Abstract: Historical glass, especially non-durable mediaeval glass, can undergo corrosion. This sometimes results in the formation of dark-coloured manganese-rich inclusions or stains that reduce the transparency of the glass. A conservation treatment with reducing or chelating agents may be considered with the aim of improving the transparency. In this paper, high-resolution desktop microcomputed tomography (µCT) is used in combination with element-specific twodimensional imaging methods for in situ monitoring of manganese removal by hydroxylamine hydrochloride from an archaeological stained-glass sample suffering from manganese browning and from artificially corroded model glass samples. µCT also proved itself useful for the study of the (re-)penetration of manganese into the gel layer during artificial corrosion of a model glass.

Abstract: Historical glass, especially medieval glass, can undergo weathering under the influence of time and environmental conditions. The aim of this investigation was to better understand the processes involved in this natural degradation process by studying artificially altered glass samples prepared for the use of evaluation of conservation methods. Non-durable glass sensors produced by the Fraunhofer Institute (type M1.0) were used as a starting material for artificial alteration. These were immersed in acidic (pH = 0, 2, 4) and neutral solutions (1 h – 8 h). In a second stage the glass samples were immersed in a 0.5 M MnCl2 solution (24 h, 48 h and 72 h), allowing intrusion of Mn from the solution into the gel layer. The samples were characterized at different stages with reflectance FTIR spectroscopy, mu XRF mapping and mu XANES. All measurements were carried out at ESRF, beamline ID21. Reflectance FTIR spectroscopy measurements were performed in the 800 4000 cm(-1) range. Cluster analysis of the resulting maps evidenced the rapid growth of the gel layer in strong acidic conditions. The average spectra for each cluster feature show for the original glass a strong Si-O- stretching band between 900 and 1000 cm(-1), whereas the gel layer could be identified by the increasing Si-O-Si bands around 1100 and 1250 cm(-1). mu XRF maps were recorded at different stages of the experiment at energies around the Mn-K edge (6.539 keV) and with a step size of 2 by 2 m. These confirm the leaching of K+ and Ca+2 from the glass and the intrusion of Mn from the solution. Mn was found throughout the entire gel layer, but with a concentration gradient peaking at the surface. XANES point measurements were recorded at various points where Mn was present. No spatial variation was found, but linear combination fitting of the spectra with various Mn reference compounds indicated that Mn2+Mn23+O4 is the main Mn compound in the gel layer, as was hypothesised by Watkinson et al. The standard corroded glass samples studied here can be used for the evaluation of conservation treatments in follow-up experiments. (C) 2013 The Authors. Published by Elsevier B.V. Selection and peer-review under responsibility of the IA-CS (Italian Association of Conservation Scientists) and University of Antwerp

Abstract: This paper critically compares the performance of four non-invasive techniques that match the accuracy, flexibility, time-efficiency, and transportability required for in situ characterization of leaded glass windows: macroscopic X-ray fluorescence imaging (MA-XRF), UV-Vis-NIR, Raman spectroscopy, and infrared thermography (IRT). In order to compare the techniques on equal grounds, all techniques were tested independently of each other by separate research groups on the same historical leaded window tentatively dated to the 17th century, without prior knowledge. The aim was to assess the ability of these techniques to document the conservation history of the window by classifying and grouping the colorless glass panes, based on differences in composition. IRT, MA-XRF and UV-Vis-NIR spectroscopy positively distinguished at least two glass groups, with MA-XRF providing the most detailed chemical information. In particular, based on the ratio between the network modifier (K) and network stabilizer (Ca) and on the level of colorants and decolorizers (Fe, Mn, As), the number of plausible glass families could be strongly reduced. In addition, UV-Vis-NIR detected cobalt at ppm level and gave more specific information on the chromophore Fe2+/Fe(3+)ratio. Raman spectroscopy was hampered by fluorescence caused by the metal ions of the decolorizer in most of the panes, but nevertheless identified one group as HLLA.