An assessment of prominent lines in inductively-coupled argon plasmas with special reference to spectrographic general survey analysis

Abstract

About 600 prominent lines for inductively-coupled plasma-atomic emission spectrometry (ICP-AES) using an argon ICP were studied with the triple objective of (1) compiling a library of prominent lines for spectrographic general survey analysis in this laboratory, (2) comparing the detection limits and sensitivities of the lines with literature data, in particular those in the author's Line Coincidence Tables for ICP-AES [Pergamon Press, Oxford (1980)], and (3) providing a list of lines with “universal” sensitivities that can be generally used for semiquantitative analysis. A 50 MHz ICP operated under compromise conditions and spectrographic detection were used. The latter limited the wavelength region of the prominent lines covered to a range between 235 and 446 nm. For 598 lines listed in the Line Coincidence Tables (LCT) the paper presents the visually estimated detection limits and an assessment of band interferences in regions of ±0.05nm about the prominent lines. Similar data are presented for 88 computer predicted, supplementary prominent lines [Spectrochim. Acta 36B, 169 (1981)] for which the predictions came true. For the 383 prominent lines located between 252 and 446 nm densitometrically determined detection limits and sensitivities are given. The sensitivities were converted to a “universal scale”, independent of the spectral characteristics of the optics and the detector. The results of the visual and densitometric determinations are compared mutually and with the data given in the LCT, which are primarily based on the experimental work of Winge et al. [Appt. Specirosc. 33, 206 (1979)]; consequently the present paper also assesses the extent to which data for prominent lines can be transferred from the one ICP to the other, if both are operated under compromise conditions. The conversion of sensitivities on the “universal scale” to a scale applicable to a particular apparatus using the corresponding spectral distributions of the background intensity is discussed and illustrated with examples. The experimental results for 383 lines in the 252–446 nm region are shown to confirm the appropriateness of the data in the LCT for the great majority of the lines; 18 exceptions are specified and discussed. Similarly, for 29 lines below 252 nm or above 446 nm discrepancies with the LCT data are reported and discussed. The use of the complete set of results (including those on the supplementary prominent lines) as a basis for a library of prominent lines for spectrographic general survey analysis is explained. It is shown that in such a library, owing to the limitations that photographic detection imposes on the accessibility of the lower wavelength region, an order of prominence (“ranking”) has to be accepted which differs drastically from that in a “complete” library usable if no wavelength limitations are dictated by the spectroscopic apparatus. In this connection, the paper finally gives a brief outline of the operation of a computer software package used in this laboratory for line selection in general and for spectrographic survey analysis in particular.