Effect of visible light on chrome yellow pigments: towards safer illumination of paintings

28-07-2015

Knowledge of the effect of visible light on light-sensitive pigments such as chrome yellows is of great importance for the long-term conservation of unique masterpieces including those by Vincent van Gogh. A combination of high spatial resolution µ-XRF and µ-XANES spectroscopic methods at beamline ID21 and electron paramagnetic resonance allowed a team of scientists from Italy and Belgium to establish that visible light of a blue-green colour is the most effective in stimulating the darkening process of chrome yellows. This occurs via a Cr(VI)→Cr(III) reduction process that also involves Cr(V)-species as intermediates.

  • Share

The darkening of chrome yellows is a serious conservation problem, threatening iconic late 19th-20th century works of art such as paintings by Vincent van Gogh.

Our previous Cr K-edge µ-XRF and µ-XANES investigations at beamline ID21 on micro-samples taken from Van Gogh paintings and on artificially aged paint models demonstrated that chrome yellows (generally denoted as PbCr1-xSxO4, with 0≤x≤0.8 due to the variable sulfate content) are prone to darkening due to a light-induced reduction of the original chromate ions to Cr(III)-compounds [1-4]. This process is more favoured for sulfate-rich, lemon-yellow, orthorhombic PbCr1-xSxO4 varieties of the pigment (with x>0.4) orange-yellow, monoclinic PbCrO4 that is the most lightfast of these materials [1,3].

The alteration phenomenon makes it challenging to establish the optimal museum conditions to ensure a safe display of unique masterpieces, especially in view of possible employment of emerging illumination devices such as white light emitting diodes (WLEDs) that often emit a considerable amount of violet-blue radiation.

Building upon knowledge acquired through previous studies, our latest Cr K-edge µ-XRF and µ-XANES investigations at beamline ID21, combined with laboratory UV-Vis and electron paramagnetic resonance (EPR) spectroscopic analyses, aimed to (i) evaluate the effects of the violet-blue-green light (400-560 nm) emitted by several white sources and (ii) determine the wavelength dependence of the alteration process of different types of chrome yellows.

As Figure 1 illustrates, the darkening of the surface of a series of laboratory-prepared paint models depends on the illumination device used for aging.

Photographs of the paint models before and after aging with different white sources and irradiance profiles of the lamps used for the aging.

Figure 1. a) Photographs of the paint models before and after aging with different white sources. b) Irradiance profiles of the lamps used for the aging.

After aging, the abundance of Cr(III)-alteration within the superficial layer shows a correlation with the total colour change (expressed as ΔE*) of the paint surface while not being dependent on the radiant flux (Figure 2a-b). The lightfast PbCrO4 paints contained up to 10-15% of Cr(III), irrespective of the light sources employed (results not shown). For light-sensitive PbCr0.2S0.8O4 samples, the Cr(III)-abundance gradually increased with increasing violet-blue-green radiation (400-530 nm) emitted by the white light sources (Figure 2c).

RG composite Cr(VI)/Cr(III) chemical state maps obtained from PbCr0.2S0.8O4 paints after exposure to different lamps.

Figure 2. a) RG composite Cr(VI)/Cr(III) chemical state maps obtained from PbCr0.2S0.8O4 paints after exposure to different lamps [step sizes (h×v): 0.8×0.3 µm2; dwell time: 100 ms/pixel]. Green labels indicates the fraction of Cr(III)-species obtained from the linear combination fitting of the line of Cr K-edge µ-XANES spectra. Plots of b) ΔE*final and c) the sum of the photon counts in the 385-415 nm and 485-515 nm range vs. the average Cr(III) relative amount of the upper first micrometre.

The sensitivity of this pigment in both spectral ranges was unequivocally confirmed by similar investigations carried out on a series of PbCr0.2S0.8O4 paints exposed to a selection of monochromatic light (Figure 3). Darkening is visible for all paints and it becomes progressively less significant for aging wavelengths above 500 nm (Figure 3a). The reduction process is activated not only by λ≤460 nm but also in the 500-530 nm range (Figure 3b).
The efficiency of green light to induce the alteration process could be explained via EPR identification of Cr(V)-compounds. These species may absorb radiation in this spectral range and may play a key role in driving the photo-reduction forward, towards the formation of different Cr(III)-compounds.

Photographs of PbCr0.2S0.8O4 paints before and after exposure to different wavelenghts of monochromatic light.

Figure 3. a) Photographs of PbCr0.2S0.8O4 paints before and after exposure to different wavelengths of monochromatic light. b) RG composite Cr(VI)/Cr(III) chemical state maps of PbCr0.2S0.8O4 after exposure at 288 nm and 500 nm [step sizes (h×v): 0.8×0.3 µm2; dwell time: 100 ms/pixel]. Green labels indicates the fraction of Cr(III)-species obtained from the linear combination fitting of the line of Cr K-edge µ-XANES profiles.

The information gained from this study indicates that exposure to violet-blue-green radiation should be minimised as much as possible for the safe display of paintings containing chrome yellows.

 

Principal publication and authors
Synchrotron-based X-ray spectromicroscopy and electron paramagnetic resonance spectroscopy to investigate the redox properties of lead chromate pigments under the effect of visible light, L. Monico (a,b), K. Janssens (b), M. Cotte (c), A. Romani (a), L. Sorace (d), C. Grazia (a), B.G. Brunetti (a), C. Miliani (a), J. Anal. At. Spectrom. 30, 1500 (2015).
(a) CNR-ISTM and Centre SMAArt c/o Department of Chemistry, Biology and Biotechnologies-University of Perugia (Italy)
(b) Department of Chemistry, University of Antwerp (Belgium)
(c) ESRF
(d) Department of Chemistry “U. Schiff” and INSTM RU - University of Florence (Italy)

 

References
[1] L. Monico, G. Van der Snickt, K. Janssens, W. De Nolf, C. Miliani, J. Verbeeck, H. Tian, H. Tan, J. Dik, M. Radepont, M. Cotte, Anal. Chem. (Washington, DC, U.S.) 83, 1214 (2011).
[2] L. Monico, G. Van der Snickt, K. Janssens, W. De Nolf, C. Miliani, J. Dik, M. Radepont, E. Hendriks, M. Geldof, M. Cotte, Anal. Chem. (Washington, DC, U.S.) 83, 1224 (2014).
[3] L. Monico, K. Janssens, C. Miliani, G. Van der Snickt, B.G. Brunetti, M. Cestelli Guidi, M. Radepont, M. Cotte, Anal. Chem. (Washington, DC, U.S.) 85, 860 (2013).
[4] L. Monico, K. Janssens, F. Vanmeert, M. Cotte, B.G. Brunetti, G. Van der Snickt, M. Leeuwestein, J. Salvant Plisson, M. Menu, C. Miliani, Anal. Chem. (Washington, DC, U.S.) 86, 10804 (2014).

 

Top image: Light-sensitive chrome yellow paints exposed to different wavelengths of monochromatic light.