FALL 2016
Recent Developments in the Cleaning of Modern Paints
By Bronwyn Ormsby and Tom Learner
Former Tate Conservation Science Fellow Josefina López evaluates new microemulsions

The last decade has seen a considerable amount of research in the scientific study of modern paints. One area where significant advances have been made is the development and evaluation of potential cleaning treatments. Cleaning accumulated dirt off an unvarnished painted surface is arguably the most commonly executed conservation treatment on modern and contemporary paintings and was identified by a number of researchers by 2005 as a subject in desperate need of research and focus. Before this research, most conservators had very few options for cleaning modern painted surfaces—typically a handful of basic aqueous systems and a range of erasers and sponges.

Perhaps more critically, there was no real understanding of the effects these cleaning systems might have on the paints themselves, especially over the long term. What conservators did know was that cleaning dirt directly off an unvarnished modern paint surface was highly problematic, with an uncomfortably high incidence of paints that were vulnerable to pigment removal, surface abrasion, gloss changes, patchy cleaning, or other forms of undesirable change or damage.

In 2006 the “Modern Paints Uncovered” symposium—held at Tate Modern and organized by Tate, the National Gallery of Art, and the Getty Conservation Institute (GCI)—brought together over two hundred conservation professionals to take stock of current thinking about the conservation of modern paints. An entire section of the symposium was devoted to research into measuring the effects of simple cleaning systems on acrylic dispersion (often called emulsion) paints. Some broad trends were noticed: water-based systems tended to be reasonably decent cleaners, but they would typically cause a significant level of swelling in the paint and would also readily dissolve the migrated surfactant on the surface of the paint film, one of the more abundant additives still present in the dried film. On the other hand, mineral spirits (aliphatic hydrocarbon solvents) seemed to prompt little swelling and were far less likely to disrupt the migrated surfactant. However, they were not particularly good cleaning systems. In summary, neither system looked ideal. Moreover, there was no overall consensus on how to usefully evaluate these observations. In fact, different researchers drew different conclusions about what type of change was acceptable, especially concerning the removal of migrated surfactant. It also became clear that many conservators felt they should wait (presumably very patiently!) until the scientific research had progressed further.

Since then, research has flourished, and approaches have changed as a result of this increasing body of information. A vastly expanded range of potential cleaning materials have now been proposed, developed, and tested by research groups across Europe and North America, and a broader range of paint types has been included in many studies, including acrylics, PVA (polyvinyl acetate), oil-modified alkyd paints (fast-drying oils), and, of course, modern oil paints. Interestingly, many of the nonaqueous systems are now being evaluated for use on a range of water-sensitive materials, including lacquer and water-gilded surfaces.

There is also growing acceptance that alongside scientific researchers, conservators themselves play a crucial role in developing safer cleaning systems. Progress relies on incorporating feedback on recently developed options, experiences using cleaning systems on works of art, and methods of successfully manipulating cleaning options.


Conservators face considerable challenges when cleaning works of art made with modern paints. One major cause is the complexity of modern paint formulations. Gone are the days of simply grinding a pigment powder into a binder. Modern paints contain all kinds of additives essential for (among other things) creating an optimum consistency, a long shelf life, or an appropriate drying time, as well as resisting mold growth. Water-based dispersion paints are particularly complicated, and it is not uncommon for such paints to contain more than ten components in addition to the pigment and binder. Each additive, although perhaps only present at a fraction of a percent by weight, may have a significant effect on how a particular paint responds to cleaning systems. Complicating things even further, the identities of these additives are among the secrets most guarded by all paint manufacturers, making any systematic study completely impossible.

Former Tate Conservation Science Fellow Josefina López evaluates new microemulsions

Then there is the inherent challenge of creating meaningful test samples for use in determining the relative safety of a particular cleaning treatment. For comparative measurements to be made, large swatches of uniformly applied paint samples are necessary, but these will react differently than will forty-year-old paint layers on works of art, particularly where an artist might have manipulated the paint, thinned it with solvent, mixed in other materials, and then allowed it to age naturally. There are of course methods that mimic the ways objects change with age, but concerns remain over the validity of accelerated aging procedures. With cleaning, dramatically different results can also be achieved by applying different pressure, movement, or application methods, even when using the same cleaning agent. All of these practical issues are very real and need to be factored into any testing and evaluation—or at the very least the limitations of testing must be openly discussed. Otherwise, results can be meaningless and of little use to the practicing conservator.

For unvarnished painted surfaces, there are also risks associated with not removing dirt from them. If dirt remains on a paint surface for too long, it can become embedded in the paint film, resulting in permanent disfigurement and possibly initiating deterioration at the paint surface. This is particularly true of the relatively soft acrylic dispersion paints and of modern oil paints that have developed a soft surface layer. Unfortunately, preventive conservation practices such as glazing may not be appropriate for all modern and contemporary paintings since many of them are large or unframed, or because of other aesthetic considerations. There may also be concerns about the formation of microenvironments that may enclose volatile, potentially harmful materials within the glazed frame.


Most of the recent advances in cleaning are the result of research into acrylic dispersion paints. As noted earlier, acrylic dispersion paints are vulnerable to swelling caused by water, as well as by many polar and aromatic hydrocarbon–based organic solvents. Water, which is often the most efficacious general solvent for surface cleaning, can promote the swelling of paint films, the removal of migrated surfactants from the surface, and microlevel extraction of paint constituents, such as surfactants (used primarily in the acrylic medium) and pigment dispersants (used to help make a smooth paint with well-distributed pigment). Research over the last ten years or so has highlighted the many factors that can influence the amount and speed of this removal and/or extraction, which vary according to paint pigment type, age, quality, and thickness of the paint, as well as the cleaning application method, exposure time, and solvent type. It has therefore been difficult to identify many general trends, but we know that the risk of paint swelling and surfactant extraction decreases with paint age and that migrated surfactant on paint surfaces can degrade with light exposure alone.

Josefina López using stereomicroscopy

Substantial progress in research on the cleaning of acrylic painted surfaces has, in part, advanced through the use of methods that test large numbers of variables in short periods of time. The ongoing collaboration of the GCI, Tate, and the Dow Chemical Company is an example of this approach, although other researchers have adopted similar methods. Dow made available their range of high-throughput (HTP) analytical devices, which allowed the rapid screening of extraordinary numbers of potential cleaning solutions, where specific details on the effects of varying the percentage of each additive on the efficacy of the overall cleaning system could be fully explored. Part of the HTP setup included some instruments that monitored changes to paint gloss and color and some that could detect changes in flexibility. Others helped monitor the effects of the cleaning systems on migrated surfactant removal and detected any residues left at the surface after cleaning. The promising cleaning systems were then subjected to more relevant testing procedures, including hand-held swabbing, working directly with conservators to understand and evaluate some of the factors that could not be built into the HTP protocols.

One of the first results of this approach came from exploring the influence of pH on the cleaning ability of aqueous cleaning systems. Recent research by Richard Wolbers has furthered our understanding through exploring how adjusting the pH and conductivity of deionized water can help minimize the swelling and extraction potential of aqueous cleaning options when used alone or in combination with other solvents or gels.1

Other HTP research included the testing of combinations of additives such as surfactants and chelating agents, which increased the cleaning efficacy of both aqueous and nonpolar (in this case hydrocarbon-based) cleaning systems, resulting in the introduction of some new surfactants to the profession. More recently, a whole range of silicone solvents has also been introduced; they seem to offer effective cleaning power without causing undue paint swelling. Silicone solvents—which display extremely low polarity and surface tension and also have very low toxicity—were introduced to the conservation field by Wolbers in 2009. Since then, they have been increasingly used in conservation, notably for cleaning water-sensitive paint and other surfaces.2

Finally, a new class of cleaning systems has been developed and tested. Known as “water-in-oil” microemulsions, they combine the respective benefits of aqueous and nonpolar (hydrocarbon or silicone) solvent systems. These systems—in which water is dispersed within a nonpolar solvent and stabilized with surfactants—exploit the high cleaning power and adaptability of the aqueous environment and the relatively low swelling environment offered by nonpolar hydrocarbon or silicone solvents. At this point, four microemulsion cleaning series have been produced through our ongoing collaboration with Dow. Each iteration was designed for certain conservation situations and modified after feedback from trials performed by practicing conservators.3


A serendipitous outcome of the research into developing and evaluating cleaning systems for acrylic dispersion paints is that the same systems appear to be promising for removing deposited dirt and other unwanted materials, such as varnishes, from other water-sensitive surfaces, including many modern oil paintings (loosely defined as those from the twentieth and twenty-first centuries). Water sensitivity in modern oils is an intriguing issue; it is often hard to predict, and it can be exhibited by both thin and impasto paints, as well as on both glossy and matte surfaces.

Conservation image

Modern oil paintings are, in fact, beginning to present a whole range of challenging conservation issues in addition to water sensitivity. Regularly observed phenomena include the formation of soft, sticky, and vulnerable skins of oil medium on paint surfaces; surface efflorescence; insoluble thin surface crusts; solvent sensitivity; and the increasing frequency of paint regaining fluidity, possibly resulting in drips flowing down the painting surface.

Some of the factors contributing to these changes include the formation and migration of degraded oil components to paint surfaces and the use in artists’ oil paints of semi- or nondrying oils such as safflower. Interviews with artists’ paint manufacturers have helped identify additives (such as driers) that may contribute to water sensitivity directly, or indirectly through negatively influencing paint drying and aging processes. Finally, the current disuse of lead compounds, for environmental and health reasons, may also contribute to the instability and sensitivity of modern oil paints, which do not benefit from their stabilizing effects.

The current Cleaning Modern Oil Paints (CMOP) project—a collaboration of Tate, several other EU partners, and the GCI—and the newly formed Modern Oils Research Consortium (MORC) aim to explore several aspects of modern oil paints, including further characterization of the causes and mechanisms of paint deterioration that may lead to water sensitivity. Also to be studied are the effects of solvents on sensitive oil paints and the development of low-risk surface cleaning systems. These processes will be systematically evaluated through trials on test samples and, eventually, on case study works of art following the collaborative model successfully developed for acrylic dispersion paints.4


Exploring what may be regarded as “acceptable” change after conservation treatments such as surface cleaning is also gaining attention, especially at a time when analytical techniques have become so sensitive that they can detect changes in materials far below the threshold of detection by the human eye, prompting the “So what?” question to creep into discussions. Although change of any description, arguably at any scale, during a cleaning treatment is never desirable, the changes detected need to be placed in context and may be perhaps most usefully evaluated in a risk assessment. For example, if a trace of original material is removed from a painting during cleaning, but this removal is completely invisible to a viewer and does not adversely affect the flexibility of the paint, then can that change be tolerated if the overall cleaning of the work results in significant aesthetic improvement?

A Cleaning of Acrylic Painted Surfaces (CAPS) workshop exercise

More research, engagement, debate about research results, and discussion are clearly required among scientists, conservators, artists, collectors, and paint manufacturers to further advance this important area of investigation. At the same time, conservation treatment strategies need to be designed with awareness of the unique properties of each of these modern paints, knowledge of the likely aesthetic effects of cleaning, and informed understanding of the risks associated with each cleaning system.

Similarly, as we move forward with modern oil paint research, increased study of the sometimes alarming changes occurring with oil paints will lead to a better comprehension of the changes taking place with time, with exposure to the environment, and with treatment, which should lead to more sophisticated, more appropriate, and lower-risk treatment strategies.

The July 2016 CAPS workshop

A key element in making meaningful progress with research into cleaning treatments over the last ten years has been the importance placed on receiving and integrating input from experienced conservators throughout the process, in addition to the scientific testing and systematic evaluations by conservators involved in the research. This can be done at many different levels, including ongoing dialogue between scientists and conservators, through collaborative case study treatments, and through a range of continuous professional development workshops, which, for this research, have proven to be particularly beneficial. The Cleaning of Acrylic Painted Surfaces (CAPS) series of workshops offered by the GCI is one such endeavor, in which the most up-to-date research is disseminated to a small number of experienced conservators, and, where practical, feedback is garnered during the workshop itself. Such discussion of interim (and even negative) results has facilitated useful dialogues regarding the pros and cons of various systems, helped to identify areas requiring further research or modifications that could be made to improve cleaning, and, most significantly, helped to determine if the outcomes were proving useful.

Looking ahead, it is expected that the various active research groups will continue to suggest and develop useful new options for conservators for cleaning a whole range of modern paints, each of the options being systematically evaluated and modified after receiving feedback through trials and professional workshops and from conservators using these systems in the studio. While this type of research will never produce the ultimate cleaning system for any one type of modern paint, we have already reached the point where it is unlikely that conservators will complain about having too few options available to them—and that really is progress.

Bronwyn Ormsby is principal conservation scientist at Tate in London. Tom Learner is the head of Science at the GCI.

1. Courtney E. Dillon, Anthony F. Lagalante, and Richard C. Wolbers, “Acrylic Emulsion Paint Films: The Effect of Solution pH, Conductivity, and Ionic Strength on Film Swelling and Surfactant Removal,” Studies in Conservation 59, no. 1 (January 2014): 52–62.
2. Chris Stavroudis, “Silicone-Based Solvents in Conservation. As Free Solvents and Components of Gel Systems and Microemulsions,” Colore e Conservazione, November 13–14, 2015, Politecnico di Milano (Padova: CESMAR7/Il Prato, 2016): 176–84.
3. Bronwyn Ormsby, Melinda Keefe, Alan Phenix, Eleanor von Aderkas, Tom Learner, Christopher Tucker, and Christopher Kozak, “Mineral Spirits-Based Microemulsions: A Novel Cleaning System for Painted Surfaces,” Journal of the American Institute for Conservation 55, no. 1 (2016): 12–31.
4. Bronwyn A. Ormsby, Alexia Soldano, Melinda H. Keefe, Alan Phenix, and Tom Learner, “An Empirical Evaluation of a Range of Cleaning Agents for Removing Dirt from Artists’ Acrylic Emulsion Paints,” AIC Paintings Specialty Group Postprints 23 (Washington, DC: AIC, 2010): 77–87.