Rbeida et al. 2023 — Pb, tAs, Cd, Ni in 12 cosmetic samples (Henna leaves, Henna paste, baby Talcum powder, Khol stones), Tripoli, Libya
This Mediterranean Journal of Pharmacy & Pharmaceutical Sciences study quantifies four heavy metals (lead, total arsenic, cadmium, nickel) by atomic absorption spectroscopy (Shimadzu AAS-6300) in 12 cosmetic samples purchased from public markets in Tripoli, Libya in winter 2022: three brands each of Henna leaves, Henna paste, baby Talcum powder, and Khol stones. Khol stones carry the highest lead by an order of magnitude relative to the other three product categories: one sample reaches 102.1 µg/g, roughly 51× the BVL Germany cosmetic limit (2.0 µg/g), 10× the WHO/Health Canada limit (10 µg/g), and 200× the EU limit (0.5 µg/g). One Henna paste sample reaches 41.2 µg/g Pb; one baby Talcum powder sample reaches 33.6 µg/g Pb. Cd and tAs are present at lower but still appreciable levels in henna and kohl; Ni is largely below the limit of detection in henna and talc but reaches 1.5 µg/g in one kohl sample. The paper’s central argument is that international cosmetic heavy-metal guidelines diverge widely (FDA, EU 1223/2009, WHO, Health Canada, German BVL) and that the Libyan National Centre for Standardization and Metrology (LNCSM) lacks national specifications for cosmetic heavy-metal limits; the authors call for harmonization.
Key numbers
Per-sample concentrations from Table 3 (p. 5; µg/g; the ± column is reported as “±SD” in the table header but the paper text states “the percentage of RSD or (±cd) was calculated” so these values are interpreted as %RSD or RSD-derived dispersion estimates rather than absolute SDs — see Verification notes; ND = Not Detected):
| Sample | Pb (µg/g) | tAs (µg/g) | Ni (µg/g) | Cd (µg/g) |
|---|---|---|---|---|
| LEAV1 (Henna leaves) | 21.3 ± 0.9 | 1.4 ± 2.1 | ND | 0.8 ± 3.1 |
| LEAV2 (Henna leaves) | 8.2 ± 2.9 | 1.7 ± 0.4 | ND | 0.04 ± 2.3 |
| LEAV3 (Henna leaves) | 11.9 ± 2.1 | 0.8 ± 1.8 | 0.08 ± 2.2 | 0.09 ± 4.1 |
| PAS1 (Henna paste) | 29.2 ± 3.1 | 3.5 ± 1.9 | 0.09 ± 2.3 | 1.7 ± 3.6 |
| PAS2 (Henna paste) | 5.6 ± 3.2 | 2.9 ± 3.1 | ND | 1.2 ± 4.5 |
| PAS3 (Henna paste) | 41.2 ± 1.2 | 1.4 ± 0.9 | 0.1 ± 4.1 | ND |
| BWD1 (baby Talcum powder) | 9.1 ± 0.8 | 0.8 ± 3.3 | ND | ND |
| BWD2 (baby Talcum powder) | 33.6 ± 1.6 | 1.9 ± 0.7 | 0.08 ± 4.8 | 0.8 ± 4.9 |
| BWD3 (baby Talcum powder) | 7.6 ± 3.4 | 2.2 ± 2.5 | ND | ND |
| Koh1 (Khol stones) | 54.4 ± 3.1 | 3.4 ± 1.1 | 0.05 ± 5.9 | 0.07 ± 3.1 |
| Koh2 (Khol stones) | 102.1 ± 1.4 | 4.5 ± 3.1 | 0.2 ± 3.1 | 1.2 ± 5.1 |
| Koh3 (Khol stones) | 89.8 ± 0.6 | 1.4 ± 2.5 | 1.5 ± 5.5 | ND |
Per-category ranges (computed from Table 3 by the wiki, not printed in the paper):
- Henna leaves (n=3): Pb 8.2–21.3 µg/g; tAs 0.8–1.7 µg/g; Ni ND–0.08 µg/g; Cd 0.04–0.8 µg/g.
- Henna paste (n=3): Pb 5.6–41.2 µg/g; tAs 1.4–3.5 µg/g; Ni ND–0.1 µg/g; Cd ND–1.7 µg/g.
- Baby Talcum powder (n=3): Pb 7.6–33.6 µg/g; tAs 0.8–2.2 µg/g; Ni ND–0.08 µg/g; Cd ND–0.8 µg/g.
- Khol stones (n=3): Pb 54.4–102.1 µg/g; tAs 1.4–4.5 µg/g; Ni 0.05–1.5 µg/g; Cd ND–1.2 µg/g.
Distribution-frequency ordering across all 12 samples (Results, p. 5): Pb > tAs > Cd > Ni.
Regulatory reference limits compared by the paper (Table 1, p. 3 — BVL Germany updated values; and Table 2, p. 3 — multi-jurisdiction comparison; µg/g unless noted):
| Standard | Pb | tAs | Cd | Ni |
|---|---|---|---|---|
| US FDA (cosmetics) | 5.0 | 5.0 (C) / 3.0 (CA) | NK | NK |
| WHO | 10 | 3.0 | 0.3 | 0.1 |
| Health Canada | 10 | 1.0 | 3.0 | 3.0 |
| BVL Germany (cosmetics) | 2.0 ᵃ | 0.5 (toothpaste) | 0.1 (toothpaste) | 0.1 |
| EU | 0.5 | ** | 0.5 | ** |
Footnotes from the paper’s tables: ᵃ = “for the products make-up powder, rouge, eye shadow, eyeliner, Kajal, carnival make-up”; (C) = cosmetics; (CA) = color additives; “NK” = not known/not specified; ”**” = not specified in EU. Table 1 (BVL Germany) additionally lists Hg cosmetic limit 0.1 mg/kg and Sb 0.5 mg/kg, neither of which this study measured.
Exceedance counts that the paper reports narratively (Results, p. 5-6; numbers cross-checked by the wiki against Table 3):
- BVL Germany Pb (2.0 µg/g): all 12 samples exceed.
- WHO Pb (10 µg/g): 8 of 12 detected-and-exceed (LEAV1 21.3, LEAV3 11.9, PAS1 29.2, PAS3 41.2, BWD2 33.6, Koh1 54.4, Koh2 102.1, Koh3 89.8); 4 of 12 are below 10 (LEAV2 8.2, PAS2 5.6, BWD1 9.1, BWD3 7.6). The paper narrative on p. 6 states “samples containing high levels of Pd and passed the level of 10 ppm except for three samples” — this implies 9 above / 3 below, which under-counts the “below” by one. The Table 3 enumeration is taken as authoritative: 8 above, 4 below. See Verification notes for this narrative-vs-table discrepancy.
- Health Canada Pb (10 µg/g): same cut-off as WHO; 4 samples below 10, 8 above. Paper narrative p. 6 states “for Pd all the samples contained high levels of Pd, except four samples” — this row aligns to Table 3 (4 below).
- WHO tAs (3.0 µg/g): 3 samples exceed (PAS1 3.5, Koh1 3.4, Koh2 4.5).
- WHO Cd (0.3 µg/g): 3 samples are detected-and-below 0.3 (LEAV2 0.04, LEAV3 0.09, Koh1 0.07) and 4 samples are ND for Cd (PAS3, BWD1, BWD3, Koh3) — for a total of 7 of 12 samples at-or-below 0.3 µg/g. Paper narrative p. 6 states “only four samples are below the limit of 0.3 ppm.” The most plausible reading is that the paper’s “four below” refers to the 4 ND samples (treating ND as below-LOD-below-0.3), with the 3 detected-low samples treated as a separate category in the narrative; alternatively the narrative under-counts the at-or-below-0.3 set. Either way, Table 3 enumeration is authoritative: 3 detected-below + 4 ND = 7 at-or-below 0.3 µg/g; 5 above 0.3 µg/g (LEAV1 0.8, PAS1 1.7, PAS2 1.2, BWD2 0.8, Koh2 1.2). See Verification notes.
- WHO Ni (0.1 µg/g, paper text also writes 1.0 ppm — discrepancy with Table 2 which lists WHO Ni = 0.1): the paper narrative reports “only three samples are below the limit of 1.0 ppm of Ni” but Table 2 lists WHO Ni limit as 0.1 µg/g. At the 0.1 µg/g cut-off, two samples exceed (Koh2 0.2, Koh3 1.5); at the 1.0 µg/g cut-off, only Koh3 (1.5) exceeds. The narrative-versus-table discrepancy on the WHO Ni reference value is reproduced verbatim from the paper; see Verification notes.
Methods
Cosmetic products were purchased from different shops in Tripoli, Libya in winter 2022: three brands of talcum powder, Henna leaves and paste from various stockholders and distributors (Henna paste includes one locally produced homemade sample; the remaining Henna leaves, Henna paste, and Khol stones are predominantly imported from Sudan, India, and Pakistan). Henna leaf samples were dried at 85 °C in an air-ventilated oven for 12-24 hours, ground with a mixer grinder, and sieved through a 250 µm mesh.
For each sample, 1.0 g was weighed into Teflon vessels (to avoid metallic contamination) and digested with 6.0 mL of 65.0% nitric acid plus 2.0 mL of 30.0% hydrogen peroxide; the mixture was evaporated near dryness on a hot plate for 10 min, then filtered through Whatman No. 42 filter paper followed by 0.45 µm millipore filter paper, transferred quantitatively to a 50.0 mL volumetric flask, and topped up with deionized water. The 50.0 mL filtrate was transferred into an acid-rinsed polyethylene container for analysis. Glassware was cleaned with 6.0% nitric acid for 24 hours; Teflon beakers were treated with 6.0% nitric acid and washed with ultrapure water.
Quantification was performed on a Shimadzu AAS-6300 atomic absorption spectrophotometer at the Drug and Food Control Center laboratory (Tripoli). Standards were prepared from certified stock solution (1000 ppm, Merck, Darmstadt, Germany) and from Fluka Analytical (Sigma Aldrich, UK). Calibration concentration ranges: Pb 5.0–250 µg/mL; Ni 0.5, 1.5, 8 µg/mL; tAs 2.5, 10, 15, 20 µg/mL; other metals 0.5–60 ppm. Linear correlation coefficients (r²) were reported in the range 0.9822–1.000. The paper defines limit of detection (LOD) as the lowest calibration concentration for each metal and limit of quantification (LOQ) as 3 × LOD; specific per-metal LOD/LOQ values in finished-sample units (µg/g) are not enumerated. Recovery from spiked standard solutions across three concentrations was reported in the range “0.89%–99.5%” (printed verbatim; the 0.89% lower bound is anomalous and probably an 89% typo — see Verification notes). All solutions were prepared in double-distilled water; dilution correction was applied where samples were diluted or concentrated. Duplicate determinations per sample were averaged and the percentage of RSD (or ±cd) was calculated.
Speciation: total arsenic and total cadmium/lead/nickel only — no inorganic/organic As speciation, no inorganic Pb speciation, no Cr or Hg measurement. Mercury and antimony are listed as regulatory reference analytes (Table 1, BVL Germany updated limits) but not measured in this study.
Implications
-
Certification (HMTc): Direct occurrence evidence for at least three HMTc Category 13 (Cosmetics and Personal Care — Leave-on) rows and one HMTc Cat 2 (Children Personal Care) routing destination. The three baby Talcum powder samples (BWD1-3) contribute occurrence data for
baby-talcum-powder(Cat 2 children-personal-care routing) — Pb 7.6–33.6 µg/g, with BWD2 at 33.6 µg/g substantially exceeding the BVL Germany 2.0 µg/g cosmetic-Pb limit and meaningfully above the FDA 5.0 µg/g and WHO/Health Canada 10 µg/g caps. The three Khol stone samples (Koh1-3) contribute occurrence data foreye-makeup(Cat 13 Row 7) — Pb 54.4–102.1 µg/g, with the maximum at ≈ 51× the BVL Germany limit. The three Henna paste samples (PAS1-3) and three Henna leaves samples (LEAV1-3) contribute occurrence data for skin/body leave-on applications (henna paste applied to arms and thighs per the paper’s Introduction, p. 1) and for the leaves as the raw material that becomes paste or hair preparation. Sample frame is very small (n=3 per product category), method recovery range is wide, and the reported ±SD values are likely %RSD rather than absolute SD; these limitations make this source one contributing entry in a multi-source pool rather than a standalone anchor for any threshold. -
Courses: Useful as a teaching case for (a) the regulatory-divergence problem (FDA, EU, WHO, Health Canada, and German BVL diverge by 4-200× on the cosmetic Pb cap alone), (b) traditional cosmetic categories (henna, kohl) that fall outside most Western regulatory frameworks but carry the highest Pb concentrations, and (c) the limits of small-n AAS quality-control studies as a basis for population exposure inference.
-
App: This paper is a leave-on cosmetic source (not food) and does not change any food-ingredient
contamination_profileblock. If the future Heavy Metal Index app extends to a leave-on personal-care category — particularly traditional cosmetics like kohl, henna, and ethnic baby talcum powders — the per-sample occurrence values from Table 3 are direct inputs. -
Microbiome: Not addressed by the paper.
Wiki pages this source may touch
- lead
- arsenic-total
- cadmium
- nickel
- baby-talcum-powder
- eye-makeup
- body-hand-leave-on-skin-care
- leave-on-hair-preparations
Verification notes
- 2026-05-18 fresh ingest (Claude Opus 4.7, autonomous v2.0 manual-fetch skill, daemon tick): NEW path. Three identity checks against
wiki/sources/returned no hits: DOI10.5281/zenodo.8136836not present; raw_handleMFK_quality-control-cosmetics-libyanot present; cite-key stemrbeida2023not present. PDF SHA-256d6928cf09c76b4f130fc2535eeaf4aea37175b9fdc22261e28117f03119981c3. metals: [Pb, tAs, Cd, Ni]usestAs(notiAs) per Part 14 speciation discipline: the paper measures total arsenic by AAS with no inorganic-vs-organic speciation step (Materials and Methods, p. 4-5; Table 3 reports a single “Arsenic” column). The paper text repeatedly writes “Ar” for arsenic (e.g., Abstract, p. 1; Introduction, p. 2) — this is a paper-internal abbreviation idiosyncrasy, not a different analyte. The paper also misspells the symbol Pb as “PB” (Table 1, p. 3) and “Pd” (Results narrative, p. 5-6); both refer to lead per context, and the wiki page uses the standardised abbreviation Pb throughout.evidence_tier: Brather than A: the paper is peer-reviewed (Mediterranean Journal of Pharmacy & Pharmaceutical Sciences is indexed and applies review per its policy page) and reports analytical method and instrument, but several methodological items are weak: (a) the printed recovery range “0.89%–99.5%” is implausibly wide (the lower bound is almost certainly a typo for 89%); (b) per-metal LOD/LOQ values in finished-sample units are not enumerated; (c) the dispersion values in Table 3 are labelled “±SD” but the paper text describes “percentage of RSD or (±cd)” being calculated — these are likely %RSD, which makes the dispersion values misleadingly large relative to the reported concentrations (e.g., LEAV2 Cd 0.04 ± 2.3 reads as a 5750% RSD if taken at face value, or as approximately 2.3% RSD if interpreted as RSD%); (d) no reference material spike-recovery validation against a certified cosmetic reference material is reported; (e) n=3 per product category limits statistical power; (f) the paper-internal narrative contradicts its own Table 3 on several maxima (narrative p. 5 says “Pb in Henna leaves and paste the maximum was 41.1 µg/g” but Table 3 PAS3 reads 41.2; narrative says “As in Henna products reached to 2.9 µg/g” but Table 3 PAS1 reads 3.5; narrative says “in Talc baby powder it was 3.2 µg/g” for As but Table 3 maxima for BWD samples is 2.2 at BWD3; narrative says “Ni in Henna products was 0.9 µg/g” but Table 3 maximum Ni in henna samples is 0.1 µg/g at PAS3). Table values are taken as the primary data here (tables are usually source-of-truth over results-narrative); narrative discrepancies are reproduced here in this verification log rather than silently corrected. The B-tier designation reflects this overall pattern of imprecise/inconsistent presentation rather than evidence of fabrication.license: "CC BY 4.0": the article footer on each page states “Copyright© 2023. This open-access article is distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.” This is read as CC BY 4.0 (the default current version when “Creative Commons Attribution License” is named without further qualification); if a future check of the journal’s policy page indicates a different CC BY version, this field should be updated.jurisdictions: [LY]: samples were purchased from public markets in Tripoli, Libya (Materials and methods, p. 4). The paper extensively compares results to US, EU, German, Canadian, and WHO regulatory limits, but those are reference jurisdictions; only Libya is the sample-source jurisdiction.matrices: [cosmetic-personal-care]: the established bare-string matrix vocabulary for cosmetic-occurrence sources. Noexposure-modelingbecause this paper is a quality-control study of finished-product concentrations and does not compute dermal SED, MoS, HQ, HI, or LCR.products:routing rationale:baby-talcum-powder(Cat 2 children-personal-care) — direct evidence for the BWD1-3 samples described in the paper as “baby Talcum powder” used “frequently for babies in large quantities” (p. 2). Note the legacybabycare_03_Powder_Lotion_Oil_Diapersource folder name aligns with the children-personal-care framing for this source’s placement, even though the henna and kohl samples in the same paper are adult cosmetics.eye-makeup(Cat 13 Row 7) — direct evidence for the Koh1-3 Khol stones described in the paper as “applied to eyebrows, around the eyes, and as mascara for the eyelashes” (p. 1) — Khol is the traditional eye cosmetic that the eye-makeup row covers (eyeliners, eyebrow products, mascara).body-hand-leave-on-skin-care(Cat 13 Row 4-equivalent) — direct evidence for the PAS1-3 Henna paste samples described in the paper as “applied to the arms and thighs” (p. 3) and as body-decoration cosmetic (Introduction, p. 1). Henna paste in this study is the body-art / leave-on-skin application form.leave-on-hair-preparations(Cat 13 Row 11-equivalent) — direct evidence for the LEAV1-3 Henna leaves samples. Henna leaves are the raw material from which paste or hair-coloring preparations are made; the genus Lawsonia is most widely used as a hair dye/conditioner globally, and the paper describes Henna as “present as a powder, pasta, shampoo or a plant extract” (p. 1). The wiki routes LEAV samples to leave-on hair preparations as the predominant use-case for Henna-leaf raw material; if a future source-page schema adds ahenna-leavesortraditional-cosmeticsslug, this routing should be revisited.
- Brand firewall (Part 12, strict reading locked 2026-05-17): the paper does NOT name individual cosmetic product brands in Table 3 — the samples are coded as LEAV1-3, PAS1-3, BWD1-3, and Koh1-3 with no manufacturer identification attached. No brand-attribution stripping was necessary. The methods section names instrument vendor (Shimadzu AAS-6300), reagent stock vendor (Merck, Darmstadt, Germany), and standards vendor (Fluka Analytical / Sigma Aldrich, UK) — these are scientific-method vendor names retained per Exception 2 of the Part 12 strict reading (locked 2026-05-17).
- Wiki/HMTc firewall (Part 2): the paper’s Conclusion advocates “the need for unified international guidelines” and updating Libyan national specifications via LNCSM. The wiki page reports this advocacy as a paper-internal claim without adopting it as the wiki’s voice and without proposing HMTc threshold values. The Implications section frames the source as one entry in a multi-source pool for cosmetic-product occurrence data, not as a sole anchor for any HMTc threshold.
- Internal paper inconsistencies catalogued (reproduced verbatim above; documented here for the audit trail):
- Narrative p. 5: “Pb in Henna leaves and paste the maximum was 41.1 µg/g” — Table 3 PAS3 reads 41.2 µg/g. Minor rounding/typo.
- Narrative p. 5: “As in Henna products reached to 2.9 µg/g” — Table 3 maximum henna tAs is PAS1 at 3.5 µg/g (PAS2 is 2.9). Narrative cites a non-maximum.
- Narrative p. 5: “in Talc baby powder it was 3.2 µg/g” for As — Table 3 BWD maximum tAs is BWD3 at 2.2 µg/g. Narrative number is not in the table.
- Narrative p. 5: “Ni in Henna products was 0.9 µg/g” — Table 3 maximum henna Ni is PAS3 at 0.1 µg/g. Narrative is off by 9×.
- Narrative p. 6: “for Ni only three samples are below the limit of 1.0 ppm” — Table 2 lists WHO Ni limit as 0.1 µg/g, not 1.0 µg/g. The narrative and Table 2 disagree on the WHO Ni reference value.
- Narrative p. 6: “for Pd all the samples contained high levels of Pd, except four samples” (Health Canada cut-off 10 µg/g) — wiki count confirms 4 below the 10 µg/g cut-off (LEAV2, PAS2, BWD1, BWD3). The narrative for the WHO 10 µg/g cut-off says “except for three samples,” which under-counts by one.
- Throughout: lead is spelled “PB” in tables and “Pd” in narrative (the symbol for palladium); arsenic is spelled “Ar” (not “As”) and the abstract additionally writes “mercury (Mg)” which is the symbol for magnesium. These are paper-internal symbol inconsistencies, not different analytes.
- Recovery range “0.89%–99.5%” is reproduced verbatim from p. 4; the 0.89% lower bound is almost certainly an “89%” typo (a 0.89% recovery would make the entire analytical method invalid).
- Data availability statement (p. 6): “The raw data that support the findings of this article are available from the corresponding author upon reasonable request.” No supplementary data repository linked.
- Audit subagent (2026-05-18, fresh-context general-purpose, autonomous v2.0 manual-fetch skill): REVISE verdict; 3 findings applied, 1 rejected as false positive.
- CHECK 1 ⚠️ WHO Pb (10 µg/g) exceedance count: wiki originally said “9 of 12 exceed” but enumerated only 8 sample codes; auditor recomputed against Table 3 and confirmed 8 of 12 exceed (LEAV1 21.3, LEAV3 11.9, PAS1 29.2, PAS3 41.2, BWD2 33.6, Koh1 54.4, Koh2 102.1, Koh3 89.8), 4 below (LEAV2 8.2, PAS2 5.6, BWD1 9.1, BWD3 7.6). Verified independently against the table; the prior wiki count was arithmetically wrong. Corrected the Key numbers bullet to read “8 of 12 detected-and-exceed” and clarified that the paper narrative’s “except for three samples” under-counts the below-set by one (the narrative-vs-table discrepancy is now logged here in this verification list rather than masked by a wrong wiki count).
- CHECK 1 ⚠️ WHO Cd (0.3 µg/g) “four samples below” count: wiki originally claimed “the wiki count of 4 detected-and-below matches” but the actual count of detected-and-below 0.3 is 3 samples (LEAV2 0.04, LEAV3 0.09, Koh1 0.07), with 4 additional ND samples (PAS3, BWD1, BWD3, Koh3) — total 7 at-or-below 0.3 µg/g, not 4. Verified independently against Table 3. Corrected to enumerate 3 detected-below + 4 ND = 7 at-or-below, with the most plausible reading of the paper’s “four below” being a reference to the 4 ND samples specifically (treating ND as the “below limit” category in the narrative).
- CHECK 5 ⚠️ Inline bold emphasis on “Khol stones carry the highest lead by an order of magnitude”: Part 15 / writing-style discipline prohibits inline bold for emphasis. Verified against
docs/writing-style.md. Bold removed; the same factual claim is retained without emphasis markup. - CHECK 5 ⚠️ Cross-source synthesis sentence “The lead-in-kohl signal is consistent with the broader literature on traditional kohl/surma being a major Pb-exposure vehicle” in Implications: this is a “consistent with the broader literature” cross-source claim that Part 2 / audit-prompt Check 5 explicitly identifies as a slip. The kohl/surma cross-literature claim is real (multiple published reviews establish kohl as a major Pb-exposure source globally), but stating it in a single-source page mixes synthesis into ingest. Sentence removed from Implications; the cross-source pattern, if relevant, should surface in the metals/lead page or in a synthesis pass per Part 9, not on this single-source page.
- CHECK 2 ⚠️ FALSE POSITIVE —
matrices: [cosmetic-personal-care]flagged as not in any controlled vocabulary visible to the auditor. Verified againstwiki/sources/:cosmetic-personal-careis the established bare-string matrix vocabulary in use across multiple cosmetic-occurrence sources (arshad2020 Pakistan, attard2022 review, bashir2025 Kashmir, opss2023 UK, sccs2023 EU SCCS, jitareanu2025, li2021). The matrices vocabulary is not enumerated indocs/gpt-collaboration/taxonomy-snapshot.mdbecause that snapshot covers ingredients/products/metals/regulations only; matrices is a separate controlled vocabulary tracked by use-count acrosswiki/sources/. Finding rejected; no change. - CHECK 1 (Table 3 per-sample values, per-category ranges, Table 1 BVL Germany values, Table 2 multi-jurisdiction limits), CHECK 3 (speciation tAs/Pb/Cd/Ni correct, methods accurately reflect PDF Methods section), and CHECK 4 (brand firewall — paper-side anonymisation preserved, method-vendor names retained per Exception 2): ✅ clean.
Page history
The five most recent substantive edits to this page. The full version history lives in git; when DOI minting comes online (see schema docs), each entry below will also link to a version-pinned DataCite DOI.