Open AccessArticle

Er:YAG Laser Cleaning of Micromosaics from the Rosalinde and Arthur Gilbert Collection at the Victoria and Albert Museum

Julia Brand

Carmen Vida

and

Lucia Burgio

Conservation Department, Victoria and Albert Museum, Cromwell Road, London SW7 2RL, UK

Author to whom correspondence should be addressed.

Heritage 2024, 7(12), 7309-7324; https://doi.org/10.3390/heritage7120338

Submission received: 20 November 2024 / Revised: 5 December 2024 / Accepted: 13 December 2024 / Published: 23 December 2024

(This article belongs to the Special Issue The Conservation of Glass in Heritage Science)

Download

Browse Figures

Figure 1
Cross-section of micromosaic with base or cassina material (stone, metal), putty paste in which glass filati are embedded, grouting around the filati, and several layers of wax on the surface. "> Figure 2
St. Peter’s Square bonbonnière (accession number LOAN:GILBERT.924-2008). (a) Overview of the bonbonniere with broken corner; (b) close-up on the broken corner showing the structure of the micromosaic with filati inserted in the putty; (c) broken fragment used for the trials. The structure of the micromosaic is clearly visible in the cross section, with the underlying putty into which the glass filati are inserted, and the grouting in between the filati. "> Figure 3
Before and after irradiation test on the fragment of the St. Peter’s Square bonbonnière. (a) Before; (b) after 3 laser shots at 0.56 J·cm−2, alteration circled in red; (c) close-up on the altered surface; (d) SEM image of the altered surface showing melting; (e) cross-section of the filati before laser irradiation; (f) cross-section of the filati after irradiation at 0.56 J·cm−2 showing no change; (g) surface of the filati before laser irradiation; (h) surface of the filati after laser irradiation at 0.56 J·cm−2 showing no change; (i) surface of the stone before laser irradiation; (j) surface of the stone after laser irradiation at 0.56 J·cm−2 showing no change; (k) surface of the grouting before laser irradiation; (l) surface of the grouting after laser irradiation at 0.56 J·cm−2 showing no change (slightly different spot, no changes in texture, inclusions). "> Figure 4
Before and after cleaning test on the bonbonnière. (a) Before laser and solvent cleaning; (b) after softening with the laser at 0.56 J·cm−2 and swabbing with ethyl acetate. The red frame indicates cleaned area; (c) before solvent cleaning; and (d) after cleaning with ethyl acetate only. "> Figure 5
Microscopic images of the glass filati before and after treatment. (a) White filati of the sky before treatment; (b) white tiles after laser and solvent cleaning showing excellent removal of dirt and preservation of the tinted wax and no alteration of the filati or grouting; (c) white filati (different area) after solvent cleaning only showing good removal of dirt but dirt remaining at the corners. "> Figure 6
The Beautiful Sky of Italy tabletop (accession number LOAN:GILBERT.894-2008) before treatment, showing dirt in the sky and concentric black cracks with losses. "> Figure 7
Using Evolon® to pre-wet the surface prior to irradiation. (a) The Evolon® following application. Note the yellow colour immediately picked up by the textile; (b) the three stages of cleaning: 1—Evolon® applied to the surface; 2—after removal of the Evolon® but before the laser; 3—after laser irradiation and swabbing. "> Figure 8
Images before and after treatment of the tabletop. (a) Overview before treatment; (b) overview after treatment showing good removal of dirt and brightening; (c) before treatment of the Milan and Genova sections; (d) after treatment of the Milan and Genova sections showing excellent removal of dirt and yellowing and brightening of colours; (e) before treatment, zoom on the Milan section; (f) after treatment showing better impression of 3D and brightening of colours. "> Figure 9
Microscopic images of the surface of the sky before and after cleaning. (a) Before cleaning, showing significant amount of dirt between filati; (b) after cleaning, showing reduction of the dirt, removal of the yellowing, and slight improvement around the crack; (c) before and after treatment of blue filati, showing good removal of dirt and preservation of the blue tinted wax; (d) before and after treatment of blue filati in a different area, showing good removal of dirt and preservation of the blue tinted wax. "> Figure 10
Microscopic images of the surface after treatment, checking for laser-induced damage; (a) close-up on the Milan duomo showing no alteration to the grouting or filati; (b) potential laser-induced damage on the dark green tower separating the Milan and Venice sections, showing possible laser damage circled in white dotted lines. "> Figure 11
Cleaning The Flora of Two Sicilies tabletop (LOAN:GILBERT.190-2008). (a) Overview of the tabletop; (b) a detail before cleaning; and (c) after cleaning. "> Figure 12
Successful examples of laser-assisted cleaning. (a) Bonbonnière depicting a mountain (LOAN:GILBERT.490-2008); (b) before and after cleaning image showing the difference between clean and dirty surface. (c) Bonbonnière depicting a goldfinch on a branch before cleaning (LOAN:GILBERT.203-2008) and (d) after cleaning; (e) microscopic image before treatment showing dirt between the filati and (f) after treatment showing reduction of dirt and improvement in the appearance of the dark cracks. ">

Versions Notes

Abstract

Conserving micromosaics from the Gilbert collection at the Victoria and Albert Museum is a challenge due to their complex structure and fragility. Their highly polished decorative surface prevents access to the substrate, yet deterioration can affect both, and manifests as cracking, failure and losses of the substrate and grouting, and as yellowing and accumulation of dirt on the uppermost organic coating. In order to minimise any possible damage to the substrate while cleaning the surface with solvents, laser cleaning using an Er:YAG laser was investigated on various micromosaic objects. Tests were first conducted on a non-displayable bonbonnière, and digital microscopy and scanning electron microscopy were performed to investigate the effects of the laser on the different materials; the results were then compared to solvent-cleaning only. The combination leading to the most effective cleaning results was found to be laser irradiation at fluences up to 0.71 J·cm⁻² on the surface, followed by gentle swabbing with solvent. The surface was successfully cleaned with no changes induced on the materials, and a much lower amount of solvent was used. Following these successful preliminary tests, cleaning was undertaken on other pieces of the collection, and the outcome is presented in this study. The results show a great variability in terms of response of the materials to solvents, emphasising the need to consider each item on a case-by-case basis.

Keywords:

laser cleaning; Er:YAG laser; micromosaic; conservation; cleaning; Rosalinde and Arthur Gilbert Collection

1. Introduction

The Rosalinde and Arthur Gilbert Collection, held at the Victoria and Albert Museum (V&A) and displayed in the eponymous dedicated galleries, is one of the most comprehensive collections of European decorative arts. Conservation of its nearly 1200 items is taking place before the refurbished galleries reopen in 2026. The collection includes gold and silver objects, enamel miniatures, gold boxes, stone mosaics, and one of the largest micromosaic collections in existence, with over 240 objects [1].

Micromosaics are highly decorative, layered, composite objects of extraordinary workmanship, valued for their visual appeal. Their manufacture technique is well understood [2,3], but the exact materials used and the variability that may exist from one object to another is not well known. The highly polished surface precludes access to the substrate, yet deterioration can affect both, and manifests as cracking, failure and loss of the substrate and grouting, and as yellowing and accumulation of dirt on the topmost organic coating. Many objects also present a complex history of interventions throughout time in the form of waxing, repairs, and restorations.

Micromosaic plaques were initially made as sophisticated souvenirs, set into mementos such as jewellery, pictures, boxes, or even larger furniture items such as tables, which were created as statement pieces to impress tourists completing their Grand Tour [4]. Unlike paintings or prints, the colours of micromosaics would never fade or change over time, ‘making them the perfect souvenir from Rome, the Eternal city’ [4].

Unlike other mosaics, which are made by cementing the individually cut tesserae into a mortar or adhesive surface, micromosaics are made by inserting minute millimetre-thin rods of coloured glass (called smalti filati or simply filati) into a mastic or putty made up of linseed oil, slaked lime, waxes, and other materials (see Figure 1). The putty is applied in small quantities onto a tray-like base called cassina made of copper, iron, stone or glass depending on the object. The filati are made by hand, so that their colour and shape are unique and cannot be reproduced exactly. Each filato is painstakingly inserted in the putty with tweezers to form a pattern, scenery, object, or even portrait. Once the design is complete, hot wax is poured to grout the filati and fill the interstices between them. The micromosaic is then polished repeatedly with increasing grades of grit and more layers of wax, to level all the individual rods into a single smooth surface. The upper layers of the wax are often coloured with ground glass that is painted in with a brush to match the colour of the area, before the whole surface is polished to a high shine [5].

In contrast to our knowledge of the overall manufacturing technique, little is known about the specific materials used for each of these valuable items other than very generically, since the recipes were kept secret and passed down generations of makers. Initial analysis conducted on a tabletop at the V&A in 2018 revealed that the thin glass filati contain Si, K, Ca and Pb [2], with metal oxides giving the colour (Pb, Cu, As, Fe, Sb, Mn, Co). Other additives such as As and Sb, or As and Sn, could be found as opacifying agents. According to historical written accounts [2,3], the putty is made of boiled linseed oil, powdered slaked lime, travertine, and sometimes colophony. The surface is covered by a coating, often beeswax tinted with smalti powder (powdered coloured glass). Nevertheless, the materials used to make the micromosaics may have varied between different workshops, and can also vary from one piece to another, making it complex to predict the interactions during conservation treatments (with lasers or solvents). The full scientific analysis of the micromosaics is rarely possible, as sampling would be too destructive to be permitted. Therefore, throughout this article, the more generic term ‘organic coating’ is used, except when specifically referring to waxes found as grouting or to the coloured waxes found at the top of the grouting between filati.

The Gilbert micromosaics appear to be, overall, in fair structural condition. As expected, the oils in the substrate have oxidised and cross-linked over time, creating cracks and losses, but only a few of the objects exhibit bulging or areas that sound hollow when tapped indicating substrate loss. Visually, however, yellowing of the final coating and dirt accumulation are frequent, as are material additions in the form of restorations, fills or waxes.

Conserving these objects is a challenge due to their complex structure and fragility, and to the difficulty in establishing what the specific materials present in each object may be and how they may react to the conservation treatment. Intense mechanical action during cleaning can generate losses of the grouting and filati and is time-consuming. Moreover, both the putty and the grouting appear to be sensitive to most solvents [2], making traditional cleaning difficult. The organic coating and the grouting can have similar solubilities; therefore, finding a suitable solvent for the coating that will not affect the grouting is challenging. Consequently, finding an alternative cleaning method, gentler and with reduced use of solvents, has extremely significant implications for the collection.

In a collaboration between the V&A Sculpture Conservation and Conservation Science sections, laser cleaning was investigated as a potential solution on various micromosaic items: bonbonnières, plaques, and tabletops. Laser cleaning is now a well-established method for the cleaning of various heritage items and addresses significant challenges in the preservation of intricate artworks such as micromosaics: it minimises chemical and physical risks associated with conventional cleaning methods, provides better control, and enhances the efficiency of dirt removal.

Previous research on laser cleaning of glass has focused on the removal of corrosion crusts, organic coatings (polymers) and biolayers from historic stained glass using different lasers (excimer or Nd:YAG) [6,7,8,9]. Laser cleaning of glass was reported as unsuitable with the 1064 nm wavelength of the Nd:YAG but feasible with UV wavelengths [7,9], provided special care was taken to lower the risk of overcleaning due to the similarity in ablation threshold of the glass and contaminants [7,8,9]. It was also reported that the increase in temperature during laser ablation could generate thermal stresses on the glass surface, which could potentially lead to cracking [10].

Removal of beeswax was studied by Pan et al. on granite surfaces at various wavelengths and it was reported that the 1064 nm wavelength of the Nd:YAG was not viable as it damaged the granite, the 532 nm harmonic was inefficient, and UV harmonics appeared appropriate [11,12].

Nd:YAG lasers remain the most commonly used lasers in conservation, particularly for the cleaning of stony materials [13]. Er:YAG lasers have received less attention, despite offering promising results on a range of materials, especially in the cases of otherwise problematic surfaces for which conventional cleaning methods were found ineffective or unacceptable [14]. Er:YAG lasers operate very differently to Nd:YAG lasers. Generally, Q-switched Nd:YAGs cause material to be explosively ablated and ejected from the surface during cleaning [14], whereas Er:YAGs rely on near-surface thermal effects and absorption of the mid-IR wavelength by OH bonds to disrupt and weaken the unwanted material, providing a gentler interaction. Therefore, an Er:YAG laser is generally not considered as a standalone cleaning tool but is used in combination with other cleaning methods [15].

In this study, the Er:YAG laser was selected, emitting at 2940 nm, with 100 µs pulse duration, and various pulse diameters depending on the size of the objects and the fluence required. Tests were first conducted on a non-displayable bonbonnière, and digital and scanning electron microscopy were used to investigate the effects of the laser on the different materials of the object; the results were then compared to solvent-cleaning only. Following these successful preliminary tests, cleaning was undertaken on other pieces of the collection, and the results are presented in this study.

2. Methods

2.1. Methodology

In response to the conservation challenges associated with micromosaics, and with the ambition to create a cleaning protocol applicable to different items of the Gilbert collection, initial tests were conducted on a non-displayable item with a broken fragment that gave access to all the different layers of the object, including the putty, the grouting, the filati, and the wax layers. Solvents and laser cleaning were combined with digital microscopy and scanning electron microscopy to assess not only visible but also invisible changes induced on the materials of the item and shape the development of the cleaning method.

The collaboration of conservation practitioners and scientists ensures the best preservation of the objects. By combining the intimate understanding of the historical and cultural significance of the objects and their materiality with the expertise in the physical and chemical properties of the materials, actionable insights relevant to conservation practice were generated.

First, it must be stated that laser cleaning is not a universal solution but can be a viable one when conventional cleaning methods are too challenging. The laser was used here in combination with solvents to clean the dirty and yellowed superficial coatings. Consequently, a preliminary step consisted of solubility tests to find an acceptable solvent. Conventional wet cleaning was evaluated and, if acceptable results were obtained (removal of dirt without removal of coloured wax and without the need for intense mechanical action), conventional cleaning was undertaken without laser. If cleaning seemed aggressive or the surface response was too variable and unpredictable, then an initial laser exposure was investigated to check if it provided a viable alternative. Laser irradiation was then undertaken on an inconspicuous part of the object or on a fragment, if available.

2.2. Laser Parameters

Laser irradiation was performed using an Er:YAG laser Fotona Fidelis XS emitting at 2940 nm and equipped with an articulated arm and pen delivery system. The Very Short Pulse duration was selected (100 μs). The repetition rate was set to 2 Hz for testing, then 5 Hz for cleaning. The pulse diameter varied, either 3 mm or 5 mm depending on the size of the area selected. Laser exposure was tested at low fluence first (0.56 J·cm⁻²). Depending on the OH content in the coating, the response could vary from no change at all to subtle softening. The number of laser passes was determined empirically by testing, trying to minimise the number of passes while obtaining satisfactory cleaning results afterwards. A couple of passes were generally applied on the surface. In some cases, the lower 0.56 J·cm⁻² was sufficient to provide adequate cleaning, but if no effect was noted, the fluence was increased slightly (0.71–0.86 J·cm⁻²). The collimated beam of the Fotona laser ensures the fluence is relatively constant even when the distance between the surface and the laser handpiece is changed.

The laser was applied first, followed by swabbing with a solvent. Alternatively, the surface was pre-wetted with a small amount of solvent, before applying the laser, followed by swabbing with solvent to clear the disrupted material. Pre-wetting was carried out as it was reported in the literature to have beneficial impacts on the cleaning, increasing the effectiveness of treatment and keeping the surface temperature lower, providing a safer process overall [13].

2.3. Choice of Solvent

The use of the Er:YAG laser was tested in combination with a variety of solvents to identify the best pairing and method of application. In previous research into the degradation process of linseed oil, one of the main components of the putty [2], Sonntag identified several solvents that can aid saponification of the oils and cause their swelling or solubilisation, and, therefore, should be avoided. Water, industrial methylated spirits (IMS), acetone and ammonia were all reported to cause swelling after some time, and methyl ethyl ketone (MEK), dimethylformamide, and dimethyl sulfoxide were found to solubilise the dried oil. On the other hand, ethyl acetate and mineral spirits did not seem to interact with the dried linseed oil [2]. For this reason, cleaning tests initially focused on ethyl acetate. Mineral spirits (or white spirit) would also have been safe for the putty, but it was deemed unsuitable for cleaning the surfaces because the waxes in both the top coatings and in the grouting between filati could have become easily solubilised by this solvent. As an alternative, benzyl alcohol (C₇H₈O) was tested. Benzyl alcohol is an aromatic alcohol that combines polarity with a benzene ring in its molecular structure. This makes it likely to not affect the waxes yet allows for some level of cleaning. Benzyl alcohol was tested on a small putty sample and did not affect or solubilise it after 48 h. It was, therefore, used as the second solvent of choice in cleaning tests.

Prior to conducting any cleaning, solvent tests were conducted on each object to ensure the removal of the dirt and any yellowed coating, whilst preserving the coloured waxes in the grouting and materials underneath.

In all cases, tests started with ethyl acetate as the preferred choice, followed by benzyl alcohol. In some cases, additional tests were required to find another solvent mixture through the Modular Cleaning Program (MCP) [16], with benzyl alcohol as the solvent element of the solution. The MCP is both a database system and an approach for the cleaning of objects that was developed to assist conservators in their approach to cleaning with water-borne systems, solvents, solvent gels, emulsions and microemulsions [16].

A second area of testing focused on the mode of application and use of solvents in combination with laser irradiation. The main purpose of these trials was to find the most effective cleaning method while reducing as much as possible the amount of solvent and mechanical action. Several modes were investigated:

Laser irradiation followed by gentle swabbing of the surface with the chosen solvent.
Pre-wetting of the surface with a swab and solvent, followed by laser irradiation and then swabbing of the surface with the same solvent.
Pre-wetting of the surface with a piece of Evolon^® CR textile from Deffner & Johann GmbH (Röthlein, Germany) dampened with solvent, followed by laser irradiation and then by swabbing the surface with solvent. Evolon^® CR textile has been increasingly used for the removal of varnish from oil paintings and offers the benefit of limiting solvent exposure and its penetration into substrate layers, and of reducing mechanical action [17].

Both the choice of solvent and the method of application were different from object to object. In some cases, several methods of application were used on the same object to maximise results or in response to the different condition of areas within the object.

2.4. Analysis

Examination was carried out before and after cleaning using digital microscopy with a HIROX HRX-01 digital benchtop microscope, with the wide objective lens at 20× magnification, and mid objective lens at 140× magnification (and up to 600×) for finer details. Digital microscopy was undertaken for objects small enough to fit on the stage of the microscope (boxes, plaques). Some microscopy was undertaken on larger objects using a flexible arm HIROX HRX-01 digital microscope with a wide objective lens at 20× and 50× magnification, and a diffuse reflectance attachment as the surfaces were reflective.

Scanning electron microscopy (SEM) was undertaken during preliminary testing with a Hitachi TM4000Plus tabletop SEM, in low vacuum without coating, back-scattered electron mode (BSE) and 15 kV voltage in spot 3, with 100× to 500× magnification. The working distance was 8–12 mm. The Shadow 2 image preset was selected. Energy-dispersive X-ray spectroscopy (EDX) spectra were acquired using the Oxford AZtecOne software version 4.3, and 20 kV voltage.

3. Results

3.1. Preliminary Tests on Non-Displayable Bonbonnière

3.1.1. Initial Laser Irradiation Test

To assess the suitability of laser cleaning on micromosaics, tests were performed on a non-displayable bonbonnière, broken in a corner with a detached fragment on which laser irradiation could be investigated (see Figure 2). Preliminary analysis of the bonbonnière fragment was carried out with digital and scanning electron microscopy. Then the laser was tested on the fragment and the bonbonnière. Finally, microscopy was carried out again to compare the surfaces before and after and check for damage.

Three laser shots were fired at 0.56 J·cm⁻² on the side of the fragment, where the putty, the stone, and the filati are visible. This was enough to cause changes in the putty, which instantly appeared discoloured and altered (see Figure 3b,c). Melting was suspected and confirmed by SEM (see Figure 3d). The melted areas show small bubble-like patterns at the centre of the irradiated area. The filati and stone appear unaltered and unchanged: no change in surface roughness, no signs of ablation in Figure 3e,f, and no laser-induced microcracks. Their damage threshold is likely to be higher than the investigated range of fluences.

Additional shots were fired on the top of the fragment at 0.56 J·cm⁻², where the surface is covered by the organic coating. In this case, the filati were unaltered (Figure 3g,h). Note that the two images appear different from each other due to different contrasts, but no sign of damage is visible. The stone appeared unaltered as well (Figure 3i,j). The grouting between the filati was thoroughly checked all over the sample and no sign of damage could be observed anywhere—same texture, similar amount of inclusions, no bubbling, no melting (Figure 3k,l). This was a promising sign for laser cleaning.

It is important to note here that it is unlikely that we would encounter an area with a portion of exposed putty on the surface of the micromosaics, as the surface is covered by a coating. The situation encountered on the fragment, therefore, presented a unique opportunity for research. As the filati and stone were unaltered, laser cleaning was still a viable option. However, care should be taken to avoid exposing the putty to the laser (which means additional care should be taken on areas with pre-existing damage).

3.1.2. Cleaning Trials

Following the initial tests conducted on the fragment, cleaning was tested on the bonbonnière, in the right corner of the sky just above the building, on the white filati, in an area exhibiting dirt between the tiles. The fluence was first lowered to 0.20 J·cm⁻² with a 5-mm spot diameter. No effect was observed at all after several laser passes. It was then decided to increase the fluence back to 0.56 J·cm⁻² with the 3 mm spot diameter. Several passes were applied, and no alteration was observed with the naked eye. Swabbing with ethyl acetate followed and seemed to improve the appearance of the area by removing some of the dirt. It also seemed that the yellowish tinge of the coating was removed, bringing back the vivid colours of the surface. The results are presented in Figure 4.

A second test was undertaken just above the first one at 0.71 J·cm⁻² with the 3 mm spot diameter. No alteration was observed after three passes with the laser. Swabbing was then undertaken with ethyl acetate and gently rubbing the area. Dirt was noticeably removed by the laser from the area, which appeared cleaner immediately, and the solvent removed additional dirt. The best results were obtained with 0.71 J·cm⁻²; 0.86 J·cm⁻² was also trialled but did not improve the results so it was decided to continue at 0.71 J·cm⁻².

These results were compared with those obtained by cleaning with solvents only (Figure 4c,d). This was carried out to assess if and to what extent the removal was facilitated by the laser, or if it was caused by the solvent. Solvent cleaning was undertaken on the left side of the bonbonnière in the sky just above the architectural details, in an area affected by dirt. Solvent-only cleaning required a higher quantity of solvent, and more rubbing compared to the combination of laser and solvent. After cleaning, the results appeared similarly satisfactory to the naked eye. They were then compared under the microscope.

Digital microscopy was carried out to analyse the effectiveness of treatment and check for any damage. Figure 5 shows white filati in the sky before and after treatment with the laser and solvent swabbing, compared to solvent cleaning only. In both cases, dirt was removed between the tiles, and the grouting is cleaner and whiter. The yellow appearance was also removed. No difference was observed for the filati or the grouting (no melting, no discoloration, no heat-related effects, no differences in relief).

With solvent only, dirt residues were still visible on the edges of the tiles, compared to cleaning with the laser + solvent where the tiles’ edges were clean. This shows that the laser leads to a more homogeneous cleaning compared to solvent only.

An area of the surface with red tiles was tested for potential discoloration, as iron-rich (and copper-rich) pigments can react strongly to laser radiation and can discolour dramatically. When checked before and after irradiation with the laser only (no solvent), the tiles did not discolour, and they appeared unaltered. It is important to note that it will be necessary to test different colours on each object as the composition of the pigments is likely to vary from one object to another, and this object being safe to clean does not mean that all objects will be equally safe.

To summarise, no damage was observed with digital microscopy (and SEM where possible) in all the areas tested on the bonbonnière. Good cleaning effectiveness was observed for the combination of laser at 0.71 J·cm⁻² and ethyl acetate swabbing. Solvent cleaning alone was shown to be less effective than pre-wetting followed by laser cleaning. Care should be taken when undertaking any work with the laser. Surfaces on the same object may react differently, and similar surfaces from different objects may react differently too.

After these successful tests, cleaning was undertaken on various objects from the collection following the procedure described in the previous sections.

3.2. Cleaning the Beautiful Sky of Italy Tabletop

3.2.1. Introduction and Cleaning Trials

The table, presented in Figure 6, commissioned by Francis Needham, Earl Kilmorey and made by Michelangelo Barberi in 1845, is a circular micromosaic tabletop depicting a central group of four angels in the sky, with the attributes of painting, music, architecture, and sculpture [18]. Around the circumference are representations of eight Italian cities: Piazza del Duomo, Milan; St. Mark’s Square, Venice; Piazza della Signoria, Florence; St. Peter’s Square, Rome; the Colosseum, Rome; the Riviera di Chiaia, Naples; the Cathedral, Palermo; and the Harbour at Genoa. The table base is gilt bronze and consists of four legs surmounted with rams’ heads. It is 98.7 cm in diameter and weighs 86 kg. This table was exhibited in London at the Great Exhibition of 1851 where it received a Council Medal, the highest honour awarded [18].

The sky was the most affected part of the tabletop, with a significant amount of dirt darkening the blue colours and making the filati stick out individually as opposed to appearing as a painting-like surface; significant dark circular cracks with losses were also visible. Several layers of wax were suspected over the surface: a layer of original beeswax, potentially covered by another layer of more recent wax. It was suspected that the dirt, rather than being loose, was ingrained in the topmost layer of wax. This layer yellowed over time, reducing the visual appearance of the colours. The scenery of the cities appeared dull and flat, having lost its effect of perspective.

Tests were conducted before cleaning to find an appropriate solvent that would remove the softened dirt and superficial layer of wax without dissolving the original coloured wax. As for the previous items, ethyl acetate was trialled but did not lead to any significant removal of dirt. Other solvents were, therefore, tested with different combinations of laser. The best results were obtained by lightly pre-wetting the surface with benzyl alcohol, irradiating with the laser, then clearing the surface with benzyl alcohol.

Tests were also made using Evolon^® CR microfilament textile for pre-wetting the surface with benzyl alcohol. The textile would be dampened with the solvent and left for a few minutes on the surface before starting irradiation with the laser. Impressive results were obtained, with the Evolon^® textile instantly removing some of the yellow colour of the non-original wax (see Figure 7a). Some areas were left under the Evolon^® for longer periods than others, but no noticeable differences in the cleaning were observed. Cleaning was completed by gentle swabbing to lift off the dirt and excellent results were achieved with this method both in terms of cleaning levels and reduction in the treatment time.

3.2.2. Treatment of the Tabletop

Most of the surface of the tabletop was treated as follows. The area was pre-wetted with benzyl alcohol for a few seconds (with or without Evolon^®), the laser was applied for a couple of passes at 0.71 J·cm⁻² and 5 Hz, then the dirt was lifted off by swabbing with benzyl alcohol. The surface was left to air dry before buffing with cotton. No changes in shine or gloss, or loss of colour from the original tinted wax were noticed after treatment.

The final results of treatment of the tabletop are presented in Figure 8.

Excellent removal of dirt and yellowing was achieved, leading to the retrieval of the original blue colour of the sky. However, care should be taken as minor amounts of coloured wax can be removed when cleaning too intensely. The difference between clean and dirty areas was clear and cleaning made a considerable improvement.

Swabbing was tested on the outer ring of the tabletop where the names of the cities are written, to brighten up the dark maroon surface. The coloured wax was instantly removed, and it was decided to leave the outer ring in its current condition to avoid removing the original coloured wax. The laser was not tested in this area.

The results were not as uniform as the ones obtained on the bonbonnières and other smaller objects from the collection because the original surface was not as homogeneous. Because of existing cracks, some filati were at slight angles or sticking out. The dark circular cracks showed losses of material, and cleaning could reduce their dark appearance in some areas, but additional conservation work was needed to fill in the gaps especially around the angels or in the sky of Genova.

Overall, the treatment of the tabletop was very successful and gave excellent results, both in terms of effectiveness of removal of dirt and yellowed coating, and preservation of the original materials. The use of the laser greatly increased the efficiency of treatment. Once the testing phase was completed, the rest of the table was cleaned in approximately 18 h. A considerably reduced use of solvent and mechanical action was possible compared to conventional cleaning using solvent only, and the process was much more controllable than with solvent only.

3.2.3. Evaluation of Treatment

Digital microscopy images of the treated areas are shown in Figure 9. Before treatment, a significant amount of dirt was noticeable on the surface, mainly between the glass tiles, making them stand out individually. A dark crack with losses runs vertically in Figure 9a,b, and cleaning led to the removal of superficial dirt, but additional conservation work was needed to fill it and reduce its visual impact.

After treatment (Figure 9b), the surface was brighter, the yellow coloration had been removed, and dirt was significantly reduced between the glass tiles, making the surface smoother and the filati better integrated. The swabs used to clean the surface did not show any traces of blue pigment with the naked eye or under the microscope. The treatment was successful. Figure 9c,d show some areas in the sky before and after cleaning. The removal of dirt and yellowing led to a brightened and more homogenous surface after treatment.

Microscopy was also undertaken to check for potential laser-induced damage on the surface. Figure 10a shows details of the Milan duomo after cleaning. Images were taken here to check for discoloration or unwanted interactions with darker colours. The duomo appeared unaltered overall.

Figure 10b shows some possible laser-induced damage (in white) on the green filato of the dark green tower separating the Milan and Venice section. This is the result of one laser pulse on the surface. Irradiation was immediately stopped, and no further cleaning was carried out in this area. It is suspected that the area was already damaged and had been repaired as some drops of wax were visible and reacted strongly to the laser. This reinforces our caution that interactions with the laser can vary strongly depending on the material and extra care is necessary when treating areas that are pre-damaged or have undergone prior repairs.

The combination of Er:YAG laser at 0.71 J·cm⁻² on surfaces pre-wetted with benzyl alcohol and swabbing with benzyl alcohol to clear out the softened material was found to be very successful for the removal of dirt from the tabletop. The colours were brightened, the yellow hue of the coating was removed, and no significant damage was observed under the microscope (except for minor signs of alteration in pre-damaged areas). The laser’s use greatly increased treatment efficiency and reduced the mechanical action required to dislodge the encrusted dirt.

Assessing the effect of the laser on this type of object was complex, because they often exhibited pre-existing damage that could be confused with laser-induced damage during the cleaning process. The use of imaging to assess pre- and post-test conditions was vital to make specific treatment choices and is to be recommended whenever possible. Laser-assisted cleaning carried out in this way offers a valuable option to achieve cleaning that either eliminates or, as far as possible, limits damage, and greatly improves the visual appearance of the surfaces.

3.3. Treatment of Other Objects from the Collection

Various other items of the Gilbert Collection were selected for trial and cleaning: bonbonnières, plaques, and a second tabletop. The combined treatment was found suitable for most of the micromosaic items.

A second tabletop, The Flora of Two Sicilies, made by Michelangelo Barberi in 1850 (LOAN:GILBERT.190-2008) and presented in Figure 11, was treated with the laser, as solvent cleaning only was found uncontrollable and required excessive mechanical action. The laser was set at 0.71 J·cm⁻², and ethyl acetate and benzyl alcohol were tested. The combination of pre-wetting with benzyl alcohol, laser irradiation at 0.71 J·cm⁻², and gentle swabbing with benzyl alcohol led to the removal of most of the dirt and yellowed coating, greatly improving the appearance of the table without the need for intense rubbing (see Figure 11b,c). It is interesting to note that the best cleaning conditions for this tabletop were the same as the ones used on the other tabletop made by Barberi.

Other successful examples include two bonbonnières (LOAN:GILBERT.490-2008, LOAN:GILBERT.397-2008), and a goldfinch plaque (LOAN:GILBERT.203-2008), presented in Figure 12. The combination of Er:YAG laser at 0.56–0.81 J·cm⁻² on pre-wetted surfaces together with swabbing with ethyl acetate was found to be very successful for the removal of dirt from the objects. The colours were brightened, the yellow hue removed, and no significant damage was observed under the microscope.

The results also highlight the uniqueness of each object and the need to understand better the variability of the materials used. The conditions used when working on one item may not be suitable for another, making it impossible to completely streamline the process and making it necessary to approach each item on a case-by-case basis. This applied to both laser irradiation (requiring adjusting the fluence slightly) and choice of solvent. This also shows the importance of testing the laser parameters and solvents before applying them to the entire object and observing the effects of the laser after treatment.

4. Conclusions

The conservation of micromosaic items is challenging because of their inherent fragility. Solvent cleaning is often unsuitable due to the risk of penetration of the solvent into the materials, and to the damage to the fragile surfaces through intense mechanical action, potentially generating additional losses of original materials. Laser cleaning was, therefore, investigated as an alternative method as it addresses several challenges faced during the conservation of micromosaics: an increased control, reduction in solvent use, reduced mechanical action, and overall better preservation of the original composite materials.

The Er:YAG laser by itself was found to dislodge some dirt off the surface, but the best combination for all items was pre-wetting with either ethyl acetate or benzyl alcohol, applying the laser at low fluences between 0.55 and 0.86 J·cm⁻² and 3-5 mm spot diameter for a couple of passes depending on the items, then swabbing gently with the solvent. Successful results were obtained for most tested items, with removal of a significant amount of dirt off the surface and removal of the yellow hue of the superficial coating, revealing the bright colours underneath and retrieving the 3D perspective effect in the case of the tabletops. The process was found to be safe for the original materials overall and offered a valuable option to achieve cleaning that either eliminated or, as far as possible, limited damage, and greatly improved the visual appearance of the surfaces. To further document the impact of laser cleaning, it would be worth investigating potential chemical or structural changes of the coating using Fourier-transform infrared spectroscopy.

The use of the laser significantly increased the efficiency of the treatment, which was particularly beneficial for conserving large objects where traditional cleaning methods may be time-consuming and labour-intensive. The mechanical action to dislodge the encrusted dirt was reduced, therefore improving the conservation of these complex items overall. Additionally, laser cleaning reduced the amount of solvent needed, which made the process more precise and controllable. By minimising the use of solvents, laser cleaning lowers the exposure of both conservators and the environment to problematic chemicals, reducing health risks and the chemical footprint of conservation practice.

Each item is unique and the materials and their response to the laser and solvents varied. Conditions working on one item might not be suitable for another, making it impossible to completely streamline the process and making it necessary to approach each item on a case-by-case basis. This also shows the importance of testing the laser parameters and solvents before applying them to the entire object and observing the effects of the laser after treatment.

Author Contributions

Conceptualization, C.V. and J.B.; methodology, J.B. and C.V.; validation, J.B., C.V. and L.B.; formal analysis, J.B.; investigation, J.B. and C.V.; resources, C.V. and L.B.; data curation, J.B.; writing—original draft preparation, J.B.; writing—review and editing, L.B. and C.V.; visualization, J.B.; supervision, L.B.; project administration, J.B. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by Ed and Anne Teppo.

Data Availability Statement

Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.

Acknowledgments

The authors would like to thank Ed and Anne Teppo for the donation of the Er:YAG laser to the Victoria and Albert Museum and for funding the research fellowship of one of the authors (J.B.). We are grateful to the UKRI Arts and Humanities Research Council (reference AH/V012134/1), through the Capability for Collections Fund (CapCo) for funding the refurbishment of the V&A science laboratory, including the purchase of the Hirox HRX-01 digital microscope and Hitachi TM4000Plus Desktop SEM used for the digital microscopy and scanning electron microscopy in this study.

Conflicts of Interest

The authors declare no conflicts of interest.

References

V&A Collection Database. Gilbert Collection. Available online: https://www.vam.ac.uk/collections/gilbert-collection (accessed on 6 November 2024).
Sonntag, M. Micromosaics from the Gilbert Collection: Conservation-guided research into materials and techniques. AIC Objects Spec. Group Postprints 2019, 26, 41–67. [Google Scholar]
Rudoe, J. Mosaico in piccolo: Craftsmanship and Virtuosity in Miniature Mosaics. In The Gilbert Collection: Micromosaics; British Museum: London, UK, 2000; pp. 27–48, Note 44. [Google Scholar]
Victoria and Albert Museum. Magical Micromosaics. Youtube Video. 4 April 2019. Available online: https://www.youtube.com/watch?v=DmB8-RGTd-Q (accessed on 1 September 2024).
Petochi, D. I Mosaici Minuti Romani dei Secoli XVIII e XIX; Abete: Italy, Rome, 1981; ISBN 887-047-018-0. [Google Scholar]
Fekrsanati, F.; Hildenhagen, J.; Dickmann, K.; Troll, C.; Drewello, U.; Olaineck, C. UV-laser radiation: Basic research of the potential for cleaning stained glass. J. Cult. Herit. 2000, 1, 155–160. [Google Scholar] [CrossRef]
Fekrsanati, F.; Hildenhagen, J.; Dickmann, K.; Mottner, P.; Drewello, U. Feasibility studies on applying UV-lasers for the removal of superficial deposits from historic glass. Stud. Conserv. 2001, 46, 196–210. [Google Scholar] [CrossRef]
Römich, H.; Dickmann, K.; Mottner, P.; Hildenhagen, J.; Müller, E. Laser cleaning of stained glass windows—Final results of a research project. J. Cult. Herit. 2003, 4, 112–117. [Google Scholar] [CrossRef]
Fekrsanati, F.; Klein, S.; Hildenhagen, J.; Dickmann, K.; Marakis, Y.; Manousaki, A.; Zafiropulos, V. Investigations regarding the behaviour of historic glass and its surface layers towards different wavelengths applied for laser cleaning. J. Cult. Herit. 2001, 2, 253–258. [Google Scholar] [CrossRef]
Maingi, E.M.; Treil, V.; Abad, M.P.A.; Angurel, L.A.; Rahman, M.A.; Chapoulie, R.; Dubernet, S.; Schiavon, N.; de la Fuente, G.F. Challenges in laser cleaning of cultural heritage stained glass. J. Phys. Conf. Ser. 2022, 2204, 012079. [Google Scholar] [CrossRef]
Pan, A.; Chiussi, S.; Serra, J.; González, P.; León, B. Excimer laser removal of beeswax from galician granite monuments. J. Cult. Herit. 2009, 10, 48–52. [Google Scholar] [CrossRef]
Pan, A.; Chiussi, S.; González, P.; Serra, J.; León, B. Comparative evaluation of UV–vis–IR Nd:YAG laser cleaning of beeswax layers on granite substrates. Appl. Surf. Sci. 2011, 257, 5484–5490. [Google Scholar] [CrossRef]
Pereira-Pardo, L.; Korenberg, C. The use of erbium lasers for the conservation of cultural heritage. A review. J. Cult. Herit. 2018, 31, 236–247. [Google Scholar] [CrossRef]
Teppo, E. Introduction: Er:YAG lasers in the conservation of artworks. J. Inst. Conserv. 2020, 43, 2–11. [Google Scholar] [CrossRef]
Pereira-Pardo, L.; Melita, L.N.; Korenberg, C. Tackling conservation challenges using erbium lasers: Case studies at the British Museum. J. Inst. Conserv. 2020, 43, 25–43. [Google Scholar] [CrossRef]
Stavroudis, C. The Modular Cleaning Program. 2024. Available online: https://modularcleaningprogram.com/background.html (accessed on 6 November 2024).
Vergeer, M.; van den Berg, K.J.; van Oudheusden, S.; Stols-Witlox, M. Evolon^® CR Microfibre Cloth as a Tool for Varnish Removal. In Conservation of Modern Oil Paintings; van den Berg, K.J., Bonaduce, I., Burnstock, A., Ormsby, B., Scharff, M., Carlyle, L., Heydenreich, G., Keune, K., Eds.; Springer: Cham, Switzerland, 2019. [Google Scholar] [CrossRef]
V&A Collections Database. The Beautiful Sky of Italy Table. Available online: https://collections.vam.ac.uk/item/O157742/the-beautiful-sky-of-italy-table-michelangelo-barberi/ (accessed on 6 November 2024).

Figure 1. Cross-section of micromosaic with base or cassina material (stone, metal), putty paste in which glass filati are embedded, grouting around the filati, and several layers of wax on the surface.

Figure 2. St. Peter’s Square bonbonnière (accession number LOAN:GILBERT.924-2008). (a) Overview of the bonbonniere with broken corner; (b) close-up on the broken corner showing the structure of the micromosaic with filati inserted in the putty; (c) broken fragment used for the trials. The structure of the micromosaic is clearly visible in the cross section, with the underlying putty into which the glass filati are inserted, and the grouting in between the filati.

Figure 3. Before and after irradiation test on the fragment of the St. Peter’s Square bonbonnière. (a) Before; (b) after 3 laser shots at 0.56 J·cm⁻², alteration circled in red; (c) close-up on the altered surface; (d) SEM image of the altered surface showing melting; (e) cross-section of the filati before laser irradiation; (f) cross-section of the filati after irradiation at 0.56 J·cm⁻² showing no change; (g) surface of the filati before laser irradiation; (h) surface of the filati after laser irradiation at 0.56 J·cm⁻² showing no change; (i) surface of the stone before laser irradiation; (j) surface of the stone after laser irradiation at 0.56 J·cm⁻² showing no change; (k) surface of the grouting before laser irradiation; (l) surface of the grouting after laser irradiation at 0.56 J·cm⁻² showing no change (slightly different spot, no changes in texture, inclusions).

Figure 4. Before and after cleaning test on the bonbonnière. (a) Before laser and solvent cleaning; (b) after softening with the laser at 0.56 J·cm⁻² and swabbing with ethyl acetate. The red frame indicates cleaned area; (c) before solvent cleaning; and (d) after cleaning with ethyl acetate only.

Figure 5. Microscopic images of the glass filati before and after treatment. (a) White filati of the sky before treatment; (b) white tiles after laser and solvent cleaning showing excellent removal of dirt and preservation of the tinted wax and no alteration of the filati or grouting; (c) white filati (different area) after solvent cleaning only showing good removal of dirt but dirt remaining at the corners.

Figure 6. The Beautiful Sky of Italy tabletop (accession number LOAN:GILBERT.894-2008) before treatment, showing dirt in the sky and concentric black cracks with losses.

Figure 7. Using Evolon^® to pre-wet the surface prior to irradiation. (a) The Evolon^® following application. Note the yellow colour immediately picked up by the textile; (b) the three stages of cleaning: 1—Evolon^® applied to the surface; 2—after removal of the Evolon^® but before the laser; 3—after laser irradiation and swabbing.

Figure 8. Images before and after treatment of the tabletop. (a) Overview before treatment; (b) overview after treatment showing good removal of dirt and brightening; (c) before treatment of the Milan and Genova sections; (d) after treatment of the Milan and Genova sections showing excellent removal of dirt and yellowing and brightening of colours; (e) before treatment, zoom on the Milan section; (f) after treatment showing better impression of 3D and brightening of colours.

Figure 9. Microscopic images of the surface of the sky before and after cleaning. (a) Before cleaning, showing significant amount of dirt between filati; (b) after cleaning, showing reduction of the dirt, removal of the yellowing, and slight improvement around the crack; (c) before and after treatment of blue filati, showing good removal of dirt and preservation of the blue tinted wax; (d) before and after treatment of blue filati in a different area, showing good removal of dirt and preservation of the blue tinted wax.

Figure 10. Microscopic images of the surface after treatment, checking for laser-induced damage; (a) close-up on the Milan duomo showing no alteration to the grouting or filati; (b) potential laser-induced damage on the dark green tower separating the Milan and Venice sections, showing possible laser damage circled in white dotted lines.

Figure 11. Cleaning The Flora of Two Sicilies tabletop (LOAN:GILBERT.190-2008). (a) Overview of the tabletop; (b) a detail before cleaning; and (c) after cleaning.

Figure 12. Successful examples of laser-assisted cleaning. (a) Bonbonnière depicting a mountain (LOAN:GILBERT.490-2008); (b) before and after cleaning image showing the difference between clean and dirty surface. (c) Bonbonnière depicting a goldfinch on a branch before cleaning (LOAN:GILBERT.203-2008) and (d) after cleaning; (e) microscopic image before treatment showing dirt between the filati and (f) after treatment showing reduction of dirt and improvement in the appearance of the dark cracks.

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

Share and Cite

MDPI and ACS Style

Brand, J.; Vida, C.; Burgio, L. Er:YAG Laser Cleaning of Micromosaics from the Rosalinde and Arthur Gilbert Collection at the Victoria and Albert Museum. Heritage 2024, 7, 7309-7324. https://doi.org/10.3390/heritage7120338

AMA Style

Brand J, Vida C, Burgio L. Er:YAG Laser Cleaning of Micromosaics from the Rosalinde and Arthur Gilbert Collection at the Victoria and Albert Museum. Heritage. 2024; 7(12):7309-7324. https://doi.org/10.3390/heritage7120338

Chicago/Turabian Style

Brand, Julia, Carmen Vida, and Lucia Burgio. 2024. "Er:YAG Laser Cleaning of Micromosaics from the Rosalinde and Arthur Gilbert Collection at the Victoria and Albert Museum" Heritage 7, no. 12: 7309-7324. https://doi.org/10.3390/heritage7120338

APA Style

Brand, J., Vida, C., & Burgio, L. (2024). Er:YAG Laser Cleaning of Micromosaics from the Rosalinde and Arthur Gilbert Collection at the Victoria and Albert Museum. Heritage, 7(12), 7309-7324. https://doi.org/10.3390/heritage7120338

Article Menu