In situ elemental analysis and imaging of painted works at defined depth by simultaneous confocal and macro - X-ray fluorescence – DepthPaint

The methods used are X-ray fluorescence (XRF) based techniques:
-Macro-XRF: consists of a head mounted on a motorized X-Y stage allowing movement along the work. The head includes an X-ray generator, an X-ray SDD detector and a microscope-type video camera. A collimator is located at the end of the X-ray source to project a parallel, wide energy range beam. The detector is placed at a distance of 15 mm from the surface of the paintings. The data (a matrix of pixels each containing a spectrum) are stored in EDF format, and processed with the open source program PyMCA. With this program, elementary black and white maps in PNG format can be obtained.
Confocal XRF (CXRF): using a confocal system composed of two X-ray optics, one in the excitation channel and the other in front of the detector, an ellipsoidal micro-volume can be defined by the overlapping foci of the two optics. Moving the artwork into the probing volume with micrometer precision, a depth scan can be performed, allowing exploration of the elemental distributions within the sample. In this device, the object is placed on a motorized X-Y-Z stage, and away from the measuring head. A full-lens polycapillary lens is installed at the exit of the X-ray tube, and a poly-CCC lens in front of the detector. The working distances are 6-11mm. The energy range of the systems is limited by the transmission of the optics between 3 and 20 keV. The acquisition time per spectrum at each analysis depth is around 100 s. The XRF spectra obtained for each depth value are adjusted using the PyMCA program to obtain a normalized representation of the intensities of the elements in each depth, indicating the location of the different layers.
Portable XRF: is an efficient, fast, non-invasive, safe and well-established method to measure single points on works of art and obtain information on elemental composition, using a small diameter X-ray beam and a detector.
Synchrotron microXRF: similar to point XRF with better detection limits uses a synchrotron as the source of the X-rays produced, typically with a beam diameter of a few µm.
-SEM-EDX analysis of stratigraphic sections: cross-sections are analyzed using a scanning electron microscope coupled with energy dispersive X-ray spectroscopy. Samples are polished and coated with carbon to avoid surface charging.
-Hyperspectral imaging: technology also called Reflectance Imaging Spectroscopy that allows to obtain a large array of images resolved in a large number (usually more than a hundred) of spectral bands. Thanks to the representation of the spectral bands in two-dimensions the distribution of pigments can be imaged on the painting.

In these first 18 months were reached most of the objectives that had been set:
-Recruitment of the doctoral student and his training in the different XRF equipment and data processing. Familiarization with the analysis of precious works of cultural heritage (field trip to the Font-de-Gaume cave to discuss the approach to be followed in the next analyses, training with portable XRF with large paintings at the C2RMF laboratory in Versailles, analysis of prehistoric samples at the PUMA line of the SOLEIL synchrotron)
- Constitution of the state of the art in the field of CXRF analysis. Comparison of the existing equipment in different laboratories
- Turn on of the old LouX3D CXRF equipment and testing its performances with thin metallic samples, showing that the device continued to work properly, and analysis of a test painting.
-First complete analysis of a painting under restoration at C2RMF «Virgin and Child surrounded by saints and donor«. Reports on the combined MA-XRF and confocal analyses show in a non-invasive way the complex modifications of this painting. Information in three areas of interest of the painting was found with the analysis of four stratigraphic sections analyzed with SEM-EDX and lateral scan (y) with the LouX3D CXRF prototype, the process of the data obtained with MA-XRF, 22 depth profiles (z) with the LouX3D CXRF set-up and hyperspectral imaging. The various results were shown at the inART 2022 and MA-XRF 2022 conferences, and a paper is being submitted (proceedings inART).
-Processing of micro-XRF and X-ray absorption chemical analysis and imaging data obtained in 2020 at the PUMA line of the SOLEIL synchrotron from limestone samples of the prehistoric cave of Font-de-Gaume and realization of a poster presented at the GMPCA 2022 conference.
- Development of the equipment design and selection of technical solutions (XRF tube, mechanical movements)
-Progressive purchase of the components. For the moment, an X-ray SDD detector has been purchased, and a PUMA public purchasing contract has been launched for the acquisition of two polycapillary lenses and an X-ray tube.

Nevertheless, difficulties also arose. Here are some of the difficulties encountered, with the solutions taken:
-The planning of the purchase of the components is not totally mastered due to the fact that the manufacturers have difficulties respecting the delivery times (due to the context of the current crisis).
-There were additional delays when the LouX3D CXRF device was restarted in order to make it compliant with radiation safety regulations.
-Difficulty establishing a perfect planning of the analyses of the test works in the long term due to their irregular and short-term availability at C2RMF

In the section « Main issues raised & general objectives « can be consulted the objectives, and in the section «Main results«, those that we arrived at. For the next year (October 2022-October 2023 period), our perspectives are to be able to:
- Analyze other paintings with the stationary confocal XRF LouX3D as well as with the MA-XRF to extend the diversity of the studied paint layer systems. This study should have a duration of six to eight weeks.
-Publicate the paper submitted to EPJ+ after the inART 2022 conference on the results obtained on the painting «Virgin and Child surrounded by saints and donor« with different techniques, including MA-XRF and FCXRF.
- Purchase the missing components, and progressively assemble the instrument as the parts arrive.
-Programme the software of the instrument (data acquisition control and processing routines).
- Set up the partnership with the company BRUKER on a possible joint development of some parts.
- Test it on model samples once the instrument is up and running. To begin, with fine metal samples, but also the test painting used when the LouX3D was restarted. This study should have a duration of around four weeks.
- Analyze samples from caves in Northern Spain with µFX at the PUMA line, SOLEIL synchrotron. The analysis time is a few days, but with the data processing it would be around four to five weeks.
-Participate in two scientific conferences (about instrumentation, scientific data evaluation, cultural heritage) at national and international level.
-Compare the performance of our new instrument with other CXRF prototypes. This would take three to four weeks.
- Write the technical notice and the manual of the instrument.
-Certify the instrument with the radiation protection regulations
- Submit a first paper presenting the first results of the application of the instrument. This would take four to six weeks.
- Carry out a mission in the Font-de-Gaume cave with the developed instrument. The mission will have a duration of one week, with data processing taking two to four weeks. If possible, a mission in a museum to analyze easel paintings will also be realized.

- Submission of a paper in EPJ+ following the inArt conference (Tapia et al.). Title: «Efficiency of combined MA-XRF and CXRF to give non-destructive insights about changes of a historical painting«.
-Oral presentation at the MA-XRF conference (Delft, September 26-27, 2022). Title: «Advantages and limits of MA-XRF on paintings with important iconographic changes«.
-Poster presented at the inART conference (Paris, 28June-1July 2022). Title: «Non-destructive confocal XRF depth profiling combined with MA-XRF and SEM-EDX of a historical painting gives new insights about its history
-Poster presented at the GMPCA conference (Chambéry, 2-6 May 2022). Title: «Analysis and chemical imaging by micro-fluorescence and X-ray absorption of calcareous samples from the prehistoric cave of Font-de-Gaume«.
-Presentation of a poster at the ED 388 PhD students day, 17/06/2022. Title: « Non-destructive confocal XRF depth profiling combined with MA-XRF and SEM-EDX of a historical painting gives new insights about its history «.

This project aims at developing in situ confocal X-ray fluorescence spectrometry (C-XRF) combined with macro-XRF scanning (MA-XRF) for depth-resolved elemental analysis and imaging on prehistoric figures in decorated caves and on historical paintings in museums. The analytical system that will be developed is a portable prototype for C-XRF that is additionally equipped with an XRF detector and X, Y, Z linear stages which allows for MA-XRF. This experimental device has a light and mobile head, which allows better control of conducting an in-depth analysis of variations in the composition of the painting, without contact and without moving the works. Our research hypothesis is that we can improve the characterization of complex heritage materials by using mobile instruments that allow for more accurate sample selection by better choosing the studied area and controlling the analytical volume. The analyses carried out in this way make it possible to precisely determine, in a qualitative and semi-quantitative way, the chemical composition layer by layer of the painted works. More reliable information on the creation and sequence of creation of prehistoric figures can be obtained with these methods. For easel paintings, an improved vision of their complex paint stratigraphy will be gained by combining portable MA-XRF and C-XRF and will considerably reduce the samples to be taken. The new device will be tested for its performances and validated. A semi-automatic spectrum acquisition and evaluation procedure will be developed. In order to allow a sustainable analysis and by limiting as far as possible the radiation damage on the works, the safe boundaries of the analyzes will be defined before carrying out analyzes and imaging in situ on the one hand the paleolithic parietal works of the Font- cave de-Gaume in Dordogne and on the other hand the painted works of the national museums studied by the C2RMF. These new data will be integrated into the C2RMF database and will thus be openly accessible to other scientists.