Abulude et al. 2007 — Pb, Cd, Ni and Cr in 19 soaps and detergents from Akure (Ondo State), Nigeria

A short three-page Academic Journals Inc. paper quantifying lead, cadmium, nickel and total chromium in nineteen soap and detergent products purchased from open-market vendors in Akure, Ondo State, southwestern Nigeria, in July 2006. The sample mix spans toilet (bath/hand) soaps, multi-purpose bar soaps, powdered laundry detergents, a liquid laundry detergent, and one locally produced hand-mould soap. Quantification used dry ashing at 500 °C followed by 2 M HCl dissolution and analysis on a SP 9 Pye Unicam atomic absorption spectrophotometer at triplicate determinations per sample. The paper reports only pooled-distribution statistics (minimum, maximum, mean, standard error of the mean across all nineteen samples) for each metal; per-sample concentrations are not tabulated. Cadmium was below the detection limit in every sample (no LOD value is reported). Pooled means across the nineteen-sample distribution were Pb 3.78 mg/kg, Ni 2.20 mg/kg, and total Cr 1.33 mg/kg on a dry-product, as-purchased basis. The authors do not propose threshold comparisons against any cosmetic, household-cleaning or food regulatory ceiling; the framing throughout is descriptive (“the values obtained for the trace heavy metals determined were low, there is a cause for vigilance”) and the discussion section places the Akure results in context against other authors’ values for unrelated matrices (tea leaves, deer tissue, street dust, shrimp, coots) rather than against soap- or detergent-specific reference data.

Key numbers

Pooled-distribution statistics across all 19 samples (Table 2, p. 104; units mg/kg of finished product on a dry-ashed basis):

Cadmium was below the detection limit in every sample; the paper reports no numeric LOD value and no recovery, blank, or certified-reference-material data against which to interpret the “ND” floor.

Sample-composition breakdown (Table 1, p. 103; brand columns suppressed per Part 12):

Pearson correlation coefficients (Results & Discussion narrative, p. 103) between the analytes and the full sample set at 95 % confidence level:

• Pb: r = 0.969
• Ni: r = 0.835
• Cr: r = 0.806

(Cd is omitted because it was below detection limit throughout.) The paper does not specify what the correlation is between — the wording is ambiguous (“good correlation between the metals and all the samples”); a literal reading is correlation between the rank of the metal value and the rank of the sample index, but the authors do not state the second variable and the wiki does not impute one.

Authors’ verbal cluster statement (Results & Discussion, p. 103, brand attribution stripped per Part 12): the paper reports that “Pb, Ni and Cr contents were significantly higher” in five specific brand-named products (a subset of industrial toilet soaps and industrial powdered detergents) compared to the rest of the sample, and concludes that “the samples varied in the uptake of metals” while noting that “the type of package and manufacture had no significant influence on other variable.” No statistical test is named to substantiate the “significantly higher” claim and no per-sample values are tabulated to verify the cluster.

Methods (brief)

Sample acquisition. Nineteen soap and detergent samples were purchased from open-market vendors in Akure, Ondo State, Nigeria in July 2006. The paper does not state per-sample purchase locations, lot numbers, manufacturing dates, retail prices, or vendor identities. Each sample is identified in Table 1 by brand name plus a four-category type label (toilet soap, bar soap, hand-mould, detergent, liquid detergent), a three-category manufacture label (local, industrial, imported), and a three-category package label (cellophane, paper, plastic, or none).

Sample preparation and digestion. Approximately 0.5 g of each sample was dry-ashed in a muffle furnace at 500 °C for 1 hour. The ash was dissolved in 2 M hydrochloric acid, filtered, and made up to 50 cm³ with 2 M HCl in a volumetric flask.

Instrumentation. A SP 9 Pye Unicam atomic absorption spectrophotometer (Mercedes, 2002, cited as method reference) was used to quantify Pb, Cr, Ni and Cd. Per-element wavelengths reported in the Materials and Methods section:

• Cd: 217 nm
• Cr: 262 nm
• Pb: 283.3 nm
• Ni: 260 nm

The reported wavelengths for Cd (217 nm), Cr (262 nm) and Ni (260 nm) do not correspond to the standard primary atomic-absorption resonance lines for these elements (Cd 228.8 nm; Cr 357.9 nm; Ni 232.0 nm); the Pb wavelength (283.3 nm) is the conventional secondary Pb line. The wiki reports the wavelengths as published and flags the discrepancy from standard practice under Limitations rather than substituting external values.

Replicates. “At least triplicate determinations were carried” on each sample (p. 102). The paper does not report whether the triplicate is at the digestion or instrument-aspiration stage, and does not propagate the within-sample variance into the Table 2 standard-error reporting.

Quality control and detection limits. No certified reference material, internal standard, blank, recovery percentage, instrumental detection limit, method detection limit, calibration range, or measurement-uncertainty value is reported. The “ND” floor for Cd is asserted without numeric anchoring.

Speciation. Total metals only. The paper measures total chromium and does not distinguish Cr-VI from Cr-III. No arsenic or mercury speciation issues arise because neither analyte is measured.

Basis. Concentrations are reported as mg of metal per kg of finished product on a dry-ashed, as-purchased basis. The paper does not report moisture content, ash content, or any factor that would allow conversion between the dry-ashed basis and the as-used (lathered, diluted) basis of these products.

Implications

• The dataset adds a small (n = 19) regional view from southwestern Nigeria (Akure, Ondo State) to the very thin global literature on trace metals in commercial soaps and detergents. Three of the four analytes measured (Pb, Cd, Ni) are on the HMTc certification analyte panel; total Cr is measured rather than the priority-panel sub-form Cr-VI, so the chromium values inform the broader chromium literature but do not directly bind Cr-VI threshold-setting.
• Cadmium below the detection limit in every sample is the most reportable finding for HMTc purposes, but the absence of a numeric LOD makes the Cd result unusable for quantitative bounding. The Pb (mean 3.78, max 5.80 mg/kg) and Ni (mean 2.20, max 3.10 mg/kg) ranges measured here are appreciably higher than the per-kg concentrations reported in other corpus papers on the Nigerian bathing-soap matrix; whether the apparent gap reflects genuine market-and-vintage differences (Akure 2006 vs Delta 2024), method differences (dry-ashing at 500 °C vs aqua-regia hot-plate digestion), or reporting-quality differences across source venues is a Part 9 synthesis-pass question and quantitative comparators are deferred to that workflow rather than tabulated here.
• The paper does not report per-sample values, so the dataset cannot contribute to a percentile distribution of Nigerian soap/detergent metal concentrations; only the pooled minimum, maximum, mean and standard error are recoverable. Downstream pooling will use the four marginal summary statistics for each metal as a single distribution-summary record per metal per source, not as a per-sample contribution.
• The authors propose no regulatory threshold comparison, no HMTc-relevant exposure calculation, and no margin-of-safety analysis. The framing is exploratory and descriptive. No threshold values are proposed by this source page.

Limitations

• Small-journal venue with thin reporting standards. Research Journal of Environmental Toxicology (Academic Journals Inc., Lagos / Wuhan / Online), ISSN 1819-3420, has been flagged in several index-quality assessments as a low-rigour venue with limited editorial peer-review depth. The article itself shows characteristic features (no QC/recovery/LOD data, no method validation, no statistical-test naming for the “significantly higher” cluster claim, multiple grammatical and typesetting inconsistencies) consistent with that venue. Treated as C-tier evidence on the HMI ladder accordingly.
• No per-sample values. Table 2 reports only the pooled-distribution minimum, maximum, mean and SEM across all nineteen samples for each metal. The cluster statement in the discussion text (“Pb, Ni and Cr were significantly higher in [five named brands]”) therefore cannot be verified against tabulated data; the wiki reports the cluster statement only as the authors’ verbal claim with brand attribution stripped.
• No detection-limit anchor. “Cd was below detection limit” is reported with no numeric LOD value. The result is therefore qualitative: it bounds Cd at “less than whatever the SP 9 Pye Unicam achieves under these digestion conditions at 217 nm”, which cannot be translated to a regulatory-comparable ceiling.
• Non-standard AAS wavelengths reported. The reported per-element wavelengths for Cd (217 nm), Cr (262 nm) and Ni (260 nm) are not the conventional primary atomic-absorption resonance lines (Cd 228.8 nm; Cr 357.9 nm; Ni 232.0 nm). The Pb wavelength (283.3 nm) is a conventional secondary line. The paper offers no calibration or sensitivity rationale for the chosen wavelengths. If the printed wavelengths reflect actual instrument settings, the sensitivity and selectivity of the measurements for Cd, Cr and Ni are likely degraded relative to a standard configuration; if they are typesetting errors for the standard lines, the documentation is unreliable. The wiki reports the wavelengths as published and flags the issue rather than substituting external values.
• No replicate variance propagation. “Triplicate determinations” are stated to have been carried out but the within-sample variance is not reported alongside the pooled standard-error-of-the-mean statistic in Table 2.
• No certified reference material, blank, or recovery data. None reported.
• Total chromium only. The paper measures total Cr and does not distinguish Cr-VI from Cr-III, so the chromium values cannot be applied directly to Cr-VI threshold-setting.
• Ambiguous correlation reporting. The paper reports correlation coefficients (Pb r = 0.969, Ni r = 0.835, Cr r = 0.806) without naming the second variable in the correlation. A literal reading is correlation between metal-rank and sample-index, but the authors do not state this and the wiki does not impute the second variable.
• Brand-by-brand attribution in the discussion section. Five specific brand names are listed in the discussion as having “significantly higher” Pb/Ni/Cr concentrations. The wiki strips brand attribution per Part 12 and reports only the product-form-and-provenance cluster (“a subset of industrial toilet soaps and industrial powdered detergents”). The underlying per-brand values are not tabulated in the source and cannot be recovered.
• Folder-vs-content mismatch. The source PDF was filed in raw/manual-fetch/Kimi_Agent_Download Corruption Issue/household_papers/02_Toilet_Bathroom_Descaler/, but the paper is about soaps and laundry detergents, not toilet-bowl / bathroom descalers. The folder-level mis-categorisation is a manual-fetch-tracker note, not a content defect.

Provenance

• Source PDF: raw/manual-fetch/Kimi_Agent_Download Corruption Issue/household_papers/02_Toilet_Bathroom_Descaler/Abulude_2007_Pb_Cd_Ni_Cr_Soaps_Detergents_Nigeria.pdf
• SHA-256: e37c78817569980591b0a844d9e1e2f77c5dabc641c7cb47545f3ea12a5628b5
• File size: 201 KB; 4 pages (cover, abstract+methods p. 102, results p. 103, references p. 104).
• Journal: Research Journal of Environmental Toxicology, Volume 1, Issue 2, pages 102–104 (Academic Journals Inc.; ISSN 1819-3420).
• Publication date: 2007.
• DOI: none assigned (Academic Journals Inc. did not mint Crossref DOIs for this article as of access date).
• License: open-access per Academic Journals Inc. policy; the journal publishes under terms permitting redistribution with attribution. Quality of editorial oversight is independently disputed (see Limitations); the access terms themselves are open.
• Access date: 2026-06-03.
• Acquisition path: included in Kimi Agent Download Corruption Issue (KADC) folder household_papers/02_Toilet_Bathroom_Descaler/. (Folder naming reflects raw-asset provenance, not topical fit — the paper is a soap-and-detergent study filed alongside toilet-and-bathroom-descaler literature in the KADC batch.)

Wiki pages this source may touch

• lead — Nigerian soap-and-detergent Pb 0.0 (ND) – 5.80 mg/kg, mean 3.78 mg/kg on a dry-ashed, as-purchased basis (n = 19, pooled distribution).
• cadmium — Cd below the (un-quantified) detection limit in all 19 Akure soap-and-detergent samples; qualitative-only contribution.
• nickel — Nigerian soap-and-detergent Ni 1.30 – 3.10 mg/kg, mean 2.20 mg/kg on a dry-ashed, as-purchased basis (n = 19, pooled distribution).
• chromium — Nigerian soap-and-detergent total-Cr 0.50 – 2.70 mg/kg, mean 1.33 mg/kg on a dry-ashed, as-purchased basis (n = 19, pooled distribution). Total Cr only; not Cr-VI-specific.
• bathing-soaps — direct evidence for the toilet-soap and hand-mould subset (n = 13) within the pooled distribution; per-sample values not recoverable.
• hand-soap — broad-product-context for hand-washing toilet-soap subset (n = 12 toilet soaps).
• laundry-detergents — direct evidence for the powdered- and liquid-detergent subset (n = 4); per-sample values not recoverable.

Verification notes

• Identity-check results (DOI / raw_handle / cite-key) on 2026-06-03 returned no matching wiki source page; ingested as NEW.
• SHA-256 of the source PDF was computed from disk on 2026-06-03 and recorded above.
• The reported AAS wavelengths for Cd, Cr and Ni do not match the standard primary atomic-absorption lines (Cd 228.8 nm; Cr 357.9 nm; Ni 232.0 nm). Verified by reading the Materials and Methods section verbatim (“Each value was recorded at 217 nm for Cd, 262 nm for Cr, 283.3 nm for Pb and 260 nm for Ni (Mercedes, 2002)”). The wiki reports the printed values as the authors stated them, with the discrepancy flagged under Limitations; no external substitution is performed.
• Pb min = ND, max = 5.80 mg/kg, mean = 3.78 mg/kg, SEM = 1.31 mg/kg per Table 2. Ni min = 1.30, max = 3.10, mean = 2.20, SEM = 0.50 mg/kg per Table 2. Cr min = 0.50, max = 2.70, mean = 1.33, SEM = 0.61 mg/kg per Table 2. Cd reported as ND with no numeric LOD anchor. All values cross-checked against the abstract’s narrative restatement (“The mean results of the nineteen samples ranged thus: 3.78, 2.20 and 1.33 mg kg⁻¹ of Pb, Ni and Cr, respectively”).
• Five specific brand names appear in the discussion section as having “significantly higher” Pb/Ni/Cr concentrations. Per Part 12 strict reading (locked 2026-05-17), the brand names were stripped and replaced with the product-form-and-provenance descriptor “a subset of industrial toilet soaps and industrial powdered detergents”. The underlying per-brand values are not tabulated in the source; the cluster statement is preserved as a verbal claim with attribution removed, not as quantitative data.
• Brand names in Table 1 (the sample-inventory table) were similarly suppressed; the sample-composition breakdown above retains only the four-axis type / provenance / package descriptors, not the brand labels. Product-form-and-provenance descriptors retain the analytical-chemistry-relevant information (cellophane vs paper vs plastic packaging; industrial vs imported vs local provenance; toilet vs bar vs hand-mould vs detergent form) without naming brands.
• Mercedes (2002) is cited in the methods section as the reference for the AAS wavelengths and sample-digestion protocol. The Mercedes (2002) reference in the References list is “Mercedes, R., 2002. Heavy metals and arsenic uptake by wild vegetation in the Guadiamar river area after the toxic spill of the Aznalcollar mine. J. Biotech., 98:125-137” — this is a soil/vegetation paper, not a soap/detergent or AAS methods paper, so the citation as a methods reference is itself anomalous. Flagged here for awareness; no wiki action.
• Evidence tier set to C on the basis of: (i) Academic Journals Inc.’s documented index-quality concerns; (ii) n = 19, single-city, single-laboratory, single-batch design with no per-sample values reported; (iii) absence of QC / recovery / LOD / CRM / blank / detection-limit reporting; (iv) non-standard AAS wavelengths for three of four analytes; (v) total Cr only, not Cr-VI; (vi) brand-attribution and unsubstantiated “significantly higher” cluster claim in the discussion. The paper’s basic analytical chemistry (dry ash + acid dissolution + AAS) is conventional and not in itself a tier downgrade.
• Folder-vs-content mismatch (paper filed under 02_Toilet_Bathroom_Descaler/ but content is soaps and detergents) recorded for the manual-fetch tracker; no wiki action required.
• Audit subagent (2026-06-03) flagged the sample-composition breakdown as not matching Table 1 (Check 1, ❌). Verified independently against the Table 1 image on PDF p. 103: the initial ingest mis-grouped four entries — Klin is “Detergent / Industrial” not a toilet soap, USA is “Toilet soap / Imported” not a detergent, and the unnamed Table 1 row alongside Teepol is a second imported liquid detergent (n=2, not n=1). Corrected sample-composition table to: toilet soaps (4 industrial + 4 imported + 1 local = 9); bar soaps (1 industrial + 2 local = 3); hand-mould (1 local); powdered detergents (4 industrial); liquid detergents (2 imported). Total 19 ✓. Updated the sample_population frontmatter narrative to match. The discussion-narrative “significantly higher” cluster (5 brand names = 2 industrial toilet soaps + 3 industrial powdered detergents) remains correctly described as “a subset of industrial toilet soaps and industrial powdered detergents” without specific count attribution.
• Audit subagent (2026-06-03) flagged Implications bullet 2 as carrying cross-source quantitative comparators that belong in the Part 9 synthesis pass, not on the source page (Check 5, ⚠️). Verified — the Oviri 2024 and Barkodia 2020 ppb ranges and the structured causal-explanation hypotheses (a/b/c) constitute synthesis work. Stripped the specific comparator values and the explanatory hypotheses; rewrote as a brief qualitative observation that the Akure 2006 Pb/Ni ranges are appreciably higher than other Nigerian bathing-soap corpus papers and deferring quantitative cross-source resolution to Part 9.
• Audit subagent (2026-06-03) flagged products/bathing-soaps as missing from docs/gpt-collaboration/taxonomy-snapshot.md (Check 2, ❌). Verified — the slug does not appear in the snapshot file. However, the live wiki has a provisional bathing-soaps product page (wiki/products/bathing-soaps.md carrying provisional_scaffold: true, created 2026-06-02 by the oviri2024-nigeria-bathing-soaps-metals ingest, which routes to the same slug). The routing audit completes cleanly for this source (0 unresolved). Finding rejected as a false positive driven by an out-of-date taxonomy snapshot; no slug change applied. Surfaces a snapshot-refresh task (Karen-side) but not a wiki-page defect.
• Audit subagent (2026-06-03) flagged matrices: [cosmetic-personal-care, laundry-detergent, household-cleaning-product] as not appearing in the taxonomy snapshot’s enumerated examples (Check 2, ⚠️). Verified — the snapshot lists products slugs but does not enumerate a matrices controlled vocabulary; the three matrices in use here all have prior-art precedents in the corpus (cosmetic-personal-care in balogun2024-body-creams-delta-nigeria and many other cosmetics papers; laundry-detergent and household-cleaning-product in barkodia2020-jhansi-detergent-heavy-metals and ecetoc1992-tr045-ni-co-cr-consumer-products). Finding rejected as a false positive on the precedent basis; no matrix change applied.
• Audit subagent (2026-06-03) confirmed ✅ on the numerical fidelity of Table 2 statistics (Pb / Cd / Ni / Cr min/max/mean/SEM), the correlation coefficients (Pb 0.969, Ni 0.835, Cr 0.806), sample size n=19, methods (dry-ashing at 500 °C for 1 h, 2 M HCl dissolution to 50 cm³, SP 9 Pye Unicam AAS, triplicate determinations, non-standard wavelengths for Cd/Cr/Ni flagged correctly without external substitution), brand-firewall stripping (Imperial Leather, Omo, Premier, Klin, Elephant Extra all removed from body; no Table 1 brand names leaked into wiki), and the absence of HMTc threshold proposals or consumer advisories.
• Audit verdict: REVISE. Three findings applied (sample-composition table correction; sample_population narrative correction; Implications bullet 2 cross-source comparator strip). Two findings rejected as false positives (bathing-soaps slug; matrices vocabulary).

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.