Vibrational Spectroscopy on Show

Visible-near infrared reflection spectroscopy and X-ray fluorescence are by far the most popular point-analysis spectroscopic methods for non-invasive investigations of works of art. From a qualitative point of view, these methods are popular because they are easy to use and give immediate, readable results. But...

Reflection Fourier-transform infrared (FTIR) and Raman spectroscopy are a competitive choice because they are non-destructive and at the same time provide the high specificity needed for identifying chemical compounds, mixtures of pigments or organic dyes, or uncolored substances, such as binders. Conservator-restorers can be skeptical of these techniques, because they output complex spectra – an attitude that’s encountered more frequently in museums that don’t have on-site analytical laboratories. Additionally, some scientists like to emphasize the disadvantages of both techniques; for example, highlighting the weak signal intensity and fluorescence emissions of Raman spectroscopy, and noting the presence of peaks due to different components found in the material under investigation (binders, varnishes, fibers, and so on) in FTIR. In my view, however, we ought to be using both of these vibrational spectroscopic methods for non-destructive analysis of cultural heritage materials in the field, and we should be testing them under different environmental conditions and on different sorts of artifacts.

There are challenges of course; the first being how we mount the instruments or probes. The apparatus must adapt to different geometries and, in our experience, a tripod equipped with a counter-balanced extension arm provides a very stable mount. An XY stage completes the mount when using a Raman microprobe with a 50× objective. Our setup enables micro-Raman spectroscopy for in-situ examination, giving satisfying results for a range of works of art – from frescoes to illuminated manuscripts and easel paintings (1). Using the technique we can recognize most inorganic pigments even in the presence of organic binders and varnishes that create fluorescence; for this purpose we use two microprobes with two different laser sources – 532 and 785 nm – chosen according to the color of the pigments or the intensity of the fluorescence background. This equipment is also useful for obtaining laser-excited fluorescence spectra from organic dyes, for example, which are difficult to identify in situ with other techniques.

As far as reflection FTIR spectroscopy is concerned, the acquisition of spectra is not as critical as with Raman, while their elaboration and interpretation appear more demanding. For example, the contribution of both specular and diffuse reflection to the collected radiation does require you to make a decision whether to apply an algorithm to the spectrum to make it comparable with transmission spectra found in most databases.

We have seen two different situations with illuminations in ancient manuscripts, where mainly specular reflection was obtained, allowing us to use the Kramers-Kronig transform, and with organic dyes found in historical textiles. These reflect diffuse radiation to give spectra that can be compared with transmission spectra of pure dyes, because the fibers “dilute” the colorant. In northern-Italian illuminated codices, dating to the 16th century, both pigments and binders, such as gum Arabic and egg white, could be identified; this information is usually gathered using laboratory-based analyses (2). For ancient Caucasian textiles dating from 17th to 18th century, we used a library-search method based on the correlation algorithm and on the subtraction of the prevailing spectrum from the textile fiber, which allowed us to identify many different dyes ranging from madder to indigo and tannins (3).

In a rather daring application of the technique, we suspended the spectrometer from a scaffolding at a height of more than four meters to obtain FTIR spectra from deterioration patinas (calcium and magnesium sulfates, and calcium oxalates) in a low relief frieze decorating a Baroque arcaded courtyard (4).

I hope these examples of how we use FTIR and Raman spectroscopy for analyzing artistic and archaeological materials encourage you to experiment with these techniques inside and outside your laboratories for obtaining as much information as possible in a non-invasive manner. In my mind, they certainly deserve their place as analytical tools for helping to conserve important works of art – and no doubt, there are applications in a great number of other fields.