Direct determination of 10 trace metals in 50 μL samples of coastal seawater using desolvating micronebulization sector field ICP-MS

Understanding the trace metal marine geochemistry of temporally variable coastal systems requires intensive sampling programs with attendant analytical burdens. Most established techniques for multi-element trace metal determinations are slow, require a skilled chemist, and are not easily automated. Advances in sample introduction systems and ICP-MS instrumentation now provide marine chemists with the sensitivity and mass resolution necessary to determine many trace metals at natural concentrations in coastal seawater. A new method has been developed for the rapid (10 samples h(-1)) determination of V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Cd and Pb in diluted seawater, requiring just 50 mu L of seawater and no reagents other than pure nitric acid. A sensitivity of 800 000-1 200 000 cps ppb(-1) Sr-86 in a 10% sea water matrix is obtained when microconcentric desolvating nebulization is combined with a shielded torch and hot plasma high resolution ICP-MS. Analyses are standardized by a matrix-matched external calibration curve with variations in sensitivity corrected by normalizing to the natural internal standard Sr, a conservative ion in seawater. The method thus depends on mass bias stability for each analyte relative to Sr, which was examined as a function of forward power and matrix and found to be optimized at 1100-1350 W. Precision and accuracy are limited by appropriate correction for blanks, which derive mainly from the ICP-MS introduction system, and are equivalent to about 10% of typical coastal seawater concentrations for these metals. Preliminary evaluation of a new low-flow nebulizer (mu Flow, Elemental Scientific, Omaha, NE, USA) suggested lower blanks and compatibility with solutions high in total dissolved solids compared with standard microconcentric designs. Determination of dissolved concentrations in reference seawater (CASS-3) demonstrate very good agreement with certified values (within 95% confidence limit) and a precision of 3-12% (1 sigma) for all elements except Cr (15%). The utility of the method is demonstrated by the determination of spatial trends for these metals in a transect of seawater samples from shelf waters off southern New Jersey, USA. The new technique is sufficiently sensitive to determine some of these metals in open ocean seawater and, with minor modifications, should be applicable to a larger suite of analytes in a wide variety of natural waters.