Richard Wolbers
Abstract
The range of tools available to the conservator for cleaning gilded surfaces has increased enormously in the past decade or so; traditional aqueous methods for surface grime removal, and solvent-solution methods for coatings removal have been augmented now with a variety of additional new formulations, that include surfactants, chelates, and other compounds or reagents, to help extend what the conservator is able to effectively, and selectively, solubilize. Cleaning systems which may prove useful in the context of gilded surfaces, however, may be more problematic when applied to surfaces decorated in faux-gilded techniques. Bronze or brass leaf, toned silver leaf, or more exotic metal leaf or powder containing decorative surfaces may be adversely effected by some of these materials. Corrosion prevention in these decorative metals is of chief concern to the conservator formulating cleaning systems with the desired efficacy. A body of literature from outside of the field can be drawn on to avoid corrosive systems or materials. Specific test methods have been developed by cosmetic chemists to evaluate the corrosivity of surfactants and other related materials in metallic packaging systems. Essentially, an adaptation of one of these test methods (Ziegler) is used in the present study to evaluate the corrosivity of multicomponent cleaning systems on test bronze and silver leaf substrates. The method is an electrical one; where current and potential (voltage) are measured for the metal leaf, along with a reference potential against the tested cleaning preparation. The short circuit potential is also measured, and the three potential values along with the current observed are plotted. The slope of the lines connecting the three points allow one to determine whether the substrate metal is dissolving, as well as the intensity of the reaction. Among the aqueous materials tested, a range of surfactant types were evaluated; cationic, anionic, and non-ionic. The non-ionics tested appeared to be less corrosive as a whole more than the anionics or cationics tested ostensibly because of the intrinsically low conductivities of these preparations. Anionics tested included lauryl sulfate and palmitic acid; depending on the counter-ion paired with these surfactants, a range of stability was noted as well. The triethanolamine salts of these surfactants in particular appeared to be essentially passivating to the test metal leafs. The most corrosive of the surfactants tested were the cationics, especially as the Cl– or Br– salts. These same cationics however, when co-precipitated onto polyacrylic acid, and suspended in various solvent systems, essentially exhibited no corrosive reaction to the test substrates. A parallel series of experiments were performed where dilute aqueous solutions of the surfactants were applied to the test leaf substrates and allowed to dry as residual materials; these samples were subjected to accelerated aging conditions, re-wet, and their potentials again measured, along with any measurable corrosion current. Surprisingly, some of the nonionics tested did exhibit slight corrosive tendencies on re-wetting. Other studies by this author have tended to suggest that the mechanism of deterioration on accelerated aging for ethoxylated surfactants is primarily hydrolytic, evolving ethanolic and glycolic moieties on hydrolysis. On complete hydrolysis, fatty ethoxylates tend to produce the concomitant fatty acid or fatty alcohol, which may account for the increased tendency towards corrosiveness post accelerated aging in these materials.