Anselm et al. 2022 — Artificial-sweat formulation and temperature effects on dermal bioaccessibility of metals in e-waste-contaminated soil
This Geosciences paper compares five artificial-sweat formulations (BSI/EN 1811:2011, NIHS 96-10, Altkofer, Ariza, Cheng) at five temperatures (17-47 °C) for dermal bioaccessibility of As, Cd, Cr, Cu, Fe, Mn, Ni, Pb, Zn in <45 µm topsoil from a Lagos, Nigeria e-waste recycling site (Alaba International Market). Inter-formulation variability was the dominant signal: at 37 °C the NIHS 96-10 formulation extracted at least an order of magnitude more of every analyte than any other formulation except As and Cu, and 1700-fold more Pb than the next-most-effective extractant. The order of release was NIHS 96-10 > CHE > ALT > ARI > BSI > DIW. A general trend of increasing release with temperature was observed for all analytes except As, with statistically significant differences (p < 0.05) for As, Cd, Mn and Pb between the lowest temperatures (17 or 27 °C) and 47 °C; no significant difference was found between 32 °C and 37 °C for any analyte. The paper establishes that dermal-bioaccessibility estimates are critically dependent on the artificial-sweat recipe and that ambient temperatures encountered at tropical e-waste sites can exceed the 32-37 °C range typically used in bioaccessibility tests.
Key numbers
Test substrate (Table 3, mg/kg dry weight, n=3)
Pseudototal PTE concentrations in the bulk composite <45 µm topsoil from Alaba International Market by aqua regia digestion:
| Analyte | Found (mg/kg) | World soil [Alloway 2013] | Background [SQuiRT] | Dutch target | Dutch intervention | Contamination factor |
|---|---|---|---|---|---|---|
| As | 10.1 ± 1.8 | 0.1 | 5.2 | 29 | 55 | 2 (moderate) |
| Cd | 78.3 ± 5.4 | 1.1 | 0.87 | 0.8 | 12 | 90 (very high) |
| Cr | 88.6 ± 12.0 | 42 | <37 | 100 | 380 | 2 (moderate) |
| Cu | 13,700 ± 843 | 14 | 17 | 36 | 190 | 806 (very high) |
| Fe | 73,500 ± 5260 | — | 18,000 | — | — | 4 (considerable) |
| Mn | 1120 ± 92 | 418 | 330 | — | — | 34 (very high) |
| Ni | 355 ± 26 | 18 | 13 | 35 | 210 | 27 (very high) |
| Pb | 3340 ± 132 | 25 | 16 | 85 | 530 | 209 (very high) |
| Zn | 11,100 ± 618 | 62 | 48 | 140 | 720 | 231 (very high) |
Soil pH 7.98; organic matter 21.0% (loss on ignition at 550 °C). Contamination factor CF = pseudototal / background; CF >6 = very high contamination (Hakanson 1980).
Bioaccessible concentrations at 37 °C (Table 4, mg/kg, mean ± SD, n=3)
| DIW | BSI | NIHS | ALT | ARI | CHE | |
|---|---|---|---|---|---|---|
| As (mg/kg) | 0.0419 ± 0.0361 | 0.0939 ± 0.0016 | 1.17 ± 0.06 | 0.134 ± 0.013 | 0.0930 ± 0.0107 | 0.140 ± 0.001 |
| Cd (mg/kg) | 0.000924 ± 0.001401 | 0.479 ± 0.036 | 19.7 ± 0.8 | 1.36 ± 0.05 | 1.32 ± 0.06 | 0.779 ± 0.013 |
| Cr (mg/kg) | 0.0379 ± 0.0298 | 0.0763 ± 0.0108 | 2.59 ± 0.08 | 0.114 ± 0.011 | 0.0321 ± 0.0122 | 0.229 ± 0.009 |
| Cu (mg/kg) | 8.77 ± 6.52 | 87.4 ± 1.8 | 4540 ± 62 | 405 ± 28 | 132 ± 3 | 717 ± 11 |
| Fe (mg/kg) | 4.37 ± 3.35 | 12.5 ± 0.4 | 859 ± 7 | 31.2 ± 1.6 | 21.5 ± 0.4 | 38.6 ± 3.8 |
| Mn (mg/kg) | 0.545 ± 0.413 | 8.61 ± 0.30 | 242 ± 0 | 24.2 ± 1.1 | 20.0 ± 0.5 | 15.8 ± 0.1 |
| Ni (mg/kg) | 0.0749 ± 0.0568 | 0.711 ± 0.0223 | 42.1 ± 0.5 | 2.74 ± 0.11 | 1.86 ± 0.05 | 2.26 ± 0.03 |
| Pb (mg/kg) | 0.0580 ± 0.0372 | 0.155 ± 0.019 | 579 ± 3 | 0.343 ± 0.162 | 0.185 ± 0.157 | 0.109 ± 0.010 |
| Zn (mg/kg) | 0.0478 ± 1.3222 | 9.98 ± 6.21 | 3190 ± 19 | 157 ± 18 | 69.3 ± 5.8 | 34.2 ± 0.3 |
% bioaccessibility at 37 °C (Table 4, %)
| DIW | BSI | NIHS | ALT | ARI | CHE | |
|---|---|---|---|---|---|---|
| As | 0.415 | 0.929 | 11.5 | 1.32 | 0.920 | 1.39 |
| Cd | 0.00118 | 0.611 | 25.1 | 1.74 | 1.69 | 0.995 |
| Cr | 0.0428 | 0.0860 | 2.93 | 0.128 | 0.0362 | 0.259 |
| Cu | 0.0640 | 0.638 | 33.1 | 2.95 | 0.962 | 5.23 |
| Fe | 0.00595 | 0.0170 | 1.17 | 0.0425 | 0.0293 | 0.0525 |
| Mn | 0.0486 | 0.769 | 21.6 | 2.16 | 1.79 | 1.41 |
| Ni | 0.0211 | 0.200 | 11.9 | 0.773 | 0.524 | 0.637 |
| Pb | 0.00173 | 0.00466 | 17.3 | 0.0103 | 0.00553 | 0.00327 |
| Zn | 0.000431 | 0.0900 | 28.7 | 1.41 | 0.625 | 0.309 |
Headline magnitude finding (Section 3.3): NIHS 96-10 gave 17.3 % Pb bioaccessibility, while every other formulation gave <0.011 % — at least three orders of magnitude lower for Pb. Across all analytes except As and Cu, NIHS 96-10 was an order of magnitude or more above the next-most-effective extractant. Cu and Zn were the most bioaccessible elements overall; As, Cd, Cr, Ni and Pb gave low bioaccessible concentrations (<5 mg/kg) in all formulations except NIHS 96-10.
Solution pH before and after soil contact (Table 5)
| Solution | pH before use | pH of soil extract |
|---|---|---|
| DIW | 5.00 | 7.96 |
| BSI | 6.48 | 7.67 |
| NIHS | 4.68 | 5.41 |
| ALT | 5.30 | 7.43 |
| ARI | 4.49 | 7.34 |
| CHE | 6.00 | 8.00 |
NIHS 96-10 retained the lowest extract pH after soil contact (5.41 vs 7.34-8.00 for other formulations); authors attribute this to the higher concentrations of acetic and lactic acid giving NIHS greater buffering capacity against the slightly alkaline soil.
Temperature dependence (Section 3.4)
ANOVA across the five temperatures showed:
- No statistically significant temperature dependence (p > 0.05) for Cr, Cu, Fe, Ni or Zn.
- Statistically significant difference (p < 0.05) between the lowest (17 or 27 °C) and highest (47 °C) temperatures for As, Cd, Mn and Pb.
- No statistically significant difference between 32 °C and 37 °C for any analyte or formulation.
- As behaved opposite the other analytes: largest amounts released at 17 °C, similar release between 27 and 47 °C for a given formulation.
- Greatest temperature sensitivity: Cd and Mn (almost all formulations); Cr and Pb under the ALT formulation; Pb under ALT at 47 °C (visible spike in Figure 3); As under all formulations (decreasing with temperature).
Lagos monthly average air temperatures 29-34 °C, with reported maxima up to 38 °C; authors note dermal bioaccessibility tests in the 32-37 °C range may underestimate exposure for workers at Alaba and other warm-climate metal-processing sites, particularly given that skin permeability also increases at higher temperatures (Murthy et al. 2004).
Sweat-formulation composition (Table 1, grams reagent per 100 mL unless stated)
| Reagent | BSI (EN 1811:2011) | NIHS 96-10 | ALT (Altkofer) | ARI (Ariza) | CHE (Cheng) |
|---|---|---|---|---|---|
| Acetic acid | — | 0.25 | — | — | — |
| Ammonia | — | — | — | — | 0.0343 |
| Ammonium chloride | — | 1.75 | 0.04 | — | — |
| Lactic acid | 0.1 | 1.5 | 0.3 | 0.1 mL | — |
| Potassium chloride | — | — | 0.03 | 0.12 | — |
| Sodium chloride | 0.5 | 2.0 | 0.45 | 0.75 | 0.468 |
| Sodium lactate | — | — | — | — | 0.6 |
| Sodium sulfate | — | — | 0.03 | — | — |
| Urea | 0.1 | 0.5 | 0.02 | 0.1 | 0.0516 |
| Initial pH | 6.5 | 4.7 | 5.3 | 4.5 | 6.0 |
Extraction protocol
- 0.1 g soil + 10 mL artificial sweat (or DIW control) in flat-bottom HDPE tube.
- Soil-to-extractant ratio 1:100 (based on 1 mg soil per cm² skin loading and 0.1 mL sweat per cm² skin volume).
- 100 rpm orbital incubator (Stuart SI500, Cole-Parmer), 8 h, at 17, 27, 32, 37 or 47 °C; triplicate.
- Centrifuged 4200 g for 15 min; supernatant filtered through 0.45 µm Acrodisc cellulose acetate (Sigma Aldrich); syringes and filters pre-washed with 0.032 M HCl to remove trace Cu and Zn.
- Filtrates diluted to 2% HNO3; stored at 4 °C in HDPE prior to ICP-MS.
Quality control
- Pseudototal digestion: GLA URM secondary reference material (EU URBSOIL EVK4-CT-2001-00053), all nine analytes within two standard deviations of indicative values, precision <12% (n=3); recoveries 81.1-102 %.
- Spike recoveries on sweat solutions without soil (mean ± SD, n=3, across 17/27/32 °C): As 92.6 ± 3.7; Cd 88.6 ± 4.2; Cr 89.7 ± 5.9; Cu 86.2 ± 3.8; Fe 89.7 ± 2.6; Mn 91.7 ± 4.0; Ni 81.2 ± 5.7; Pb 95.7 ± 5.7; Zn 87.2 ± 10 %. Spike levels: 10,000 µg/L Fe; 250 µg/L for all other analytes.
- Authors note Ni solubility in NIHS 96-10 appeared to decrease slightly in the presence of extraction reagents (consistent with Leal et al. 2018).
Dutch intervention exceedance under NIHS 96-10 alone (Section 3.3)
At 37 °C, the dermal bioaccessible concentrations alone — without considering total soil concentrations — exceeded Dutch Intervention Values for total soil PTE for Cd (19.7 mg/kg > 12), Cu (4540 mg/kg > 190), Pb (579 mg/kg > 530) and Zn (3190 mg/kg > 720). Authors flag this as evidence that further investigation of dermal bioaccessibility at the Alaba site is “urgently needed.”
Methods (brief)
Bulk composite topsoil (0-5 cm) sampled 2016-2017 from the main recycling area of Alaba International Market in Lagos, Nigeria (West Africa’s largest e-waste recycling site, established 1978, receiving an estimated five million electronic items per day from Europe and Asia). Soil sieved to ≤45 µm with a BS410/1986 sieve (Endecotts Ltd., London, UK). pH determined by BSI method (BS ISO 10390:2005) at 1:5 soil:water; organic matter by loss on ignition at 550 °C.
Pseudototal element concentrations by microwave-assisted aqua regia digestion in a MARS Xpress system (CEM, Buckingham, UK): triplicate 0.5 g portions digested in 20 mL freshly prepared aqua regia, 800 W, ramp from room temperature to 160 °C in 30 min, hold at 160 °C for 20 min. Contamination factor calculated against SQuiRT screening reference values (Buchman 2008) and Cd background from Isimekhai et al. 2017.
Dermal bioaccessibility tests with five artificial-sweat formulations (BSI/EN 1811:2011, NIHS 96-10, ALT, ARI, CHE) plus deionized water control, at five temperatures (17, 27, 32, 37, 47 °C), 8 h, triplicate. ICP-MS analysis on an Agilent 7700× with ASX-500 autosampler (Agilent Technologies Ltd., Cheshire, UK), 1550 W and 27 MHz. Calibrants in 2% nitric acid from commercial 10 mg/L multi-element standard and 1000 mg/L Fe solution (Qmx Laboratories, Thaxted, UK). Quantification on ⁷⁵As, ¹¹⁴Cd, ⁵³Cr, ⁶⁵Cu, ⁵⁷Fe, ⁵⁵Mn, ⁶⁰Ni, ²⁰⁸Pb, ⁶⁶Zn; cross-checked against a second isotope where available. Internal standard ¹¹⁵In.
Statistical analysis in Minitab Version 18: ANOVA across temperature and formulation, with Tukey pairwise comparison where significant.
Limitations stated/implied: static 8 h batch extraction (no sequential or kinetic profiling); single bulk composite soil sample (no inter-site or intra-site spatial variation captured); only inorganic-extracting reactants tested (no synthetic sebum or sweat-lipid components); no certified reference material is available for PTE extractable by these dermal bioaccessibility tests, so QC for the bioaccessibility step is restricted to spike recoveries on spiked sweat solutions without soil.
Implications
- Certification (HMTc). The 17.3 % vs <0.011 % Pb-bioaccessibility split between NIHS 96-10 and every other formulation makes the ”% bioaccessible” of Pb in soil functionally unreportable without specifying the recipe. HMTc work that incorporates dermal-bioaccessibility corrections — for soil-tracked-into-home pathways, for outdoor-play scenarios, or for any product that may carry soil residues — must lock the artificial-sweat recipe and the extraction temperature, otherwise the same physical sample produces certifications differing by orders of magnitude.
- Courses. Teaching reference for occupational and environmental health course modules on inter-laboratory inconsistency in bioaccessibility testing and on the methodological assumptions buried in regulatory “safe” thresholds.
- App. Not relevant to ingredient
contamination_profileblocks (no food matrix). Relevant background if a future children’s outdoor-play module incorporates dermal exposure to soil residues; the temperature dependence (32-37 °C standard underestimates tropical-climate exposure) would need to be carried into any climate-aware exposure model.
Wiki pages this source may touch
Verification notes
- Single-sample study (n=1 soil composite); the dominant findings are formulation and temperature effects on dermal bioaccessibility, not site-specific exposure estimates. Numerical concentrations in this page are reported for the one bulk composite, not as population estimates.
products: []andingredients: []are correct for this paper: it is a soil bioaccessibility methodology study, not a food or personal-care product study. The matrix is soil from an e-waste site; the artificial sweats are extractants, not sampled matrices. The advisory routing-malformed entry on this source reflects empty products/ingredients lists; not a frontmatter defect for a methodology paper.- Prior page revision (2026-05-17 earlier) included
artificial-sweatas a matrix; corrected to[soil]because matrices are sampled matrices (the soil), not extractants (the sweat solutions). - Prior page revision included only abstract-level claims in Key numbers; expanded to include Table 3 pseudototal concentrations, Table 4 bioaccessible concentrations and %BA values across all five formulations at 37 °C, Table 5 pH values, the Section 3.4 statistical analysis results, and the spike-recovery QC. This is required for downstream HMTc threshold work and for the Cochrane bar (exact figures, not paraphrases).
- Speciation: paper measured total As (⁷⁵As by ICP-MS), no inorganic/total split; total Cr (⁵³Cr), no Cr-VI split. Metals tagged
tAs(per CLAUDE.md Part 14 speciation discipline: when paper does not separate iAs/tAs, usetAs) and bareCr(no separatetCrconvention). - Audit subagent (2026-05-17) flagged bare
AsvstAsand missing[[metals/iron]]wikilink; both verified against source —Ascorrected totAsin frontmatter;[[metals/iron]]added to “Wiki pages this source may touch” (Fe is reported in Tables 3 and 4 alongside the other eight analytes). - Brand-firewall compliance: method-section vendor names (Agilent ICP-MS, CEM MARS Xpress digester, Sigma Aldrich Acrodisc filters, Stuart SI500 incubator, Qmx Laboratories calibrants, Minitab software) retained per Part 12 Exception 2 (scientific reproducibility). No sampled-product brand names present.
- Provisional ingredient/product scaffolds not applicable (no product/ingredient declarations).
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.