Evaluation of graphite furnace atomic absorption signals using the peak shape and position

Abstract

A computer program was developed for use with graphite furnace atomic absorption spectrometry (AAS). This program routinely determined the peak area and height and quantified the peak shape and position for each atomisation by measuring the times of six peak parameters; the time of peak appearance, the maximum, the decay to 1/2, 1/e and 1/e² of the maximum, and the end (t_app., t_max., t_1/2, t_1/e, t_1/e² and t_end). A five-point sliding average filter and a nine-point quadratic-cubic filter gave the same precision observed with no filtering and caused only a slight distortion of the peak shape. Acquiring data at a faster rate (112 Hz versus 56 Hz) had no effect on the peak time parameter precisions with platform atomisation, but offered a 25% improvement in the precisions for atomisation from the wall. In general, the peak parameter precisions improved as the volatility of the element increased and were better for wall than platform atomisation. For Cu and Pb, the relative precisions for peak areas integrated from t_app. to t_1/e² were 30% better than the relative precisions for areas integrated over the whole atomisation cycle and were 10% better than those for peak heights. Collectively, the six peak time parameters were successful indicators of the accuracy of the peak areas and heights for the determinations of ten elements (Al, Co, Cr, Cu, Mn, Mo, Ni, Pb, V and Zn) in the presence of five concentrations each of MgCl₂ and NaCl. The accuracy of the peak areas and heights were predicted correctly more than 90% of the time by comparing the time parameters of the standard in the chloride matrix with the parameters obtained for the same standard concentration with no matrix. Under more realistic analytical conditions, the accuracy of the peak areas were predicted 90% of the time by comparing the time parameters of the standard in the matrix with the parameters of the calibration standards.