Öztürk et al. 2024 — Aluminium in black, green, and white tea infusions: teapot material and infusion time effects

This study measured Al concentrations in infusions of black, green, and white tea brewed in five teapot materials (aluminium, copper, glass, steel, porcelain) at three infusion times (5, 10, 15 min) using HPLC with fluorescence detection. Al concentrations across all conditions ranged from 38.46 ± 5.08 to 845.75 ± 11.14 µg/L. Tea type was the dominant determinant of infusion Al: green tea (mean 715.53 ± 97.32 µg/L) reached roughly 14-fold higher infusion Al than black tea (mean 50.36 ± 22.78 µg/L), with white tea intermediate (mean 271.13 ± 48.76 µg/L). Teapot material and infusion time both exerted statistically significant but secondary effects (p < 0.001 for each main effect; the three-way tea × teapot × time interaction was also significant at p < 0.001). Hazard quotient calculations based on JECFA’s 1 mg Al/kg body weight/day tolerable daily intake indicated that green tea consumption exceeds HQ = 1 for all teapot/time combinations, white tea exceeds HQ = 1 only for the 15-min glass-teapot condition, and black tea remains well below HQ = 1 across all conditions.

Key numbers

Source: Table 1 (Al concentrations µg/L, mean ± SD; p. 4) and Table 2 (EDI and HQ; p. 5). All values reported as mean ± SD across triplicate analyses.

Al concentrations in tea infusions (µg/L, mean ± SD), by tea type, teapot, and infusion time (Table 1):

Black tea — overall mean 50.36 ± 22.78 µg/L

• Aluminium teapot: nd (5 min), nd (10 min), 46.4 ± 5.2 (15 min)
• Copper teapot: 68.8 ± 2.9 (5 min), 58.8 ± 4.2 (10 min), 38.5 ± 5.1 (15 min)
• Glass teapot: 69.3 ± 2.4 (5 min), 65.7 ± 5 (10 min), 47.2 ± 3.4 (15 min)
• Steel teapot: 48.5 ± 4.4 (5 min), 53.2 ± 2.3 (10 min), 47.9 ± 1.5 (15 min)
• Porcelain teapot: 68.6 ± 5.1 (5 min), 66.6 ± 2.4 (10 min), 75.4 ± 1.9 (15 min)

Green tea — overall mean 715.53 ± 97.32 µg/L

• Aluminium teapot: 598.9 ± 3.1 (5 min), 630 ± 6.9 (10 min), 628.1 ± 12.2 (15 min)
• Copper teapot: 573.2 ± 21.9 (5 min), 659.3 ± 7.1 (10 min), 700.5 ± 12.2 (15 min)
• Glass teapot: 822.8 ± 9.9 (5 min), 714.4 ± 10.7 (10 min), 587.2 ± 2.7 (15 min)
• Steel teapot: 711.6 ± 10.6 (5 min), 830.1 ± 11.6 (10 min), 844.8 ± 10.9 (15 min)
• Porcelain teapot: 823.9 ± 10.4 (5 min), 797.5 ± 9.5 (10 min), 810.3 ± 19.7 (15 min)

White tea — overall mean 271.13 ± 48.76 µg/L

• Aluminium teapot: 227.5 ± 4.7 (5 min), 188.2 ± 4.6 (10 min), 268.2 ± 10.5 (15 min)
• Copper teapot: 182 ± 11.9 (5 min), 261.1 ± 15.4 (10 min), 257.1 ± 9.4 (15 min)
• Glass teapot: 303 ± 7.2 (5 min), 232.6 ± 11 (10 min), 352.7 ± 5.9 (15 min)
• Steel teapot: 322.5 ± 3.2 (5 min), 336.8 ± 7.5 (10 min), 288.3 ± 8.6 (15 min)
• Porcelain teapot: 274 ± 4 (5 min), 295.7 ± 10.8 (10 min), 277.5 ± 4.2 (15 min)

Overall range across all 43 quantified conditions: 38.46 ± 5.08 to 845.75 ± 11.14 µg/L (paper body text, p. 3; Table 1 shows the highest tabulated condition mean as 844.8 ± 10.9 for green tea/steel/15 min — this is a paper-internal precision difference noted as published).

Infusion time main-effect means (across all teas and teapots), µg/L:

• 5 min: 339.62 ± 285.12
• 10 min: 346.04 ± 291.4
• 15 min: 351.36 ± 284.38

Teapot material main-effect means (across all teas and times), µg/L:

• Aluminium teapot: 287.44 ± 255.98
• Copper teapot: 311.07 ± 254.4
• Glass teapot: 355.04 ± 280.05
• Steel teapot: 387.15 ± 316.67
• Porcelain teapot: 387.67 ± 317.53

Hazard quotients (Table 2; based on 70 kg body weight, 10 g tea/day, JECFA tolerable Al intake of 1 mg/kg/day):

• Black tea: HQ ranges from 0.00 (aluminium teapot, 5 and 10 min; nd values yield zero EDI as tabulated) to 0.22 (porcelain teapot, 15 min). HQ < 1 for all teapot types and infusion times.
• Green tea: HQ ranges from 1.64 (copper teapot, 5 min) to 2.41 (steel teapot, 15 min). HQ > 1 for all teapot types and infusion times.
• White tea: HQ ranges from 0.52 (copper teapot, 5 min) to 1.01 (glass teapot, 15 min). HQ > 1 only for the 15-min glass-teapot condition.

Methods (brief)

HPLC with fluorescence detection (Shimadzu prominence series: LC-20AT pump, DGU-20A5 degasser, SIL-20A autosampler, CTO-10ASVP column oven, RF-20A fluorescence detector). Analytical column: GL Sciences InertSustain C18, 250 × 4.6 mm, 5 µm. Method per Shibukawa et al.; derivatisation with 5-sulfoquinoline-8-ol (HQS) and Bis–Tris buffer; mobile phase 10 mM Bis–Tris, 0.3 mM HQS, 7.0 mM TBABr, 70 mM NaCl, 25% (v/v) acetonitrile. HPLC-FD conditions: excitation 366 nm, emission 510 nm, column temperature 25 °C, flow rate 1 mL/min, injection volume 20 µL. Linear range 10–1000 µg/L (r = 0.9983).

Tea brewing protocol: 100 mL ultra-pure water boiled in each teapot type; 5 g tea (black, green, or white) added; infused for 5, 10, or 15 min. Analyses in triplicate. All tea samples commercially purchased from a local supermarket in Turkey.

Statistical analysis: 3-way ANOVA with Bonferroni correction for multiple comparisons (α = 0.05); IBM SPSS v16.

Method note: this HPLC-FD approach measures total dissolved Al in the infusion, not speciated Al forms. No dry-leaf Al concentrations are reported; all reported values are infusion concentrations (µg Al per litre of brewed tea). The paper does not report a method detection limit below the calibration range (10 µg/L); black-tea aluminium-teapot conditions at 5 and 10 min are reported as “nd” (could not be determined) in Table 1.

Implications

Certification: Green tea infusions reach mean Al concentrations of ~715 µg/L under standard household brewing conditions, with HQ > 1 across every teapot type and brewing time tested. For HMTc certification purposes, green tea (and by extension matcha and green-tea-containing products) is a meaningful Al exposure vector that should be evaluated against the JECFA TDI of 1 mg Al/kg body weight/day at expected consumption rates. Black tea infusions present substantially lower Al concentrations (mean ~50 µg/L) with HQ < 1 in every tested condition; white tea is intermediate, with HQ exceeding 1 only at the 15-min glass-teapot extreme.

Courses: The dataset illustrates that tea-leaf processing (which differs between black, green, and white tea production) and the resulting infusion chemistry are the dominant drivers of Al transfer into the brewed beverage — the ~14-fold gap between black and green tea Al under identical brewing conditions is the headline pedagogy point. Teapot material and infusion time both produced statistically significant but secondary effects.

App: For consumer-facing exposure estimation, infusion-basis Al concentrations are approximately 50 µg/L for black tea, 715 µg/L for green tea, and 271 µg/L for white tea (means across all teapot/time conditions). These are aqueous infusion concentrations (µg/L), not dry-leaf concentrations (mg/kg); they must not be conflated with dry-leaf values from other tea sources. Hazard quotient interpretation depends on body weight and daily tea volume; the paper’s calculations use 70 kg body weight and 10 g dry leaf/day (which at the brewing ratio used corresponds to approximately 200 mL of infusion).

Microbiome: Not addressed in this paper.

Wiki pages updated on ingest

• aluminum
• tea
• tea-infusions

Verification notes

Merge-enhance pass 2026-05-20 (Claude Opus 4.7, autonomous ingest cycle; existing page updated: 2026-05-13 predated the 2026-05-14 schema cutoff):

• Corrected metal wikilink slug. Prior revision linked [[metals/aluminium]] (British spelling); the canonical metal page is wiki/metals/aluminum.md (US spelling, per the existing taxonomy). Changed the wikilink to [[metals/aluminum]]. The frontmatter symbol Al was already correct.
• Stripped invented ingredient sub-slugs. Prior ingredients: array included [[ingredients/green-tea]], [[ingredients/black-tea]], and [[ingredients/white-tea]] — none of those pages exist in the current taxonomy (only wiki/ingredients/tea.md, which lists green-tea, black-tea, and oolong-tea as aliases). Per CLAUDE.md Part 10 (5-paper threshold for sub-variant page creation) and the verification checklist taxonomy rule, the array now contains only the existing umbrella slug [[ingredients/tea]]. The tea-type granularity (black/green/white) is fully preserved in the Key numbers prose and Table 1 rendering. The “Wiki pages updated on ingest” list was correspondingly trimmed.
• Corrected truncated raw_path. Prior revision ended at “Determination of aluminium concentrations in black, green, and wh.pdf”; the actual filename ends with “…white tea samples_ effects of different infusion times and teapot species on aluminium.pdf”. Restored the full filename.
• Expanded Key numbers to a per-condition rendering. Prior revision reported only overall, time-margin, and teapot-margin means. The merged page lays out all 43 tabulated condition-level means (Table 1), preserving the paper’s full 3 × 5 × 3 design. This makes the per-cell numerical fidelity auditable and reproducible without back-reference to the PDF.
• Documented paper-internal precision discrepancy. The body text on p. 3 reports the overall range as “38.46 ± 5.08 to 845.75 ± 11.14 µg/L”; Table 1 shows the green-tea/steel/15-min cell as “844.8 ± 10.9”. This is a paper-internal precision difference, not a wiki-side error. The page now notes the discrepancy explicitly. The abstract reports “38.46 ± 5.08 to 844.75 ± 10.86 µg/L”, consistent with the table at low precision. Following the precedent set by other autonomous-cycle merges, the page reports the body-text overall range and notes the discrepancy in line.
• Tightened Implications language for Part 2 firewall compliance. Prior text included “HMT&C product assessment should account for cumulative Al from all ingredients” — directive language toward the certification program. The merged Certification block frames green tea as “a meaningful Al exposure vector that should be evaluated against the JECFA TDI” — occurrence data, not a threshold proposal, per the Part 2 boundary.
• Method note added on what HPLC-FD measures and what “nd” means. Clarified that the assay quantifies total dissolved Al, that no dry-leaf concentrations were measured, and that “nd” in Table 1 (black-tea aluminium-teapot at 5 and 10 min) means below the calibration range, not confirmed absence.
• Updated sample_population to surface the experimental design (5 teapot materials × 3 infusion times × 3 tea types, triplicate) more explicitly.
• Preserved cite_key, raw_handle, license, evidence_tier, sample_n, jurisdictions, matrices, products array, and the overall page structure. No new ingredient, product, regulation, or metal pages proposed. No brand names appear in the source; no Part 12 brand-firewall actions were needed.

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.