Characterisation of a prehistorical ceramic object

J Therm Anal Calorim (2014) 116:641–645 DOI 10.1007/s10973-014-3712-8 Characterisation of a prehistorical ceramic object Moon-shaped idol, by means of thermal analysis Alexandra Kloužková • Martina Kohoutková Petra Zemenová • Zdeněk Mazač • Received: 1 October 2013 / Accepted: 17 February 2014 / Published online: 18 March 2014  Akadémiai Kiadó, Budapest, Hungary 2014 Abstract Archaeological ceramics findings are usually characterised by a combination of methods to provide as many as possible information about their origin, use and deposition. Unique moon-shaped idol approximately 0.5 m long from the Final Bronze Age was studied by XRD, XRF, Raman spectroscopy and thermal analyses (TG, DTA, DSC). Special attention was given to processes occurring during its thermal treatment, which were specified by thermal analysis. It was proved that the process of kaolinite dehydroxylation proceeded less intensively in the central part of the object and the maximum of peak was shifted to lower temperature compared to the border parts. It is supposed that the moon-shaped idol was thermally treated not until its use, and the border parts of the object were exposed to lower temperatures compared to the central part. Keywords Prehistoric ceramics  Moon-shaped idol  Thermal analysis  XRD  XRF A. Kloužková  P. Zemenová (&) Department of Glass and Ceramics, Faculty of Chemical Technology, ICT Prague, Technická 5, 166 28 Prague, Czech Republic e-mail: petra.zemenova@vscht.cz A. Kloužková e-mail: alexandra.klouzkova@vscht.cz M. Kohoutková Central Laboratories, ICT Prague, Technická 5, 166 28 Prague, Czech Republic e-mail: martina.kohoutkova@vscht.cz Z. Mazač Regional Museum in Kolı́n, Brandlova 35, 280 02 Kolı́n, Czech Republic e-mail: mazac@muzeumkolin.cz Introduction Archaeological ceramic findings from various historical periods have been a subject of many studies in recent years. Results of analytical methods enable to characterise ceramic materials and have become an important aspect for their interpretation. Ceramic shards are usually characterised by a combination of method such as XRF, XRD, OM, IR, Raman and TA [1–3]. Chemical composition is the most frequently determined by XRF analysis, sometimes supplemented by ICP. Mineralogical composition is usually characterised by XRD analysis and by optical microscopy providing additional information about microstructure of a material. Thermal analyses (DTA/DSC-TG) help to describe processes proceeding in a studied material (dehydration, dehydroxylation, transformations, decompositions, formation of new phases, etc.). In case of prehistoric low-firing ceramics containing clay minerals or their non-crystalline residues, processes of dehydration, dehydroxylation and rehydroxylation, respectively, are considered. Process of dehydroxylation, when kaolinite transforms to an unstable non-crystalline product, metakaolinite (Al2Si2O7), after calcination at 450–600 C, is described by following equation [4]: Al2 O3 2SiO2 2H2 O ! Al2 O3 2SiO2 þ 2H2 O: A process of rehydroxylation is reverse to dehydroxylation, it is induced by the influence of moisture and accompanied by volume expansion of a ceramic body. This process can be simulated in laboratory conditions by a hydrothermal treatment in autoclaves and is called hydrothermal ageing [5]. Morphology of a possible surfacing is usually studied by optical or electron microscopy. Colouring, binders or 123 642 impurities can be identified by IR or Raman spectroscopy [6, 7]. Moon-shaped idols are large ceramic objects of trapezoidal shape usually finished by horns at the borders. These artefacts or their remains have been found in the Central Europe in the context of settlements of the Final Bronze Age. Later they became a part of burial equipments in the form of miniature imitations in the Late Iron Age. Owing to their unusual shape, various theories exist about their possible functions. The objects could have a cult-religious character or they could have a practical use. The use as an andiron (a device which hold up the firewood and so allowing proper burning) is often discussed [8–10]. Prehistoric ceramics were produced from natural raw materials from local sources. They were fired at temperatures below 1,000 C and consequently they contained mainly quartz, micas, feldspars and carbonates. Clay minerals (kaolinite, illite, etc.) are usually present in the form of reactive non-crystalline phases, but their presence in crystalline form is also possible depending on the firing temperature [11]. On the basis of the presence of particular minerals, the approximate firing temperature can be estimated [12]. The aim of this work was to characterise the composition, homogeneity and surfacing of a unique archaeological ceramic finding—the moon-shaped idol from Přemyšlenı́. The presence of accessory substances in the prehistoric ceramic material and the processes occurring during its thermal treatment were specified by thermal analysis. Experimental The moon-shaped idol from Přemyšlenı́–Zdiby, near Prague in the Czech Republic was found in a store hole during an archaeological excavation of a prehistoric settlement from the Final Bronze Age (Reinecke HaB stage). This period in central Bohemia is characterised by the Štı́tary culture. Moon-shaped idols of the Final Bronze Age are known especially from settlements, in a small scale also from burial grounds, of the Western and Middle Europe [8, 13, 14]. Moon-shaped idols of the Štı́tary culture are usually undecorated, long ceramic blocks, trapezoidal in section, with horns at both ends [9, 10]. The function of these objects was not so far satisfactorily resolved. It is supposed that their use could be ritual or practical (e.g. as an andiron) [13, 14]. Considering its size and wholeness, it is regarded as a unique finding (Fig. 1) [9, 10]. Samples for analyses were taken from five different sites of the object—three body samples from the central (A) and border parts (B, C) (Fig. 1) and two compact samples from surfaces of the central and border parts. The body samples 123 A. Kloužková et al. Fig. 1 The picture of moon-shaped idol from Přemyšlenı́–Zdiby (taken by Z.Mazač) A, B, C were ground in an agate mortar to the form of fine powder, well dried and used for the following analysis: • • • • X-ray fluorescence analysis (XRF), sequential WD-XRF spectrometer ARL 9400 XP?, X-ray diffraction analysis (XRD) in the range of 5–602h, diffractometer PANalytical X0 pert Pro with Cu anode, XRD patterns were evaluated by the program X0 pert High Score Plus, Infrared spectroscopy (IR) in the range of 400–4,000 cm-1, spectrometer Nicolet IS 10, Thermo Scientific, ATR crystal of ZnSe, Thermal analysis (DTA, TG, DSC)–STA Setaram Setsys Evolution 16 system using 25 ± 0.02 mg of a sample and the heating rate of 10 C min-1 in the temperature range of 20–1,200 C in argon flow. H2O and CO2 release was measured by mass spectrometer Omni Star, Pfeiffer Vakuum, in the range of 300 AMU. Compact samples from the surfaces were used for Raman spectroscopy (RS)–Labram HR spectrometer Jobin–Yvon, with the resolution 100–4,000 cm-1. Hydrothermal ageing (HA) was induced by hydrothermal conditions in Teflon lined autoclaves at 230 C (2.77 MPa) for 10 and 100 h. A solid/solution ratio 1 g sample/20 mL water was used in the experiments. The samples from border parts were calcined at 600 C for 2 h before the hydrothermal treatment to simulate conditions closely after firing of the archaeological object. [5]. Results and discussion The chemical compositions of the samples were determined by XRF. Measured data of the main elements were adjusted according measurement error, recalculated to 100 % and are presented in oxides (Table 1). Chemical compositions of the samples were very similar. Small differences were noticed only in the contents of CaO, MgO and K2O, which demonstrates a good homogeneity of the raw material. Characterisation of a prehistorical ceramic object 643 Table 1 Chemical composition (main elements) of the samples determined by XRF [mass %], presented in oxides and recalculated to 100 % SiO2 Al2O3 Fe2O3 TiO2 CaO MgO K2O Na2O A (center) 68.28 22.30 3.49 0.88 1.52 1.08 2.26 0.20 B (border) 68.28 22.80 3.38 0.90 1.10 1.28 2.15 0.16 C (border) 68.01 22.63 3.50 0.90 1.27 1.28 2.23 0.22 showed in Fig. 3. In case of the border parts, the loss of water vapour was monitored in two temperature areas: quartz (1) mica/illite feldspars (2) kaolinite 10 20 30 40 50 60 Position/°2θ Fig. 2 XRD patterns of the original samples from the central (A) a border parts (B, C) The phase composition was determined by XRD analysis which is presented in Fig. 2. The main crystalline phases in all three samples were quartz, K-feldspar and micas (muscovite, biotite). The principal difference between the central and border parts was in the presence of the clay mineral kaolinite, which was identified only in the border parts of the object. The presence of kaolinite in the border parts (B, C) was confirmed by simultaneous thermal analysis (TG, DTA) The sample from the central part (Fig. 4) showed only a very small effect in the dehydroxylation stage (400–600 C). Accordingly, the presence of an unstable product of dehydroxylation–metakaolinite is expected. DTA curves of all three samples showed endothermic peak related to a transformation of quartz (-570 C), then an endothermic peak related to mica dehydroxylation (-880 C) and an exothermic peak, which is related to the crystallisation of a spinel structure (-940 C). The release of CO2 and the corresponding exothermic effect indicated the presence of organic substances. Considering the differences between samples from the central and border parts in the intensity and position of the dehydroxylation peak at DTA curves, more detailed DSC measurements concerning hydrothermally aged samples were performed. Figure 5 shows DSC curves of the original and hydrothermally aged (HA) samples. The endothermic peak, CO2 H2O 0 Exo 1.2 × 10–11 TG 0 DTA 2.0 × 10–12 –5 –2 –10 Heat flow/μV Fig. 3 DTA, TG and H2O and CO2 release curves of the border part (B) 100–200 C related to the loss of remaining loosebound water (in pores) and to a dehydration of kaolinite (removal of physically absorbed water), 400–600 C related to a dehydroxylation of kaolinite (removal of chemically bound water). 8.0 × 10–12 –15 –4 –12 1.5 × 10 Mass/% Zdiby A Zdiby B Zdiby C –20 –6 H2O –25 CO2 4.0 × 10–12 –30 0 200 400 600 800 1,000 1.0 × 10–12 –8 1,200 Temperature/°C 123 644 A. Kloužková et al. Fig. 4 DTA, TG and H2O and CO2 release curves of the central part (A) H2O 0 Exo CO2 –11 1.2 × 10 TG 0 DTA 2.0 × 10–12 –5 –2 8.0 × 10–12 –15 –4 –12 1.5 × 10 Mass/% Heat flow/μV –10 –20 –25 –6 H2O CO2 4.0 × 10–12 –30 0 200 400 600 800 1,000 1.0 × 10–12 –8 1,200 Temperature/°C 0 0.45 Zdiby A – center Zdiby C – border 0.40 0.35 Absorbance/a.u. Heat flow/μV mg–1 –0.01 –0.02 –0.03 –0.04 –0.05 300 Zdiby A – center Zdiby A – center + HA (10 h 230 °C) Zdiby C – border Zdiby C – border – 600 °C (2 h) + HA (10 h 230 °C) Zdiby C – border – 600 °C (2 h) + HA (100 h 230 °C) 350 400 450 500 550 0.30 0.25 0.20 0.15 0.10 0.05 600 Temperature/°C 0 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 Wavenumber/cm–1 Fig. 5 DSC curves of the original and treated samples Fig. 6 IR spectra of the original samples which is characteristic for the dehydroxylation, of the original sample from the central part (A) is shifted to lower temperature compared to the original sample from border part (C). The dehydroxylation peak of the sample A after hydrothermal treatment (10 h at 230 C) is shifted to higher temperature and has higher intensity compared to the original sample. Similar peak was measured for the sample C calcined 2 h at 600 C and subsequently hydrothermally treated 10 h at 230 C. The shift in the dehydroxylation temperature is related to the difference in the OH–OH distance and to the disorder of the kaolinite structure, i.e. dehydroxylation temperature decreases with the increasing disorder of the kaolinite structure. The DSC curve of the sample C after calcination and hydrothermal ageing at 230 C for 100 h shows that the structure of the rehydroxylated kaolinite gets closer to the original structure with the increasing time of the hydrothermal treatment. IR spectroscopy (Fig. 6) confirmed the presence of kaolinite in the samples from the border parts via the characteristic positions of OH bonds in kaolinite between 3,620 and 3,690 cm-1 [15]. The moon-shaped idol had a dark colouration on the surface of the border parts (Fig. 1). This surface layer was identified as a graphite coating by Raman spectroscopy (Fig. 7). Graphite coating was intentionally used as a type of decoration already from prehistoric times. On the contrary, amorphous carbon was identified in the central part down to a depth of at least 50 lm, where the measurement was performed (Fig. 8). It could be of secondary origin caused by the use of the object near an open fireplace. 123 Characterisation of a prehistorical ceramic object 645 primary source. It is supposed that the moon-shaped idol was thermally treated not until its use near to an open fireplace and the border parts were exposed to apparently lower temperatures compared to the central part. The surface of the object was decorated on the borders by a graphite coating. 700 1,319 Intensity/a.u. 600 1,577 500 400 300 Acknowledgements Financial support from specific university research (MSMT No. 20/2013). 200 100 0 400 600 800 1,000 1,200 1,400 1,600 1,800 Wavenumber/cm–1 Fig. 7 Raman spectra of graphite in the sample of the surface of border part (B) 2.000 1,351 1,584 1.800 Intensity/a.u. 1.600 1.400 1.200 1.000 800 600 400 200 0 400 600 800 1,000 1,200 1,400 1,600 1,800 Wavenumber/cm–1 Fig. 8 Raman spectra of amorphous carbon in the sample of the surface of central part (A) Conclusions A moon-shaped idol was characterised by XRF, XRD, IR, RS and thermal analysis. It contained crystalline phases which are typical for low-firing ceramics, mainly quartz, feldspars, micas and clay minerals. The clay mineral kaolinite was identified only in the border parts of the examined archaeological object by XRD. It was proved that the process of kaolinite dehydroxylation proceeded less intensively in the central part of the object and the maximum of peak was shifted to lower temperature compared to the border parts. It was concluded that the dehydroxylation of kaolinite due to a primary firing of the object occurred only in its central part. Kaolinite, which was identified in the central part, was a product of rehydroxylation caused by long time storage of the object under the ground. On the other hand, kaolinite identified in border parts was from the References 1. Campanella L, Favero G, Flamini P, Tomassetti M. Prehistoric terracottas from the Libyan tadrart acacus: thermoanalytical study and characterization. J Therm Anal Calorim. 2003;73:127–42. 2. Papadopoulou DN, Lalia-Kantouri M, Kantiranis N, Stratis JA. Thermal and mineralogical contribution to the ancient ceramics and natural clays characterization. J Therm Anal Calorim. 2006;84(1):39–45. 3. Ion RM, Dumitriu I, Fierascu RC, Ion MC, Pop SF, Radovici C, Bunghez RI, Niculescu VIR. Thermal and mineralogical investigations of historical ceramic. J Therm Anal Calorim. 2011;104:487–93. 4. Wang H, Li C, Peng Z, Zhang S. Characterisation and thermal behavior of kaolin. J Therm Anal Calorim. 2011;105:157–60. 5. Zemenová P, Kloužková A, Kohoutková M. Ageing of low-firing prehistoric ceramics in hydrothermal conditions. J Process Appl Ceram. 2012;6(1):59–64. 6. Edward HGM, Chalmers JM. Raman spectroscopy in archaeology and art history. Cambridge: The Royal Society of Chemistry; 2005. 7. Striova J, Lofrumento C, Zoppi A, Castellucci EM. Prehistoric Anasazi ceramics studied by micro-Raman spectroscopy. J Raman Spectrosc. 2006;37:1139–45. 8. Mazač Z. Moon-shaped idols of the Late Bronze Age and the Early Iron Age in Bohemia. Bachelor thesis. Charles University Prague; 2000. 9. Zemenová P. Evaluation of microstructure of ceramic moon shaped idols. Diploma thesis. ICT Prague; 2010. 10. Mazač Z, Kloužková A. (2012) Crescent-shaped pedestals. In: M. Kuna, A. Němcová et al., editors. The evidence of settlement discard. Finds from the Final Bronze Age at Roztoky and the depositional analysis of archaeological context. The Institute of Archaeology of Academy of Sciences of the Czech Republic, Prague v.v.i., Prague. 11. Hanykýř V, Kloužková A, Bouška P, Vokáč M. Ageing of historical ceramics. Acta Geodyn Geomater. 2009;6(1):59–66. 12. Rasmussen KL, de La Fuente GA, Bond AD, Mathiesen KK, Vera SD. Pottery firing temperatures: a new method for determining the firing temperature of ceramics and burnt clay. J Archaeol Sci. 2012;39:1705–16. 13. Hager D Einführung in die Aufarbeitung und Auswertung prähistorischer Keramik. Skulpturen der Spätbronzezeit. Seminararbeit. Universität Basel; 2007. 14. Steuer H (1994) Feuerböcke. In: J. Hoops et al.,editors. Reallexikon der Germanischen Altertumskunde Bd. 8, 390–8. 15. Madejová J. FTIR techniques in clay mineral studies. Vib Spectrosc. 2003;31:1–10. 123