4 RESULTS—POLY(ISOAMYL METHACRYLATE), 27H

TWENTY-FIVE PAINTINGS believed on the basis of National Gallery of Art records to be coated with 27H were tested. Only 5 paintings were found to actually be coated with 27H. This unexpected result is possibly attributable to incomplete documentation of subsequent treatments of the other 20 paintings before the establishment of a formal conservation facility at the National Gallery. An additional 5 of the 25 paintings were coated with varnishes that were identified as methacrylate polymers, perhaps 27H. In these cases, the resolution of the infrared spectra obtained on samples of the varnishes was not sufficient to permit a determination of which methacrylate polymer was present. One of these 5 paintings, which was of expendable quality, had been exposed to unfiltered sunlight and temperature extremes above the gallery laylights as a test for 14 to 24 years. As previously stated, exposure in the “attic” was estimated to be about 15 times that in the gallery, making comparisons with other samples difficult but possibly of interest. The coating was in such an advanced state of degradation that no solvent could remove it. Scrapings of a small quantity of the varnish were used in the infrared analysis to identify the material as a methacrylate polymer. The infrared spectrum of the coating on another of these five paintings revealed that it is a mixture of a methacrylate polymer and a natural resin.

Twelve of these 25 paintings were found to have poly(vinyl acetate) as a coating; in many of these the coating had been applied over a natural resin. One of the coatings was found to be a poly(vinyl acetate) over a layer of poly(vinyl alcohol). The poly(vinyl alcohol) layer was swollen with methyl ethyl ketone and removed mechanically. It was disturbingly difficult to remove. This painting had developed white areas over the surface. Feller (1961) has suggested that these areas are caused by poor adhesion of the poly(vinyl alcohol) layer. Two of the paintings were found to have a natural resin varnish. One of the paintings was not tested because the paint layer was extremely sensitive to solvents.

On the five paintings that were found to actually be coated with 27H, gel formation occurred with application of a mixture of 60% cyclohexane, 40% toluene. In four cases, the gel was soft and rubbery, although in one case it was particulate. Application of 100% toluene caused the formation of additional gel in all of the cases, with the gel forming more rapidly in the more polar solvent. In two of the cases, toluene removed all of the 27H. In the other three, mixtures of toluene and acetone removed the 27H as a gel on two paintings, whereas pure acetone was required for complete varnish removal on one painting, as determined by infrared spectroscopy on cleaning swab residues. In all of the cases, the 27H was found over a coat of either a natural resin or a Rembrandt varnish. These varnishes were usually removed in acetone-toluene mixtures or pure acetone.