Everyone’s a critic

Say you’re the curator of an art gallery or museum and you want to send some of your valuable, old paintings to another museum for a special exhibit; how can you be sure that the paintings are robust enough to survive the trip without any damage? Until recently, you’d just have to take a chance, but now a team of British, Spanish and Slovenian scientists has shown that NIR spectroscopy can help provide an answer, by determining the physical condition of old paintings.

Although the front of a painting is usually the main focus of any chemical analysis, the back of the painting is more important when determining a painting’s physical condition, which is determined by the state of the canvas. Specifically, it depends on the type of cellulose fibres making up the canvas, whether linen, hemp, ramie, cotton or jute, and how degraded they are. But even when studying the back of a valuable, old painting, you don’t want to damage it by cutting off a piece of the canvas for analysis.

‘While distressed paint layers can be evaluated visually from the front, there has been no method available so far to evaluate the fragility of canvas without actually cutting off a piece of it, which is certainly unacceptable,’ explains Irene Civil, head of the Conservation Department of the Gala-Salvador Dalí Foundation in Figueres, Spain.

So the team led by Matija Strlič at University College London, UK, decided to turn to NIR spectroscopy, as it’s inherently non-destructive. To start with, they studied 199 canvas samples from the 19^th and 20^th centuries using microscopy and various destructive chemical analysis techniques. They did this in order to determine the degree of polymerisation (DP) of the cellulose fibres, as a measure of degradation, and the acidity of the canvases, as acids can break down the fibres. Then they asked several experts to assess the fragility of each of the canvases, finding that the fragility was inversely related to the DP, with those canvases assessed as ‘very fragile’ having a DP of less than 600.

Next, they analysed the canvases with NIR spectrometry, using partial least square regression and linear discriminant analysis to relate the NIR data with the DP and acidity scores and with the type of cellulose fibre used in the canvas. They found that not only could NIR spectroscopy predict the DP and acidity of the canvases, but it could also determine the type of cellulose fibre in each canvas, allowing it to provide an accurate assessment of their physical condition.

Finally, the scientists used NIR spectroscopy to determine the condition of 12 paintings by Salvador Dalí held by the Gala-Salvador Dalí Foundation. As they report in Analytical Methods, this revealed that although 11 of the paintings were in good conduction, one was very fragile, with a DP below 600. Interestingly, this was one of Dalí’s early paintings, when he used cheap cotton canvases rather than the more expensive linen canvases he used later in his career.