Trace Elements Analysis of Tunisian and European Extra Virgin Olive Oils by ICP-MS and Chemometrics for Geographical Discrimination

Overview

This study measured 17 elements in 42 extra virgin olive oils from Tunisia, Spain (Basque Country), and southern France, along with 9 paired soil samples from Tunisia, by quadrupole ICP-MS after microwave-assisted digestion. The primary aim was geographical discrimination via chemometrics (PCA, LDA). The secondary aim was to correlate trace element concentrations in oil with the bioavailable fraction in paired Tunisian soils.

Four elements (Mn, Sr, Mg, Ni) showed statistically significant Pearson correlations between soil exchangeable fraction and olive oil (p < 0.05; Table 5). Lead concentrations in oil were in the low µg/kg range, with the highest value (2.16 µg/kg) found in oil from Jendouba sampled near a lead-zinc mining deposit.

Key numbers

Element concentrations in olive oil (Table 4, p. 10; µg/kg, reported as median with range):

Literature comparison column (Table 4): cadmium (Cd) 0.001-0.15; lead (Pb) 0.18-6.40; arsenic (As) 0.2-26.6.

Lead, specific findings (p. 10):

“In general, in the European oils, the concentrations of Pb did not exceed 1.35 µg/kg.” “The highest value was equal to 2.16 µg/kg, and this was found in olive oil from Jendouba (T16) sampled near a Pb-Zn mining deposit (Table 2).”

Cadmium (p. 10):

“The values obtained for Cd were below the LOD in all olive oil samples.” When quantified in literature, Cd concentrations in olive oil varied between 0.001 and 0.15 µg/kg.

Iron, Tunisian outliers (p. 12):

“Concentrations of Fe in Tunisian samples ranged from 169 µg/kg to 1310 µg/kg. The median value was equal to 1248 µg/kg.” Oils extracted at OTIT using the Abencor method showed the highest Fe, attributed to direct contamination by the stainless-steel extraction process.

Soil-oil correlations (Table 5, p. 13):

Sr: r = 0.85, p = 0.006. Mg: r = 0.78, p = 0.02. Mn: r = 0.63, p = 0.01. nickel (Ni): r = -0.8, p = 0.04 (inverse). Cd: not detected (no correlation calculable). Pb: r = 0.08, p = 0.84 (not significant).

LOD/LOQ (p. 9 and Supplementary Materials):

LODs between 0.0002 and 0.313 µg/kg; LOQs between 0.0007 and 1.042 µg/kg. “Except for Cd in olive oil, all the measured concentrations were greater than LOD and LOQ.”

CRM recovery (Table 3, p. 9):

NIST SRM 2387 peanut butter used for certified elements: Ca recovery 102%, Cu 86%, Fe 101%, K 101%, Mg 101%, Mn 97%, Zn 96%. Non-certified elements measured in SRM 2387: As 0.13 ± 0.07, Ba 1.43 ± 0.13, Cd 0.051 ± 0.002, Co 0.024 ± 0.002, chromium (Cr) less than LOQ, Ni 0.78 ± 0.04, Pb less than LOQ, Rb 5.66 ± 0.23, Sr 2.95 ± 0.10, V less than LOQ (all mg/kg).

Evidence Fitness

EF-2 (reconstructable dataset). Table 4 provides median and full range for 17 elements in olive oil, classified by geographical origin, with sample sizes. Paired soil data and Pearson correlations add mechanistic context. Cd was below LOD in all oil samples. Lead data in the low µg/kg range with the highest near a mining site. The study supports the finding that some elements in olive oil are correlated with bioavailable soil fractions and that proximity to mining increases lead in oil, but does not report a deliberate reduction intervention.

Methods

Instrument: ICP-MS Plasma Quant Elite (Analytik Jena, 07745 Jena, Germany). Standard mode and integrated collisional reaction cell (iCRC) with He and H.

Sample preparation: Multi-stage mineralization. 0.5 g olive oil pre-mineralized with H2O2 overnight, then HNO3 added, microwave-assisted digestion (Ultrawave SRC, Milestones, Italy) up to 250 °C for 20 min (P max 110 bar), evaporated to dryness (Savillex PFA vials; Evapoclean closed-medium evaporation), re-dissolved in 5 mL 2% HNO3. Total dilution factor 10.

Soil extraction: BCR sequential extraction (ammonium acetate at pH 7, exchangeable fraction only).

Calibration: External calibration at 8 concentrations (0.01 to 500 µg/L); Y, Rh, Ir internal standards.

LOD/LOQ: LODs 0.0002 to 0.313 µg/kg; LOQs 0.0007 to 1.042 µg/kg (Supplementary Table S2).

Precision: RSD 1% to 14% on olive oil triplicates.

Limitations: Abencor method used for OTIT-extracted samples may introduce Fe contamination from stainless steel surfaces (acknowledged by authors, p. 12-13). No certified reference material exists for trace elements in olive oil; peanut butter (NIST SRM 2387) was used as the closest matrix match. Cd was below LOD in all olive oil samples.

Implications

Certification: Provides multi-element occurrence data for olive oil across three Mediterranean origins. Lead concentrations were low (median 0.53 to 0.88 µg/kg), consistent with other literature. The Jendouba sample near mining (2.16 µg/kg) illustrates the orchard-siting risk. Cadmium was below detection in all oils. Iron in Tunisian oils was an order of magnitude higher than European, attributed to the extraction method rather than inherent contamination.

Courses: The paired soil-oil correlation data and PCA geographical classification make this useful for teaching provenance-trace element relationships in edible oils.

Provenance notes

Open access (CC BY 4.0). MDPI journal (Foods), DOAJ listed. Two authors (Nasr and Donard) overlap with Nasr et al. 2022. Deposit: HMI-MIT-olive-oil. Units in this paper are µg/kg throughout.

Related evidence

• Olive Oil (Section 2: mitigation and inputs)

Update history

The five most recent substantive edits to this page, classified major (evidence or structure moved), correction (a published value or statement was wrong and has been fixed), or minor (narrative rewritten without changing the underlying evidence). Each description is derived from what the edit did to this page; the linked commit is the authoritative record, routine regeneration passes are excluded, and the full version history lives in git.