PLoS ONE (Jan 2023)

Applying thermal demagnetization to archaeological materials: A tool for detecting burnt clay and estimating its firing temperature.

  • Yoav Vaknin,
  • Ron Shaar,
  • Oded Lipschits,
  • Adi Eliyahu Behar,
  • Aren M Maeir,
  • Erez Ben-Yosef

DOI
https://doi.org/10.1371/journal.pone.0289424
Journal volume & issue
Vol. 18, no. 10
p. e0289424

Abstract

Read online

Burnt materials are very common in the archaeological record. Their identification and the reconstruction of their firing history are crucial for reliable archaeological interpretations. Commonly used methods are limited in their ability to identify and estimate heating temperatures below ~500⁰C and cannot reconstruct the orientation in which these materials were burnt. Stepwise thermal demagnetization is widely used in archaeomagnetism, but its use for identifying burnt materials and reconstructing paleotemperatures requires further experimental verification. Here we present an experimental test that has indicated that this method is useful for identifying the firing of mud bricks to 190⁰C or higher. Application of the method to oriented samples also enables reconstruction of the position in which they cooled down. Our algorithm for interpreting thermal demagnetization results was tested on 49 miniature sun-dried "mud bricks", 46 of which were heated to a range of temperatures between 100⁰C to 700⁰C under a controlled magnetic field and three "bricks" which were not heated and used as a control group. The results enabled distinguishing between unheated material and material heated to at least 190⁰C and accurately recovering the minimum heating temperature of the latter. Fourier-Transform Infrared Spectroscopy (FTIR) on the same materials demonstrated how the two methods complement each other. We implemented the thermal demagnetization method on burnt materials from an Iron Age structure at Tell es-Safi/Gath (central Israel), which led to a revision of the previously published understanding of this archaeological context. We demonstrated that the conflagration occurred within the structure, and not only in its vicinity as previously suggested. We also showed that a previously published hypothesis that bricks were fired in a kiln prior to construction is very unlikely. Finally, we conclude that the destruction of the structure occurred in a single event and not in stages over several decades.