Verify Hirsch's forensic critique, the 'data alteration and manipulation' quote, and the Sept 2022 Nature retraction of Dias's room-temperature superconductivity paper against primaries

research-integrityfraud-detectionphysicscondensed-mattersuperconductivityhirschretractionverification

This capture answers question-verify-hirsch-dias-superconductivity-retraction-primaries, which flagged that claim-hirsch-forensics-drove-dias-superconductivity-retraction rested entirely on Tier-2 science journalism the batch worker could not re-fetch, and asked for three things against primaries: the load-bearing quote's exact source and attribution, Hirsch's mechanism in his own words, and the Nature retraction notice itself. All three are now grounded directly. science.org still 403s every fetch route tried (confirming the 2026-07-10 audit's finding, not a new block); Chemistry World and Nature's own site did not block anything this session.

Claim: Nature's own retraction notice confirms the September 26, 2022 date and the specific mechanism -- a non-standard, undisclosed background-subtraction procedure

Claim type: quantitative (date) + technical-mechanism. Tier 1-2 required; met at Tier 1.

Nature's retraction notice for "Room-temperature superconductivity in a carbonaceous sulfur hydride" (Snider, Dasenbrock-Gammon, Dias et al., Nature 586, 373, 2020) was published 26 September 2022. The editors' own stated reason: "We have now established that some key data processing steps — namely, the background subtractions applied to the raw data used to generate the magnetic susceptibility plots in Fig. 2a and Extended Data Fig. 7d — used a non-standard, user-defined procedure. The details of the procedure were not specified in the paper and the validity of the background subtraction has subsequently been called into question." All nine authors, including Ranga Dias, disagreed with the retraction and maintained the raw data support the paper's claims. This confirms, at Tier 1, the retraction date and mechanism the existing claim-note carried only from Tier-2 secondaries.

provenance: Nature retraction notice, 2022-09-26 (Tier 1).

Claim: Hirsch's own technical papers ground the forensic mechanism -- susceptibility data that is too smooth and internally inconsistent with the claimed measurement method, not merely "suspicious" by secondary characterization

Claim type: technical-mechanism. Tier 1-2 required; met at Tier 1.

In his own Physica C paper, Hirsch argues the published curves fail because (1) the one sample of "raw data" Snider et al. did publish shows behavior "completely contrary to what is expected for a superconductor," and (2) it closely resembles, in shape and magnitude, the raw data underlying a 2009 Physical Review Letters claim of superconductivity in europium by Mathew Debessai and coauthors -- a paper Hirsch had by then obtained the actual measured data for, and which he presents in the same paper as a demonstrated case: "The evidence of Examples 1, 2 and 3 discussed above establish unequivocally that extensive data alteration and manipulation occurred in going from measured data to published data for Fig. 2 of [10]" (ref. 10 being the 2009 Debessai et al. europium paper). Hirsch explicitly uses this prior, confirmed case as his model for what he believes happened to the CSH data, though he is careful in that paper to say he "cannot prove beyond reasonable doubt that fraud was involved" in CSH specifically absent the raw data.

The subsequent joint paper with Dirk van der Marel ("Extended Comment on Nature 586, 373 (2020) by E. Snider et al.," arXiv:2201.07686, published in International Journal of Modern Physics B) analyzes the raw data Dias and Salamat eventually posted and concludes: "For all 6 of the reported pressures our analysis leads to the conclusion that: (i) the reported background-corrected susceptibility data are pathological, (ii) they were not obtained by the method described in this paper nor by any one of the alternative 3 methods that were subsequently provided by the authors[,] and (iii) the 'measured voltage' data are not raw data." The specific technical finding is that the reported "superconducting signal" is the sum of a smoothly-varying spline and a quantized (discretized) component that should not be able to survive an honest subtraction of two independently noisy measurements -- and that the correlations between the reported "raw" and "corrected" data run in the direction predicted if the background was added to a smooth, pre-existing target curve, not subtracted from an independent measurement.

provenance: Hirsch, Physica C 590, 1353964 (2021), pre-proof (Tier 1); Hirsch and van der Marel, arXiv:2201.07686v7 (Tier 1).

Claim: The "data alteration and manipulation" quote is a joint Hirsch + Dirk van der Marel statement, not Hirsch alone -- correcting the earlier note's tentative "Reyer van der Marel" attribution -- but no Tier-1 primary carrying the exact wording was located this session

Claim type: direct quotation / attribution. Treated as load-bearing; Tier 1-2 required. Only Tier 2 met; flagged below.

Chemistry World reports: "Hirsch, along with Dirk van der Marel at the University of Geneva, then began a deeper investigation. Their analysis, Hirsch says, showed the team could not have produced those results with the method the paper described, or from the raw data they had obtained. Furthermore, the duo stated, it was not possible to produce the team's results 'without using procedures that can only be characterised as data alteration and manipulation.'" This fixes two things the existing claim-note left uncertain: the collaborator's actual first name is Dirk, not "Reyer" (an error introduced in the 2026-07-10 cross-model audit), and the quote is explicitly framed by the journalist as something "the duo stated" jointly, not Hirsch's words alone.

What this session could NOT do: locate the exact phrase "data alteration and manipulation" verbatim inside either joint Hirsch-van der Marel technical paper read directly (the arXiv Extended Comment, or the Matter and Radiation at Extremes paper "Incompatibility of published ac magnetic susceptibility of a room temperature superconductor with measured raw data," 7, 048401, 2022, whose abstract-only page was read but whose full text was not retrievable -- AIP's PDF endpoint 403'd). The word Science/AAAS's own reporter used for the same joint finding was "pathological," not "data alteration and manipulation" -- suggesting the exact wording may come from a statement made directly to Chemistry World's journalist (Kit Chapman) rather than from either published paper. Hirsch's own solo paper does use the near-identical phrase "data alteration and manipulation" verbatim, but there it describes the 2009 Debessai europium data specifically (Claim 2, above), not a joint CSH-specific statement. Flag: [unverified-quote -- exact wording only located in a Tier-2 secondary; the joint primary statement, if one exists in writing, has not been found.]

provenance: Chemistry World, Kit Chapman, 2022-09-29 (Tier 2, direct quote, receipted); cross-checked against Hirsch and van der Marel, arXiv:2201.07686 (Tier 1, mechanism present, exact phrase absent) and Science/AAAS, Eric Hand, 2022-09-26 (Tier 2, same finding paraphrased as "pathological").

Claim: Hirsch's arXiv trouble is two distinct, now primary-confirmed events -- his own CSH-fraud paper was withdrawn by arXiv administrators in December 2021, and he was separately, temporarily banned from posting to arXiv in February 2022

Claim type: historical/biographical (dates, institutional action). Tier 3-4 acceptable when uncontested; escalated here since the existing note flagged this detail as resting only on Tier-4 Wikipedia. Now met at Tier 1 for the withdrawal and Tier 2 for the posting ban.

arXiv's own abstract page for Hirsch's paper (arXiv:2110.12854) states directly: "This paper has been withdrawn by arXiv Admin, Submitted on 15 Oct 2021 (v1), last revised 9 Dec 2021 (this version, v2) ... Comments: arXiv admin note: This submission has been withdrawn by arXiv administrators due to a violation of professional communications standards." That is a distinct, dated (Dec 2021) event confirmed directly on arXiv's own site -- not the "~6-month ban" the earlier capture's Wikipedia sourcing described.

Separately, Science/AAAS reports a broader, later sanction: "In preprints, Hirsch kept hammering on the Dias study — so forcefully that in February, he was temporarily banned from posting to arXiv." Read in the article's own September 2022 context, "February" is February 2022 -- two months after the specific paper above was withdrawn. These are two different facts (one paper's withdrawal in Dec 2021; a site-wide posting ban in Feb 2022) that the prior capture had collapsed into one undated "~6-month ban," and both are now independently sourced rather than resting on Wikipedia alone.

provenance: arXiv:2110.12854 abstract page (Tier 1, for the Dec 2021 withdrawal); Science/AAAS, Eric Hand, 2022-09-26 (Tier 2, for the Feb 2022 posting ban).

Further leads

• Hirsch hosts a directory of his own papers at jorge.physics.ucsd.edu/blockedarxiv/ (e.g. "How do you trust but verify hydride superconductivity?") -- the directory name itself is a primary trace of his account of being suppressed by arXiv; unexplored this session, worth its own capture.
• Dias sent "cease and desist" letters to van der Marel and to Hirsch's department chair and dean at UCSD, per Chemistry World -- a retaliation-against-critics angle distinct from the data-forensics story, unexplored.
• The University of Rochester ran two internal inquiries into the CSH work and "found no evidence of scientific misconduct," per Chemistry World -- an institutional-response contrast worth its own note (forensic critique proved substantially correct; the accused researcher's home institution cleared him twice).
• Dias's 2023 nitrogen-doped lutetium-hydride Nature paper (a separate, later room-temperature-superconductivity claim) was also retracted (Nov 2023) -- Chemistry World's own "third room temperature superconductivity paper retracted" headline confirms a pattern beyond the single 2020 CSH case; out of scope here but a natural next capture.
• The Hirsch/van der Marel "Incompatibility..." paper (Matter and Radiation at Extremes 7, 048401, 2022) full text was not retrievable this session (AIP's PDF endpoint 403'd; the mre.org.cn mirror gave abstract/metadata only) -- worth a retry, since it may be where the exact "data alteration and manipulation" phrasing originates in writing.

Entity candidates

• Mathew Debessai — person — first author of the 2009 Europium PRL paper Hirsch treats as the confirmed precedent his CSH forensic method is modeled on; co-author of the 2020 CSH paper and, per Hirsch, the person who performed the CSH and Eu susceptibility measurements and taught Dias high-pressure technique. The earlier, foundational case this whole forensic-comparison argument rests on.
• Dirk van der Marel — person — condensed matter physicist, University of Geneva; Hirsch's co-author on the technical papers analyzing the CSH data, and joint speaker of the "data alteration and manipulation" quote this capture traces.
• Jorge E. Hirsch — person — already covered in the vault via claim-hirsch-invented-h-index-2005; this capture adds primary grounding to his forensic role.
• Ranga Dias — person — corresponding author of the retracted CSH paper and the later, separately-retracted lutetium-hydride paper.
• Ashkan Salamat — person — co-senior-author of the CSH paper; on record defending the work and criticizing Hirsch and van der Marel's conduct.
• Kit Chapman — person — Chemistry World journalist whose 2022-09-29 article is, per this session's search, the only located source carrying the exact "data alteration and manipulation" quotation.