Reconsidering the accuracy of X-ray fluorescence and ion beam based methods when used to measure the thickness of ancient gildings

Since at least 5000 years ago, gilding techniques have been used for decoration purposes or to make artefacts appear as made of solid gold. Investigation of ancient gildings inevitably requires measuring their thickness, which is usually done either through the observation of cross-sections or by methods such as Rutherford backscattering spectroscopy, particle-induced X-ray emission and X-ray fluorescence. Whereas the former method requires sampling, the latter ones are non-invasive and therefore preferable. These non-invasive methods, however, measure the number of atoms per unit area, which can be converted into thickness only if the density is known. So far this has not been considered a problem and, in most cases, density has been assumed as that of bulk gold. This paper aims at highlighting the real extent of the issue by comparing the results obtained by scanning electron microscopy and the above-mentioned non-destructive methods on a group of samples gilded by electroplating. As a result of this work, the density of the gold layers has been found to be three times lower than the theoretical density, presumably due to the presence of micro-porosity, which brings to underestimate the thickness by the same factor. Caution is therefore recommended when providing results on intrinsically porous gildings such as fire gilding and depletion gilding, where the presence of voids precludes a realistic estimation of density. © 2017 Elsevier B.V.