Pigment Discoloration DOI: 10.1002/anie.201303977 Formation of Metallic Mercury During Photodegradation/ Photodarkening of a-HgS: Electrochemical Evidence** Willemien Anaf, Koen Janssens, and Karolien DeWael* Since antiquity, red mercury sulfide (a-HgS), called cinnabar in its natural form and vermilion in its synthetic form, has commonly been used as a pigment. [1] An undesirable phe- nomenon is the degradation of this bright-red material in the presence of light, chloride ions, and humidity, [2] causing it to turn black. This degradation phenomenon has been observed on the surface of frescoes at important heritage sites such as Pompeii, and of paintings from famous masters such as P.P. Rubens and P. Brueghel. An appropriate conservation of this valuable and irreplaceable heritage requires a profound knowledge on the degradation mechanism of the pigment. Several methods, such as X-ray diffraction (XRD), X-ray absorption near-edge spectroscopy (XANES), and secon- dary-ion mass spectrometry (SIMS) have been employed to identify the degradation products: HgCl 2 , Hg 2 Cl 2 , HgSO 4 , Hg 2 SO 4 , and Hg 3 S 2 Cl 2 . [3–5] Yet, none of these compounds has a dark color that can explain the blackening of a-HgS in a convincing manner. Several hypotheses for the decompo- sition and discoloration have been proposed, some assuming the formation of black b-HgS [6, 7] and others that of metallic mercury. [3, 8–10] While volatile mercury has already been observed in photodegradation experiments on mercury ore, [9] neither b-HgS or metallic mercury as a deposit have been detected on naturally and artificially degraded HgS paint. Moreover, the role of chloride ions is not fully understood. This study presents the results of electrochemical experiments that demonstrate for the first time the formation of metallic mercury as a degradation product of HgS induced by the joint action of light and chloride ions. A degradation mechanism consistent with these findings is proposed. First, cyclic voltammetry from 250 to 100 mV (scan rate : 50 mV s 1 ) in 0.1m NaOH was employed to unravel the HgS redox transformation(s) (Figure 1). The behavior of a (red) a- HgS j Pt electrode, preconditioned for 20 seconds in a 1m NaCl solution at a potential of 1750 mV (II), was compared to a non-pretreated equivalent (I). Additionally, another a- HgS j Pt electrode (III) was illuminated for 20 min with a blue laser (405 nm) in a 1m NaCl solution. With electrode I, no oxidation or reduction processes are visible in the supporting 0.1m NaOH electrolyte. The electrochemical pretreatment at 1750 mV induces the formation of metallic mercury at the surface of the a-HgS j Pt, which is due to the applied negative potential (preconditioning). The reoxidation becomes appar- ent during the first cyclic voltammetric scan (Figure 1, II) immediately obtained after the electrochemical pretreatment. The observed oxidation peaks at about 50 mV correspond to the oxidation of Hg 0 to Hg 2+ . [11–13] The double peak is attributed to the underpotential deposition of mercury when applying strongly negative potentials. [14, 15] The waves at about 100 and 200 mV are the corresponding reduction of Hg 2+ formed during the previous oxidative sweep. After photo- degradation of an a-HgS j Pt electrode in a Cl-rich environ- ment that was not electrochemically pretreated at 1750 mV, an oxidation peak arises at a similar potential, indicating the presence of Hg 0 (Figure 1, III). The spot where the laser irradiated the electrode is clearly distinguishable by the disappearance of the red color (blackening; Figure 2). Secondly, the degradation behavior of a-HgS with and without chloride ions and with and without illumination was compared. To focus on the metallic mercury detection (the oxidation of Hg 0 ) at the surface of the degraded a-HgS j Pt electrodes, linear sweep voltammetric (LSV) experiments were performed in the 90 mV to 45 mV range, with 0.1m NaOH as supporting electrolyte. In an initial experiment, an a-HgS j Pt electrode was pretreated by laser irradiation for 30 min, either in a 1m NaCl solution or in deionized water, and subsequently measured with LSV. In deionized water, no Figure 1. Cyclic voltammetric scans (first scans) of a-HgS j Pt electro- des (different pretreatments) in 0.1 m NaOH. I) Without pretreatment, II) electrochemical pretreatment in 1 m NaCl; III) illumination pretreat- ment in 1 m NaCl. [*] W. Anaf, Prof. Dr. K. De Wael Environmental Analysis, University of Antwerp Groenenborgerlaan 171, 2020 Antwerp (Belgium) E-mail: karolien.dewael@uantwerpen.be W. Anaf, Prof. Dr. K. Janssens Antwerp X-ray imaging and instrumentation laboratory University of Antwerp Groenenborgerlaan 171, 2020 Antwerp (Belgium) [**] The authors acknowledge L. Klaassen for valuable discussions and providing samples. We acknowledge financial support from the SDD programme (S2-ART project) of the Belgian Federal Gover- ment. . Angewandte Communications 12568  2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2013, 52, 12568 –12571