Llorent-Martínez et al. 2014 — Fate of metals during virgin olive oil extraction, Spain

This methods-and-fate paper developed and validated a microwave-assisted acid digestion (HNO3/H2O2) plus ICP-MS procedure for quantifying eleven metals (Al, V, Cr, Fe, Co, Ni, Cu, tAs, Cd, Sb, Pb) in raw olive fruits (including the stone), olive pomace, and virgin olive oil (VOO). Olives of three Andalusian cultivars (Picual, Hojiblanca, Arbequina) from a single irrigated orchard in Jaén, Spain were processed through an Abencor laboratory mill (milling, malaxation at 26 °C for 30 min, centrifugation at 3,500 rpm), then analyzed as paste, pomace, and VOO. In natural (unspiked) olive fruits only Al, Cu, and Fe were above method detection limits (MDLs); As, Cd, Co, Cr, Ni, Pb, Sb, and V were all below MDL. Spiking olive paste at 2,400 ng/g per element (Hojiblanca) and at 3,000–10,000 ng/g for Cu and Fe (Arbequina) showed that on average at least 90 % (the conclusion states 93 % across all elements) of the metals introduced into the paste partitioned into the olive pomace, with the residual VOO levels remaining below the International Olive Council (IOC) maximum levels for As, Cu, Fe, and Pb and below the EU Regulation (EC) No. 1881/2006 maximum level for Pb in VOO (0.1 µg/g = 100 µg/kg) in every spike condition. The paper concludes that the Abencor extraction pathway delivers high-quality VOO from a metal-content standpoint even when the parent fruits carry an appreciable metal burden.

Key numbers

Method detection limits (MDLs, ng/g; equivalent to µg/kg):

• Olive pomace / olive fruit matrix: Al 100, As 50, Cd 35, Co 25, Cr 160, Cu 50, Fe 1,800, Ni 50, Pb 12, Sb 10, V 50
• VOO matrix (larger 0.65 g aliquot): Al 30, As 15, Cd 12, Co 9, Cr 50, Cu 15, Fe 600, Ni 15, Pb 4, Sb 3, V 15

Maximum levels (MLs) cited by the source for VOO:

• IOC trade standard COI/T.15/NC No. 3/Rev. 7: As 0.1 µg/g (100 µg/kg), Cu 0.1 µg/g (100 µg/kg), Fe 3 µg/g (3,000 µg/kg), Pb 0.1 µg/g (100 µg/kg)
• EU Regulation (EC) No. 1881/2006: Pb 0.1 µg/g (100 µg/kg) in VOO

Method validation:

• Oily multi-element standard certified at 100 µg/g; measured values 95–102 µg/g across Cd, Cr, Cu, Fe, Ni, Pb, V; Student’s t at p = 0.05 not significantly different from the certified values; RSDs <6 % (n = 3).
• Aqueous-spike recoveries (Table 2) 88–112 % with RSDs <10 % across all eleven analytes in VOO (60 and 200 ng/g spikes, with Fe spiked at 2,000 ng/g), olive pomace (2 µg/g) and olive fruit (2 and 10 µg/g). Inter-day RSDs <10 % and intra-day RSDs <4 % (n = 5).

Picual variety, natural (unspiked) — concentrations as ng/g ± SD (n = 3):

• Olive fruit: Al 600 ± 200; Cu 1,800 ± 50; Fe 6,000 ± 400. As, Cd, Co, Cr, Ni, Pb, Sb, V all <MDL.
• Olive pomace: Al 450 ± 100; Cu 1,700 ± 60; Fe 6,600 ± 300.
• VOO: Al 150 ± 100; Cu 50 ± 6; Fe <MDL.

Hojiblanca variety, paste spiked at 2,400 ng/g per element — concentrations as ng/g ± SD (n = 3):

• Olive fruit (natural background, pre-spike): Al 650 ± 150; Cu 1,850 ± 50; Fe 5,800 ± 350; remaining analytes <MDL.
• Olive pomace (post-extraction): Al 2,850 ± 300; As 2,200 ± 200; Cd 2,100 ± 200; Co 2,100 ± 300; Cr 2,300 ± 300; Cu 4,100 ± 100; Fe 8,100 ± 500; Ni 2,200 ± 200; Pb 2,400 ± 100; Sb 2,100 ± 200; V 2,250 ± 300.
• VOO: Al 250 ± 50; As 50 ± 4; Cd 150 ± 30; Co 110 ± 20; Cr 250 ± 60; Cu 50 ± 7; Fe <MDL; Ni 130 ± 30; Pb 50 ± 6; Sb 100 ± 20; V 100 ± 20.

Arbequina variety, Cu and Fe spiked — concentrations as ng/g ± SD (n = 3):

• Olive fruit natural background: Cu 2,400 ± 20; Fe 9,700 ± 200.
• Cu series: unspiked → pomace 2,380 ± 80, VOO <MDL; paste spike 3,000 → pomace 5,200 ± 400, VOO <MDL; 4,500 → 6,300 ± 500, VOO 55 ± 9; 6,000 → 8,350 ± 400, VOO 60 ± 8.
• Fe series: unspiked → pomace 9,900 ± 300, VOO <MDL; paste spike 5,000 → pomace 15,000 ± 1,000, VOO <MDL; 7,500 → 18,000 ± 1,000, VOO <MDL; 10,000 → 20,000 ± 1,000, VOO <MDL.

Partition summary stated by the authors: on average at least 90 % (Conclusion section: 93 %) of the metals present in or introduced to the olive paste remained in the olive pomace after Abencor extraction, leaving VOO concentrations below the IOC maximum levels for As, Cu, Fe, and Pb under every tested condition, including paste spikes well above natural background.

Methods

Olive sampling: 700 g of fruit per variety taken in January (2012/13 season) from an irrigated orchard in Jaén, Andalusia; fruits drawn 1.5–2.5 m up on the outer canopy with no more than 10 fruits per tree; stored at −20 °C until analysis. Soils characterized as slightly acidic, dominated by entisols, alfisols, and inceptisols.

Abencor extraction (laboratory mill mimicking industrial VOO production): hammer milling of 500 g paste aliquots; malaxation 30 min in a thermo-malaxer with the water bath at 26 °C; centrifugation twice at 3,500 rpm for 1 min to separate olive pomace (solid) from the oily must (liquid), with VOO decanted off the oily must. All wetted surfaces stainless steel. The vegetation-water fraction was analyzed but found to contain negligible metal and is excluded from the partition results.

Digestion: 0.25 g of olive paste or pomace, or 0.65 g of VOO, weighed into 50 mL PTFE PFA microwave vessels with 6.3 mL HNO3 (sub-boiled, 65 %, Sigma) plus 0.7 mL H2O2 (ultratrace grade, Sigma). MARSXpress (CEM, via Gilson) microwave digester: ramp to 200 °C over 15 min, hold 15 min, 1,000 W with automatic regulation. Each batch = blank + oil multi-element standard + 11 samples in triplicate. Digestates were diluted to 50 mL with ultrapure water (Milli-Q 18.2 MΩ cm) and spiked with In to 1 µg/L as the internal standard.

Determination: Agilent 7500a quadrupole ICP-MS with Babington nebuliser, Peltier-cooled quartz spray chamber, and standard 2.5 mm i.d. torch. Plasma 1.2 kW; plasma Ar 15 L/min, auxiliary 0.89 L/min, carrier 0.95–1.0 L/min; nickel sampler (1.0 mm) and skimmer (0.4 mm) cones. Tuned daily with 10 ng/mL Li/Y/Ce/Tl. Isotopes monitored: 27Al, 51V, 53Cr, 57Fe, 59Co, 60Ni, 65Cu, 75As, 111Cd, 115In (internal standard), 121Sb, 208Pb. Internal-standard comparison (Ge, In, Tb, Y) selected In for best long-term drift and matrix-effect correction, particularly for As, Cd, and Sb. Calibration in 10 % v/v HNO3 from a 100 µg/mL SCP Science multi-element standard, seven points 0.2–600 ng/mL, r ≥ 0.998 on all curves.

Validation: (i) oily multi-element standard (Merck 109469 Multielement Standard II Oil Dissolved) analyzed in triplicate against the 100 µg/g certified value; (ii) aqueous-standard spike recoveries on a pool of the three varieties at 60 and 200 ng/g in VOO (Fe at 2,000 ng/g) and at 2 and 10 µg/g in fruit/pomace. Polyatomic interferences from Cl, K, Mg, Na were judged negligible at the matrix levels encountered, based on the group’s prior characterization of the same instrument.

Speciation: As measured at 75As only (total arsenic; no inorganic/organic speciation); Cr measured at 53Cr only (total chromium; no Cr-VI/Cr-III speciation); no Hg or Sn isotopes in the panel.

Implications

The paper provides a quantitative fate map for eleven metals through Abencor-style VOO extraction. Under the tested conditions, even paste loadings of 2,400 ng/g per element (Hojiblanca) and Cu/Fe loadings up to 6,000 / 10,000 ng/g (Arbequina) yielded VOO concentrations below the IOC and EU 1881/2006 maximum levels for As, Cu, Fe, and Pb. For wiki product-page contamination profiling, the key takeaway is that olive pomace concentrates roughly 90 %+ of the parent-fruit metal burden, so VOO occurrence statistics derived from market-survey data should not be interpreted as a direct readout of olive-fruit contamination. Conversely, olive-pomace oils (produced by solvent or further centrifugation of the residual pomace) inherit the concentrated metal pool, consistent with prior reports of higher Cu in pomace oils than in VOO.

This source contributes Spanish (Andalusian) occurrence data for olive-oil ingredient profiling on Al, Cu, and Fe (natural levels at 150 / 50 / <MDL ng/g in VOO across cultivars under non-contaminated orchard conditions) and provides validated method detection limits for ICP-MS measurement of all eleven analytes in olive-fruit, olive-pomace, and VOO matrices. The authors note in their Discussion (p. 1829) that their own earlier work (ref [7], Llorent-Martínez et al. 2011 in Food Control) reported higher Cu in olive-pomace oils than in VOO, sometimes above the IOC ML, attributing this to the same pomace-retention effect documented here.

The paper does not propose certification thresholds, does not measure individual brand products, and does not address pesticide-derived metal contributions beyond noting that Cu fungicides are a known elevated-Cu pathway in olive groves.

Wiki pages this source may touch

• olive-oil
• olives
• olive-oil
• lead
• cadmium
• arsenic-total
• copper
• iron
• aluminum
• nickel
• chromium
• cobalt
• antimony
• vanadium
• eu-1881-2006-contaminants-superseded

Verification notes

• Source identity confirmed against DOI 10.1007/s11746-014-2511-5 (J Am Oil Chem Soc 2014, 91:1823–1830). Three identity checks (DOI grep, raw_handle grep, cite-key grep on wiki/sources/) returned no prior page; this is a fresh ingest.
• Arsenic speciation: paper monitored 75As only (total). Frontmatter uses tAs accordingly; no iAs claim is made.
• Chromium speciation: paper monitored 53Cr only (total Cr). Frontmatter uses Cr; no Cr-VI claim is made.
• Mercury and tin are not in the analyte panel and are not claimed.
• No brand names appear in the source’s contamination data — cultivar names (Picual, Hojiblanca, Arbequina) are botanical varieties, not brands, so Part 12 does not apply. Instrument and reagent vendors (Agilent 7500a, CEM MARSXpress, Sigma, Merck, SCP Science, Millipore Milli-Q, Gilson, Savillex) appear in Methods only and fall under the Exception 2 scientific-method allowance.
• IOC trade standard COI/T.15/NC No. 3/Rev. 7 is cited by the paper but is not present in the current regulations taxonomy snapshot; not linked. EU Regulation (EC) No. 1881/2006 maps to the existing regulations/eu-1881-2006-contaminants-superseded slug.
• olives and olive-oil are both present in the ingredients taxonomy snapshot (and have wiki pages); used as-is. The matrices vocabulary in docs/gpt-collaboration/system-prompt.md lists common matrices but explicitly allows adding new terms when none fit, flagged here: olives and olive-oil are reported as matrices because they are the analytical phases the paper measured (raw olive fruit including stone; the extracted VOO). olive-pomace is a new matrix term added here for the third analytical phase; no ingredients/olive-pomace or products/olive-pomace slug exists and none is invented.
• Conclusion-paragraph average retention is stated as “at least 90 %” in the abstract and “93 %” in the Conclusion; both are preserved in the Key numbers summary to keep the source-stated wording intact.
• Audit subagent (2026-06-01) flagged [[ingredients/olives]] as not in the taxonomy snapshot. Verified against docs/gpt-collaboration/taxonomy-snapshot.md line 40 — olives IS in the Ingredients snapshot and wiki/ingredients/olives.md exists. Finding rejected as false positive; slug retained.
• Audit subagent (2026-06-01) flagged the Arbequina natural-baseline pomace rows (Cu 2,380 ± 80 ng/g, Fe 9,900 ± 300 ng/g, both with VOO <MDL) as omitted from Key numbers; verified against Table 5 and added.
• Audit subagent (2026-06-01) flagged “consistent with prior reports of higher Cu in pomace oils than in VOO” in Implications as cross-source synthesis; verified against the paper’s own Discussion (p. 1829) where the authors cite their 2011 paper (ref [7]) for exactly this comparison. Rewritten to attribute the comparison explicitly to the source’s own citation, removing the wiki-side synthesis framing.

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.