Meng et al. 2023 — Electrochemical sensors for heavy metal detection in food

Meng and colleagues survey the state of electrochemical sensor (ECS) technology for detecting heavy metals in food matrices through roughly 2022. The review organizes the literature by electrode-modification material — inorganic nanomaterials (Au, Ag, Bi, Pt nanoparticles; metal-oxide nanoparticles including Fe₃O₄, TiO₂, ZnO, MnO₂, MgO, Co₃O₄, Co₂O₃; carbon nanotubes; graphene), organic materials (small molecules such as GSH, DTT, NAC, MPA; polymers including PGA and PPy; metal-organic frameworks), and biomaterials (nucleic-acid aptamers and DNAzymes; enzymes including urease, choline oxidase, catalase) — and assesses each class for sensitivity, selectivity, linear range, and limit of detection. The authors recommend Fe₃O₄/graphene/nucleic-acid composites as the optimal materials combination for electrode modification, balancing economy, sensitivity, specificity, and stability.

Key numbers

Table 1 — Au and AuNPs-based electrodes for heavy-metal detection in foods (paper Table 1, p. 3)

Reproducing the LOD and linear range columns exactly as the paper reports them. Concentrations in mixed unit systems; preserved as the paper presents them.

Key sensor performance points called out in the text

• Fe₃O₄-AuNPs / SPCE (Li et al., 2018) for As³⁺ in drinking water: sensitivity 9.43 µA·ppb⁻¹, LOD 0.0215 ppb. Among the lowest As LODs reported for any food-relevant matrix in the corpus (p. 5).
• MWCNT/tetradecanedioic-carboxamide ligand on PIGE (Selvan & Narayanan, 2019) for simultaneous Cd²⁺ and Pb²⁺: linear range 8.3–115 nmol/L, LOD Cd 2.7 nmol/L, Pb 0.9 nmol/L; validated in tap water and rice (p. 5).
• MWCNT/poly/bismuth-modified CNT (Chamjangali et al., 2015) for simultaneous Cd²⁺ and Pb²⁺: LODs 0.2 µg/L (Cd) and 0.4 µg/L (Pb) in well water (p. 5). The Hwang et al. 2008 Bi-CNT sensor cited in the same paragraph reports Pb²⁺ / Cd²⁺ detection in stream water but the review does not give LODs for that system.
• rGRO-SPCE (Jian et al., 2013) for Pb²⁺: linear range 5–200 ppb, LOD 1 ppb; validated in water, fruit juice, eggs, tea.
• Lanthanide MOF (ZJU-27) modified GCE (Ye et al., 2020) for simultaneous Cd²⁺ / Pb²⁺ in leaf vegetables: linear range 0.5–2.0 µmol/L, LODs 1.66 nmol/L (Cd) and 1.10 nmol/L (Pb) (p. 8).
• Porphyrin-MOF (porph@MOF) + DNAzyme (Zhang et al., 2020) for Pb²⁺ in spinach: LOD 5 pmol/L (Fig. 9, p. 8).
• Urease-inhibition Pt/CeO₂ sensor (Gumpu et al., 2017) for simultaneous Pb²⁺ / Hg²⁺ in water: LOD 0.019 µmol/L (Pb), 0.018 µmol/L (Hg); linear ranges 0.5–2.2 µmol/L (Pb) and 0.02–0.8 µmol/L (Hg) (p. 9).
• Choline-oxidase / MWCNT / GCE (Magar et al., 2017) for Pb²⁺: LOD 0.04 nmol/L, linear range 0.1–1.0 nmol/L (p. 10).

Authors’ synthesis recommendation (Section 6)

Fe₃O₄ / graphene / nucleic-acid composites are presented as the optimal materials combination for sensor electrodes “taking comprehensive consideration of economy, sensitivity, specificity, and stability” (p. 11). The composite’s strengths attributed to the three components individually: Fe₃O₄ — green, low-cost, accessible; graphene — robust mechanical strength, high surface area, high thermal conductivity, excellent electron transport; nucleic acids — high sensitivity and target specificity.

Material comparison (paper Table 2, advantages/disadvantages)

The review’s Table 2 (p. 11) ranks the seven material classes by qualitative advantages and disadvantages. Key disadvantages flagged: metal nanoparticles — expensive, hard to mass-produce, ion competition during electrodeposition; metal oxide nanoparticles — same plus stability concerns; carbon nanotubes — complicated preparation, uneven distribution; graphene — water-absorption character affects detection; organic small molecules — low stability, some toxicity; organic polymers — toxicity to humans and environment; MOFs — water dispersibility, limited electron affinity, recyclability and reusability; nucleic acids and DNAzymes — complicated and slow modification processes; aptamers — typically single-analyte, sample-pretreatment requirements, limited reusability; enzymes — rigorous fabrication conditions, poor stability, hard and expensive to obtain.

Methods (brief)

Narrative critical review. The paper reviews the electrochemical-sensor literature (115+ primary references) for detection of heavy metals — primarily Pb, Cd, Hg, As, Cr (including Cr-VI), Cu, Zn, and Ag — in food and food-related matrices including water, vegetables, cereals (rice), peanuts, tea, fruit juice, milk, eggs, fish, and wine. Section 2 introduces the three-electrode electrochemical cell (working, reference, counter). Sections 3–5 are organized by electrode-modification class: §3 inorganic nanomaterials (metal NPs, metal-oxide NPs, carbon-based NPs); §4 organic materials (small molecules, polymers, MOFs); §5 biomaterials (nucleic acids, DNAzymes, aptamers, enzymes). The dominant analytical techniques covered are stripping voltammetry — square-wave anodic stripping voltammetry (SWASV), differential-pulse voltammetry (DPV), linear-sweep voltammetry (LSV), cyclic voltammetry (CV), square-wave voltammetry (SWV) — with screen-printed carbon electrode (SPCE) and glassy carbon electrode (GCE) the most common substrate platforms. No formal PRISMA-style protocol, study-quality scoring, or heterogeneity assessment is performed; the review is narrative and selective. Performance figures (LOD, linear range) are passed through from the primary studies as the underlying authors reported them.

Implications

Certification (HMTc): A methods-landscape reference, not occurrence evidence. The review establishes that several electrochemical-sensor configurations achieve LODs at the µg/L (ppb) to nmol/L level in spiked or simplified matrices for the HMTc analyte set (Pb, Cd, iAs, tHg, Cr-VI). For HMTc auditor guidance the review is supportive evidence that the technique class can in principle approach regulatory limits; it is not a substitute for documented method-validation data in real food matrices (full sample digestion, recovery testing against certified reference materials, head-to-head comparison with ICP-MS) which the review itself flags as the open gap in the ECS food-safety literature (Section 6, p. 12). No certification-threshold value should rest on this review alone.

Courses: The technique-by-technique organization and Table 2 advantages/disadvantages map make this a suitable backbone reading for a course module on the landscape of food-safety heavy-metal detection methods, particularly for brand QA, supply-chain, and regulatory-affairs audiences who need to evaluate vendor claims about non-ICP detection platforms.

App: Not applicable. The review does not produce ingredient-level contamination figures.

Microbiome: Not addressed.

Limitations

B-tier industry-style methods review. CC BY 4.0 license. Coverage is restricted to electrochemical sensors; ICP-MS, ICP-OES, AAS, AFS, XRF, and other reference analytical platforms are not in scope and only mentioned in the introduction as the high-instrumentation comparators. Performance figures are pulled from primary studies and the review does not independently re-validate any of them. Sample matrices in the cited primary studies are heavily weighted toward water and spiked/simplified food matrices; performance in unspiked real food with full sample digestion is sparsely documented across the cited literature, a limitation the authors themselves flag in Section 6.

Wiki pages this source may touch

• index
• lead
• cadmium
• arsenic-inorganic
• mercury-total
• chromium
• chromium-hexavalent
• copper
• zinc
• silver
• rice
• peanuts
• tea
• leafy-vegetables
• water
• fish
• fruit-juice

Verification notes

• 2026-05-18 merge-enhance from v0 (updated: 2026-05-15, raw_handle papers-cube, source folder PCMF). The v0 page captured the framing and the FGP/AuNC + Fe₃O₄-AuNPs headline sensors correctly but compressed Table 1 (the paper’s main tabular result) into three bullets and omitted Table 2 (materials-comparison summary). Both tables are now in Key numbers.
• raw_handle: corrected from papers-cube to PCMF_article-1-copy-4 to match the manual-fetch handle convention (MFK_/MR_/PCMF_ prefixes). The v0 handle was a folder slug, not an instance handle.
• metals: corrected from [Pb, Cd, tAs, tHg, Ni, Cr] to [Pb, Cd, iAs, tHg, Cr, Cr-VI, Cu, Zn, Ag]. The paper consistently writes As³⁺ (inorganic arsenic species) when describing detection targets; tAs was incorrect. Cr-VI is detected explicitly (azacrown monolayer/Au sensor, Table 1) and is called out separately from total Cr. Cu, Zn, and Ag are detection targets in multiple cited sensors (FGP/AuNC for Cu, Zn; DNA-modified Fe₃O₄@AuNPs for Ag). Ni was removed: it is not a detection target in the review (no LOD or sensor reported for Ni); the v0 page appears to have miscoded it from the introductory list of “common contaminant metallic elements” (Section 1, p. 2) which includes Ni in passing.
• ingredients: populated from [] to a list of seven slugs covering the food matrices in which the cited sensors were validated (rice, peanuts, tea, leafy-vegetables, fish, water, fruit-juice). All slugs verified against wiki/ingredients/. Per the routing-layer convention this signals where the source touches the corpus; it does not claim that this paper measures contamination in those ingredients.
• products: reduced from ["[[products/tea]]", "[[products/peanuts]]"] to []. The v0 entries overstated the source’s relevance to product-category pages: the review cites sensor studies that used tea and peanut matrices but does not itself report contamination figures suitable for a product-page contamination synthesis. Routing should treat the source as testing/methods evidence, not product-occurrence evidence.
• matrices: cleared from [food, water, vegetable, tea, peanut, cereal] to []. The v0 values were a mix of overly broad (food), inconsistent (singular peanut vs plural vegetable), and ingredient-name-collisions with the ingredients vocabulary. For a methods review the matrices field has no clean value; left empty.
• license: clarified from CC BY to CC BY 4.0 (article footer specifies 4.0).
• jurisdictions: added [CN, WHO]. The review repeatedly benchmarks sensor LODs against WHO and Chinese GB (5th China Total Diet Study) standards and is authored from Chinese institutions (Guizhou University, Sichuan University, Minzu University of China, Southwest Forestry University).
• Body restructured: Key numbers expanded with Table 1 (Au/AuNPs sensors) and Table 2 (materials advantages/disadvantages); Methods (brief) extended to name the analytical techniques and substrate platforms; Implications restated without proposing any HMTc threshold value (Part 2 firewall); Limitations preserved; ## Wiki pages updated on ingest renamed to ## Wiki pages this source may touch (current naming) and expanded with metals and ingredients lists matching the new frontmatter.
• Brand firewall (Part 12): the review is a methods review and does not name food brands at the value level. Methods-section instrument and material names (PerkinElmer, NRCC reference materials, etc.) are not relevant to this review’s content. No firewall edits required.
• Part 2 firewall: Implications restates the methods-landscape position without proposing HMTc thresholds or harmonizing with HMTc certification criteria. The review’s recommendation of Fe₃O₄/graphene/nucleic-acid composites is reported as the authors’ synthesis, not endorsed.
• Speciation discipline: As³⁺ → iAs (inorganic arsenic). Hg²⁺ → tHg (the cited sensors measure total Hg as Hg²⁺, not MeHg). Cr(VI) → Cr-VI distinct from total Cr.
• Audit subagent (2026-05-18) flagged the Bi-CNT bullet under Check 1 as having a Cd/Pb LOD attribution transposition (proposed swap to Pb=0.2 / Cd=0.4). Verified against p. 5 of the source: the actual text reads “the detection of Cd²⁺ and Pb²⁺ with LOD of 0.2 and 0.4 µg L⁻¹” — positional matching gives Cd=0.2 / Pb=0.4, which is what the wiki already had. The subagent’s specific value-swap finding is a false positive (the subagent appears to have inverted the source’s species order in its quoted excerpt). However, the verification surfaced a real but different error: the LOD pair was misattributed to Hwang et al. 2008 when the source actually attributes it to Chamjangali et al. 2015 (the MWCNT/poly/bismuth-modified CNT well-water electrode); the Hwang 2008 reference in the same paragraph reports a separate stream-water Bi-CNT sensor without LOD values. The bullet has been rewritten to correct the citation attribution while leaving the (correct) Cd=0.2 / Pb=0.4 assignment unchanged.

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.