JOURNAL OF RAMAN SPECTROSCOPY J. Raman Spectrosc. 2005; 36: 829–833 Published online 27 June 2005 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jrs.1371
Micro-Raman spectroscopic study of pottery fragments from the Lapatsa Tomb, Cyprus, ca 2500 BC M. Sendova,1∗ V. Zhelyaskov,1 M. Scalera2 and M. Ramsey1 1 2
Division of Natural Sciences, New College of Florida, Sarasota, Florida 34243, USA The John and Mable Ringling Museum of Art, Sarasota, Florida 32243, USA
Received 7 February 2005; Accepted 22 March 2005
Micro-Raman spectroscopy was applied to the mineralogical characterization of Bronze Age Cypriot ceramic fragments from the Lapatsa Tomb collection of the Ringling Museum of Art. Micro-probing was carried out on the surface and the results were compared with those of micro-probing from the crosssection of the two samples studied. Significant statistics were collected from 500 locations on each sample. Various phases were identified: quartz, albite, calcite, ilmenite, anatase, rutile, hematite and maghemite. It was determined that the different red coloration of both pieces is due not only to different hematite concentrations, but also to different grain sizes. The presence of low-temperature albite and anatase suggests low firing temperatures. The presence of hematite suggests firing in an oxidizing atmosphere. Copyright 2005 John Wiley & Sons, Ltd.
KEYWORDS: micro-Raman spectroscopy; Cypriot pottery; firing; inorganic pigments
INTRODUCTION In 1973, the Ringling Museum of Art received 38 pieces of Bronze Age pottery from the Lapatsa Tomb 15 situated on the north slopes of the Kyrenia mountain range on the north coast of Cyprus. The artifacts were excavated during several expeditions under the auspices of the University of Melbourne, Australia. Overall, the excavated site vessels are diverse. They vary from a planned use in the after-life with tools for hunting or war to textile weaving implements such as spindle whorls. According to the archeologists, the works vary from hand-made to wheel-made. The clays may be described as originating from well-mixed, fine, sandy, sedimentary clays, fired, surface slipped then painted or polished.1 Micro-Raman spectroscopy is a powerful analytical method for non-destructive material characterization. In the past 7 years, there has been a notable increase in the number of studies applying the Raman spectroscopic technique to archeological artifacts ranging from Egyptian funerary artifacts,2 mummies,3 to prehistoric rock art,4 Byzantine hagiography,5 medieval paintings,6 manuscripts and cantorals7 and various objects of fine art from more recent history. This research has contributed a large body of Ł Correspondence
to: M. Sendova, Division of Natural Sciences, New College of Florida, Sarasota, Florida 34243, USA. E-mail:
[email protected] Contract/grant sponsor: US Department of Education; Contract/grant number: P116Z040038.
scientific facts to art historians, curators and conservationists. Much research has been devoted particularly to examining the chemical, mineralogical composition, pigments and glazes8 of ancient Greek,9 Roman,10,11 Chinese12,13 and Ottoman14 ceramic artifacts, medieval luster potteries and faience15 – 17 and Medici porcelain.18 The main questions which have to be answered are about the nature of the raw materials and the technology implemented for production of the ceramic artifacts, e.g. the firing temperature and the nature of the firing atmosphere. This knowledge will give us a better understanding of the civilization that created the ceramics, while providing a substantial advantage to conservation for improving the restoration and preservation techniques. The aim of this work was to implement micro-Raman spectroscopy to identify the mineralogical composition of fragments from the Lapatsa pottery collection (Plate 1) in order to learn more about the technological conditions, such as firing temperature and firing atmosphere under which these vessels were produced.
EXPERIMENTAL Raman spectra were taken from various locations on two samples. The two fragments originate from different vessels: samples A and B are shown in Plate 1(a). Sample A has a vibrant red lustrous surface, whereas sample B has an orange–red matte-looking surface. The two shards are part
Copyright 2005 John Wiley & Sons, Ltd.
M. Sendova et al.
spot diameter on the sample was around 16 µm, and the total laser power incident on the specimen was kept below 1.2 mW. Micro-probing was carried out on the surface and the results were compared with those of micro-probing from a cross-section of each of the samples studied. Significant statistics were collected from ¾500 locations on each sample.
of the objects found in the Lapatsa tomb, several of which are shown in Plate 1(b). The atmospheric conditions to which the artifacts were exposed during the many centuries of weathering underground have favored a possible flourishing of various microorganisms on the surface of the artifacts. To avoid any interference from the Raman spectra of these organic residues, the surfaces studied were cleaned by brief exposure to hydrogen peroxide solution with subsequent rinsing with water. Where possible, the cleaning was facilitated by ultrasonicating the samples. A fluorescence background was observed in almost all areas from which Raman spectra were recorded. As the intensity of the fluorescence spectra varied across the sample, the Raman spectra shown in this study are from places where the Raman to fluorescence intensity ratio was high enough to allow an accurate assignment of the Raman bands. Whenever it was possible, the real fluorescence background was subtracted from the raw spectra. Fluorescence background spectra without any Raman bands were collected from the immediate vicinity of the Raman observation spot, hence the form of the background spectra was kept unaltered. For some spectra the fluorescence background was subtracted by using a multipoint baseline. Moreover, as a result of this subtraction the Raman band wavenumbers in the background-corrected spectra remained within 2 cm1 of those in the raw spectra. All numerical manipulations of the spectra were done using GRAMS/AL 7.02 (Thermo Electron). Raman spectra of the samples were obtained using a Leica DMLP microscope (50ð objective) coupled to a Raman system manufactured by Kaiser Optical Systems. The Raman Rxn1 analyzer incorporates a TE-cooled CCD detector for maximum sensitivity, an Invictus NIR semiconductor laser (785 nm) and holographic grating to provide fast, simultaneous full spectral collection of Raman data. The laser
RESULTS AND DISCUSSION
a-Quartz Quartz is commonly found in ancient ceramics.19 ˛-Quartz is a three-dimensional polymer in which the SiO4 tetrahedral units are linked throughout the crystal. Trace I in Fig. 1 shows the Raman spectrum of a crystal of quartz on the surface of sample A. A micrograph of the crystal, from which the spectrum was obtained, is shown to the right of the trace. The spectrum is characterized by a prominent band due to the symmetric bending vibration (Si–O–Si) at 463 cm1 , medium to weak bands due to lattice modes at 126, 199 and 262 cm1 and weak bands at 356 and 401 cm1 due to asymmetric bending modes of the silica tetrahedra. It has also weak bands due to an Si–O–Si bending mode at 804 cm1 and to an asymmetric stretching mode of the silica tetrahedra at 1085 cm1 . The observed bands are in good agreement with the characteristic Raman modes of quartz, (D33 ) space group, reported in the literature.20 Small traces of anatase with a characteristic strong band at 146 cm1 were detected in the same spectrum, indicating the close proximity of these compounds on the surface (within 16 µm). The strongest quartz band of 463 cm1 can be detected on the surface and in the cross-section of both samples.
Titanium dioxide The low-temperature crystal form of TiO2 , anatase, is commonly found in both samples. It is characterized by
(I)
Raman Intensity
830
(II)
10 µm
10 µm
(III) 10 µm
(IV) 200
400
600
800
1000
1200
10 µm
Wavenumber/cm-1
Figure 1. Raman spectra of (I) ˛-quartz, (II) anatase from sample B, (III) rutile from sample A and (IV) calcite from sample B. Micrographs of the crystals from which the spectra were collected are shown next to the traces.
Copyright 2005 John Wiley & Sons, Ltd.
J. Raman Spectrosc. 2005; 36: 829–833
Micro-Raman spectroscopy of 2500 BC pottery
Plate 1. (a) Photograph of the samples analyzed. (b) Photograph of some of the vessels from the Lapatsa tomb.
Copyright 2005 John Wiley & Sons, Ltd.
J. Raman Spectrosc. 2005; 36
Micro-Raman spectroscopy of 2500 BC pottery
order of increasing wavenumber: 226 (s), 246 (vw), 294 (vs), 411 (s), 505 (w), 612 (m) cm1 . The broad band at 499 cm1 in spectrum II together with the broader and less intense bands around 350 and 737 cm1 marked with asterisks, are most likely a result of -Fe2 O3 , maghemite. Because of the structural difference between hematite (trigonal symmetry) and maghemite (cubic symmetry), their Raman spectra are easily distinguished. The three characteristic bands for maghemite are in good agreement with those reported by other workers24,25 implementing either 632.8 or 785 nm Raman excitation wavelengths. Optical images of the hematite grains from which spectrum I (sample A) and spectrum II (sample B) were collected are shown next to the Raman spectra. The size of the hematite grains shown in these two images is representative of that of the grains in samples A and B. We established that the average grain size of the hematite in the lighter colored sample B, around 20 µm, is different to that of the more vibrant red and lustrous sample A, in which we detected a higher concentration of hematite grains with size