A STUDY OF ACRYLIC DISPERSIONS USED IN THE TREATMENT OF PAINTINGS
Michael C. Duffy
ABSTRACT—A series of simple tests studied the properties of five acrylic emulsion adhesives: Plextol B500, Rhoplex AC 33 and AC 234, and Lascaux 360 HV and 498 HV. Properties such as peel strength, yellowing, and solubility were investigated to assess the practicality of these products' use in the treatment of paintings with regard to their aging characteristics.
ACRYLIC AND OTHER POLYMER DISPERSIONS have been used as adhesives and in paint media for more than 20 years. These synthetic polymers have found sporadic use in the conservation field as well, specifically for consolidating fragile materials and lining paintings. Properties such as permanence, stability, and reversibility, important to the conservator, are not always important in industry; yet often these materials have been adapted for use in conservation without extensive testing.
This study focused on the practical reliability of five polymer dispersions used frequently in paintings conservation: Plextol B500, Rhoplex AC 33 and AC 234, and Lascaux 360 HV and 498 HV. These adhesives were chosen because they are readily available to the conservator. Properties of peel strength (reversibility), color change, and swelling of dried films (solubility) were examined. Both aged and unaged samples were tested and examined for flexibility. Consideration of these data provided information about the long-term stability of these products.
In a recent paintings conservation refresher course sponsored by the FAIC, these materials were demonstrated to have advantages in their low toxicity, ease of short-term reversibility, and compatibility with other materials. Disadvantages included the risk of moisture reactivity when used as a wet laminate, some application difficulties with large fabrics, changing of product formulation without notice, and, most important, the lack of comparative testing.1
2 LITERATURE REVIEW
THE PRIMARY SOURCE FOR INFORMATION on acrylic polymers and their properties remains Robert Feller's “Polymer Emulsions,” first published in 1966 and included as an appendix in the pioneering On Picture Varnishes and Their Solvents (1985). This study includes fundamental information on polymer emulsions, such as manufacturing methods, drying phenomena, and film formation. Physical properties such as minimum film forming temperature (MFT) and glass transition temperature (Tg) are also noted for some acrylic copolymer emulsions. Feller's study is essential reading before undertaking further research into acrylic dispersions.
A review of recent conservation literature in AATA revealed the first mention of polymer dispersions used as lining adhesives in an article by Mehra published in 1972.2 Mehra's study on the use of Plextol B500 as a lining adhesive included data on peel strength and sheer strength. In articles published in 1974 and 1975 Mehra described the use of Plextol B500 in conjunction with a low pressure cold-lining table he had developed.3 Other practical applications of polymer dispersion adhesives were addressed by Volkmer et al.4 Accelerated aging of acrylic dispersion Plextol D360 was done in 1976 by Ketnath, but apparently no test results were reported.5
Not until 1984 were comprehensive test results of these materials published in the conservation literature. Howells et al. examined the aging properties of 14 different dispersions and considered weight change, color change, solubility, tensile mechanical properties, response to heat, and pH changes.6 They found that the most significant changes occurred with thermal aging as opposed to natural, sunlight, and fluorescent aging. DeWitte's study demonstrated that certain dispersion additives sometimes influence the physical properties of a synthetic resin.7 These articles are useful precedents because they begin to examine the long-term stability of these materials being used by conservators today. In addition, Falvey8 and Barclay9 have published case studies in which the application of acrylic dispersion adhesives has been useful in treating difficult problems.
There is much useful information published outside the conservation literature. Excellent information on polymer dispersions can be obtained from Martens' Emulsion and Water–Soluble Paints and Coatings. For information on adhesives and adhesive theory, Wake's Adhesion and the Formulation of Adhesives and Shield's Adhesives Handbook are indispensable. The most current information on advancements in industry can be found in journals such as Colloid and Polymer Science.
3 SAMPLE PREPARATION
EACH OF THE FIVE ACRYLIC DISPERSIONS were prepared as follows:
3.1 Cast Films
The undiluted adhesives were cast onto sheets of silicone release mylar using a doctor knife. This tool spreads an even film of uniform thickness (0.015″). When dry, these cast films could be peeled off for testing or left on the mylar carrier. The Plextol B500 had to be mixed with toluene (toluene 1:10 Plextol) to facilitate spreading; otherwise it beaded up on the mylar surface (for consistency, all samples were prepared using toluene “thickened” Plextol). The Plextol AC 33 and 489 HV dried down to a hard, clear film overnight, while the AC 234 and 360 HV remained tacky and soft. Samples not subjected to accelerated aging were kept under room conditions (approximately 68° F and 50% RH).
3.2 Wet Laminate Lining Mockups
The five adhesives were applied onto 9″ × 7″ pieces of polyester sailcloth10 fabric using a squeegee to achieve a film of uniform thickness (approximately 1/16″). Pieces of the same size polyester sailcloth were placed in contact with these samples to simulate a wet laminate lining. These samples were then placed on the vacuum table in a mylar envelope at 1″ Hg pressure for one hour to assure even adhesion. The mockups were allowed to air dry under room conditions over a 48-hour period and then cut into 1″ × 7″ strips in preparation for peel strength testing.
3.3 Reactivated Lining Mockups
The adhesives were applied to sheets of polyester sailcloth as described in section 3.2 and allowed to dry for 48 hours under room conditions. Toluene was sprayed repeatedly onto the adhesive surface to swell and reactivate the tack. Approximately 500 ml of toluene was required to reactivate five 9″ × 7″ samples. While the samples were still tacky, a second piece of sailcloth was applied to the surface. The samples were then placed under vacuum for one hour at 1″ Hg to simulate lining conditions. After “lining” the samples were cut into 1″ × 7″ strips.
4.1 Aging of samples
HALF OF THE PREPARED LINING SAMPLES—both wet and reactivated—plus the cast films were placed into an aging chamber.11 Both thermal and light aging occurred during the samples' exposure. The chamber was a Plexiglas box fitted with a General Electric 275 Watt Sunlamp Bulb regulated by a rheostat. A small fan kept air circulating in the chamber, and a Heathkit Relative Humidity Indicator monitored humidity inside the chamber.12 An attempt was made to regulate the humidity inside the chamber with a saturated solution of sodium iodide salts. Even so, the humidity fluctuated significantly, mostly in the 36%–40% range. Temperature averaged 46°–58° C (115°–136° F). Light falling on the tray holding the samples was about 500 footcandles. Over a one-month period the samples received 124.25 hours of exposure, equaling approximately 62,125 footcandle-hours. Even in this relatively short period the samples showed noticeable yellowing in both the clear films and the lining mockups. Control samples were kept under room conditions while the others were aging.
4.2 Peel Strength Testing (ASTM D-903-49)
A Scott CRE/500 peel strength tester was used to determine peel strengths for all samples.13 The machine was equipped with interchangeable load cells of 5 or 500 pounds. Where possible the 5-pound load cell was used. Calibrating the machine with a 500-gram weight made it possible to get an accurate estimate of the samples' peel strengths. The weight was hung from the clamp, and a reading was recorded on the same chart paper being used for the sample. All strengths were recorded in grams per linear inch. To get a measurement, the loose ends of the strips were clamped into the opposing jaws of the tester. Turning the tester on caused the clamps to move away from each other and the sample to be peeled apart. A recording device was used to chart the peel strengths so they could be graphically compared.
4.3 Measuring Change in Color
Yellowing was measured using a Minolta Chroma Meter CR-100.14 Data were collected from the aged and unaged samples of each type of adhesive and compared after statistical analysis.
4.4 Swelling of Cast Films
Samples of both aged and unaged cast films were cut into sections weighing approximately 0.5 grams each. These were weighed into glass vials using an analytical balance. Three solvents were tested for their effect on the adhesives: distilled water, toluene, and isopropanol. About 5 ml of the chosen solvent was added to the vials and left under room conditions for 24 hours. Excess solvent was decanted off, and the samples were left to dry for another 48 hours in the fume hood. The samples were weighed again. Percent swelling was calculated with the formula:15
5.1 Aging Results
AFTER MORE THAN 100 HOURS IN THE CHAMBER, the cast films showed slight yellowing. The polyester fabric of the lining samples was yellowed as well, but the adhesive sandwiched in between remained clear. Since light and thermal aging were occurring simultaneously, it is hard to say which was the cause of yellowing. It may be significant, however, that the adhesives exposed to light and heat did discolor while those exposed to heat only did not. Lengthier exposure would have certainly caused more significant differences in discoloration among the samples. No measurable differences in flexibility could be discerned between the aged and unaged samples. All were extremely flexible before and after aging took place.
5.2 Peel Strengths
Some significant trends emerge from the peel strength test data. Average values for the peel strengths of each sample type were calculated from the Scott Tester charts (see table 1). Representative charts from each sample type were chosen for evaluation. Random high and low points were chosen, and an overall average value was attained. Some peel tests gave uneven results, probably because of bubble inclusions in the dried dispersion or unevenness in application that were not apparent during sample preparation. The slightly ribbed texture of the sailcloth fabric was also a factor contributing to variable adhesion. For practical considerations, peel strengths of under 100 grams were considered failures.
Table 1. Peel Strength Values
Peel Strength Values
Differences in peel strengths after aging were most apparent with the Rhoplex films. Aged samples of both wet laminate and reactivated samples had higher peel strengths than unaged samples. This result may indicate a degree of polymer cross-linking that occurred after exposure to heat and light. Moreover, the excessive heat of the chamber, sometimes exceeding the Tg of the samples, may have caused a more tenacious bond. The solvent reactivated samples of AC 33 failed across the board, although the aged samples had slightly higher peel strengths. When used as a wet laminate, the AC 33 held more tenaciously, giving peel strengths of about 450 grams in the unaged samples and about 1,545 grams in the aged samples. AC 234 had slightly higher peel strengths in the wet laminate samples but failed in the reactivated category. AC 234 wet laminate samples had much higher (+1020) values than their unaged counterparts.
Aging also caused higher peel strengths in the 498 HV samples. There was about a 1,000 g higher average peel strength for the aged wet samples for this adhesive. Reactivated unaged samples came close to failing, while the aged reactivated samples had slightly higher peel strengths (about +100 g). This adhesive performed much like the Rhoplexes in peel tests.
The most consistent peel strength values were given by the 360 HV, with very high peel strengths in all categories. There was very little change in values between aged and unaged samples or even between wet and reactivated samples. This result may be explained by the fact that the 360 HV remains tacky when dry and bonds readily to most surfaces. Even the reactivated lining samples formed tenacious bonds with values two to four times higher than those of the other adhesives.
Plextol also gave similar values in the aged and unaged categories, differing only by 300 g or so. The reactivated samples, however, failed in both the aged and unaged trials. Of the five adhesives, the Plextol B500 performed the most consistently between aged and unaged samples without the high bonding of the 360 HV.
5.3 Color Change
To compare changes in color, the L*a*b* system of the Minolta Chroma Meter was used. Chroma (b*) values were chosen for statistical analysis because increases in this number correspond to an increase in yellowing of the sample. Six measurements were taken, and their averages and standard deviations were calculated (see table 2). Confidence levels were established for comparison of selected groups of data. The degree of freedom equaled 10 for the two sets of six samples.
Table 2. Chroma (b*) Values
Chroma (b*) Values
Of the five samples tested, Plextol had the most discoloration, with a difference of 3.04 units between aged and unaged samples. Lascaux 360 HV had the least discoloration, with only a 1.11 difference in Chroma values. Comparison of these two values using the T-Test established that they differed in discoloration significantly, measurable at the 99.9% confidence level. Other discoloration values could not be distinguished from each other at that level of confidence. Comparison of the Rhoplex AC 33 and AC 234 values yielded a less than 50% confidence level. It should be noted that although these are measurable differences, the degree of yellowing appeared to the eye to be about the same for all adhesive samples—the Plextol was not noticeably more discolored than the others. Aged and unaged samples could easily be distinguished, however.
5.4 Percent Swelling
With the addition of water the film samples swelled slightly, turning white and opaque. Swelling in the unaged samples of 360 HV and 498 HV was significantly greater (16% and 10% respectively) than for the aged samples (0.4% and 0.6%), suggesting that aged films are much less permeable to moisture than young films (see table 3). Swelling with water alone would not reverse a polymer film.
Table 3. Swelling Values
Percent swelling of the films in isopropanol was also very small. The samples became semi-opaque with the addition of this solvent. The greatest amount of swelling occurred in the aged and unaged 360 HV samples at around 7.5%. Unaged 498 HV swelled to about 5%, while values for all the others were under 3%. What was surprising was that, on average, greater swelling occurred in distilled water than in isopropanol.
The addition of toluene to the films caused them to swell completely and dissolve into a clear gel. For the most part, unaged samples swelled at a higher percentage than aged films. The least swelling occurred in Plextol and AC 33 films, while swelling of the AC 234, 360 HV, and 498 HV was especially high—in the 30%–50% range.
MATERIALS USED IN CONSERVATION should be stable over an extended period of time. This does not seem to be the case for the polymer dispersions tested. Discoloration (the formation of conjugated double bonds) occurred in measurable amounts in all samples after a relatively short exposure to aging conditions. Changes in peel strengths related to the aging of films indicate increased difficulty of reversibility in the Rhoplexes and 498 HV. While the precise peel strength desirable would depend on a number of factors not examined here extreme highs and lows were judged to be undesirable. The extremely high peel strength value for HV 360 makes its reversibility a harsh undertaking for a fragile art object. Conversely, reactivated samples with low peel strengths do not have the adhesive strength to keep a fabric securely adhered to another and so threaten delamination. Changes in technique of reactivating the adhesive may be required if it is to be used in the lining procedure.
Of the five adhesives tested, the Plextol B500 appears to be the most resistant to peel strength changes over time. Unfortunately, it also sustained the most discoloration of any of the adhesives. Since the adhesive would be hidden from light in a lining, yellowing may not be a factor under consideration. Accepting a material's negative properties to gain the advantage of its positive properties is a compromise frequently made in conservation. Certainly the discoloration of any of these materials is inevitable and must be considered vis-ll
l3mbined adhesives could be tested to determine if a mixture of two adhesives would yield more favorable properties than either one used alone. Comparison of peel strength vs. sheer strength values would further enhance understanding of these materials. There are many avenues for further study of the use of these materials in the conservation field.
WITH THESE RESULTS, it appears that Plextol B500 or Lascaux 498 HV would be the most suitable for lining paintings where high adhesive strength is required. Where a lighter bond is needed—perhaps for paper objects or fine fabrics such as silks—the Rhoplexes would be more appropriate. The very high peel strength of the 360 HV would seem to limit its application alone, but it should prove useful in combination with 498 HV to improve tack. Of the solvents tested, toluene was the best choice for swelling these films.
Unfortunately, the long-term reliability of these materials appears questionable, considering the short aging exposure, resulting discoloration, and increased peel strengths. Further testing with longer-term exposure and interim testing are warranted to determine if an actual trend toward irreversibility with aging exists. In the meantime, the results put forth here should be considered when these adhesives are applied to works of art.
THANKS ARE DUE TO LOWELL PERKINS AND JANET SCHRENK of the Winterthur Museum Program in Art Conservation at the University of Delaware, who served as advisers on this research project. At the University of Delaware, Jeremiah Weaver generously provided assistance in gathering and interpreting peel strength data. Samuel Hudson from DuPont's textile division also provided assistance and encouragement for the project.
1. See FAIC, Williamstown Painting Refresher Notes (Washington, D.C.: FAIC, 1983).
2. V. R. Mehra. “Comparative Study of Conventional Relining Methods and Materials and Research Towards Their Improvement.” Interim Report ICOM Committee for Conservation, 3rd Triennial Meeting (Madrid, 1972). 21/72/4. 28pp.
3. V.R. Mehra, “Nap-bond and Cold Lining on a Low Pressure Table,” Maltechnik Restauro 81, no 2 (1974): 87–95; and “Further Developments in Cold-Lining (Nap-Bond System) In ICOM Committee for Conservation (Venice, 1975), 75/11/5, 1–26.
4. J. Volkmer et al. “An Examination of Lining Materials and Methods for Special Problems in Painting Conservation,” AIC Bulletin 15, no., 2 (1975): 103–5.
5. A. Ketnath, “Acrylic Resins for the Conservation of Paintings,” Meddeleser an Konservering 2, nos. 7, 8 (1976): 223–35.
6. R. Howells et al., “Polymer Dispersions Artificially Aged,” in Adhesives and Consolidants, London: IIC, 1984, 36–43.
7. E. DeWitte et al., “Influence of the Modification of Dispersions on Film Properties,” In Adhesives and Consolidants, London: IIC, 1984, 32–35.
8. D. Falvey, “Practical Cold Lining Developments,” ICOM Committee for Conservation, 7th Triennial Meeting (Copenhagen, 1984) 84/2/12–15 ff.
9. M. H. Barclay, “The Final Report on a Stabilizing Treatment for an Unusually Large and Heavy Contemporary Oil Painting on Canvas.” ICOM Committee for Conservation, 7th Triennial Meeting (Copenhagen, 1984). 84/2/3.ff.
10. The sailcloth was chosen because it is nonabsorbent and would not react with the adhesives being tested. It is a woven fabric that has been cross-linked with melamine formaldehyde for high stiffness.
11. Samuel Hudson and Lowell Perkins were of great assistance in modifying the aging chamber, which was originally designed by Marion Mecklenberg.
12. Heath Co., Benton Harbor, Mich. 49022.
13. The Scott Tester was made available by Dr. J. Weaver, chairman of the University of Delaware Textiles Department.
14. Minolta Corporation, Ramsay, N.J. 07446.
15. S. Blackshaw and S. Ward, “Simple Materials for Use in Conservation,” In Resins in Conservation (Edinburgh: SSCR, 1983). 2-1 to 2–15.
Berger, G.“The Procedure of Developing an Adhesive for Paintings: The Importance of Valid Tests,” In Adhesives and Consolidants, 13–17. Paris: IIC1984.
Berger, G.“Testing Adhesives for the Consolidation of Paintings.”Studies in Conservation17 (1972): 173–194.
Bjarnhof, M.“Removal of Damp Blotches by the Aid of a Low Pressure Table,” In ICOM Preprints, 84.2.10ff. Copenhagen, 1984.
Boissonnas, P., and W.Percival-Prescott. “Some Alternatives to Lining,”ICOM Committee for Conservation, 7th Triennial Meeting. 84/2/35–44. Copenhagen, 1984.
Bradley, S.“Strength Testing of Adhesives and Consolidants for Conservation Purposes,” In Adhesives and Consolidants, 22–25. London: IIC, 1984.
Crafts Council. Adhesives and Coatings. London: Crafts Council, 1983.
Down, J.“Adhesive Testing at the Canadian Conservation Institute” In Adhesives and Consolidants, 18–21. Paris: IIC, 1984.
Feller, R. L., N.Stolow and E.Jones. On Picture Varnishes and Their Solvents. revised and enlarged. Washington, D.C.: National Gallery of Art, 1985.
Martens, C.R.Emulsion and Water-soluble Paints and Coatings, Rheinhold Publishing Corp.1964.
Phenix, A., and G.Hedley. “Lining without Heat or Moisture.” In ICOM Preprints, 84.2.38ff. Copenhagen, 1984.
Sander, I.“The Use of Synthetic Resins in Picture Conservation,” In Resins in Conservation, Edinburgh: SSCR, 1983.
Sheilds, J.Adhesives Handbook.Cleveland: CRC Press, c. 1970.
Snedecor, G.Statistical Methods. Ames, Iowa: Iowa State University Press, 1967.
Wake, W.Adhesion and the Formulation of Adhesives, New York: Applied Science Publishers, 1976.
Roehm GmbH, Postfach 4242, 61 Darmstadt, West GermanyRhoplex AC 33
Roehm & HaasRhoplex AC 234
Independence Mall West, Philadelphia, Pa. 19105, USA
Lascaux 360 HV.
Lascaux FarbenfabrikLascaux 498 HV.
Department Restauro, CH-8306 Bruetisellen, SwitzerlandSailcloth
Bainbridge-Aquabatten, 252 Revere Street, Canton, MA 02021