The whole picture: ACCN asks the Canadian Conservation Institute what chemical techniques can reveal about paintings.

When considering pigments, it is easy to see how chemistry plays an integral part in the creation of artistic paintings. Early pigments were derived from the earth and from minerals. Later pigments were made from animal and vegetable dyes. More modern pigments are synthetic in nature, (1) can include newer elements, and therefore have different physical properties compared to earlier pigments.

What is less obvious is the integral part chemistry plays in the understanding of paintings. Chemical techniques can be used to probe the differences in pigment physical properties, identifying atoms, compounds, and entire pictures present in paintings that cannot be seen by eye. Combined with extensive art knowledge, these techniques can reveal much about a painting's provenance, often telling a story that would otherwise remain hidden.

Marie-Claude Corbeil has many of these stories to tell. She is a senior researcher in the analytical research laboratory at the Canadian Conservation Institute in Ottawa, ON. Having received her PhD in inorganic chemistry at l'Universite de Montreal, Corbeil is a specialist in X-ray diffraction. She was happy to discuss a variety of the work done in the analytical lab.

ACCN: What is the role of your lab in the Canadian Conservation Institute?

Marie-Claude Corbeil: The role of the analytical research laboratory is to analyze materials from museum objects and works of art found in Canadian public collections to see what they are made of. We answer questions related to their conservation and authenticity in the case of paintings, or time of manufacture in the case of archaeological objects, for example. We examine and analyze all types of objects, from paintings to scientific instruments.

ACCN: Why are paintings analyzed?

M-C. C.: Paintings are analyzed typically for two reasons==to answer questions related to their conservation or to their authenticity. In the first case, we could be looking for fugitive pigments--pigments that are affected by light used in paintings that require special conditions when exhibited. Or we might try to detect overpaints, which are touch-ups done by other painters or conservators. In the second case, analysis could help determine if the painting is from a given period or not. That information would then help art historians link the painting to a given artist. Our role is not to attribute paintings to given artists but to determine if the materials and techniques are consistent with a certain time period.

[FIGURE 1 OMITTED]

ACCN: What techniques do you use to analyze paintings?

M-C. C.: Typically, paintings are first examined and analyzed using non-invasive techniques that do not require removing samples. First, it is important to understand the structure of the painting and to see beneath the surface. We use techniques such as X-radiography, infrared reflectography, and ultraviolet fluorescence photography to detect changes in composition, preparatory drawings, and overpaints. This step is very important as it guides the subsequent steps. For example, if there are overpaints, we will be careful not to take samples from these areas as we want to analyze the original paint, not paint that was added later to cover a damaged area.

After this radiographic and photographic examination, the painting is analyzed by X-ray fluorescence spectrometry--a non-invasive method that will provide information about chemical elements present in various areas. By detecting chemical elements that are found in pigments that were introduced only in the 19th century (such as cadmium) or 20th century (such as titanium), we can often easily determine that an old painting is not what it claims to be.

If nothing anachronistic is found after the non-invasive examination, microscopic samples are taken to determine the actual composition of paint of various colours, and to establish the layer structure of the paint. The composition and layering of the paint have varied a lot over centuries and from region to region. It therefore helps in determining when and where a painting was made. Samples are analyzed by techniques such as X-ray diffraction, infrared spectroscopy, and gas chromatography with mass spectrometry (GC-MS). These techniques are standard analytical techniques, but are tailored to the analysis of microscopic samples, such as those taken from precious works of art and museum objects.

X-ray diffraction is used mainly to detect inorganic pigments and fillers. Infrared (IR) spectroscopy is used to detect organic pigments and determine the type of binder (oil, egg, protein), while GC-MS is used to determine more precisely the nature of the binder--the exact oil (linseed oil, poppyseed oil, etc.), or the exact nature of a natural resin varnish (dammar, mastic, etc.). Of course, these techniques over lap. IR spectroscopy can also detect inorganic compounds, and GC-MS can sometimes detect organic pigments. But most of the time, several techniques must be used to get the complete makeup of the paint.

[FIGURE 2 OMITTED]

ACCN: How do you prove a fake?

M-C. C.: It is actually easier to detect a fake than to confirm that a painting is original. Forgeries are often produced just one time, just to fool people into making a sale. They are not particularly carefully made, and there is always something that will prove to be wrong, anachronistic. I hate that romantic idea of the faker being a genius able to fool the experts! There are a lot of technical details about painting and mistakes are easily made, such as the use of an anachronistic pigment.

Figure 1 shows a painting that was analyzed in our lab as part of an art fraud investigation. The painting, signed "Vincent," was attributed to Vincent van Gogh. The infrared reflectogram (Figure 2) revealed that the current landscape was painted over a previous portrait. Titanium white (TiO2), a pigment developed in the early 20th century, was identified. This indicated that the painting could not have been painted by van Gogh, who died in 1890.

When we are asked to examine a painting tentatively attributed to Rembrandt or van Gogh, it' s usually not that difficult to spot mistakes if the painting is a fake or a copy. However, even if the materials and techniques are consistent with the period, we cannot conclude that the painting is by Rembrandt or van Gogh. In the case of Rembrandt, for example, a painting could be from the period during which Rembrandt was active, but it may have been painted by another Dutch artist of the same period. When a painting is not attributed to a specific artist or period, sometimes it is only possible to determine a fairly broad period during which the painting may have been painted. So science cannot answer all questions.

For more information on chemical techniques used for art investigation, visit the Canadian Conservation Institute Web site at www.cci-icc.gc.ca.

References

(1) Kirsh, Andrea, and Levenson, Rustin S. Seeing Through Paintings New Haven and London: Yale University Press, 2000.

Jennifer Clarke, MCIC, is a freelance writer based in the Ottawa area.