Notoriously nervous to clean its most famous Leonardo, the Louvre has collaborated with the Laboratoire du Centre de Recherche et de Restauration des Musées de France and the European Synchrotron Radiation Facility to perform a quantitative chemical analysis on the Mona Lisa, 1503-06, and the faces of seven other works by the artist. Using x-ray fluorescence, researchers analysed the composition and thickness of the painting’s layers. They found that he used up to 30 layers, each roughly half the thickness of a human hair.
Meanwhile, the National Gallery in London has just cleaned one of its most celebrated Leonardos after spending more than 15 years investigating paintings by the artist and his Milanese associates and assistants in its collection. In early 2009 the museum embarked upon an 18-month project to clean one of its most celebrated works, The Virgin of the Rocks, about 1491-1508.
“The treatment was driven by the need to improve the work’s aesthetic quality. The varnish had become something between the viewer and the painting,” said Larry Keith, the museum’s director of conservation, referring to varnish applied by the museum in 1948-49, which had become yellowed and cracked.
Keith, in conjunction with the museum’s scientific department, used gas chromatography-mass spectrometry to confirm that the 1940s varnish was a mixture of mastic and turpentine with linseed oil. Organic solvents were then used to reduce the varnish. Keith was keen to point out that the varnish was not fully removed: “The treatment wasn’t about getting off all of the varnish. We wanted to achieve an aesthetic aim to be able to see the depth of the dark colours, which entailed a substantial reduction of the varnish. When we were satisfied with that, the cleaning was finished.”
Was there concern that the cleaning might remove one of the work’s multiple pigment layers, a characteristic of Leonardo’s work in which he applied opaque light colours over darker underlayers to build up fleshtones and create a smokey, sfumato effect? “The paint layers are thin and very delicate but that does not mean that they are inherently more soluble. All our research indicates that he used either walnut or linseed oil and there is nothing to suggest that this work is particularly more vulnerable relative to any other Renaissance painting,” said Keith.
“I feel really positive about this restoration. A really good restoration is one that doesn’t draw too much attention to itself,” said Keith, who added: “There are still open questions, but it’s better to leave them open than to impose answers.”