Quantitation of mixtures from two-dimensional data sets using orthonormal functions.

Abstract

The use of orthogonal polynomials in mixture analysis has been extended to two-dimensional data sets. Orthonormal two-dimensional functions, constructed from Legendre polynomials, are fitted to the standard and test data sets. Weighted combinations of coefficients are used to correct for interferences from other components whilst quantifying the analyte of interest. The general procedure can be extended to data sets of higher dimensionality or to different orthogonal basis sets.