Title: Mineralogy and geochemistry of Mediterranean sand deposits as a raw material for Roman natron glass production.
Other Titles: Mineralogie en geochemie van Mediterrane zandafzettingen als grondstof voor Romeinse natron glasproductie.
Authors: Brems, Dieter
Issue Date: 7-Sep-2012
Abstract: During the Late Roman and Byzantine period, natron glass was made from its raw materials in a limited number of primary production centres in Egypt and Syro-Palestine. For the earlier Hellenistic and Roman period, no primary furnaces have been found and the location of primary production during this era remains unclear. Ancient authors such as Strabo and Pliny the Elder suggest that glassmaking sands were found near the River Belus (Israel), in Egypt, near the mouth of the Volturno River (Italy) and also in Spain and France. However, primary production in the western part of the Mediterranean is not supported by any direct archaeological evidence and possible sand raw materials from these regions have never been evaluated for their suitability to produce glass.In this study the possible existence of a Roman primary glass industry in the western Mediterranean is investigated, based on the occurrence of suitable sand raw materials. 178 beach sands from Spain, France and Italy are evaluated for their suitability for glass production by calculating the composition of hypothetical glasses made from these sands and comparing them to Roman natron glass. The results show that good glassmaking sands are far from common. Only a very limited number of the 178 beach sands analysed would produce a glass with major and minor elemental compositions within the ranges of Roman imperial natron glass. The rest of the sands analysed are unsuitable for glass production in their present form due to their insufficient SiO2, high Al2O3 and Fe2O3, and either too low or too high CaO contents. If the sand raw material was too low in CaO, extra lime could be added to the glass batch in the form of shell or limestone. This was taken into account in a second calculation.Overall, six limited areas could be define where suitable sand raw materials would have been available to the Roman glassmaker. Good glassmaking sands occur in the Basilicata and Puglia Regions (SE Italy), and Tuscany (W Italy). After the addition of an extra source of lime, sands from the Huelva Province (SW Spain), the Murcia Region (SE Spain) and from the Provence (SE France) would also produce glasses with a typical Roman composition. The suitability of the suggested sands was checked by performing a series of melting experiments, which provide further insights in the way Roman glass makers would have had to prepare their raw materials and the used batch recipes.Because of the complex nature of Roman natron glass, the provenance determination of this material remains one of the most challenging problems in the field of archaeometry. In the past decades, several attempts have been made to determine the provenance of ancient glass based on major, minor and trace elemental compositions, rare earth element patterns and isotopic signatures of O, Pb, Sr and Nd. Unfortunately, the major elemental composition of natron glass is relatively uniform and specific objects can almost never be uniquely assigned to their origin. Although the use of Sr and Nd isotope ratios and trace element signatures as an indication of provenance has proven promising, there are still unknown factors.The bulk of the Sr in natron glass is believed to have been incorporated with the lime bearing material. Where the lime was derived from Holocene seashell, the Sr isotopic composition of the glass would reflect that of modern seawater. Where the lime was derived from ‘geologically agedÂ’ limestone, the signature would reflect that of seawater at the time the limestone was deposited, possibly modified by diagenesis. However, also other minerals in the sand, such as feldspar and mica, can influence the Sr budget of the glass batch.Nd in natron based glass originates from the non-quartz mineral fraction of the silica raw material, which is most probably sand. It is well known that the εNd value of deep-sea sediments in the eastern Mediterranean varies significantly due to the varying sediment influx from the Nile (fluvial), the Sahara (eolian) and the European continent (fluvial) and shows a pronounced E-W gradient from as high as -1 at the mouth of the River Nile and the coasts of Egypt and Israel, to -12 south of Sicily. In the western Mediterranean, the distribution of the Nd isotopic composition is less well known. Only very few results for river particulates and deep sea sediments are published and they all show rather low εNd value of around -10. Although it is not possible to directly compare the Nd isotopic signature of glass to sea-floor sediment, such significant differences between the easternmost part of the Mediterranean Sea and the rest of the basin offer a great potential in tracing the origins of sand raw materials used in primary glass production.Trace elements in raw natron glass are related to the accessory minerals in the glassmaking sand. The mineralogy of the heavy mineral fraction of a sand deposit is determined by the composition of the source rocks and is often a good indication of its origin. However, also the other raw materials, such as the natron and the lime, introduce traces of specific elements to the glass batch. Especially the transition metals, such as Mn, Co, Cu, Sb and Pb, are strongly influenced by the addition of colouring or decolouring agents. Recycling of glass cullet, including small quantities of coloured glass, results in elevated concentrations of these elements.In this study, the applicability of Sr and Nd isotopes and trace elemental analysis for provenancing natron glass is evaluated. A database of Sr and Nd isotopic compositions of possible sand raw materials from the western Mediterranean is presented, as a means of comparison for the growing number of isotopic studies on ancient glass. It is shown that the 87Sr/86Sr ratio in natron glass is significantly influenced by the silicate fraction of the sand used and does not always provide a clear indication of the lime source used. The 143Nd/144Nd isotope ratio and εNd value of sands is a good indicator for their geological (and sometimes geographical) provenance.The variation of the isotopic composition of Nd in beach sands along the Mediterranean coasts of Spain, France and Italy is presented, and the relation between the Nd isotopic signatures and geologically distinct regions is explained. The use of the isotopic signature of Nd as a proxy for the source of silica in glass is, however, not always straightforward because of the possible overlap of signatures from different suppliers. However, when used in combination with trace element geochemistry, raw natron glass and sand raw materials from the known production centres in the Levant can be readily distinguished from the suitable glassmaking sands identified in the western part of the Mediterranean. Ti, Cr, Sr, Zr and Ba are shown to be the most useful elements for provenancing natron glass.A comparison between the range of the Nd isotope ratios of suitable glassmaking sands and archaeological natron glass from around the Mediterranean, indicates primary production in both the east and west. Hellenistic and Roman (2nd century BC to 4th century AD) natron glasses show a wide range of Nd isotopic compositions. This may suggest raw glass production all around the Mediterranean Sea but can also reflect intense recycling of glass from a few big glass producers scattered around the Mediterranean. During the Late Roman, Byzantine and Early Islamic period (5th to 7th century AD) large scale raw glass production appears to be restricted to the eastern Mediterranean.
ISBN: 978-90-8649-546-7
Publication status: published
KU Leuven publication type: TH
Appears in Collections:Division of Geology

Files in This Item:
File Status SizeFormat
PhD Brems Dieter.pdf Published 2930KbAdobe PDFView/Open Request a copy

These files are only available to some KU Leuven Association staff members


All items in Lirias are protected by copyright, with all rights reserved.