Laser Cleaning of Wall Paintings

Summary of an article from yesterday:

The Laser Cleaning of Wall Paintings

Maria Carolina Gaetani, Ulderico Santamaria; J. Cult. Heritage 1 (2000) S199-S207

 The authors attempt to determine the proper laser parameters for cleaning of frescoes via direct experimentation. The laser cleaning of paintings is an order of magnitude more complicated than the cleaning of stone because of the presence of pigments. In order to maintain the painting, obviously one desires to preserve the pigments. However, the difference between constituent pigments in the paint layers along with their thicknesses and the superimposition of consolidants/protective films requires significant and case-specific laser parameters. 

 The authors attempted cleaning with an Nd:YAG laser at both 532 and 1064 nm. The pulse width was 5-6 ns. Their delivery system could use either a focusing or diffracting optic, leading to irradiation areas of .018 or .38 cm^2. The authors also counted on direct particulate absorption, which depends on the stratigraphic (strata) composition and chromatic characteristics of the materials. The authors tested their results by monitoring pH and specific conductivity, as well as measuring the surface temperature after ablation and carrying out colorimetric tests.

 The first painting the authors attempted to clean was a wall painting representing The Visitation. The laser treatment was applied to a portion of the sky which was in a relatively good state of preservation, but which was blackened due to the stratification of various layers superimposed over the original image during previous restorations. Organic solvents were attempted, but did not allow a selective removal of layers. The authors found that the optimal cleaning condition was 532nm, 10 pulses/s (#pulses unstated) fluence of 1.5 J/cm^2 neither focused nor diffused.  The laser removal did, however, lead to yellowing in one of strata due to photo-oxidation of an oily layer. To completely remove that layer required fluences far superior to the damage threshold of the original paint layer, so a solvent was used. The yellowing was, however, less for 532 nm than for 1064.  The authors attempted the use of water as a cleaning moderator, but found that the bubbles formed damaged the paint layer.

 The second painting was a representation of the Nativity. Mary's cloak had been repainted several times during previous restoration, and consequently had a fragmentary surface. The authors found that using a diffuser, 532nm light at a fluence of .56 J/cm^2 and a 10 pulse/s rep rate yielded the best cleaning without harming the underlying original paint layer.

 The last painting was a representation of The Circumcision. The non-original substances were mainly removed with a pH alkaline saline solution, but small residues remained firmly attached to the surface in tiny fragments. The authors found that the optimal operating conditions were 1064nm light with a fluence range of .8-1.4 J/cm^2. Fluences above 1.4 J/cm^2 caused damage to the original substrate. 532nm light produced no appreciable results. 

 The authors claim that laser cleaning is a valid tool for painting restoration, but requires piece-specific calibration.