Anal. Chem. 2005, 77, 1261-1267
Pigment Identification by Spectroscopic Means: Evidence Consistent with the Attribution of the Painting Young Woman Seated at a Virginal to Vermeer Lucia Burgio,†,‡ Robin J. H. Clark,*,† Libby Sheldon,§ and Gregory D. Smith†,|
Christopher Ingold Laboratories, University College London, 20 Gordon Street, London WC1H 0AJ U.K., and Department of History of Art, University College London, Gower Street, London WC1E 6BT U.K.
Technical examination of the painting Young Woman Seated at a Virginal by cross section and polarized light microscopy, chemical tests, surface microscopy, energydispersive X-ray analysis, and Raman microscopy has led to the identification of the pigments lead tin yellow (type I), lazurite, vermilion, calcite, lead white, red and yellow iron oxides, umber, lamp black, and green earth on the canvas. These pigments are entirely typical of Vermeer’s palette and are consistent with a large body of other technical and art historical findings on paintings by Vermeer and other Dutch 17th century artists. While not authenticating the painting as being by Vermeer, the results provide further critical material that is consistent with this attribution. This case study also provides an opportunity to outline the role of analytical and forensic sciences in the examination and attribution of art objects. There are about 35 paintings currently considered to be by the celebrated Dutch (Delft) artist Johannes Vermeer (1632-75). A further work, Young Woman Seated at a Virginal (Figure 1), was at one time considered to be by Vermeer and held in the collection of Sir Alfred Beit throughout the early 20th century. However, in 1945 the expert forger Han van Meegeren revealed that he had sold seven forged Vermeers to unwitting museums and collectors between 1937 and 1943. In consequence, a number of works previously believed to be by Vermeer, including that specified above, were deattributed. The leading authority at the time, Arie de Vries, subsequently attempted to have this particular painting reinstated as a Vermeer, but without success, and it fell into near obscurity in private hands. The present investigation was undertaken in an attempt to ascertain by spectroscopic and other analyses the palette of the painting and whether this is consistent with that of other accepted paintings by Vermeer. Most art historical evidence now favors the view that the painting is indeed by Vermeer,1 and no contrary evidence was uncovered in * Corresponding author: (e-mail)
[email protected]. † Christopher Ingold Laboratories. ‡ Current address: Conservation Department, Science Section, Victoria & Albert Museum, South Kensington, London SW7 2RL, U.K. E-mail l.burgio@ vam.ac.uk. § Department of History of Art. | Current address: Art Conservation Department, Buffalo State College, 230 Rockwell Hall, 1300 Elmwood Ave., Buffalo, NY 14222. 10.1021/ac048481i CCC: $30.25 Published on Web 01/25/2005
© 2005 American Chemical Society
the scientific analysis reported herein: Vermeer scholars have agreed, and the painting, which is the only accepted Vermeer not held by a museum or royal collection, was sold recently for £16.2 million to an anonymous bidder.2 The small painting, which measures 25 cm × 20 cm, reveals a young woman with hands outstretched while playing a virginal-a keyboard instrument. Her cream-colored skirt appears from beneath a yellow shawl, which envelops her upper body, with only the white cuffs of an undergarment showing at her elbows. Her body is viewed sideways on, her arms reaching across to touch the keyboard, while her face looks directly at the viewer. The chair back is seen in dark profile, with nothing but a blank wall behind the seated woman and the virginal. The interior is lit by what appears to be natural light from an unseen source in the upper left of the painting, allowing reflections in the polished wood of the front of the instrument (Figure 1). This celebrated case study provides an ideal framework from which to highlight the increasingly important role that analytical chemistry plays in the art world. As collectors and galleries become willing to pay more dearly to compete for newly discovered masterpieces and rare auction items, so too has the forger become more adept at the creation of forgeries intended to satisfy the market. Museum staff and their collaborators have developed newsor redirected existingsanalytical techniques and imaging technologies in order to detect inappropriate materials and procedures used in the generation of forgeries. Thus, the relatively simple examination technique by UV radiation is now easily fooled by forgers who incorporate fluorescent dopants, such as ZnS, into their inpainting medium and varnish in order to give the appearance of age and to mask significant restorations or outright forgery. In consequence, potential purchases by museums or private collectors need to be subjected to ever more sophisticated modes of examination such as those offered by Raman microscopy. EXPERIMENTAL SECTION Examination Procedure. The painting was examined using a variety of analytical techniques, which are described below. Some of the examination techniques were utilized to discern the artist’s methods used in the creation of the painting while others (1) Sheldon, L.; Costaras, N. Oud Holland, in press. (2) Sotheby’s Catalogue, 8 July 2004. The painting, owned by the late Baron Frederic Rolin, Belgium, was sold on this date.
Analytical Chemistry, Vol. 77, No. 5, March 1, 2005 1261
Figure 1. The painting Young Woman Seated at a Virginal (before restoration).
were aimed at identifying specifically the materials of the work.1 The discovery of atypical working methods or the identification of anachronistic pigments (for instance, modern synthetic ones) on the painting would of course undermine any claims for authenticity. In practice, only artistic techniques consistent with those of known Vermeer paintings could be documented and only period pigments known to be part of Vermeer’s palette could be identified. Some techniques such as Raman microscopy have the advantage of not requiring samples to be taken. Stereomicroscopy. The surface of the painting was examined under a Kyowa Trinocular examination stereomicroscope at up to ×90 magnification. Microphotographs were taken of various areas to record the surface appearance of the different colors. In particular, photographs were taken of the areas sampled for both pigment content and cross-section analysis. Both mixtures of pigments and individual pigment particles could be seen at this magnification.1,2 1262 Analytical Chemistry, Vol. 77, No. 5, March 1, 2005
Cross-Section Analysis. Microscopic samples were prepared as cross sections in clear-setting polyester resin (clear casting AM, Tiranti) before being cut and polished for examination under a Zeiss Axiolab metallurgical microscope using reflected and transmitted light at up to ×1000 magnification. Details of the sample location and areas cleaned are available in ref 1 and also in four reports at the Paintings Analysis Unit, UCL, L. Sheldon and C. Hassall, Reports C1156, C1360, C1639, and F1775, 19972004. The same cross sections were also examined under a Zeiss Jenamed 2 ultraviolet (UV) illuminated fluorescence microscope at ×250 magnification. Pigments could be seen within the layers of paint, a few of which could be identified by their physical appearance (color, fracture, transparency, particle size). Polarized Light Microscopy (PLM). Dispersions were made of some pigment samples from the painting. This method is useful for the identification of the different pigments present in a paint mixture and is a common preliminary optical procedure before
other forms of analysis. PLM can play a crucial role, for example, in distinguishing smalt from the 19th century painter’s pigment cobalt blue. The pigments were set in Meltmount resin, refractive index 1.662 (Cargile) and were examined under a polarized light microscope (Zeiss Axiolab) at up to ×1000 magnification in both plane and crossed polars and by partly reflected light. Standard procedures for identifying pigments were undertaken, and the particle sizes of some pigments were noted. Energy-Dispersive X-ray (EDX) Analysis. Cross sections and pigments set on aluminum stubs were examined by EDX analysis on a scanning electron microscope (JEOL superprobe 733) at various stages in the analytical procedures. EDX analysis was conducted with an Oxford Instruments ISIS system (15 kV accelerating voltage; peaks checked against known standards of silicates and oxides). Yellow, blue, and green pigments within various cross sections were selected and analyzed, with area scanning of the electron beam allowing an assessment of the elements involved in a mixture of pigments making up a particular color. X-Radiography and Infrared Reflectography. Two X-radiographs of the painting were taken at different stages of the technical examination of the painting, the second taken under different operating conditions to clarify the underlying image (illustrated in refs 1 and 2). The painting was also examined by infrared reflectography. Equipment used in this study for the infrared reflectography/digital image processing was a Peca 1010 camera fitted with an English Electric XQ1615 infrared 2/3-in. Leddicon (sensitive to 2200 nm, also with autogain circuitry that was employed), a Fujinon C6 × 17.5B zoom lens, and a range of cmount extension tubes (1-10 mm) for producing the macro images. A 150-W tungsten-halogen lamp was used as the infrared source and a 1000-nm cutoff filter employed to remove visible light. Chemical Tests. A test for lead was performed on small loose samples removed from the painting as well as on samples mounted as cross sections. A test to distinguish lazurite from smalt was also carried out. Raman Microscopy. Fourier transform (FT) Raman spectra were collected on a Bruker RFS 100/S FT-Raman system equipped with a Nikon microscope and a Nd:YAG CW laser operating at 1064 nm and an output power of 1 W. Raman spectra taken with visible exciting lines were recorded using a Renishaw RM1000 system configured with a Leica optical microscope, holographic rejection filters, a dispersive grating (1800 grooves/mm), and a thermoelectrically cooled charge-coupled device detector. Excitation was provided by the 632.8-nm line of a He-Ne laser with a power of
