Laser cleaning for historic buildings
The introduction of laser-based cleaning techniques in the early 1990s has led to a significant advance in the quality of conservation carried out on artworks.
Lasers are now used routinely in many conservation studios throughout Europe to remove many types of unwanted surface accretions (such as pollution crusts, corrosion layers, paint and previous conservation treatments) from a wide range of substrates (such as stone, marble, terracotta, plaster, wood and bronze) in an extremely selective manner.

The development of robust, portable and flexible laser cleaning systems has meant that the technique has found applications outside the sheltered environment of the studio, in large-scale historic building projects. Such projects have included Amiens Cathedral (France), St. Stephan's Cathedral (Vienna, Austria), the Tomb of the Unknown Soldier (Warsaw, Poland), Rotterdam Town Hall (Netherlands), St Frederiks Church (Copenhagen, Denmark), the Church of the Maddalena (Venice, Italy) and the Palace of Westminster in London. In France and Italy in particular, the use of laser cleaning in historic building conservation is now commonplace.

The idea of using laser radiation to clean artworks can be traced back to the pioneering work of the American scientist John Asmus in the early 1970s.1 Asmus demonstrated that pulsed laser radiation could remove hard black pollution encrustations from fragile marble sculpture with a degree of selectivity not possible using more conventional methods of cleaning. Having demonstrated the feasibility of the technique, Asmus developed a prototype cleaning tool, which was employed in the conservation of a number of important artworks in Venice. However, although the technique appeared extremely promising in principle, the technology was largely undeveloped and its practical application was difficult. Laser systems were temperamental and immobile, and cleaning was slow and awkward. At that time other techniques, such as microairabrasive cleaning, provided a more practical and affordable solution. It was not until the late 1980s that the conservation world began to look once more at laser technology. Developments in laser design and laser beam delivery (such as articulated arms and optical fibres) meant that laser systems were now far more reliable and practical.

Small-scale laser cleaning of a Roman marble bust in a studio. The system shown delivers approximately two watts average power through a flexible optical fibre.
Initial work during this exciting period concentrated on refining the removal of hard black pollution encrustations from fragile limestone and marble sculpture. Researchers in the UK, France and Italy (working independently of each other) established the most suitable type of laser for this kind of work. Working closely with conservators, they demonstrated the high level of control and precision offered by laser cleaning. Prototype laser cleaning systems were developed and tested on a wide range of materials, including stone, terracotta and plaster. The success of these tests encouraged collaboration between conservators, scientists and laser system suppliers, which led to the introduction of the first commercial laser cleaning systems almost one decade ago.

There are now at least eight suppliers of laser cleaning systems in Europe, and in excess of fifteen models for conservators to choose from. Since the initial work of Asmus, research and development of laser cleaning in the field of conservation has tended to concentrate on sculpture. It is not surprising that this is where the technique is now most widely accepted (investigations into its application in other areas, such as paper, parchment, textile and painting cleaning are well underway). Initial laser cleaning systems were developed for studio work and were suitable for work of a scale up to about life-size; the development of these systems was driven by the extremely high quality of cleaning that could be achieved using a laser, rather than by the speed of cleaning (although cleaning rates were usually faster than comparable techniques).

The sensitivity of the technique inevitably led to the deployment of such systems on site, where they were generally used to clean sculpture and architectural detail on historic buildings.2,3 This work led to the development of more portable and robust systems, suitable for work in site conditions. During the past five years, a lot of work has gone into developing higher power lasers for larger-scale cleaning applications: public monuments and building facades.4

A laser produces light in an extremely intense and pure form. The light is also highly directional. These unique properties have meant that the laser, although originally touted as a solution looking for a problem shortly after its first demonstration in 1960, has found applications in all walks of life. Lasers are used at home in compact disc players, at the supermarket to read barcodes, in industry to cut, drill, weld and engrave materials, and also in the medical and military fields. Lasers are used to deliver energy (in the form of light) to a surface very precisely and in a form which is very controllable.

Plaster model (approximately 1.5 m tall) for the Port Sunlight war memorial (Wirral, Merseyside) during cleaning. Laser radiation has been used to remove dirt layers from the plaster without disturbing areas of original shellac.
The most widely used laser cleaning systems in conservation emit light in very short pulses (typically 10-8 s duration) at a wavelength in the near infrared region (1.06 mm).5 At this wavelength, many types of dirt absorb energy very strongly, whereas the surfaces of many types of stone, terracotta, plaster and other materials absorb relatively weakly. This means that there is often a strong degree of selectivity inherent in the cleaning process, and often it is possible to work at a level which is strong enough to remove dirt layers but too weak to damage the substrate. Overcleaning is not a problem in such cases, since the removal process will stop as soon as the clean surface has been exposed. In effect, the laser beam is able to discriminate between the clean and soiled surfaces.

The absorption of laser radiation in the surface of a dirt crust will cause a thin volume of material to heat up rapidly. This leads to a very fast expansion of the heated material, which in turn generates forces sufficient to eject particles of dirt away from the surface of the artwork. During laser cleaning, these processes are so fast that the dirt is removed before there is time for significant heat to pass into the artwork. The removal of hard black pollution encrustations from limestone carving, for example, can be achieved in an extremely selective manner. The patina of the stone is left intact, preserving fine surface detail such as tool markings. It is a noncontact process, and therefore suited to work on both sound and fragile surfaces (which can usually be cleaned without pre-consolidation).

The laser beam is usually delivered to the surface through a handheld pen, via an articulated arm (which uses mirrors to guide the beam through jointed hollow tubes) or a flexible optical fibre. Work on sculpture and other important carved surfaces is usually carried out by trained conservators, who use their skills and experience to determine the appropriate level of cleaning. Although there is a high degree of selectivity inherent in the laser cleaning process, it is important to remember that the laser is simply a tool. The technique is clean (the only waste product is the dirt removed from the surface, which is removed from the workplace by an extractor), involves no chemicals and causes very little disruption to the running of the building.

Cleaning rates depend on a range of factors, the most important being the nature of the surface layer being removed (hard black pollution crust, light soiling, paint etc.), the nature of the substrate (plain or carved, good or poor condition, porosity etc.) and the average power of the laser. The average power of a pulsed laser is defined as the number of pulses emitted per second multiplied by the energy in each pulse. It is a measure of the rate at which energy is delivered to the surface being cleaned. The cleaning rate increases as average power increases. A typical smallscale system emits pulses of 300 millijoules energy at a rate of 10 pulses-per second (an average power of three watts).

Typical cleaning rates using such a system would be 0.6 m2 /h for a black layer of pollution on brickwork, 0.2 m2/h for a light pollution crust on limestone ashlar. Public monuments and carved stonework on building facades can and have been cleaned with such systems, but higher power systems, such as 20 watts, are more appropriate for this type of work. A seven times increase in average power will probably lead to a five times increase in the rate of cleaning (there is always some loss of efficiency at high average power levels due to the manual control of the process). During the last five years largescale laser cleaning systems have been developed for work on building facades.

St Frederick's Church, Copenhagen, Denmark. High quality laser cleaning was used to clean much of the architectural and sculptural detail during the recent programme of work.
An 80 watt average power laser system is currently being used to clean the facade of Rotterdam Town Hall (5000 m2). Cleaning is carried out by an operator standing on a cherry picker. The laser system remains at ground level and the laser beam is delivered to the surface via a 30 m long optical fibre. Cleaning rates of 4 m2/h (black pollution crust on sandstone) have been reported. Of course, the overall time required to clean a facade can be reduced proportionately by using more than one system.

The development of laser cleaning systems in conservation has been driven by a desire to refine methods of cleaning employed by conservators. Quality rather than speed has, over the years, been the primary concern of conservators and scientists working in this field. This is the main reason that lasers have often been restricted to the cleaning of sculptural and architectural detail on historic buildings, rather than large areas of plain stonework. The development of higher power systems during the past five years has seen a significant increase in cleaning rates (and consequently a decrease in cost/m2), which has led to deployment of the technique over much larger areas. However, the cost per square metre of laser cleaning is still considered to be higher than that for alternative cleaning techniques (although this difference is often significantly less when all aspects of the cleaning process are taken into account, such as waste collection and disposal, disruption to the running of the building and reliability of equipment), which tends to limit the surface area of a historic building that can be cleaned in this way.

Test cleaning (using laser) carried out on heavily polluted Swedish Oland limestone.
A lack of awareness of the technique during the specification- design stage of the conservation work has also been a stumbling block at times. This problem is being overcome through the introduction of training courses, organisation of demonstrations, seminars, international conferences (the fourth international conference on Lasers in the Conservation of Artworks, LACONA IV, took place in Paris in September) and setting up of European networks (such as the European Commission's COST Action 'Artwork Conservation by Laser'), but it is a relatively slow process.

The widespread use of lasers for large-scale cleaning of buildings (as opposed to a facade or portal or simply important sculptural and architectural elements) in the future will depend on whether the technique can be developed to enable cleaning rates comparable with alternative methods. Maximum cleaning rate depends on using the energy delivered from the laser as efficiently as possible, using just enough pulses to obtain a clean surface but not wasting pulses on cleaned areas. At present, the most reliable means of delivering high average power to a surface is by using lasers emitting relatively low energy pulses at high speed. However, manual control of such a system invariably leads to inefficiency as the operator simply cannot respond fast enough to what is happening on the surface.

By using robotic systems, it would be possible to scan the laser beam across the surface at the optimum rate so as to achieve high quality cleaning at maximum speed. The process could be monitored from ground level. The laser and robotics technologies required for this already exist, so this form of semi-automated cleaning for large areas of relatively plain surface is a real possibility. Although this part of the cleaning would be semi-automated, the cleaning of more detailed carved surfaces would remain a manual process. It will take time and money to integrate the various systems and develop a working prototype. But in the future we may see lasers being used to clean the entire surface area of some of our wonderful historic buildings.

1 Asmus, J. F., Murphy, C. G. and Munk, W. H. (1973). "Studies on the interaction of laser radiation with art artifacts". Proceedings of SPIE, 41, 19-27.
2 Weeks, C. (1998). "The 'Portail de la Mere Dieu' of Amiens Cathedral: its Polychromy and Conservation". Studies in Conservation, 43, 101-108.
3 Beadman, K. and Scarrow, J. (1998). "Laser cleaning Lincoln Cathedral's Romanesque Frieze". Journal of Architectural Conservation, 4 (2), 39-53.
4 Salimbeni, R. and Bonsanti, G. (eds.) (2001). LACONA Ill, Proceedings of the third international conference on Lasers in the Conservation of Artworks. Florence, April 1999. Journal of Cultural Heritage, Elsevier, Paris.
5 Cooper, M. 1. (1998). Laser Cleaning in Conservation: an introduction. Butterworth-Heinemann, Oxford.