Novel approach of signal normalization for depth profile of cultural heritage materials

The investigation of cultural heritage materials is always complex and specific because unique. Materials are most often heterogeneous and organized in several layers such as mural paintings or corrosion products. The characterization of a complete artwork's stratigraphy is actually one of the questions of science conservation. Indeed, the knowledge of these layers allows completing the history of the work of art and a better understanding of alteration processes in order to set up an appropriate conservation action. The LIBS technique has been employed to study the stratigraphy of an artwork thanks to the ablation laser. However, as we know, atomic information could be insufficient to characterize two materials composed by the same based elements. Therefore, an additional molecular analysis, like Raman spectroscopy; is sometimes necessary for a better identification of the material in particular for organic coatings in cultural heritage. We suggest in this study to use Standard Normal Variate (SNV) as a common normalization for different kinds of spectra (LIBS and Raman spectroscopy) combined with a 3D colour representation for stratigraphic identification of the different layers composing the complex material from artwork. So in this investigation, the SNV method will be applied on LIBS and Raman spectra but also on baseline Raman spectra often considering as nuisance. The aim of this study is to demonstrate the versatility of SNV applied on varied spectra like LIBS, Raman spectra as well as the luminescence background. This original work considers the SNV with a 3D colour representation as a probable new perspective for an easy recognition of a structure layered with a direct overview of the depth profile of the artwork. - Highlights: • Influence of SNV normalization on net Raman spectra to identify the different layers of a sample • Influence of SNV normalization on luminescence background from Raman spectra to characterize the interfaces of a sample • Combination of both methodologies to perform an efficient identification of layers for stratigraphic samples • Complementarity of LIBS and Raman spectra after SNV normalization for stratigraphic analyses