Development of a Monte Carlo - Library Least-Squares code package for the EDXRF inverse problem

The Monte Carlo-Library Least-Squares (MCLLS) approach has now been developed, implemented, and tested for solving the inverse problem of EDXRF sample analysis. It consists of a linear library least-squares code and a comprehensive Monte Carlo code named CFARXRF that is capable of calculating the unknown sample spectrum, all the elemental library spectra in the sample, and the differential operators for each library spectrum with respect to each element. Two code,, with graphical user interface have been designed to implement the MCLLS approach and benchmark results are presented for the two stainless steel samples, SS304 and SS316. The results are accurate, the system is easy to use, and all indications are that this approach will be very useful for the EDXRF practitioner. (c) 2005 International Centre for Diffraction Data,