JAIC 2003, Volume 42, Number 2, Article 9 (pp. 341 to 362)
JAIC online
Journal of the American Institute for Conservation
JAIC 2003, Volume 42, Number 2, Article 9 (pp. 341 to 362)





In 1967, Gustav Berger initiated a research program investigating lining adhesives for paintings. These investigations focused on wax-resin combinations and culminated in the development of BEVA 371 solution (Berger 1984). A series of formulations was developed, composed of a combination of ethylene vinyl acetate (EVA) resins with a variety of waxes and ketone resins. All of these materials were believed to be stable. After subjecting the combinations to accelerated aging environments involving heat-and light-aging, the effects of the various additions and substitutions on adhesive strength and cross-linking were studied, and the best combination was selected—BEVA 371. The original formulation consisted of 500 g Elvax Resin grade 150 and 170 g A-C Copolymer 400 (two ethylene vinyl acetate copolymers of different vinyl acetate contents), 300 g Ketone N Resin (a condensation product of cyclo-hexane), 40 g Cellolyn 21 (a phthalate ester of technical hydroabietyl alcohol), and 100 g oil-free paraffin. The elements were blended to form 38% solids in a solvent mixture of aromatic and nonaromatic solvents (e.g., 1,250 g toluene or xylene and 750 g aliphatic mineral spirits) (Berger 1975). The solid compounds remain the same today (Ketone N is now replaced by Laropal K80, a condensation product of cyclohexanone with the same melting-point range). Only minor variations were made in the composition and proportion of the solvent mixture (Berger 1976) and in the dilution, presently at 40% solids.

As explained by Berger (1976), the key to a hot-melt adhesive such as BEVA 371 is the combination of high and low molecular weight polymers. The high molecular weight elements of BEVA 371 include the ethylene vinyl acetate copolymers: A-C Copolymer 400 and Elvax Resin grade 150. These materials provide structural strength and toughness but have high softening temperatures, 83C and 116C respectively. With its softening point at 75–85C, the lower molecular weight resin Laropal K80 (or originally Ketone N) decreases the mixture's viscosity and softening point. At the activation temperature, it behaves like a solvent that dissolves the EVA polymers and achieves wetting, which is essential in creating a good bond. The plasticizer-tackifier Cellolyn 21 and the wax, each with softening points at 65C, further reduce the softening point and improve wetting. The wax also affects setting speed and is believed to protect the mixture from irreversible cross-linking, ensuring BEVA 371's future removability (Berger 1976).

During the development of BEVA 371 solution, Berger conducted tests on the adhesive's strength, compatibility, effectiveness, reversibility, and aging qualities in accordance with its intended use in paintings conservation, with positive results. The adhesive's mechanical performance was characterized by testing for peel strength, resistance to shear, and flexural properties. These tests were accomplished on naturally and artificially aged samples (Berger 1972). In testing the aged samples, Berger was able to compare the characteristics of BEVA 371 solution over time with those of other aged samples of conventional adhesives used in painting linings. It is beyond the scope of this article to discuss these analyses in detail; however, it can be generally concluded that BEVA 371 was extensively analyzed and good results were obtained. It was stated that cross-linking of the adhesive would not occur over time and that, if used properly, it could be easily removed from fragile materials without staining or damage (Berger 1970).


The following section gives an overview of several studies and assessments of BEVA 371 used as a backing adhesive. Details of the types of assessment and the results obtained are presented in table 1 by date of publication.

As part of the Adhesive Testing Program carried out at the Canadian Conservation Institute (CCI), BEVA 371 solution was analyzed isolated from lining fabrics or supports (Down et al. 1996). This study focused on the properties of the adhesive before and

Table . Results of Select Beva Assessments Organized by Date of Publication
Table .
Table .
after dark and light aging. Results showed that the pH of the adhesive remained within the acceptable range before and after aging, that minimal volatile emissions occurred over time, and that the adhesive remained flexible upon dark and light aging. Notably, the BEVA 371 solution yellowed significantly upon light aging. Though this may not be an issue aesthetically, yellowing can indicate an instability or degradation of the adhesive (Down et al. 1992). A future CCI research project is to examine the interaction of a range of adhesives over an extended aging period on substrates such as silk, paper, and possibly skin and leather. The purpose would be to investigate whether any deterioration of the substrate occurs due to presence of the adhesive.

Several other studies focused on BEVA 371's bond strength when used as a lining adhesive in painting conservation. These studies investigated how the bond strengths (using peel or lap-shear tests) varied as a function of treatment factors such as application method, temperature, time of exposure to the activation temperature, thickness of the adhesive layer, dilution of the adhesive, solvent used, pressure, type and state of the canvas being lined, lining fabric, and pretreatment of the lining fabric. Katz (1985), Hawker (1987), Pullen (1991), Hardy (1992), Gayer (1992), and Daly Hartin et al. (1993) investigated some of these factors for the BEVA 371 solution. Hawker (1987) accomplished similar investigations for an Adam Chemical Company “BEVA” film of 3.5 mil (this company modified the original formulation of BEVA film and is no longer authorized to produce it), and Forest (1997) focused her research on the BEVA 371 film of 2.5 mil thickness. These studies give useful insight into the wide range of bond strengths that can be achieved with BEVA 371 depending on how it is prepared and applied, as well as on the object and lining fabric it is in contact with. The latter variables will be further discussed in section 3.2. Of the above studies, only Daly Hartin et al. (1993) investigated the effects of natural aging and found no significant trends in change in strength.

Daly Hartin et al. (1993) also found that identically prepared samples of BEVA solution showed a wide range of bond strengths in spite of the high degree of consistency in sample preparation. Thus even small differences in preparation or application, which one may tend to overlook especially in a practical application, can significantly affect bond strength. Interestingly, the film has provided more consistent bond strengths in terms of a lower standard deviation and coefficient of variation as compared to films made with the solution (Forest 1997), a result that the author explains as due to its even thickness and its ease in application with no delays due to solvent evaporation. In other words, the film provides a better control of the range of bond strengths that can be obtained with BEVA 371. Daly Hartin et al. (1993) give a useful discussion on what is considered “adequate bond strength,” pointing out how the concept varies with the lining fabric, the object, and, subjectively, the practitioner.

Benjamin (1994) conducted a practical study focusing on reversal of the adhesive and found that BEVA 371 used for lining may not be as easily reversed as one might hope, the problems being high strength and adhesive transfer to the painting substrate. The author noted that the project was entirely empirical in nature, and she concluded that a more in-depth examination of the process was necessary. Berger refuted her claim that reversibility of BEVA-lined fabrics is not easy, arguing that Benjamin had not closely followed outlined instructions for its use (Berger 1995).

Once the application of the two BEVA 371 products expanded from paintings conservation to the treatment of skins and leathers, related assessments were carried out that focused on practical issues such as ease of use and compatibility. Fenn (1984) describes BEVA 371 solution as an excellent adhesive when used cold and at full strength on native-tanned skin, as well as on oiled and waxed skins. Calnan et al. (1991) noted several advantages of BEVA 371, whether film or solution: it has one of the lowest activation temperatures (65C) of the conservation choices of thermoplastic adhesives, it does not remain sticky upon cooling, and it produces flexible bonds. BEVA 371 (solution and film) was among the five adhesives that Calnan et al. found most suitable for bonding support fabrics to upholstery leather, and it produced the strongest heat-activated bonds. Alter-natively, the bond created by BEVA 371 used wet at room temperature was too weak to be measurable on a tensometer. Additionally, the authors set up preliminary accelerated-aging experiments focusing on the BEVA 371 solution isolated from a lining fabric, and they found little change in tensile strength or percent elongation after five weeks at 50C and 0–5% RH.

Though the use of adhesives in textile conservation is infrequent and can have serious drawbacks, they have been employed when textile conservators are confronted with problematic artifacts, such as degraded painted banners or powdery silk flags. If an adhesive is to be utilized, BEVA 371 solution has been found to have certain advantages: it can be applied sparingly to form a nap-bond without impregnating the fibers, and it is quite flexible. There are reports of its use in the 1970s and 1980s in this context, and published observations can give helpful insights to conservators treating skins and leathers. Of particular interest to this review are reports in which treatments using BEVA 371 were revisited years later. Mailand (1998) used BEVA 371 solution to line flags, costumes, and textile upholstery, and one of these treatments, accomplished in 1984 on a 1890s silk costume, was reexamined in 1996. During the later examination, it was found that the areas treated with BEVA were stable in that they had not yellowed or attracted particulate matter, but areas left untreated had continued to deteriorate, showing increased fracturing, probably from movement of the textile. On the other hand, Shore (1994) states that when conservators attempted reversal of BEVA 371 solution–impregnated treatments on cotton and silk, the adhesive had become less soluble, or cross-linked, in as little as five years. Upon rinsing the fabric with solvent, a dark gray discoloration resulted, visible at both the front and back surfaces.

Copyright 2003 American Institution for Conservation of Historic & Artistic Works