Specifically, we aimed to first perform Raman spectroscopy towards carbon-bearing pottery and yields insights on the application of this technique for estimating maximum firing temperatures of Late Bronze Age vessels from Upper Nubia, comparing two site-specific data sets of samples, from the 18th Dynasty (1550–1290 BCE) at Sai Island and Dukki Gel (Kerma).
In testing Raman spectroscopy, our principal aims were to search for differences in producing technique and firing temperatures between the Nubian- and Egyptian-style samples; between the samples from Sai Island and those from Dukki Gel; and eventually among the different ceramic wares and types.
Now, as it is often the case, one idea leads to another and while some of the archaeological questions posed in the initial objectives of the work have not yet been fully answered, new exciting questions have arisen during the research and led us to further expand our Raman project into a new spin-off project focused on the investigation of the effects of oxidative weathering in relation to site-specific depositional environment and time.
Generally speaking, oxidative weathering is known to cause significant alteration modifications of the ceramic body after deposition resulting in distorting Raman spectra and potentially leading to an overestimation of the firing temperatures (Deldicque et al., 2023). Although all archaeological ceramics experienced oxidative weathering, our case study showed that the weathering effect was possibly more intense on the Dukki Gel than on the Sai Island samples hence to affect maximum firing conditions.
The new set of samples for which thin-sections were produced to perform Raman Spectroscopy.
In the coming months, together with Fabian Dellefant, we will particularly focus on this aspect of our research and perform Raman Spectroscopy on new samples, including ceramic sherds that have undergone extremely intensive oxidative weathering and experimental replica which were never exposed to any depositional and post-depositional processes.
Stay tuned to know more about the development of our work!
Reference
Deldicque, D., Rouzaud, J., Vandevelde, S., Medina-Alcaide, M.A., Ferrier, C., Perrenoud, C., Pozzi, J., Cabanis, M., 2023. Effects of oxidative weathering on Raman spectra of charcoal and bone chars: consequences in archaeology and paleothermometry. Comptes Rendus. Geoscience 355, 1–22. https://doi.org/ 10.5802/crgeos.186.
At the beginning of June, we have announced the launch of a new pilot study – as part of the WP 3 of the DiverseNile project – implementing and testing the potential of the Raman Spectroscopy technique on a selection of ceramic samples coming from our reference collection from the sites of Sai Island (SAV/S-samples) and Dukki Gel, Kerma (DG-samples).
This study is currently in progress as part of our cooperation with the Department of Earth and Environmental Sciences of the LMU, and namely with Fabian Dellefant, co-author of this post, who is a geoscientist and doctoral student under the supervisors of Prof. Dr. Trepmann and Prof. Dr. Gilder.
Due to fast measurements and its non-destructive approach with only little sample preparation, Raman spectroscopy can be easily applied to ancient ceramic materials, answering various technological questions, in particular on the manufacturing stages of production and firing of the pots.
For our investigation, we analysed so far a total number of 8 samples/thin sections (namely samples DG-18, DG-23, DG-29, DG-35, SAV/S 02, SAV/S 14, SAV/S 17, and SAV/S 51). Of all these, micro photos were primarily taken under the petrographic microscope with both transmitted and reflected light in order to select the areas of the sample to be examined with Raman Spectroscopy (normally two different spots including the clay matrix and particular organic components, both within the inner portion or core of the sample and on the rim area).
These samples are either locally produced cooking pots or other local ware manufactured both according to the so-called Nubian (DG-18, DG-23, DG-29, and SAV/S 02) and Egyptian style (DG-35, SAV/S 14, SAV/S 17, and SAV/S 51). All of them consist of a non-calcareous optically active clay matrix with dark cores and red or buff oxidised surfaces. In some specimens, the oxidised margins are narrow and well defined, while in others the red-black zonation appears larger and less regular. The Egyptian style samples normally show a kind of “sandwich” structure consisting of a dark core enclosed, both above and below, by red oxidised surfaces (Fig. 1).
Fig. 1 Photos of the fractures from thin section scanning of samples SAV/S 02 (left) and SAV/S 17 (right). Note the large amount of organic inclusions which have been totally or partially carbonized and are surrounded by voids. Both samples show a dark core due to insufficient penetration of oxygen during firing.
All these samples contain, in a different extent, organic matter either plant remains (chaff, straw, grass and possibly various cereals components), and probably herbivore manure (those finely divided straw particles). The organics are either totally or partially carbonized so that the plant inclusions are often preserved as black carbonized relics into the voids.
The carbonaceous core (dark-grey zone in the center of the ceramic samples) can be the result of insufficient firing under oxidizing conditions. It is also related to the use of a paste of high organic component. During the firing of the pot, the combustion of the organics acts indeed as a reducing agent, taking away oxygen from the firing environment (Velde and Druc 1999: 126-127, see also Quinn 2013).
In organic chemistry, the process of thermal decomposition, obtained by the application of heat and in the complete absence of an oxidizing agent is known as pyrolysis or graphitization.
Pyrolysis-GC/MS to ceramics which are conspicuously black or exhibit a black inner core from incomplete burn-out has been applied for the assessment of molecular properties of organic matter in archaeological pottery matrix (see Kaal et al. 2013).
In Raman Spectroscopy, vibrational modes of specific crystallographic components are used to determine a specific crystallographic structure. In our case, the temperature-dependent formation of graphite is used to quantify the highest temperature the sample has experienced.
The lab setup consists of an optical microscope with different magnifications and a computer software, which handles data acquisition (Fig. 2). Measurements are conducted by using a laser with a 532 cm-1 wavelength directly on the thin section which has been first well-polished and cleaned with ethanol. In the lab, temperature is kept constant at 18° C degrees with the lights turned off so as not to interfere with the measurement.
Fig. 2 Lab of the Museum Mineralogia in Munich with the optical microscope (right) and the computer (left), which were used for the investigation.
In the investigated ceramics, the precursor of the measured graphite can be either organic material, such as grass and straw, or dung of herbivores, which was mingled into the clay before heating. Furthermore, in some samples firing ash could have been added as well. Our preliminary results show that graphite can be clearly detected in the sample material. Interestingly, a group of samples showed graphite formation only within the organic components, which is interpreted as being the relicts from plant inclusions. Other samples clearly show graphite spectra also within the clay matrix, which could have been added to the clay as ash in the first place (see e.g., SAV/S 02, Fig. 3).
Fig. 3 Optical microphotograph with reflected light of sample SAV/S 02. Datapoint 12 (pink) marks an organic component in a void and refers to the Raman spectra SAV_2_1-r12 in Fig. 4. Datapoint 22 (blue) characterizes the ceramics matrix and refers to the Raman spectra SAV_2_1-r22 in Fig. 4.Fig. 4 Raman spectra of a datapoint from the matrix and an organic component. The spectrum of the matrix refers to datapoint 22 and the spectrum of the organic component refers to datapoint 12 of Fig. 3.
The interpretation of the maximum temperature the sample experienced is based on the ratio of two Raman peaks, which have a wavenumber of ~1390 cm-1 and 1606 cm-1. Given the dataset shown in Fig. 4, the maximum temperature can be estimated to ~600 °C after Guizani et al. 2017.
In the following weeks, we will proceed to the data processing and potential grouping based on the various Raman spectra collected from our pottery samples (we measured on average up to 20-25 datapoints for each sample). This will allow us to develop our preliminary interpretations and come to more specific conclusions on the quality of the organic material added to the paste and the heating temperatures reached during the firing. Eventually we might get insights on the type of clay sources selected to make the pots.
We can maybe spoil a bit things for you, anticipating that possibly some of the examined samples experienced a more homogeneous firing than others, these latter showing otherwise varying temperatures!
References
Guizani, C., Haddad, K., Limousy, L., and Jeguirim, M. 2017. New insights on the structural evolution of biomass char upon pyrolysis as revealed by the Raman spectroscopy and elemental analysis. Carbon 119:519–521. http://dx.doi.org/10.1016/j.carbon.2017.04.078.
Kaal, J., Lantes-Suárez, O., Martínez Cortizas, A., Prieto, B., and Prieto Martínez, M. P. 2013. How Useful is Pyrolysis-GC/MS for the Assessment of Molecular Properties of Organic Matter in Archaeological Pottery Matrix? An Exploratory Case Study from North-West Spain. Archaeometry 56 (S1): 187–207. https://doi.org/10.1111/arcm.12057.
Quinn, Patrick S. 2013 Ceramic Petrography. The Interpretation of Archaeological Pottery and Related Artefacts in Thin Section, Oxford.
Velde, Bruce and Druc, Isabelle C. 1999. Archaeological Ceramic Material. Origin and Utilization, Berlin.
If there is something that the Covid-19 pandemic has taught us is resilience, work flexibility and mostly the capacity to design alternative solutions to meet the various physical restrictions and newly shaped work conditions and needs. Further, we learned the importance of networks and acquiring skills even in remote formats, and that online (and/or hybrid) classes and conferences can give virtuous outputs as those in presence. Within the framework of our project, a successful example of this is certainly represented by our online Diverse Nile Seminar Series 2021 Cultural Diversity in Northeast Africa.
For me operating within the Work package 3 of the project and principally dealing with laboratory analysis on the material data – ceramic samples – collected in the field, the pandemic has inevitably meant that I had to shift my main focus from the study of fresh excavation data to the study of reference collections. Hence, in the last months my work schedule has been mainly centred on documentation, database archive, and comparison among the various ceramic datasets. Also, the obligatory permanence in Germany (missing the field and the warmness of the Sudanese sun) together with the need to work often via remote or, whenever possible from the lab, pushed me to convey my working goals to search for new theoretical approaches and interpretative inputs, eventually enlarging the spectrum of the analytical competencies and methodologies devoted to the study of the ceramic samples.
In these circumstances the idea was born together with our PI and other colleagues from the Department of Earth and Environmental Sciences of the LMU to cooperate and expand the networking between our departments hence to test together a new analytical methodology for archaeological ceramic material, namely Raman Spectroscopy.
This technique, which took its unusual name after the Indian physicist C. V. Raman who was the first to observe Raman scattering in 1928 and won a Nobel Prize in Physics in 1930 for this discovery, is a molecular spectroscopy procedure which provides information about vibration and rotational states of molecules. It works using the interaction of a source of monochromatic light, normally an intensive monochromatic laser radiation, and the matter of the sample. The largest part (99.99%) of the laser light radiates through the sample, a very small proportion is scattered in all spatial directions (so-called Rayleigh scattering), finally an even smaller part is scattered inelastically (so-called Raman scattering). This latter contains information about the sample, its molecular structure (no the single chemical elements) and specific characteristics of the material (see among others, Spieß et al. 1999; also What is Raman Spectroscopy? | Raman Spectroscopy Principle (edinst.com); Raman spectroscopy – Wikipedia).
For the study of archaeological samples like ceramics, Raman spectroscopy has the advantage of being a non-destructive (only a minimum portion of the sample as the same slide of the thin section is needed), rapid and relatively low-priced technique. However, the high potential of this methodology may collide with the natural heterogeneity of most of the ancient, especially hand-made, ceramic manufactures (Medeghini et al. 2014; Vandenabeele & Van Pevenage 2017; see also Legodi & de Waal 2007). This is why, at the moment, our goal consists primarily to observe the methodological potentials of Raman and discern its use for our specific research questions.
For our trial study, we selected ten samples (of which six are ceramics from Sai Island and four from the Dukki Gel’s reference collection). All of them are either locally produced cooking pots or other local ware manufactured both according to the so-called Nubian and Egyptian style (Figure 1). In testing this new analytical technique, our main aims are the following: to search for differences in producing technique and firing temperatures/regimes 1) between the Nubian and Egyptian-style samples; 2) between the Nubian samples from Sai Island and those from Dukki Gel; 3) between the Egyptian-style samples from Sai Island and those from Dukki Gel; 4) among the different Nubian types (cooking pots with basketry impressions, incisions, and others). In addition, we also want to look at the behaviour of the organics and their carbonization and check for a possibility of a better characterisation of some opaque mineral phases.
Figure 1 – Examples of Nubian cooking pots with basketry impressions from Sai Island (left) and Dukki Gel; Kerma (right).
In the last days, together with the colleague Fabian Dellefant, geoscientist and doctoral student at the Department of Earth and Environmental Sciences of the LMU, we have realized high resolution scans of the selected ceramic thin sections and photographed them at the petrographic microscope under different light conditions (both transmitted cross polarized and plane polarized light, and also reflected light) in order to describe and document the areas which we are ultimately going to analyse by Raman.
Stay tuned to know more about our ongoing work and first results!
Selected references and links
Legodi, M. A. &, de Waal, D. 2007. Raman spectroscopic study of ancient South African domestic clay pottery, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 66, Issue 1, 135-142.
Medeghini et al. 2014. Micro-Raman spectroscopy and ancient ceramics: applications and problems. Journal of Raman Spectroscopy, 45, Issue 11-12, Special Issue: Raman in Art and Archaeology 2013, 1244-1250.
Vandenabeele, P. & Van Pevenage J. 2017. Raman Spectroscopy and the Study of Ceramic Manufacture: Possibilities, Results, and Challenges. In Hunt, Al (Ed.) The Oxford Handbook of Archaeological Ceramic Analysis.
