Sol-gel deposition of silica films has been successfully used to protect modern and historical soda-lime glasses. It was previously found that the coatings of silica films reduce the leaching phenomena on soda-lime glasses treated in climatic chamber. In previous works an acid solution of tetraethyl orthosilicate (TEOS) in ethyl alcohol and deionised water, with small amounts of hydrochloric acid as catalyst, was used as precursor sol-gel solution to coat soda-lime glasses. It has been found that the same precursor solution did not produce good results on lead silicate glasses, because the lead ions, contained in the glasses, are easily leached in acid environment. To avoid these phenomena, several tests were carried out. It was observed that the hydrolysis reaction took place even with 1/100 of the amount of acid previously used. However no hydrolysis has been detected with lower amounts of acid after several weeks and no film has been formed on soda-lime glasses without acid catalyst. On glasses with large amounts of lead (66 or 45.3 wt-%lead oxide) coating films of good quality have been obtained by long dipping times (24 hours), temperatures over 20°C, without acid catalyst. Likewise films of good quality have been obtained on soda-lime glasses by the addition of lead salts instead of hydrochloric acid. For this reason a role of the lead ion as catalyst is proposed. The films have been characterised by optical microscopy, AFM, XPS, and SIMS.

L’accélérateur de particules AGLAE du C2RMF a eu un rôle majeur dans l’émergence du projet européen Eu-ARTECH. Après un bref rappel des propriétés des méthodes d’analyse par faisceau d’ions, la spécificité d’AGLAE est soulignée. Les activités impliquant l’accès transnational (TNA) à cette installation sont alors présentées, révélant la grande diversité des thèmes scientifiques abordés. Des exemples d’études sont donnés, qui illustrent la contribution d’AGLAE à la connaissance des œuvres d’art et d’archéologie et tout particulièrement celles en céramique et en métal.

For three decades, ion beam analysis (IBA) in external mode was considered as the best choice for the characterisation of cultural heritage materials, as it combines excellent analytical performance and non-invasive character. However, in recent years, other analytical techniques arose as serious competitors, such as those based on synchrotron radiation (X-ray absorption, fluorescence or diffraction) or those using portable instruments (XRF, micro-Raman). It is shown that nevertheless IBA remains unmatched thanks to two unique features, namely the analysis of light elements and the high-resolution 3D chemical imaging.

The comparative leaching behavior of a soda-lime glass and of a simulated nuclear glass has been investigated through H and Na depth profiles obtained with resonant nuclear reaction analysis. It is shown that glass surface hydration involves both H+/Na+ ion exchange and permeation of molecular water, the first process being dominant at low temperature and solution ionic strength and the second for high values of these two parameters. The shape of the H profile as well as the response of H retention to thermal treatment indicate that on the nuclear glass, most of the molecular water is more weakly bonded than on the soda-lime glass.

The combination of particle-induced X-ray emission (PIXE) and Rutherford backscattering (RBS) is particularly fruitful for the study of cultural heritage objects. Several set-ups have been developed at the AGLAE facility of the Louvre Laboratory to implement these techniques with an external beam. Successively have been tested the simultaneous use of PIXE and RBS with a single beam of protons, the sequential application of PIXE with protons and RBS with 4He2+ ions and finally the simultaneous implementation of PIXE and RBS with high-energy 4He2+ ions. Several examples illustrate the benefits of these combinations of techniques.

The application of IBA to cultural heritage mostly relies on the use of PIXE because of its high sensitivity and its ease of implementation at atmospheric pressure. The need for depth information not easily available with this technique has conducted to associate RBS also in external beam mode. We have progressively developed a set-up that permits such a combination of techniques either simultaneously or sequentially. The set-up is currently further improved to permit NRA measurement (depth profiles of light elements) in addition to PIXE and RBS. The coupling of all these techniques provides a wealth of information on cultural heritage objects, not easily attainable with any other single method.

To hinder the phenomenon of weathering of ancient stained glass, the present work proposes the application of sol–gel coatings to the glass surface. Previous investigations [1], [2], [3], [4], [5], [6], in fact, show that sol–gel silica coatings do not change the appearance of artistic glasses when deposited on their surface. Moreover, the film thickness is so small (around 200nm) and its composition and structure so compatible with that of the original glass, that the characteristics of the coating and original glass are not distinguishable. In this work, several recipes used to produce sol–gel coatings have been tested in order to understand their behavior when adopted for covering ancient weathered glass. The coatings are made of sol–gel silica prepared with two different catalysts: H+, Pb2+ and without catalyst. All the investigated samples show a good adhesion of the coating to the glasses used to simulate the behavior of ancient artefacts. The sol–gel silica coatings have been studied before and after accelerated ageing to test the resistance of the protective coatings to weathering. Another important index to test of the efficiency of the sol–gel coatings for the protection of an ancient glass is the lead ion mobility. In ancient stained glass, in fact, this element is present in the metallic lead network, in the grisaille paintings and constitutes a main component of many glass tesserae. The action of water on this highly mobile ion involves the degradation of the glass itself and the release of the ion in the rain solution. Ageing tests show the efficiency of H+ and Pb2+ catalyzed coatings and the inefficiency of the non-catalyzed sol–gel layers.

Ion beam induced luminescence (IBIL) has been exploited for the first time in the analysis of inorganic painting pigments. The elemental constituents of the different compounds have been determined by particle induced X-ray emission (PIXE). The acquisition time of each spectrum ranges from 100ms to a few seconds, depending on the luminescence intensity. The luminescence features are fingerprints of the different compounds, thus identifying the provenience of pigments of the same nominal composition. Organic varnish layers do not affect the IBIL features, allowing the identification of pigments, like lapis-lazuli, whose identification with PIXE is hindered by the varnish. IBIL proved to be a technique complementary to PIXE in the archeometry and cultural heritage analysis fields.

The impact of energetic ions on solid samples gives rise to the emission of visible light owing to the electronic excitation of intrinsic defects or extrinsic impurities. The intensity and position of the emission features provide information on the nature of the luminescence centers and on their chemical environments. This makes ion beam induced luminescence (IBIL) a useful complement to other ion beam analyses, like PIXE, in the cultural heritage field in characterizing the composition and the provenience of art objects. In the present paper, IBIL measurements have been performed on inorganic pigments for underlying the complementary role played by IBIL in the analysis of artistic works. Some blue and red pigment has been presented as case study.

Many lead silicate historical glasses suffer degradation phenomena often observed as color changes and iridescence caused by lead ions leaching from the outer layers of the glass. In order to repair and to prevent these phenomena, glasses with large amounts of lead (6.7 and 14.3at.% of lead) have been coated with silica films at neutral pH by dipping them in a precursor solution of TEOS (tetraethyl orthosilicate), ethyl alcohol and deionized water without any other acid or basic catalyst. Experiments with long dipping times (24h) and temperatures around 20°C have been performed to evaluate the role of lead ions of the glass as a catalyst. Silica films of very good quality and optical transparency have been also obtained on lead-free, soda-lime glasses by adding catalytic amounts of Pb(NO3)2 instead of HCl to the precursor solution. The films have been characterized by optical microscopy, AFM (Atomic Force Microscopy), XPS (X-ray Photoelectron Spectroscopy) and SIMS (Secondary Ion Mass Spectrometry).

Two paintings by Murillo from the Louvre Museum entitled Agony in the garden and Penitent St. Peter kneeling before Christ and the column were analysed by PIXE to identify the nature of their unusual dark mineral backing. Considered until now as black marble, this support turns out to be obsidian, with an almost identical elemental composition for the two works. This composition was compared to that of six Mesoamerican unpainted obsidians labelled “smoking mirrors” with comparable size and shape from the Paris Museum National d’Histoire Naturelle and Musée de l’Homme and to that of geological samples from five Mexican sources. The trace element contents of Murillo’s obsidians, in particular those of Mn, Zr, Sr, Y, Rb, Zr, Nb and Zn appeared to be very similar to that of four smoking mirrors and to that of obsidians from the Ucareo-Zinapécuaro source in central Mexico, an important complex of obsidian quarries exploited since pre-Hispanic times. A literature survey showed no such similarity with obsidians from other Mesoamerican sources or even from Mediterranean and surroundings source-areas. This study points out that Murillo, although living in Sevilla, had occasionally employed for his paintings materials shipped from the New World to Spain.

In the occidental world, Gum Arabic has been extensively used as a binder in the preparation of iron-gall ink. Its impact on iron-gall ink corrosion was investigated based on elemental analysis of seventeen original documents. Most of these originals had the same origin, which meant that they had been stored under the same climatic conditions. SEM imaging showed that the ink had formed a thick covering layer on well preserved samples. On corroded samples almost no matter could be observed in the inked areas. Looking at samples containing the same amount of iron, it was observed that the best preserved documents contained the highest quantity of potassium and calcium. If we attribute the potassium and calcium contents to the gum, it appears that gum-rich inks are less corrosive. We also proposed a possible scenario to explain the visual appearance of original manuscripts: the gum may physically protect the paper because the corrosive components of the ink are embedded in this medium. However, this protective effect is limited because the gum interacts chemically with iron. When it is destroyed, the corrosive components of the ink are free to migrate into the paper.

The possibility of non-destructive elemental analysis makes PIXE a very attractive technique in archaeological provenance studies. This technique has been fruitfully implemented on two different facilities to address the issue of obsidian provenance in the Mediterranean and in surrounding regions. At C2RMF, we took advantage of the possibility to analyze large archaeological pieces with the external micro-beam set-up. At CENBG, we used the nuclear microprobe providing a 5 mu m beam diameter in large scans (700 x 700 mu m(2)) to control the homogeneity of elemental distribution. In both cases we dosed the same set of 13 elements: Na, Al, Si, K, Ca, Ti, Mn, Fe, Zn, Ga, Rb, Sr and Zr. While at C2RMF, two Si(Li) detectors were used simultaneously to measure all elements at once with 3 MeV protons, at CENBG where only one detector was available, the light elements Na to Fe were determined with a 1.5 MeV beam, and the heavy ones, including Fe, with a beam energy of 2.7 MeV.In Western Mediterranean, it is possible with PIXE to differentiate all obsidian sources of archaeological significance. Examples are given of obsidian provenances from Neolithic sites of France and from the islands of Corsica and Sardinia. In the Near East, we can differentiate the Cappadocian and Eastern Anatolian obsidian sources used during the early Neolithic. This is illustrated by examples taken from Neolithic sites of the Middle Euphrates Valley (Syria). (c) 2005 Elsevier B.V. All rights reserved.

Elemental composition fingerprinting by PIXE technique is very attractive for obsidian provenance studies as it may proceed in a non-destructive mode, even if a more complete elemental characterization can be obtained by ICP-MS and/or ICP-AES. Only few studies have compared results obtained by both methods for solid rock samples. In this work, elemental compositions were determined by ICP-MS/-AES for international geochemical standards and by ICP-MS/-AES and PIXE for inter-laboratory reference obsidians. In addition 49 obsidian source samples and artefacts were analysed by both ICP-MS/-AES and PIXE. Instrumental work and measurement quality control performed for obsidian chemical characterization, underline that PIXE and ICP-MS/-AES provide reproducible, accurate and comparable measurements. In some volcanic districts the limited number of elements dosed by PIXE is sufficient for the discrimination of the potential raw sources of obsidians. Therefore, PIXE can be an advantageous substitute to ICP-MS/-AES techniques for provenance studies.

This paper reports on the use of PIXE and RBS (resonant and non-resonant RBS) implemented with proton beams to simultaneously analyse light and heavy elements in materials of cultural heritage significance, as exemplified by Russian icons or in lead seals. It is shown that in spite of its poor mass resolution RBS with protons can provide useful information when combined with PIXE. In the case of Russian icons, it is possible to discriminate between an Au–Ag bilayer and an alloy of these metals in the gilds. However, when applied to lead seals RBS with protons encounters a significant limitation due to some deficiency in the available computer programs used for spectrum processing.

Elemental analysis by X-ray fluorescence and particle induced X-ray emission is applied to the study of several Mesoamerican mural samples containing blue pigments. The most characteristic blue pigment is Maya blue, a very stable organo-clay complex original from Maya culture and widely used in murals, pottery and sculptures in a vast region of Mesoamerica during the pre-hispanic time (from VIII century) and during the colonization until 1580. The mural samples come from six different archaeological sites (four pre-hispanic and two from XVI century colonial convents). The correlation between the presence of some elements and the pigment colour is discussed. From the comparative study of the elemental concentration, some conclusions are drawn on the nature of the pigments and the technology used.

For more than 14 years, an IBA facility has been operated in the Louvre for the study of works of art and archaeology. The choice of this equipment derives from the non-destructive character of IBA techniques which has been further strengthened by designing an external beam line permitting the in-air analysis of large or fragile works of art without sampling. Successive improvements have markedly extended the analytical capability of the set-up. The measurements were originally restricted to PIXE–PIGE combination using an external millimetre-sized-beam. By adding a focusing system and an ultra-thin exit window we were able to obtain external beams of protons and alpha particles of respectively 10 and 50 μm in diameter, with low energy straggling. These features have permitted to apply in external beam mode other IBA techniques including RBS, NRA and more recently ERDA. Moreover, elemental maps can be drawn in PIXE and PIGE modes by mechanically scanning the sample under the fixed beam within a lateral range much larger than conventional nuclear microprobes. This facility is used for both short investigations at the request of museum curators and extensive research works in art history and archaeology. Several examples are given to highlight the impact of this tool on cultural heritage.

An open-air furnace has been designed to study in real time the high-temperature surface transformation of materials by Rutherford backscattering spectrometry (RBS) in external beam mode. A device previously designed for the observation of the high-temperature oxidation of galena has been re-designed in order to analyse massive samples and to reach a temperature range up to 700 °C and a better temperature regulation (±2 °C). Experiments are carried out to measure, by RBS using a 3 MeV 4He2+ external micro-beam, the dynamic growth of oxide layers in air on the surface of copper–tin alloys heated at temperatures varying from 250 to 650 °C. The results obtained demonstrate the usefulness of this approach: actually one single measurement series permits to obtain at the same time the composition of oxide layers built on the metal and their growth kinetic laws. In the particular study of bronzes, the growth kinetics obtained by this method show the large influence of tin concentration on the oxidation mechanism. These results are verified by measurements on samples oxidised in an independent closed furnace. The results show also the occurrence of different oxidation mechanisms as a function of temperature. A diffusion mechanism governs the oxidation kinetics only at some temperatures.