Aplicación de la calibración multivariada a la corrección del efecto matriz combinado en la espectrometría de emision atómica de plasma inductivamente acoplado .

Abstract

Inductively-coupled plasma atomic emission spectrometry (ICP-AES) has proved to be an excellent tool for simultaneous multi-element determinations in natural waters. In these samples, the matrix elements have a variable and relatively high concentration depending on their origin. These variations could produce spectral interferences and (or) matrix effects teat affect the accuracy of the determination of trace elements. On the other hand, mathematics modelation have used as correction methods for these inter-element effects factorial designs and multivariate calibration. The latter was fundamentally directed to the correction of spectral interference. However, in other optical methods of analysis its efficiency for the correction of matrix effects has been proven. Due to the possibilities of these mathematical tools and the problem caused by the analysis of samples with high relative contents of Ca and Mg, a study of multivariate calibration application for the correction of combined matrix effects produced by these major components in the ICP-AES determination of 10 trace elements was undertaken. The multivariate calibration technique of Ordinary Least Squares (OLS) was used. With this technique, the classic and inverse process was applied for two series of signal and concentration data, one of them with background corrections and the other one without it. The standard error of prediction (RMSEP) was used to evaluate the method's performance. According to our results, the multivariate calibration is a satisfactory solution as a correction method of combined matrix effects in ICP-AES .