By Valerie Dorge

One of the most important and sometimes controversial stages of conservation treatment is the surface cleaning of museum objects—such as paintings, decorative arts, or archaeological materials—and of monuments.

Painted surfaces especially present difficulties. From an aesthetic point of view, decisions have to be made regarding partial or complete removal of varnish or other coatings and/or overpaint layers. Technical considerations include selection of a method that allows a great deal of control in the cleaning process, so that undesired layers can be removed without damage to underlying ones. Artifacts with porous or unpainted surfaces, such as marble or terracotta, can also present difficulties when grime, stains, or nonoriginal decorative layers are removed. In addition to aesthetic and technical considerations, there are ethical ones as well (see "Finding a Certain Balance: A Discussion about Surface Cleaning,").

Traditional cleaning methods include mechanical removal with scalpels or the use of organic solvents or alkali-based aqueous solutions. With the last two methods, a number of techniques provide more control—in particular, by slowing the solvent's evaporation rate or reducing its migration to surrounding areas. Conservators of paintings have used wax-solvent pastes, while conservators of objects have used thickening agents, including paper pulp, waxes, clays, and various types of cellulose-based materials.

In the early 1980s, alternative cleaning systems were introduced to the conservation community by Richard Wolbers of the University of Delaware. The cleaning systems have an aqueous gel base composed of a polymer resin that thickens with the addition of water, and a surfactant—also a thickening agent—which improves the gel's contact with the surface to be cleaned. Any number of cleaning agents can be added to this gel base. These include solvents that dissolve the varnish or undesired paint layer, enzymes that chemically break down oils, and resin soaps that remove degraded varnishes (the resin soaps have a chemistry similar to that of the varnish, and they work on the principle of like-dissolves-like).

The chemical composition of these systems increases the conservator's control over the cleaning process. The systems can be prepared for specific cleaning tasks. Their effectiveness can be further improved when the conservator has in-depth information on the chemical composition of the surface to be cleaned; on the paint-layer stratigraphy; and on the nature of the coating, dirt, or paint layer to be removed. In addition, they significantly reduce the conservator's exposure to volatile toxic organic chemicals. Because of these important advantages, the new cleaning systems have been widely incorporated into conservation lab practice.

Nevertheless, a number of concerns have been raised by some conservators and conservation scientists regarding the possible long-term effects on surfaces, particularly painted surfaces, of cleaning with these systems. The most pressing concern has been whether or not any residue of the gels is left on the treated surfaces—and, if so, if the residue might pose a danger to the surface. These dangers would include increasing the solubility of the paint or altering the paint's chemistry, thereby possibly accelerating its deterioration. A number of institutions have undertaken studies of various aspects of possible long-term effects from use of the gels systems. However, because of the specific focus of these studies, the questions regarding residue have not been fully answered and remain an impediment to more widespread use of solvent-based gels.

The Gels Cleaning Research Project

Conservation image

To provide the conservation community with definitive answers on "the residue question," the GCI incorporated research on this subject into a broad GCI scientific research project on surface cleaning, begun in 1997. The GCI gels project is in collaboration with the Analytical Laboratory of the Winterthur Museum, Garden, and Library, the Winterthur - University of Delaware Program in Art Conservation, and the Department of Chemistry, California State University, Northridge (CSUN). The Getty research team includes scientists and conservators from the GCI and the Getty Museum.

At the start of the project, three main research areas were identified: (1) quantitative measurement of gel component residues; (2) aging characteristics of the surfactant components and investigation of the interaction between gel residue and paint layers; and (3) analysis of the surface of paintings cleaned during the past 10 years with the gels systems. In the course of research, a fourth subject was added: a study of solvent residue left on and in paint layers following traditional cleaning with only solvents or solvent mixtures.

To measure the amount of residue left after use of the gels systems, the project team developed a highly sensitive methodology that uses radioactive materials to label the gel components. Four chemically identical gel formulations were used, each with one major component labeled. This methodology was applied to a cleaning experiment carried out in November 1998 at CSUN. An international group of conservators and conservation scientists participated in the experiment (see Conservation, vol. 14, no. 1).

Conservation image

The preliminary results of this study, which is nearly complete, were presented to the conservation community at the biennial congress of the International Institute for Conservation of Historic and Artistic Works (IIC), which took place in Melbourne, Australia, in October 2000. The study showed that there was very little gel residue left on the cleaned test surfaces. To put this finding into perspective, the quantities were equivalent to the average amount of material transferred to a surface by touching it up to 10 times with a finger. The results of this study will contribute to development of parameters for an optimal cleaning procedure.

Current work focuses on identification of decomposition products of the gel surfactants that have been found to be unstable under ultraviolet irradiation. This study is performed with the aid of a residual gas analyzer, which identifies gaseous molecules that are generated during exposure of the gel residue to ultraviolet light. Interaction of the residue with the paint film will be studied as well. Sample films of the four most common types of paint surfaces—casein, egg tempera, distemper, oil—will be subjected to cycles of cleaning, varnishing, and aging. The study will attempt to determine if gel residue is encapsulated during the revarnishing process and, if so, whether it is removed in a subsequent cleaning. Surface distribution of the gel residue will be studied through two-dimensional autoradiography—a technique that locates the distribution of the gel residue in a sample—in combination with laser profilometry, which measures the topographical features of the cleaned sample surface. To date, this study has concentrated on the potential residue on a representative painting—fragments from a large 1911 painting on canvas by Frank Linton that had been vandalized and subsequently donated to the Winterthur - University of Delaware Program in Art Conservation.

Because the gels cleaning systems are widely used for cleaning painted or unpainted objects, the question of gel residue is also applicable for these materials. Therefore, the existence of residue and its potential long-term effect for this class of objects is now being studied. As a first stage, four materials have been identified by the conservators in the Decorative Arts and the Antiquities departments of the Getty Museum. These materials—gilded wood, unglazed terracotta, marble, and plaster—are representative of surfaces commonly found on museum objects. They will be subjected to a cleaning experiment similar to that carried out on the test painting samples. Modification of the methodology developed for the initial experiment, especially with regards to sample preparation, posed a challenge for the project team. The characteristics of these materials—for example, the hardness of the marble and the friability of the terracotta—made separation of individual samples from the prepared panels much more difficult than for the painted canvas samples. (Separation is needed, as each sample is placed in a small vial that is then inserted into a scintillation counter that measures the radioactively labeled components of the residue.) After a number of ways were tried to separate out the samples, separation was achieved by scoring (within a millimeter of the surface) the underside of the sample before cleaning.

Other Studies in the Project

Although the chemical properties of surfactants are known, their long-term stability under natural and artificial aging conditions has not been studied. For that reason, two parallel studies are being carried out in the Analytical Laboratory of the Winterthur Museum, Garden, and Library, to evaluate the deterioration products of a number of surfactants used in surface cleaning on representative films, including a linseed oil film.

To date, the films have been analyzed at 72-hour cycles of artificial aging—which represent 20 to 30 years of normal museum lighting conditions—to identify the degradation process. The rate of change of the nonvolatile surfactants to more volatile degradation products is currently being examined. Based on the project's already completed work measuring the amount of gel component residues, the artificial aging tests are being repeated, with sampling at shorter time intervals. Preliminary results indicate that the surfactants degrade rather quickly. These studies will provide conservators with the necessary information to help them select an appropriate surfactant for a gel formulation.

Conservation image

In a complementary accelerated aging test at the GCI, a quantity of cleaning gel containing the surfactant Ethomeen was applied to a sample; half of the sample was then covered with aluminum foil, and the full sample was exposed to ultraviolet radiation. The amount of radioactivity decreased tenfold during the eight-week test period of continuous exposure. This indicates that the Ethomeen is prone to decomposition by ultraviolet radiation and that some low-molecular products of the decomposition process evaporate from the irradiated paint surface. This interesting finding confirms the Winterthur experiments and will help to further the understanding of what happens to Ethomeen when it is left on the paint surface after cleaning. Further work on Ethomeen residues remains to be done.

The third part of the project will involve analysis of the surface of paintings cleaned with gels systems during the past 10 years. Because these systems were first used in the mid-1980s on museum artifacts, analysis of the surfaces of some of these artifacts provides the potential for determining if any optical or chemical change has taken place on or within the surface layers over time. It is hoped that the project's analyses can identify the source of any such degradation as a way to determine if the degradation can be linked to residue from the gel cleaning components.

Conservation image

The two main components of any residue will be the surfactant Ethomeen and the gelling resin Carbopol. Investigation of a potential analytical method to detect Carbopol in the residue on the surface of paint samples was successful for a model situation in which a large amount of sample was available. However, it was not successful for the small amount of sample that could be taken from a museum object. Therefore, Ethomeen now is being investigated as a marker for detecting residue, as the available analytical methods can identify it more easily. Analysis of samples collected from objects treated with gels over the last 10 years is just beginning. The samples were taken from seven 18th- and 19th-century paintings and decorative art objects.

The ongoing studies of gel residue being conducted will provide an insight into the amount and chemical composition of residue left on the surfaces of objects following gel cleaning. The fourth part of the project will compare these residues with possible residues left by traditional solvent-cleaning methods. A series of experiments is being performed to provide information on any solvent residue in the paint layers following solvent cleaning. The solvents selected are those most frequently used in conservation practice today in the United States and in Europe: acetone, benzyl alcohol, dimethylformamide, dodecane, ethanol, isopropanol, methanol, N-butylamine, toluene, and xylene. This study also includes an investigation of the potential for even the highest-purity grades of organic solvents to introduce to the cleaned paint layer impurities that may not evaporate and which, therefore, may contribute to any degradation of the paint layer.

An important objective of the gels cleaning research project is to make the results of its studies available to the conservation community. In addition, the project seeks to provide conservators with some recommendations to help them prepare gel formulations that will be the most effective in the cleaning process or in removing layers—while minimizing the risk of future damage to surfaces through degradation from residue or other chemical reactions.

As already noted, the preliminary results from the quantification of gel residue were presented at the IIC Congress in Melbourne. Other dissemination efforts are under way. An article entitled "A Survey of the Conservation Literature Relating to the Development of Aqueous Gel Cleaning on Painted and Varnished Surfaces" by project team member Narayan Khandekar will appear this year in volume I of Reviews in Conservation, a peer-reviewed journal published by the IIC. A full report on the project—including the methodology developed for the experiments, the data obtained, the conclusions reached, and the pertinent recommendations—will be published by the GCI.

Valerie Dorge is a project specialist with the GCI. Other members of the gels cleaning project assisted in the preparation of this article.

Members of the Gels Cleaning Research Project Team

Getty Conservation Institute
Valerie Dorge
Nora Ibarra
Narayan Khandekar
Herant Khanjian
Dusan Stulik
Alberto de Tagle

J. Paul Getty Museum
Brian Considine, Decorative Arts Conservation
Joe Godla, Decorative Arts Conservation
Mark Leonard, Paintings Conservation
Jeff Maish, Antiquities Conservation
Eduardo Sanchez, Antiquities Conservation
Yvonne Szafran, Paintings Conservation

Winterthur Museum
Janice Carlson, Analytical Laboratory
W. Christian Petersen

Winterthur - University of Delaware Program in Art Conservation
Richard Wolbers

California State University, Northridge
David Miller, Department of Chemistry


Components of Gels Cleaning Systems

The gels cleaning systems have an aqueous gel base that includes a polymer resin and a surfactant. The surfactant is usually Ethomeen (an ethoxylated [15] cocoalkylamine), and the water-based resin is usually Carbopol, a polyacrylic acid.

To this gel base, any number of cleaning solvents can be added. The most common solvents used in the solvent-based gel formulations include acetone, isopropanol, ethanol, toluene, xylene, and benzyl alcohol, or mixtures thereof.