Zhang et al. 2010 — Pb, Cd, Cr, Ni, Cu, Zn, Fe, Mn, Ti, Se in four Beijing-market detergents by ICP-MS

A four-page short communication in Asian Journal of Chemistry reporting simultaneous ICP-MS quantification of ten heavy metals — titanium, chromium (total), manganese, nickel, copper, cadmium, lead, iron, zinc and selenium — in four detergent brands purchased at a Beijing supermarket. The products are framed by the authors as manual-use detergents for washing tableware, vegetables and fruits, a single product category in the Chinese household market. Titanium dominated the metal mass in three of the four brands (740-872 ng/g) while one brand whose raw material is described as plant-tissue-derived (Brand 3) returned an order-of-magnitude lower Ti (6 ng/g) and the lowest values for nearly every other metal. The authors recommend that Chinese hygiene authorities widen monitoring of detergent products and that consumers reduce detergent use and prefer lower-metal brands. The contribution to the heavy-metals literature is a small but unusually multi-element occurrence snapshot for the manual-dish/produce-wash detergent category in the Chinese retail market circa 2009, with the caveat that n=4 and no certified reference material, recovery percentage or interlaboratory comparison is reported.

Key numbers

Per-brand heavy-metal content of four Beijing-market detergents (Table 3, p. 4779; units ng/g of finished product, equivalent to µg/kg = ppb on an as-purchased basis; Brand 3 is the plant-tissue-derived formulation per p. 4779 narrative; “Brand 1/2/3/4” are the source’s own anonymised labels):

Cross-brand summary (p. 4779 narrative): the four highest by across-brand maximum were Ti (max 871.676 ng/g, Brand 2), Pb (max 249.835 ng/g, Brand 1), Fe (max 174.450 ng/g, Brand 2) and Se (max 147.340 ng/g, Brand 2). The detergents contained little cadmium across the panel (Cd range 0.025-0.206 ng/g). Brand 3, whose raw material is described as derived from plant tissue, produced the lowest metal levels for eight of the ten analytes (Ti, Cr, Mn, Ni, Cu, Cd, Pb, Fe); Brand 3 was not the lowest for Zn (Brand 3 = 31.057 ng/g; Brand 2 = 12.737 ng/g was lower) or Se (Brand 4 = 102.060 ng/g was lower than Brand 3 = 117.590 ng/g).

Analytical-method performance (Table 2, p. 4778):

Correlation coefficients of all calibration curves exceeded 0.940; limits of detection were below 1.00 µg/g for every analyte except Ti (1.06 µg/g); relative standard deviations of the five instrumental replicates ranged 1.20 % (Cd) to 8.91 % (Se).

Unit-scale check. The reported per-brand Table 3 values are stated in ng/g, while the Table 2 LOD column is in µg/g. The Table 3 maxima (e.g., Brand 2 Ti 871.676 ng/g ≈ 0.872 µg/g) sit below the corresponding Table 2 LOD (Ti 1.06 µg/g) and most other measurements also sit below their stated LODs when both are read in µg/g (e.g., Pb max 249.835 ng/g = 0.250 µg/g vs LOD 0.05 µg/g — Pb is above LOD; Cd max 0.206 ng/g = 0.000206 µg/g vs LOD 0.07 µg/g — Cd is well below LOD). The values are reported by the authors without LOD-censoring annotation. This unit-vs-LOD relationship is noted as a paper-internal data-quality flag and is propagated through the synthesis route rather than silently recomputed; see Verification notes.

Methods (brief)

Sample acquisition. Four detergent brands (“detergents”) were purchased from the Chaoshifa supermarket in Beijing (p. 4778). The paper frames the product category as products for washing tableware, vegetable and fruits. Per-capita Chinese detergent consumption is contextualised in the Introduction as >2 kg/year (citations 1, 2 — Y. Lan 2003; X.-Y. Lan 2003), versus 10 kg/year in Japan and 29 kg/year in the United States, motivating attention to product safety. No brand names, manufacturer identities, batch lot numbers, package sizes or purchase dates are given. Sampling appears to be a single time point in 2009 (manuscript received 26 October 2009).

Sample preparation. Microwave digestion per the method of Li et al. 2008 (Spectrosc. Spec. Anal., 28, 2181), cited as reference 9 (p. 4778). The Zhang paper does not reproduce digestion mass, acid mixture, temperature programme or vessel volume; these parameters live in the cited Li 2008 method paper.

Metal analysis. Inductively coupled plasma mass spectrometry on a PerkinElmer ELAN DRC II (“PQ Excell instrument (ELAN DRCII, PE Company, USA)”; p. 4778) with the operating parameters listed in Table 1: 1250 W RF power, 13.0 L/min plasma flow, 0.98 L/min nebuliser flow, 1.25 L/min auxiliary flow, 2.0 mm sampling depth, 180 scanning times, 0 V exit plate, 20 rpm pump rate, 10 s rinse time, 5 replicates per sample read with 8 s replicate read time, 8 s instrument stabilisation, 60 s sample-delay uptake. Instrument parameters cited as adapted from references 10-12 (Rui et al. 2007a, 2007b, 2006). Ten heavy metals (Ti, Cr, Mn, Ni, Cu, Cd, Pb, Fe, Zn, Se) are determined simultaneously.

Method validation parameters reported (Table 2). Calibration linearity (correlation coefficient ≥0.940 for every analyte; ≥0.99 for Ti, Mn, Cu, Cd, Pb, Fe, Zn), instrumental LOD (range 0.05-1.06 µg/g), instrumental relative standard deviation across five replicate reads (range 1.20-8.91 %). The paper does not report a certified reference material recovery, spike-recovery percentage, method-detection-limit (as distinct from instrumental LOD), reagent-blank values, or an interlaboratory comparison. The paper does not report digestion-derived recovery or matrix-matched calibration. Cr is reported as total Cr only; Cr-VI is not speciated. Hg and As are not in the ten-metal panel. Sb, Al, Sn are also not measured.

Acknowledgements. Funded by the National Project of Scientific and Technical Supporting Programs of the Eleventh 5-year Plan (Ministry of Science & Technology of China, project 2006BAD25B02) and Beijing Emphasis Construction Subjects (project PXM2009_014207_078172). Detection performed by Ms. Wang Xiaoyan, School of Public Health, Peking University.

Implications

• Lead spread across four Beijing dish/produce-wash detergent brands. The 0.041-249.835 ng/g (µg/kg, ppb) Pb range across four brands — with one brand (Brand 1) registering 249.835 ng/g, three to four orders of magnitude above the other three brands — informs the Pb panel cell on dish-soaps-manual for the CN jurisdiction. The within-category Pb variability (factor of >6000 between Brand 3 and Brand 1) is the dominant single observation in this paper and motivates the authors’ call for monitoring across the product category rather than reliance on category averages.
• Cadmium as a low-prevalence trace contaminant. Cd across the four brands ranged 0.025-0.206 ng/g (= 0.025-0.206 µg/kg, parts per trillion at low end, low-ppb at high end). Every value is well below the µg/g-scale instrument LOD listed in Table 2 (0.07 µg/g), so these readings are sub-LOD interpolations from the calibration curve and should be flagged as such in the dish-soap Cd panel cell.
• Chromium total, not speciated. Cr (total) ranged 0.000-73.938 ng/g (Brand 3 reported as 0.000 — the authors give no LOD-censoring annotation; this is presumably below the 0.08 µg/g instrument LOD). The Cr-VI HMTc analyte cell cannot be informed by this dataset because speciation was not performed. The total-Cr data inform the metals/chromium umbrella context only.
• Nickel as a moderate-prevalence trace contaminant. Ni ranged 0.502-45.936 ng/g; Brand 3 (plant-tissue feedstock) was the low value and the three conventional-feedstock brands clustered 26-46 ng/g. Feeds the Ni panel cell on dish-soaps-manual.
• Plant-tissue feedstock signal. Brand 3 — the one brand described by the authors as “raw material came from plant tissue” — produced the lowest values for eight of ten analytes (Ti, Cr, Mn, Ni, Cu, Cd, Pb, Fe) and the second-lowest for Se. The two analytes where Brand 3 did not rank lowest (Zn and Se) showed only modest across-brand spread. The data point is a single-brand single-time-point observation and is not on its own grounds for a category-wide feedstock recommendation, but it is consistent with the broader pattern that surfactant-grade plant-derived feedstocks tend to carry lower transition-metal residuals than mineral-sourced or petrochemical feedstocks. Worth flagging as a supplier-class hypothesis to test as more dish-detergent datasets accumulate.
• Titanium as the dominant metal mass (not an HMI panel metal). Ti ranged 6-872 ng/g across the four brands and was the highest single-element value in three of four. Ti is not on the HMTc Category-1 analyte list and is not an HMI synthesis priority metal; the data are recorded here for completeness and to anchor the paper’s narrative finding that Ti “is much higher than any other heavy metal” (p. 4778). The probable Ti source is TiO2 used as opacifier / pigment in formulated dish detergent — common industry practice — though the paper does not name the source.
• No regulatory comparison anchor. The paper states that “all detergent brands meet safety standards” (p. 4779) but does not name the standard, the limit value, or the regulatory document being referenced. China GB standards for detergent heavy-metal limits exist but are not cited. The wiki page does not propagate the “meets safety standards” claim to the regulatory anchor cells; the source did not supply a citable limit.

Wiki pages this source may touch

• lead (Pb 0.041-249.835 ng/g in four Beijing dish/produce-wash detergents; one brand at 249.835 ng/g drives the within-category range)
• cadmium (Cd 0.025-0.206 ng/g; values sit below the paper’s stated instrument LOD of 0.07 µg/g)
• nickel (Ni 0.502-45.936 ng/g; plant-tissue-feedstock brand at the low end)
• chromium (Cr total 0.000-73.938 ng/g; not speciated; does not feed Cr-VI)
• copper (Cu 12.703-52.097 ng/g)
• zinc (Zn 12.737-94.224 ng/g)
• iron (Fe 3.293-174.450 ng/g)
• manganese (Mn 0.000-29.897 ng/g)
• titanium (Ti 6.162-871.676 ng/g; dominant metal mass; not an HMTc analyte)
• selenium (Se 102.060-147.340 ng/g; narrow across-brand range, near the instrument LOD of 0.58 µg/g)
• dish-soaps-manual (per-analyte four-brand panel for Beijing manual dishwashing / produce-wash detergent, ng/g basis; primary product-form route)

Verification notes

• Brand-firewall handling (Part 12). The source paper does not name the four detergent brands; it uses anonymised labels “Brand 1, Brand 2, Brand 3, Brand 4” throughout, with one formulation descriptor (“raw material came from plant tissue”) attached to Brand 3 (p. 4779). The “Brand 1/2/3/4” labels are reproduced in this wiki page exactly as in the source because they are not actual brand names and they carry the dataset’s only structural attribution (brand-by-brand). No brand identification is introduced by this page.
• Folder placement is misleading; product routing is dish-soap, not laundry. The PDF is filed in the source corpus under 08_Laundry_Bleach_Fabric/, but the paper itself describes the product category explicitly as detergents for “washing tableware, vegetable and fruits” (p. 4777, Introduction) — i.e., manual dishwashing and produce-wash liquid, not laundry detergent. The frontmatter products: field therefore routes to dish-soaps-manual only, not laundry-detergents. The folder name reflects the corruption-folder triage bucket, not the paper’s content. Flagged so a future corpus-organisation pass can rebucket the PDF.
• Unit / LOD relationship is paper-internal and propagated unaltered. Table 3 reports per-brand concentrations in ng/g while Table 2 reports method limits of detection in µg/g (i.e., 1000× higher unit). For most analytes the Table 3 reported values sit below the Table 2 LOD when both are converted to a common unit: Brand 2 Ti 871.676 ng/g = 0.872 µg/g vs LOD 1.06 µg/g (below); Brand 1 Cr 73.938 ng/g = 0.074 µg/g vs LOD 0.08 µg/g (below); Brand 2 Mn 29.897 ng/g = 0.030 µg/g vs LOD 0.12 µg/g (below); Brand 2 Ni 45.936 ng/g = 0.046 µg/g vs LOD 0.23 µg/g (below); Brand 1 Cu 52.097 ng/g = 0.052 µg/g vs LOD 0.58 µg/g (below); Brand 1 Cd 0.206 ng/g = 0.000206 µg/g vs LOD 0.07 µg/g (far below); Brand 2 Fe 174.450 ng/g = 0.174 µg/g vs LOD 0.42 µg/g (below); Brand 4 Zn 94.224 ng/g = 0.094 µg/g vs LOD 0.36 µg/g (below); Brand 2 Se 147.340 ng/g = 0.147 µg/g vs LOD 0.58 µg/g (below). Only Pb max (Brand 1, 249.835 ng/g = 0.250 µg/g vs LOD 0.05 µg/g) is unambiguously above LOD. This is a paper-internal data-quality concern: either the per-brand values are extrapolations from below the validated instrumental LOD (and should be treated as semi-quantitative), or the unit reported in one of the tables is mislabelled in the published article. The wiki page preserves the source’s stated units verbatim and flags the relationship here; downstream synthesis should treat sub-LOD values as semi-quantitative.
• n=4 single-time-point. Four brands × one finished-product sample × five instrumental replicate reads. The dataset is structurally a single-laboratory single-shopping-trip characterisation. Not a population-level survey; not a temporal-trend dataset. Synthesis should weight accordingly.
• No CRM / recovery / blank. The paper does not report a certified reference material, spike-recovery percentage, method-detection-limit (as distinct from instrumental LOD), reagent-blank values, or an interlaboratory comparison. This is a B-tier paper with peer review and a documented method but minimal QA reporting, not A-tier.
• No DOI assigned. Asian Journal of Chemistry papers from 2010 in this issue carry an article identifier (AJC-8474 for this paper) rather than a CrossRef DOI. no_doi_assigned: true is set in frontmatter.
• License. Asian Journal of Chemistry is a copyrighted paywalled journal published by Asian Publishing Corporation. The wiki page reproduces only table-level summary data and methodological description consistent with fair-use / scholarly-citation norms.
• Jurisdiction scope. Beijing retail market, single shopping trip at the Chaoshifa supermarket, late 2009. Feeds the CN jurisdiction signal on dish-soaps-manual.

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.