X-ray Raman scattering for analysing molecular features in artists' paint samples
We have successfully designed, prepared, and conducted a comprehensive series of XRS experiments at the Stanford synchrotron radiation light source (SSRL) and European synchrotron radiation facility (ESRF) synchrotron facilities. These experiments were specifically dedicated to the in-depth study of mixed-media artists' paints, enabling us to explore their complex chemistry. High-resolution XRS experiments performed at SSRL, with spectral resolution approaching the core-hole lifetime of the element under investigation, enable the detailed characterization of subtle features in the XRS spectrum. This level of precision is instrumental in identifying variations in molecular bonding within complex materials.
The tempera grassa and protein-coated oil paints, prepared using identical starting materials, ultramarine blue, egg yolk, or egg white, and drying oil, exhibit notable differences in their carbon K-edge spectra. We expect these differences to arise because of the different preparation techniques employed, which significantly influence the spatial organization of the organic phases within the paints.
Such variations lead to the formation of different interfaces among the constituents, including protein-oil, oil-pigment, and protein-pigment interfaces. These interfacial differences impact the chemical interactions and reactions that occur during the drying and aging processes of the paint films. Consequently, the observed differences in the carbon K-edge spectra provide valuable insights into how preparation methods affect the physicochemical properties of paints.
X-ray induced radiation damage. During our initial experiments at the ESRF, we identified that prolonged acquisition times caused significant radiation damage, altered the paint chemistry, and resulted in spectra that did not accurately reflect the original chemical state of the samples. This finding underscored the importance of refining our experimental approach for the collection of high-quality XRS data from historical materials.
In subsequent experiments at SSRL we focused on mitigating radiation-induced damage and ensuring the collection of meaningful, undistorted data. To achieve this, we are developing a comprehensive protocol that optimized the experimental conditions. In addition, we are in the process of establishing stringent sample preparation and analysis requirements. This includes defining the optimal area available for analysis, determining the most effective mounting conditions, specifying acquisition times, and setting the number of spectra to be collected per sample. A critical focus is to ensure that each spectrum is acquired in a fresh, undamaged spot to preserve the integrity of the samples. These measures are being carefully tailored to achieve the required signal-to-noise (S/N) ratio while safeguarding the original chemistry of the paints.
Scanning transmission x-ray microscopy
During our initial STXM experiments on the HERMES beamline (SOLEIL) we investigated the organic phases of three paint samples from paintings by the Italian Renaissance painter Carlo Crivelli (c. 1430–1495) (Figure 1). Ultrathin sample sections (~100 nm thick), required for successful STXM experiments, were meticulously prepared using focused ion beam (FIB).
The analysis revealed distinct chemical signatures that differentiated regions within individual paint layers and across multiple layers. The characteristic core-electron transitions at the carbon K-edge provided critical insights into the chemical bonding of organic species. These findings highlighted significant chemical heterogeneity within the paint layers and at the pigment-binder interfaces. This heterogeneity is the result of complex interactions among organic and organic-inorganic phases during the formation and aging of the paint film. Such phenomena are challenging to document using conventional laboratory-based techniques because of their limited spatial resolution, underscoring the importance of nanoscale imaging approaches. Following the STXM experiments, we performed transmission electron microscopy (TEM) at University of Pisa to investigate the elemental composition and nanoscale distribution of inorganic phases to achieve a complete characterization of the ultrathin painting samples.
Deep UV photoluminescence (PL) imaging spectroscopy
Photoluminescence (PL) imaging spectroscopy, performed at DISCO beamline at the SOLEIL synchrotron facility, deciphered the spatial distribution of oils and proteins in complex mixed-media paints by utilizing their autofluorescence properties. Deep-UV excitation provided the chemical selectivity and submicrometer lateral resolution necessary for the precise identification and localization of distinct components within the mixtures. When the excitation and detection ranges were optimized to match the unique absorbance and emission properties of each material, the luminescence of specific species was enhanced, ensuring clear differentiation. This approach established a direct correlation between the paint preparation techniques and the resulting microstructure of the paint layers.
Samples under study. During our beamtime at DISCO beamline we analysed a range of model samples, including oil paints, tempera paints (prepared with egg yolk and egg white) and mixed-media paints such as tempera grassa and protein-coated oil paints (also prepared with egg yolk and egg white). The ultramarine blue pigment was selected for its non-autofluorescent properties and its ability to preserve the fluorescence of the binder, making it ideal for studying binder interactions without interference.
