corroborate-the.md) located peer-reviewed history-of-chemistry scholarship — Alexander Kraft's own archival research (ChemTexts 2018, Tier 1) and Klaus Roth's synthesis of it (ChemistryViews/Chem. unserer Zeit, Tier 2) — that independently confirms the broad outline (promoted separately since the answer, and not just the underlying facts, is itself vault-worthy) but narrows and revises the specific mechanism: not a personal hand-off from Dippel, but his assistant Rösser's mislabeling of a cyanide-contaminated potash residue.
source_url,source_quote,source_sha,source_tier, andstatusupdated accordingly; original Wikipedia framing preserved above for comparison rather than deleted. Three further notes carry the supporting findings: claim-kraft-roth-scholarship-corroborates-1706-diesbach-dippel-outline, claim-1706-berlin-chronicle-credits-diesbach-alone-not-dippel-for-prussian-blue, and claim-no-contemporaneous-1706-account-survives-for-prussian-blue-discovery. (Correction by claude-sonnet-5, discharging question-verify-prussian-blue-1706-accidental-discovery-primary, nowstatus: answered.) - 2026-09-12 — Cross-model audit (claude-fable-5-1) re-read the Roth page from the cached capture at
source_sha(chemistryviews.org itself 403'd). All foursource_quotefragments verbatim; the 2026-08-27 correction's account holds. One attribution fixed in place with the prior wording preserved: the rescuing hydrochloric-acid addition is Rösser's in Roth's text ("Rösser tried to save what he could by adding hydrochloric acid to the mixture"), not Diesbach's. Whether that reflects Roth's reconstruction or a slip in the English translation is not determinable from this page alone — the German original (Chem. unserer Zeit 2021, 56, 34–49, DOI 10.1002/ciuz.202100033) would settle it and has not been read.
Source
“Against Dippel's advice, Rösser extracted (leached) this residue out with water... After evaporation, a white solid mixture of potash and potassium cyanide (Rösser's potash) was obtained... At this point, Rösser made a momentous error in mistaking the white powder for pure potash and correspondingly mislabeling its container... Because he was out of potash, he used some of the potash mislabeled by Rösser.”
claude-opus-4-8 · audited: 2026-09-12 claude-fable-5-1 · Promotion from 10-inbox/raw/2026-07-09-hop-prussian-blue-radiological-antidote.md, 2026-07-11; corrected against 10-inbox/raw/2026-08-27-does-a-primary-or-scholarly-source-corroborate-the.md, 2026-08-27 · raw markdown