Slag Production | Archaeomagnetism of Slag | Results

A project undertaken by Ilana Peters, Dr. Erez Ben-Yosef, and Professor Lisa Tauxe focused on using archaeomagnetic experiments in order to date the age of copper smelting activity in the southern Levant at Site 34 ("Slaves' Hill") and Site 30a from the central Timna Valley. The experiments dated different slag deposits at Site 34 and 30a via reconstructing their ancient geomagnetic intensities as recorded by the individual slag samples at the time of their formation. The correlation between the location of the slag mounds and their dates reflects varying socio-economic and political dynamics of the region.  Furthermore, in comparing the new data with previous archaeomagnetic studies from the nearby Site 30, we can review whether there was simultaneous copper production at Sites 30, 30a, and 34, which would give further support to the claim of intense smelting in the central Timna Valley during the early Iron Age.


Fig. 1. Timna Sites 34 and 30 in red (Rothenberg 1990)

Pictured below are the slag mounds of Site 34, numbered 1 through 19, and Site 30a:


Fig. 2. Site 34 (Photograph provided by Uzi Avner)



Fig. 3. Site 30a (Photograph provided by Uzi Avner)


Slag Production

Slag is produced as part of the copper-smelting process; it is the waste by-product of the smelting process, as shown in the diagram below. Within the oven, after being heated to 1100°C, the denser metal sinks to the bottom, whereas the siliceous waste floats to the top and solidifies as slag.


Fig. 4. Copper metal production process

The project uses the rapid changes in the Earth's magnetic field in order to date the slag. Earth's geomagnetic field represents a vector quantity, which has both a direction and magnitude.  These data are represented by three geomagnetic attributes, (1) inclination, (2) declination, and (3) intensity, which together produce a vector representation of the geomagnetic field specific to a time and place (Sternberg 1997).  Inclination and declination are two directional coordinates, whereas intensity is not a spatial coordinate.  Additionally, each coordinate changes over time in a manner independent from the other two.  Therefore, each type of coordinate can outline a range of potential dates for the object. 

Archaeomagnetism of Slag


Fig. 5. Magnetic field (Glazmaier et al. 1995)

However, this project concentrates solely on intensity.  The other coordinates, being directional, are possible to measure only if the object has not moved since its last heating event, for example an intact kiln.  Intensity is the sole coordinate that may be measured after an object has moved since its last heating (e.g., Selkin and Tauxe 2000). 

Calculating ancient intensity values is based on the principle that the magnetization acquired by an object during cooling is linearly proportionate to the intensity of the Earth’s ancient geomagnetic field. We can calculate the ancient geomagnetic field's intensity with the following equation:


The ancient magnetization of the artifact (in this project, slag) is first measured. Then the artifact is reheated in a known magnetic field (called “lab magnetic field” in this equation) and the artifact’s magnetization is remeasured, giving us the lab magnetization of the artefact. Thereafter we can solve for the last variable, the ancient geomagnetic field.

Fig. 6. Placement in oven-regular measurements

The artifacts are placed in the oven as shown in the diagram above. Each cylinder pictured holds a maximum of 24 specimens. Each specimen is placed in the same spot inside the oven during every heating, with the last being cylinder (numbered 3) being placed directly above the first two cylinders.
However, the experiment is complicated since reheating the artefact may change its ability to acquire magnetization due to alterations in its crystal lattice (Tauxe 2010). In order to determine whether this has occurred, the artifact is heated in a step-wise fashion (Thellier and Thellier 1959). This project uses a step-wise method called IZZI, an experimental protocol based on the Thellier-Thellier method (Tauxe and Staudigel 2004). As explained above, this method involves heating to a temperature at a known magnetic field (an in-field), then at a magnetic field of zero (a zero field), the IZ step (Figure 14). The ZI step is heating the object in a zero field, after which it is heated in an in-field. 

Included in this protocol are several tests to determine sample quality, namely the pTRM checks and pTRM tail checks (Tauxe 2010, Chapter 10). A pTRM check is conducted in order check whether the specimen's capacity to acquire magnetism has changed during the experiment. A visual representation of these steps is shown below. The samples are heated to a low temperature (T1) in an in-field, reheated to a higher temperature (T2) and then reheated to the low temperature (T1) at the same magnetic field as before. For example, the samples were heated to 100°C at an in-field of 40 mT, measured, reheated to 250°C in a zero field, and measured once more. They were then reheated to 100°C at an in-field of 40 mT, which should give the same data as the first measurement. In contrast, a pTRM tail check is to check whether the magnetization has disappeared. In this check, the sample is heated to a temperature (T1), and then again heated to the previous temperature (T1) but in a zero field.

Fig. 7. IZZI (Ben-Yosef et al. 2008a)

Further tests, anistropy measurements, are taken when the specimen produces results that indicate it is not isotropic. This means the specimen reacted unequally in some directions to the Thellier-Thellier experiment. This is visible in the equal area plot as pictured below. In this example, specimen TP01A4's equal area plot has the points converging towards the center, and so anistropic measurements are unnecessary. However, a different specimen pictured on the right, TP01B3, the points converge to the right of the center of the equal area plot, which is not ideal and this specimen should therefore undergo anistropy measurements.

Fig. 8. Examples of equal area plots

Anistropy measurements are produced by heating the chosen specimens at six different orientations to the in-field. The specimens are heated in six positions to the highest temperature used when measuring magnetism, with a seventh reheating in the same position as the first in order to check whether the specimens' ability to acquire magnetism has altered.

Fig. 9. (1) Aerial view of the stand that holds the specimens, here seen with the magnetic field (β) on the right side


Fig. 10. (2) Specimen (the grey circle) in a vial and the orientation of the vial




Fig. 11. (3) Orientations of a vial during the experiment, in an aerial viewpoint

As seen in the three figures above, the vials are placed in a container holding them upright. They are turned 90° every heating four times, with an additional heating in the same position as the pervious measurements with the sample facing away from the magnetic field, and one heating with the sample facing towards the magnetic field. Shown in one (orientations anisotropy) is the aerial view on how the specimens are positioned during +x, -x, +y, and -y part of the experiment. For example, if the specimens are meant to be orientated to -x, then the inscribed line on the vial will face towards the magnetic field. If the sample is meant to be orientated to +x, then the inscribed line will be turned away from the magnetic field. The temperature and field strength remain the same at each step of the anisotropic experiment. In this experiment, we used a temperature of 580°C and a magnetic field strength of 40 mT.


Once the experiment is completed, we can compare our results to known archaeomagnetic data, as seen below. This slag piece is compa red to intensity values during the early Iron Age calculated from Timna Site 30 and Khirbet en-Nahas in Jordan. We can see that this slag piece dates to throughout the early Iron Age. In this project, samples are compared to a graph similar to the one shown below, but instead of the graph showing just the early Iron Age, the final graph encompasses several periods in which the copper-smelting may have taken place at Timna Site 34 and Site 30a, ranging from the Chalcolithic to the Islamic period. From these graphs we will be able conclude during which period copper-smelting took place at these sites, and consequently during which time period copper-smelting was most intense in the central Timna Valley.

Fig. 12. Example of slags archaeointensity data during early iron age (based on Shaar et al. 2011)


For more in-depth description of the CTV research on Archaeomagnetism view our 2013 poster submission for the Timna Park International Conference HERE

See also: Peters, I., Tauxe, L., Ben-Yosef, E., 2017. Archaeomagnetic dating of pyrotechnological contexts: a case study for copper smelting sites in the central Timna Valley, Israel. Archaeometry (DOI: 10.1111/arcm.12322)