Influence of pulse duration on the effects induced by three Nd:YAG lasers operating at 1064 nm to tempera paintings mock-ups

Abstract

Laser cleaning of soiled wall paintings is a conservation practice that has been adopted for decades. However, the increasingly widespread availability of laser technology highlights the need to optimize the choice of laser systems and, consequently, the parameters used on a case-by-case basis. In this study, three Nd:YAG lasers (1064 nm) with different pulse durations were tested: a short free-running system (110 μs), long Q-switched system (100 ns), and Q-switched (8 ns). These lasers were tested on tempera painting mock-ups made of one pigment (i.e. malachite, azurite, lead white, orpiment, or cinnabar) mixed with an organic binder (i.e. egg yolk or rabbit glue) applied to a lime mortar. Each laser was also evaluated in tablets composed solely of one pigment. The reference and irradiated temperas were studied using stereomicroscopy, spectrophotometry, attenuated total reflectance Fourier transform infrared spectroscopy (ATR-FTIR), and scanning electron microscopy with energy-dispersive X-ray spectroscopy (SEM-EDX). The pigment tablets were also analyzed with micro X-ray diffraction (μXRD). All samples suffered blackening after laser irradiation, the intensity of which depended on the nature of the pigment and the binder. Overall, pigment susceptibility was broadly comparable across the three pulse-duration lasers. Cinnabar exhibited the highest level of susceptibility, followed by copper carbonates and orpiment, while lead white proved to be the least affected pigment. The binders exhibited a protective effect on the pigments, although this effect was more intense in egg yolk paintings. Regarding pulse duration, it was found that at low fluences, the laser with the longest pulse duration generated the lowest color changes, while at high fluences, this same laser generated the greatest impact on the temperas. Physical effects were observed in the samples (mineral melting and binder loss), though these were not proportional to the intensity of the blackening.