The article is devoted to the evaluation of analytical capabilities of a small-size atomic absorption spectrometer, which differs from traditional atomic absorption devices in that it allows to register the whole absorption spectrum of the sample in one cycle of electrothermal atomization, thus significantly increasing the efficiency of the analysis. The evaluation of the analytical capabilities of the developed prototype was based on the example of wastewater analysis, because this type of analysis has the least stringent requirements for detection limits and the number of elements to be determined, but is also a very important task. Based on the results obtained, it has been concluded that a small-size atomic absorption spectrometer is capable of detecting 9 out of 32 elements normalized in wastewater with detection limits substantially below the MPC. Determination of another 3 elements will be possible after increasing the resolution of the spectrometer to 22-24 pm.

The work is devoted to the study of atomic vapor absorption spectra obtained in the electrothermal atomizer (graphite furnace) and registered by the high-resolution spectrometer "Grand-2000" (spectral resolution 4 pm). Comparison of the obtained spectra with the spectra recorded by the "Grand-2" spectrometer, which is a part of the "Grand-AAS" atomic absorption spectrometer, indicated the presence of unresolved spectral interferences in the latter, as well as a number of unresolved doublets of analytical lines, both already presented in the database of spectral absorption lines of chemical elements, and missing in it. Based on the results of the work, we can talk about the suitability of the "Grand-2000" spectrometer for adjusting this database.

The traditional widespread method for the atomic-absorption spectral analysis (Walsh's method) now is consecutive since demands use of the selective light source and reorganization of wavelength of the monochromator for definition of each element as a part of test. This fact seriously reduces analysis productivity, however, this method almost completely excludes emergence of the problems connected with spectral hindrances. Using a method of the atomic-absorption spectral analysis with a continuous spectrum source and a spectrometer with low resolution for the simultaneous analysis of multielement samples, spectral hindrances of analytical lines of the elements which are contained in the analyzed substance are inevitable. For carrying out the correct qualitative and quantitative analysis it is necessary to identify, to divide and to consider spectral hindrances. In this work we presented application of a method of identification and accounting of spectral hindrances on a spectral complex of low resolution "Kolibri-AAS".

One limitation of atomic absorption spectrometry is the narrow range of measurable concentration (1–2 orders of magnitude). In simultaneous multi-element analysis, this may require multiple dilutions of the sample to determine several elements with different concentrations in the sample. Possible ways to expand the range are to linearize the calibration curve by correcting the integral of the absorption signal or to use absorption values on the line wing as an analytical signal and plot several graphs along one line at different distances from its center. Both methods have their drawbacks. We propose another method for expanding the measurable concentration range by using less sensitive absorption lines of elements. A number of elements are identified that have a sufficient number of lines with different sensitivities. The proposed method is compared with the method of linearization of the calibration graph and the calculation of the absorption signal on the line wing. Using as an example Co and Ni, which have sufficiently rich absorption spectra, we have shown the possibility of expanding the measurable concentration range by using several absorption lines: for cobalt, the range is expanded to six orders of magnitude, and for nickel, to five orders of magnitude. Calibration curves are plotted in the concentration ranges 0.24–250.000 μg/L for cobalt and 1.9–250.000 μg/L for nickel. The calibration error is lower than that of the linearization method: 5% against 25 % for cobalt and 4% against 24% for nickel. Thus, the proposed method can be used for simultaneous multielement determination in a wide range of concentrations without diluting the sample.