Interferences in ICP-AES

The contribution of interference elements can be estimated by performing spectral line interference corrections. Calibrators are prepared in which mutually interferent elements are not present in the same solution. These solutions are then used to calibrate the system. Apparent concentrations are obtained by analyzing the ultrapure single element solutions (or solids). The interference coefficients are calculated by dividing the apparent concentration by the concentration of the interferent. In ICP-AES, the corrections are generally linear and thus a single element solution suffices to determine the correction factor. In spark and DC arc emission spectrometry, several samples are required. In practice, the determination of an element may be influenced by several other sample concomitants, and the final corrected concentration must be the summation of all the in-terferents. To complicate matters further, an iterative procedure must be used to deal with mutual interferences. 

ICP-AES and ICP-MS analyses are hampered in almost all cases by the occurrence of sample matrix effects. The origins of these effects are manifold, and have been traced partly to physical and chemical aerosol modifications inside sample introduction components (nebulisation effects). Matrix effects in ICP-AES may also be attributed to effects in the plasma, resulting from easily ionised elements and spectral background interferences (most important source of systematic errors). Atomic lines are usually more sensitive to matrix effects than are ionic lines. There exist several options to overcome matrix interferences in multi-element analysis by means of ICP-AES/MS, namely ... 

Several authors [386,387] have discussed the spectroscopic and nonspectroscopic (matrix) interferences in ICP-MS. ICP-MS is more susceptible to nonspectroscopic matrix interferences than ICP-AES [388-390]. Matrix interferences are perceptible by suppression and (sometimes) enhancement of the analyte signal. This enhanced susceptibility has to be related to the use of the mass spectrometer as a detection system. In fact, since both techniques use the same (or comparable) sample introduction systems (nebulisers, spray chambers, etc.) and argon plasmas (torches, generators, etc.), it is reasonable to assume that, as far as these parts are concerned, interferences are comparable. The most severe limitation of ICP-MS consists of polyatomic... 

Physical interferences in FLAA analysis are of the same nature as the ones in ICP-AES analysis high dissolved solids contents in samples change the sample viscosity and alter the aspiration rate. 

Laboratories use this test in ICP-AES and AA analysis of new and unknown or unusual matrices to determine if non-linear physical and chemical interferences may be obscuring the target analytes. 

Are results of interference check samples in ICP-AES analysis acceptable ... 

The analysis of environmentally-relevant samples is a major field of application. Based on the work of Garbarino and Taylor [421], a method has been proposed by the US EPA (Environmental Protection Agency) [422] and later by DIN [423] for waste water analysis. The latter, standardized procedure describes the sample decomposition, the analytical range for 22 elements and frequent interferences of ICP-AES in waste water analysis. For the analysis of natural waters, hydride generation [424], preconcentration based on liquid-liquid extraction of the dithiocarbamate complexes [425], adsorption of trace elements onto activated carbon [426] and also co-precipitation [e.g. with In(OH)2] [427], etc. have been reported and special emphasis has been given to speciation (as given in the Refs, in [428]) and on-line preconcentration [134]. 

Although the relative intensities of spectral lines in the ICP differ from those observed in the DC arc and AC spark, the published tables are invaluable for the selection of analyte lines in ICP sources, and the identification of spectral interferences in the spectrometer bandwidths. However, spectral lines are emitted by ICP sources that are not emitted by DC arcs and sparks. In order to facilitate spectral line selection in ICP-AES, numerous spectral line atlases are now available which list the best analytical lines and the potential interferences due to coincidences from major and minor constituents. Simulated... 

In addition to mass overlap, signal suppression may hamper analysis. In ICP-AES, ionizable elements such as Na, K, and Ca are generally tolerated up to 1% before ionization interferences predicted from the Saha equation become significant. However, ICP-MS is more prone to these interferences, and mechanisms other than ionization equilibria must be involved [39]. Although the origin is still controversial, suppression has been explained as arising primarily from... 

For each element, signals from its different isotopes are obtained. Their intensity ratios correspond to the isotopic abundances in the sample. This fact can be used in calibration by isotopic dilution with stable isotopes, and in tracer experiments. Isotopic abundance is also useful for the recognition of spectral interference. The spectra in ICP-MS are simpler than in ICP-AES,... 

Power of Detection. For optimum power of detection, the analyte density in the plasma, the ionization, and the ion transmission must be maximized. The necessary power is 0.6-2 kW with the sample ca. 10-15 mm above the tip of the injector. The detection limits, obtained at single element optimum conditions, differ considerably from those at compromise conditions, but are still considerably lower than in ICP-AES (Table 6). For most elements they are in the same range, but for some they are limited by spectral interference. This applies to As ( As" with Ar CF), Se ( Se with Ar Ar ), and Fe ( Fe with Ar O ). The acids present in the measurement solution and the material of which the sampler is made (Ni, Cu, etc.) may have considerable influence on these sources of interference and the detection limits for a number of elements. The detection limits for elements with high ionization potential may be even lower when they are detected as negative ions (for CF Cl = 5 ng/mL and for Cr Cl = I ng/mL). 

---