Koga et al. 2021 — Microwave digestion method for Sn, As, and Pb in soft drinks

Single-laboratory method development and validation paper from the Fukuoka Institute of Health and Environmental Sciences. The study evaluates a microwave digestion sample-preparation procedure coupled to ICP-AES (Sn) and ICP-MS (As, Pb) for soft drinks other than mineral water. The paper reports no native concentrations in commercial product; all reported analytical values are spike-recovery on synthetic matrices.

Published as a note (“ノート”) in Shokuhin Eiseigaku Zasshi (Food Hygiene and Safety Science, Japan), Vol. 62, No. 3, pp. 100–104. Received 2 November 2020; accepted 16 April 2021.

Key numbers

Performance parameters were estimated against the MHLW validation guideline (target ranges: trueness 90–110%, repeatability and intralaboratory reproducibility RSD <15%). The Sn results below are for the procedure that includes a post-digestion addition of 1 mL H2SO4 (Fig. 1A in the source); the alternative procedure without H2SO4 (Fig. 1B) gave Sn recoveries of only 40–90% (mean 63%, n=4) and was rejected.

The performance design was 2 operators × 2 replicates × 3 days (n=12 per analyte) on a canned coffee beverage matrix containing milk and sugar. Sample blanks did not exceed the lowest calibration point for any analyte (Sn <0.1 mg/L, As/Pb <0.0001 mg/L in measurement solution). Calibration regression coefficient of determination R² >0.999 for each element.

Recovery on six additional high-organic-load matrices (n=3 each) ranged Sn 92–96%, As 92–99%, Pb 92–102%. The six matrices were: a plant-extract fermented beverage, a cola, a reconstituted orange juice, a vegetable juice cocktail, a soy milk drink, and a liquid protein milk beverage (the last is classified as a milk product rather than a soft drink in Japan but was included to test method robustness against high organic content).

Reagent blanks (ultrapure water 0.5 g substituted for sample) and sample blanks (unspiked beverage) showed no positive interferences within the calibration range. The internal-standard yttrium count rate ratio across H2SO4-containing samples relative to H2SO4-free blanks was 0.94–0.95, indicating that the 0.83% H2SO4 in the final measurement solution did not materially depress the ICP-AES signal under the chosen internal-standard configuration.

Japanese regulatory context cited by the authors (MHLW Notification No. 370 of 1959, “Standards for Foods, Additives, etc.”, as cross-referenced in the relevant notification test method): soft drinks other than mineral water carry a Sn limit of 150.0 ppm (only when packed in metal containers), and “not detected” limits for As and Pb. The notification test methods translate “not detected” to operational thresholds of 0.2 mg/kg as As2O3 for As and 0.4 mg/kg for Pb. The authors chose a 0.2 mg/kg spike level for both As and Pb to align with the As detection threshold.

Methods (brief)

Sample preparation: 0.5 g aliquot of beverage, 2 mL HCl (Ultrapur grade), 5 mL HNO3 (Ultrapur grade), digested in a Milestone ETHOS One (Milestone General K.K.) closed-vessel microwave digester at 200 °C for 10 minutes (temperature profile in Fig. 3 of the source; ramp roughly 20 min to 200 °C, hold 10 min, cool).

For Sn evaluation: 1 mL concentrated H2SO4 added to the digestate, transferred to a PTFE beaker, heated on a hotplate at 160 °C until volume reduced to approximately 1 mL, cooled, reconstituted with 10 mL ultrapure water + 5 mL HCl, heated 100 °C for 5 min, made up to 50 mL in a glass volumetric flask. A 5 mL aliquot was then diluted with 5 mL ultrapure water, 1 mL HCl, and 1 mL Y (10 mg/L) internal standard for ICP-AES measurement.

For As and Pb evaluation: digestate transferred to PTFE beaker, evaporated to ~0.5 mL at 160 °C without H2SO4 addition, reconstituted with 10 mL ultrapure water + 1 mL HNO3, heated 100 °C for 5 min, made up to 50 mL. A 10 mL aliquot was diluted with 0.2 mL HNO3 and 1 mL In (0.01 mg/L) internal standard for ICP-MS measurement.

Instruments: Agilent 720 ICP-AES (RF 1.2 kW; plasma Ar 15 L/min; aux Ar 1.5 L/min; Sn 189.925 nm, Y 371.029 nm internal-standard line). Agilent 7900 ICP-MS (RF 1.4 kW; carrier Ar 1.0 L/min; aux Ar 0.8 L/min; He collision gas 7.5 mL/min; As m/z 75, Pb m/z 208, In m/z 115). No As speciation; reported as total As. No additional reference material is named.

Calibration: Sn series 0.1, 0.5, 1, 5 mg/L from a 50 mg/L working standard (1000 mg/L Sn stock, FujiFilm Wako). As/Pb series 0.0001, 0.0005, 0.001, 0.005, 0.01, 0.025 mg/L from a 10 mg/L mixed standard (GL Sciences, custom). PP digestion tubes (Digi TUBEs, GL Sciences) used for dilution and volumetric work in the calibration series.

The H2SO4 addition is the methodological pivot of the paper. The authors attribute the recovery failure of the H2SO4-free Sn procedure to volatilisation of Sn under chloride conditions and precipitation in concentrated HNO3 (mechanism reported previously by Higurashi 2015 for dry ashing). The higher boiling point of H2SO4 (340 °C versus 121 °C for HNO3) drives off the nitric acid completely before precipitate formation or volatile-Sn losses can occur, while the residual sulphate is then re-acidified with HCl to keep Sn in solution.

Implications

For testing methodology: this paper documents a method for Sn, As, and Pb in soft drinks that achieves MHLW validation-guideline performance with substantially shorter sample-preparation wall time than the wet-digestion procedure the same group previously published (Kashiwabara et al. 2017). Wet digestion of high-organic-load beverages such as green tea, coffee, or enzyme drinks took 2 days to 1 week of operator-attended work; microwave digestion completed in approximately 5.5 hours per batch with most of that time as unattended cooling and concentration. The procedure as published applies to non-mineral-water soft drinks; the spiked-matrix validation suggests it should generalise across plant-extract, cola, fruit-juice, vegetable-juice, and soy-milk matrices.

The H2SO4 addition step in the Sn procedure is non-obvious and is a useful methodological detail for any laboratory replicating Sn analysis in chloride-containing matrices.

The paper provides no native contamination measurements and therefore contributes no occurrence data for soft-drinks product pages. Its value to this wiki is as a methods reference, not as evidence in any ingredient or product contamination profile.

Verification notes

• Source is a Japanese-language note with an English abstract; key numbers and tables verified directly from PDF pages 1–5 (the full article). Body text is Japanese; technical content extracted from the English abstract, the figures and tables (which are in English), and the equation/numerical sections of the Japanese body.
• DOI not printed on the PDF but reconstructed from the journal’s deterministic J-STAGE DOI scheme 10.3358/shokueishi.<volume>.<startpage> (verified against three other Shokuhin Eiseigaku Zasshi source pages in the wiki: watanabe2017 → 58.80, kataoka2018 → 59.269, takanashi2018 → 59.275). For vol 62, p. 100 this resolves to 10.3358/shokueishi.62.100.
• All applicability-test sample types in Table 2 (cola, reconstituted orange juice, vegetable juice cocktail, soy milk drink, liquid protein milk beverage, plant-extract fermented beverage) are described as commercial products selected for the validation but no brand attribution is given in the source. No Part 12 brand-firewall risk.
• Matrices set to analytical-method per the precedent established by nassar2025-spectrophotometric-cadmium-method and wang2024-gsh-spr-pb-sensor: the paper reports method-validation parameters, not native contamination in any specific product category. Products and ingredients arrays left empty for the same reason.
• Metals: tAs used (not iAs) because the paper measures total As only (m/z 75 by ICP-MS without speciation chromatography). Sn used (not tin-inorganic) because no Sn speciation was performed.
• Spike levels: Sn 150 mg/kg was chosen to align with the Japanese regulatory limit of 150.0 ppm; As and Pb both spiked at 0.2 mg/kg. The 0.2 mg/kg Pb spike is half the operational “not detected” threshold of 0.4 mg/kg, which the authors noted in the methods section.
• Japanese regulatory context cited (MHLW Notification No. 370 of 1959 for the Sn 150 ppm soft-drinks limit) is not yet a wiki regulation slug; flagged as a potential future regulation page but not proposed here.
• No previous wiki source page existed for this paper (DOI grep null because DOI null; raw_handle grep negative; cite-key grep negative). NEW path.
• Audit subagent (2026-06-06) flagged the methods text “CEM ETHOS One (Milestone General)” as a vendor misattribution: ETHOS One is a Milestone instrument, not a CEM one (CEM produces the MARS-series digesters). Verified against PDF page 2 (“マイクロ波分解装置（ETHOS One，マイルストーンゼネラル（株）製）”) — finding confirmed and corrected to “Milestone ETHOS One (Milestone General K.K.)“.

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.