The analytical performance of a high transmission interface (S mode), inductively coupled plasma-quadrupole mass spectrometer (the VGE Plasma Quad 3) was evaluated for multitrace element analysis of geological and environmental materials. The sensitivity, limits of detection (LODs), effect of Ca and Na and other major elements on mass response, background, percentage (CeO+)-Ce-156/Ce-140(+), Ce-70(++)/Ce-140(+),and long- and short-term variations were compared with those obtained with the conventional mode (normal mode). Normal mode sensitivities varied from 20 MCPS ppm(-1) (millions of counts per second per ppm) for Be-9(+), 70-80 MCPS ppm(-1) for Co-59(+), 90 for In-115(+) and 40-50 MCPS ppm(-1) for the heavy masses. S mode sensitivities were 180 MCPS ppm(-1) for Co-59(+), 350-380 for In-115(+) and Ce-140(+), 300 MCPS ppm(-1) for Pb-208(+), and 150 MCPS ppm(-1) for Th-232(+) and U-238(+), i.e. enhancements amounting to 7. Three sigma normal and S mode LODs are mainly in the 1-2 and 0.1 ppt range, respectively. S mode LODs are enhanced relative to the normal mode, for masses >80 amu, by factors ranging from about 10 to 50. S mode LODs are depressed relative to normal mode LODs for masses <50 amu by a factor of 10, and the extent of depression is related linearly to mass. In the high- and mid-mass ranges, backgrounds were about 10. They were not affected by sample composition: at 8 amu the S mode background for real samples amounted to about 20, whereas at 220 amu it amounted to four counts. S and normal mode percentage (CeO+)-Ce-156/Ce-140(++) and percentage Ce++/Ce+ ratios were about 1.5%, and temporal variations were insignificant. The percentage RSDs of normal and S mode Sr+, Ag+ and Pb+ isotope ratios were about 0.1%, with the exception of S mode Pb-208(+) and Pb-208(+)/Pb-206(+) ratios in the presence of NaCl, which were degraded by a factor of about 2. Normal and S mode long-term variations for continuous aspiration of 0.1% NaCl for periods of up to 13 h were mass dependent, varying from 2.5-4% for Li-7(+) and Be-9(+) to about 2% for the mid-mass range, increasing slightly to about 3% for high masses. Most of this variation occurred during the first 100-150 min of the analysis during cone priming. With compensation, normal and S mode long-term percentage RSDs and drift were reduced to 1-2%. These variations indicate that extended periods of S mode analysis can be conducted without periodic recalibration. A calibration procedure, based on spiked HNO3, was validated by analysing spiked NaCl solutions, standard water and geological standard reference material (SRM) solutions with internal standardization using conventional solution delivery and how injection. The agreement of the S mode data and the certified and literature values for ultratrace elements, including ppt levels of rare earth elements in the water standards, was satisfactory. An important conclusion is that ion sampling effects in the S mode are minimal and that the enhanced ion transmission interface is not only beneficial for microanalysis using laser ablation, but for geological and environmental type solutions as well. Sensitivity enhancements were preserved and matrix effects were approximately +/-20% for solutions containing 0.1-0.2% total dissolved salt concentration. These variations were reduced to <5% with internal standards matched in mass and ionization potential. (C) 1998 Elsevier Science B.V. All rights reserved.
