Chen et al. 2001 - Arsenic in edible fats and oils from Taiwan

Chen et al. developed a focused microwave digestion and atomic fluorescence spectrometry method for arsenic in edible fats and oils, then applied it to 21 market samples. The market-sample results ranged from 0.005 to 0.027 ppm arsenic, all below Taiwan’s 0.1 ppm maximum permissible concentration for edible fats and oils cited by the authors. The paper is useful both as total-arsenic occurrence evidence for fats and oils and as a method-validation source for oil-matrix digestion.

Key numbers

Market-sample arsenic results

Table 3 reports arsenic in ppm for edible fat and oil samples. The source does not report speciation, so these are total arsenic values.

The source-stated overall range is 0.005-0.027 ppm arsenic across 21 samples.

Method validation

The atomic fluorescence instrument detection limit was 50 pg/mL, calculated as three times the standard deviation of twelve reagent blanks. The calibration curve used arsenic standards at 300, 500, and 700 pg/mL and reported Y = 0.2154X + 0.645 with R2 = 0.999.

Table 2 reports recovery for spiked samples:

Methods (brief)

The method weighed 0.5 g of fat or oil into a digestion vessel, added 10 mL sulfuric acid and 5 mL nitric acid, and used hydrogen peroxide during focused microwave digestion until the digest turned clear. The digest was diluted, treated with potassium iodide, reacted with hydrochloric acid and sodium borohydride to form volatile hydrides, and analyzed by hydride-generation atomic fluorescence spectrometry. The optimized conditions used 30% hydrochloric acid, 1% sodium borohydride, and argon carrier gas at 250 mL/min.

The occurrence data are total arsenic in source-reported ppm for edible fats and oils. The paper discusses arsenic species in general toxicology background, but the analytical method and results table do not speciate inorganic versus organic arsenic.

Implications

Certification: The study provides Taiwan-market total arsenic occurrence values for edible fats and oils. The data are useful for oil-matrix context but should not substitute for inorganic arsenic and should not be silently pooled with US-market oil benchmarks without a market decision.

Courses: Useful as a compact case study for oil-matrix digestion and hydride-generation AFS, including recovery data in palm oil, olive oil, and lard.

App: Adds low-level total arsenic context for several oil and fat categories. Matrices without exact ingredient pages, such as teaseed oil, salad oil, lard, winter oil, and palm oil, remain source-specific matrix descriptors rather than invented ingredient slugs.

Microbiome: Not addressed.

Wiki pages this source may touch

• arsenic-total
• peanut-oil
• sesame-oil
• olive-oil
• sunflower-oil
• butter
• corn-oil
• rapeseed-oil
• vegetable-oil
• vegetable-oils
• cooking-oils-other

Verification notes

• PDF metadata, article header, and journal citation identify this as Journal of Food and Drug Analysis volume 9, issue 2, pages 121-125, with DOI 10.38212/2224-6614.2799.
• The first page lists receipt on June 12, 2000 and acceptance on February 27, 2001; the publication year is 2001.
• The PDF does not state a Creative Commons or equivalent reuse license, so the frontmatter records license: "not stated in PDF".
• The paper reports ppm; this page preserves source units and does not convert them into a product-row basis.
• Total arsenic is used in frontmatter (tAs) because no arsenic speciation is reported.
• No brand names are reported or needed for the source page.
• Source-internal sample-count discrepancy: the abstract, results body, conclusions, and Chinese-language summary state “Twenty-one market samples of edible fats and oils were analyzed,” but the per-matrix sample counts enumerated in Table 3 (peanut 2, sesame 2, olive 2, sunflower 2, salad 3, teaseed 1, butter 1, lard 1, corn 1, winter 1, canola 1, vegetable 1, palm 1) sum to 19. This page keeps sample_n: 21 to match the paper’s narrative count and the Chinese summary, but readers extracting per-matrix occurrence rows from Table 3 should treat the enumerable count as 19 unless additional samples are recoverable from a supplementary source. Audit subagent 2026-06-02 flagged the discrepancy; the contradiction is in the source itself, not introduced by the wiki page.
• Source-internal spike-level label ambiguity: the abstract reports recovery spike levels as “0.05, 0.10 and 0.20 ppm,” while the Materials and Methods section reports spikes as “1 mL of arsenic standards of various concentrations (0.025, 0.05, or 0.1 µg/mL)” added to “0.5 g” of oil. The Materials and Methods values resolve mathematically to 0.05, 0.10, and 0.20 ppm in the spiked sample, matching the abstract. Table 2’s column header reads “Added(ppm)” with the values 0.025, 0.050, and 0.100, which corresponds to the standard-solution concentrations from Materials and Methods rather than the final spiked-sample concentrations from the abstract. This page reproduces Table 2 as the paper printed it; readers comparing spike levels across studies should treat the spiked-sample concentrations (0.05, 0.10, 0.20 ppm) as the meaningful values.
• Calibration-coefficient precision: the body text on page 124 reports the calibration curve as Y = 0.215X + 0.645, while Figure 4 prints Y = 0.2154X + 0.645. This page reproduces the more precise Figure 4 value.
• Matrices vocabulary: the matrices: array lists thirteen source-specific oil and fat descriptors plus edible-fats-and-oils. Several entries (salad-oil, teaseed-oil, winter-oil, lard) are source-specific descriptors that do not have wiki ingredient or product pages of their own; they are preserved in matrices for source-fidelity and are recognized as a known limitation rather than invented controlled-vocabulary slugs.

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.