JOURNAL OF RAMAN SPECTROSCOPY J. Raman Spectrosc. 2004; 35: 633–639 Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jrs.1216 Micro-PIXE and micro-Raman spectrometry applied to a polychrome wooden altarpiece from the 16th century Sandrine Pag ` es-Camagna * and Thomas Calligaro Centre de Recherche et de Restauration des Mus ´ ees de France, CNRS UMR-171, Palais de Louvre, 6 rue des Pyramides, 75041 Paris cedex 01, France Received 5 September 2003; Accepted 24 March 2004 The aim of this work was to identify the materials used for the colours and the metallic parts of a polychrome wooden altarpiece from the 16th century. Since restoration work on the object was almost completed, no sampling was allowed and the analysis had to be performed in a fully non-destructive manner. Micro-Raman spectrometry was employed to identify the pigments and the ground layer. Taking advantage of micro-cracks, it was possible to show that the ground layer was made of calcite and the blue decorations were made of an azurite layer on top of a carbon black underlayer. Lead white was employed for the white areas and for the carnation. Elemental analysis directly performed in air by PIXE with the external nuclear microprobe of the AGLAE accelerator confirmed the results obtained by Raman spectrometry for the pigments and the preparation. It allows the determination of the composition and thickness of the metallic foils and the nature of the underlying support (bolus), two cases where Raman spectrometry is inefficient. The two techniques proved to be fully complementary for the characterization of the materials used for the decoration of this precious altarpiece. Raman spectrometry permits the identification of all pigments including those containing carbon (soot, azurite or lead white), not accessible to PIXE. Copyright  2004 John Wiley & Sons, Ltd. KEYWORDS: PIXE; micro-Raman; polychromy; metallic foils; pigments; altarpiece INTRODUCTION The scientific investigation of materials constituting works of art can provide important information related to the provenance, dating, authenticity and conservation state of such objects. The aim of this work was to determine the constituents of a precious altarpiece: preparation layers, pigments, inks, metallic parts, etc. The high quality and good preservation state of the polychromy of this wooden sculpture excluded any sampling. Two non-destructive techniques were directly applied to the work to identify the pigments and the metals employed for the decorations: elemental analysis by particle induced x-ray emission (PIXE) performed in air with an external nuclear microprobe and structural characterization by Raman micro-spectrometry. The two techniques proved to be fully complementary for the characterization of the inorganic and metallic materials. The altarpiece representing the Trinity, kept in the Mus´ ee de la Renaissance, Ecouen (France), is a remarkable and particularly well-preserved example of a Dutch work of art L Correspondence to: Sandrine Pag` es-Camagna, Centre de Recherche et de Restauration des Mus´ ees de France, CNRS UMR-171, Palais de Louvre, 6 rue des Pyramides, 75041 Paris cedex 01, France. E-mail: sandrine.pages@culture.fr dated from the first part of the 16th century (Fig. 1). This specific design, with two articulated panels attached to the central case by two hinges, is called triptych. Its small size (30 cm high, 20 cm wide, 15 cm deep) suggests that it was used as a portable triptych or portable oratory. An inscription in the back of this anonymous masterpiece indicates that it was offered to Sister Perrette Dobray in 1542 by her brothers and sisters. EXPERIMENTAL Raman set-up The Raman spectra were collected with an Infinity microspec- trometer (Jobin-Yvon), equipped with a charge-coupled device (CCD) detector, using laser excitation at 532 and 633 nm. The horizontal external output was employed to focus the laser beam directly on the altarpiece, placed in front of the spectrometer. A motorized sample holder with stepper motors was used for the positioning of the triptych. A laser power in the 3–10 mW range was used to avoid any possible damage to the sample, with integration times of 20–100 s and 5–30 accumulations. The 50ð long distance objective provided a 5 μm diameter laser beam, allowing Copyright  2004 John Wiley & Sons, Ltd.