Potortì et al. 2026 — Chemical composition and safety of commercial tahini (Italy)

This characterisation of seven commercial tahini brands sold on the Italian market is the first to assess toxic and potentially toxic inorganic elements — Cd, Pb, As, Hg, Ni — alongside proximate composition, fatty acid profile, sterol profile, and essential minerals. One sample (T-1, Turkey-origin organic hulled-seed tahini) substantially exceeded EU Regulation 2023/915 limits for both Cd (0.35 mg/kg vs. the 0.10 mg/kg oilseeds limit, 3.5× over) and Pb (0.50 mg/kg, five times the 0.10 mg/kg fats-and-oils limit) and was the only sample with detectable arsenic (tAs 0.20 mg/kg). The remaining six brands were at or near LOQ (0.010 mg/kg) for Cd, Pb, and As. Total mercury was below LOQ in every sample. Nickel was quantified in all seven brands at 0.02–0.10 mg/kg.

Key numbers

Toxic and potentially toxic elements (mg/kg, n = 7 brands, 3 analytical replicates each; Table 6, p. 15):

Be, Li, Cr, Mo, Sn, Sb were below LOQ in every sample (Table 6).

Regulatory anchors cited by the authors (Section 3.5, p. 17): EU Reg. 2023/915 sets 0.10 mg/kg Cd in oilseeds and 0.10 mg/kg Pb for fats and oils (tahini classified as a high-lipid food). T-1 exceeded both; the authors report T-1 Pb was “five times higher than the regulatory limit.” Method detection / quantification limits for Cr, Mo, Cd, Pb, As, Hg, and Ni: LOD 0.003 mg/kg and LOQ 0.010 mg/kg (Section 3.5, p. 14). NIST SRM 1570a spinach leaves served as certified reference material; Cd recovery 100.75%, K recovery 91.15% (lowest); intraday RSD < 1.2%, interday RSD < 1.3% (Table S1, summarised p. 14).

Macro- and trace-element ranges across the seven brands (mg/kg, Table 6): P 7,951.51–9,434.88; K 4,159.36–5,191.27; Ca 1,299.89–5,105.53; Mg 2,094.11–2,925.25; Na 30.95–1,986.47 (T-4 lowest, T-1 highest — wide spread the authors attribute to NaCl addition during processing); Fe 47.40–83.13; Zn 54.75–86.80; Cu 9.59–15.00; Mn 13.51–21.03; Se 0.72–1.11. Order of macro-element abundance for most brands: P > K > Ca > Mg > Na.

Mercury note: tHg was measured by direct mercury analyser (DMA-80, thermal decomposition / amalgamation / atomic absorption at 253.7 nm), not by ICP-MS. No methylmercury speciation was performed. All seven samples returned <LOQ (0.010 mg/kg).

Arsenic note: total arsenic was determined by ICP-MS on isotope 75As. The study does not perform inorganic-arsenic speciation; the reported 0.20 mg/kg for T-1 is tAs.

Exposure estimate (Table 7, Section 3.5, p. 18): authors compute element uptake assuming 1 g tahini/day (FAOSTAT sesame-seed proxy, no tahini-specific intake reference) and 70 kg body weight, benchmarked against TDI/TWI/BMDL01 values (As 0.3 µg/kg-bw/day, Hg 4 µg/kg-bw/day, Ni 22 µg/kg-bw/day, Pb 2.5 µg/kg-bw/week, Cd 2.5 µg/kg-bw/day from EFSA/JECFA). At 1 g/day, uptake of every quantified toxic element is ≤1% of the reference value for every brand. Authors explicitly flag the 1 g/day assumption as a methodological limitation — actual tahini consumption can exceed this — and call for tahini-specific intake data.

Methods (brief)

Seven commercial tahini brands purchased online in 2025 for the Italian retail market; stored at 4 °C until analysis; n = 3 analytical replicates per brand. Microwave acid digestion in ETHOS 1 mineraliser (Milestone, Bergamo, Italy) with 1 mL 30 % H₂O₂ and 7 mL 69 % HNO₃, heating 0 → 200 °C over 10 min, 200 °C hold 20 min, 20 min cool, then dilution and 0.45 µm filtration. ICP-OES (HORIBA ULTIMA 2) for Ca, Fe, K, Mg, Na, Zn, P (P at 213.618 nm); ICP-MS (Thermo iCAP-Q) for Li, Be, V, Cr, Mn, Co, Ni, Cu, As, Se, Mo, Cd, Sn, Sb, Pb (isotopes ⁷Li, ⁹Be, ⁵¹V, ⁵²Cr, ⁵⁵Mn, ⁵⁹Co, ⁶⁰Ni, ⁶³Cu, ⁷⁵As, ⁸⁰Se, ⁹⁸Mo, ¹¹¹Cd, ¹²⁰Sn, ¹²¹Sb, ²⁰⁸Pb); DMA-80 (Milestone) for total Hg via thermal decomposition / amalgamation / AAS at 253.7 nm. Seven-point calibration curves 0.5–50 µg/L for ICP-MS, 1–100 µg/L for Hg, R² > 0.9990 across all elements. Certified reference material NIST SRM 1570a (spinach leaves), spiked with 5 mg/kg Fe and 2 mg/kg Cr/Mo/Be/Li/Sn/Sb where absent. Recovery range 91.15 % (K) to 100.75 % (Cd). Intraday/interday RSD < 1.2 %/1.3 %. Statistical analysis: SPSS 19.0, one-way ANOVA at p < 0.05; analytes below LOQ in <50 % of samples excluded from ANOVA; in the few below-LOQ cases reported in tables, LOD/2 was assigned. Main limitations stated by the authors: small sample size (n = 7 brands, single purchase round, no lot-to-lot replication), and the 1 g/day intake assumption borrowed from FAOSTAT sesame-seed consumption rather than measured tahini intake.

Implications

Certification: T-1’s Cd (0.35 mg/kg, 3.5× the EU 2023/915 oilseed limit) and Pb (0.50 mg/kg, 5× the EU 2023/915 fats-and-oils limit) make tahini a category where occasional supplier-level exceedances can be order-of-magnitude events, while six of seven brands sat at or below LOQ for Cd and Pb. The data set is small but supports treating tahini as a category where origin/supplier screening rather than category-wide assumptions is the prudent posture. tAs was detectable only in T-1; tHg never. Ni was uniformly low.

Courses: First comprehensive contaminant screen for tahini sold in Italy. Illustrates that “organic” labelling did not predict lower heavy-metal content in this small dataset — T-1, the only sample exceeding any EU limit, was the certified-organic Turkey-origin brand. Useful teaching example for the soil-uptake mechanism in sesame and for the limits of certification-as-proxy for contamination.

App: Adds direct concentration data for Cd, Pb, tAs (only T-1 quantified), tHg (all <LOQ), and Ni in commercial tahini. Distribution is strongly right-skewed with one high outlier across both Cd and Pb. For category-level occurrence modelling, use the full seven-brand distribution including the LOQ values rather than the maximum alone; for outlier-risk modelling, the T-1 contamination pattern (high Cd + high Pb + detectable As in the same brand) is the signal.

Wiki pages updated on ingest

• lead
• cadmium
• arsenic-total
• mercury-total
• nickel
• sesame
• tahini
• nut-seed-butters-other
• eu-2023-915-cadmium
• eu2023-contaminants-maximum-levels

Verification notes

Merge-enhanced 2026-05-28 from prior 2026-05-14 revision. Changes: (1) corrected raw_path to actual filename (prior revision truncated the “(Sesamum indicum L.)_ Analytical,” suffix and would not resolve on disk); (2) added access_url and no_doi_assigned to align with current MFK frontmatter schema; (3) rewrote opening prose to remove the prior ambiguous “Mercury and arsenic were detectable only in T-1” wording — only arsenic was detectable in T-1, mercury was below LOQ in every sample; (4) added cross-checked exposure-estimate paragraph (Table 7) and macro/trace-element ranges (Table 6) to Key numbers; (5) added explicit Hg and As speciation notes; (6) expanded Methods (brief) with the actual digestion protocol, instrument vendors/models, isotopes monitored, calibration ranges, certified reference material details, RSD figures, and the authors’ stated limitations; (7) expanded “Wiki pages updated on ingest” to include metals, product, and the two EU 2023/915 regulation slugs the paper invokes. No brand names appear anywhere in the source (samples are coded T-1…T-7 in the paper), so no Part 12 redaction was needed.

Audit subagent (2026-05-28) flagged [[ingredients/tahini]] (and by extension [[ingredients/sesame]]) as out of the 2026-05-18 taxonomy snapshot. Verified: neither page exists yet in wiki/ingredients/, both are freq-1 references (only this source declares them), and the routing audit accepts them without unresolved entries. Auto-stub handles ingredients at freq ≥ 2, so these slugs remain in frontmatter as the correct routing destinations even though no stub yet exists. Removing [[ingredients/tahini]] per the audit’s recommendation would lose the most specific routing target for the paper (a tahini-specific characterisation, not a generic sesame study), so the slug is retained.

Audit subagent (2026-05-28) also noted that the paper’s Section 2.9 (p. 6) reports the cadmium reference value as “2.5 µg kg-b.w./day”. Verified against the PDF: the paper does print “Cd (2.5 µg kgb.w./day)” verbatim. EFSA’s actual cadmium TWI is 2.5 µg/kg-bw/week (per EFSA Scientific Opinion 2009/2011, ratified by the JECFA PTMI of 25 µg/kg-bw/month). The “/day” in the paper is a likely paper-side transcription error; if read as written, the Cd reference would be ~7× more permissive than EFSA’s actual TWI. The wiki page reproduces the paper’s number verbatim in the exposure-estimate paragraph but flags this caveat here for any downstream synthesis that pulls the Cd exposure-percentage figure.

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.