Abstract

The use of standard additions in the presence of instrument drift and the optimum size of the added spike relative to the unknown analyte concentrations have been investigated for ICP-MS. In particular, a bracket approach, where the spiked sample is measured between two different measurements of the sample, has been investigated. The average of the two sample measurements is used in the standard additions formula to estimate concentration. Several multi-element analyte solutions with single element matrices (Na, Cs, Ba), were analyzed using both the bracket approach and regular standard additions. It was found that the bracket approach led to better results where drift was significant. In addition, optimum spike size was investigated. Simple models predict that determinations would be more precise with larger spikes if the instrument response was linear and RSD was constant. These results show that while the use of larger spikes (from 7 to 50× the unknown concentration) did not yield the better precision predicted by the models, the precision was no worse than for spikes of size equal to the unknown concentration. The Autonomous Instrument project is an approach to the automation of ICP-MS based on choosing an appropriate analytical calibration methodology for an unknown sample. The method of standard additions is the most accurate analytical methodology considered by the Autonomous Instrument. These results have implications for the Autonomous Instrument, suggesting that bracket standard additions should be considered the best method, followed by regular standard additions. In addition, the spike size results imply that in automatic determination, the long linear range of ICP-MS allows the addition of large analyte spikes with minimal prior knowledge of the sample.