The literature

Every source behind the Miller-Urey article, with what it measured, whether anyone here opened it, which way it cuts, and the sources considered and not used.

Every source the article rests on is listed here with what it measured, whether anybody here read it, and which way it cuts.  The larger catalogue for the whole origin-of-life series is on the parent page, the origin of life: the literature.  This page lists only what this article uses or deliberately does not use.

Direction.  supports strengthens the case that the origin of life is unsolved;  undercuts weakens it;  complicates cuts both ways.  Assigned by what the source shows, not by what its author believes.

Access.  "Opened in full" means somebody on this project read the file.  "Abstract" means the abstract was verified verbatim from a database record and the body was not read.

Sources cited

Source What it is Access Direction
Miller 1953, Science 117:528 The original two-page note. CH4/NH3/H2 over water, sparked one week. Three amino acids identified with confidence, two of them proteinogenic, milligram yields. Records the colloidal silica from the glass. Opened in full (JSTOR scan) complicates
Miller & Urey 1959, Science 130:245 Summary of the 1955 analysis. Table 2 accounts for 15% of input carbon; glycine 2.1%. Does not itself say where the other 85% went. Opened in full incl. Table 2 supports
Lazcano & Bada 2003, OLEB 33:235 Fiftieth-anniversary retrospective. "a relatively small number of compounds were produced in surprisingly high yields," on 1950s chromatography. Opened in full complicates
Cleaves et al. 2008, OLEB 38:105 Neutral CO2/N2 spark. Yields on input N: 0.013% unbuffered, 0.19% with CaCO3 buffer, 2.5% with buffer and ascorbate. Names Wells 2000 as the objection answered. Ascorbate is not prebiotic; pyrite gave 10× in a separate 1 ml test, so no buffer-plus-pyrite yield exists. Few free amino acids before hydrolysis; HCN 0.0% in the buffered run. Earlier neutral yields "low but not zero." O2 run still gives racemic glycine and alanine. No tar in neutral runs. Opened in full incl. Tables 1-2 undercuts
Johnson et al. 2008, Science 322:404 Miller's archived volcanic-apparatus vials: 22 amino acids, 5 amines, racemic. Reducing gases still. Absolute yields unrecoverable. Table S1 is an image and was not read. Opened in full incl. SOM undercuts (venue), complicates (yield)
Parker et al. 2011, PNAS 108:5526 Miller's 1958 H2S run: 23 amino acids, 4 amines, 7 organosulfur compounds, racemic; trace L-excess in three low-abundance species flagged as contamination. Opened in full (PMC) complicates
Trail, Watson & Tailby 2011, Nature 480:79 Cerium-in-zircon oxybarometry: Hadean melts at FMQ, so outgassing would not give a reducing atmosphere. Abstract (Europe PMC) supports the geological premise
Cleaves 2012, Evo Edu Outreach 5:342 Review. CO2 is "almost as good" as methane only with a high H2/C ratio; Hadean H2 level unclear. Opened in full complicates
Wollrab et al. 2016, OLEB 46:149 FT-ICR MS on a Miller-type broth: 668 substances in m/z 150-300, "we cannot identify any preferred reaction product." Opened in full supports
Catling & Zahnle 2020, Sci Adv 6:eaax1420 Review of the Archean atmosphere. Surface O2 < 10-6 PAL from S-MIF; end-Hadean atmosphere anoxic, CO2/N2. Opened in full (CC BY) undercuts the O2-rich claim
Zahnle et al. 2020, PSJ 1:11 Impact-generated reducing atmospheres from late-veneer iron; states the paradox; biggest impacts leave the surface at 400 K. Opened in full complicates
Criado-Reyes et al. 2021, Sci Rep 11:21009 Borosilicate vs Teflon: 48 vs 31 compounds; alanine, valine, Asp, Glu, isovaline and glycylglycine only with glass; yield factor 3.3 buffered, 1.6 unbuffered. Opened in full (CC BY) complicates
Wogan et al. 2023, PSJ 4:169 Impactor of 570-1330 km needed for CH4/CO2 > 0.1; four to seven of the smaller, zero to two of the larger; >360 K surfaces; cyanide-stockpiling proposal "assuming that all molecules are preserved." Opened in full (CC BY) complicates
Jenewein et al. 2025, PNAS 122:e2413816122 Silica-induced HCN-polymer films with cell-shaped structures; warning about abiotic microfossil mimics. Two corrections read, both administrative. Opened in full (PMC) complicates
Meng et al. 2025, Sci Adv 11:eadt8979 Microlightning between sprayed water droplets makes HCN, glycine, urea, uracil in a CH4/NH3/CO2/N2 mixture. Energy source, not atmosphere. Opened in full (PMC) undercuts the frequency objection only
Paschek et al. 2025, arXiv:2503.15479 / ApJ Serpentinization alone can reduce a CO2 atmosphere enough for nucleotide synthesis. A model. Preprint, abstract and introduction undercuts
Diukarev et al. 2025, Angew Chem 64 Carell group: HNO + formaldehyde + SO2 under weakly reducing conditions gives precursors to amino acids and all four nucleosides. Abstract (Europe PMC) undercuts
Houchin et al. 2026, PNAS 123:e2525466123 Uranium XANES on Jack Hills zircon: FMQ−1 to +1, "near-modern redox states by 4.15 Ga." Post-cutoff. Abstract (Crossref JATS) supports the geological premise
Ohmoto & Ferry 2026, Sci Rep 16 Thermodynamic case that the prebiotic atmosphere was CH4/NH3/H2-rich after all. Minority dissent, models not proxies. Post-cutoff. Opened in full (PMC) complicates
Moldogazieva et al. 2026, Comm Chem 9 Review: "weakly oxidizing, but transiently reducing." Post-cutoff. Abstract verbatim, body skimmed undercuts
Benner 2014, OLEB 44:339 The paradoxes paper; the "asphalt" framing set against Wollrab's optimism. Opened in full supports
Bergman 2004, J. Creation 18(2):28 "was probably O2-rich," traced through his reference list to a PBS tie-in book; "some O2" sourced to 1984-85. Opened in full (AiG reprint) apologetics primary; the claim audited
Sarfati 2024, "15 loopholes" Loophole 1, 2024 wording, now hedged with "in the absence of reducing gases"; the free-oxygen claim and Rosing & Frei 2004. Opened in full from disk apologetics primary; the claim audited
Sarfati 2015/2024, "The Miller-Urey experiment revisited" CMI feedback page. Reads Trail 2011's magma oxidation state as an oxidizing atmosphere; correct on the non-protein amino acids in Parker 2011; the cold-trap objection. Fetched 4 Sept 2026 via bin/cdp-fetch.py (creation.com blocks curl) apologetics primary; the claim audited
Wells 2000, Icons of Evolution, ch. 2 (pp. 9-27) The Intelligent Design statement of the objection. Photodissociation oxygen "never resolved" (p. 17); Walker's "A reducing atmosphere is required" as evidence of circular reasoning (p. 18); neutral CO2/N2/H2O "yielded no amino acids at all" and "it doesn't work" (pp. 21-22); textbook captions (p. 25). Wrong on oxygen and on "no amino acids at all"; fair on the 1970s circularity and on the textbook captions. Read 27 Oct 2026 by full-text search of archive.org iconsofevolution0000well (printdisabled), bin/iapdsearch.py, printed page = leaf - 18, anchored on running heads at leaves 29, 31, 37, 39 apologetics primary; the claim audited
Gishlick 2003, NCSE, "Icon 1: The Miller-Urey Experiment" The published reply to Wells. Right that oxygen was "essentially absent"; overstated that atmosphere revisions "have little effect" on yields, which were 0.013% until 2008. Opened in full (sources/web/gishlick-2003_*) undercuts the apologetic; itself overstated
Thaxton, Bradley & Olsen 1984, The Mystery of Life's Origin, chs. 4 and 6 Dose's ≤10-7 M amino-acid ceiling and "much too dilute" (pp. 60-61); traps "a legitimate means of collapsing time," but the ocean is not clearly a trap (pp. 102-103); "no geological evidence indicates an organic soup" (p. 66, not used: no Hadean rock survives to test it). Read 27 Oct 2026 from archive.org mysteryoflifesor0000thax (printdisabled), bin/iapdsearch.py, printed page = leaf - 14, anchored at leaves 57, 81, 115, 125; OCR exponents recovered from the prose supports
Meyer 2009, Signature in the Cell, pp. 334-335 Sets the atmosphere aside; the tar ("melanoidin, the curse of the prebiotic chemist") and investigator intervention are his argument. The intervention inference is weak for the one-pot 1953 run. Read 27 Oct 2026 from archive.org signatureincelld0000meye, bin/iapdsearch.py, offset 12, running heads at leaves 69, 346, 347 supports (tar); complicates (inference)
Stribling & Miller 1987, OLEB 17:261 Spark energy yields for HCN and H2CO; ocean HCN 4 × 10-6 M and amino acids 3 × 10-4 M at steady state, vent destruction only. The mainstream counterpart to Dose's ceiling; scaled to a neutral sky in research/dilution.ts. Abstract (PubMed 2819806); body paywalled undercuts the dilution argument as TBO priced it; supports once scaled to a neutral sky
Wogan et al. 2023, PSJ 4:169 (further passages) Abstract: Hadean air "relatively oxidizing with negligible photochemical production of prebiotic molecules"; HCN < 105 cm-2 s-1 when CH4/CO2 < 0.1. p. 8: lightning ≤ 104. p. 10: volcanic HCN "seems challenging"; pond HCN 4 × 10-11 M. p. 11: 0.01 M needed. Opened in full supports
Cleaves 2012 (further passages) p. 347: CO2/N2 air turns split N into NOx, not HCN. p. 351: ocean HCN 2 × 10-6 M at 0 °C, 7 × 10-13 M at 100 °C. p. 342: products "would need to be concentrated." Opened in full supports; the reply (local concentration) complicates

Considered and not used

  • Miller 1955, J. Am. Chem. Soc. 77:2351-2361.  The full product analysis with the polymer description.  Not obtained.  Everything from it here is relayed through Miller & Urey 1959, Cleaves 2008 and Johnson 2008.  It belongs in the library.

  • Schlesinger & Miller 1983, J. Mol. Evol. 19:376-382.  The baseline Cleaves 2008 overturns and the source of the ~5.5% reducing-atmosphere figure.  Not obtained.  The article states only that the two yields are on different denominators.

  • Flowers 1998, A Science Odyssey, p. 173.  Bergman's apparent source for "O2-rich."  Not read;  archive.org holds it (scienceodyssey100000flow).

  • Rosing & Frei 2004, EPSL 217:237-244.  Sarfati's and Bergman's source for early oxygenic photosynthesis.  Not opened.  The article rests on Sarfati's own description of what it claims and on the S-MIF constraint.

  • Wogan, Catling & Zahnle 2024, Astrobiology 24, "Timing and Likelihood of the Origin of Life."  Abstract only.  Its 90% / 20% figures are conditional on assuming prebiotic chemistry required a post-impact atmosphere, so they were not used.

  • Faubel 2025, Natural Sciences, commentary on Meng et al.  Wiley blocks automated fetches and Crossref has no abstract.  Its existence is noted in the article;  its content is not characterised.

  • Menor-Salván & Ruiz-Bermejo 2026, Astrobiology.  Abstract only (Liebert paywall).  Puts the open problem at nucleosides and polymers rather than the atmosphere;  consistent with the article but not needed.

  • Yang et al. 2026, PNAS 123 (mineral-facilitated HCN from amino acids);  Scherf et al. 2026 Research Square preprint (formaldehyde rainout);  Boden et al. 2026, Sci Adv (abiotic fixed nitrogen).  Surfaced in the 2026 search, not opened, not used.

  • The ScienceDaily press release of 30 November 2011 that Sarfati quotes for "an oxidation state closer to present-day conditions."  Not fetched.  The article attributes the sentence to the report, as Sarfati does, and checks it against Trail's abstract instead.

  • Chyba & Sagan 1992, Nature 355:125.  Considered for global organic production rates by atmosphere type.  Paywalled;  the library holds only a YAML.  Wogan 2023 and Stribling & Miller 1987 cover the same ground with numbers that could be read.

  • Hill 1991, OLEB 22:277.  Johnson 2008's only reference for reducing gases in eruption plumes.  Not opened;  the article reports only that Johnson's case is one sentence.

  • Farquhar, Bao & Thiemens 2000, Science 289:756;  Miyakawa et al. 2002;  Toner & Catling 2019;  Pearce et al. 2022.  Cited only as Catling 2020, Cleaves 2012 and Wogan 2023 report them.  Not opened.

  • Thaxton, Bradley & Olsen 1984, p. 66, "no geological evidence indicates an organic soup."  Considered and not used:  no Hadean sedimentary rock survives, so the absence of a soup layer is expected on either view and cannot test it.

  • Thaxton, Bradley & Olsen 1984, p. 102, "two days of sparking represent an energy input ... comparable to some 40 million years."  Not used:  a quotation whose source is in an endnote the search route does not return.

Open questions

  1. Johnson 2008 Table S1 (per-compound abundances) is an image and has not been read.
  2. Nobody appears to have characterised the insoluble 85% of a Miller-type run at compound level. If that is right it should be stated more strongly than the article dares to.
  3. The cerium-in-zircon calibration has a critical literature ("revised Ce/Ce*") that was not surveyed. The article cites two proxies agreeing, which is the honest hedge, but the precision should not be overstated.
  4. Bergman's footnote attribution to Flowers 1998 should be checked against the print Journal of Creation 18(2).
  5. The rain-out step, from spark to sea, is unpriced: how much of what lightning makes reaches water before the same energy destroys it. The published ocean estimates (3 × 10-4 M, Stribling & Miller 1987; ≤ 10-7 M, Dose via Thaxton et al. 1984) differ mainly in which destruction they count.
  6. Stribling & Miller 1987's body would show which HCN rate the 3 × 10-4 M estimate uses; only the abstract was read.