Search

The spectral variability of reflectance properties of pictorial layers can be caused by chemical parameters, such as the nature of pigments and binders, or by morphological parameters, such as the pigment concentration, the thickness of the layer or the size distribution of pigments. This work proposes a study of the influence of morphological parameters on reflectance spectra using both hyperspectral measurements on physical samples and reflectance spectra calculated using an optical model based on the four-flux model. The spectral variability is evaluated by comparing the spectra to a reference spectrum, thanks to the Kullback–Leibler Pseudo-Divergence.

Yellow to red feathers in ornithological and ethnographical museum collections are known to be sensitive to light exposure. However, only a few studies on the interaction of light with the color of feathers have been published. This work provides fresh insight into the sensitivity of feathers for better conservation and exhibit of ethnographic collections, with a specific focus on carotenoids, psittacofulvins and porphyrins-based feathers. A first approach was to evaluate the use of Fiber Optics Reflectance Spectroscopy (FORS) in the visible range to characterize the chemical nature of the bio-pigments. It was implemented on 19 ornithological specimens preserved in the Museum national d’histoire naturelle (MNHN) in Paris, and on assigned feathers from more than 40 objects from the collections of the musee du quai Branly - Jacques Chirac (mqB-JC). Then, microfading tests were carried out on molting feathers or picked up on dead birds. Two different setups were used in order to determine a reproducible protocol for these 3D materials. Several variables were considered including the positioning of the feather and the duration of measurements. In addition to the visible range, the near-infrared (NIR) region was also investigated to determine if the color fading is concomitant to any structural modification. Then the developed protocol was applied on some loose lumps of feathers from objects of the mqB-JC collections. The data were compared against each other and to the Blue Wool Standards fading rates.

International audience; During initial conservator's investigation of paneled wooden vault of the church in Le Quillio, 17th century paintings, hidden behind painted canvas nailed on the panels were discovered. A closer look on these paintings and some exposure windows showed the presence of an older paint layer. This layer corresponds to the first vault decoration, dating back all the way at the end of 15th century, existence of which was attested by written sources. However, the conservation state of the original pictorial layer was unknown, so it was not possible to make a decision to uncover it without further information acquired, preferably, non-invasively. Examination of different layers was performed by extracting, processing and cross-referencing the information from the complementary data obtained from MA-XRF scanning as well as multi-and hyperspectral imaging. This allowed to characterize the mineral composition of different layers and to attribute them respectively to the end of 15th century original painting, the 17th century overpaint and to the later 19th century interventions corresponding rather to a large retouching or restoration. Our study also showed the particularities of the mounting of panels, as some parts on the first decoration's paint seem to be partially discontinued between successive panels and may attest to the exchange of panels between different parts of the vault. This pilot study was conducted on an only small part of the large vault (0.4% of the 32 m 2). Further, more extensive investigations will be required to better understand this complex object and its complex history.

International audience; Over the last quarter of the 20th century, chromogenic colour process was the predominant photographic process on the market and has been used by the professionals, artists and the public at large. Today it represents a precious part of our photographic heritage. Unfortunately, some colour prints are very fragile and discolour during display or storage faster than others; it is well known that some brands and production times correspond to different thermal and light-ageing behaviour for prints. Being able to identify a colour process, a brand and possibly a period of printing may inform us not only about the provenance of the artefact but also its sensitivity to the environment. With this aim, this paper investigates the possibility of using near-infrared spectral signatures of chromogenic colour photographs combined with statistical analysis to identify photographic paper manufacturers. To carry out the study, a batch of a few hundreds of amateurs colour photographs printed on Agfa, Fuji or Kodak paper produced from the 1960's to the 2000's was gathered from private and public collections. Non-invasive fiber optical reflectance spectroscopy (FORS) in the range of 1000–2500 nm was performed on those images in order to implement a reflectance spectra database. Then statistical data treatments were tested to evaluate the possibility of retrieving a print manufacturer with the lowest error possible. The use of linear discriminant analysis (LDA) combined with principal component analysis (PCA) allows reaching from 82 to 96% of correct identification depending on the paper type. This shows a possibility of implementing automated attribution procedures for chromogenic colour photographic prints.

Photographic negative images on paper are certainly the earliest types of photographs ever produced, giving access to multiple photographic prints using a unique matrix. As early as 1841 in France, many variants of the paper negatives processes have been developed by skilled chemists and photographers covering a wide spectrum of practical and aesthetic concerns. Nowadays, due to their inherent historical and esthetical values, those negatives are getting an increasing interest from the art and museum community. However, their materiality has been poorly studied despite a large variety of processes involving many different chemicals and organic coatings; limiting our possibilities of identification, attribution or even preservation. The aim of this work was to develop a methodology to better assess the way those images were produced based on physical and chemical characteristics. A non-invasive approach combining optical, vibrational and X-ray spectroscopies was implemented on a collection of 138 historical images between 1841 and 1856 from French collections. A survey was carried out on these negatives to provide a series of relevant physical data (dimensions, thickness, weight, etc). Non-invasive reflectance FTIR showed great potentiality in revealing the presence of organic sizing (gelatin) or impregnation (wax). Multivariate analysis was applied on XRF data to help clustering negatives that have similar elemental composition and highlight relationships between makers. Finally, the morphological, physical and chemical results were all combined to answer historical questioning on the paper negative process.

The paintings by edouard Manet in The Courtauld Gallery Dejeuner sur l'herbe (1863-68), Marguerite de Conflans en Toilette de Bal (1870-1880), Banks of the Seine at Argenteuil (1874), and A Bar at the Folies-Bergere (1882) were investigated for the first time using a range of non-invasive in situ analyses. The aims of the study were to investigate the painting techniques and materials used for this group of works and to critically evaluate the technical evidence derived from the integrated use of imaging techniques and portable spectroscopic methods in this context. The paintings were investigated by means of macro X-ray fluorescence (MA-XRF), reflection spectral imaging, portable UV-Vis-NIR spectroscopy, portable Raman spectroscopy, and reflection FTIR. MA-XRF and reflection spectral imaging allowed visualising elements in the compositions that were not visible using traditional methods of technical study. For example, MA-XRF analysis of Dejeuner sur l'herbe revealed elements of the development of the composition that provided new evidence to consider its relationship to other versions of the composition. The study also highlighted questions about the interpretation of elemental distribution maps and spectral images that did not correspond to the reworking visible in X-radiographs. For example, in A Bar at the Folies-Bergere Manet made numerous changes during painting, which were not clearly visualised with any of the techniques used. The research has wider implications for the study of Impressionist paintings, as the results will support technical studies of works by other artists of the period who used similar materials and painting methods.

The Leman Album is Europe's earliest group of dated designs for woven silk known to survive. For the past two years it has been studied by scientists, conservators, curators, archivists, makers and historians, using a truly interdisciplinary approach under the auspices of the V&A Research Institute (VARI), generously funded by the Andrew W. Mellon Foundation. This paper focuses on the scientific analysis of the designs, which pursued multiple targets, such as investigating the nature of degradation processes, e.g., fading or darkening of the pigments and dyes; uncovering the underdrawing and any amendments; and identifying the artists' materials, in order to inform conservation, storage and handling procedures. The analyses took place at the Victoria and Albert Museum (V&A) and involved V&A staff as well as researchers from other institutions through the EU-funded consortium IPERION-CH. Selected results will be made available to the wider public through a web-based platform: the enhanced facsimile.

In this work, we are focusing on the development of metrological features for the analysis of hyperspectral measurements. We propose to define the entropy of a hyperspectral dataset or image. As the processing cannot be developed in the spectral acquisition space, the proposed entropy is processed using histograms of spectral differences. The metrological properties are induced by an adapted spectral difference, the Kullback-Leibler pseudo-divergence. Results are proposed for different color scales and show how the entropy can be used to select spectral references for hyperspectral image analysis or processing.

International audience

International audience

Luminescence multispectral imaging is a developing and promising technique in the fields of conservation science and cultural heritage studies. In this article, we present a new methodology for recording the spatially resolved luminescence properties of objects. This methodology relies on the development of a lab-made multispectral camera setup optimized to collect low-yield luminescence images. In addition to a classic data preprocessing procedure to reduce noise on the data, we present an innovative method, based on a neural network algorithm, that allows us to obtain radiometrically calibrated luminescence spectra with increased spectral resolution from the low-spectral resolution acquisitions. After preliminary corrections, a neural network is trained using the 15-band multispectral luminescence acquisitions and corresponding spot spectroscopy luminescence data. This neural network is then used to retrieve a megapixel multispectral cube between 460 and 710 nm with a 5 nm resolution from a low-spectral-resolution multispectral acquisition. The resulting data are independent from the detection chain of the imaging system (filter transmittance, spectral sensitivity of the lens and optics, etc.). As a result, the image cube provides radiometrically calibrated emission spectra with increased spectral resolution. For each pixel, we can thus retrieve a spectrum comparable to those obtained with conventional luminescence spectroscopy. We apply this method to a panel of lake pigment paints and discuss the pertinence and perspectives of this new approach.

Vibrational spectroscopy (Raman and FTIR) has played an important role in identifying pigments, substrata and deterioration products in rock art studies worldwide: in the laboratory and on-site. However, the detection of organic binders and carrying agents has so far been scarce and the quality of many spectra recorded on-site inadequate. In this study, possible pigments (charcoal, ochre, raptor faeces, thermally treated ostrich egg shell, etc.), binders (fat, egg, blood) and carrying agents (saliva, gall, egg, water) were selected based on artistic considerations and analysed with FTIR and Raman (514.6 and 785 nm excitation, both available in mobile instruments) spectroscopy in order to determine usable marker bands for each ingredient. The resultant marker bands were then used to analyse five ten year old San replica paints. It was found that FTIR spectroscopy is very efficient to identify organic compounds as there is no fluorescence but the broadness of the bands inhibits the exact assignment of many ingredients. A high fluorescence background experienced for many natural products prevented the recording of Raman spectra for all ingredients, in many instances though the sharp peaks usually associated with Raman spectra make identification easier than with FTIR spectroscopy. Most of the ingredients in the paints could be identified, but it is clear that better results are obtained when more that one technique is used.

International audience; The development of the laser technology for stone cleaning purposes has been undermined in the recent years because the technique occasionally conveys a yellow hue to the cleaned surfaces. Several components of black gypsum crusts have been suggested to be potentially responsible for the phenomenon, more specifically some granular fractions of their components. The contribution of these particles to laser yellowing is studied here on powdery deposits collected at Saint-Denis Basilica (France). The powder is characterized by optical and scanning electron microscopy, X-ray diffraction and Raman spectroscopy before and after the laser irradiation. Irradiation of the samples results in a significant yellowing, along with particle fragmentation, while no mineralogical change is evidenced by powder diffrac-tometry. Hematite and goethite micro-Raman signatures ob-V. Verg es-Belmin Centre de recherche en conservation : Laboratoire de recherche des monuments historiques, (CRC:LRMH), CNRS/USR 3224-Sorbonne Universités-29, rue de Paris-77420 Champs sur Marne-FRANCE Tel.: +33 1 60 37 77 80/95 Fax: + 33 1 60 37 77 99 tained before laser irradiation are replaced after irradiation by a strong fluorescence in addition to weak goethite signatures in Raman spectra. The powdery deposit is further analyzed by diffuse reflectance spectrometry after a specific preparation which increases both its cohesion and its content in new phases potentially formed. The powdery material shows a phase with a spectral signature similar to goethite which subsides after laser irradiation. The yellow appearance caused by laser irradiation seems to be due, in part, to a selective elimination of the black components (soots) from the powdery deposit. This elimination reveals a colour which could be related to a pre-existing phase containing iron, possibly goethite or another phase containing iron and having the same spectral signature in the visible range.

On-site Raman spectroscopy is used to investigate four mosque lamps and a bottle dating back to the 13–14th centuries (Syria and/or Egypt, Mamluk period) from the Department of Islamic Art, Musee du Louvre, Paris. The pigments and the enamelled glass matrix have been identified in order to discuss their technology. A comparison is made with one masterpiece from the 19th century made by Brocard. The results obtained for blue (lapis lazuli or Co-coloured glass), yellow (Naples yellows or zinc/chrome yellow), green (mixture of the aforementioned blue and yellow pigments or lead chromate for restorated foot) red (hematite), white (cassiterite or arsenate) and pink (hematite and cassiterite) pigments are presented. The consistency of the pigments period of use/technology with the datation proposed for each artefact is evaluated. For one lamp, the detection of quartz in some places of the hematite-rich glassy layer supporting the gold is related to the search of a physical bonding with the artefact body. Carbon residues in such enamel are consistent with the use of an organic medium to place the enamel powder before the firing. Copyright © 2012 John Wiley & Sons, Ltd.

The comparison of the low wavenumber of polarized Raman spectra (50–300 cm–1) from Bombyx mori (fresh cocoons fibres, hand-stretched ‘Crins de Florence’ strings from the gland content, dried gland, regenerated silk films) and Nephila madagascarensis silks reveals the high polarisation of fibre modes and the absence of polarisation for dried gland and regenerated silk films. This is consistent with X-ray diffraction measurements. The orientation of the fibroin/spidroin chains is due to the stretching during production, as for advanced synthetic fibres. The bandwidth of the ‘ordered chains’ signature is almost the same for the different fibres. However, the degree of polarisation seems to be higher in the case of spider fibre. The huge bandwidth of low wavenumber components of regenerated films indicates high disorder. Measurements along the fibre point out conformation changes with a periodicity (~20 mm) related to the silkworm head motion during the fabrication of the cocoon. Copyright © 2012 John Wiley & Sons, Ltd.

Thousands of glass trade beads were excavated over a 75 year period on Mapungubwe Hill and at K2, two archaeological sites in the Limpopo valley, South Africa. An assemblage of 175 beads that appeared to be different in shape, size and colour (red, yellow, green, blue, white, black, pink, plum) was studied with Raman scattering. At least seven different chromophores or pigments (lazurite, lead tin yellow type II, Ca/Pb arsenate, chromate, calcium antimonate, Fe–S “amber” and a spinel) have been identified. Pigment identification allows the recognition of specific productions and indicated that many of the pigments colouring the beads excavated on Mapungubwe hill were manufactured after the 13th century, confirming the presence of modern beads in the archaeological record, some dating from the 19th century. This date is in agreement with the last occupation date for the hill suggested by the earlier excavators and raises questions about the revision of this date to 1290 AD by archaeologists in the 1970’s.

The assignment of Asian bronzes and brasses is difficult because the copies of ancient pieces have been made a long time ago in different countries. A selection of 30 Japanese and/or Chinese bronzes/brasses are studied on-site in the storage and exhibition rooms of the Cernuschi Museum, in Paris, using portable Raman and X-ray fluorescence spectrometers. Attempts are made to identify specific Raman signatures of the patina to detect similarity in a nondestructive procedure. X-ray fluorescence measurements allow an identification of two brass artefacts and different types of bronze, including lead-rich and mixed lead–tin–zinc-rich compositions. The following phases are identified: CuO, Cu2O, Cu2S, tin oxides, ZnO, Cu3(OH)4SO4, Cu4(OH)6SO4, 3PbOPbSO4 H2O, PbSO4/PbO, Pb(AsO4)3Cl, HgO/HgS. Relics of the mould, retained as concretions at the artefact surface are also identified: TiO2, SiO2, and different soda(−lime) glasses. Patina with very similar colour and habit may have different Raman signatures. This study offers tools to classify the artefacts in a noninvasive way. Copyright © 2011 John Wiley & Sons, Ltd.

About 200 coloured glass beads (red, yellow, green, blue, white, black, pink, plum) excavated on Mapungubwe hill and at K2, archaeological sites in the Limpopo valley South Africa, were studied with Raman scattering. This is also the most southern site in Africa where evidence for glass reworking has been found. The glass matrix of the beads was classified according to its Raman signature into three main subgroups and corroded glass could also be identified. At least seven different chromophores or pigments (lazurite, lead tin yellow type II, Ca/Pb arsenate, chromate, calcium antimonate, Fe–S ‘amber’ and a spinel) were identified. Many of the pigments were manufactured after the 13th century, confirming the presence of modern beads in the archaeological record. This calls for further research to find a way to reconcile the carbon dating of the hill, which currently gives the last occupation date on the hill as 1290 AD with the physical evidence of the modern beads excavated on the hill. The results are discussed in terms of the glass production origin of the beads (Europe, Mediterranean area, India, China). Copyright © 2011 John Wiley & Sons, Ltd.

Six French miniature portraits on ivory and paper dating back to the 18th to 19th centuries have been studied by Raman spectroscopy. The miniatures have been investigated through their protective glass cover using different operating conditions: various laser wavelengths (1064, 785 and 532 nm), spectrometers (fixed and mobile) and objectives (10× to 200×). The results obtained for black (carbon), grey (carbon + chalk), red (lead oxide, vermillion), blue (Prussian or phthalocyanine blue), green (emerald or Paris green) and white (lead white) pigments are presented. The consistency of the period of use of the pigments with the proposed dating for each artwork is evaluated. Attention is paid to the analysis of the ivory substrates (assigned to Asian elephant ivory) and to the protective glass. Copyright c � 2011 John Wiley & Sons, Ltd.

Ancient mosaic tesserae are a range of materials of very varied and complex nature, including pottery, stone and glass. Raman spectroscopy is a powerful tool for the analysis of all these kinds of materials. In the particular case of glasses, this technique can be used both for a study of surface weathering and for the characterization of bulk structure, but it has not yet been extensively used for the characterization of mosaic glass tesserae. We carried out Raman analyses on a set of Roman and Late Antiquity period mosaic glass samples, which allowed a good characterization of both the glass matrix and the crystalline inclusions. All the samples show the typical Raman signatures of soda-lime-silicate glasses. Several crystalline phases were also identified, being relics of raw materials used during the glass manufacturing process, such as quartz and feldspars, or linked to the glass color/opacification, such as bindheimite and cuprite. The analyses also led to the identification in some blue, turquoise and green tesserae of calcium antimonate, whose Raman signature has only recently been recognized in the scientific literature on mosaic glasses. Some emphasis is given to the analysis of red lead-containing tesserae, colored with Cu+ ions or even Cu0 (or Au0) metal nanoparticles. Samples with peculiar compositions, as well as “modern” (and restoration) samples, could quite easily be distinguished from the ancient ones by their Raman spectra.

A selection of red-flashed and red-coated medieval potash lime and 19th century soda lime stained glass pieces were studied in the laboratory using different instruments and laser wavelengths, both at lower and higher than the surface plasmon resonance (SPR) of copper. The Raman signatures of the transparent glass matrix and the red glass layers are discussed and compared with those recorded on model glasses containing a dispersion of Cu0 nanoparticles. Evidence is given that the conformation of the silica network in the vicinity of the metal nanoparticles differs from that of the glass matrix. Hematite and carbon traces are evidenced in the Cu0-rich layers, which is consistent with the use of a combination of a reducing atmosphere and redox couples to control the growth of metallic copper particles. Copyright © 2009 John Wiley & Sons, Ltd.

Functional titania coatings on glass microballoons (GMBs) and cenospheres have a broad range of potential environmental applications, primarily in purification of drinking water and treatment of industrial wastewater. The heterogeneous photocatalytic capabilities of titania films and particles have been extensively examined in the literature as effective alternatives to current technologies. Although the chemistry of titania films for photocatalysis has been studied, titania-coated GMBs have not yet been extensively considered and the materials science aspects of the titania-GMB and titania-cenosphere systems have not been addressed. We have examined the microstructure, morphology, and mechanical properties of titania coatings on both cenospheres and commercially produced GMBs. Scanning electron microscopy was used to examine coating coverage and defects. Energy dispersive X-ray spectroscopy and Raman spectroscopy were used for element and phase identification, respectively. Hardness and modulus measurements of the titania coatings and the GMB and cenosphere materials were done by nanoindentation. Additionally, the photocatalytic activity of the titania-coated GMB system was tested on Procion Red dye using two different types of mixing, a magnetic stirrer and an aeration bubbler apparatus. The titania coatings showed good coverage and retention except in the case of magnetic stirring, where significant coating loss was observed.

A selection of 23 rare glass objects, mostly enameled, of various provenance and age, from the 5th century BC to the 19th century AD including the Western and Islamic Middle Ages but with a focus on 16th–18th century Venetian and French ‘facon de Venise’ artefacts, have been studied on-site at the Sevres museum or at the laboratory. The Raman signatures of the transparent or opacified glass matrix and of enameled decorations are discussed and compared to those previously recorded on ceramics and stained glasses. The Raman parameters allow discrimination between 2 groups (with some variations) of glass bodies, belonging to mixed Ca–Na and Ca-containing Na-rich silicates, with some exceptions. Most enamels are instead lead-based glasses, but we also found enamels having a composition close to that of the glass body. Most of the pigment signatures are similar to those recorded on ceramic glazes, which proves the link between the two technologies. A particular emphasis was given to the identification of white opacification techniques. Very specific signatures could question the authenticity of some artefacts, and at least in two cases, arguments have been found to identify a fake or embellished artefact. Copyright © 2008 John Wiley & Sons, Ltd.

Ancient to modern glass used for stained glass windows, representative of the main composition types (Ca-rich, Na- and Ca-rich and K-rich) were treated with boiling concentrated H2SO4 (338 °C) and/or in molten NaOH (320 °C, with or without water addition). Subsequent annealings in air or in molten KNO3 (334 °C) were also performed. Microstructures and corrosion habits were analysed by optical microscopy. Modifications of the silicate network and corrosion products were studied by Raman and IR spectroscopies, dilatometry, microhardness, and TGA. Only K-rich silicate glasses are drastically modified by concentrated sulphuric acid. The alkali leaching takes place in three steps: i) K+–H+ (not H3O+) exchange, partly reversible, inducing glass contraction and cracking, ii) a complete, irreversible H+/K+ ion exchange, iii) a (small) hydration of the formed porous, multicracked lixiviated silica-rich solid crust. The thermal treatment of the crust induces a water loss (~ 0.6 wt.% at

Red and yellow natural (roasted) pigments have been used for millennia. This paper reports on the Raman characterization of about 50 pigments of the Okhra ‘Matheriautheque’ collection from different origins: violet, red, orange, apricot, yellow ochre, Umbrian and Siena earths from France (Puisaye, Ardennes, Vaucluse), Italy (Siena, Sardinia, Venice, etc.), Germany, Sweden, Cyprus and India (Madras). Comparison is made with ‘Bengara’ Japanese pigments. Goethite is the chromophore of yellow pigments, and haematite (sometimes with magnetite) that of red ochre and earths. Umbrian pigment (‘Terre d'Ombre’) contains additional manganese oxides. Color shades are related to the content of secondary phases: iron oxides (haematite, goethite, magnetite), quartz, gypsum and anhydrite, calcite, dolomite, alums (jarosites), and to the Fe/M substitution in the haematite structure. The set of secondary phases can be used to confirm a provenance signature (Supplementary Material available). Copyright © 2008 John Wiley & Sons, Ltd.

The experimental procedures and most important conclusions of the first on-site Raman study of the stained glass windows in the upper chapel of the Sainte-Chapelle, Paris are discussed. Some of the windows suffered damage during/after the French Revolution and were partly restored in the 19th century. Measurements were performed with a new portable Raman instrument on colourless, blue and green stained glasses. We illustrate how the Raman signature of the glass makes it possible to distinguish between medieval K/Ca or 19th century restored Na/Ca-based silicates and to determine their weathering degree. This is achieved by means of the extraction of vibrational parameters and then processing them, using chemometric approaches, principal components and cluster variation analyses with varying degrees of complexity. The Raman scattering intensity of weathered glasses is used to determine their relative age. The results differ from those presented in the Corpus Vitrearum Medii Aevi (obtained by visual inspection) and therefore demonstrate the need for updating these reports with modern methods such as in situ Raman spectroscopy.

The aim of this investigation was to establish a Raman procedure for on-site identification of stained-glass windows and their deterioration level. This study of a representative series of stained-glass pieces dating from the thirteenth to seventeenth, nineteenth and twenty-first centuries made it possible to classify the glasses into four types: Type 1 (rare, Ca-rich Na silicate), Type 2 (rare, Na-rich, Ca silicate), Type 3 (more frequent, KCa silicate) and a variant of Type 3, referred to as 3bis. Ion exchange of the alkali cations was made in boiling sulfuric acid and related to structural and compositional changes, which were analyzed with IR and Raman spectroscopy, optical microscopy and EDS. Measurements of the film thickness showed a great discrepancy between the ion-exchange rates of KCa (Type 3: 60 µm/h) and NaCa (Type 2: 15 µm/h) silicates with that of Type 1 Ca-silicate (0.05 µm/h). IR and Raman spectra provided proof that the nanostructure of the glass was modified, chiefly by the downward (Raman) and upward (IR) shift of the position of the main SiO stretching peak and a decrease in the intensity of the Boson peak, as well as the Raman 580 cm−1 SiO bending peak. The decrease in intensity of the narrow ∼950 cm−1 Raman peak, assigned to (earth)alkali nano crystallites in the glassy network, is correlated with the K/Na ion loss and glass weathering. Consequently Type 3bis samples are in fact glasses, in which the surface is depleted of K/Na ions. Optical micrographs support the macroscopic compositional and structural heterogeneity of the ancient glasses. This work demonstrates the potential of Raman scattering for in situ measurements of the degree of weathering of ancient stained glasses. Copyright © 2006 John Wiley & Sons, Ltd.

As an optical method, Raman microspectroscopy offers a great advantage over most other techniques in that it can be performed without any contact with the studied artefact, both at the laboratory using high-resolution, large spectral window instruments, and on site using medium resolution, portable instruments. Six years of experience on various materials has enabled us to propose a tentative guide to identify different types of glassy silicates and to classify them as a function of their composition. In previous papers, different families were recognised empirically using a limited set of samples and the relationship between the peak area ratio (A500/A1000) of the SiO bending (∼500 cm−1) and stretching (i.e. ∼1000 cm−1) envelopes, and the different components of the latter peak were established for porcelain glazes. In this paper, we extend the procedure to a larger set of samples (30 representative samples with known compositions selected from hundreds of spectra). Additional relationships between the origin of the material, the Raman parameters and the glass composition (fluxing oxide content, ionicity ratio, etc.) are discussed. Finally we propose different procedures with different degrees of complexity (from two to multivariate treatment) to identify the glass composition. The method is illustrated with an example, namely, the differentiation between Iznik and Kutahya productions. Copyright © 2006 John Wiley & Sons, Ltd.