Correction of ultraviolet-induced fluorescence spectra for the examination of polychromy.

Ultraviolet-induced fluorescence spectroscopy is a commonly used technique for the characterization and identification of painting materials, such as organic binders and colorants. Its interpretation is strictly connected to both the experimental setup and an understanding of the physical and chemical interactions among materials in paint layers, which are commonly composed of a fluorescent organic binder and a pigment. When irradiated with ultraviolet radiation, the light emitted by fluorophores present in the organic binder undergoes several types of interactions, in particular scattering and absorption by neighboring pigmented particles and auto-absorption. As a result of scattering and absorption phenomena, the emission spectrum is deformed according to the physical properties of the surrounding pigmented particles. This can lead to shifts of the emission maxima and/or to the formation of apparent new emission bands. The extent of the modifications to the emission spectra, caused by auto-absorption and selective absorption phenomena, may lead to the erroneous characterization or identification of the fluorescent materials. As a consequence, the interpretation of the emission signal can be greatly compromised. A correction based on the Kubelka-Munk theory is proposed to evaluate the extent of the spectral distortion and is assessed on modern replicas of wall paintings of known composition. Although the model cannot be applied to all cases, qualitative distinctions between real and apparent emissions are achieved.

Analysis of protein-based media commonly found in paintings using synchronous fluorescence spectroscopy combined with multivariate statistical analysis.

The spectrofluorimetric analysis of protein-based binding media, which are commonly found as painting materials, is based on the detection of emissions from amino acids, as well as fluorescent degradation products that develop with aging. Laser-induced fluorescence spectroscopy, fluorescence excitation emission spectroscopy, and time-resolved fluorescence spectroscopy have all been employed in efforts to discriminate between commonly found proteinaceous binding media, including egg white, egg yolk, milk, and casein, as well as collagen-based glues from rabbit skin, ox bone, parchment, and fish. However, synchronous fluorescence spectroscopy (SFS), a rapid means of recording fluorescence properties of samples, has not been reported for the differentiation between binding media. This work focuses on the analysis of a large set of naturally aged films of different protein-based binding media using SFS with a range of different offsets between excitation and emission monochromators between 30-60 nm. An interpretation of synchronous fluorescence spectra of binding media is presented and is followed by an assessment and classification of a database of recorded spectra using multivariate analysis. Importantly, following SFS analysis of films of binding media, principal component analysis is used to differentiate among all the proteinaceous media considered on the basis of clustering of data. This application is thus a novel and nondestructive means for differentiation between protein-based binding media.

Stratigraphic analysis of organic materials in wall painting samples using micro-FTIR attenuated total reflectance and a novel sample preparation technique.

Wall paintings typically contain low concentrations of organic materials within a largely inorganic matrix and are characterised by their high porosity and long-term exposure to severe environmental conditions. The identification of organic materials within specific paint or plaster layers is challenging and the inherent characteristics of wall painting samples present further complications. Embedding materials (such as epoxy, polyester and acrylic-based resins) used to produce cross-sections often infiltrate porous and leanly bound samples, and compromise the interpretation of Fourier transform infrared attenuated total reflectance (FTIR-ATR) spectra and the qualitative identification of natural organic materials. An alternative method for the preparation of cross-sections of wall painting samples was developed using cyclododecane (C(12)H(24)) as a temporary consolidant and barrier coating to encapsulate the sample, and to provide necessary support to produce a cross-section through microtoming. Impacts of traditional and novel sample preparation techniques on the identification of organic materials with micro-FTIR-ATR were examined for both replica and real wall painting samples.

Time-resolved fluorescence spectroscopy and imaging of proteinaceous binders used in paintings.

The differentiation of proteins commonly found as binding media in paintings is presented based on spectrally resolved and time-resolved laser-induced fluorescence (LIF) and total emission spectroscopy. Proteins from eggs and animal glue were analysed with pulsed laser excitation at 248 nm (KrF excimer) and 355 nm (third harmonic of Nd:YAG) for spectrally resolved measurements, and at 337 nm (N2) and 405 nm (N2 pumped dye laser) for spectrally resolved lifetime measurements and fluorescence lifetime imaging (FLIM). Total emission spectra of binding media are used for the interpretation of LIF spectra. Time-resolved techniques become decisive with excitation at longer wavelengths as fluorescence lifetime permits the discrimination amongst binding media, despite minimal spectral differences; spectrally resolved measurements of fluorescence lifetime have maximum differences between the binding media examined using excitation at 337 nm, with maximum observed fluorescence at 410 nm. FLIM, which measures the average lifetime of the emissions detected, can also differentiate between media, is non-invasive and is potentially advantageous for the analysis of paintings.

Raman spectra of proteinaceous materials used in paintings: a multivariate analytical approach for classification and identification.

This work presents Raman spectra obtained from thin films of protein materials which are commonly used as binding media in painted works of art. Spectra were recorded over the spectral range of 3250-250 cm(-1), using an excitation wavelength of 785 nm, and several bands have been identified in the fingerprint region that correspond to the various proteins examined. Differences in the C-H vibrations located between 3200 and 2700 cm(-1) can be accounted for with reference to the amino acid composition of the protein-based binding media as well as the presence of fatty acid esters, in the case of egg yolk. In addition, the discrimination of different proteins on the basis of variations in spectra between 3200 and 2700 cm(-1) can be achieved following multivariate analysis of a large data set of spectra, providing a novel and nondestructive alternative based on Raman spectroscopy to other methods commonly used for the analysis of proteins.

Analysis of protein-based binding media found in paintings using laser induced fluorescence spectroscopy.

Laser induced fluorescence (LIF) spectroscopy of intrinsic fluorophores from organic media found in paintings (casein, animal glue and egg proteins) provides novel non-invasive means of characterisation of general classes of media on the basis of fluorescence emission arising from the presence of certain amino acids and their degradation byproducts. Proteins from traditionally employed binding media include collagen, casein, albumin and other egg proteins, of animal sources (skins, milk and egg respectively). Wavelength dependence of the spectra is presented for analyses of thin films of protein-based binding media.