Component Three: An Evaluation of Cleaning Methods for Acrylic Paintings
A final and large-scale aspect of this research into modern paints involves an in-depth analysis of the effects of cleaning artists' acrylic emulsion media, by far the most common contemporary artists' paint. This component of the research is being mainly conducted by Bronwyn Ormsby, Leverhulme Fellow at Tate and Gregory Smith, Samuel L Golden Fellow at the NGA (between 2002-2004).

Cleaning—usually meaning the removal of surface dirt and/or varnish from the surface of a paint film—is arguably the most routine treatment carried out on painted surfaces. Although much is now known about the relative efficiency and safety of various cleaning techniques for traditional oil paints, an equivalent awareness of the short- or long-term effects on modern and contemporary works of art, especially those executed with synthetic paints, does not yet exist. Research into the cleaning of modern paints is therefore urgently required, in particular to find effective methods for removing surface dirt (the majority of modern paintings are not varnished) and to evaluate the possible consequential long-term damage to the paint film as a result of cleaning.

This component will focus on three main areas:

• the aesthetic effects of cleaning treatments on acrylic emulsion paintings;
• chemical analysis of the paints before and after cleaning treatments; and
• the resultant changes in physical properties for those films.

In addition, changes in the composition and properties of these paints as a result of aging and cleaning will be investigated by both Tate and the NGA. At the GCI, supplemental studies of solvent extracts will be carried out with the analytical methods employed in Component One for characterizing additives. It is intended that the results of these tests will then be used to guide actual treatments of acrylic paintings.

Work Completed

At Tate, over 600 samples of acrylic emulsions, both pigmented and clear, have been painted out onto a range of substrates, including canvas, primed canvas, and solid supports. A selection of these paint samples has then been subjected to accelerated ("artificial") aging, including both thermal and light aging regimes. A range of various wet and dry cleaning techniques have been applied to the dried films, and consideration of each cleaning treatment has included close examination of the surfaces of paint films before and after cleaning to assess gloss, color, and surface texture changes, as well as the measurement of changes in the films' physical properties, such as strength, hardness, and brittleness.

A range of analytical techniques have been assessed for potential monitoring of changes that might occur with cleaning treatments, many of which have been capable of detecting the presence of polyethylene oxide surfactants on the surfaces of acrylic films and their rapid removal with aqueous cleaning systems. Changes have also been measured in the thermal properties of acrylic emulsion films. Experimental work has focused on whether the glass transition temperature (T_g) of a paint film is altered by cleaning treatments (by either increasing softness or increasing rigidity). This has been measured by dynamic mechanical analysis (DMA) at the University of Exeter (UE) in the UK, and by differential scanning calorimetry (DSC) at the GCI. The results have indicated that although minor changes to T_g do occur with both aging and extreme cleaning treatments, such as immersion in water, the changes are generally minimal, with relatively little loss of film flexibility.

The DMA experiments have also been used to approximate the effect of swelling the paints when fully immersed in water. The results show that the paints become far more flexible when immersed in water, and can stretch to a significant degree when subjected to a constant end-to-end force. However, the extent of the response depends on a number of factors such as:

• the age of the film;
• the ambient temperature and relative humidity of the testing room and the paints themselves;
• the pigment type;
• the thickness of the film;
• the support (i.e., canvas);
• the brand/formulation of the paint;
• the evaporation rate of the solvent(s);
• the exposure time to the solvent(s); and,
• the type of solvent used.

Nonetheless, trends have appeared in the data—including the consistently more pronounced response of one brand of acrylic emulsion over the others.

The most useful finding confirmed that the exposure time of any of the water-based solvents is critical, in that the swelling/softening of these films increases substantially with increased exposure. It may be, therefore, that the paints may benefit from being dried quickly after treatment with water-based systems to prevent unnecessary swelling. The tests also confirmed that the more polar solvents, such as ethanol and water, have the greatest effect, while the predominantly hydrocarbon-based solvent, white spirits / Stoddard's solvent, had the least.

Changes to the chemistry of acrylic emulsion paint films after cleaning and aging has been primarily documented using FTIR, with an ATR attachment (at both Tate and the NGA). This attachment facilitates the gathering of data from the surface layer of the paint films, and hence reflects changes to surface chemistry only. The experimental work carried out thus far is in agreement with some initial findings that the surfactants are water-soluble and can be removed by swab rolling with water and by water immersion. The aging studies also confirmed that the evaporative action of both light (enhanced museum conditions) and thermal (heat) aging regimes resulted in either a decrease or the complete removal of the surfactants. This is supported by the DMA tests, which confirmed that only minor or negligible changes in film flexibility occur after the complete removal of the surfactant layer.

Building on the analytical protocols developed at the NGA, as described in Component One, the NGA has also initiated the examination of actual materials that can be extracted from acrylic emulsion paints by aqueous solutions and solvents.

Work in Progress

The investigations into potential changes in physical, optical, and chemical properties of acrylic emulsion paints are ongoing. Current work involves checking repeatability of some of the measurements, assessing a broader range of the paint samples, and interpreting the data that has been gathered so far.

Further assessments of analytical techniques are being carried out to monitor potential changes to paint films after cleaning treatments, including dynamic mechanical analysis (DMA) as a powerful tool to monitor changes in thermal properties, gas chromatography/mass spectrometry (GC-MS) and size-exclusion chromatography (SEC) to identify soluble extracts after cleaning and atomic force microscopy (AFM) and/or environmental scanning electronic microscopy (ESEM) to assess surface changes.