K-edge filter subtraction technique used for mapping elemental distribution on paintings

Abstract Several techniques exist for mapping the element pigments on painting using the synchrotron light or quasi monochromatic sources. K-edge subtraction technique is a well known technique and based on discontinuity in the attenuation coefficient due to photoelectric K-edge of the absorbing materials. A couple of filters with slightly different K -edge energies, if are with balanced thicknesses, can isolate spectra in a narrow energy band. Quasi monoenergetic X-rays with a spectral width Ek that is the difference in K-edge energies will be taken by subtracting images or spectral data. Thus by using filter foils with appropriate thicknesses with the K-edge absorption just below and above the K-edge energy of an element allows us evaluating the distribution of that element in whole painting. The aim of this thesis has been application of K-edge subtraction technique to determine pigment composition in paintings by using common X-ray sources with balanced filters. The K-edge digital subtraction technique actually is investigated for estimation of Cadmium element. By choosing the Silver Cadmium and Indium as filter materials with properly thickness and with K-edge energies very close to the Cadmium element, determination of cadmium content has been presented as qualitative and quantitative result. The technique is tested by using a RadEye200 CMOS sensor. Theoretical simulations considering the experimental setup of technique has been presented as preliminary investigation. Complementary analyses are preformed with micro X-ray fluorescence spectrometry (μXRF), in order to validate the technique for cadmium element with the same samples. Theoretical aspects of K-edge imaging, the algorithms used for elaboration of images, detailed information for cadmium pigments and samples tests, are described in first chapters. An evaluation of X-ray performance of RadEye200 CMOS sensor has been performed in terms of MTF, NPS and DQE. Preliminary theoretical simulation is performed to determine the range where KES imaging technique with balanced filters response linear with the element content in paintings. In chapter 5 after description of spectral characteristics of the tungsten X-ray tube with additional K-edge filtering, imaging and steps of elaborating images has been described. Determination of Cadmium element in its pigments it is performed by KES and Lehmann algorithm applied on digital images. Results obtained on different samples are presented by a couple of images, graphically, tabulated data, and discussed. In chapter 6 introduction in theoretical aspects of XRF spectrometry is presented. Measurements for sampling area of μXRF ArtaxBruker200 and for Cadmium content in same samples used for imaging are presented. In chapter 7 the results carry out from KES technique and XRF for cadmium element in different samples are compared. The comparison between KES and XRF technique performed with the same sampling area of two detection systems (CMOS sensor with SDD solid drift detector) on the same area of the test sample shows a very good correlation among them.