Effect of sample volume on the limit of detection in flow injection hydride generation electrothermal atomic absorption spectrometry

Abstract

The analytical performance of methods for the determination of hydride forming elements has been improved recently by the development of procedures in which the hydride is trapped on the interior surface of a graphite furnace atomizer. The signal for a given concentration increases with increase in sample volume and it is often implied that a decrease in the limit of detection may also be achieved by increasing the sample volume. To evaluate this claim, a simple equation was derived which predicts the relationship between detection limit and sample volume when all the contributions to the blank are proportional to sample volume. A time-based approach to the variation of sample volume was developed to ensure that the analyte introduced from reagent contamination was, in fact, proportional to sample volume. Detection limits were measured for a series of sample volumes between 156 and 1560 µl. As the sample volume was increased, the detection limit improved significantly from 0.3 to around 0.05 µg l^–1 up to a volume of about 500 µl. Between 500 and 1000 µl, a further improvement, to around 0.02 µg l^–1, was obtained, but for volumes larger than 1000 µl no further significant improvement was obtained. Good agreement between the predicted and experimentally determined variations in detection limit with sample volume was obtained and thus the underlying inverse proportionality of the relationship between detection limit and sample volume was confirmed. This rectangular hyperbolic relationship has practical consequences for the extent to which detection limits can be improved by increasing the sample volume, even when the blank is very low or zero.