Fan et al. 2025 — Heavy metals in green tea from Hangzhou area (n=120, ICP-MS)

This study is the most analytically comprehensive recent survey of heavy metals in Hangzhou-area green tea, measuring 15 elements (Al, As, Ba, Cd, Cr, Cu, Hg, Li, Mn, Ni, Pb, Sb, Se, Sn, V) in 120 tea samples by ICP-MS. All 120 samples were below Chinese standard limit values for As, Cd, Cr, Hg (NY659-2003 agricultural industry standard for tea) and Pb (GB2762-2022 national food safety standard). The comprehensive pollution index (P = 0.094) and all single-factor pollution indices were well below 0.7, placing Hangzhou-origin tea in the “safe/clean” classification. Health risk assessment yielded a cumulative hazard index (HI) of 0.42, with Mn the dominant contributor (HQ = 0.40); the study concludes that green tea consumption from this region poses no meaningful health risk.

Key numbers

All concentrations in tea leaves (mg/kg dry weight, Table 4):

• Al: range 126.00–975.00, mean 276.50 ± 148.50, median 227.00 (detection 100%)
• As: range 0.013–0.17, mean 0.054 ± 0.030, median 0.047 (detection 100%)
• Ba: range 4.01–103.00, mean 21.40 ± 15.92, median 16.75 (detection 100%)
• Cd: range 0.0093–0.17, mean 0.040 ± 0.024, median 0.034 (detection 100%)
• Cr: range 0.044–3.52, mean 0.62 ± 0.55, median 0.42 (detection 100%); total Cr, no Cr-VI speciation
• Cu: range 7.48–31.00, mean 13.80 ± 3.68, median 13.25 (detection 100%)
• Hg: range ND–0.017, mean 0.0033 ± 0.0034, median 0.0028 (detection 69.2%); total Hg, no MeHg speciation
• Li: range 0.025–1.02, mean 0.18 ± 0.15, median 0.15 (detection 100%)
• Mn: range 202.00–2010.00, mean 836.00 ± 274.76, median 814.50 (detection 100%); highest-mass element measured
• Ni: range 2.03–29.90, mean 10.70 ± 4.37, median 10.35 (detection 100%)
• Pb: range 0.051–1.71, mean 0.45 ± 0.30, median 0.37 (detection 100%)
• Sb: range ND–0.22, mean 0.022 ± 0.022, median 0.018 (detection 98.3%)
• Se: range ND–0.18, mean 0.076 ± 0.040, median 0.076 (detection 93.3%)
• Sn: range ND–12.80, mean 0.18 ± 1.22, median 0.022 (detection 94.2%)
• V: range 0.016–0.71, mean 0.11 ± 0.11, median 0.083 (detection 100%)

Chinese regulatory limits applied (Table 2, mg/kg dry leaf):

• NY659-2003 (agricultural industry standard for tea): As 2.0, Cd 1.0, Cr 5.0, Hg 0.3
• GB2762-2022 (national food safety standard): Pb 5.0

Comprehensive pollution index (Table 6): Pi single-factor — As 0.027, Cd 0.040, Cr 0.12, Hg 0.011, Pb 0.090; P comprehensive = 0.094 (all far below 0.7 “safe/clean” threshold).

Estimated daily intake and hazard quotient (Table 7; tea ingestion rate 3.6 g/person/day, body weight 64.36 kg, exposure duration 77.4 years):

• Al: EDI 3.13 µg/kg-bw/day, HQ 0.0031
• As: EDI 0.00061, HQ 0.0020
• Ba: EDI 0.24, HQ 0.0012
• Cd: EDI 0.00045, HQ 0.00025
• Cr: EDI 0.0070, HQ 4.7×10⁻⁶
• Cu: EDI 0.16, HQ 0.0039
• Hg: EDI 0.000037, HQ 0.00012
• Li: EDI 0.0020, HQ 0.0010
• Mn: EDI 9.48, HQ 0.40
• Ni: EDI 0.12, HQ 0.0061
• Pb: EDI 0.0051, HQ 0.0014
• Sb: EDI 0.00025, HQ 0.00062
• Se: EDI 0.00086, HQ 0.00017
• Sn: EDI 0.0020, HQ 3.4×10⁻⁶
• V: EDI 0.0012, HQ 0.00018
• Cumulative HI = 0.42 (Mn contributes 95% of HI)

Methods (brief)

Inductively coupled plasma mass spectrometry (Agilent 8900 ICP-MS) with microwave digestion (CEM, USA). Sample mass 0.3 g tea leaves digested with 5 mL HNO₃ (two-stage program: 140°C 2 min hold, then 180°C 30 min hold); digestate diluted to 25 mL with ultrapure water. Internal standard: indium (In) 3 µg/L. Monitored isotopes: ⁷Li, ²⁷Al, ⁵¹V, ⁵²Cr, ⁵⁵Mn, ⁶⁰Ni, ⁶³Cu, ⁷⁵As, ⁷⁸Se, ¹¹¹Cd, ¹¹⁸Sn, ¹²¹Sb, ¹³⁷Ba, ²⁰²Hg, ²⁰⁶ᐟ²⁰⁷ᐟ²⁰⁸Pb. Calibration r > 0.9995–0.9999; LODs 0.0003–0.03 mg/kg. Method validated against standard reference material GBW10052 (GSS-30 green tea biocomposition); recoveries 86.4–97.5%. Method validation followed European Commission SANCO/3103/2000 guidelines. ND values handled per WHO: substitution depends on proportion of below-LOD results (0 if <80% below LOD, LOD if 60–80% below LOD; middle-bound = LOD/2). Samples collected from supermarkets (67 of 120), local markets (43 of 120), and planting sites (10 of 120) across seven Hangzhou-area districts: Xihu (33), Yuhang (26), Fuyang (18), Linan (13), Chunan (12), Jiande (9), Tonglu (9).

Limitations

Geographic scope limited to Hangzhou area — cannot be generalised to all Chinese green tea. Market samples dominate (supermarket + local market = 110 of 120); origin-labelling reliability not assessed. Total Cr reported; Cr-VI not speciated. Total Hg reported; MeHg not speciated (acceptable for tea leaves where inorganic Hg dominates, but limits direct comparison to seafood Hg literature). Mn contributes 95% of the HI but Mn is generally not considered a regulatory concern in tea at these concentrations. No leaching/infusion data — all values are dry-leaf concentrations; as-consumed exposure depends on leaching fractions reported elsewhere. Tea ingestion rate (3.6 g/person/day) is a Zhejiang Province average and may understate exposure for heavy tea-drinking sub-populations.

Implications

• Certification: Hangzhou-origin green tea (Longjing/Dragonwell production region) is clean by Chinese standards and by health-risk metrics. Mean Pb 0.45 mg/kg and mean Cd 0.040 mg/kg (dry leaf) are well within NY659-2003/GB2762-2022 limits. Al at mean 276 mg/kg (dry leaf) is the highest non-Mn metal by mass and is relevant for any HMT&C product category that includes Al-bearing tea ingredients.
• Synthesis input: provides a 120-sample baseline for green tea at a recognised premium origin (Hangzhou/West Lake). Contributes occurrence data to camellia-sinensis for tAs, Cd, Cr-total, Pb, tHg, Al, Ni, Cu, Mn, plus the longer-tail elements Ba, Li, Sb, Se, Sn, V.
• App: dry-leaf basis values; conversion to as-consumed (infusion) requires leaching fractions from companion sources (e.g., Brzezicha-Cirocka 2016 reports ~30–45% leaching for Pb, ~10–44% for Cd in green tea infusions). Flag for synthesis pass on tea.

Wiki pages this source may touch

• lead
• cadmium
• arsenic (total As only)
• mercury-total
• aluminum
• chromium (total Cr only)
• nickel
• manganese
• copper
• camellia-sinensis
• tea

Verification notes

2026-05-20 (Claude Opus 4.7 merge-enhance from MFK ingest cycle):

• License corrected from cc-by to CC BY-NC-ND 4.0 per the article’s explicit Open Access statement (“This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License”, PDF page 7).
• Added raw_handle: MFK_occurrence-exposure-and-health-risk-assessment-of- and re-pointed raw_path to the canonical MFK location in 07_Processed_Foods_Snacks_Beverages; the raw/studies/ duplicate is recorded under near_duplicates. Both files have identical sha256 (eef5ee6c…dda1).
• Metals frontmatter corrected: removed Zn (not measured by this study; the analyte list per the abstract is Al, As, Ba, Cd, Cr, Cu, Hg, Li, Mn, Ni, Pb, Sb, Se, Sn, V), added Ba, Li, Se, V. Kept speciation flags tAs (paper measures total As only), tHg (paper measures total Hg only), and unspeciated Cr (paper reports total Cr).
• Key numbers expanded to include Ba, Li, V (previously omitted) and EDI µg/kg-bw/day values per metal alongside HQs (Table 7). Per-metal detection rates added.
• Methods section corrected: the prior version said samples came from “six Hangzhou districts (Xihu, Fuyang, Linan, Chunan, Jiande, Tonglu/Jiande)” — the source actually lists seven districts and Yuhang was the second-largest contributor (26 of 120 samples). Corrected to the full seven-district enumeration with sample counts. Added the SANCO/3103/2000 validation reference and the WHO ND-handling rule reported in the Statistical analysis section.
• Regulatory citations disambiguated: NY659-2003 (agricultural industry standard for tea) covers As, Cd, Cr, Hg limits; GB2762-2022 (national food safety standard) covers Pb. Previously lumped together.
• Section heading “Wiki pages updated on ingest” replaced with current convention “Wiki pages this source may touch” (the ingest does not edit downstream pages directly per Part 9 synthesis-workflow separation).
• No brand-firewall (Part 12) issues found in the source: paper reports aggregate distributions across 120 samples with no brand attribution; values were not converted into a brand ranking on the wiki page.
• No wiki/HMT&C firewall (Part 2) issues: Implications section names the literature contribution and does not propose HMT&C thresholds.

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.