Muniz et al. 2018 — Metals in tomato sauces by packaging type, Brazil

Twenty commercial tomato sauce samples across four packaging types (plastic, metallic/canned, cellulosic, and glass) from Rio de Janeiro supermarkets were analysed by ICP-MS for As, Cd, Pb, Cr, Ni, Sb, and Sn. The study found that As, Cd, and Pb were all below Brazilian regulatory limits across all packaging types, and no significant difference between packaging types was found for most metals. The critical findings were that Cr exceeded Brazil’s maximum permissible level (0.1 mg/kg) in all packaging types (mean values 0.27–0.39 mg/kg), and Sn was significantly higher in metallic/canned packaging (0.125–0.258 mg/kg) than in plastic, cellulosic, or glass containers, with tin migration from can linings confirmed by simulant solution testing.

Key numbers

Metal concentrations by packaging type (mg/kg, n=20 total):

Key contrasts:

• Sn and Ni were the two elements with significantly different concentrations between packaging types (p<0.05). Metallic cans had 3–5x higher Sn than non-metallic packaging, and Ni in metallic packaging (0.100 ± 0.010 mg/kg) was roughly double that in plastic (0.049 ± 0.007 mg/kg). As, Cd, Pb, Cr, and Sb showed no significant packaging-type effect.
• Cr mean values were 2.7–3.9x above Brazil’s 0.1 mg/kg limit in every packaging type tested. The authors attribute this to chromium in the raw material rather than packaging migration (Cr migration from packaging simulants was below LOQ).

Packaging composition by XRF: Cr detected in plastic (155 ± 23 mg/kg), cellulosic (74 ± 4 mg/kg), and metallic can inner surface (0.05% w/w); Sn detected only in metallic can (1.32% w/w). Pb detected only in glass at very low concentration (0.003% w/w). Cd, Sb, Ni not detected in any packaging.

Simulant solution migration (GF-AAS): Sn migration from metallic can simulant: 0.067 mg/L. Cr migration from all packaging types: below LOQ (5 µg/L). Conclusion: Sn in canned tomato sauce comes from can lining dissolution; Cr contamination originates in the raw material, not from packaging.

Analytical methods: ICP-MS (Nexion 300D Perkin Elmer) for food samples, LOD/LOQ: As 0.095/0.32, Cd 0.014/0.047, Pb 0.15/0.50, Cr 0.034/0.11, Ni 0.40/1.33, Sb 0.18/0.61, Sn 0.17/0.57 µg/L. GF-AAS (PinAAcle 900Z Perkin Elmer) for simulant solutions. XRF (XL3t Thermo Scientific) for packaging composition.

Recovery ranges: As 74–94%; Cd 84–88%; Pb 77–93%; Cr 93–95%; Ni 102–106%; Sb 86–92%; Sn 54–83%.

Note on As: ICP-MS measured total arsenic (tAs). No speciation performed; iAs fraction not reported.

Methods (brief)

Cross-sectional analytical study. Twenty tomato sauce samples in four packaging types from Rio de Janeiro supermarkets (two brands: brand A with 12 samples in plastic/cellulosic/metallic; brand B with 8 samples in plastic/metallic/glass). Microwave digestion with HNO3 65%; ICP-MS using isotopes 75As, 111Cd, 208Pb, 52Cr, 60Ni, 121Sb, 118Sn, internal standard Rh at 10 µg/L. Simulant solution tests per Brazilian ANVISA legislation for each packaging type (acetic acid/NaCl/sucrose/tartaric acid as appropriate). XRF for packaging composition. No brand differences found for same packaging type, so results pooled by packaging. Recovery for Sn (54–83%) was at the low end; Sn values should be interpreted with some caution.

Implications

Certification: Cr in tomato sauce consistently exceeded the Brazilian ANVISA limit for Cr in food (0.1 mg/kg) across all four packaging types studied (mean 0.276–0.39 mg/kg). EU Regulation 1881/2006 sets ML values for fresh and raw vegetables but does not list a Cr ML specific to tomato sauce or other processed tomato products. Sn in tomato sauces stored in cans (mean 0.186 mg/kg; range 0.125–0.258 mg/kg) was 3–5x higher than Sn in non-metallic packaging and was the only metal whose simulant-solution test confirmed migration from packaging (0.067 mg/L from cans). Even so, the highest measured Sn (0.258 mg/kg) is roughly three orders of magnitude below the 250 mg/kg Brazilian tolerable limit for Sn. Concentrations of Cd, Pb, As, Ni, and Sb were below all applicable Brazilian and EU ML values in every sample.

Courses: Illustrates the packaging-migration angle (Sn from cans, Cr from raw material) versus contamination-at-source distinction. The XRF packaging analysis plus simulant testing is a methodological teaching case for migration studies.

App: Tomato sauce Cd: ~0.004–0.006 mg/kg (4–6 µg/kg); Pb: ~0.016–0.049 mg/kg (16–49 µg/kg); Cr (total): ~0.27–0.39 mg/kg (270–390 µg/kg); Sn: ~0.020–0.258 mg/kg depending on packaging (canned 125–258 µg/kg). Note: total As ~0.029–0.035 mg/kg; no iAs speciation.

Wiki pages updated on ingest

• tomato
• tomato-sauce
• food-packaging

Verification notes

2026-05-28 enhance pass (Claude): re-read source PDF in full and reconciled frontmatter to current schema. License corrected from “unknown” to “CC BY” — the Food Sci. Technol. (Campinas) issue page bears a Creative Commons BY logo on page 1 of the article. Added access_url (DOI URL) and no_doi_assigned: false. Fixed an internal Key-numbers contradiction in the previous revision: bullet 1 claimed “Sn was the only element with significantly different concentrations” while bullet 2 stated Ni was also significantly higher in metallic packaging; the abstract is explicit that both Sn and Ni differed significantly (p<0.05), with all other elements showing no packaging-type effect. Bullet rewritten to match. All numerical values in the Key numbers table were verified against Table 5 of the source and left unchanged.

Note on Cr: the source prose on p. 387 reports “the lowest mean value found was 0.27 mg kg⁻¹ in products stored in cellulosic packaging” while Table 5 reports cellulosic Cr = 0.322 ± 0.05 mg/kg (the actual lowest mean is metallic at 0.276). The wiki tracks Table 5 values, which are the authoritative measurements. The summary range “0.27–0.39 mg/kg” is retained because it spans both the prose figure and the table extremes and is internally defensible.

2026-05-28 audit-application pass (Claude, fresh-context subagent verdict REVISE):

• ❌ Sn Glass range corrected 0.057–0.059 → 0.057–0.058 (Table 5 reads 0.057–0.058; prior wiki had a single-digit transposition on the upper bound). Verified against source PDF page 5 Table 5.
• ⚠️ SD precision restored to source values across the Key numbers table: Cd Metallic 0.001 → 0.0008, Cd Glass 0.001 → 0.0008, Cr Glass 0.04 → 0.036 (Table 5); Cr in cellulosic packaging “74 mg/kg” → “74 ± 4 mg/kg” (Table 7).
• ⚠️ Part 2 wiki/HMTc-firewall language tightened in the Implications block: “compliance gap in European/global markets” softened to a statement of fact about EU Regulation 1881/2006’s scope, and “Cd, Pb, As pose no regulatory concerns” reworded to “below all applicable Brazilian and EU ML values” (literature-native phrasing, no risk-advisory framing).
• ⚠️ Subagent flagged matrices: [tomato-sauce, food-packaging-simulant] and [[supply-chain/food-packaging]] as not in the taxonomy-snapshot vocabulary. Rejected as false positives: the matrices field and supply-chain/* references are not enumerated in the snapshot’s controlled vocabulary (which covers ingredients, products, metals, regulations only) and function as descriptive metadata, not routing targets. No change.
• ⚠️ Subagent noted Sn row uses range-only presentation while other metals use mean ± SD. Rejected as stylistic: the source emphasizes Sn ranges in its abstract and discussion, and the range-only format mirrors that emphasis.

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.