A procedure for overlapping deconvolution and the determination of its confidence interval for arsenic and lead signals in TXRF spectral analysis

Abstract

We studied the applicability and validity range of a mathematical procedure on the basis of the principia of maximum likelihood for the identification and quantification of arsenic and lead. This procedure has showed to be appropriate for analyzing two or more signals that interfere with each other or for the quantification of a small signal in a very noisy environment, but a complete study of determination of its confidence interval in samples in which one element in the overlap is present in a concentration lower than the other elements was not studied previously. The identification implemented provides very exact values of relative concentrations for cases that are difficult to process using other adjustment methods. The proposed procedure is applied to experimental cases, analyzing liquid solutions through total reflection X-ray fluorescence to determine a range of concentrations for the detection of traces of arsenic signals in the presence of large lead interfering signals and vice versa.