REVIEW Surface characterisation techniques in the study and conservation of art and archaeological artefacts: a review A. Giumlia-Mair*1, C. Albertson2, G. Boschian3, G. Giachi4, P. Iacomussi5, P. Pallecchi4, G. Rossi5, A. N. Shugar2 and S. Stock6 Key issues related to surfaces and materials in the study and conservation of archaeological, artistic and historical objects are presented and illustrated with case studies. The materials cover a relatively broad chronological and compositional range. An important objective of the review is to inform the materials science and engineering community of the problems and needs of conservators, archaeologists, conservation specialists, art historians, archaeometrists and researchers in the field of ancient materials. In this way, improved technical information on methods designed to identify surface treatments and surface finishes, development of a common language among humanities and scientific researchers, and awareness of new applications appropriate in archaeometric studies can be promoted. Keywords: Archaeometry, Ancient materials, Surface analyses, Surface treatments, Reviews Introduction In a time in which research, scientific discoveries and technical approaches are evolving almost daily, scholars working on art and archaeological objects feel that more useful work might be done if a closer connection can be established with the world of technology and with scientific and engineering communities. This review presents examples and case studies related to the surfaces of ancient finds and of objects of art, and of solutions and approaches to a series of problems and questions arising in different fields of study which may be of interest to a large number of scholars confronted with similar problems. However, the aim is also to provide an indication of the wishes and points of view of conservators, archaeologists and conservation specialists to a larger scientific and materials engineering community, to stimulate new possibilities and the application of emergent technologies, efficacious and adequate to represent a solution to some of our everyday problems. Finally, the authors wish to promote this emerging interdisciplinary activity to a wider audience, and at the same time draw attention to the importance of finding the right technology to solve the problems encountered in their work. Important efforts have been made in recent years to improve the 1 AGM Archeoanalisi, Merano (BZ), Italy Art Conservation Department, Buffalo State College, Buffalo, NY 14222, USA 3 Dip. di Scienze Archeologiche, Università degli Studi di Pisa, Pisa, Italy 4 Soprintendenza Beni Archeologici della Toscana, Firenze, Italy 5 Istituto Nazionale di Ricerca Metrologica (INRIM), Torino, Italy 6 Conservation, Royal Ontario Museum, Toronto, Canada 2 *Corresponding author, email giumlia@yahoo.it ß 2010 W. S. Maney & Son Ltd. Received 4 August 2009; accepted 10 September 2010 DOI 10.1179/175355510X12850784228001 study of surface features in different fields of research and to collect information on this topic.1,2 This review also aims to promote broader diffusion of these studies. It is first appropriate to comment on some current trends in the field of archaeometry. Very often, the aim of studies is simply the wish to apply a novel technology, developed for other purposes, to the ancient materials and art objects without there being a real need to use this technology. Regrettably, in many cases, extremely sophisticated and highly expensive technologies are employed for simple tasks such as identifying inclusions in ceramics or even just distinguishing different alloys or repairs on copper-based objects. These tasks can cheaply, quickly and effectively be carried out by using current and inexpensive analysis methods such as scanning electron microscopy (SEM) and X-ray fluorescence spectroscopy (XRF). There is no need to employ complex methods to solve simple questions, but new technologies are badly needed to provide solutions to problems that cannot be solved with current methods. Surface methods of analysis in archaeometry One of the most common, effective and widely employed techniques for the analysis of ancient materials is energy dispersive XRF. The elemental composition of samples can be determined non-destructively, with short analysis time. A great advantage of the method is also the simultaneous multi-element analysis of a wide range of elements. The Xray spectrum from an irradiated sample displays multiple peaks of different intensities which immediately give an idea of the composition and of the quality of the data collection, so that, if needed, the measurement can be immediately Materials Technology 2010 VOL 25 NO 5 245 Giumlia-Mair et al. Surface characterization techniques in art and archaeology repeated or a better area for measurement can be found. Latest generation mobile instruments have greatly increased their precision and accuracy and are particularly advantageous in the many cases in which it is not possible to transport to the laboratory the objects to be analysed, or for example on excavation, for the quick identification of materials. If carefully used, this method is extremely useful and versatile, but care must be taken in the choice of dedicated standards for each material – ideally a selection of all possible material compositions – and in the evaluation of the results. The user must be experienced in the field of ancient materials from the analytical and historical point of view, have a wide knowledge of ancient technology, be aware of possible alterations due to age, prolonged burial and past treatments, and, obviously, of the difficulties in the determination of some elements, overlaps of lines, spurious peaks, too high concentrations results in the case of some elements due to interference with other elements, etc. X-ray diffraction is a well known method in archaeometry and it is employed on crystalline materials to determine their atomic-scale structure. This method is based on the fact that when crystalline matter is irradiated with X-rays it produces a diffraction pattern consisting of sharp peaks (Bragg diffraction peaks) which allow the identification of the three-dimensional (3D) arrangement of atoms in the analysed material. The SEM is widely used in archaeometry whenever the objects to be studied fit into the instrument chamber. The signals give information on the sample topography and composition. The images achieved have a very high resolution, a very large depth of field and can show details at nanoscale, around 250 times the magnification limits of a light microscope. Back-scattered electrons also show the distribution of different elements in the sample and can be extremely useful in the identification of inhomogeneous materials with inclusions and surface layers of different kinds. Characteristic X-rays allow the determination of the elements present in the specimen and quantify them. The limits of this method are the low sensitivity and the size of the chamber. Furthermore, metal samples do not require preparation (except for some cleaning), but non-conductive organic samples must be coated with a thin film of gold to be observed in the SEM. Transmission electron microscopy (TEM) has now been used by several generations of scientists and the performance of these instruments has been continually improved. In the last few years, there have been significant advances in the performance of this technique. For example the correction of the aberration in the magnetic lenses has reached a resolution which in the recent past was unthinkable and it is now possible to have an atomic resolution, under 0?1 nm. Resolution limits of better than 0?2 nm are now available in 200 or 300 keV commercial instruments and are sufficient for atomic-scale images, which open up exciting perspectives for some specific problems in archaeometry.3 Atomic-scale tomography and single atom identification are not impossible and might be applied to the study, and in particular the conservation, of ancient artefacts in very special and most complicated cases. A further important tool in the investigation of archaeological and art objects is Raman spectroscopy, a technique based on the change in frequency produced by an incident photon inelastically scattered by a 246 Materials Technology 2010 VOL 25 NO 5 material in a laser beam. The Raman spectrum is the unique chemical fingerprint of the analysed material. Surface enhanced Raman spectroscopy is a further development and uses a roughened metal film to enhance the Raman scattering of the analysed material. While XRF shows the elemental composition surface enhanced Raman spectroscopy gives information even on small amounts of degraded material for instance it can allow the reconstruction of the original aspect of ancient and/or altered painted surfaces.4 The following sections present an array of different material-related issues. The surface characterisation methods employed are those listed above and some more specific techniques that will be specifically mentioned in the appropriate place. Italian Neolithic pottery surface treatments and materials The fashioning of pottery in prehistoric times went well beyond the basic aim of producing even and smooth vessels; these were often embellished by features that were not strictly functional, like complex-shape handles or surface ‘decorations’. The meaning of these features is not relevant to the subject of this review, but probably transcended a purely aesthetic scope, resulting from complex social and psychological interactions within the human groups that produced the pottery. Literature on the archaeometry of pottery is rich and of high quality, the works deal mainly with the compositional characterisation of the artefacts, aiming to identify their provenance and – from the technological point of view – to determine the temperature of firing. Conversely, information about the technology of the surfaces is rather sparse, even if some concern about this topic dates back to the late 1970s,5,6 while most of the scientific literature and of the manuals deal with more recent productions. As to the methods employed in the archaeometric study of pottery, all the most up-to-date techniques have been tested and applied, the choice often carried out on the basis of the most easily available device. Nevertheless, it must be pointed out that optical microscopy has often proved to be the fastest and cheapest basically reliable method in the study of the surface treatments, provided that a good set of reference samples is available. Stereoand petrography microscope observations can be carried out respectively on the raw objects and on thin sections or slide-mounted crushed samples. Plain and reasonably smooth surfaces were simply obtained by spreading by hand or by a platy tool the clay paste on the surface of the vessel. In such cases, the aspect of the surface is rather dull, moderately porous, and may preserve fingerprints or strip-like/faceted traces of the tools. The basic decoration types were obtained by simply removing or compressing more or less large areas of the ceramic paste near the surface, in order to create a pattern of voids contrasting with the smoothness of the surface. Techniques such as scratching, impressing or stamping were common, and a wide range of tools were used, such as thin sticks, bone (or wood?) points and spatulae, shells, fingers and fingernails. Stamping the walls of the vessels before the clay paste was dry was typical of some cultures of the Early Neolithic of Southern Italy (Impressa ware) that date back to about 6000–5900 years BC; the most Giumlia-Mair et al. 1 ‘Impressa ware’ from Ripa Tetta (Northern Apulia, Italy); about 5800 BC. Brown pottery with handsmoothed surface and Cerastoderma edule impressions (image reproduced courtesy Boschian/Colombo) characteristic decorations (Fig. 1) were obtained by printing the rim with Cerastoderma edule, a widespread sort of clam. The removal of thin layers of the ceramic paste after the firing – a much more recent technique called scratching – was less common and produced typically rough surfaces within predefined areas characterised by a complex pattern of incised/excised areas. In some cases, the depressed parts of the stamped/ scratched decorations were filled by a paste of different colours, usually reddish or white, contrasting with the overall aspect of the vessel. The red pastes are mostly made up of a fine mix of clay, haematite and other complex Fe3z-oxides that occur commonly in the red soils of the Mediterranean region, and their extremely fine grain-size suggests that some physical separation of the soil material was carried out, probably by settling velocity. The white pastes included in most cases finely ground calcite (CaCO3) or other whitish rocks. In most cases, it is not clear whether the pastes were added before or after firing; in fact, these often occur just as traces preserved within the voids, suggesting that they did not adhere well to the surface. This may be because they were added after firing, or because they were poor in clay – as in the case of CaCO3 pastes – so that the adhesion to or the sintering with the ceramic paste was poor. In a few cases, the red paste is made of cinnabar (HgS);7 this peculiar pigment was certainly added after firing, because this mineral is decomposed at temperatures around 300uC and the Hg sublimates. Burnishing is a sort of polishing by which the surface of a vessel becomes glossy. This finishing technique appears already in the early phases of the Italian Neolithic. A smooth object, like a pebble8 or a bone spatula, is carefully but firmly rubbed on the surface of the vessel before firing, when the paste has partly dried and is no longer soft. In this way, the coarse temper grains – if present – are pushed below the surface and Surface characterization techniques in art and archaeology covered by a thin film of clay paste; moreover, the clay crystals lying within a surface layer 100–200 mm thick are rearranged, so that their larger faces become parallel to the surface. The glossy aspect of the burnished surface is therefore due to the reflection of light by a very large number of very small (less than 2 mm) crystal faces that are oriented in the same way. This pattern is very evident in thin section under crossed polars light, as the strongly continuous optical orientation of the clay crystals gives the micromass of the surface layer a typically striated birefringence fabric. If the polishing tool is wetted, the effect is enhanced and the surface becomes brightly shiny and lustrous, but with irregularities consisting in thin bands of slightly different colours that develop parallel to the direction of the movement of the tool. With reference to the post-depositional weathering of the sherds, a weakness locus may develop at the bottom of the burnished layer, which detaches in small flakes. In such a case, it may be mistaken for an added layer (as for example a slip) but the thin section examination should show the difference because of the birefringence pattern. Slipped ware is extremely rare in the Italian Neolithic; it can easily be recognised in thin section, because the surface layer was added by simply dipping the vessel into a diluted clay dispersion before firing, so that its birefringence fabric is not evidently striated. Moreover, the limit between the slip layer and the rest of the vessel body is very sharp and evident. Painting was a way to change the colour of the surface of a pottery object that appeared somewhat later during the Neolithic, about 5700 years BC, and became widespread between 5500 and 4500. A pigment, finely ground and dispersed in water (or other more viscous liquid?), was laid on the surface of the object in a 50–70 mm film, probably by brushes; as suggested by some secondary patterns occurring in the painted areas,9 burnishing could be combined with painting to give the colour a brighter aspect. The so-called Catignano Culture and the more recent Ripoli Culture date back to this period and produced large assemblages of very high-quality pottery decorated by burnishing and by typical zig-zag red bands on light brown background (Fig. 2). The list of available pigments is relatively short; these were usually made up of minerals or of mixtures of minerals with particular colours. Red was usually made up of amorphous Fe3z-oxides, mostly haematite (Fe2O3), and some Fe-stained clay (ochre) that could be easily obtained from red soils; it is noteworthy that iron oxides change to bluish grey if fired in reducing atmosphere, but this was more often the result of misfiring than a desired effect. Black was usually made of amorphous manganese oxides (MnO2, pyrolusite), when combined with red colours which had to be fired in an oxidising environment; conversely, intense black hues were obtained by firing in a strongly reducing atmosphere and by smudging the kiln/bonfire, so that minute C particles could penetrate the ceramic body, whose colour was also darkened by the reduction processes. White- or whitish-painted pottery is quite rare; the pigment could be made of calcium carbonates, which usually adheres poorly to the surfaces, or much more frequently of kaolinite-rich clay. Materials Technology 2010 VOL 25 NO 5 247 Giumlia-Mair et al. Surface characterization techniques in art and archaeology 3 Beads of gold between fingers of gold figure brought to Royal Ontario Museum for identification (image reproduced courtesy S. Stock) 2 ‘Catignano pottery’ from Catignano (Abruzzo, Central Italy); about 5500 BC. Light brown burnished ‘figulina’ (very fine) pottery, with haematite and clay red bands (image reproduced courtesy Boschian/Colombo) Recognition of unusual features on precious metal surfaces The identification of fabrication techniques is an important part in the assessment of the authenticity and provenance of historical artefacts and antiquities. Conservators must frequently rely on visual or microscopic examination of surfaces to identify them, because other methods, such as thin sections, would require destructive sampling. Furthermore, many artefacts cannot be considered for sampling because of their value (aesthetic or monetary); condition; or because their examination is part of an assessment for evaluation and authentication prior to acquisition. As technological advances in microscopy are made, we are not only able to see, but also to record, features on the artefact surface which would have been ‘invisible’ to the conservator in the past. Because of this, older collections can be re-examined to reassess their provenance and to distinguish collectibles and archaisms from genuine antiquities. As forgers improve, their product becomes more sophisticated making it increasingly difficult to distinguish between antiquities and archaisms or outright fakes. Therefore, the study of surfaces, which enables us to recognise and differentiate between marks left by various fabrication techniques, becomes an important element in the examination of an artefact. The study and identification of corrosion products and the relation between corrosion and metal substrate is another useful diagnostic tool. Corrosion products on the surface of and within the metal substrate, and their orientation to one another and to the metal surface, are also an important means of assessing an artefact. Although some corrosion products have been replicated successfully by chemical means with the purpose of deception, they can still be distinguished microscopically with careful visual examination. The texture, density and intergranular corrosion developed over centuries of burial in fluctuating burial environments are difficult to replicate. Several small gold figures and Islamic coins were brought to the Royal Ontario Museum (ROM), 248 Materials Technology 2010 VOL 25 NO 5 Toronto, Ontario, Canada for identification. The artefacts were purported to be from Iran, and to have been discovered among a deceased parent’s possessions, their origins unknown. Although they appeared ancient, possibly Elamite figures and Islamic coins, their identification was problematic. The objects were brought to the conservation department for detailed microscopic examination and photography using a stereo microscope, Olympus SZ40 with external fibre optic light. All photos were taken with a Nikon Coolpix 4500 through the eyepiece. The thin gold figures had no surface features that could be considered technologically diagnostic. There was no evidence for either casting or cold working and minimal surface corrosion/tarnish and dirt. However, visible under magnification, between the fingers on the hand of the male figure, were unusual microscopic beads or granules of gold (Fig. 3). This feature is not indicative of casting (would have shown casting flash, not perfect beads) or cold working (hammering or filing would have removed the beads), but similar features are seen on electroformed objects. For example, an examination of the ROM electroform (Fig. 6) of a silver Byzantine ring with Menorah and inscription, ROM 986?181?15, (Fig. 4) revealed the same type of beads or granules which are not present on the original (Fig. 5). Furthermore, surface corrosion and dirt in the detail of the original (Fig. 5 see arrow) were copied in the mould and transferred into the electroformed copy (Fig. 6 see arrow 2) as surface metal. The surface examination of the figure and the identification of the beads exposed it as modern ‘replica’. The silver tetradrachm of Melqart, Tyre, 74/3 BC, ROM 925?2?66 (Fig. 7) exhibits surface features resulting from the die stamping of a silver blank. Die flaw marks can be seen over the nose (Fig. 7 see arrow 1) and die striking-flow marks (Fig. 7 see arrow 2) are visible around the edge of the nose. Marks on the surface of the gold coins (Fig. 8) of the same owner as the gold figure, seemed to simulate the appearance of die flaws or die flow marks. But under magnification the marks on the ‘visitor’ gold coin (Fig. 9 see arrow 1) are clearly in the actual surface of the metal, probably as a result of a poor quality electroform. Mechanical finishing has left file marks (Fig. 9 see arrow 2) along an extremely fine edge. These features are not typical of a struck coin. Possibly, these ‘coins’ are imitations created as decorations for Ottoman bridal veils.10 Giumlia-Mair et al. 4 Silver ring with menorah 986?181?15, Byzantine 400–500 Bezel 1?4461?29 cm. Gift of Government of Ontario (image Stock) and inscription, ROM AD, D hoop 2?60 cm, ROM Membership and reproduced courtesy S. Surface characterization techniques in art and archaeology 1: die flaw marks on nose; 2: stretch marks from stamping by edge of nose 7 Detail coin, Silver tetradrachm, ROM 925?2?66, Tyre, Phoenicia mint, 74/3 BC, O/head of Melquart (Heracles), R/eagle on brow of boat (image reproduced courtesy S. Stock) 8 Gold coin brought by visitor to ROM for identification (image reproduced courtesy S. Stock) 5 Detail of lower edge of ring, ROM 986?181?15, arrow marks area of corrosion which has been transferred in the mould to the electroform positive as metal (image reproduced courtesy S. Stock) 1: pseudo die flaw marks; 2: file marks 9 Detail edge of gold ‘visitor’ coin (image reproduced courtesy S. Stock) 1: gold beads on electroform which do not exist on original ring, 2: mould transfer of corrosion into metallic silver 6 Detail of lower edge of electroform of ROM 986?181?15 (image reproduced courtesy S. Stock) Noble metals, unalloyed, such as gold and silver do not develop the complex surface corrosion typical of other archaeological metal surfaces. However, during burial, chlorides will cause corrosion of silver resulting in a diagnostic waxy lavender to pink brown corrosion product, silver chloride, chlorargyrite – AgCl – which can develop into a rather thick waxy layer (Fig. 10). This corrosion product is very difficult to replicate and is unlikely to be found on archaisms and replicas. If it has not been burnished onto the metal surface (Fig. 7 see cheek area), as was common years ago during cleaning, the edges can be lifted mechanically (Fig. 11) to reveal the original metal surface or what is a pseudomorph of this surface composed of complex silver corrosion Materials Technology 2010 VOL 25 NO 5 249 Giumlia-Mair et al. Surface characterization techniques in art and archaeology 10 Silver tetradrachm, ROM 949615?413, Syria, Seleukos III, 226-223 BCE. O/Seleukos III R/Apollo. 28?6 mm. Before cleaning in 2006. Area between arrows is area of massive silver chloride corrosion overburden. Corrosion overburden on inscription on pR and around face (image reproduced courtesy S. Stock) (Fig. 12). The condition of the surface will vary depending on metal structure and burial conditions. As visible on Fig. 12, where the chloride was removed, the detail of the inscription is now readable but the surface has a granular texture, which over the years will be lost through handling and polishing the soft silver. A further feature on silver objects which often goes unrecognised is firescale (Fig. 13). The silver tureen by Laurent Amiot, ROM 992?291?1?1 c. 1795, was covered with common silver tarnish, silver sulphide, acanthite – Ag2S. After polishing, a mottling was still visible over the exterior of the bowl, even though the ‘surface’ was clean. After many efforts to clean a small area of the underside of the bowl, a silversmith, Ellen Stock, was consulted. She interpreted this non-removable mottling as corrosion of the copper in the silver alloy, known to the trade as firescale. Firescale is produced by annealing the silver during the process of cold working. Hammering embrittles the metal during working and after each pass with the hammer, the metal is annealed to soften the structure to allow further working. Without annealing the metal would crack, but the heating involved oxidises the copper in the alloy to cuprite, turning it visibly black. Finishing 11 Detail edge of silver tetradrachm, ROM 949615?413. Mechanically lifting silver chloride corrosion from surface (image reproduced courtesy S. Stock) 250 Materials Technology 2010 VOL 25 NO 5 12 Silver tetradrachm, ROM 949615?413. Silver chloride corrosion overburden removed around inscriptions and face and detail on coin surface revealed (image reproduced courtesy S. Stock) the work by pickling removes this scale on the surface to leave a silver enriched zone, but the firescale deep in the body of the metal is not removed. Over the years, polishing removes the outer silver enriched material, exposing the mottled firescale zone. Firescale is not usually seen on objects of good quality and professional workmanship, but repeated polishing over the years can result in the firescale reappearing on the surface. The examination of the surface of metal artefacts reveals information which can ultimately alter our perception of an artefact. Conservators learn much about the preservation of materials, but are increasingly asked to evaluate these same materials for authentication. Today, a good understanding of fabrication techniques and the impact they have on materials is as invaluable as an understanding of the properties of the metals themselves. Characterisation of archaeological paintings: some examples from the north Mediterranean area Testimonies of human painting activity go back to very remote times. Just to name a few: the Aurignacian paintings found in the cave of Fumane, Lessini Mountains in the Italian region Veneto, are dated 35 000–32 000 BP;11 the famous caves of Chauvet (Ardèche, France) are dated to 32 000–31 000 BP,12 those of Lascaux (Dordogne, France) to 22 000–17 000 BP13 and the Franco-Cantabrian caves are dated to a time span ranging 17 000–13 000 years BP. Greek, Etruscan and Roman paintings are much more numerous, not only on walls but also on valuable objects. In all cases, their value 13 Silver tureen, ROM 992?219?1?1. Laurent Amiot (1764– 1839) c. 1795. Quebec City, Canada. H: 27 cm; W: 40?5 cm; D: 21?9 cm. Detail of firescale (image reproduced courtesy S. Stock) Giumlia-Mair et al. involves both technology and aesthetic aspects. The range of available raw materials and tools is narrow; however, ancient paintings demonstrate a ‘richness’ which can be better estimated through the knowledge of the pigments and their preparation techniques, the binders and the thickness of the paint layers. When present, preparatory layers and the methods of applying them are also important. All this information is necessary to explain the chromatic rendering of the colours. In order to collect information, different analytical tools are employed. There are different and increasingly specialised ways to investigate structure and composition of the painting layers. Usually sampling is needed for the study of the coloured surfaces; however, the principle of minimal invasiveness is respected, and some non-invasive techniques are also available. The first step is the observation of the samples through a stereomicroscope. The sample is then prepared in thin or polished sections and examined by optical (reflected and transmitted polarised light) and/or electron microscopy (SEM; environmental scanning electron microscope (ESEM); TEM respectively). These instruments allow us to distinguish the layers of the painting and the number of colours and preparatory coatings, to measure their thickness and to observe the granulometry and morphology of their components. The composition can be determined by energy dispersive X-ray microanalysis (EDX to identify chemical elements with n.a. >9 and collect a semi- or quantitative evaluation of the elements with n.a. >11) and by non-destructive X-ray fluorescence analysis (XRF to identify elements with n.a. >15) for the qualitative and quantitative determination of chemical elements. Fourier transform infrared (FTIR) and Raman spectroscopy are employed for the identification of the compounds and X-ray diffraction (XRD) for the determination of the mineralogical phases present in the sample. The aim is the determination of a compositional ‘fingerprint’ on specific areas of the archaeological finds or on samples of small size (down to K mm2) by using these instruments in combination. A good example are micro-FTIR and micro-Raman techniques, the latter can also be coupled to a micro-beam XRF.14 Where present, the organic components of the paint (i.e. lacquers and binders) are determined after appropriate extractive separations by using FTIR and/or gas chromatographic techniques, such as pyrolysis coupled to gas chromatography and mass spectrometry (Py-GC/ MS which distinguishes the various classes of organic products), gas chromatography coupled to mass spectrometry (GC/MS which recognises, after appropriate wet treatments, proteinaceous, lipidic and terpenic materials), or high-pressure liquid chromatography with UV-vis absorption and fluorescence detection (HPLCUV-fluo, for the separation and identification of the organic components of coloured lacquers). The analytical survey of the oldest mentioned paintings, those of Fumane, was carried out by following these steps. The analyses showed that red ochre was applied on a substrate consisting of limestone probably without any organic binder.11,15 A binder was found in paintings of the Epigravettian shelter of Dalmieri, where beeswax, identified by the FTIR, was used to disperse and apply an ironbased red pigment.16 A binder was also recognised in the Paleolitic paintings of Ariège (France) where GC/MS allowed the detection of an oil of vegetable or animal origin.17 The palette of paintings in the Cantabrian area Surface characterization techniques in art and archaeology (including Lascaux and Altamira) appears to be more various. The reds and blacks of the Lower Paleolithic were enriched by yellow in the Middle Palaeolithic and by white and brown in the Upper Palaeolithic. The compositional analysis of these paintings, which also utilised PIXE-PIGE instrumentation (with a detection limit lower than EDX) showed the use of haematite, sometimes mixed with goethite, of red ochre to obtain a red colour, and of manganese oxide and mineral, and wood and bone charcoal for black.18 Moreover, the analysis demonstrated that in the Upper Paleolithic the red pigment was also artificially produced by heating goethite.19 The Etruscans used more advanced painting techniques. The colour was applied over a preparatory layer which improved adhesion to the substrate and chromatic rendering. The tombs of Sarteano and Chiusi (Siena, Italy), dated sixth to fifth century BC, can be taken as a good example. In this case, the observation of stratigraphic sections of the painting showed how the stone substrate was properly smoothed by applying a layer of clay to close the porosity and level the surface of the rock, where the tombs were dug. In the hypogean paintings, the Etruscans added the blue pigment deriving from the Egyptian tradition and which remained in use for many centuries. Indeed in Roman times, in Pompeii, Egyptian blue (cuprorivaite) was called Pompeian blue or caeruleum. The identification of this kind of blue pigment, as well as the chemical analysis of the constituents (CaCuSi4O10), is provided by its crystallographic habitus (bipiramidal-tetragonal) which is easily recognisable in electron microscopy. In Sarteano e Chiusi the application of pigments (hematite for red, calcite for white, Egyptian blue for blue and charcoal for black) was made after dispersion in egg (tempera painting).20 One more example of a multianalytical survey is provided by the characterisation of the paintings found in the Etruscan necropolis of Sovana (Grosseto, Italy). The preliminary results lead to the identification of a careful research in the preparation of the colours. A complete diagnostic survey on the Necropolis of Sovana by the Soprintendenza per i Beni Archeologici della Toscana (Superintendence for Archaeological Heritage of Tuscany) is now in progress. The preparatory layer of clay was replaced by a coating of different materials. In the Ildebranda Tomb (third century BC) the application of paint on columns (not on sculptured walls) was done on a calcium carbonate based intonachino. On the modelled eardrum the preparation consisted of a thin layer of amorphous silica. In the same tomb, egg and animal glue were used as paint binders, and were probably differentiated depending on coloured or preparatory layers.21 Optical and electronic microscopy, EDX and wavelength dispersive X-ray spectroscopy microanalysis, PIXE, XRD and FTIR were utilised for the characterisation of a large number of wall paintings in the preRoman tombs in southern Italy and Macedonia (Greece). Here the study of the preparatory and coloured layers revealed the use of different painting techniques: fresco, secco, with lime and also with tempera, employing proteinaceous and saccharidic binders, with an increased variety of pigments.22,23 The analyses carried out to study the painting technique of the Sarcophagus of the Amazons (IV cent. BC), attributed to Etruscan or Magna Graecia workers,24 identified the organic and inorganic components of the whole stratigraphy (Figs. 14 and 15).25 Materials Technology 2010 VOL 25 NO 5 251 Giumlia-Mair et al. Surface characterization techniques in art and archaeology 14 Sarcophagus of the Amazons. Detail of decoration (image reproduced courtesy Soprintendenza Archeologica Toscana) The HPLC-UV-fluorescence analysis showed that the violet colour was purpurissum, a lacquer produced with a dye extracted from molluscs of the genus Murex.26 This purple-coloured lacquer is also present in the pigments found in Pompeii and its nature was matter of debate. A multi-analytical work including the already mentioned techniques, but also direct exposure electron ionisation coupled with mass spectrometry (DE-MS), a technique which is able to search for possible brominates chromophores, (i.e. 6, 69-dibromoindigotine, the molecule constituting the purple colour of the lacquer) and TEM coupled with microanalysis (TEM-EELS, TEMEDX, TEM-SAED) evidenced the use of organic dyes of animal (purple) and plant (madder) nature, absorbed by amorphous silica (allophane) or crystalline silicate (kaolin).27 Micro-Raman, XRD and FTIR28 were also used to characterise the white pigment present in some Etruscan polychromies coming from some ceramics of Cerveteri: kaolin was used on painted plates for parietal coating and pottery (sixth century BC). Among the ancient artefacts analysed, there is a classical Greek marble basin (320–280 BC). The pigments used for its decoration were investigated by means of polarised light microscope, XRF, XRD, ESEM with EDX, EPMA. The organic components of the purple lacquer were analysed with ultraviolet-visible spectroscopy, fluorescence spectroscopy and thin layer chromatography.29 The results showed the use of 15 Polished section (right) of the blue stratigraphy with a EDX spectrum of cuprorivaite, b XRD of cerussite with calcite, c FTIR of blue painting versus cerussite standard and cuprorivaite standard (images reproduced courtesy Soprintendenza Archeologica Toscana) 252 Materials Technology 2010 VOL 25 NO 5 Giumlia-Mair et al. 2010 VOL 25 NO 5 253 Surface characterization techniques in art and archaeology Materials Technology 16 Atomic force microscopy topographic rendering and 3D rendering of the acrylic painting control sample, suction block sponge treated sample and the Mr Clean sponge treated sample. The variation of surface topography has been reduced with cleaning. The extensive alterations in the Mr Clean sponge cleaned sample is too large to fully appreciate using AFM (images reproduced courtesy Shugar/Albertson) 2010 Surface characterization techniques in art and archaeology Materials Technology Giumlia-Mair et al. 254 VOL 25 NO 5 17 Atomic force microscopy topographic rendering and 3D rendering of the oil painting with a dammar varnish control sample, suction block sponge treated sample and the Mr Clean sponge treated sample. The variation of surface topography has clearly been altered. Scratches are evident on the suction block sponge cleaned sample and deep grooves and alterations are clearly visible on the Mr Clean sponge cleaned sample (images reproduced courtesy Shugar/Albertson) Giumlia-Mair et al. 2010 VOL 25 NO 5 255 Surface characterization techniques in art and archaeology Materials Technology 18 Optical profilometry 3D rendering, 3D mesh and section profile of the acrylic painting control sample, rubber chemical sponge treated sample, suction block sponge treated sample and the Mr Clean sponge treated sample. All three cleaning techniques have reduced the natural topography of the control sample. Smearing of the paint and scratches can be seen in the suction block sponge and Mr Clean sponge samples. The deep wide grooves created by the Mr Clean sponge are clearly evident (images reproduced courtesy Shugar/ Albertson) 2010 Surface characterization techniques in art and archaeology Materials Technology Giumlia-Mair et al. 256 VOL 25 NO 5 19 Optical profilometry 3D rendering, 3D mesh and section profile of the oil painting with a dammar varnish control sample, rubber chemical sponge treated sample, suction block sponge treated sample and the Mr Clean sponge treated sample. The reduction in topography and scratching is evident in all cleaning techniques (images reproduced courtesy Shugar/Albertson) Giumlia-Mair et al. cinnabar for red, Egyptian blue, cerussite for white, jarosite for yellow and madder for purple. Cerussite and cinnabar were also detected in the Sarcophagus of the Amazzoni and jarosite in Pompeian pigments.25,30 Roman and Medieval wall paintings were also investigated. In this period the colour was applied over a layer of fresh intonachino in several coatings.31,32 A final note concerns the painting on wood in Roman times. In Pisa (Italy), in the archaeological site called ‘of the Ancient Ships’ the ships C (first century BC to first century AD) and E (second century AD) still have limited areas of paint: red and white, the former, and only white, the latter. The characterisation of pigments shows the presence of hematite and cerussite in the C and kaolinite and calcite, in the E hulls. For the application on wood, the colours were dispersed in a water resistant mixture of beeswax and Pinacee resin.33 The results confirm what Pliny wrote on paints used for ships.34 The recent studies on Lombardic frescoes in Italy are also worthy of mentioning.35 Potential damage of sponge cleaning treatments of paintings: an evaluation in the light of surface investigations Traditional surface cleaning techniques of paintings typically consist of swabbing with mild enzymatic solutions or other aqueous mixtures.36,37 Although proven and highly effective, these methods can be extremely time consuming. There are often instances in which it may be more appropriate to use a rubbing sponge especially with tight time constraints and/or when burdened with an exceptionally large painting/mural. Where these sponges have been used successfully in conservation, they clean by abrasive action and can potentially damage a painted surface on the micro-scale. Various aqueous cleaning methods have been well documented and described with respect to their interaction with varnish and paint surfaces,36–38 but investigations as to their effect on the topography of the painting has been extremely limited.39–41 Several studies of sponge cleaning have focused on their effectiveness at cleaning42–44 and identifying potential residual material left on the paintings after treatment.45 Only recently have the effects of sponge cleaning been investigated topographically46 with recommendations that further work needs to be done to better assess the damage. This study compared the topographical surface effects that a series of cleaning sponges has on two types of painting surfaces, an acrylic and an oil painting varnished with a dammar resin. Three sponge surface cleaning techniques were investigated: a vulcanised rubber chemical sponge or soot removal sponges, a suction block sponge, and a Mr Clean ‘magic eraser’ sponge. Several techniques were used to characterise the abrasion caused by the sponges. Qualitative analysis was collected by reflected light microscopy (RLM) and SEM, while quantitative analysis was performed by optical profilometry and atomic force microscopy (AFM). Methodology Sample preparation Two paintings were chosen from the Buffalo State College Art Conservation Department’s collection of expendable paintings for research. These were an unvarnished acrylic and an oil coated with a dammar varnish. A 100 6 65mm inch portion was cut from each of the paintings and Surface characterization techniques in art and archaeology four sections were blocked off, one for each of the three sponges to be used and one control. Strip 1 was left as the control sample. Strip 2 was rubbed with a rubber chemical sponge, also known as a soot removal sponge. Strip 3 was rubbed with a suction block sponge (Saugwunder-King of Suction Block) in deionised water. Strip 4 was rubbed with a Mr Clean Magic Eraser in deionised water. Analytical techniques Analysis of the samples included both qualitative and quantitative assessments of abrasion caused by the sponges. To qualify the abrasion via visual observation a Zeiss Axio Imager A1 RLM was used to image the scratches on the surface of the samples and to roughly measure the average scratch width. A Hitachi S-4000 field emission SEM-EDS was used to confirm and detail surface morphology. In this study, SEM was used for visual examination and comparison of the surface damage as witnessed by the other methods of analysis since SEM requires that a sample be removed for investigation which is typically inappropriate for works of art. For quantitative analysis two newer techniques within the field of art conservation were employed. First, a Nanosurf EasyScan 2 atomic force microscope provided 3D topography at the nanometre/ micrometre scale using contact mode.47,48 Typical applications for AFM include: surface roughness, hardness measurement, corrosion, surface tension and surface inspection. It can image any number of surfaces: polymers, ceramics, glass, and biological samples. Images were taken of the cleaned areas and a 3D topography scan was computed. Second, a Nanovea ST400/3D non-contact profilometer was used. Optical profilometry measures the topography of a surface with sub-micrometre vertical resolution by scanning the surface using a white light probe and measuring the reflected light. Coupled with a computer for recording the data, highly accurate 3D topographic maps of the surface can be created and statistically studied. This technique is well documented and has been used successful in the art world to look at items from engraved astrolabes to characterising varnish layers on paintings.49 Results/discussion During cleaning the visual changes seen on the sponge and on the paintings were recorded. The rubber chemical sponge showed slight grime. The suction block sponge showed the least effect, picking up little if any grime. The Mr Clean sponge was the most aggressive and it began to pick up paint and/or rip through the varnish layer. In addition, when the oil painting was cleaned a resinous smell was created, probably from the abrasive action on the dammar varnish. Using RLM the extent of damage became more apparent. The rubber chemical sponge only very slightly scratched or abraded the acrylic and oil paintings. The suction block sponge produced scratches on both samples. Finally, the Mr Clean sponge produced extensive scratches, abrasions and worn surfaces on both paintings. The visual observations were confirmed by scanning electron microscopy which showed that the rubber sponge was the least harsh, but also had the least cleaning power. The suction block sponge also caused some damage to the acrylic painting. There were only three scratches (average width 43 mm) visible which may be due to slight contamination running across the surface. The Mr Clean sponge exhibited intense abrasion with numerous scratches (average width 21 mm). The oil samples all Materials Technology 2010 VOL 25 NO 5 257 Giumlia-Mair et al. Surface characterization techniques in art and archaeology experienced varying degrees of scratching. The visible scratches were smaller in size but greater in scale than those of the acrylic samples. The rubber chemical sponge had an average scratch width of 8 mm, the suction block sponge an average scratch width of 4 mm, while the Mr Clean sponge had innumerable scratches with average scratch width 5 mm. Atomic force microscopy was able to discern clear differences between the different sponge cleaners. The main concern with AFM is that its measurement range is designed for nano-scale resolution48 and the scratches made by the sponges are considerably larger. Nonetheless, clear measurements could be made and were helpful in assessing the damage. The limited damage by the rubber sponge was not clearly definable and was therefore not included. Both the acrylic and oil paintings show signs of damage. The acrylic painting had greater scratch width than the oil painting average scratch width of 3?8 mm for the suction block sponge and 8?6 mm for the Mr Clean sponge (Fig. 16). The oil samples had average scratch width of 4?5 mm for the suction block sponge and 4?7 mm for the Mr Clean Sponge (Fig. 17). For both sample sets there were innumerable scratches from the Mr Clean sponge and very few from the suction block sponge. The degree of damage to the acrylic was not clearly discernible using AFM. The damage to the oil painting was exceptionally clear. The naturally rough surface of the painting was reduced to a flat landscape using the suction block sponge, while the Mr Clean sponge ground the surface and created heavy striations on the painting. Optical profilometry provided excellent resolution for determining the level of damage produced by the sponges (Fig. 18). The peaks and troughs on the acrylic control ranged between 10 and 20 mm. The rubber chemical sponge flattened the painting’s landscape and reduced the range to 5–10 mm. There were almost no apparent scratches from this cleaning sponge but its effect on the topography of the painting was clear. The suction block sponge introduced a smeared surface with minute scratches of y5 mm. Although there appeared to be several more scratches than those produced by the rubber chemical sponge, the topography was not altered as greatly. The Mr Clean sponge showed the greatest damage entirely reforming the topography of the painting and creating wide deep grooves 0?35 mm wide and 20 mm deep. The oil painting showed similar effects (Fig. 19). The peaks and troughs on the oil control ranged from 1 to 1?5 mm. The rubber chemical sponge had the same effect on the oil painting as the acrylic. The surface was smoothed and the range of peak and trough height was reduced slightly to 0?7 mm. Several scratches were evident across the surface. The suction block sponge altered the surface to the same extent as the rubber chemical sponge. Several scratches were approximately 0?03 mm wide and 1?5 mm deep. Other smaller scratches ranged in depth from 0?1 to 0?25 mm. Again the Mr Clean sponge showed the greatest damage. The surface was burnished down and extensive scratching can be seen ranging from 0?025 to 0?04 mm wide and 0?25 to 0?75 mm deep. Using AFM, optical microscopy, SEM, and optical profilometry it was demonstrated that the Mr Clean sponge was extremely damaging to the paint surfaces. The suction block sponge was the second most damaging while the rubber chemical or soot removal sponge caused the least amount of surface abrasion. Both AFM and profilometry 258 Materials Technology 2010 VOL 25 NO 5 were found to be excellent tools for quantiative determination of the damage to paint surfaces by sponge cleaning. Profilometry might have a slight edge on AFM as it is able to look at larger areas yet still provides the level of resolution required to record the damage done by these sponge cleaning techniques on paintings. Applications of colorimetry in the field of cultural heritage Colour is one of the most important part of human expressions in art and cultural heritage (CH). The science of colour characterisation and measurement, i.e. colorimetry, has had a notable growth in the last 10 years in CH. Nowadays, most museums or conservation laboratories have their own colorimeter or portable spectrophotometer to objectively measure colour. This is mainly due to an increase in scientific analysis and to the development of portable instruments at low prices. Unfortunately there is a lack in measurement methodologies, measurement uncertainty evaluation and set-up. Due to the wide range of application and materials it is not possible to define the best system and methodology for all CH applications. However, for each application and required accuracy, it is possible to establish what is better and what is to be avoided. The past experience in the study of CH materials is fundamental in this work.50 Here we present different applications based on different measurement methodologies and instruments. Colorimetric characterisation Scientifically, colour is identified in the Commission International de l’Eclairage (CIE) colorimetric spaces (CIE 1931 or CIE Lab).51 With these systems each colour can be expressed as an univocal combination of three numbers, obtained from a mathematical elaboration of the spectral radiance of the observed radiation. For objects viewed in reflection this parameter depends on the spectral reflectance factor of the sample and the spectral radiance of the lighting source. If the source is known or standardised (i.e. the CIE illuminant set51) the colour can be considered a property of the sample. The instruments were especially developed to perform in one step this type of measurement. The measuring is fast and economical but the definition of accuracy and reliability of data requires a deeper knowledge of measurement methods and systems. Usually colours are expressed in chromaticity coordinates CIE 1931 x, y, z; where xzyzz51 x~ X X zY zZ y~ Y X zY zZ z~ Z X zY zZ and X~ 830ðnm Sl x(l) r(l) dl 360 nm Y~ 830ðnm Z~ 830ðnm Sl y(l) r(l) dl 360 nm Sl z(l) r(l) dl 360 nm where r(l) is the spectral reflection factor, Sl is the relative Giumlia-Mair et al. Surface characterization techniques in art and archaeology 20 Spectral reflectance factor of Egyptian statue of diorite (image reproduced courtesy Iacomussi/Rossi) spectral distribution of the incident energy radiation, and xðlÞ,yðlÞ,zðlÞ are the CIE colour matching functions. The chromaticity coordinates in the CIE Lab system can be obtained from XYZ with a simple algorithm. Usually the spectral reflectance factor is measured with a spectrophotometric device, in a defined geometrical condition of incidence and observation angles. The most common instruments are portable with an internal light source set up to measure in the geometrical conditions of 8/d, (8u of incidence, diffuse observation) or 45/0 (45u of incidence, 0u of observation) or vice versa. A microprocessor elaborates the acquired data and provides the colorimetric results. Usually portable instruments have lower accuracy than instruments in labs. The instrument accuracy is reduced when the linearity of the instrument is strongly involved and low signals are detected. In Fig. 20, the spectral reflectance factor of an Egyptian statue of diorite is shown.52 The measurement was carried out with a portable instrument: a step around 600 nm and two different trends are clearly visible. This behaviour is common with portable spectrophotometers when very low signal levels are measured. Spectro-colorimetric characterisations can be performed to evaluate the condition of the artefact or the influence of restoration. In Fig. 21, the spectral and colorimetric characteristics of several Egyptian paintings are shown. Restoration and cleaning can radically modify the spectral and colorimetric properties of the sample. The colour coordinates can also be evaluated by using special devices able to filter the reflected light with the colour matching functions. These devices are called tristimulus colorimeters and are only able to provide the colour coordinates for one given light source. If the device is not equipped with tristimulus colour filters, but with RGB filters, like scanners or digital cameras, it is necessary to provide a special calibration matrix M 2 3 2 3 X R 6 7 6 7 4 Y 5~M4 G 5 Z B to convert the measured RGB coordinates, of the digital device, in XYZ CIE 1931 coordinates. The definition of the calibration matrix M is not an easy task, the full procedure is described elsewhere.53 The principal difficulties arise from the colorimetric Gamut of the RGB device, i.e. the measurable RGB colours are a subset of all the colours of human perception, and from device linearity. Unfortunately with such devices it is not possible to recover metameric effects (i.e. materials with different reflectance spectra, but with the same colour coordinates, under a defined source). On the other hand, this type of instruments has several advantages and shows an increase in applications and feasibility. A key point is the availability, at low cost, of digital images with accurate colorimetric information in absolute values, if suitable characterisation and calibration methods are used. With an instrument of this type, in 2000 and 2002 the INRIM team performed the first full scanning of the Turin Holy Shroud.54 The reproducibility of the instruments (i.e. the agreement between the results of measurements of the same measurand carried out under changed conditions,55 must be verified to monitor the conservation conditions of an object, and should be high, so as to ascribe the measured differences only to the artefact and not to the instrument or measurement method. Instruments able to acquire images, but equipped with dispersive devices (gratings) or with several selective elements (i.e. interferential filters or a tunable filter) should be used to reach higher accuracy and overcome gamut problems. One of these instruments was developed by INRIM, and used, for the first time, to characterise Giotto Frescoes in Cappella degli Scrovegni Padova.56 The instrument (called MIR) consists of two main elements: an imaging spectrograph equipped with a CCD nitrogen cooled detector, with very high spectral accuracy and low spatial resolution, and a second CCD camera equipped with a tunable filter, with low spectral accuracy and high spatial resolution. In this last case, the spectral reflection factor and colorimetric data of a significant number of points were used with the aim of designing lighting installations with optimum colour rendering at low illuminance levels. As an example of Materials Technology 2010 VOL 25 NO 5 259 Giumlia-Mair et al. Surface characterization techniques in art and archaeology 22 Chromaticity coordinates of thirty sample points under the reference illuminant D65 and TL965 lamp 21 a spectral reflection factor and b colorimetric data of some Egyptian red frame paintings. Note the differences between the untouched and the restored pieces (images reproduced courtesy Iacomussi/Rossi) obtainable results in Fig. 22 the measured chromaticity coordinates of selected points under a fluorescent lamp (TL965) and a reference illuminant CIE D65 are shown. New development in colour measurement instruments is focused on the improvement of RGB filter equipped devices adding one or two ad-hoc filters and dedicated mathematical algorithm in order to enlarge the subset of measurable colours and reconstruct the spectral reflection factor of the artefact. This last operation in the cultural heritage context, because the spectral reflection factor of old pigments is limited in possible shapes and not strongly selective in wavelengths: it is mathematically demonstrated that a basis of six (RGBzRGB filtered) different and independent vectors is adequate to evaluate a large number of pigment spectra with sufficient accuracy. References Conclusions It is possible to envisage further progress being made in the field of conservation and many other surface connected research areas on ancient materials with the help of next-generation high resolution electron microscopy and associated spectroscopic chemical analyses. For example, new generation SEM instrumentation such as the ESEM or field emission gun SEMs operated at low voltage allow the observation of unprepared organic materials.1 One of the most advanced methods is point on line based confocal scanning. This technique can collect datasets that allow measurements down to 260 Materials Technology nanometre scale. Recent improvements on digital imaging measuring devices have made it possible to get information on colour and to measure topographies in high resolution. This method is now applied to industrial material and productions, but in the near future it might perhaps be advantageously applied to artistic and archaeological materials to solve special problems. The aberration corrected scanning TEM now makes it possible to identify individual atoms and any atomic scale defects in the materials, and might therefore be useful for a precise interpretation of surface alterations in the field of conservation or the mechanisms behind the development of different kinds of patina. Video-rate scanning tunnelling microscopy might even show the actual growth of the patina as a dynamic process,57 while in the near future it might even be possible to manipulate the atoms so as to induce new characteristics and properties and, for example, stabilise altered materials, by leaving the bulk unchanged. Arrays of metal nanoparticles are now used to further amplify Raman scattering from samples of different kinds and allow the analysis of tiny amounts of materials such as organic dyes.4 Traditional XRD has evolved into highenergy XRD and together with atomic pair distribution function data analysis can now be used on the nanoscale.58 It is not yet possible to say how these techniques and methods will change perspectives when they are applied in the study of art and archaeological artefacts, but their potential is exciting. 2010 VOL 25 NO 5 1. A. Giumlia-Mair (ed.): ‘Arts and surfaces’, Surf. Eng., 2001, 21, (5– 6), 329–479. 2. A. Giumlia-Mair (ed.): ‘Arts and surfaces’, Surf. Eng., 2008, 24, (2), 81–165. 3. D. J. Smith: ‘Ultimate resolution in the electron microscope’, Mater. Today, II, (New Microscopy Special Issue), 2008, 30–38. 4. C. Sealy: ‘Looking over the artist’s shoulder’, Mater. Today, 2008, 11, (11), 40–44. 5. T. Mannoni: ‘Analisi delle ceramiche dipinte neolitiche di Lipari’, Meligunı̀s Lipàra, 1980, 4, 870–875. 6. T. Mannoni: ‘Caratterizzazioni mineralogico-petrografiche e tecniche di alcune ceramiche del Tavoliere’, in ‘Passo di Corvo e la civiltà neolitica del Tavoliere’, (ed. S. Tiné), SAGEP Libri & Comunicazione, 94–97; 1983, Genova. Giumlia-Mair et al. 7. G. Boschian: ‘Appendice 1. Grotta Patrizi. Analisi petrografica di frammenti ceramici in sezione sottile’, in ‘La grotta Patrizi al Sasso di Furbara’, (ed. R. Grifoni Cremonesi and A. M. Radmilli), Bullettino di Paletnologia Italiana, 91–92, 63–120; 2000–2001, Cerveteri, Roma. 8. B. Zamagni: personal communication, 2004. 9. M. Colombo: ‘La Cultura di Catignano. Radici, fasi, evoluzione e contatti tra le oposte sponde del Mare Adriatico’, PhD dissertation, XXII Cycle, University of Pisa, Italy. 10. E. Keall: personal communication, March, 2009. 11. A. Broglio, G. Giachi, F. Gurioli and P. Pallecchi: Coll. Int.. sur ‘Le chemins de l’art aurignacien en Europe’, Harald Floss and Nathalie Rouquerol eds., Museeforum d’Aurignac, Toulouse, France, September 2005, 157–170. 12. J. Clottes and J. M. Geneste: Coll. Int.. sur ‘Le chemins de l’art aurignacien en Europe’, Harald Floss and Nathalie Rouquerol eds., Museeforum d’Aurignac, Toulouse, France, September 2005, 363–378. 13. A. Leroi-Gourhan and J. Evin: Supplément à Gallia Préhistoire, 1979, 12, 81–84. 14. A. Zoppi, C. Lofrumento and E. M. Castellucci: Proc. 6th Infrared and Raman Users Group Conference (IRUG6), Florence, Italy, March–April 2004, IFAC, 246. 15. M. P. Colombini, G. Giachi, F. Modugno, P. Pallecchi and E. Ribechini: ‘Preistoria Alpina’, nr. speciale, 50–54; 2005, Verona, Museo Civico di Storia Naturale di Verona. 16. P. Rosanò and S. Pellizzaro: ‘Preistoria Alpina’, nr. speciale, 150– 155; 2005, Verona, Museo Civico di Storia Naturale di Verona. 17. C. Pepe, J. Clottes, M. Walter and P. Menu: Comptes rendus à l’Académie des Sciences, 1991, 312, 929–934. 18. J. Clottes, M. Menu and P. Walter: Bulletin de la Société Préhistorique Française, 1990, 87, (6), 170–192. 19. M. P. Pomiès, M. Menu and C. Vignaud: Eur. J. Solid State Inorg. Chem., 1998, 35, 9–25. 20. M. P. Colombini, G. Giachi, P. Pallecchi and E. Ribechini: Atti del Convegno su ‘Pittura etrusca: Problemi e prospettive’ A. Minetti, Sarteano-Chiusi, Italy, Ottobre 2003, 162–167. 21. M. P. Colombini, G. Giachi, E. Ribechini, P. Pallecchi and U. Bartolucci: Atti XXI Conv. Int. Scienza e Beni Culturali ‘Sulle Pitture Murali’, ed. Arcadia Ricerche, Marghera (VE), Bressanone, Italy, Luglio 2005, 667–675. 22. C. Meucci: Boll. ICR., Nuova serie, 2001, 2, 115–144. 23. H. Breculaki: ‘L’esperienza del colore nella pittura funeraria dell’Italia pre-romana V-III secolo a.C.’, 102; 2001, Napoli, Electa. 24. A. Bottini: in ‘Il Sarcofago delle Amazzoni’, (ed. A. Bottini and E. Setari), 98; 2007, Milano, Electa. 25. G. Giachi, P. Pallecchi, M. P. Colombini, F. Modugno, A. Zoppi and E. M. Castellucci: in ‘Il Sarcofago delle Amazzoni’, (ed. A. Bottini and E. Setari), 164–165; 2007, Milano, Electa. 26. A. Andreotti, I. Bonaduce, M. P. Colombini and F. Modugno: in ‘Il Sarcofago delle Amazzoni’, (ed. A. Bottini and E. Setari), 157– 162; 2007, Milano, Electa. 27. M. Clarke, P. Fredrickx, M. P. Colombini, A. Andreotti, J. Wouters, M. van Bommel, N. Eastaugh, V. Walsh, T. Chaplin and R. Siddal: Proc. 8th Int. Conf. on ‘Non Destructive investigations and microanalysis for the diagnostics and conservation of the cultural heritage’, University of Lecce, Lecce, Italy, May 2005. 28. G. Trojsi, F. Bordignon, P. Postorino, P. Dore and V. Bellelli: Atti Conv. AIAr ‘Colore e arte. Storia e tecnologia del colore nei secoli’, Firenze, Italy, Febbraio 2007, Patron Editore, 353–361. 29. A. Wallert: Studies Conserv., 1995, 40, 177–188. 30. G. Giachi, E. De Carolis and P. Pallecchi, 2009, Materials and Manufacturing Processes, Special Issue: Manufacturing Techniques from Prehistory to the Renaissance, A.Giumlia-Mair ed., Vol.24, no.9, September 2009, 1015–1022. 31. P. Pallecchi: in ‘S. Pier Scheraggio. Gli scavi archeologici nell’ala di levante degli Uffizi’, (ed. M. Salvini), 231–233, 2005, Firenze, CA Editore. 32. G. A. Mazzocchin, M. Del Favero and G. Tasca: Annali di Chimica, 2007, 97, (9), 905–913. Surface characterization techniques in art and archaeology 33. M. P. Colombini, G. Giachi, F. Modugno, P. Pallecchi and E. Ribechini: Archaeometry, 2003, 45, (4), 649–664. 34. C. S. Plinius: ‘Naturalis Historia’, I cent. AD, (XXXV, 19 and 31): Plinio, Storia Naturale, 1988 (reprint), vol. V, Torino, Einaudi. 35. A. Zucchiatti: Surf. Eng., 2007, 24, 162–165. 36. R. White and A. Roy: ‘GC-MS and SEM studies on the effects of solvent cleaning on old master paintings from the National Gallery, London’, Studies Conserv., 1998, 43, (3), 159–176. 37. R. Wolbers: ‘Cleaning painted surfaces: aqueous methods’; 2000, London, Archetype Publications Ltd. 38. A. Burnstock and T. Learner: ‘Changes in the surface characteristics of artificially aged mastic varnishes after cleaning using alkaline reagents’, Studies Conserv., 1992, 37, (3), 165–184. 39. P. B. Eipper and G. Frankowski: ‘Cleaning of painted surfaces and examination of cleaning by 3D-measurement technology at the August Deusser Museum, Zurzach’, in ‘Lasers in the conservation of artworks: Lacona VI proceedings’, Vienna, Austria, September 2005, Springer proceedings in physics, 116, (ed. J. Nimmrichter et al.), 473–486; 2007, New York, Springer-Verlag. 40. B. A. Ormsby and T. J. S. Learner: ‘Wet cleaning acrylic emulsion paint films: an evaluation of physical, chemical, and optical changes’, Modern Paints Uncovered, Los Angeles, The Getty Conservation Institute, 2007. 41. R. Ploeger and A. Murry: ‘Morphological changes and rates of leaching of water-soluble material from artists’ acrylic paint films during aqueous immersions’, Modern Paints Uncovered, Los Angeles, The Getty Conservation Institute, 2007. 42. J. Hackett: ‘Observations on soot removal from textiles’, The Textile Specialty Group, postprints of papers delivered at the Textile Subgroup Session: American Institute for Conservation Annual Meeting, 1998, 8, 63–69. 43. S. Spafford-Ricci and F. Graham: ‘The fire at the Royal Saskatchewan Museum, Part 2: removal of soot from artefacts and recovery of the building’, J. Am. Inst. Conserv., 2000, 39, (1), 37–56. 44. P. B. Eipper: ‘Trockenschwämme zur Oberflächenreinigung (Dry sponges for surface cleaning)’, Museum aktuell, 2003, 91, 3898–3902. 45. E. Moffatt: ‘Analysis of ‘chemical’ sponges used by the commercial fire clean-up industry to remove soot from various surfaces’, Bulletin (IIC-CG), 1992, 17, (3), 9–10. 46. P. B. Eipper: ‘Problemfall Acrylfarbenoberflächen’, in ‘Restauro, 2,’ 112–122; 2009, München. 47. G. Binnig, C. F. Quate and C. Gerber: ‘Atomic force microscope’, Phys. Rev. Lett., 1986, 56, (3), 930–933. 48. G. Binnig, C. Gerber, E. Stoll, T. R. Albrecht and C. F. Quate: ‘Atomic resolution with atomic force microscopy’, Europhys. Lett., 1987, 3, 1281–1286. 49. W. Wei, J. Frohn and M. Weber: ‘Characterisation of the varnishobject interface using white light confocal profilometry’, in ‘Optics for arts, architecture, and archaeology’, (ed. C. Fotakis et al.), Munich, Germany, June 2007, SPIE no. 6618. 50. P. Iacomussi and G. Rossi: Surf. Eng., 2007, 17, 90–97. 51. Commission Internationale de l’Eclairage: Colorimetry, CIE 15?3; 2004, Wien, CIE. 52. P. Iacomussi and G. Rossi: ‘Luce per Faraoni e Dei, Riflessi di Pietra’, 137–138; 2006, Milano, Skira. 53. L. MacDonald and R. Luo: ‘Colour imaging’; 1999, Chichester, Wiley. 54. M. Flury-Lemberg (ed.): Sindone 2002; 2003, Torino, ODPF. 55. ‘International vocabulary of basic and general terms in metrology’, ISO/IEC Guide 2003 56. P. Iacomussi, G. Rossi, P. Soardo and L. Fellin: ‘A new method for studying lighting design for works of art’, Proc. 9th LUX Europa Conf., Reykjavı́k, Iceland, June 2001, Icelandic Illuminating Engineering Society. 57. G. Schitter and M. J. Rost: ‘Scanning probe microscopy at videorate’, Mater. Today, 2008, II, (New Microscopy Special Issue), 40– 48. 58. V. Petkov: ‘Nanostructure by high-energy X-ray diffraction’, Mater. Today, 2008, 11, (11), 28–38. Materials Technology 2010 VOL 25 NO 5 261