Determination of the total arsenic concentration in human urine by inductively coupled plasma mass spectrometry: A comparison of the accuracy of three analytical methods

Measurement of arsenic (As) in biological samples such as urine has important clinical applications and is being undertaken more frequently in epidemiologic studies because of concern about the carcinogenicity of low to moderate levels of As exposure. The objective of this study was to evaluate and improve the accuracy of As determination in urine by inductively coupled plasma mass spectrometry (ICP-MS). Determination of As in biological samples by ICP-MS is difficult for two reasons: the formation of the molecular ion ⁴⁰Ar³⁵Cl, which overlaps with monoisotopic As at a mass-to-charge ratio (m/z) of 75 (causing spectral interference), and signal enhancement due to organic matrix (nonspectral interference). Available procedures were examined, including the application of different correction procedures using ⁴⁰Ar³⁷Cl and ¹⁶O³⁵Cl molecular-ion formation; the addition of N₂ into plasma or nebulizer gas flows; and the addition of organic molecules to the sample and to calibration standards to eliminate or correct for interference due to molecular-ion formation. The accuracy and precision of determination of As [m/z 75, ionization potential (IP) 9.81 eV] with use of an internal standard was also investigated. Three elements were studied as candidate internal standards: germanium (Ge: m/z 74, IP 7.90 eV, indium (In: m/z 115, IP 5.79 eV), and tellurium (Te: m/z 128, IP 9.01 eV). It was found that these three elements performed more or less equally well with Ar-N₂ plasma; it was also found that accuracy was significantly improved when Te was used as the internal standard instead of Ge or In for ethanol-added samples. Our results indicate that accurate and precise measurement of As in urine by ICP-MS can be obtained by either of two methods (< 5% error, ≃ 2% RSD, limit of detection 0.1 ng ml^-1): (1) the addition of 1% N₂ to plasma gas flow or 3% N₂ to nebulizer gas flow, along with use of any of the internal standards tested, or (2) the addition of ethanol to the sample and to calibration standards, with use of Te as the internal standard. The most accurate results (< 1% error) for National Institute of Standard and Technology Standard Reference Material (NIST SRM) 2670 (toxic elements in urine) were obtained with Ar-N₂ plasma with either Te or In as the internal standard.