The Complex Characterization of Ceramics from Lalatepe Archeological Site

doi:10.21203/rs.3.rs-2628746/v1

Download PDF

Research Article

The Complex Characterization of Ceramics from Lalatepe Archeological Site

https://doi.org/10.21203/rs.3.rs-2628746/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

In the present work, four ancient ceramic shreds from the archaeological site Lalatapa, Fizuli region of the Republic of Azerbaijan, and two local raw ceramic pastes were characterized by powder X-ray diffraction (PXRD), scanning electron microscopy and thermal analysis (TG-DTG) techniques. XRD analysis of ceramic sherds reveals that all investigated samples contain similar minerals: quartz, feldspar, and clay. Three samples out of four contain calcite. Based on the traditional approach, it has been assumed that the firing process in these samples stopped before 700°C. The mass loss ratios of samples of ancient ceramics also indicate that reversible dehydroxylation took place in all four samples, thus indicating the initial mild firing conditions. SEM analysis also did not reveal the signs of deep vitrification. The summary of all the applied methods indicates that the ceramic samples were made using a similar manufacturing technology. According to XRD analysis, samples N1 and N4 contain diopside, and samples N2 and N3 maghemite indicate the different origins of the ceramic shreds. Analysis of the raw ceramic mass also did not reveal the presence of these minerals, which may indicate a discrepancy between the origin of ancient ceramic sherds and modern ones.

Ceramics

Thermogravimetry

X-ray diffraction

ancient ceramic

firing temperature

The complex characterization of ancient ceramics most often includes a description of its mineralogical, chemical, and thermal properties [1] [2][3][4][5]. These data can be provided in an interdisciplinary approach using thermogravimetric (TG), thermoluminescence (TL), X-ray diffraction (XRD), and X-ray fluorescence (XRF) methods. The interpretation of the analytical results, in combination with the relevant archaeological information, can lead to essential conclusions in studies of the origin of ceramics [6] and the ancient technologies used [7][8][9].

During firing, ceramics' mineral components undergo phase transitions characteristic of their firing temperatures [10][11]. The results of thermal analysis in combination with powder X-ray diffraction (PXRD) are usually used to estimate the initial firing temperature of ancient ceramics [1], [12]. XRF reveals the chemical composition of ancient ceramics and can be successfully used for the classification of ceramics into groups of similar composition [13].

In this work, TG/DTG, PXRD, and SEM methods were used to analyze four fragments of ancient pottery from Lalatepe archaeological site in the Fizuli district of Azerbaijan and two modern raw ceramic paste samples. This work aims to characterize ancient pottery in terms of its mineralogical composition, as well as to study its thermal behavior and determine the firing temperature. Based on stratigraphic estimations, Lalatepe is a settlement of the Neolithic era, attributed to the first half of the VI millennium BC. It is one of the most significant Neolithic-Eneolithic monuments in the area known as the Caucasus. The choice of this monument as a base for our studies was primarily because it is considerably larger than all the monuments known in this area, and we perceive it to be the center of the nest of settlements.

The Institute of Archeology, Ethnography, and Anthropology of ANAS provided the samples of single fragments of ceramics found during the "Karabakh Neolithic-Eneolithic Expedition" in the territory of Lalatapa, Fizuli region of the Republic of Azerbaijan (samples N1, N2, N3, and N4). Most of these specimens are believed to be from the Neolithic period and may have been used for cooking or preserving food. The samples were air-dried overnight at 50°C before analysis and finely powdered in an agate mortar. Two raw ceramic paste samples were also analyzed as reference material (Samples CP1 and CP2).

PXRD was performed using a D2Phaser (Bruker) diffractometer with Ni-filtered CuKα radiation on randomly oriented samples. The samples were scanned at the region of 5≤2Ɵ ≤75° at a scanning speed of 1.2°/min. Semi-quantitative estimates of the abundance of the mineral phases were derived from the PXRD data, using the intensity of specific reflections, the density, and the mass absorption coefficients of the elements for CuKα radiation.

Thermogravimetric and differential thermal analysis of ceramic powders were carried out in a Perkin Elmer STA6000 Simultaneous Thermal Analyzer with the following parameters: heating ranges from ambient to 950◦C, heating rate 5◦C, balance sensitivity- 0.1µg, and nitrogen gas flow-20 ml/min.

Ceramic samples were investigated using scanning electron microscopy Quattro-C, Thermo Scientific, USA, coupled with EDX microanalysis. The operating regime for the microscope is the following: voltage of 20 kV, the working distance is 10 mm, and the current is 1 nA. At the same time, the entire surface of the samples is scanned at magnifications from 100 to 30,000 times (the microscope's resolution at the last magnification is 0.1 µm).

Row ceramic paste consists of clay (smectites, kaolinites) and tempering materials like sand, calcite, and organic fillers. Because the tempering materials are thermally more stable than clay, one observes changes only in the clay material under mild firing (meaning the firing temperature is ≤700 °C) [14]. It is generally accepted that the loss of mass during clay heating occurs due to (I) dehydration (≤350 °C), (II) decomposition of hydroxyls (350–600 °C), and (III) decomposition of carbonates, mainly calcite (600–850 °C) [15] if it is present in the source material. Changes also occur at higher temperatures. Upon firing above 1000°C, the ceramic paste turns into a glassy substance containing particles of added temper materials. This change is irreversible, and the resulting product has little in common with the initial ceramic paste. Here we should note that this confirmation is correct only if montmorillonite predominates as a clay material since it contains both bound water and hydroxyl groups in its composition.

In contrast, kaolinite begins to lose mass at temperatures of about 400°C and higher due to dehydroxylation [15]. Mass loss after dehydroxylation also occurs at higher temperatures, but this depends on the type of clay mineral. During the firing of montmorillonite and kaolinite, clay minerals are transformed into an anhydrous amorphous phase. Even this brief review indicates that the processes under consideration are incredibly complex, and the specified temperature intervals are conditional. Nevertheless, the authors of [14] managed to establish some patterns in the ratios of mass loss in these temperature ranges, which will also be used in the context of this work. Fig. 1 shows the results of TG/DTG experiments carried out on a raw ceramic paste sample with two mass loss steps indicating that the investigated material does not contain calcite. This is a representative TG/DTG analysis figured out of two ceramic paste samples. The DTG curve is a derivative of mass loss (dm/dt) and indicates the steps of mass loss more clearly. Mass loss values for both samples are summarized in Table 1. Samples lose 8.06 % (CP1) and 15.0 % (CP2) upon dehydration and 3.16 % (CP1) and 5.45 % (CP2) upon decomposition of hydroxyls.

The results of thermogravimetric TG/DTG measurements of ancient ceramic sample N1 are shown in Fig. 2. Compared to the ceramic paste in Fig. 1, the dehydration and dehydroxylation peak for ceramics is clearly less than for ceramic paste. However, dehydroxylation occurs when ancient pottery is heated.

Table 1 Mass-loss of ancient ceramic samples and raw ceramic pastes in the temperature intervals

Sample	Mass loss, %			m1	m2	m3	m2/m1
Sample	≤350◦C	≤600◦C	≤850◦C	m1	m2	m3	m2/m1
N1	98.7	97.9	96.32	1.3	0.8	1.58	0.62
N2	96.47	94.4	91.72	3.53	2.07	2.68	0.59
N3	97.33	95.59	94.64	2.67	1.74	0.95	0.65
N4	97.98	96.56	94.75	2.02	1.42	1.81	0.70
CP1	91.94	88.78	87.68	8.06	3.16	1.1	0.39
CP2	85	79.55	78.61	15	5.45	0.94	0.36

Table 1 also summarizes the results of TG measurements of four ceramic samples found during archaeological excavations in Lalatepe. The results of TG measurements of ceramic samples represented in terms of the mass loss in the temperature ranges of ≤350°С (m₁-dehydration), 350÷600°С (m₂-dehydroxylation), and 600÷850°С (m₃-decomposition of carbonates, micas, etc.). These samples' TGs differ from each other and from the ceramic paste samples.

There are different approaches in the literature for determining the firing temperature of ancient pottery [2]. The basic idea of the thermogravimetric method is that only reversible thermal transformations will be detected if the sample is heated a second time. Upon reheating, transformations not observed in the previous heating will be detected only at temperatures above the upper-temperature limit of the first heating. The irreversibility of thermal transformations in clay occurs due to chemical transformations with the release of gaseous products, the formation of new minerals, or irreversible phase transformations.

The endothermic peaks in differential thermal analysis enable the identification of the upper limit of temperature intervals of loss of chemically combined hydroxyl groups by clay minerals such as smectites and kaolin. It has been assumed that [1] “when this peak is present, it shows that the pottery has not been heated above this temperature previously; when it is absent, the pottery has been fired above this temperature ."However, this statement is confirmed only for freshly prepared ceramic products. In the DTA of these products, no endothermic peaks are detected at temperatures from 400°С to 600°С [16]. Numerous experimental data indicate that most samples of ancient ceramics lose water, i.e., the hydroxyl group when heated to 550°C [15]. In this case, one has to assume that either the firing temperature of ancient ceramics was less than 550°C or the decomposition of hydroxyls is reversible. However, many works indicate that the process of dehydration and dehydroxylation of clay fired up to 700°С (so-called mild firing) is a partially reversible process (for instance,[10] [17][18]). The authors of [19] have shown that the dehydroxylation of clay heated up to 700-800°С is not only a reversible process but can even be used for dating ancient ceramics after re-hydroxylation at ambient conditions. Therefore, dehydroxylation upon reheating ancient pottery indicates a low or short-term firing temperature.

XRD analysis of ceramic sherds reveals that all investigated samples contain similar minerals: quartz, feldspar, and clay (Fig.3 and Table 2). Raw ceramic pastes (CP1 and CP2) are the modern examples presented by a pottery master

working near the archaeological site using local raw materials. Table 2 shows that the samples of raw ceramic paste do not contain calcite in their composition. XRD patterns of raw ceramic paste indicate the presence of clay minerals and quartz, and felspars as a tempering material (Fig. 4). Calcite is also not contained in sample N1. According to XRD analysis, samples N1 and N4 contain diopside, and samples N2 and N3 maghemite. Analysis of the raw ceramic mass did not reveal the presence of these minerals, which may indicate a discrepancy between the origin of ancient ceramic shards and modern ones.

Feldspars (in our case, albite, microline, anorthite) can be introduced into the ceramic mass as a hardening or be present in the composition of the original clay as a natural admixture since the clays themselves are considered to be the weathering product of feldspar. Feldspars are stable when heated to 950°C [20]. Alkaline feldspars remain in a glassy state for a long time when melted, while anorthite crystallizes relatively quickly. So, the presence of feldspars says little about the firing temperature of ceramics.

Quartz is a significant component of tempering materials and also exists in raw clay as a natural mixture. Quartz undergoes a phase transition around 573°C when heated, but this process is reversible, and no signs of previous heating could be detected after cooling.

Calcite is the most common “fingerprint” for determining the provenance of ceramics and, to some extent, for determining the firing temperature since it can be added to ceramic paste or found in the original clays as a natural impurity. The presence of calcite in ancient pottery is considered today the sign of low-temperature firing at about 700°C [7]. But here, too, not everything is so clear. Silicates that were produced from the clay minerals possess the ability to absorb carbon dioxide selectively from the atmosphere [21], [22]; in other words, it is quite a reversible process in a long-term period and, thus, cannot be used for the estimation of firing temperature [15].

Table 2. Mineral composition of ancient ceramic samples and raw ceramic pastes

Sample	Quartz,mass %	Feldspar, mass %	Calcite, mass %	Clay minerals, mass %	Other minerals, mass%
N1	26.7	Albite -31.3 Microline-22.6	0	Muscovite -8.9	Diopside -10.4
N2	54.8	Anorthite -29.3	3.7	Muscovite -10.3	Maghemite -1.9
N3	40.0	Anorthite-29.8	1.9	Muscovite -27.1	Maghemite -1.2
N4	29.1	Albite-39.0	8.6	Muscovite -7.3	Diopside -16.0
CP1	42.6	Albite -9.5	0	Muscovite -40.5 Montmorillonite -7.3
CP2	42.0	Albite -14.7	0	Muscovite -36.2 Montmorillonite -7.3

The concentration of calcite in samples N2, N3, and N4 is 3.7, 1.9, and 8.6%, respectively (Table 2); therefore, according to the traditional interpretation, the firing temperature of these samples was in the range of 700°C [23]. X-ray phase analysis did not reveal the calcite in sample N1, and it can erroneously be assumed that this sample was fired at a temperature above 800°C. But the analysis of the raw ceramic paste CP1 and CP2 showed that calcite may not be in the original material; therefore, in this case, its absence in ceramics cannot be an indicator of the firing temperature. At the same time, the presence of calcite cannot be used to determine the origin of ceramics, suggesting that it can accumulate over time under natural storage conditions.

X-ray phase analysis did not detect mullite either in the raw ceramic paste or ceramic shreds. The absence of this mineral is also a criterion for assessing the firing temperature of ancient ceramics since mullite is not introduced into the composition of the clay fraction of ancient ceramics as a tempering mineral. Since it occurs during the high-temperature firing of kaolinite, it can be a signature for the upper limit to the firing temperature of ancient ceramics.

Montmorillonite can also serve as a guideline for determining the lower limit of the firing temperature of ceramics since, at 600°C, this mineral turns into an amorphous phase[16].

The degree of vitrification was studied using environmental scanning electron microscopy (ESEM) on the original shreds at x800 magnification, following the procedure described in [24]. At higher firing temperatures, vitrification of the ceramic mass can be observed, but no traces of melting were found in these samples under a microscope. Based on the changes found in the original and re-fired samples, it was shown (Fig. 5) that the firing technology of the studied samples did not provide the level of vitrification that occurs at temperatures of about 900°C.

The “average” composition of the ceramic sample of N2 was studied by energy-dispersive X-ray diffraction analysis at several points, and the results are shown in Fig. 6.

Table 3 illustrates the chemical composition of sample N2, presented in oxides.

Table 3. The chemical characterization of the sample N2

Composition	Na₂O	MgO	Al₂O₃	SiO₂	SO₃	K₂O	CaO	TiO₂	FeO
Mass, %	2.85	3.83	14.80	56.59	2.36	4.62	7.73	0.79	5.90

The presence of the iron-containing mineral is also confirmed by XRD and TGA analysis. Differential thermal curves of maghemite show a peak at 815°С, and XRD analysis attributed this peak to the recrystallization of maghemite or hematite [25].

To reveal the differences between the thermogravimetric data for various samples of ancient ceramics, the authors of [14] developed a mass-loss diagram. The mass-loss diagram is an alternative way that visualizes the variations in the tempering and degree of thermal transformations of ancient ceramics [14]. This is not a description of experiments on the reconstruction of the firing of ancient pottery but rather a way to identify principal components for group differences among the studied samples.

Figure 7 shows the mass-loss diagram for the investigated samples based on the data given in Table 1. It can be seen from the figure that the mass loss ratios of samples of ancient ceramics lie almost in the same line as for raw ceramic pastes. Although the authors of [14] warned about the consequences of misuse of the diagram, we still allow the possibility of asserting that, according to the diagram, the samples under study restored the original hydroxyl cover over time while in the ground. And according to stratigraphic data, this time was long since the archaeological site belonged to the Neolithic period. Another conclusion is that the initial firing conditions were relatively mild; otherwise, rehydroxylation would not have been possible.

In the present work, four ancient ceramic shreds from the archaeological site Lalatapa, Fizuli region of the Republic of Azerbaijan, and two local raw ceramic pastes were characterized by powder X-ray diffraction (PXRD), scanning electron microscopy and thermal analysis (TG-DTG) techniques. XRD analysis of ceramic paste reveals that all investigated samples contain similar minerals: quartz, feldspar, and clay. Three samples out of four contain calcite. Based on the traditional approach, it has been assumed that the firing process in these samples stopped before 700°C. The mass loss ratios of samples of ancient ceramics also indicate that irreversible dehydroxylation took place in all four samples, thus indicating the initial mild firing conditions. SEM analysis also didn't reveal the signs of deep vitrification. The summary of all the applied methods indicates that the ceramic samples were made using a similar manufacturing technology. XRD analyses reveal the presence of accompanying minerals in the composition of ceramics. The difference in the concentration of calcite and accompanying minerals (diopside and maghemite) indicates that the sources of investigated ceramic samples were somewhat different.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this Manuscript.

S. Meyvel, P. Sathya, and G. Velraj, “Thermal characterization of archaeological pot sherds recently excavated in Nedunkur, Tamilnadu, India (Caracterização térmica de fragmento cerâmico arqueológico,” Ceramica, vol. 58, no. 347, pp. 338–341, 2012.
V. Molodin, L. Mylinkova, N. Shtertser, I. Durakov, and V. A. Drebushchak, “Thermogravimetry in the Studies of Ancient Technical Ceramics,” Chem. Sustain. Dev., vol. 27, no. 1, pp. 101–108, 2019, doi: 10.15372/csd20190116.
C. Yuan, C. Zhao, F. Wang, and S. Yuan, “Characterization of ceramic from the Early Bronze Age Xinzhai site, Henan Province, China, by using a multi-analytical approach,” J. Archaeol. Sci. Reports, vol. 44, p. 103, Aug. 2022, doi: 10.1016/j.jasrep.2022.103551.
A. Iordanidis, J. Garcia-Guinea, A. Strati, A. Gkimourtzina, and A. Papoulidou, “Thermal, mineralogical and spectroscopic study of plasters from three post-Byzantine churches from Kastoria (northern Greece),” J. Therm. Anal. Calorim., vol. 103, no. 2, pp. 577–586, 2011, doi: 10.1007/s10973-010-1055-7.
A. Kloužková, M. Kohoutková, P. Zemenová, and Z. Mazač, “Characterisation of a prehistorical ceramic object,” J. Therm. Anal. Calorim., vol. 116, no. 2, pp. 641–645, May 2014, doi: 10.1007/s10973-014-3712-8.
L. A. Ortega, M. C. Zuluaga, A. Alonso-Olazabal, X. Murelaga, and A. Alday, “Petrographic and geochemical evidence for long-standing supply of raw materials in neolithic pottery (mendandia site, spain),” Archaeometry, vol. 52, no. 6, pp. 987–1001, 2010, doi: 10.1111/j.1475-4754.2010.00523.x.
D. N. Papadopoulou, M. Lalia-Kantouri, N. Kantiranis, and J. A. Stratis, “Thermal and mineralogical contribution to the ancient ceramics and natural clays characterization,” J. Therm. Anal. Calorim., vol. 84, no. 1, pp. 39–45, 2006, doi: 10.1007/s10973-005-7173-y.
D. E. Arnold, “Does the Standardization of Ceramic Pastes Really Mean Specialization?,” J. ofArchaeological Method Theory, vol. 7, no. 4, pp. 333–375, 2000, [Online]. Available: http://www.jstor.org/stable/20177426
D. Vlase, O. Rogozea, C. Moşoiu, G. Vlase, R. Lazău, and T. Vlase, “Thermoanalytical investigations of some ceramics dated from the Neolithic period, discovered at Oxenbrickel, Sânandrei, Romania,” J. Therm. Anal. Calorim., vol. 138, no. 3, pp. 2145–2157, 2019, doi: 10.1007/s10973-019-08767-8.
S. Shoval, M. Boudeulle, and G. Panczer, “Identification of the thermal phases in firing of kaolinite to mullite by using micro-Raman spectroscopy and curve-fitting,” Opt. Mater. (Amst)., vol. 34, no. 2, pp. 404–409, 2011, doi: 10.1016/j.optmat.2011.08.031.
M. A. Legodi and D. de Waal, “Raman spectroscopic study of ancient South African domestic clay pottery,” Spectrochimica Acta - Part A: Molecular and Biomolecular Spectroscopy, vol. 66, no. 1. pp. 135–142, 2007. doi: 10.1016/j.saa.2006.02.059.
V. A. Drebushchak, L. N. Mylnikova, and V. I. Molodin, “Thermogravimetric investigation of ancient ceramics: Metrological analysis of sampling,” Journal of Thermal Analysis and Calorimetry, vol. 90, no. 1. pp. 73–79, 2007. doi: 10.1007/s10973-007-8478-9.
I. Weaver, G. E. Meyers, S. A. Mertzman, R. Sternberg, and J. Didaleusky, “Geochemical evidence for integrated ceramic and roof tile industries at the Etruscan site of Poggio Colla, Italy,” Mediterr. Archaeol. Archaeom., vol. 13, no. 1, pp. 31–43, 2013.
V. A. Drebushchak, L. N. Mylnikova, and T. N. Drebushchak, “The mass-loss diagram for the ancient ceramics,” J. Therm. Anal. Calorim., vol. 104, no. 2, pp. 459–466, 2011, doi: 10.1007/s10973-010-1230-x.
V. A. Drebushchak, L. N. Mylnikova, and T. N. Drebushchak, “Thermoanalytical investigations of ancient ceramics,” J. Therm. Anal. Calorim., vol. 133, no. 1, pp. 135–176, Jul. 2018, doi: 10.1007/s10973-018-7244-5.
V. A. Drebushchak, L. N. Mylnikova, T. N. Drebushchak, and V. V. Boldyrev, “The investigation of ancient pottery,” J. Therm. Anal. Calorim., vol. 82, pp. 617–626, Sep. 2005, doi: 10.1007/s10973-005-6913-3.
Y. Gallet and M. Le Goff, “Rehydration and Rehydroxylation in Ancient Ceramics: New Constraints from Mass Gain Analyses Versus Annealing Temperatures,” J. Am. Ceram. Soc., vol. 98, no. 9, pp. 2738–2744, 2015, doi: 10.1111/jace.13674.
V. Š. Fajnor and K. Jesenák, “Differential thermal analysis of montmorillonite,” J. Therm. Anal., vol. 46, no. 2, pp. 489–493, Feb. 1996, doi: 10.1007/BF02135026.
G. T. Barrett, “Rehydroxylation Dating: Assessment for Archaeological Application Ph.D. thesis, Queen’s University Belfast.,” 2015. [Online]. Available: http://pure.qub.ac.uk/portal/files/61474855/G_Barrett_PhD_RHX_Full.pdf
I. Stubna, R. Podoba, P. Bacík, and Ľ. Podobník, “Romanesque and Gothic bricks from church in Pác – estimation of the firing temperature,” Epa. - J. Silic. Based Compos. Mater., vol. 65, no. 2, pp. 48–51, 2013, doi: 10.14382/epitoanyag-jsbcm.2013.11.
A. M. Kalinkin, E. V. Kalinkina, and O. A. Zalkind, “Mechanosorption of carbon dioxide by Ca- and Mg-containing silicates and alumosilicates. Sorption of CO2 and structure-related chemical changes,” Colloid J., vol. 71, no. 2, pp. 185–192, Apr. 2009, doi: 10.1134/S1061933X09020069.
A. M. Kalinkin, E. V. Kalinkina, O. A. Zalkind, and T. I. Makarova, “Mechanochemical interaction of alkali metal metasilicates with carbon dioxide: 2. The influence of thermal treatment on the properties of activated samples,” Colloid J., vol. 70, no. 1, pp. 42–47, 2008, doi: 10.1134/s1061933x08010079.
G. Cultrone, C. Rodriguez-Navarro, E. Sebastian, O. Cazalla, and M. J. De La Torre, “Carbonate and silicate phase reactions during ceramic firing,” Eur. J. Mineral., vol. 13, no. 3, pp. 621–634, 2001, doi: 10.1127/0935-1221/2001/0013-0621.
M. Chatfield, “Tracing firing technology through clay properties in Cuzco, Peru,” J. Archaeol. Sci., vol. 37, no. 4, pp. 727–736, Apr. 2010, doi: 10.1016/j.jas.2009.11.003.
R. M. Taylor and U. Schwertmann, “Maghemite in Soils and Its Origin II. Maghemite Syntheses at Ambient Temperature and pH 7,” Clay Mater., vol. 10, no. 4, pp. 299–310, 1974.

Download PDF

Version 1

posted

You are reading this latest preprint version

The Complex Characterization of Ceramics from Lalatepe Archeological Site

Status:

Version 1

Abstract

Figures

Introduction

Materials And Methods

Result And Discussions

Conclusions

Declarations

References

Status:

Version 1