Spectrum stripping in rapid sequential atomic emission spectrometry with the inductively coupled plasma

Citation
D. Merten et al., Spectrum stripping in rapid sequential atomic emission spectrometry with the inductively coupled plasma, SPECT ACT B, 54(9), 1999, pp. 1377-1382
Citations number
13
Categorie Soggetti
Spectroscopy /Instrumentation/Analytical Sciences
Journal title
SPECTROCHIMICA ACTA PART B-ATOMIC SPECTROSCOPY
ISSN journal
05848547 → ACNP
Volume
54
Issue
9
Year of publication
1999
Pages
1377 - 1382
Database
ISI
SICI code
0584-8547(19990913)54:9<1377:SSIRSA>2.0.ZU;2-G
Abstract
Rapid sequential atomic emission spectrometry with the 'Zig Zag' acquisitio n system (ISA Jobin Yvon, France) in the case of the inductively coupled pl asma is described. The 'Zig Zag' system is shown to enable the application of spectral stripping, as both the reproducibility of the intensities measu red, for several lines of Fe, and the reproducibilty of the wavelength posi tioning (errors below 3 pm in the 2nd order and 5 pm in the 1st order) were found to be high. Using an integration time of only 2.5 ms/point, the prec ision of the intensities expressed as the relative standard deviation is be low 5%. These features are shown to allow spectrum stripping from the contr ibution of a Zr matrix, which has a rather line-rich atomic emission spectr um. Spectrum stripping was applied in the wavelength range from 237.9 to 24 1.1 nm, where 10 of the most sensitive emission lines of Fe were found. For sample solutions with 500 mg/l Zr only, 10 mg/l Fe only 500 mg/l Zr as wel l as 10 mg/l Fe, and for a blank solution, respectively, five spectral scan s were averaged and the averaged spectral scan for a blank solution was sub tracted. The spectral scan obtained for a solution containing 500 mg/l Zr o nly was subtracted from the one obtained for the solution containing both 5 00 mg/l Zr and 10 mg/l Fe. The result of spectral stripping is comparable t o the experimentally obtained spectral scan for the solution containing 10 mg/l Fe only after blank subtraction. The recovery for Fe was found to be 9 3 +/- 2% and is therefore comparable to the one obtained by chemical separa tion methods using solvent extraction. When using an integration time of 50 ms per measurement point instead of 0.5 ms/point, as it is possible in the case of the 'Zig Zag' approach, the detection limit for Fe in the measurem ent solutions can be improved from 20 down to 3 mu g/l, or from 40 to 6 mu g/g Zr, respectively. (C) 1999 Elsevier Science B.V. All rights reserved.