The accurate determination of the first row transition metals in water, urine, plant, tissue and rock samples by sector field ICP-MS

Magnetic sector ICP-MS was used to measure the first row transition metals Sc, V, Cr, Mn, Fe, Co, Ni, Cu and Zn in certified reference materials, viz., NIST SRM 1643d Water, NIST SRM 1573a Tomato Leaves, NIST SRM 1566a Oyster Tissue, Bio-Rad Level 1 Urine, and in geological standard rocks BHVO-1, BCR-1 and AGV-1. Solid samples were prepared for ICP-MS analysis via high temperature and pressure digestion, followed by simple dilution (1 + 99 or 1 + 999). The water sample was analysed without further treatment, while the urine sample was diluted 1 + 9. With the exception of Zn in the urine matrix, indium was found to be a suitable internal standard for most of the isotopes of interest in the samples considered. All samples were analysed using both low (300) and medium (3000) resolutions (m/Delta m at 10% peak height). The transition metals are well known often to suffer major interference from polyatomic species, and the extent of this interference was clearly evident when using spectral resolution 300. Depending on the sample type and matrix, accurate results could not be guaranteed for all isotopes when using low resolution. Without mathematical correction, extra sample preparation measures, or modified instrument operation, the same problems would be encountered in quadrupole ICP-MS studies. However, magnetic sector ICP-MS using medium resolution mode was found to overcome adequately polyatomic interferences associated with the transition metal isotopes considered, for all sample types. To a general approximation, measured elemental concentrations agreed to within +/- 5-10%, or better, of certified values when a spectral resolution of 3000 was employed [the accuracy found for individual elements varies depending on sample type, analyte concentration, isotope considered, and the relative success of the sample digestion procedure (if required)]. Results are presented taking into consideration both instrumental and method detection limits.