A Genetic Algorithm for Model-Free X-ray Fluorescence Analysis of Thin Films

Single-layer films and multiple-layer films can be quantitatively analyzed using X-ray fluorescence. Due to the absence of adequate standards, most methods are based on the calculation of theoretical X-ray fluorescence intensities from fundamental parameter methods. These methods require as initial...

Full description

Saved in:
Bibliographic Details
Published in:Analytical chemistry (Washington) 1996-07, Vol.68 (14), p.2419-2425
Main Authors: Dane, Adrie D, Timmermans, Patrick A. M, van Sprang, Hans A, Buydens, Lutgarde M. C
Format: Article
Language:English
Subjects:
Algorithms
Analytical chemistry
Chemistry
Condensed matter: structure, mechanical and thermal properties
Defects and impurities: doping, implantation, distribution, concentration, etc
Exact sciences and technology
Fluorescence
Genetics
Low-dimensional structures (superlattices, quantum well structures, multilayers): structure, and nonelectronic properties
Physics
Spectrometric and optical methods
Surfaces and interfaces
thin films and whiskers (structure and nonelectronic properties)
Thin film structure and morphology
X-rays
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Single-layer films and multiple-layer films can be quantitatively analyzed using X-ray fluorescence. Due to the absence of adequate standards, most methods are based on the calculation of theoretical X-ray fluorescence intensities from fundamental parameter methods. These methods require as initial estimates the exact qualitative sample structure and accurate starting values for both concentrations and layer thicknesses. This paper proposes a fundamental parameter method that uses a genetic algorithm as an optimization procedure. The relaxation of the requirements on the description of the sample due to this robust optimization is discussed. Preliminary results are presented indicating possible applications as well as areas for further research.
ISSN:0003-2700
1520-6882
DOI:10.1021/ac951152t