Determination of Nonmetallic Inclusions in Metal Alloys by the Method of Spark Atomic Emission Spectroscopy (Review)

Abstract

The literature on the determination of nonmetallic inclusions in metal alloys by the method of atomic emission spectroscopy with single-spark spectra registration is reviewed. The main advantage of this method is its high rapidity (~1 min per measurement), which makes it useful for direct production control. When the spark discharge hits a nonmetallic inclusion, it leads to a high peak (outlier) in the intensity of spectral lines of the elements contained in the inclusion, since the content of these elements in the metal matrix is usually much lower. The intensity distribution of the spectral line of the element obtained from several thousand single-spark spectra consists of two parts: (i) the Gaussian function corresponding to the content of the element in the dissolved form, and (ii) an asymmetric residue in the region of high intensity values due to the inclusions. The quantitative determination of inclusions is based on the assumption that the intensity of the spectral line of an element in the single-spark spectrum is proportional to the content of this element in the mass of the substance ablated by the spark discharge. Thus, according to the calibration curve, which is obtained using the samples with the certified total content of the element, it is possible to determine not only the fractions of the dissolved and undissolved element but also the size of individual inclusions. However, the determination of the sizes is limited to a range of 1 to 20 μm. Furthermore, at present, only Al-containing inclusions can be quantitatively determined. Difficulties arise both with elements that are practically insoluble in steels (O, Ca, Mg, S), and with those whose content in the dissolved form is usually high (Si, Mn). Also, it is still not possible to determine carbide and nitride inclusions in steel using the spectral lines of C and N. The use of time-resolved spectroscopy can reduce the detection limits of the inclusions containing Si and, possibly, Mn. The use of an internal standard in the determination of inclusions also reduces the detection limits, but may distort the results. The use of solid-state linear radiation detectors instead of photomultipliers made it possible to develop a more reliable internal standard based on the background in the neighborhood of the spectral line. Verification of the results of the analysis is difficult owing to the lack of inclusion content reference samples. Further studies can expand the method’s detectable inclusion types list.