Moussa et al. 2024 — Heavy metals in spices and herbs on the Lebanese market

Moussa et al. measured Pb, Cd, total As, and total Hg in 96 composite samples representing 13 herb and spice categories sold in Lebanon (black pepper, cinnamon, cumin, oregano, paprika, red chili, white pepper, thyme, garlic powder, thyme mix, sesame, sumac, and dried mint), purchased across four source categories: unpackaged bulk, imported (France), locally packaged with a Food Safety Management System (FSMS), and locally packaged without FSMS. Lead was the most problematic analyte, exceeding the Codex Alimentarius maximum permissible limit (MPL) of 0.3 mg/kg in 20% (19/96) of samples overall, with unpackaged bulk samples showing the highest exceedance rate (31%) and imported samples close behind (28%). Cadmium exceeded its MPL (0.2 mg/kg) in 4% (4/96) of samples; arsenic and mercury were within Codex limits across all 96 samples.

Key numbers

Lead (Pb). Range across all samples 0.052 to 0.486 mg/kg. Overall exceedance of Codex MPL (0.3 mg/kg): 20% (19/96). By source/packaging category (Table 1): bulk 31% (8/26), imported 28% (5/18), local with FSMS 12% (3/26), local without FSMS 12% (3/26). By spice/herb type (Table 2): thyme 100% (8/8), thyme mix 100% (6/6), and garlic powder 100% (8/8) above MPL; paprika 75% (6/8); cinnamon, cumin, white pepper, red chili, and oregano each 50% (4/8); dried mint 33% (2/6); sumac 33% (2/6); black pepper 25% (2/8); sesame 0% (0/6). Highest reported mean: unpacked paprika at 0.486 mg/kg; lowest: locally packed red chili without FSMS at 0.052 mg/kg. Lebanon-wide mean Pb levels differed significantly across the four packaging categories (ANOVA p = 0.0007).

Cadmium (Cd). Range 0.002 to 0.345 mg/kg. Overall exceedance of Codex MPL (0.2 mg/kg): 4% (4/96). By source/packaging category (Table 1): bulk 8% (2/26), local with FSMS 4% (1/26), local without FSMS 4% (1/26), imported 0% (0/18). By spice type (Table 2): cinnamon 25% (2/8), thyme mix 17% (1/6), red chili 12% (1/8); all other types compliant. Highest reported mean: unpacked thyme mix at 0.345 mg/kg; lowest: imported white pepper at 0.002 mg/kg. Cd levels differed significantly across spice categories (Kruskal-Wallis p = 0.0394).

Total arsenic (tAs). Range 0.017 to 0.242 mg/kg. All 96 samples below Codex MPL (0.5 mg/kg). No significant difference between packaging categories (p = 0.5661). Highest reported mean: oregano locally packed with FSMS at 0.242 mg/kg; lowest: unpacked white pepper at 0.017 mg/kg. Speciation was not performed; values are total arsenic determined by GF-AAS.

Total mercury (tHg). Range 0.003 to 0.033 mg/kg. All 96 samples below Codex MPL (0.1 mg/kg). No significant difference between categories (p = 0.7832). Highest reported mean: unpacked thyme at 0.033 mg/kg; lowest: locally packed paprika without FSMS at 0.003 mg/kg. Values are total mercury by direct mercury analyzer; methylmercury speciation was not performed.

Sampling design. Nine spice/herb types were assessed across all four categories (bulk, imported, FSMS, NFSMS); four types (sumac, sesame, dried mint, thyme mix) were assessed across only three categories because no imported option was available in the Lebanese market. 480 individual samples (9×5×4×2 + 4×5×3×2) were collected, then combined into 96 composites by pooling the 5 brands within each (spice × category × period) cell. Brands were not analyzed individually.

Methods (brief)

Analysis at the Lebanese Agricultural Research Institute (LARI), ISO/IEC 17025 accredited. Composite sampling: five brands per (spice × category × collection period) combined into a single composite. Sample prep: approximately 0.5 g per composite digested in 1 mL H₂O₂ + 7 mL HNO₃ in Teflon containers using a microwave digester (Milestone Ethos Plus, 1000 W, two-stage to 200°C), then diluted to 25 mL with 1% HNO₃. Pb, Cd, and As determined by graphite-furnace atomic absorption spectroscopy (Thermo Fisher Scientific AAS thermal M series — Graphite Furnace GF95Z Zeeman Furnace) at 217.0 nm (Pb), 228.8 nm (Cd), and 193.7 nm (As). Hg determined by direct mercury analyzer (DMA-80, Milestone, Sorisole, Italy) at 253.7 nm; quartz furnace, 0.1 g sample, no separate digestion. LOD / LOQ (mg/kg): Hg 4×10⁻⁵ / 1.2×10⁻⁴; Cd 3.6×10⁻⁵ / 1.2×10⁻⁴; As 2.62×10⁻⁴ / 8.76×10⁻⁴; Pb 2.5×10⁻⁴ / 8.33×10⁻⁴. Recovery 90–103% across all analytes (validated by spike recovery on cinnamon at three concentration levels because certified reference material was unavailable). Statistical analysis: IBM SPSS v22.0; Kruskal-Wallis for non-normal data (Cd, Hg, As), ANOVA for normal data (Pb). Two collection periods (June and September 2020) to account for seasonal variability. Basis: as-sold dried spice/herb (dry-weight matrix). No arsenic speciation; no mercury speciation.

Implications

Certification: Spices and dried herbs are intermediate food ingredients used across many product categories; their Pb and Cd load contributes to dietary intake at the rates documented here. The 20% Pb exceedance rate against the Codex MPL of 0.3 mg/kg establishes an occurrence-data baseline for Lebanese-market spices and herbs. Bulk/unpackaged sourcing shows the worst Pb compliance profile (31%) and imported product is close behind (28%), indicating that channel-level surveillance and supplier-level controls — rather than packaging alone — are the loci where Pb load is determined.

Courses: Useful as a case study in how packaging, FSMS status, and sourcing channel interact with contamination outcomes. The finding that Pb and Cd exceedance rates were identical for locally packaged categories with and without FSMS (both 12% for Pb, 4% for Cd) supports the authors’ inference that FSMS programs in this setting do not routinely monitor heavy metals as part of food-safety management. Also illustrates the use of composite sampling (5 brands pooled per cell) to screen a market — including the limitation that individual high-contamination samples can be diluted by pooling.

App: Spice/herb ingredients carry meaningful Pb load (range 0.052 to 0.486 mg/kg as-sold; multiple categories with mean exceeding the Codex MPL of 0.3 mg/kg) and lower Cd load (max mean 0.345 mg/kg, mostly compliant). Total As and total Hg are low in this category relative to Codex MPLs.

Regulations: Benchmarked against Codex Alimentarius CXS 193–1995 (general standard for contaminants and toxins in food; Pb MPL 0.3 mg/kg, Cd MPL 0.2 mg/kg, As MPL 0.5 mg/kg, Hg MPL 0.1 mg/kg applied to spices and dried herbs) and Codex CXS 328–2017 for dried thyme. Lebanon has no national metal limits for most spices and herbs; the Lebanese Standards Institution (LIBNOR) NL 677:2017 covers thyme and thyme mixes and aligns with Codex.

Wiki pages this source may touch

• lead
• cadmium
• arsenic-total
• mercury
• cinnamon
• black-pepper
• herbs-and-spices
• herbal-botanicals
• spices

Verification notes

• Merge-enhanced 2026-05-29 against the manual-fetch PDF in raw/Manual Fetch Kimi /condiment_papers/05_PB_Vanilla_Spices/. The PDF is the same paper as the Marker-converted raw/markdown/FM_11343411/FM_11343411.md already referenced in raw_path; raw_handle preserved as the FM_ identity.
• Expanded ingredients: from a single umbrella (herbal-botanicals) to include the umbrella plus the specific commodity slugs that actually exist in the wiki and that the paper measures distinctly (cinnamon, black-pepper, herbs-and-spices). The other 10 measured spices (thyme, thyme mix, paprika, cumin, oregano, garlic powder, red chili, white pepper, dried mint, sesame, sumac) do not yet have wiki ingredient pages and are not pre-emptively created here per Part 10 / skill stop conditions.
• Added per-spice exceedance percentages from Table 2 for completeness (the prior page summarized only the highest-exceedance and most-affected categories; the full table is now in the page).
• Added the sampling-design paragraph clarifying the 480 → 96 composite-sampling structure that drives the n/N denominators in Tables 1 and 2.
• Brand firewall (Part 12, strict): the paper does not name any of the 5 brands used per composite; composite values are reported only at the spice × category × collection-period level. Method-vendor names (Thermo Fisher Scientific GF95Z, Milestone DMA-80, Milestone Ethos Plus, Sigma-Aldrich reagents, BOECO UV/UF water system, IBM SPSS) are retained under the Part 12 scientific-method exception.
• Wiki/HMTc firewall (Part 2): no threshold proposals; the Certification implication describes the occurrence baseline against the Codex MPL but does not recommend an HMTc value.
• Speciation note (Part 14): paper uses GF-AAS for As (no inorganic speciation) and DMA-80 for Hg (no methyl speciation); metals listed as tAs and tHg accordingly.
• Audit subagent (2026-05-29) flagged “BOECEL” as a vendor-name transcription error; verified against PDF p.3 §2.3 — paper text is “BOECO pure (Boeckel, Hamburg, Germany)” — corrected to BOECO.
• Audit subagent (2026-05-29) noted that matrices: [spices, dried-herbs] cannot be verified against an enumerated controlled vocabulary; both slugs are used consistently across the merge-enhanced sibling spice-source pages (huff2025-spices-lancaster-pa, tinggi2025-spices-herbs-queensland, munarso2024-vanilla-indonesia-quality) and are retained as-is.

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.