Innovative foam-based cleaning concepts for historical objects

1 Introduction

The historical surfaces of artistic and cultural assets can be regarded as the ˋfacesˋ of these pieces of art. The surfaces are often soiled as a result of either wear and tear or long-term exposure to environmental influences. In the long run these soilings can cause serious damage to the original surfaces, i.e. to the historical objects, and therefore have to be removed. The most important reasons for soilings are biogenic contaminations, for example microbial activity, as well as anthropogenic contaminations – especially since the beginning of industrialisation. These include soot, pesticides and oils. The latter were transferred to the surfaces through contact with skin or through care products. Cleaning these surfaces is quite a challenge because a tailor-made cleaning method needs to be developed for each individual surface during which mechanical impact such as rubbing must be avoided. Typically, mixtures of organic solvents are used that either dissolve the greasy or salty components of a soiling layer or soften the soiling layer and remove the insoluble components along with it [1]. However, solvent-based systems migrate into the surfaces and thus transport small portions of the dissolved soiling into the surface structure. The process is driven by capillary forces. In order to minimise this process, thickening agents are used, which gel the solvent [1]. However, residues of the thickening agents may remain on the surface after cleaning and removing them can cause additional problems. Furthermore, the use of solvents – regardless of whether they are used as pure solvent or as gel – represents an environmental pollution that must be prevented. The traditional cleaning methods for historical surfaces are usually used for very small areas to have a good control over the cleaning effects and the resulting side effects. Typically, small brushes or cotton swabs are used as tools for cleaning [2], which is very time- and material-consuming, especially for large works of art. Therefore, the time required to clean an art object of the size and surface complexity of a 19th century royal carriage can amount to approximately 1000 working hours. In order to reduce the use of organic solvents, aqueous solutions are also used for cleaning [2]. However, water-based cleaning agents can only be used to a limited extent, depending on the surface structure and composition, as they often wet and thus damage the surfaces. Any reduction of the amount of the used liquid improves the cleaning process considerably. Summing up, the disadvantages of the currently used cleaning techniques are (a) the use of organic solvents, (b) the remain of excess liquid on the surface, (c) difficulties of removing the gels/solvents from the objects, and (d) that they are very time- and material-consuming, i.e. very expensive.

Recent research showed that foamed detergents can clean far more efficiently than non-foamed ones [3], [4], [5], [6], [7], [8]. However, the mechanisms behind the cleaning process were not investigated and were thus not understood. This is where our research activities started. Our goal was – and still is – to develop new, foam-based cleaning methods. The use of foams for cleaning surfaces has a great potential for several reasons. Firstly, the amount of detergent can be reduced by up to 90% and the amount of water by up to 70%. Secondly, foams allow covering hard-to-access and even vertical surfaces. Thirdly, foams are easy to apply and easy to remove. Last but not least, foams generate additional physical cleaning mechanisms, namely imbibition, wiping, and drainage, which must contribute to the cleaning process, but how? In order to answer this question, we carried out an extensive study with model contaminations on glass surfaces. In one case, we used a fluorescent oil, which was sunflower oil coloured with the fluorescent dye Pigment Yellow 101 [3]. In a second case, we used a model contamination consisting of two components – a liquid and a solid. The model contamination was a combination of sunflower oil coloured with Sudan Red 7B dye and soot particles [4]. In both cases we succeeded in cleaning the surfaces with foams of a very special structure and we identified three cleaning mechanisms (see Section 3.1). Having successfully cleaned glass surfaces with model contaminations we were keen on doing the next step, i.e. cleaning the surfaces of real historical objects.

2 Experimental procedure

The foam used to clean the model and historic surfaces was produced with the double syringe technique of Gaillard et al. [9] to adjust the liquid fraction and ensure small initial bubble sizes with narrow size distribution. The composition of the cleaning solution with a technical grade surfactant and the foam generation process are described in Schad et al. [3, 4]. As model contaminated surfaces circular glass plates with a diameter of 8 cm and glass cuvettes (h = 35 mm, w = 35 mm, d = 32 mm) were used. Preparation of the contaminated model surfaces as well as cleaning-, imbibition-, and drainage-tests with the model surfaces are described in detail in [3, 4].

In order to clean historical surfaces, a foam with an average initial bubble size of r = (10–30) µm and a liquid fraction of ε = 5% was used. The bubble size was determined with an optical microscope (Reflecta Digital-Mikroskop 200 x). A spot of approximately 10 cm² was selected on the surface of the historical object. The freshly prepared foam was applied directly to the object and removed from the surface with a wet vacuum cleaner after a contact time of up to 5 min. Subsequently, the surface was gently dried with a cotton swab. To document the result, a photograph of the surface before and after cleaning was taken.

3 Results and discussion

3.1 Model surfaces

In our preliminary work we applied the cleaning foams on model glass surfaces contaminated with two model contaminations: (a) sunflower oil saturated with the fluorescent dye Pigment Yellow (Schad et al. [3]) and (b) Sudan Red 7B coloured sunflower oil with soot particles (Schad et al. [4]). By way of example, some results for the foam-based cleaning of model surfaces contaminated with fluorescent oil are shown and discussed in this Section.

Figure 1 demonstrates the cleaning effect of a foam with a liquid fraction of 5% which was applied on the top of the glass plate contaminated with 0.1 ml fluorescent sunflower oil. The cleaning process was observed under UV-light from below. The area A_oil of the fluorescent contamination on the glass surface decreases over time (Figure 1 bottom, left), while the foam bubble size increases due to foam decay (Figure 1 bottom, right). We studied the influence of liquid fraction in foam, the foam stability and the structure on the efficiency of the cleaning process considering the conditions required for the cleaning of historic surfaces (see Introduction). We identified three important mechanisms in the foam-based cleaning process: (1) imbibition, (2) wiping, and (3) drainage, which are schematically shown in Figure 2. Different combinations of the above-mentioned mechanisms play a significant role in cleaning process, depending on liquid fraction and foam stability.

Figure 1:

Results of foam-based cleaning of contaminated model surfaces. Foam with liquid fraction of 5% was applied on the top of the glass plate contaminated with 0.1 ml fluorescent sunflower oil. (top) Photographs of the cleaning tests under UV-light taken from below immediately after foam application and after 6 min, 18 min, and 30 min. (bottom, left) Time evolution of the contaminated area A_oil (in percent) during the cleaning of the glass plates. (bottom, right) Time evolution of the mean foam bubble radius <r>. The data in the bottom are taken from [3] and redrawn.

Figure 2:

Schematic drawing of the three mechanisms taking place in the cleaning process with foams. Modified from [3] with permission from Elsevier.

Mechanism I (imbibition, Figure 2, top) occurs most efficient at low liquid fractions and small bubble sizes: the contamination (e.g. oil) is sucked into the plateau borders of the foam due to capillary forces (the smaller the bubbles the stronger the capillary forces) [3, 4, 10], [11], [12]. The photographs shown in Figure 3 demonstrate the efficient imbibition in case of foam with 5% liquid fraction. In these experiments the foam was placed on top of the fluorescent oil layer in the glass cuvette and observed under UV-light. Over time the foam at the bottom of the cuvette started to fluoresce because of the imbibed oil. It should be noted that over long time the foam releases the contaminations due to foam decay and drainage [3, 4]. Therefore, the foam exposure time on the contaminated surface must be adjusted to the foam stability.

Figure 3:

View of the imbibition process from the side under UV-light. Cuvette (left) with fluorescent oil layer without foam and (center) completely filled with foam (with 5% liquid fraction) 8 min after foam application. (right) Zoomed photograph shows the foam with imbibed fluorescent oil at the bottom of foam layer.

Mechanism II (wiping, Figure 2, middle), i.e. shifting of the contact line between oil, foam and solid surface, takes place during the foam-based cleaning process independent of the liquid fraction [3, 4]. This movement is caused by the instability of the foam, in other words the foam decay, which causes the bubbles on the surface to be in constant motion. This creates a wiping movement on the surface, if the foam is sufficiently unstable. Wiping is important to reduce the adhesion of contaminations and thus detach them from the surface, i.e. to easily remove the contaminations. We showed in our experiments with a two-component model contamination (sunflower oil and soot) that this mechanism plays the crucial role in the foam-based cleaning: without wiping (by using extremely stable perfluorohexane-containing foams with different liquid fractions) no significant cleaning was achieved [4].

Mechanism III (drainage, Figure 2, bottom) is relevant at high liquid fractions: the liquid drains out of the foam underneath the contamination and lifts it from the surface [3, 4]. The photographs in Figure 4 illustrate this cleaning mechanism. In these experiments the foam with 20% liquid fraction was placed on top of the fluorescent oil layer in the glass cuvette and observed under UV-light. Several minutes after foam application the fluorescent oil swims on the top of the cleaning solution, which drains out the foam (Figure 4). It must be noted that significant drainage of the cleaning solution should be prevented in the case of sensitive historical surfaces. For this reason, foams with high liquid fractions were not applied to the art objects although they showed some good results in the cleaning of the model surfaces [3].

Figure 4:

View of the drainage process from the side under UV-light. Cuvette (left) with fluorescent oil layer without foam and (center) completely filled with foam (with 20% liquid fraction) 8 min after foam application. (right) Zoomed photograph shows that the fluorescent oil swims on the top of the cleaning solution, which drains out the foam.

Summing up the results for the model contaminated surfaces one can conclude that for efficient foam-based cleaning of historical objects (a) foams with low liquid fractions and small initial foam bubbles should be applied to benefit from the imbibition mechanism; (b) relative instable foams should be used to take advantage of the wiping mechanism (foams should be stable enough to sit on the surface, but decay over time and change their structure during the cleaning to ensure wiping); (c) foams with liquid fractions above 10% should be excluded for the cleaning of historical surfaces because of drainage within a short time interval. Consequently, the combination of imbibition and wiping is the best choice for an efficient foam-based cleaning of historical surfaces, therefore foams with 5% liquid fraction, small bubble sizes, and intermediate stability should be used.

3.2 Historical surfaces

In this study we performed cleaning experiments using foams on real historical objects. These objects are historical vehicles of former kings and emperors. The vehicles are exhibited in the Marstallmuseum in Castle Nymphenburg in Bavaria. We investigated the foam-based cleaning concept previously developed in Schad et al. [3, 4] and how the foam affects the historical surfaces. In addition, we were interested whether the discovered cleaning mechanisms of imbibition and wiping are also effective on historical objects, i.e. whether they can remove the dirt that had accumulated on the objects over the years.

3.2.1 Small fragments

To test the cleaning foams on historical objects, they were first tested on some fragments of historical objects that could not be identified, including an old gilded frame and a replica of a gilded leaf rosette. The photographs of the cleaning tests are shown in the following figures (Figures 5 and 6). The purpose of these tests was to determine how long the foam can remain on the object in order to clean most efficiently without “flowing” from the object and without drainage on the object. In addition, the best method for removing the foam should be identified. It was recognized that a cleaning time of 5 min was sufficient to clean the historical surfaces [13]. For the cleaning tests on the model surfaces (which were glass plates), a much longer exposure time was needed to clean efficiently. Note that we did not study the reason for this difference. However, it is reasonable to ascribe it to the different types of surfaces, which, in turn, change adhesion and wetting properties of the foam on the surface. Thus, the cleaning of the historical surfaces is much faster, which results in additional time savings [13].

Figure 5:

Photos of the cleaning test on a gold-plated leaf frame (left) before cleaning, (center) during cleaning with foam, and (right) after cleaning. Photos: Tamara Schad, Bavarian Palace Administration.

Figure 6:

Photos of the cleaning test on a replica of a gold-plated leaf rosette (left) before cleaning, (center) during cleaning with foam, and (right) after cleaning [13]. Photos: Tamara Schad, Bavarian Palace Administration.

The gold-plated frame has a relatively smooth surface without large ornaments (Figure 5). It can be seen that the area covered with the foam is clean, shiny and the gilding is not damaged. In addition, there are no foam residues. For the removal of the foam, the use of a cotton swab and a wet vacuum cleaner was tested. It was found that the foam can only be removed with a cotton swab in a very effort and time-intensive manner, whereas the complete foam can be removed almost residue-free with the wet vacuum cleaner. In some small areas, the cotton swab had to be used to remove small residues of the foam.

The surface of the rosette is more structured and has more ornaments in which the dirt is collected. Figure 6 shows that a significant fraction of the dirt could be removed from the area and no foam residues are left after the foam has been removed by vacuuming. The gilding of the surface was not damaged during cleaning.

3.2.2 Historical vehicles

The next step was to clean the historic surfaces of the carriages in Nymphenburg Palace in the Marstallmuseum. First, the cleaning methods currently used by restorers were compared with foam-based cleaning [13]. One traditional method is to clean the objects with cotton swabs. These are dipped in the cleaning solution and are then carefully rolled over the surface to be cleaned, which is very time-consuming. The second traditional method used for cleaning is to place a compress soaked with the cleaning solution on the surface. The compress is placed on the surface for about 5 min and then carefully removed. After cleaning, the surface is re-cleaned in both cases with ethanol and water to remove the cleaning solution [13]. The object is then considered to be clean, although the crevices and ornaments of carved wood cannot be properly cleaned using these methods. Approximately 5 ml of cleaning solution was required to clean an area of ∼ (10–15) cm² with each of the two cleaning methods. For comparison, the cleaning was performed with a foam. For an area of the same size, 20 ml of foam (5% liquid fraction) containing 1 ml of cleaning solution was used. The foam was applied to the surface and removed by vacuuming after 5 min. The foam shows a very good cleaning effect and can also clean the ornaments and cracks, i.e. the same – or even better – results can be achieved with foam-based cleaning compared to conventional cleaning methods in a shorter time and with less cleaning solution.

Figure 7 shows the cleaning of a soiled gilded feather on the historical Dress Chariot of Prince Eugène Beauharnais (Figure 8) with the foam. The foam was removed from the surface after 5 min. The feather is cleaned and no surface damage to the historical materials is visible.

Figure 7:

Historical carved and gilded feather on the historical Dress Chariot of Prince Eugène Beauharnais (left) before cleaning, (center) during cleaning with foam, and (right) after cleaning. Photos: Heinrich Piening, Bavarian Palace Administration.

Figure 8:

Historical Dress Chariot of Prince Eugène Beauharnais, Munich 1819/20, altered 1856 by Michael Mayer, Marstallmuseum Nymphenburg Palace, Munich, WAF.

After the successful cleaning of several other bigger parts of the historic vehicles, a carriage was selected which was to be completely cleaned with the foam [13]. The coronation carriage of Emperor Charles VII (Figure 9), which is to be restored, was chosen for two reasons: (1) it has a robust surface that is suitable for outdoor use and (2) it has a large size (6.85 m × 2.13 m × 3.05 m).

Figure 9:

Photo of the coronation carriage (year of construction 1721) of emperor Karl VII of the Holy Roman Empire [13]. Photo: Tamara Schad, Bavarian Palace Administration.

Figure 10 shows the cleaning of a mythical creature on the door of the carriage. With the foam it was possible to clean the complete carriage without damaging it. Moreover, the opulent carriage could be cleaned in approximately 340 working hours (including preparation and introduction) instead of 900–1000 working hours needed with conventional cleaning methods. In addition, the amount of surfactant solution used was reduced by about 90% compared to conventional methods. Thus, an efficient, time-saving, and surface-friendly method was found to clean historic vehicles while saving a large amount of cleaning agent.

Figure 10:

Historical carved and gilded mythical creature on the door of the coronation carriage (year of construction 1721) of emperor Charles VII of the Holy Roman Empire (left) before cleaning, (center) during cleaning with foam, and (right) after cleaning [13]. Photos: Tamara Schad, Bavarian Palace Administration.

4 Conclusions

We demonstrate for the first time that liquid foams can clean sensitive surfaces in a very gentle way. Note that the mechanisms which cause the cleaning effect have been unknown up until only recently when they were revealed by us [3]. The cleaning process has a high application potential especially for cleaning sensitive surfaces because (a) one can control the properties of the foam (adjustable liquid fraction and tuneable bubble size) such that the surfactant solution does not flow on the surface during cleaning, (b) foam residues can be easily removed without significant mechanical stress, (c) there is a low need for post-cleaning, and (d) the selected surfactant type is biobased and biodegradable. We showed that the time required for cleaning objects can be reduced by (50–60)% compared to conventional cleaning methods, in which brushes and cotton swabs are used. A time-saving cleaning method is of particular interest for the cleaning of large historical objects. In addition to a reduction of the working time, the amount of the cleaning solution could also be reduced by approximately 90%: instead of approximately (60–70) L only 6.5 L were used for the cleaning of the carriage of Emperor Charles VII of the Holy Roman Empire. In conclusion, with the new foam-based cleaning process sensitive surfaces can be cleaned and both working time and resources can be significantly reduced.