New chemical imaging: MA-XRF scanning of Marten and Oopjen
Rembrandt’s portraits of Marten and Oopjen are currently in the Ateliergebouw for research pending conservation treatment, and every morning when I enter the Rijksmuseum painting conservation studio they are there to be admired.
Before the actual conservation treatment will start, extensive research is taking place, led by a team of experts from both the Laboratoire de recherche des musées de France (C2RMF) in Paris, and the Rijksmuseum. The Rijksmuseum team consists of Petria Noble, Head of Painting Conservation, Gwen Tauber, Senior Painting Conservator, Susan Smelt, Junior Painting Conservator together with Rijksmuseum Conservation Scientists, Katrien Keune and Annelies van Loon and Senior Scientist, Rob Erdmann.
To understand the present appearance of the paintings, we first need to gain insights into the true condition of Marten and Oopjen and learn more about the materials and painting techniques. Therefore, a range of non-destructive imaging techniques is being used for the preliminary research, which we will discuss step by step in this and following blogs, starting with macro-X-ray fluorescence scanning
A recent imaging technique, macro X-ray fluorescence scanning (MA-XRF), provides much more information than traditional X-radiography, and is perfect for gaining a better understanding of the materials and techniques used by Rembrandt in these two portraits. This makes it possible to examine both surface and sub-surface layers in a non-invasive manner in order to visualize hidden paint layers and help identify pigments and their location in the paint layer build-up.
Using the mobile macro-XRF scanner (Bruker M6 Jetstream) developed by the University of Antwerp and Delft University of Technology, the paintings were scanned point by point, line by line, with an X-ray beam mounted on a moving XZ stand. The elements associated with the pigments in each area emit so-called X-ray fluorescence, which is registered by the X-ray detectors in the scanner. Each element emits characteristic X-ray fluorescence.
In this way chemical information of the entire painting is generated in a series of contrast images, which we refer to as elemental distribution maps. For example copper (Cu) was detected in the white of Marten’s eye, which suggests the use of azurite, a copper-based blue in the white paint of his eye. The mercury (Hg) map also indicates the use of the red pigment vermilion in the faces of both figures. In the image above you see the detail of Marten’s face on the left, and to the right first the copper (Cu) map followed by the mercury (Hg) map. Based on the detected elements, most of the pigments in the painting can be identified with this technology.
The size of the scanning window is only 60 x 80 cm, which means that for these large paintings which measure 207.5 x 132 cm, we needed to make 9 scans for each picture. The amount of time also depends on the desired step-size and the dwell time of the scanning. Considering each scan takes 20 hours, it took two weeks to scan one painting! The processing of the acquired spectra, registration and stitching of the maps also took many days.
Of course, Macro-XRF scanning is just one of the imaging techniques used, so stay tuned to this blog in which we will share highlights of the ongoing research and treatment.
Thanks to Petria, Gwen, Susan and Annelies.