EXPLOITING THE POTENTIALITY OF SURFACE-ENHANCED RAMAN SCATTERING (SERS) FOR THE IDENTIFICATION OF NATURAL DYES IN ANCIENT TEXTILES: FROM HYPHENATED TECHNIQUES TO IN-SITU ANALYSIS

Surface-enhanced Raman scattering (SERS) is nowadays a well-assessed spectroscopic tool for the identification of historical natural dyes in the scientific community of analytical chemists, even if not yet entirely established in the common practice of laboratories devoted to the chemical characterization of works of art. This technique is based on the enhancement of Raman bands that takes place when a molecule is placed in the proximity of, or possibly adsorbed on, a roughened metal surface, provided by a pre-treated electrode, a metal colloid or a metal thin film. Silver is the most frequently used metal, thanks to the relevant enhancement it provides with visible excitation wavelengths, and colloids are the preferred substrates for analytical applications, given their easiness of preparation. Researchers first composed databases of SERS spectra of natural dyes in solution obtained with different excitation wavelengths to provide suitable references for the identification of the same in ancient textiles and paintings. Subsequently, new protocols of analysis have been developed to improve the identification of dyes in complex matrices, such as micro-samples coming from valuable artistic and historical objects. In the present Ph.D. project, surface-enhanced Raman scattering is studied in order to overcome some limitations both inherent to the technique and to ancient colored samples and artifacts. The methodologies developed evolved from the knowledge of a very recent literature, with continuous cutting-edge progresses. In more detail, as suggested above, to complete the transfer of the SERS technique from the research laboratory to the specialized analytical laboratory, a further enlargement of the database is still required. Indeed, the range of natural dyes used historically worldwide in craftsmanship and art is really huge and, even if some of them were employed in different geographical areas, many others were typical of a given zone being obtained from local vegetable (or, less frequently, animal) sources. In order to contribute to accomplish this aim, a SERS study of anthocyanins by means of SERS was carried out for the first time and reported in the present doctoral thesis. The work regarded both SERS analyses of anthocyanins as pure molecules and in mixture with other compounds in plants and textiles. Moreover, the hypothesis of an anthocyanin content of the so-called folium dye, whose main chromophores are still unknown, was assessed. It is worth remembering that, even if SERS shows, as usual for vibrational spectroscopy, a remarkable specificity in the identification of an analyte, its use can be partially hindered when the sample is more properly a mixture of compounds, as frequently happens for natural dyes that contain many colored and uncolored substances and that were often used in mixtures to yield a given tint. Indeed, this is the reason why a separative technique, namely high-performance liquid chromatography (HPLC), is still the most widely employed analytical tool for this sort of studies, usually coupled with UV-visible diode-array detection (HPLC-PDA). In spite of its sensitivity, however, this kind of detection can sometimes lack in specificity. Therefore, SERS proposes itself as an alternative, or parallel, detection method for HPLC analysis of natural dyes. Few studies have been reported in the literature about the development of equipments combining HPLC and SERS techniques, mainly applied to the analysis of drugs. A part of the present project was therefore devoted to the optimization of a laboratory-made HPLC-PDA-SERS hyphenated system to be used specifically for the identification of natural organic dyes. Of course, being conceived for solutions, the application of SERS as well as of HPLC to the analysis of colorants in ancient objects requires as first step the extraction of the dyes from a sample drawn from the object itself. Dyestuffs were indeed fixed on textiles or incorporated in painting as metal complexes. Extractions are of course time-consuming, as well as obviously destructive with regard to the sample. For this reason, non-extractive protocols based on SERS analysis performed directly on the sample itself have been recently suggested in the literature. “Extractionless” analyses can be performed on textile fragments as such, or on samples previously hydrolyzed by acid vapors. In this latter case, usually a hydrolysis pre-treatment of micro-samples is executed, in order to break the dye-mordant complexes without removing the dye from textiles, followed by direct SERS analysis onto the sample. A further aim of the present project was indeed to contribute to the development of an “extractionless” protocol to be applied to the widest range of dyes and to the least amount of sample, possibly reducing problems such as the interference of the textile substrate and the fluorescence background typical of ancient samples. The use of a less-energetic source of radiation, as in the Fourier-transform Raman spectroscopy (1064 nm), was therefore investigated to match these aims. Finally, it should be noted that even extractionless methods are not completely respectful of the integrity of the artefact being examined, requiring a sample even if as small as possible. Being this an issue of the outmost importance for conservators, it is evident that an entirely non-destructive application of SERS technique would be highly desirable. Some attempts in this regard were also carried out, both at Duke University (USA) in collaboration with the research group of Prof. Tuan Vo Dinh and at the Department of Chemistry of Università degli Studi di Milano. Different solid nanoplatforms were experimented and tested for the enhancement of Raman signals of dyes directly on textiles, exploiting solid-solid interactions between the dyes and the nanomaterials with the final aim to allow the identification of these molecules without need of displacing the artefacts to which they belong from their site of storage. Finally, it is worth underlining that, as SERS protocols need to be tested not only on reference dyes, but also on ancient samples to verify their performance in real analytical problems, the present project investigated also the applicability of the SERS techniques developed for recognizing dyes in ancient textile samples from artistic collections.