Applications of Raman spectroscopy in art and archaeology

Abstract

The increasing importance of the application of Raman spectroscopy in art and archaeology is illustrated by an increasing number of research papers published each year and by the scientific conferences and sessions that have been dedicated to this research area in the past. A great number of authors consider Raman spectroscopy as the most adequate instrumental technique to identify and characterize the original and decayed compounds present in objects of art and archaeology. The biennial International Congress on the Application of Raman Spectroscopy in Art and Archaeology (RAA) is for sure the most important event dedicated to this topic. The RAA conferences promote Raman spectroscopy and play an important role in the increasing field of its application in art and archaeology. These prominent international events have a long tradition. Previously, they were held in London (2001),1 Ghent (2003),2 Paris (2005),3 Modena (2007),4 Bilbao (2009),5 Parma (2011)6 and Ljubljana (2013),7 and Wroclaw (2015).8 The ninth edition of the International Congress on the Application of Raman Spectroscopy in Art and Archeology (RAA 2017) was held in Evora (Portugal) from 24 to 27 October 2017. As in the previous editions, the scientific program was focused on the characterization of materials, conservation issues affecting cultural heritage, Raman spectroscopy of organic-based materials, and Raman applications in archaeology and forensics with authenticity research. These studies were presented along four plenary lectures, 35 oral presentations, and 32 poster presentations. The number of active participants was 96 delegates from 18 countries among the 196 authors that presented at least one work to the Congress. The high quality of the contributions is reflected in the selected 15 manuscripts covered in this special issue, ordered in topics suggested to participants when this RAA2017 edition was announced. The most interesting aspects of such manuscripts are highlighted below. Raman microscopy is a spectroscopic technique able to detect compounds that are present in the samples at minor/trace levels thanks to the precise focusing on particles of interest, although the detection also depends on the Raman scattering of the measured compound. This is especially important when dealing with samples belonging to the arts and archaeological fields. Among the different archaeological sites in the world, Pompeii is one of the most important ancient cities. In the work presented by Marcaida et al.,9 Raman microscopy supported by X-ray diffraction (XRD) was used to detect traces of mineralogical phases of volcanic origin and contaminations in red and yellow ochre pigments recovered from the archaeological site of Pompeii in their original bowls after being buried for hundreds of years. The detected mineral phases of volcanic origin were langite [Cu4(SO4)(OH)6·2H2O], jarosite [KFe3(SO4)2(OH)6], atacamite [Cu2Cl(OH)3], and anatase (β-TiO2), a polymorph of titanium oxide found in igneous rocks while some contamination compounds such as Egyptian or Pompeian blue (CaCuSi4O10) and huntite [Mg3Ca(CO3)4] were also identified. As expected, XRD could not detect the minerals at trace levels. Raman spectroscopy and time of flight secondary ion mass spectrometry (ToF-SIMS) were used by Sodo et al.10 to answer conservation questions left open after the preliminary analysis performed by the restoration team on the Bosch painting Saint Wilgefortis Triptych, within the project “Bosch in Venice.” These two techniques were used on five cross-sections to perform a detailed study of pigments and bindings degradation processes, together with the identification of organic components (binders/lakes) and the characterization of restoration products. Raman measurements have shown the presence of several degradation products: (a) calomel on the red pigment cinnabar, (b) calcium-oxalate (weddellite) in a not original external layer, and (c) lead soaps in several layers of the studied samples. Although the detection of lead soaps suggests the use of oils as binders, Raman spectroscopy did not give conclusive information about such binders, neither on the possible presence of red lakes, suggested by independent UV observations. However, ToF-SIMS measurements revealed the presence of lead palmitate and stearate in the painting layers along with miristic, palmitic, oleic, and stearic acids, confirming the use of an oil medium, likely linseed oil. Additionally, ToF-SIMS showed the presence of polydimethylsiloxane, likely from a previous restoration treatment in the 1990s, in the external layer of the investigated cross-sections. Due to experimental difficulties during the analysis of the red lake, authors could only hypothesized on the presence of alizarin. The work by Antunes et al.11 aimed to characterize materials and production techniques used in Ecce Homo painting (P1) from the beginning of the 17th century, made at the time of Iberian Union government in Goa, which belongs to Museum of Christian Art at Old Goa (India). As the knowledge of painting materials from Goa at that epoch is an almost unknown issue, the scope of the work was not only the characterization of the pigment layers, but also the materials used for the ground/priming layer, and the ascertaining of the manufacturing processes. For this study, cross sections were analyzed by confocal micro-Raman spectroscopy complemented with other techniques: optical microscopy, energy dispersive X-ray fluorescence, scanning electron microscopy–energy dispersive spectroscopy, and micro-Fourier transform infrared spectroscopy. Paintings surfaces were also examined by infrared reflectography in order to find subjacent drawings. By scrutinizing this P1 painting, scientific confirmation on materials and technique was achieved, and the comparison with two other paintings, one Goan (P2) and another Portuguese (P3), from the same period and under the same theme was established. Results allowed identification and origin of paintings under study, considered as a Goan feature the use of orange–brown earth pigments, with elevated iron oxide content, and kaolinite the typical matrix ground, found in both P1 and P2 paintings, while the Portuguese paint P3 showed a ground based on calcium sulfate. The pigments used in the analyzed Goan and Portuguese paintings were similar. This represent new data for art history and for painting conservation decision making. The Raman spectra of the Na2SO4–K2SO4–H2O system are not well-defined in the literature. Specifically, the proper identification of sodium–potassium sulfate (aphthitalite or glaserite, K3Na(SO4)2) and anhydrous sodium sulfate (thenardite, Na2SO4) is particularly problematic because their vibrational profiles present the same main Raman band at 993 cm−1 and very similar bands at low frequencies. As proved in bibliography, the similarity of their spectra can often lead to uncertain or erroneous identifications. Considering that aphthitalite and thenardite can be found as degradation products on Built Heritage materials, and the degree of danger associated to them is not the same; the second one the being most harmful, the resolution of this problem has a critical importance for conservation works in Built Heritage. This was the reason to afford the deep study of the Raman spectra of aphthitalite and thenardite, performed by Prieto-Taboada et al.,12 to identify the vibrational fingerprints enabling their correct identification. Two different Raman instruments were used to guarantee that the spectrum of aphthitalite displays characteristic secondary bands at 1,084 and 1,202 cm−1; while the bands at 1,100, 1,129, and 1,152 cm−1 seem to be characteristic of thenardite. Furthermore, when those secondary bands are not observed or mixtures of both compounds are present, the ratio between their most intense bands at 452(s) and 993(vs) cm−1 is the key for their correct characterization: Thenardite always presents a ratio between the area of 993/452 cm−1 bands around 12, whereas in the case of aphthitalite, it is around 2. On the whole, this study fills the gaps observed in literature and gives the solution for the correct identification of aphthitalite and thenardite even when secondary bands are not observed. The Room of the Beds, in the Royal Bath of Comares of the Alhambra monumental ensemble, maintains the testimony of the controversial restorations carried out in the 19th century in an attempt to imitate the lost original appearance of the authentic Nasrid plasterwork. To clarify this important issue, Arjonilla et al.13 studied those polychrome plasterwork decorations using always Raman microspectroscopy with complementary techniques: (a) dispersive X-ray fluorescence, to help in the identification of pigments and extenders, (b) scanning electron microscopy–energy dispersive X-ray spectroscopy, to gain additional information about the morphology of the painting layers, and (c) infrared microspectroscopy, to provide insight into the nature of the organic materials employed as binders. Vermillion, synthetic ultramarine blue, hematite, and carbon black were clearly identified in red, blue, brown, and black decorations. Green decorations were executed with a copper-arsenic pigment that could not be unambiguously identified although the presence of typical stretching Raman bands of arsenate suggest a possible alteration processes of copper arsenite pigments. The pictorial layer was applied over a preparation layer of white lead and barite mixed with a proteinaceous binder. The presence of anglesite and other phases related to hydrocerussite alteration was also evidenced. Finally, the comparison of materials and execution technology used in the 19th century redecoration with those identified in original Nasrid decorations revealed important differences in both aspects. Quantification with nondestructive techniques is not very well developed in the field of Cultural Heritage despite its interest especially when original and decayed compounds coexist in the same area. Although several works have been published using Raman spectroscopy for quantifying in specific cases, depending on the methodology used, the information provided by this technique is not complete, and the results could lead to misunderstanding when dealing with unknown samples. With the aim to advance in the quantification of mineral phases present in the surface of documents, artworks, or facades, a novel double quantification using Raman imaging (its representativeness would be higher than point-by-point analysis) and laser-induced breakdown spectroscopy (LIBS) analyses were proposed by Aramendia et al.14 To show the advantages of the new quantitative analytical methodology, several dolomitic marble samples, with some calcite impurities, covered or not by a calcium oxalate (whewellite, CaC2O4·H2O) layer, were analyzed in order to optimize the novel procedure. The agreement between the quantitative results from the independent analyses of the Raman Image data and the LIBS data is consistent within the uncertainty (instrumental uncertainty and uncertainty coming from the heterogeneity of the samples at micrometric level) arising from both techniques, if the sample area is the same (in this case it was 200 μm in diameter). Further, the same methodology was applied on the same samples but using point-by-point Raman analysis with portable instruments and portable LIBS information, both with 200-μm spots in diameter, showing again a great agreement between them and with the results obtained using the laboratory instruments. Thus, the proposed methodology can be used in the field and in the lab to estimate concentrations of mineral phases present in surfaces of Cultural Heritage assets. The degradation of cellulose nitrate image heritage is strongly susceptible to degradation, being a major conservation challenge. Infrared spectroscopy has been the traditional technique in the assessment of the polymer degradation, but new in situ diagnostic tools to monitor the initial stages of degradation are needed. This has been afforded by Neves et al.15 using Raman spectroscopy to follow the chemical changes of new cellulose nitrate films produced and irradiated as an accelerated aging process. Raman spectroscopy confirmed in the earlier stages of degradation the mechanisms proposed in literature for the advanced ones. Moreover, a new relevant feature that characterize them was identified: The intense Raman band at 1,046 cm−1 associated to nitric acid. These results on mock-up samples were compared with the observations on old cellulose nitrate cinematographic films, stored inside an aluminum can, concluding that the plasticizers, whose identification was more straightforward using Raman microscopy, interfere in the spectroscopic regions where chemical changes are better observed, making it difficult to draw conclusions. However, nitric acid and silver nitrate peaks were found in the Raman spectra of the stored films, confirming the unstable and noxious environment inside the aluminum can due to the degradations. The chemical identification of materials is the first step for developing and implementing recommendations for the care and display of resin cast and plastic objects in museum collections. The paper from Klisińska-Kopacz et al.16 presents the advantages as well as the limitations of portable Raman to perform such chemical identification of polymers used in cast sculptures. For that, a comparative analysis of portable and benchtop Raman instruments was performed in the study of materials found in the contemporary art collections of the National Museum in Krakow, Poland. The results of the Raman study were complemented with those obtained using other analytical techniques such as Fourier transform infrared spectroscopy, near infrared spectroscopy, and gas chromatography coupled to mass spectrometry, in order to verify their accuracy. The results of this study showed that portable Raman spectroscopy is a suitable technique for the identification of compounds in plastics directly on museum collections. A handheld Raman spectrometer was used by Saviello et al.,17 for the first time to the knowledge of authors, to perform surface-enhanced Raman scattering (SERS) analysis of felt-tip pen drawings on paper support. The use of SERS combined with near-infrared excitation wavelength allowed authors to overcome the strong fluorescence background interference so that diagnostic spectra were obtained from drawings made by 18 pens of 10 different colors and five different brands. The analysis revealed that often, mixtures of two or more dyes are used by pen manufactures in order to obtain the desired color hue, and also, that same colors from different pen brands often display different dye composition. Handheld SERS was also used to ascertain the dye content of historical felt-tip pens extensively used by the Italian film director Federico Fellini. The results from this work open the way to the use of handheld instrumentation for analysis of ink-based art works and documents, providing a starting point toward the completion of reference SERS databases for dye identification. Raman microscopy was used, in situ, by Angelin et al.18 to identify pearlescent pigments used to create luster in poly (methyl methacrylate) artworks from Ângelo de Sousa (1938–2011). Plumbonacrite and bismuth oxychloride were unequivocally characterized by comparison with reference materials, synthesized for this study, highlighting the vibrational pattern (Raman assisted by IR spectroscopy) of both pigments. Moreover, authors confirmed the differentiation between the two basic lead carbonates reported in the literature to be pearlescent: hydrocerussite Pb3(CO3)2(OH)2, characterized by two sub-bands in the ν1 stretching at 1,048 and 1,051 cm−1, and plumbonacrite Pb5(CO3)3O(OH)2, characterized by three strong different sub-bands (1,048, 1,052, and 1,056 cm−1). Raman microscopy was used for the first time as the fingerprint tool for the molecular identification of pearlescent pigments in plastic materials, having also the capability to check the presence of degradation products. Based on these findings, better informed conservation strategies for the acrylic sculptures could be developed in the near future. Limited knowledge is available on the long-term behavior of synthetic and natural polymers used in art. The work presented by Gomez et al.19 settled the foundations for studies on organic materials suffering from autoxidation processes. They proposed to use nondestructive surface-enhanced Raman spectroscopy (SERS) for the detection of commonly present low molecular weight degradation products, because (a) such products can be considered molecular markers for diagnostic investigations, and (b) SERS allows detection limits never reached before with other spectroscopic techniques. 3D aluminum (Al)-coated SERS substrates were optimized for in situ sampling of artifacts, testing different reference materials, sampling strategies, and instrumental conditions. As linseed oil is an organic material widely used in art, whose degradation mechanisms are well-known, authors performed their SERS work on linseed oil model samples (dry films with a thickness of 50 ± 10 μm), considering that many polymeric compounds used in artworks follow similar degradation pathways. Linseed oils films were subjected to natural and accelerated photo-aging, and silicone strip samplers (using different solvents with different polarity, from water to hexane) were used to remove micrometric amounts of samples for the SERS analysis to detect low molecular weight degradation products such as acids and esters but also low chain hydrocarbons. The composition of archaeological glass beads and the manufacturing techniques employed in their production was ascertained by Costa et al.20 based on a new nondestructive methodology that combines micro-Raman spectroscopy and micro-X-ray diffraction (μ-XRD), complemented by variable pressure scanning electron microscopy coupled with energy dispersive X-ray spectrometry. The work revealed that most samples belong to the alkaline glass family and identified the main colorants used in the manufacture of the studied glass beads. Dark blue and turquoise glass were colored using cobalt ions and copper ions, respectively. Amber or light brown hues were produced using the iron–sulfur amber chromophore. Iron ions were also used to produce green, yellow, cream-colored gray, and black hues. White glass was produced using calcium antimonite phases that were also used as opacifying agents. This work highlights the importance of using micro-Raman spectroscopy to study ancient glass artifacts, allowing not only the determination of the glass family of heavily degraded samples but also the identification of evidences of rearrangement in the silicate network following selective leaching. Moreover, the combined use of micro-Raman spectroscopy and μ-XRD permitted the identification of the opacifying agents and heat treatment used in the manufacture of the studied opaque glass beads. Raman spectra of 42 minerals, commonly and less frequently used as gemstones in gemmology s. s. or works of arts, were selected and analyzed by Culka and Jehlička21 using a portable sequentially shifted excitation Raman spectrometer with the aim to compare the results with this new set-up from the ones obtained with non-sequentially shifted Raman spectrometers. The analyzed minerals included chrysoberyl, diamond, emerald, garnets, magnesiotaaffeite-2 N'2S (taaffeite), magnesiotaaffeite-6 N'3S (musgravite), ruby, topaz as well as other less precious minerals. Some minerals were represented by several different gemstone samples, having different causes for fluorescence, in order to ascertain the capability of the fluorescence removal by the sequentially shifted spectrometer on a larger set of natural mineralogical samples. This novel instrument is able to effectively suppress the laser-induced fluorescence that is sometimes present in the Raman spectra obtained with non-sequentially shifted spectrometers on natural and often colored samples of mostly silicate minerals. Moreover, the portable sequentially shifted Raman spectrometer allowed the analyses of gemstones and works of arts in situ, which can be relevant in many cases. Castro Mendes et al.22 described the results of the first study of the collection of art materials, painted works, and techniques, belonging to the Lasar Segall Museum in Brazil. Those materials and documents come from the early 20th century. The paint materials from the collection consisted of over a hundred glass vases, tubes, and packets containing powdered pigments, inks, oils, and varnishes from where 50 samples were collected. Powdered and dried ink (labeled “ink for posters”) samples were analyzed using confocal Raman microscopy spectroscopy, assisted with Fourier transform infrared spectroscopy, energy-dispersive X-ray fluorescence, and X-ray diffraction. The artists palette chromophores such as ultramarine, cobalt and Prussian blue, goethite, hematite, and magnetite from natural ochres, lead basic carbonate, lead carbonate, anatase, zinc carbonate hydroxide, zinc oxide, lead chromate, carbon, cadmium sulfide, copper acetate arsenite, verditer, celadonite, serrabrancaite, vermilion, as well as the dyes PY1, PY3, PR3, and PO5 were identified mainly by Raman spectroscopy. The combination of EDXRF and Raman techniques was essential for identifying umber pigments. Nevertheless, the violets found are not mineral in nature but organic. The color palette found in paintings of Lasar Segal from 1919 is closely related to the characterized pigments from the museum. The descriptions of the results for each sample and painting were made available to the museum to assist in future conservation, dating, and authentication procedures of the works. Underwater Cultural Heritage understands all traces of human existence having a cultural, historical, or archaeological character, which have been partially or totally underwater, in periodic or continuous forms for at least 100 years, according to UNESCO. This last work of the special issue devoted to RAA2017 was performed by Estalayo et al.23 on two pieces recovered in 1999 from a sunken English shipwreck of the XVII century, located in Bakio (Basque Country, Northern Spain). The two analyzed pieces were an iron anchor and a swivel gun, both exposed in Bakios Town Hall after their restoration in 2005. The aim of the work was to study the elemental composition of the underwater archaeological metallic pieces as well as presence of decayed compounds. For that purpose, nondestructive analytical techniques were employed, namely, Raman spectroscopy and X-ray fluorescence spectroscopy to check the conservation state and to identify the raw materials that were used in the manufacture of those pieces. The analyses concluded that both artifacts were manufactured in cast iron, although the composition of the raw materials was not the same. The main decaying compound was lepidocrocite (γ-FeO (OH)), a highly reactive iron phase that increases the corrosion rate of the artifacts, though other iron oxy-hydroxides were also detected in minor amounts. This compound together with the also found akaganeite (β-Fe2(OH)3Cl) were probably responsible of the continuous oxidation of the metallic pieces, which are in a poor conservation state. Therefore, the applied treatment was not the most adequate for the conservation of these pieces. The 15 works included in this special issue are excellent examples of the innovative applications of Raman spectroscopy from prehistoric samples to present-day artefacts. Some papers highlight the possibilities of field nondestructive analysis, to obtain even quantitative information on the amount of compounds in the surfaces of Cultural Heritage assets. Others incorporate also chemical modeling and/or chemometric analysis to explain and interpret the presence of unexpected materials together with the original ones. Some of the papers incorporate new Raman spectra of compounds in downloaded databases. Other papers deal with the use of complementary nondestructive as well as micro-destructive instrumental techniques to support the Raman information. But all of them have in common Raman spectroscopy as the core of the manuscripts included in this special issue. The contribution of the people attending the RAA2017 Congress to a collaborative research among scientists from different fields (restorers, spectroscopists, chemists, geologists, biologist, environmentalists, etc.) has been clearly shown and we hope to increase such cooperation in the works to be presented in the forthcoming RAA2019 in Potsdam, Germany. We are extremely grateful to the participants and institutions that assisted in making the conference possible. In particular, we would like to thank the University of Évora in Portugal. The organizing committee expresses its thanks to the following institutions and companies who also supported the conference: Hercules Laboratory, Alentejo2020, Portugal2020, European Regional Development Fund, Hitech, WITec, BWTEK, and Dias de Sousa. J.M. Madariaga acknowledges the support of the UFI Global Change and Heritage (ref. UPV/EHU, UFI11/26) from the University of the Basque Country.