Raman spectroscopic study of ancient South African domestic clay pottery

https://doi.org/10.1016/j.saa.2006.02.059Get rights and content

Abstract

The technique of Raman spectroscopy was used to examine the composition of ancient African domestic clay pottery of South African origin. One sample from each of four archaeological sites including Rooiwal, Lydenburg, Makahane and Graskop was studied. Normal dispersive Raman spectroscopy was found to be the most effective analytical technique in this study. XRF, XRD and FT-IR spectroscopy were used as complementary techniques. All representative samples contained common features, which were characterised by kaolin (Al2Si2O5(OH)5), illite (KAl4(Si7AlO20)(OH)4), feldspar (K- and NaAlSi3O8), quartz (α-SiO2), hematite (α-Fe2O3), montmorillonite (Mg3(Si,Al)4(OH)2·4.5H2O[Mg]0.35), and calcium silicate (CaSiO3). Gypsum (CaSO4·2H2O) and calcium carbonates (most likely calcite, CaCO3) were detected by Raman spectroscopy in Lydenburg, Makahane and Graskop shards. Amorphous carbon (with accompanying phosphates) was observed in the Raman spectra of Lydenburg, Rooiwal and Makahane shards, while rutile (TiO2) appeared only in Makahane shard. The Raman spectra of Lydenburg and Rooiwal shards further showed the presence of anhydrite (CaSO4). The results showed that South African potters used a mixture of clays as raw materials. The firing temperature for most samples did not exceed 800 °C, which suggests the use of open fire. The reddish brown and grayish black colours were likely due to hematite and amorphous carbon, respectively.

Introduction

The technique of Raman spectroscopy was used to examine a series of African domestic clay pottery shards. Such a detailed examination of pottery could give valuable information on the characterisation of archaeological items by the identification of materials such as pigments, binders and clays used in the manufacturing process. The importance of this subject has already been demonstrated in the literature [1], [2], [3]. The chemical composition can be used to define the pottery of a particular area and people by determining the raw materials used. For very important artefacts the knowledge of the composition of raw materials and processing methods used will help in the reconstruction of such pieces. For instance, studies have shown that natural sources of Ca2+ in clay may be ash, bedrock clay bodies, shells (containing Ca particles), etc. while sources of P2O5 in clays are dung, blood, fats and lipids [4]. The bushmen are known to anoint their clay vessels with fat while still damp, then to smear them with gum after drying and to boil springbok blood in them after they were fired. The protein-based stem extraction product from animal bones and butcher's offal which is normally referred to as animal glue is a nitrogen-containing material. Animal glue is usually applied to pottery before colouring with metal-containing pigments [5].

There are two ways in which the pottery could be fired: in an open fire and in a kiln. The open firing methods have an upper limit of 800 °C [4]. The kilns were used for iron smelting and could reach a maximum temperature of 1230 °C.

Various techniques have been used to analyse pottery for defining non-local pottery; neutron activation, X-ray diffractometry (XRD), trace elemental analysis, spectroscopy, etc. by chemically and mineralogically characterising the clay paste [6]. However, the paste analysis is complicated by the diversity of clay minerals and trace elements in nature. Chemical analysis is applied in clay pottery for characterisation, as different clay deposits can be distinguished by their chemical composition. Optical emission spectroscopy is used for analysis of minor and/or trace elements [7]. The success of these approaches is constrained by data on regional and interregional distribution of clays and trace elements. Tempering (aplastic) agents in the pottery are often well suited for use as indicators of provenance and origin, even though they could be traded. The aplastics are often used to retain the shape of the pottery controlling the shrinkage during drying and firing [6]. Many tempering agents are not good indicators of source area, e.g. bone and shells, because they are common, but rock fragments and sand grain are better due to the fact that they may reflect geological source.

In this paper, we report on the results obtained by Raman spectroscopy during the determination of the main components of selected South African earthenware (low temperature fired clay pottery) objects in the form of domestic clay pottery shards. The purpose of the study was to characterise the samples on the basis of their chemical composition as obtained from their Raman spectra. This will help determine whether the same raw materials were used, and if so, whether the pottery samples were processed under the same conditions. It is intended that these results will help in revealing the potential of Raman spectroscopy as a tool in analysing earthenware pottery shards and their methods of production in South Africa.

Section snippets

Samples

The samples originate from four areas in South Africa, namely, Lydenburg (Fig. 1), Makahane (Fig. 2), Graskop (Fig. 3) and Rooiwal (Fig. 4). The Makahane collection was obtained from an archaeological site just outside the Kruger National Park. The shards were found at the same location back in the 1930s. These were dated back to the 13th and 14th century, and most likely belonged to the Venda- and Pedi-speaking people who lived in that region. The Graskop collection was found at a place

Results and discussion

The representative shards were Lydenburg sample no. 8, Makahane sample no. 7, Graskop sample no. 5 and the Rooiwal sample. The laser was focused in each recording on coloured areas visible on the object under the microscope. The predominant colours on the shards were black, orange, red and maroon scattered over the uncoloured background. This is normally an indication of the composite nature of clay products. A range of spectra was collected from different spots on each sample. The regions with

Conclusion

Raman spectroscopy has been successfully used to determine the composition of African clay pottery shards from a selected number of South African archaeological sites (Lydenburg, Makahane, Graskop and Rooiwal). Despite the low Raman scattering effects of clay components and fluorescence commonly associated with this type of samples, important results were obtained in this study. A total of 13 chemical phases were identified by Raman spectroscopy in the samples investigated. The identification

Acknowledgements

The financial support by the National Research Foundation in Pretoria and the University of Pretoria is gratefully acknowledged. The authors would also like to thank Annette Weitz from the Centre for Indigenous studies for the supply of samples and historical information.

References (48)

  • J. Poblome et al.

    J. Archaeol. Sci.

    (2002)
  • G.V. Robins et al.

    J. Archaeol. Sci.

    (1983)
  • C.R. Ferring et al.

    J. Archaeol. Sci.

    (1987)
  • R.L. Frost et al.

    Spectrochim. Acta Part A

    (1993)
  • H.G.M. Edwards et al.

    J. Mol. Str.

    (2000)
  • D. Hradil et al.

    Appl. Clay Sci.

    (2003)
  • V.C. Farmer
  • R.L. Frost et al.

    Spectrochim. Acta Part A

    (2000)
  • E.E. Coleyshaw et al.

    Spectrochim. Acta Part A

    (1994)
  • I.A. Degen et al.

    Spectrochim. Acta Part A

    (1993)
  • P. Valfre

    Yixing Teapots for Europe

    (2000)
  • R. Gout et al.

    Geochim. Cosmochim. Acta

    (1997)
  • M. Tabbal et al.

    Thin Solid Films

    (2004)
  • R.N. Tarrant et al.

    Diamond Relat. Mater.

    (2004)
  • B. Mihailova et al.

    J. Non-Crystal. Solids

    (1995)
  • C.D. Yin et al.

    J. Non-Crystal. Solids

    (1986)
  • J.V. Owen

    J. Archaeol. Sci.

    (1997)
  • R.R. Shagidullin et al.

    Atlas of IR Spectra of Organophosphorus Compounds (Interpreted Spectrograms)

    (1990)
  • R.J.H. Clark

    Chem. Soc. Rev.

    (1995)
  • R. Davey et al.

    J. Raman Spectrosc.

    (1994)
  • C.A. Bollang et al.

    J. Archaeol. Sci.

    (1983)
  • P.S. Peacock

    J. Archaeol. Sci.

    (1976)
  • Labspec, version 2.04, Distributed by Dilor SA & Universite’ de Reims, France,...
  • S. Verryn, Details of XRD Procedure Used at the University of Pretoria, XRD Laboratory, Personal Written Communication,...
  • Cited by (88)

    • When the time stops: The “Grotta dei Cocci” (Terni, Italy)

      2022, Boletin de la Sociedad Espanola de Ceramica y Vidrio
    View all citing articles on Scopus
    View full text