Using the Quartz Crystal Microbalance to Monitor the Curing of Drying Oils.

Drying oils such as linseed oil form a polymer network through a complex free-radical polymerization process. We have studied polymerization in this challenging class of polymers using a quartz crystal microbalance (QCM). The QCM is able to measure the evolution of polymer mass and mechanical properties as the oil transitions from a liquid-like to a solid-like state. Measurements using bulk materials and thin films provide information about the initial polymerization phase as well as the evolution of the mass and mechanical properties over the first two years of cure. The temperature-dependent response of the cured linseed oil films was also measured. These results were combined with previously published results obtained from traditional dynamic mechanical analysis to give a unified picture of the properties of these materials across a very broad temperature range.

Multiscale organisation of lead carboxylates in artistic oil binders.

The supramolecular and mesoscopic architectures of lead-saponified linseed oil, used by painters since the Renaissance, have been characterised and linked to their rheological properties. The multi-scale organization of saponified oils has been demonstrated by SAXS (Small Angle X-ray Scattering), FF-TEM (Freeze-Fracture Transmission Electron Microscopy) and DIC (Differential Interference Contrast): some of the lead soaps (formed when the oil is heated in the presence of PbO) are organized into microscopic lamellar domains, distributed in a continuous matrix made up of unorganized species (partially saponified triglycerides, glycerol, remaining soaps, etc.). The concentration of lead soaps in the oil controls the average size and interaction between the lamellar domains. Linseed oil + PbO 17 mol% is viscous and consists of aggregates of lamellar domains isolated within the continuous unorganized matrix. In contrast, in linseed oil + PbO 50 mol%, the domains are homogeneously dispersed and form what can be described as a three-dimensional network, giving the system viscoelastic properties.

Neural Networks for Hyperspectral Imaging of Historical Paintings: A Practical Review.

Hyperspectral imaging (HSI) has become widely used in cultural heritage (CH). This very efficient method for artwork analysis is connected with the generation of large amounts of spectral data. The effective processing of such heavy spectral datasets remains an active research area. Along with the firmly established statistical and multivariate analysis methods, neural networks (NNs) represent a promising alternative in the field of CH. Over the last five years, the application of NNs for pigment identification and classification based on HSI datasets has drastically expanded due to the flexibility of the types of data they can process, and their superior ability to extract structures contained in the raw spectral data. This review provides an exhaustive analysis of the literature related to NNs applied for HSI data in the CH field. We outline the existing data processing workflows and propose a comprehensive comparison of the applications and limitations of the various input dataset preparation methods and NN architectures. By leveraging NN strategies in CH, the paper contributes to a wider and more systematic application of this novel data analysis method.

Connecting Rheological Properties and Molecular Dynamics of Egg-Tempera Paints based on Egg Yolk.

Egg-tempera painting is a pictorial technique widely used in the Middle Ages, although poorly studied in its physico-chemical aspects until now. Here we show how NMR relaxometry and rheology can be combined to probe egg-tempera paints and shed new light on their structure and behavior. Based on recipes of the 15th century, model formulations with egg yolk and green earth have been reproduced to characterize the physicochemical properties of this paint at the mesoscopic and macroscopic scales. The rheological measurements highlight a synergetic effect between green earth and egg yolk, induced by the interactions between them and the structural organisation of the system. 1 H NMR relaxometry emphasizes the presence and the structure of a network formed by the yolk and the pigment.

Time-of-Flight Secondary Ion Mass Spectrometry Imaging of Cross Sections from the Bacchanals Paintings of Nicolas Poussin.

The two paintings Infant Bacchanals (Museo Nazionale d'Arte Antica, Palazzo Barberini, Rome, Italy) executed by Nicolas Poussin (Les Andelys, 1594-Rome, 1665) in around 1626 are thought to have been painted "a guazzo", which means either with a glue or with an egg binding medium. To date, this has never been confirmed through analysis. Dual-beam time-of-flight secondary ion mass spectrometry (TOF-SIMS), using a bismuth cluster liquid metal ion gun and an argon gas cluster ion beam, allows the mapping of organic and inorganic matter on paintings cross sections, with the possibility to acquire submicrometer-resolution mass spectrometry images of the sample, together with high mass resolution using a delayed extraction of secondary ions. The surfaces of cross sections from both paintings were prepared beforehand either by polishing or by microtome cutting and then cleaned with the gas cluster ion beam directly inside the vacuum chamber of the instrument. The nature of the binders in the two paintings was investigated by TOF-SIMS analyses. By considering the uneven physical properties of the heterogeneous analyzed surfaces, several high-resolution images were recorded with different instrument settings. The detection of lipids seems to point toward an oil-containing medium, rather than a glue-binding medium. An emulsion made of oil and glue is another hypothesis to be explored to better understand the artist's working methods in his early career.

Artificial Intelligence for Pigment Classification Task in the Short-Wave Infrared Range.

Hyperspectral reflectance imaging in the short-wave infrared range (SWIR, "extended NIR", ca. 1000 to 2500 nm) has proven to provide enhanced characterization of paint materials. However, the interpretation of the results remains challenging due to the intrinsic complexity of the SWIR spectra, presenting both broad and narrow absorption features with possible overlaps. To cope with the high dimensionality and spectral complexity of such datasets acquired in the SWIR domain, one data treatment approach is tested, inspired by innovative development in the cultural heritage field: the use of a pigment spectral database (extracted from model and historical samples) combined with a deep neural network (DNN). This approach allows for multi-label pigment classification within each pixel of the data cube. Conventional Spectral Angle Mapping and DNN results obtained on both pigment reference samples and a Buddhist painting (thangka) are discussed.

Materials science challenges in paintings.

Through the paintings of the old masters, we showcase how materials science today provides us with a vision of the processes involved in the creation of a work of art: the choice of materials, the painter's skill in handling these materials, and the perception of the finished work.

A 19th Century "Ideal" Oil Paint Medium: A Complex Hybrid Organic-Inorganic Gel.

British 19th century painters such as J. M. W. Turner, commonly modified the properties of their paint by using gels called "gumtions". These gels allowed them to easily tune the paint handling and drying properties. The fascinating properties of these "gumtions" were obtained by adding lead acetate to a ternary system based on mastic resin, linseed oil and turpentine. Herein, we report and investigate in depth the rheological properties of these gels as well as their structure at a molecular and supra-molecular scale.

Watching Kinetic Studies as Chemical Maps Using Open-Source Software.

A nonproprietary software package, "PyMca", primarily developed for X-ray fluorescence analysis offers an easy-to-use interface for calculating maps, by integrating intensity (of X-ray fluorescence, as well as any spectral data) over Regions Of Interest (ROI), by performing per pixel operations or by applying multivariate analysis. Here we show that, while initially developed to analyze hyperspectral two-dimensional (spatial) maps, this tool can be beneficial as well to anyone interested in measuring spectral variations over one or two dimensions, these dimensions being time, temperature, and so on. Different possibilities offered by the software (preprocessing, simultaneous analysis of replicas, of different conditions, ROI calculation, multivariate analysis, determination of reaction rate constant and of Arrhenius plot) are illustrated with two examples. The first example is the Fourier transform infrared spectroscopy (FTIR) follow-up of the saponification of oil by lead compounds. The disappearance of reagent (oil) and formation of products (lead carboxylates and glycerol) can be easily followed and quantified. The second example is a combined extended X-ray absorption fine structure (EXAFS), diffuse reflectance infrared Fourier transform spectroscopy (DRIFT), and mass spectroscopy (MS) analysis of RhAl2O3 catalyst under NO reduction by CO in the presence of O2. It is possible to appreciate, in a single shot, Rh particles' structure and surface changes and gas release and adsorption in the reaction conditions.

Rheo-SAXS characterization of lead-treated oils: Understanding the influence of lead driers on artistic oil paint's flow properties.

From the 15th century onwards, painters began to treat their oils with lead compounds before grinding them with pigments. Such a treatment induces the partial hydrolysis of the oil triglycerides and the formation of lead soaps, which significantly modify the rheological properties of the oil paint. Organization at the supramolecular scale is thus expected to explain these macroscopic changes. Synchrotron Rheo-SAXS (Small Angle X-ray Scattering) measurements were carried out on lead-treated oils, with different lead contents. We can now propose a full picture of the relationship between structure and rheological properties of historical saponified oils. At rest, lead soaps in oil are organized as lamellar phases with a characteristic period of 50 Å. Under shear, the loss of viscoelastic properties can be linked to the modification of this organization. Continuous shear resulted in a preferential and reversible orientation of the lamellar domains which increased with the concentration of lead soaps. The parallel orientation predominates over the entire shear range (0-1000 s-1). Conversely, oscillatory shear coiled the lamellae into cylinders that oriented themselves vertically in the rheometer cell. This is the first report of such a vertical cylindrical structure obtained under shear from lamellae.

Effect of the molecular structure on the hierarchical self-assembly of semifluorinated alkanes at the air/water interface.

Semifluorinated alkanes (C(n)F(2n+1)C(m)H(2m+1)), short FnHm display local phase separation of mutually incompatible hydrocarbon and fluorocarbon chain moieties, which has been utilized as a structure-forming motif in supramolecular architectures. The packing of semifluorinated alkanes, nominally based on dodecyl subunits, such as perfluoro(dodecyl)dodecane (F12H12) and perfluoro(dodecyl)eicosane (F12H20), as well as a core extended analogue, 1,4-dibromo-2-((perfluoroundecyl)methoxy)-5-(dodecyloxy)benzene) (F11H1-core-H12), was studied at the air/water interface. Langmuir monolayers were investigated by means of neutron reflectivity directly at the air/water interface and scanning force microscopy after transfer to silicon wafers. Narrowly disperse surface micelles formed in all three cases; however, they were found to bear different morphologies with respect to molecular orientation and assembly dimensionality, which gives rise to different hierarchical aggregate topologies. For F12H12, micelles of ca. 30 nm in diameter, composed of several circular or "spherical cap" substructures, were observed and a monolayer model with the fluorocarbon block oriented toward air is proposed. F12H20 molecules formed larger (ca. 50 nm diameter) hexagonally shaped surface micelles that were hexagonally, densely packed, besides more elongated but tightly interlocked wormlike structures. Conversely, F11H1-core-H12 films organized into linear rows of elongated surface micelles with comparable width, but an average length of ca. 400 nm, apparently formed by antiparallel molecular packing.

Green earth pigments dispersions: Water dynamics at the interfaces.

HYPOTHESIS: The objective is to elucidate the multiscale dynamics of water within natural mixtures of minerals, green earth pigments that are mainly composed of phyllosilicates containing large amount of iron. In particular, the interaction of water with the different kinds of surfaces has to be probed. One issue is to examine the influence of surface type, basal or edge, on the dispersion quality. EXPERIMENT: The study was carried out using 1H variable field NMR relaxometry on various green earth pigment dispersions and concentrations. To analyse the data, a new analytical model was developed for natural phyllosilicates containing large amount of paramagnetic centres. FINDING: The proposed theoretical framework is able to fit the experimental data for various samples using few parameters. It allows to determining water diffusion and residence times in complex phyllosilicate dispersions. Furthermore, it makes it possible to differentiate the contribution of the basal and edge surfaces and their respective surface area in interaction with water. Moreover, NMR relaxation profile reveals to be highly sensitive to the structural aspect of the phyllosilicates and to the accessibility of water to iron, hence allowing to discriminate clearly between two very similar phyllosilicates (glauconite and celadonite) that are difficult to distinguish by standard structural methods.

Green Earth pigments aqueous dispersions: NMR relaxation rates dataset.

The data presented here are related to the research paper entitled "Green Earth pigments dispersions: water dynamics at the interfaces". The nuclear magnetic resonance (NMR) relaxometry data are provided for various aqueous Green Earth (GE) pigments dispersions with volume fraction spanning approximately from 0.1 to 0.5. For two of them (Cyprus GE and Bohemian GE), the NMR relaxation profiles from 10 kHz to 30 MHz (1H frequency) is given for several temperatures spanning from 293 to 318K. In addition, the X-ray diffraction pattern is provided for France GE (Kremer pigments) for the identification of the main mineral component. The nitrogen gas isotherms are provided for Cyprus GE and Bohemian GE.

Multilayers quantitative X-ray fluorescence analysis applied to easel paintings.

X-ray fluorescence spectrometry (XRF) allows a rapid and simple determination of the elemental composition of a material. As a non-destructive tool, it has been extensively used for analysis in art and archaeology since the early 1970s. Whereas it is commonly used for qualitative analysis, recent efforts have been made to develop quantitative treatment even with portable systems. However, the interpretation of the results obtained with this technique can turn out to be problematic in the case of layered structures such as easel paintings. The use of differential X-ray attenuation enables modelling of the various layers: indeed, the absorption of X-rays through different layers will result in modification of intensity ratio between the different characteristic lines. This work focuses on the possibility to use XRF with the fundamental parameters method to reconstruct the composition and thickness of the layers. This method was tested on several multilayers standards and gives a maximum error of 15% for thicknesses and errors of 10% for concentrations. On a painting test sample that was rather inhomogeneous, the XRF analysis provides an average value. This method was applied in situ to estimate the thickness of the layers a painting from Marco d'Oggiono, pupil of Leonardo da Vinci.