Investigation on water content in fresco mock-ups in the microwave and near-IR spectral regions

Abstract

Water diffusion inside masonry is responsible for the majority of the decay phenomena observed in wall paintings and frescos. Thus, the diagnostics of moisture and water content and their monitoring represent a key issue. In order to preserve the integrity of surfaces of artistic interest, investigations by means of non-destructive techniques (NDT) are preferred over others. The aim of this research is to determine methodologies to quantify the moisture content (MC) of frescos by means of the integrated use of two non-invasive techniques, namely fiber optic reflectance spectroscopy (FORS) in the near-IR region and evanescent field dielectrometry (EFD) in the microwave range. The FORS technique has been employed in order to assess the amount of water adsorbed from the surface by means of an analysis of the reflectance spectra in the Vis–NIR (350-2200 nm) range. This technique investigates the electronic and vibrational transitions that are characteristic of each compound and enables their identification. The water content is evaluated on the basis of the 1920 nm and 1450 nm absorption bands. The EFD system consists of a resonant probe connected to a network analyzer. The resonance frequency of the cavity under different moisture-content conditions of frescos is in the 1.0–1.5 GHz range. The device makes it possible to compute, in real time, the MC from a measurement of the transmission coefficient (amplitude versus frequency) through the probe. Fresco mock-ups have been prepared in collaboration with the Opificio delle Pietre Dure in order to recreate most of the possible chromatic shades obtained by mixing iron oxides and hydroxide-based pigments. Measurements were performed by employing both techniques on fresco models after wet-dry cycles obtained by means of poultices with a known water content. The results obtained with these two techniques were compared, and cross relationships between the EFD and FORS data were defined.