The paint layer is flaking primarily at the wood-ground interface; approximately 75%, or 307 m2, of the ceiling requires some treatment (Fig. 7). In order to select appropriate adhesives for the treatment and to develop a comprehensive conservation strategy, a testing and planning phase was carried out.
Seven test areas representative of failed paint layers were selected. Written records and photographs documented the condition before and after treatment. A section in each test area was left untreated as a control.
The adhesives tested included acrylic resins (Rhoplex AC-33, Acryloid B-72, and Acryloid B-67); PVAc-derived polymers (Bakelite AYAF and AYAT, Butvar B90 polyvinyl butyral resin, Gelvatol polyvinyl alcohol, and BEVA 371); cellulose ethers (Methocel A4C and Klucel G); and proteinaceous glues (isinglass and gelatin). Solvents were selected to control working properties. Adhesives in water, ethanol, or mixtures of both were preferred because they softened the paint, allowing it to be returned to plane, and had low toxicity. The adhesives were introduced under and through the paint layer with a brush or syringe.
Figure 7
Ground-facing surface of a main beam from the palace hall illustrating the flaking and loss of paint.
The test areas were evaluated at intervals of one month and two years by assessing adhesion of the paint to the wood and visible alterations to the surface quality of the paint. Change in saturation of color or the matte quality of the paint was undesirable. It was found that many of the adhesives stained the wood and saturated the color, especially the layer of blue-dyed calcite and lead white.
Based on the test results, Methocel A4C (1-2% in 50/50 ethanol/ water) and Klucel G (1-2% in ethanol, and 1-2% in 60/40 toluene/ ethanol) (all w/v) were used for the treatment. Methocel A4C did not saturate the wood and paint layer, except for the blue colors; the adhesion was sufficient within the museum environment, which is maintained at approximately 21°C and 50% relative humidity Klucel G was the only adhesive tested that did not saturate the blue areas and was therefore used selectively, although in general Klucel G may be considered less stable than Methocel A4C (Feller and Wilt 1990:94-95).
The treatment procedure included removal of particulates on the surface with dry brush and vacuum, removal of particulates from behind the detached paint layer with a brush dampened with mineral spirits, injection of the adhesive with a syringe, and readhesion of the paint with a hot spatula. This procedure was modified as needed to suit the condition of the paint layer.
The project was scheduled to be completed at the end of 1996. A representative area will be left untreated for future materials analysis and as a reference for comparison to treated areas. Long-term care will include monitoring the adhesion in areas that were documented in detail, minimizing dust accumulation, and maintaining low light levels as well as a stable temperature and relative humidity.
Conclusion
The materials of the wall painting and palace hall ceiling have many parallels, including a narrow palette, raised gold ornamentation, coarse and fine grounds, and proteinaceous binders with trace levels of conifer resins. The use of different materials over time in the palace hall might suggest a change in the cost or availability of particular pigments; studies of similar architecture would provide useful comparisons.
The materials analysis and the testing and evaluation of treatment options, including a review of previous treatments, were considered essential in developing and implementing the conservation treatment of the palace hall ceiling.
Acknowledgments
The authors would like to acknowledge the following individuals for their technical support: Edward P Vicenzi, Princeton Materials Institute, Princeton University; Raymond White and Jennifer Pilc, National Gallery of Art, London; Robert Walker and Jim Lau, Hewlett-Packard, Paramus, New Jersey; Joe F. Leykam, Macromolecular Structure Sequencing and Synthesis Facility, Michigan State University; Richard Newman, Museum of Fine Arts, Boston; and Marigene Butler, Andrew Lins, Felice Fischer,
And Joe Mikuliak, Philadelphia Museum of Art. The treatment planning phase for the palace hall ceiling was supported in part by the National Endowment for the Arts, a federal agency. The conservation treatment of the palace hall ceiling was supported in part by the Women's Committee of the Philadelphia Museum of Art and by the Institute of Museum Services, a federal agency that offers conservation project support to U. S. museums.
Notes
1 Instruments used: Zeiss Universal Research polarizing microscope; Wild Leitz Laborlux S microscope; Cameca SX50 electron probe microanalyzer equipped with wavelength dispersive spectrometers (WDS) and Princeton Gamma Tech IMIX solid-state detector; Philips PW1729 X-ray generator equipped with a PW1840 diffractometer and Gandolfi cameras; Nicolet 510P FT-IR spectrometer bench and Nic-Plan microscope with a Spectra-Tech Micro Sample Plan with diamond windows; Fisions VG Trio 2000 MS equipped with a Hewlett-Packard 5850 Series II GC; Hewlett-Packard MS Engine with HP5890 Series II GC; and Waters Picotag amino acid analysis system.
2 A section of the painting that was not installed was examined by infrared reflectography. A brush stroke could be seen as a guide for the placement of the proper left eyebrow of
The figure.
3 In contrast, Hanna, Lee, and Foster (1988:34) found the gilding below, not above, the painted decoration for a painting of three bodhisattvas from 1424, Shanxi Province.
4 Proteinaceous material was characterized by FT-IR: NH stretch 3309 cm"1; amide II overtone 3078 cm"1; amide I (CO absorption) 1657 cm"1; amide II (NH2 deformation) 1533 cm"1 (Bellamy 1975:233, 250-57). Selected samples were confirmed by amino acid analysis. The amino acid composition of the paint layer included hydroxyproline and had a profile characteristic of an animal-skin glue (Halpine 1992). The coarse render contained other amino acids not accounted for by the animal-skin glue. It should be noted that hide glue was used for the facing as discussed in the treatment section.
5 GC-MS analysis in the selected ion monitoring mode (SIM) detected a molecular ion at mass 314 and the base peak at mass 239, suggesting the presence of methyl dehydroabietate, the methylated derivative of dehydroabietic acid, a diterpenoid present in conifer resins (Mills and White 1982).
6 Many wall paintings acquired by museums in North America and Europe were treated in a similar manner (e. g., Stout and Gettens 1932).
7 Two 2.54 cm diameter fragments of the painting were saved at this stage of the treatment. They are very rigid and glossy.
8 See Stout (1941:105) for a discussion of compensation, including examples from a Chinese wall painting.
9 It is unclear whether the first paint campaign on the main beams is contemporary with the first campaign on the ridge beam. For example, vermilion and gold are found in both; however, the sequence and composition of grounds are different.
10 These anomalous layers may reflect partial campaigns or a sampling bias.
11 See note 4. Amino acid analyses detected more than one proteinaceous material but included animal-skin glue.
12 The dicarboxylic acid methyl ester, dimethyl nonanedioate (azelate) molecular ion, was found with peaks observed at masses 185, 152, 111, 83, and 74, indicating the presence of an aged oil. Trace levels of methyl-7-oxo-dehydroabietate, an oxidation product of pine resin, were detected with peaks at masses 328 and 253 (Mills and White 1982).
13 GC-MS analysis of the preparation layer of the first paint campaign revealed the presence of abietic and pimaric acid methyl ester analogues. In nature, these compounds appear as the free
Acids, with trace quantities of the esters occurring in a true bled conifer resin. Significant amounts of these esters may have accrued by pyrolysis of the wood in the collection or production process (White 1993).
Materials and Suppliers
Acryloid B-72, Acryloid B-67, Rhoplex AC-33, Conservation Materials, Ltd., 1165 Marietta Way, P. O. Box 2884, Sparks, NV 89431 (manufactured by Rohm and Haas Company, Philadelphia, PA).
Bakelite vinyl acetate AYAF and AYAT, Conservation Materials, Ltd. (manufactured by Union Carbide Corporation/ Chemicals and Plastics, 270 Park Avenue, New York, NY 10017).
BEVA 371, Conservator’s Products Company, P. O. Box 411, Chatham, NJ 07928; or Adam Chemical Co. Inc., 18 Spring Hill Terrace, Spring Valley, NY 10977.
Butvar B90 polyvinyl butyral resin, Conservation Materials, Ltd. (manufactured by Monsanto Polymers and Petrochemical Co., 800 N. Lindbergh Blvd., St. Louis, MO 63167).
Gelvatol polyvinyl alcohol, Conservation Materials, Ltd.
Klucel G, Aqualon Company (A Division of Hercules), Wilmington, DE 19894.
Methocel A4C, The Dow Chemical Company, Midland, MI 48674.
World History
