The matrix effect in emission spectrometry has been treated in the literature as several separate problems. These problems are here considered in two general classifications: (1) prior to excitation, the matrix effect resulting from differences in composition and history between standards and samples, and (2) during excitation, the matrix effect resulting from changes in composition caused by preferential vaporization from the sample.
In the first classification, different working curves are frequently obtained for the analysis of a given element in a broad group of materials. These differences in matrix produce two effects: (1) changes in volatilization, and (2) changes in excitation temperature. These effects may be eliminated by using standards that are similar to the samples to be analyzed, by fusing or mixing every sample, regardless of matrix, with a buffer to produce a common matrix, and by proper choice of source conditions.
In the second classification, the composition may change by preferential evaporation of volatile elements and compounds when the sample is heated by the discharge. Hence, the composition of the vapors available in the discharge column changes with time. If the sample is a powder, several methods have been used to control the evaporation process, namely, the Stallwood air-jet method, Danielsson's tape machine, and the method of Rusanov and Khitrov of introducing the sample into the discharge by a stream of air.