1 INTRODUCTION

In the conservation of wooden artifacts, it is often necessary to repair broken wooden elements that serve a structural or load-bearing function. Such repairs must have high strength, yet be reversible in the future. Where the break in question is recent and the mating surfaces are clean and undisrupted, animal hide glue is widely believed to be a suitable adhesive, though, in practice, reversal of intact hide glue bonds can be problematic. In cases where the mating surfaces are dirty or damaged, or a gap-filling adhesive is needed, animal hide glue may have greatly reduced strength, and an alternative adhesive may be required. Bulked epoxy resins have found wide use in such instances, and some have the advantage that after setting they can be carved, sawn, sanded, and finished, allowing them to be used simultaneously as both adhesive and fill material. These qualities, along with their high strength, can recommend the use of bulked epoxies over laboratory-prepared acrylic adhesives such as Paraloid B-72 bulked with glass microspheres or fumed silica. One commercially available bulked epoxy, widely used by furniture conservators, is Araldite 1253, a carvable paste epoxy, bulked with titanium dioxide, amorphous silica, iron oxide, and phenolic resin (Ciba 2001). The primary disadvantage of using epoxies in conservation is that, once cured, the bonds formed can be extremely difficult to reverse due to the insolubility of the cross-linked polymer in solvents.

Barrier coatings add a measure of reversibility to adhesive bonds that might otherwise be impossible to release. They are applied as thin films to mating surfaces prior to the application of the difficult-to-reverse primary adhesive. In the conservation of wooden artifacts, animal hide glue has been used as a barrier material for epoxy joins due to its high strength, its ease of use, and its familiarity among furniture conservators. Hide glue, however, has certain disadvantages when used as a barrier material. First and foremost, it is not always reversible in a safe and practical manner. Reversal depends on moisture and/or heat, both of which can cause damage to wood and associated finish materials. Some promising work has been reported using microwave radiation to reverse hide glue bonds (Neher 1996). However, the equipment necessary is quite expensive, and the technique has not gained wide acceptance. In addition, hide glue is known to weaken when exposed to extremes of humidity (Buck 1990) and may degrade over long periods of time.

Recognizing the difficulty in reversing repairs made using hide glue as a barrier, and seeking an appropriate alternative, Anderson and Podmaniczky (1990) tested the suitability of Butvar B-98 (polyvinyl butyral) as a barrier layer for epoxy joins in wood. Butvar B-98 is often referred to simply as a polyvinyl butyral, but it is actually a copolymer of polyvinyl butyral, polyvinyl alcohol, and polyvinyl acetate in a ratio of approximately 40:10:1 (Horie 1987, 101– 2; Monsanto 1994). This polymer was recently shown experimentally by Spirydowicz et al. (2001) to be a suitably stable material for the consolidation of dry archaeological wood. Reporting on the results of their testing, Anderson and Podmaniczky (1990) noted that while barrier coatings should help make epoxy repairs more easily reversible, they must also maintain the overall strength of the bond. The results of their work demonstrated that Butvar B-98 dissolved in ethanol is a suitably strong barrier material when used in conjunction with the bulked epoxy Araldite 1253.

Butvar B-98, while a good alternative to hide glue as a barrier material, nevertheless has significant limitations with regard to its reversibility; it is soluble in polar solvents such as alcohols and in certain mixtures of polar and nonpolar solvents (Monsanto 1994; Spirydowicz et al. 2001). Unfortunately, many varnishes and paints used traditionally to coat wooden artifacts are also sensitive to this range of solvents, making it difficult or impossible to dissolve a Butvar B-98 barrier layer without damaging an adjacent surface coating. This is particularly true because extended exposure may be necessary to allow the solvent (present as either a liquid or vapor) to penetrate deep into the repair and dissolve the barrier. Even after Anderson and Podmaniczky's important study, therefore, a need remained for a well-tested barrier coating of high strength that could be reversed in low-polarity solvents.

In this study, the authors chose Paraloid B-72 and Paraloid B-67 for comparison with the other proven barrier adhesives because they have advantageous dissolution properties, they are readily available, and they are well known and widely used by conservators. Paraloid B-72, a copolymer of ethyl methacrylate and methylacrylate, formerly known as Acryloid B-72, is a Feller Class A material and is not known to become insoluble or degrade over time (Horie 1987, 106). Its inclusion in this study seemed obvious: it is a mainstay in the conservator's studio, and its strength, when used in combination with epoxies in the bonding of stone, has recently been clearly established and published by Podany et al. (2001). In addition, it is soluble in a wide range of solvents, ranging from low-polarity solvents such as xylenes to highpolarity solvents such as acetone. In the context of wooden artifacts conservation, this property confers the significant advantage that a barrier layer of Paraloid B-72 could be applied in acetone—a solvent that is fast-drying and low in toxicity, while remaining reversible in xylenes, a solvent that will not dissolve or damage most associated historic furniture finishes or paints.

Paraloid B-67, polyisobutyl methacrylate, formerly known as Acryloid B-67, is also considered a Feller Class A material even though it is known to cross-link over time (Horie 1987, 108). It was considered in this study because it is reversible in lowaromatic hydrocarbons that present less of a health hazard than the fully aromatic solvents necessary to reverse Paraloid B-72.

The authors conducted comparative shearstrength testing using Araldite 1253 bulked epoxy with all four of the mentioned barrier coatings (hide glue, Butvar B-98, Paraloid B-72, and Paraloid B-67). Results were compared with the measured strength of Araldite 1253 used without a barrier coating. It was hoped that if Paraloid B-72 or Paraloid B-67 proved to be of comparable strength to the other two proven barrier materials, the results of this work would provide conservators with more options in choosing a barrier coating when factors such as the solubility of an original finish need to be considered. To confirm that the adhesive bonds using barrier layers were in fact reversible, the authors also conducted empirical reversibility testing.