The following is an abstract for the selected article. A PDF download of the full text of this article is available here. Members may download full texts at no charge. Non-members may be charged a small fee for certain articles.

Effects of Air-Acetylene Flame Parameters on Simultaneous Multielement Atomic Absorption Spectrometry

Volume 36, Number 6 (Dec. 1982) Page 637-643

Harnly, James M.; Kane, Jean S.; Miller-Ihli, Nancy J.

The effects of the air-acetylene ratio and the observation height on the detection limits and analytical accuracy of 11 elements (Ca, Co, Cr, Cu, Fe, K, Mg, Mn, Na, Ni, and Zn) determined by simultaneous multielement atomic absorption spectrometry with a continuum source (SIMAAC) were investigated. Determinations were made at four observation heights and five air-acetylene ratios evenly spaced over normal ranges for these variables. For the 11 elements, detection limits and characteristic concentrations varied by factors of 1.5 to 25. The best compromise detection limits were obtained at observation heights low in the flame and with richer flames. The variations of the analytical recoveries for Co, Cr, Cu, Fe, K, Mg, Na, and Zn determined in a National Bureau of Standards (NBS) standard reference material (SRM) and two United States Geological Survey (USGS) standard rocks as a function of the flame parameters were relatively insignificant. Significant variations were observed for Ca and Mg, whereas the concentration of Ni was too low to permit any conclusions. The optimal, compromise flame parameters, for analytical accuracy, were the two lowest observation heights in lean flames when all elements except Ca and Mg were determined. Inclusion of Mg shifted the optimal parameters to a higher observation height. Four NBS SRM's and three USGS standard rocks were analyzed weekly, over a 7-week period, at a low observation height in a lean air-acetylene flame. In general, accuracies within ±5% of 100% and precisions of ±5% were obtained for Co, Cr, Cu, Fe, K, Mn, Na, and Zn. Acceptable recovery (100% ± 5%) for Ca could only be obtained with a large dilution of the sample into a 0.5% La matrix.