Next Article in Journal
An Assessment of Landscape Evolution Through Pedo-Geomorphological Mapping and Predictive Classification Using Random Forest: A Case Study of the Statherian Natividade Basin, Central Brazil
Previous Article in Journal
Two Opposite Change Patterns Before Small Earthquakes Based on Consecutive Measurements of Hydrogen and Oxygen Isotopes at Two Seismic Monitoring Sites in Northern Beijing, China
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Geosciences
Volume 15
Issue 6
10.3390/geosciences15060193
geosciences-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

A Provenance Study of Ceramic Artifacts from the Area of Makariopolsko Village, NE Bulgaria

by
Bilyana Kostova
1,*,
Svetlana Todorova
2,
Katerina Mihaylova
3,
Chavdar Lalov
2 and
Ralitza Berberova
1
1
Department of Natural Sciences, New Bulgarian University, 21 Montevideo Blvd., 1618 Sofia, Bulgaria
2
National Institute of Archaeology with Museum, Bulgarian Academy of Sciences (NAIM-BAS), 2 Saborna Str., 1000 Sofia, Bulgaria
3
Institute of Mineralogy and Crystallography “Acad. Ivan Kostov”, Bulgarian Academy of Sciences, Acad. G. Bonchev Str., Bldg.107, 1113 Sofia, Bulgaria
*
Author to whom correspondence should be addressed.
Geosciences 2025, 15(6), 193; https://doi.org/10.3390/geosciences15060193
Submission received: 17 April 2025 / Revised: 19 May 2025 / Accepted: 21 May 2025 / Published: 22 May 2025
(This article belongs to the Section Geoheritage, Geoparks and Geotourism)
Browse Figures
Versions Notes

Abstract

:
The Roman site at Makariopolsko village in Northeastern Bulgaria has been identified as a ceramic production center, featuring single- and double-chamber kilns, abundant ceramic material, and a nearby water source. Geological assessments also reveal local clay deposits. Previous archaeological studies have noted similar Roman production sites in the region, primarily focusing on the study of the kilns and the macroscopic description and classification of the ceramics. However, there has been a lack of research into the pottery’s composition and the sourcing of raw materials, which is essential for understanding the area’s cultural and economic context. This study aims to determine the raw material and firing temperature of the ceramic from the site at Makariopolsko village. Clay samples (both raw and fired at 1100 °C) and ceramic were subjected to chemical, statistical, phase X-ray structural, and thermal analyses. The findings indicate the use of calcareous illite–kaolinitic clay, sourced locally, with an added sandy component. The ceramics were fired at temperatures of 570–760 °C and 920–945 °C. These results, which support the site’s identification as a pottery production center, highlight advanced pottery skills and the dual functional capabilities of the kilns. Additionally, they pave the way for further research into regional production center relationships.

1. Introduction

Ceramic is the most ancient material created by man. It is easy to produce, inexpensive, and durable over time. It makes up the largest and most commonly found group of artifacts in archaeological excavations [1,2]. It is of great interest even when found in a fragmented state [2], as it is used to date archaeological sites [3].
The chemical and mineralogical characteristics of ceramics are also of particular interest to archaeology, as they can be used to determine the origin of the raw material and the technology used to produce it [4]. Determining the deposit from which clay was extracted for pottery production in the past provides important archaeological information, as its selection may be influenced by a variety of factors—land ownership and control, seasonal access, ground surface exposure, transportability, etc. [5]. The firing temperature of the ceramic is directly related to the kiln design and the fuel used. The experimental study of ceramic artifacts complements empirical archaeological research and allows for archaeological interpretation of both the cultural and economic nature of societies in the past [6]. The obtained data can be used by researchers worldwide, leading to the dissemination of information and the globalization of science [7], allowing for the comparison of ceramic artifacts from sites in different geographic locations. It also provides the opportunity to create contemporary ceramics that are compatible with ancient ceramics and suitable for the conservation and restoration of cultural heritage [6].
This work presents a provenance study of ceramic artifacts from the archaeological site of Makariopolsko village, as well as a study of the equivalent firing temperature of the ceramic using the capabilities of geosciences. The chemical, mineralogical, and phase compositions of clay and ceramics were compared by X-ray fluorescence (XRF) analysis, statistical processing of the XRF data by cluster analysis, powder X-ray diffraction (PXRD) measurements, and thermal analysis (TG/DTG-DSC). In addition, the changes in the clay samples upon heating were also investigated for greater accuracy in determining the equivalent firing temperature of the ceramic.
In the area around the site at Makariopolsko village, there are other described sites that have been identified as production centers and ceramic workshops. In relation to these, the site at Makariopolsko would not have been an isolated case for the area in the Roman period, in terms of possible production functions. A similar case is evidenced at the nearby villa at Beli Lom village, Razgrad region, for which there is very little information. The same goes for the large villa complex at Madara village, Shumen region. Several settlements in the Targovishte region were also involved in production activities (the sites at Kardam and Gorsko Ablanovo), Abritus at Razgrad city, and the Shumen region (within the Outer Town of Pliska at the village of Varbyane and the village of Pet mogili). These sites have already been studied in the last century, and to date, all published results have been summarized [8]. Research in this regard continues to this day, but it is mainly focused on the examination of the discovered kilns and archaeological macroscopic description and classification of the pottery. The provenance study of ceramic artifacts and data on their equivalent firing temperature are virtually absent in the literature. The results obtained are new for the studied site, as well as for the other Roman age sites in the area.

2. Study Area

2.1. Archaeological Background

The archaeological site at Makariopolsko is located 1.9 km south of the village center, on the right bank of a small river, a left tributary of the Saedinenska River. It is situated on a gentle slope facing west and has an area of 87 acres. The site was found in 2016 [9], while the first surveys were made in 2024 through partial core drilling to clarify the chronology, stratigraphy, and nature of the site’s occupation [10]. The excavations have revealed materials dating to the Bronze Age, Late Iron Age, and Roman periods. Of the structures identified at the site, the most significant are two kilns and two buried structures. The buried structures are distinguished by their larger size and the high concentration of fragmented pottery from the Roman period (2nd–mid-3rd century) (Figure 1). At this stage of the excavations, no traces of walls and roof structures of residential and/or farm buildings have been found.
Kiln 1 (Figure 1 and Figure 2) is located on relatively flat ground and has a northwest–southeast orientation (with an opening on the southeast). The kiln is two-chambered, designed for firing household pottery. Its shallow location (0.37 m from the surface) led to its partial destruction during agricultural cultivation. There is no perforated floor (single fragments were found), and the walls are preserved up to 0.46 m high (5–8 cm thick). The lower chamber is 1.3 m in diameter. The support for the perforated floor is preserved, which was built of crushed stones heavily plastered with clay (height 0.42 m, dimensions at the base 0.31 m, and thickness of the plaster 6–8 cm). The stoking channel is arch-shaped, with a width of 0.57 m and a height of 0.27 m. It extends in its easternmost part to the stoke pit. A large quantity of fragmented pottery from the Roman period was found at the bottom of the chamber, which was probably used to fill the kiln after it ceased to function. The facility belongs to the widespread-in-Antiquity furnace type Ia, according to the classification of A. Harizanov [8]. They are distinguished by a round or oval lower chamber, a single channel, and a central pillar with a circular, quadrangular, or elliptical layout [8].
Kiln 2 (Figure 3) is single-chambered and built on the slope towards the ravine. It has been heavily eroded by the working of the land, with the walls preserved to a height of 0.27 m (7/9 cm to 0.14 m thick). It has an oval shape (dimensions 1.26 × 0.91 m) and an opening from the northwest. There is no impression or trace of a central pillar or perforated floor. The bottom slopes towards the opening, with a difference of 0.18 m between the two ends. The channel is 0.35 m wide and 0.40 m long, with the floor overlapping a medium-sized flat stone at its westernmost part. In front of the furnace opening and to the northwest, an oval-shaped stoke pit, measuring 1.66 × 2.0 m and 0.22 m deep, develops. The difference between the bottom of the stoke pit and the furnace channel is 0.37 m. No archaeological materials were found in the fill. The established kiln 2 is defined as type If, according to the classification of Harizanov [8]. Due to its small size and the lack of a perforated floor and a supporting pillar, the hypothesis of its use for domestic purposes is more plausible. However, it should not be overlooked that furnaces with similar characteristics and production functions have been registered. Single-chamber kilns are usually attributed to domestic functions, since they reach a maximum temperature of 450–600 °C, but they may also be used for the production of low-quality ceramic products [11].
The recorded kilns, the two structures with large concentrations of ceramic material, and the presence of the river adjacent to the site provide archaeological grounds to suggest ceramic production within the site, although no scrap production or other structures associated with the manufacturing process have been identified at this stage. The absence of such artifacts can be explained by the drilling nature of the excavations conducted to date.

2.2. Geological Background

The archaeological site at Makariopolsko is located in an area with a ground surface exposure of Upper Cretaceous rocks of the Razgrad Formation (rK1h-b) and Pleistocene formations: eolian (eQp2–3) and eolian–alluvial––delluvial (e-a-dQp) (Figure 4).
The Razgrad Formation is composed of alternating calcareous marls with clayey limestones, among which small glauconite grains and concretions of iron hydroxides are observed. The eolian formations (eQp2–3) include a loess complex that is widespread in the studied area. The loess is a whitish–yellowish poorly cemented clayey siltstone containing single grains, crusts, or concretions of CaCO3. From north to south, the thickness of the loess cover increases along with the percentage of the clay component. Today, in the northern parts of the area, loess is used locally for brick production. The loess develops with a gradual transition over the Lower Pleistocene red sandy clays with manganese accumulations. The eolian–alluvial–delluvial formations (e-a-dQp) are revealed on significantly smaller areas in the region—only in the vicinity of the villages of Golyamo Sokolovo, Makariopolsko, and Panayot Volovo. They are represented by loess clays of small thickness (1 to 5 m), which are situated on the Razgrad Formation and are a natural continuation of the loess complex. Loess clays are yellowish to brownish–red in color, oily, dense, and structureless. They contain iron and manganese hydroxide accumulations [12,13].

3. Materials and Methods

3.1. Materials

3.1.1. Clay

Four clay samples were analyzed: R1, R2, R3, and R4. Three of the samples (R1, R2, and R3) were collected from the western part of the site near the kilns, and R4 was collected from the eastern part (Figure 1, Table 1). The clays are part of the Pleistocene eolian–alluvial loess clays (e-a-dQp) (Figure 4). All samples are slightly porous, greasy to the touch, and vary in color. R1 and R3 are light, beige-to-yellowish in color, and R2 and R4 are darker, brownish–reddish in color (Figure 5).

3.1.2. Ceramic Sherds

Six ceramic fragments were studied (Figure 1 and Figure 6, Table 1). Two of the fragments come from kiln 1 (C2 from the fill and C3 from the front of the channel), while fragments C1 and C6 are from structures where a large amount of pottery was found. C4 and C5 can be dated to the pre-Roman period and were chosen for analysis due to the following characteristics: C4 (greyware pottery) was made in a reduction setting, and handmade pottery (C5) is associated with production on site from local raw material—i.e., a product of domestic manufacture or industry [14,15].

3.2. Methods

3.2.1. Clay Sample Collection

Around 8–10 kg of clay samples were collected from each of the locations at the site (Figure 1, Table 1). The samples were quartered and homogenized to obtain the quantity required for analysis.

3.2.2. Sample Preparation

Clay—quartered samples were dried, ground in an agate mortar, and homogenized.
Ceramic sherds—the samples were mechanically cleaned by brushing and purified with distilled water. The surface part (the part that was in direct contact with the environment) of the sherds was removed in order to analyze the unaltered material [16]. The samples were then ground in an agate mortar and homogenized.

3.2.3. XRF (X-Ray Fluorescence) Analysis

X-ray fluorescence (XRF) analysis was used for the determination of the chemical composition of the samples—spectrometer WD-XRF Supermini 200—Rigaku, Japan (50 kV and 4 mA, 200 W X-ray tube with Pd-anode, 30 mm2) in a helium atmosphere. Three different X-ray detectors were used for light and heavy elements. Depending on the wavelength range, different analyzing crystals were used: LIF 200 (for Ti-U), PET (for Al-Ti), and RX25 (for F-Mg).
The XRF analysis results were normalized to 100% in oxides. To reduce the measurement error, the percentage of loss on ignition (LOI) was first determined [17]. A standard method was used to determine the LOI—heating the samples and measuring the difference in weight before and after heating due to the release of volatile components such as H2O, OH, CO2, etc. [18]. The heating temperature depends on the purpose and material studied [19]. For the purpose of the present work, all samples were heated by gradually increasing the temperature up to 900 °C continuously, until a constant weight was achieved. Rigaku’s built-in ZSX software package v. 7.67 was used for the data processing. The measured LOI values for each sample were set as the input parameters in ZSX software [20].

3.2.4. Statistical Processing

Statistical processing of the major and minor elements measured by XRF was performed using OriginPro 2018 software, and a hierarchical cluster analysis (CA) was applied. CA is the most commonly used multivariate statistical method in archaeology for the purpose of increasing research accuracy by determining similarities and differences, i.e., correlation in the chemical composition of ceramic artifacts and for provenance studies [5,21,22,23,24]. The CA results are represented graphically by a dendrogram using the average linkage of Euclidean distance. The dendrogram is a visualization of the complete sequence of steps of the hierarchical clustering algorithm, showing the order and level of the clusters and the distance between the individual samples [22,24]. The Elbow method was used to determine the optimal number of clusters.

3.2.5. PXRD (Powder X-Ray Diffraction) Measurements

Powder X-ray diffraction (PXRD) measurements were made by Empyrean, Panalytical, CuKα radiation (λ = 0.15418 nm) (operating at 40 kV, 30 mA) from 5 to 80° 2θ with a step of 0.013 2θ, 30 s/step. Phase identification was performed using the Powder Diffraction File database [25], and peak fitting was with QualX (v. 2.24).

3.2.6. Thermal Analysis (TG/DTG-DSC)

The total calcite content (%), the temperatures of volatile phase separation, and the structural decomposition of the organic and inorganic phases were determined by thermal analysis. A simultaneous thermogravimetric (TG) differential scanning calorimeter was used (DSC)—Setline STA 1100, SETARAM in the temperature range RT—1000 °C in an oxygen atmosphere, heating step—10 °C min−1, sample weight 25.0 ± 1.0 mg (resolution 0.05 µg), accuracy of temperature measurements +/−0.3 °C, and ceramic crucibles with a volume 100 µL. Calisto Processing’s built-in software v. 2.0994 was used to process the results.

4. Results

4.1. XRF

The results of the chemical analysis and the measured LOI are presented in Table 2. The major elements (with values mostly above 1.0%) are presented in oxide form (wt%). Their total amounts range from 99.62 to 99.86 wt%. The trace elements have low contents in oxide form (below 0.5 wt%) and are, therefore, presented as ppm [26]. The major elements for all samples tested were Si, Al, Ca, Mg, K, Fe, and Ti, and for sample R3, also Na. The SiO2 content was the highest and ranged from 24.02% to 44.30%. C3 and C4 are exceptions, with SiO2 contents of 59.95% and 57.05%, respectively. The values for Al2O3 range from 6.42% to 17.50%. The CaO content is high—13.83% to 34.37%. The only exceptions are C3 and C4, where CaO has significantly lower values—4.46% and 2.31%. The values of MgO vary from 1.23% to 2.48%, those of K2O from 1.37% to 4.07%, and those of Fe2O3 from 2.73% to 7.79%. The TiO2 values vary slightly between 0.33% and 0.93%. Of the rare elements, only Mn and Sr were detected in all of the samples examined and had high concentrations, with hundreds of ppm. Cr and Ni were detected in only three samples. Cu was registered in most samples, except R3 and C6. Zn was only absent in R1 and R4. Rb was only absent in R1, and Zr was only absent in R3. V was only detected in two samples, R2 and C4.

4.2. CA

Figure 7 presents the CA results. In the statistical processing of the data, the elements V and Cr were excluded due to their large variation in surface conditions [5]. The samples were distributed in three clusters. Cluster 1 includes only clays R1 and R3. Cluster 2 includes ceramic (C1, C21, C5, and C6) and clay (R2 and R4), and cluster 3 includes only ceramic (C3 and C4).

4.3. PXRD

Table 3 and Figure 8 and Figure 9 present the results of the mineral and phase composition analyses of the samples.
All of the clays (R1, R2, R3, and R4) have identical mineral composition—phyllosilicates (kaolinite, illite, muscovite, and chlorite), quartz, anorthite, and calcite. PXRD studies of all clays heated to 1100 °C show the presence of quartz, anorthite, gehlenite, and akermanite.
All ceramic samples contain quartz and anorthite. Illite and muscovite were recorded in C4, C5, and C6. Calcite was found in all samples except C4. Gehlenite was only proven in C1 and C2.

4.4. Thermal Analysis

The thermal analysis results are presented in Figure 10. The total mass losses (MLtot) in the clays range from 24.31 to 35.56% and in the ceramic from 4.64 to 23.22%. During the heating of the samples, five temperature stages with separation and/or absorption of volatile components are distinguished: 1—dehydration; 2—hydroxide dehydroxylation/magnetite oxidation and organic decomposition; 3—phyllosilicates dehydroxylation; 4—decarbonation; 5—phyllosilicates structure destruction and crystallization of high-temperature phases.
In stage 1, a dehydration process takes place at temperatures from RT to 200/220 °C. From RT to ~100 °C, absorbed water molecules are released [28], and between 100 and 200/220 °C, the dehydration of phyllosilicates, including illite, muscovite, and chlorite, takes place with the release of water from the interlayer spaces [29,30,31]. Kaolinite absorbs little water. Thus, no dehydroxylation is evident from it in this temperature range [29]. The dehydration of the samples studied proceeded with mass losses from 0.60 to 15.75%.
In stage 2 (220–450 °C), several processes take place. Two exothermic peaks on the DSC curve, recognized as the dehydroxylation of goethite (~288 °C) and manganite (~427 °C), are recorded in the clays [31]. The ceramic fragments show a peak in the DSC curve between 270 and 291 °C, which is associated with magnetite oxidation [31]. The decomposition of organic matter occurs with an exothermic peak at 320–360 °C [32,33]. Mass losses in this interval range from 0.78 to 3.67%.
In stage 3, phyllosilicate dehydroxylation processes take place. In the clay samples (R1–R4) (Figure 10), the dehydroxylation of kaolinite was recorded, with an endothermic peak on the DSC curve at 519 °C. The temperature range in which kaolinite dehydroxylates is 500–600 °C [29,34]. The reported dehydroxylation temperature at the lower limit of this temperature range is related to the influence of dry grinding during sample preparation [29]. With an endothermic DSC peak at ~570 °C, chlorite [28] and illite [31,35] dehydroxylate. Muscovite dehydroxylates with an endothermic peak at 650–690 °C [31,36,37]. The dehydroxylation of chlorite proceeds with structural breakdown [28], whereas that of illite and muscovite proceeds without structural breakdown [38,39,40]. The dehydroxylation of kaolinite and chlorite was not recorded in the ceramic samples. Only the dehydroxylation of illite and muscovite was detected, which is possible after clay firing, as these two minerals are dioctahedral 2:1 layer silicates and have the ability to rehydroxylate at low temperatures [38,39,40]. No phyllosilicate dehydroxylation was detected in samples C1 and C2. For C5, the dehydroxylation of both illite and muscovite was recorded, and for C6, only that of muscovite. For C3 and C4, only the dehydroxylation of muscovite was recorded. An entodermal peak on the DSC curve at ~573 °C is also observed in stage 3, which is related to the polymorphic α-β transition of quartz. The transition proceeds without mass loss [41]. The effect is not present in all samples, even though quartz is proven by PXRD everywhere.
In stage 4, decarbonation takes place—the release of CO2 from calcite with structural breakdown at temperatures of 700–820 °C [42]. Decarbonation was found in almost all samples. Sample C4 shows a peak of the DSC curve at 734.5 °C with minimal mass loss, which is ignored. This peak can be attributed to both the decomposition of minimal amounts of calcite and to the decomposition of manganese oxides (pyrolusite) [31], in relation to the high manganese values in this sample, −1086.81 ppm (Table 2). Due to the minimal mass loss, it is assumed that the percentage of calcite/pyrolusite in the sample is too low for either of these phases to be detected by PXRD.
In stage 5 (700/820–1000 °C), phyllosilicate structure destruction and the crystallization of high-temperature phases take place. The structure of dehydrated illite and muscovite is destroyed. The effects are recorded by the endothermic peaks of the DSC curve at ~850 °C for illite and at ~920–940 °C for muscovite [43,44]. Structural decomposition proceeds with virtually no mass loss or negligible loss due to the evaporation of alkali cations [45]. The effects were recorded for all samples, except sample C2. Two exothermic peaks at temperatures ~895–900 °C and at ~950 °C are also observed in the temperature interval and are associated with the crystallization of the new high-temperature phases [31,34,46]. These effects were recorded for all samples, except C2.
Table 4 presents the reported calcite decomposition temperatures of the samples determined at the calcite decomposition inflection point (Tinfl) and the mass loss from CO2 release during calcite decomposition (MLCO2). The amount of CaO incorporated into calcite (CaOcalcite%) was calculated from MLCO2. From the total amount of CaO (CaOtotal%) measured by XRF and the calculated CaOcalcite%, the amount of CaO that is included in the silicate and/or amorphous phase (CaOsilicates/amorphous%) is also calculated. The amount of calcite in the samples (calcite/%) was also calculated from the measured MLCO2. Calcite was not detected only in C4.

5. Discussion

5.1. Clay Composition

The studied clays have high contents of Si, Ca, and Al (Table 2), indicating a composition mainly of silicates, aluminosilicates, and/or carbonates, which is consistent with the PXRD analysis results. Their compositions are identical (Figure 8a, Table 3), containing all three mineral groups typical of clays: clay, accessory, and inclusion minerals [30]. Clay minerals are represented by kaolinite and illite, accessory minerals by muscovite and chlorite, and inclusions by quartz, anorthite, and calcite. Fe2O3 also has high contents, but iron oxide and/or hydroxide minerals were not registered due to the experimental conditions. PXRD measurements were performed with CuKα radiation. Thermal analysis recorded iron phases, goethite (α-FeOOH), and manganese hydroxide, manganite (γ-MnOOH). Manganite was not detected by PXRD, probably due to the low sample content, which is below the detection limits of the method. The thermal analysis also detected organic phases in the clays (Figure 10).
A difference was found between the major element compositions of samples R1 and R3 relative to those of R2 and R4. In the first two samples, there is an increased amount of CaO and a correspondingly decreased content of the other major elements (Table 2). These data suggest elevated carbonate content in samples R1 and R3. The percentage of calcite in the samples was determined by thermal analysis (Table 4). For R1 and R3, the values are 48.63% and 49.47%, and for R2 and R4, 23.60% and 32.91%, respectively, indicating also a different ratio of clay and accessory minerals to the inclusion minerals and/or a different ratio of the inclusion minerals quartz and anorthite to calcite. There was also a difference in the amount of Fe2O3, with R1 and R3 having a lower Fe2O3 content (2.91% and 2.73%, respectively) and R2 and R4 having a higher one (4.83% and 3.96%, respectively) (Table 2). Such variations are typical for clays of young geological age [47], as are the samples studied (Pleistocene) [12,13]. In characterizing clays, it is essential to determine their type, either non-calcareous (poor in carbonates) or calcareous (high in carbonate content). The carbonate content of the initial clay controls the vitrification processes and the formation of new high-temperature phases [48,49,50], and these, in turn, control the quality of the produced ceramic [51]. Carbonate-poor clays are defined as clays that contain CaO at less than 5% [50,52] or less than 6% [53]. Depending on the mineral composition of the initial clay, there may be other minerals and/or amorphous phases that contain Ca, in relation to which the determination of the clay type cannot be made solely on the basis of CaOtotal% [54,55]. CaOcalcite% and CaOsilicates/amorphous% contents were calculated from MLCO2 (Table 4). In all samples, the amount of CaOcalcite% was above 6%, which defines the clays as calcareous. The rare-element contents of the clays are essential for determining ceramic provenance. Rare and major elements correlate differently. For example, changes in Sr percentages are associated with differences in percentages of Ca in samples [56]; Fe contents correlate with those of Ti [56], but Fe and Mg also substitute for Al in clay minerals; K in clay minerals is substituted by Rb; and Cu associates with Fe [57]. This complex correlation makes it very difficult to use bi-plots to search for a relationship between the rare-element contents of the source clay and the ceramic and requires the use of statistical methods to determine the relationships between individual samples. CA divides the clays into cluster 1, comprising R1 and R3, and cluster 2, comprising R2 and R4, with an average Euclidean distance between the two clusters of AD = 19.57. The differentiation of the samples into two distinct clusters is mainly related to differences in the major-element ratio, which is controlled by the different proportions of clay and non-clay minerals. This difference is also revealed by visual inspection of the clays (Figure 5). Samples R1 and R3 are lighter in color due to the higher calcite content and lower iron content, and vice versa, and samples R2 and R4 are darker in color due to the higher iron content and lower calcite content. As the samples were taken from a single area, with the sampling sites located in close proximity (Figure 1), the difference in clay/non-clay mineral ratios is related to the natural alteration of clays in the same area [47]. XRF, CA, and PXRD studies define the clays as carbonate Illite–kaolinite, with variable ratios between the individual minerals.

5.2. Ceramic Provenance

The chemical composition of fragments C3 and C4 differs significantly from C1, C2, C5, and C6 (Table 2). C3 and C4 have very low CaO content (4.46% and 2.31%) and higher SiO2 content, respectively, with Al2O3 not changing significantly. These results suggest higher quartz concentrations in C3 and C4, with approximately equal aluminosilicate contents in all ceramic fragments studied. In addition, the ceramic fragments (with the exception of C6) showed higher SiO2 and Al2O3 contents and lower CaO values compared to the contents of these elements in the clay samples. In C6, the increase is only for Al2O3. Increased Fe2O3 and TiO2 levels were also observed in all ceramic fragments. In relation to these differences, the CA shows a distribution of the ceramics in two clusters: cluster 2 (C1, C2, C5, and C6) and cluster 3 (C3 and C4) (Figure 4). The separation of cluster 3 from cluster 2 is mainly related to the difference in the amount of CaO. Cluster 2 has two sub-clusters: (i) 2.1, which includes R2, R4, and C6, and (ii) 2.2, which includes only ceramic sherds (C1, C2, and C5). The average Euclidean distance between the members of the two sub-clusters is AD = 12.21. AD indicates the differences in major element contents between the two sub-clusters, and the affiliation of C6 to sub-cluster 2.1 indicates greater similarity in chemical composition with the source clay. The separation of the ceramic fragments into two sub-clusters also indicates the possibility of clay mixing from different parts of the deposit. The belonging of C1, C2, C5, and C6 to one cluster, together with R2 and R4, is determined by the rare element contents, indicating the fabrication of these fragments from carbonate clay with the increased SiO2 content compared to the source clay. The increased SiO2 content can be explained by the addition of a sandy fraction (usually composed mainly of quartz, some feldspars, and micas) to the raw clay. The addition of a sand fraction to the clay is usually associated with the use of carbonate clay. When ceramics are fired at temperatures above that of calcite’s structure breakdown, empty pore spaces are left in the ceramic, leading to its shrinkage and significantly reducing the quality of the finished product [6]. The addition of quartz and plagioclase to the starting clay preserves the shape of the fired ceramic and improves its mechanical properties [51].
Samples C3 and C4 were included in cluster 3, which was located at AD = 31.53 from cluster 1 and cluster 2 members. This distance indicates that the samples are made of non-carbonate clay from another deposit and have an elevated amount of sand component. The different source clay used for the production of the ceramic fragments from cluster 2 and cluster 3 is also revealed by visual inspection of the ceramic sherds (Table 1, Figure 6 Fragments C1, C2, C5, and C6 have a beige–red color, and C3 has a brown-red color. The beige coloration is associated with ceramics produced from carbonate clay, in which iron is more involved in newly formed silicate phases, such as those of the melilitic group, rather than in oxide compounds [58]. In accordance with the geological setting of the area and previous archaeological studies, different possibilities for the origin of C3 and C4 can be outlined: (i) the fragments were made in situ, but from a different raw material source—eolian clays (eQp2–3) are exposed on the ground surface in large areas within a radius of several tens of kilometers around the site (Figure 1) [12,13]—or (ii) the pottery was imported. As other sites with production functions are known in the area of the site at Makariopolsko village, the production of C3 and C4 fragments at some of these sites can be assumed. The production centers listed above are located in areas where Pleistocene eolian formations (eQp2–3), represented by clay loess, are exposed on the land surface. Clay loess is also exposed in the immediate vicinity of the site at Makariopolsko, suggesting the knowledge of this type of raw material in the Roman age and the possibility of its extraction and use for on-site ceramic manufacture (Figure 4).

5.3. Equivalent Firing Temperature of Ceramic

The equivalent firing temperature of the ceramic can be determined by the temperature stability of its relict minerals and the formation temperature of new phases that are recorded in its composition [52,59]. Better results can be achieved by studying the changes in mineral and phase composition during the firing of the raw clay.
To determine the equivalent firing temperature of the ceramic with an established raw material source (fragments C1, C2, C5, and C6), the phase changes of the clay at 1100 °C were analyzed (Figure 8b).
PXRD analysis registered quartz, anorthite, gehlenite, and akermanite, while thermal analysis indicated the additional presence of hydroxide minerals (goethite and manganite) (Figure 10) in all samples. Quartz and anorthite are also found in the unheated clay (Figure 8a). Quartz is a relict mineral from the source clay, as depending on the elemental impurities it contains, it is stable in the temperature range 1100–1600 °C [60]. Above this temperature interval, quartz transforms into its polymorphs, cristobalite and/or tridymite [61], which were not detected in the samples. Nevertheless, it is possible that small amounts of tridymite are present in the samples but not detectable due to the method’s low detection limit [62]. Anorthite is stable up to 1100/1200 °C [59], which defines it as a relict mineral. In carbonate kaolinitic clays, kaolinite is transformed into metakaolinite, which, after ~800 °C (decomposition temperature of the calcite structure) is successively transformed to gehlenite, which in turn, is transformed into anorthite [34,63]. Anorthite also forms after the decomposition of illite at temperatures of ~850 °C in the presence of CaO [44,52]. In this regard, in clay heated to 1100 °C, anorthite can be both a relict mineral and a newly formed high-temperature phase. High-temperature phases of the melilite group, gehlenite and akermanite, are also present in the samples. The transformation of kaolinite into metakaolinite was recorded by the DSC effect at 519 °C, and the decomposition of calcite was in the range of 790–815 °C (Table 4), indicating a crystallization of gehlenite above 815 °C [6,28]. The formation of akermanite is associated with the release of Mg upon decomposition of the minerals in which it is incorporated, with its crystallization temperature in the range of 950–1100 °C [46]. The exo-effect was recorded on the DSC curve at temperatures of ~950 °C (Figure 10).
Ceramic samples C1 and C2 contain quartz, anorthite, gehlenite, and calcite, while illite, muscovite, and akermanite are not recorded. The presence of the high-temperature phase akermanite defines the calcite as secondary, recrystallized after the ceramics were fired, and cannot be used to determine the firing temperature [64]. Similarly, no effects were recorded on the DSC curve after the decomposition temperature of the recrystallized calcite, suggesting that the firing temperature was higher than the decomposition temperature of illite but did not reach the crystallization temperature of akermanite [65]. The absence of DSC effects after 750 °C defines a lower limit on the firing temperature as muscovite’s structure breakdown temperature and an upper limit, which is the crystallization temperature of akermanite. Considering these temperatures in the clay samples, the ceramic fragments were fired in the temperature range of 920–945 °C. The determined high firing temperatures support the archaeological evidence for the production functions of the two-chambered kiln 1 (Figure 1).
Samples C5 and C6 are composed of quartz, anorthite, illite, muscovite, and calcite. No newly formed high-temperature phases were recorded, which defines the registered calcite as a relict mineral from the source clay [64]. Quartz, anorthite, illite, and muscovite are also relict. The measured decomposition temperature of relict calcite defines an upper firing limit for the ceramic of 787.8 °C for C5 and 766.5 °C for C6 (Table 4). The lower temperature firing limit can be taken as the chlorite decomposition temperature recorded by thermal analysis for samples R2 and R4, namely ~570 °C. Such low temperatures suggest the possibility of using kiln 2, a single-chamber furnace (Figure 2), which can reach temperatures up to 450–600 °C, not only for domestic but also for production purposes [8,11].
The equivalent firing temperature of samples C3 and C4 can be partially judged from the thermal analysis data. The endothermic effect recorded on the DSC curve at 880 °C reflects the breakdown of the muscovite structure, while the exothermic effect at 951 °C is associated with the crystallization of high-temperature phases—Mg-spinel, mullite, which is characteristic of non-carbonate clays [52]. These data determine that the temperature in the furnace did not reach 880 °C. Nevertheless, the lack of a proven raw material source requires further research to determine its equivalent firing temperature.

5.4. Archaeological Importance

The Roman province of Moesia, later Lower Moesia, was a contact zone in the field of ceramic production between Greek and Roman influence and local traditions. With the consolidation of Roman domination in the 1st century, thanks to the military units, imports from the western provinces penetratedthe luxury pottery terra sigillata. With the quality of the clay, exquisite in form and decoration, it became a role model. However, terra sigillata is costly, difficult to transport, and hardly accessible to a wide range of buyers, which is why it was gradually being replaced by the high-quality production of local ceramic production centers. Through their study, the nature of production, the development of forms and uses, technological processes, the organization of its distribution, and the resulting connections and relationships between indigenous peoples and settlers are elucidated [66]. In this regard, the results obtained, with their relation to the geological setting of the area, are of archaeological importance for the study of the ceramics from the sites within the territory of the province of Moesia. In addition, they allow their comparison with data from other archaeological sites located in the territories of the provinces of the Roman Empire, where production activities took place. Such a comparison allows the study of the choice of production technology in local pottery production centers, which also depended to a large extent on the raw material source used [67].

6. Conclusions

Through the conducted research, the source for the production of the ceramic was identified, and the equivalent firing temperature was determined.
The clay used to make fragments C1, C2, C5, and C6 is exposed on the ground surface within the archaeological site. The clay is calcareous (with a CaOcalcite% content between 13.26% and 27.32%) illite–kaolinitic with varying percentages of clay and non-clay minerals. It was mined from a section/sections with a lower carbonate content and, therefore, a higher iron content. An additional amount of sandy component was added to the clay during paste preparation, which eliminated shrinkage during firing and produced a stronger final product. The distinction of lower from higher calcareous clay is aided by the difference in color of the two types of raw material—darker and lighter with a beige tint, respectively. The raw material used and the way in which the paste is prepared show considerable knowledge in the field of pottery.
No raw material source has been identified for the remaining ceramic fragments examined, but they have been shown to be made from non-calcareous clay, of which none has been identified within the site. The archaeological and geological evidence suggests the possibility that these fragments were made at the investigated site, but from the raw material obtained from a location distant from the site, or that the pottery was imported.
The determined equivalent firing temperatures of the ceramic with proven source material are in two temperature ranges: 570–760 °C and 920–945 °C. The higher firing temperature range is consistent with the archaeological view regarding kiln type 1 (pottery kiln). The lower firing temperature range indicates the possibility that kiln 2 could have been used for firing low-temperature ceramics, in addition to domestic use. The studied handmade pottery (C5) dated as pre-Roman, was made from clay extracted from the site, and was fired in the low temperature range (570–760 °C), which can be seen as a continuity in the knowledge of clay extraction and pottery production over time.
The archaeological evidence (the identification of kilns with possible production functions and the discovery of handmade pottery), the geological setting of the area (the presence of different types of clays exposed on the ground surface), and the presence of a water source are supported by the experimental studies and are the basis for identifying the site at Makariopolsko village as a ceramic production center. A comprehensive survey will clarify its status (settlement identification).
Although there are other archaeologically proven workshops and ceramic centers in the study area, there has been no research on either the source material or the ceramic artifacts. The new data on ceramic provenance and equivalent firing temperature of the ceramics for the site at Makariopolsko are essential for the site itself, but also outline a clear perspective for the upcoming archaeomineralogical studies of the clay and ceramics in this area. Comparing such results will resolve both local and more regional archaeological questions by identifying:
  • the origin and production technology of fragments C3 and C4;
  • opportunities for raw material extraction from locations remote from the sites, controlled both by the ability to transport the extracted clay and by restrictions on land ownership;
  • the technology of pottery production in ceramic centers and workshops (raw material and firing conditions) in the Roman and earlier eras in this region;
  • relationships between different production centers—possibility of knowledge continuity, trade relations, etc.;
  • the possibility of comparing the production in the region with that of other ceramic centers in Bulgaria.

Author Contributions

Conceptualization, B.K. and S.T.; methodology, B.K. and K.M.; software, K.M.; validation, B.K., K.M. and S.T.; formal analysis, B.K. and K.M.; investigation, B.K. and S.T.; resources, S.T. and Ch.L.; data curation, R.B. and Ch.L.; writing—original draft preparation, B.K., S.T. and K.M.; writing—review and editing, K.M.; visualization, S.T., Ch.L. and R.B.; supervision, B.K. and S.T.; project administration, B.K.; funding acquisition, B.K. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the National Science Fund of Bulgaria under grant KP-06-N80/5 (B.K.).

Data Availability Statement

The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author.

Acknowledgments

The authors would like to thank Tsvetelina Slavkova from the Archaeological Museum “Mechislav Domaradski” for processing the ceramic material. They are also grateful to the head of research, Simeon Stanchev (History Museum—Popovo), for the shared information. Special thanks go to Zhivko Uzunov. The authors gratefully acknowledge New Bulgarian University, the Department of Natural Sciences, and the Geology laboratory—BF.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
RTroom temperature
XRFX-ray fluorescence
CAcluster analysis
PXRDpowder X-ray diffraction
TGthermogravimetric
DTGdifferential thermogravimetric
DSCdifferential scanning calorimetry

References

  1. Goffer, Z. Archaeological Chemistry, 2nd ed.; Chemical Analysis; Wiley-Interscience: Hoboken, NJ, USA, 2007. [Google Scholar] [CrossRef]
  2. Barker, D.; Majewski, T. Ceramic Studies in Historical Archaeology. In The Cambridge Companion to Historical Archaeology; Hicks, D., Beaudry, M.C., Eds.; Cambridge University Press: Cambridge, UK, 2006; pp. 205–232. [Google Scholar] [CrossRef]
  3. Orser, C.E. Ceramics as Dating Tool in Historical Archaeology. In Encyclopedia of Global Archaeology; Smith, C., Ed.; Springer: New York, NY, USA, 2014; pp. 1291–1292. [Google Scholar] [CrossRef]
  4. Sánchez Ramos, S.; Bosch Reig, F.; Gimeno Adelantado, J.; Yusá Marco, D.; Doménech Carbó, A. Application of XRF, XRD, Thermal Analysis, and Voltammetric Techniques to the Study of Ancient Ceramics. Anal. Bioanal. Chem. 2002, 373, 893–900. [Google Scholar] [CrossRef] [PubMed]
  5. Liritzis, I.; Xanthopoulou, V.; Palamara, E.; Papageorgiou, I.; Iliopoulos, I.; Zacharias, N.; Vafiadou, A.; Karydas, A.G. Characterization and Provenance of Ceramic Artifacts and Local Clays from Late Mycenaean Kastrouli (Greece) by Means of p-XRF Screening and Statistical Analysis. J. Cult. Herit. 2020, 46, 61–81. [Google Scholar] [CrossRef]
  6. Cultrone, G.; Rodriguez-Navarro, C.; Sebastian, E.; Cazalla, O.; De La Torre, M.J. Carbonate and Silicate Phase Reactions during Ceramic Firing. Eur. J. Mineral. 2001, 13, 621–634. [Google Scholar] [CrossRef]
  7. Liritzis, I.; Laskaris, N.; Vafiadou, A.; Karapanagiotis, I.; Volonakis, P.; Papageorgopoulou, C.; Bratitsi, M. Archaeometry: An overview. Sci. Cult. 2020, 6, 49. [Google Scholar] [CrossRef]
  8. Harizanov, A. Ceramic Kilns in the Territory of Modern Bulgaria from the 1st to the 6th Century AD. Ph.D. Thesis, National Archaeological Institute with Museum—BAS, Sofia, Bulgaria, 2019; p. 15. [Google Scholar]
  9. Aladzhov, A.; Dimitrov, Y. Archaeological field survey along the Hemus highway, stages 2 and 3, sections 7 and 8, km 276–327. In Archaeological Discoveries and Excavations 2015; National Archaeological Institute with Museum at the Bulgarian Academy of Sciences (NAIM-BAS): Sofia, Bulgaria, 2016; pp. 837–839. [Google Scholar]
  10. Todorova, S.; Lalov, C. Rescue archaeological investigations at site 83 near Makariopolsko village, Targovishte region. In Archaeological Discoveries and Excavations 2024; National Archaeological Institute with Museum at the Bulgarian Academy of Sciences (NAIM-BAS): Sofia, Bulgaria, in press.
  11. Klenina, E.Û. Stolovaâ i Kuhonnaâ Keramika III–VI Vekov iz Nov (Severnaâ Bolgariâ); Novae: Studies and Materials; Instytut Historii Uniwersytetu im. A. Mickiewicza: Sevastopol: Poznań, Poland, 2006. [Google Scholar]
  12. Philipov, L.; Decheva, A. Geological Maps of Bulgaria, M 1:100,000, Map Sheet Razgrad. M 1:100,000; Geology and Geophysics AD: Sofia, Bulgaria, 1992. [Google Scholar]
  13. Philipov, L.; Decheva, A. Explanatory Note to the Geological Map of Bulgaria M 1:100,000. Map Sheet Razgrad; Bolid: Sofia, Bulgaria, 1994. [Google Scholar]
  14. Borisova-Katsarova, I. Handmade Roman pottery—Questions unasked and answers unsought. Stud. Archaeol. Univ. Serdicensis 2021, 7, 153–177. [Google Scholar]
  15. Peacock, D.P.S. Pottery in the Roman World: An Ethnoarchaeological Approach; Longman Archaeology Series; Longman: London, UK; New York, NY, USA, 1982. [Google Scholar]
  16. Cardiano, P.; Ioppolo, S.; De Stefano, C.; Pettignano, A.; Sergi, S.; Piraino, P. Study and Characterization of the Ancient Bricks of Monastery of “San Filippo Di Fragalà” in Frazzanò (Sicily). Anal. Chim. Acta 2004, 519, 103–111. [Google Scholar] [CrossRef]
  17. Martin, A.; Inoue, H.; Yamada, Y.; Kohno, H. Standardless XRF Analysis for LOI-Rich Rock Samples by Scatter Fundamental Parameter Method. Geochim. Cosmochim. Acta Suppl. 2005, 69, A796. Available online: https://goldschmidtabstracts.info/2005/796.pdf (accessed on 12 March 2025).
  18. Kenkel, J. Analytical Chemistry for Technicians, 4th ed.; Taylor & Francis: London, UK, 2014. [Google Scholar]
  19. King, B.; Vivit, D. Loss-on-ignition Corrections in the XRF Analysis of Silicate Rocks. X-Ray Spectrom. 1988, 17, 145–147. [Google Scholar] [CrossRef]
  20. Shaltout, A.A.; Gomma, M.M.; Ali-Bik, M.W. Utilization of Standardless Analysis Algorithms Using WDXRF and XRD for Egyptian Iron Ore Identification. X-Ray Spectrom. 2012, 41, 355–362. [Google Scholar] [CrossRef]
  21. Pillay, A.E.; Punyadeera, C.; Jacobson, L.; Eriksen, J. Analysis of Ancient Pottery and Ceramic Objects Using X-Ray Fluorescence Spectrometry. X-Ray Spectrom. 2000, 29, 53–62. [Google Scholar] [CrossRef]
  22. Papageorgiou, I. Ceramic Investigation: How to Perform Statistical Analyses. Archaeol. Anthropol. Sci. 2020, 12, 210. [Google Scholar] [CrossRef]
  23. Rahman, M.S.; Metselaar, H.S.C.; Bushroa, A.R. XRF Data-Based Statistical Clustering of Ancient Ceramic Artifacts and Analysing Provenance with Thermo-Mineralogical Analysis. Spectrochim. Acta Part B At. Spectrosc. 2025, 224, 107090. [Google Scholar] [CrossRef]
  24. Liritzis, I.; Zacharias, N. Portable XRF of Archaeological Artifacts: Current Research, Potentials and Limitations. In X-Ray Fluorescence Spectrometry (XRF) in Geoarchaeology; Shackley, M.S., Ed.; Springer: New York, NY, USA, 2011; pp. 109–142. [Google Scholar] [CrossRef]
  25. Kabekkodu, S.; Dosen, A.; Blanton, T. PDF (Powder Diffraction File); ICDD: Newtown Square, PA, USA, 2001. [Google Scholar]
  26. Xanthopoulou, V.; Iliopoulos, I.; Liritzis, I. Characterization techniques of clays for the archaeometric study of ancient ceramics: A review. Sci. Cult. 2020, 6, 73–86. [Google Scholar] [CrossRef]
  27. Bergaya, F. Handbook of Clay Science, 1st ed.; Developments in Clay Science; Elsevier: Amsterdam, The Netherlands; Heidelberg, Germany, 2008. [Google Scholar]
  28. Ion, R.-M.; Ion, M.-L.; Fierascu, R.C.; Serban, S.; Dumitriu, I.; Radovici, C.; Bauman, I.; Cosulet, S.; Niculescu, V.I.R. Thermal Analysis of Romanian Ancient Ceramics. J. Therm. Anal. Calorim. 2010, 102, 393–398. [Google Scholar] [CrossRef]
  29. Brown, M.E.; Gallagher, P.K. Handbook of Thermal Analysis and Calorimetry. Vol. 2: Applications to Inorganic and Miscellaneous Materials; Elsevier: Amsterdam, The Netherlands, 2004. [Google Scholar]
  30. Meyers, K.S.; Speyer, R.F. Thermal analysis of clays. In Handbook of Thermal Analysis and Calorimetry. 2. Applications to Inorganic and Miscellaneous Materials; Brown, M.E., Gallacher, P.K., Eds.; Elsevier: Amsterdam, The Netherlands, 2004. [Google Scholar]
  31. Földvári, M. Handbook of Thermogravimetric System of Minerals and Its Use in Geological Practice; Occasional papers of the Geological Institute of Hungary; Geological Institute of Hungary: Budapest, Hungary, 2011. [Google Scholar]
  32. Palanivel, R.; Kumar, U.R. Thermal and Spectroscopic Analysis of Ancient Potteries. Rom. J. Phys. 2011, 56, 195–208. [Google Scholar]
  33. Labus, M.; Matyasik, I.; Ziemianin, K. Thermal Decomposition Processes in Relation to the Type of Organic Matter, Mineral and Maceral Composition of Menilite Shales. Energies 2023, 16, 4500. [Google Scholar] [CrossRef]
  34. Xue, H.; Dong, X.; Fan, Y.; Ma, X.; Yao, S. Study of Structural Transformation and Chemical Reactivity of Kaolinite-Based High Ash Slime during Calcination. Minerals 2023, 13, 466. [Google Scholar] [CrossRef]
  35. Marsh, A.; Heath, A.; Patureau, P.; Evernden, M.; Walker, P. Alkali Activation Behaviour of Un-Calcined Montmorillonite and Illite Clay Minerals. Appl. Clay Sci. 2018, 166, 250–261. [Google Scholar] [CrossRef]
  36. Hatakeyama, T.; Liu, Z. Handbook of Thermal Analysis; Wiley: Chichester, UK, 2000. [Google Scholar]
  37. Pei, Z.; Lin, M.; Liu, Y.; Lei, S. Dissolution Behaviors of Trace Muscovite during Pressure Leaching of Hydrothermal Vein Quartz Using H2SO4 and NH4Cl as Leaching Agents. Minerals 2018, 8, 60. [Google Scholar] [CrossRef]
  38. Kotryová, B.; Ondruška, J.; Štubňa, I.; Bačík, P. Thermoanalytical Investigation of Ancient Pottery. AIP Conf. Proc. 2016, 1752, 040016. [Google Scholar] [CrossRef]
  39. Hamilton, A.; Hall, C. A Review of Rehydroxylation in Fired-Clay Ceramics. J. Am. Ceram. Soc. 2012, 95, 2673–2678. [Google Scholar] [CrossRef]
  40. Muller, F.; Drits, V.; Plançon, A.; Robert, J.-L. Structural Transformation of 2:1 Dioctahedral Layer Silicates during Dehydroxylation-Rehydroxylation Reactions. Clays Clay Miner. 2000, 48, 572–585. [Google Scholar] [CrossRef]
  41. Moropoulou, A.; Bakolas, A.; Bisbikou, K. Characterization of Ancient, Byzantine and Later Historic Mortars by Thermal and X-Ray Diffraction Techniques. Thermochim. Acta 1995, 269–270, 779–795. [Google Scholar] [CrossRef]
  42. Böke, H.; Akkurt, S.; İpekoğlu, B.; Uğurlu, E. Characteristics of Brick Used as Aggregate in Historic Brick-Lime Mortars and Plasters. Cem. Concr. Res. 2006, 36, 1115–1122. [Google Scholar] [CrossRef]
  43. Lee, W.E.; Souza, G.P.; McConville, C.J.; Tarvornpanich, T.; Iqbal, Y. Mullite Formation in Clays and Clay-Derived Vitreous Ceramics. J. Eur. Ceram. Soc. 2008, 28, 465–471. [Google Scholar] [CrossRef]
  44. Pérez-Monserrat, E.M.; Maritan, L.; Garbin, E.; Cultrone, G. Production Technologies of Ancient Bricks from Padua, Italy: Changing Colors and Resistance over Time. Minerals 2021, 11, 744. [Google Scholar] [CrossRef]
  45. Derkowski, A.; Kuligiewicz, A. Thermal Analysis and Thermal Reactions of Smectites: A Review of Methodology, Mechanisms, and Kinetics. Clays Clay Miner. 2022, 70, 946–972. [Google Scholar] [CrossRef]
  46. Camara, C.A.; Lopes, G.; Schiavon, N.; Mirão, J.; Beltrame, M. Islamic Middle Ages Pottery from Muge (Portugal), Serradinho Archaeological Site—A Long-Lasting Tradition of Pottery Production. Ceramics 2025, 8, 31. [Google Scholar] [CrossRef]
  47. Chamley, H. Clay Sedimentology; Springer: Berlin/Heidelberg, Germany, 1989. [Google Scholar] [CrossRef]
  48. Papadopoulou, D.N.; Lalia-Kantouri, M.; Kantiranis, N.; Stratis, J.A. Thermal and Mineralogical Contribution to the Ancient Ceramics and Natural Clays Characterization. J. Therm. Anal. Calorim. 2006, 84, 39–45. [Google Scholar] [CrossRef]
  49. Trindade, M.; Dias, M.; Coroado, J.; Rocha, F. Mineralogical Transformations of Calcareous Rich Clays with Firing: A Comparative Study between Calcite and Dolomite Rich Clays from Algarve, Portugal. Appl. Clay Sci. 2009, 42, 345–355. [Google Scholar] [CrossRef]
  50. Badica, P.; Alexandru-Dinu, A.; Grigoroscuta, M.A.; Burdusel, M.; Aldica, G.V.; Sandu, V.; Bartha, C.; Polosan, S.; Galatanu, A.; Kuncser, V.; et al. Mud and Burnt Roman Bricks from Romula. Sci. Rep. 2022, 12, 15864. [Google Scholar] [CrossRef] [PubMed]
  51. Khitab, A.; Riaz, M.S.; Jalil, A.; Khan, R.B.N.; Anwar, W.; Khan, R.A.; Arshad, M.T.; Kirgiz, M.S.; Tariq, Z.; Tayyab, S. Manufacturing of Clayey Bricks by Synergistic Use of Waste Brick and Ceramic Powders as Partial Replacement of Clay. Sustainability 2021, 13, 10214. [Google Scholar] [CrossRef]
  52. El Ouahabi, M.; Daoudi, L.; Hatert, F.; Fagel, N. Modified Mineral Phases During Clay Ceramic Firing. Clays Clay Miner. 2015, 63, 404–413. [Google Scholar] [CrossRef]
  53. Maniatis, Y.; Tite, M.S. Technological Examination of Neolithic-Bronze Age Pottery from Central and Southeast Europe and from the Near East. J. Archaeol. Sci. 1981, 8, 59–76. [Google Scholar] [CrossRef]
  54. Kostova, B.; Dumanov, B.; Mihaylova, K. Archaeological bricks and tiles from southeast Bulgaria- determination of raw material and production technology by chemical, phase, and thermal analyses. Mediterr. Archaeol. Archaeom. 2023, 23, 1–22. [Google Scholar] [CrossRef]
  55. Kostova, B.; Mihaylova, K.; Dumanov, B.; Berberova, R. Archaeometric study of pithoi from Bulgarian archaeological sites. Sci. Cult. 2024, 10, 57–76. [Google Scholar] [CrossRef]
  56. Pollard, A.M.; Heron, C. Archaeological Chemistry, 2nd ed.; Royal Society of Chemistry: Cambridge, UK, 2008. [Google Scholar]
  57. Bauer, A.; Velde, B.D. Geochemistry at the Earth’s Surface: Movement of Chemical Elements; Springer: Berlin/Heidelberg, Germany, 2014. [Google Scholar] [CrossRef]
  58. Kornilov, A.V. Reasons for the Different Effects of Calcareous Clays on Strength Properties of Ceramics. Glass Ceram. 2005, 62, 391–393. [Google Scholar] [CrossRef]
  59. Aras, A.; Kiliç, S. The Mineralogy and Firing Behaviour of Pottery Clays of the Lake Van Region, Eastern Turkey. Clay Miner. 2017, 52, 453–468. [Google Scholar] [CrossRef]
  60. Friede, H.; Hejja, A.; Koursaris, A. Arhaeo-metallurgical studies of iron smelting slags from prehistoric sites in Southern Africa. J. S. Afr. Inst. Min. Metall. 1982, 82, 38–48. [Google Scholar]
  61. Stevens, S.J.; Hand, R.J.; Sharp, J.H. Polymorphism of Silica. J. Mater. Sci. 1997, 32, 2929–2935. [Google Scholar] [CrossRef]
  62. Bitay, E.; Kacsó, I.; Tănăselia, C.; Toloman, D.; Borodi, G.; Pánczél, S.-P.; Kisfaludi-Bak, Z.; Veress, E. Spectroscopic Characterization of Iron Slags from the Archaeological Sites of Brâncoveneşti, Călugăreni and Vătava Located on the Mureş County (Romania) Sector of the Roman Limes. Appl. Sci. 2020, 10, 5373. [Google Scholar] [CrossRef]
  63. Traoré, K.; Kabré, T.S.; Blanchart, P. Gehlenite and Anorthite Crystallisation from Kaolinite and Calcite Mix. Ceram. Int. 2003, 29, 377–383. [Google Scholar] [CrossRef]
  64. Tarhan, İ.; Massa, M.; Tuna, Y.; Şahin, F. An Archaeometric Study of the Konya Basin Metallic Ware through FTIR and XRD Analysis with Chemometrics: Central Anatolian Early Bronze Age Ceramics. Archaeometry 2023, 65, 136–150. [Google Scholar] [CrossRef]
  65. Zvereva, I.A.; Kuznetsov, V.M.; Zhukov, Y.M.; Mazurkevich, A.N.; Dolbunova, E.V. Thermal Analysis and Multi-Analytical Comparison of Samples of Neolithic Ceramics from Dnepr–Dvina and Low Don Regions. J. Therm. Anal. Calorim. 2019, 138, 1879–1886. [Google Scholar] [CrossRef]
  66. Dimitrova-Milcheva, A. Luxury ceramic—Import and local production from Thrace and Moesia in the 1st–3rd centuries AD. In Archaeology of the Bulgarian Lands; Ivanov, R., Ed.; Bibliophilia: Ivray, Sofia, 2008; Volume 3, pp. 119–161. [Google Scholar]
  67. Hunt, A. (Ed.) The Oxford Handbook of Archaeological Ceramic Analysis, 1st ed.; Oxford University Press: Oxford, UK, 2016. [Google Scholar] [CrossRef]
Geosciences 15 00193 g001
Figure 1. Geodetic survey with mapping of the samples and the kilns (made by V. Valchev, NAIM BAS, with additions made by S. Todorova).
Figure 1. Geodetic survey with mapping of the samples and the kilns (made by V. Valchev, NAIM BAS, with additions made by S. Todorova).
Geosciences 15 00193 g001
Geosciences 15 00193 g002
Figure 2. Kiln 1 (two-chambered): 1. horizontal plan; 2. cross-sections (drawings T. Todorov); 3. photo of the kiln with details of the stoking channel and the support (photo Ch. Lalov).
Figure 2. Kiln 1 (two-chambered): 1. horizontal plan; 2. cross-sections (drawings T. Todorov); 3. photo of the kiln with details of the stoking channel and the support (photo Ch. Lalov).
Geosciences 15 00193 g002
Geosciences 15 00193 g003
Figure 3. Kiln 2 (one-chambered): 1 horizontal plan (1.1 the kiln; 1.2. the stoke pit); cross-sections (drawings T. Todorov); photo of the kiln with details (photo Ch. Lalov).
Figure 3. Kiln 2 (one-chambered): 1 horizontal plan (1.1 the kiln; 1.2. the stoke pit); cross-sections (drawings T. Todorov); photo of the kiln with details (photo Ch. Lalov).
Geosciences 15 00193 g003
Geosciences 15 00193 g004
Figure 4. Simplified geological map of the area (by Phollipov ana Decheva, modified) [12]: rK1h-b—Razgrad Formation (calcareous limestones and marls); eQp2–3—eolian formations (clayey loess); e-a-dQp—eolian–alluvial–delluvial formations (loess clays).
Figure 4. Simplified geological map of the area (by Phollipov ana Decheva, modified) [12]: rK1h-b—Razgrad Formation (calcareous limestones and marls); eQp2–3—eolian formations (clayey loess); e-a-dQp—eolian–alluvial–delluvial formations (loess clays).
Geosciences 15 00193 g004
Geosciences 15 00193 g005
Figure 5. Clay samples—photographed after drying at room temperature (RT).
Figure 5. Clay samples—photographed after drying at room temperature (RT).
Geosciences 15 00193 g005
Geosciences 15 00193 g006
Figure 6. Ceramic sherds.
Figure 6. Ceramic sherds.
Geosciences 15 00193 g006
Geosciences 15 00193 g007
Figure 7. CA results—graphic representation by dendrogram.
Figure 7. CA results—graphic representation by dendrogram.
Geosciences 15 00193 g007
Geosciences 15 00193 g008
Figure 8. PXRD patterns of (a) clay samples; (b) clay samples heated at 1100 °C.
Figure 8. PXRD patterns of (a) clay samples; (b) clay samples heated at 1100 °C.
Geosciences 15 00193 g008
Geosciences 15 00193 g009
Figure 9. PXRD patterns of ceramic samples.
Figure 9. PXRD patterns of ceramic samples.
Geosciences 15 00193 g009
Geosciences 15 00193 g010
Figure 10. Thermal analysis results—TG, DTG, and DSC curves of representative samples: clay (R4) and pottery (C2, C4, and C6).
Figure 10. Thermal analysis results—TG, DTG, and DSC curves of representative samples: clay (R4) and pottery (C2, C4, and C6).
Geosciences 15 00193 g010
Table 1. Sample number and type, sampling location (for clay), and discovery location (for ceramic fragments) at the site.
Table 1. Sample number and type, sampling location (for clay), and discovery location (for ceramic fragments) at the site.
SampleTypeLocation in the Archaeological Site
R1clayarchaeological sondage 12
R2clayarchaeological sondage 18,
R3clayarchaeological sondage 21
R4clayarchaeological sondage 30, structure 14
C1amphora(?), fragment of a wall made on a wheel, beige–brown color of the surface, and pastearchaeological sondage 2, structure 1
C2vessel made on a wheel, beige color of the surface, and pastearchaeological sondage 24, structure 11, kiln 1
C3fragment of a wall made on a wheel, red–brown color of the surface, and pastearchaeological sondage 24, structure 11—stoke pit—kiln 1)
C4greyware pottery—a bowl, orifice fragment, made on a wheelarchaeological sondage 30
C5handmade pottery—pot(?) wall fragment, beige–brown surface and pastearchaeological sondage 36, structure 9
C6bowl, orifice fragment, made on a wheel, beige–brown surface and fracture, red poor varnish on the outer surfacearchaeological sondage 37, structure 12
Table 2. XRF results of clay and ceramic samples.
Table 2. XRF results of clay and ceramic samples.
Chemical CompositionSample
R1R2R3R4C1C2C3C4C5C6
Main elements (oxide/wt%)SiO224.0234.9823.5132.5142.5244.3059.9557.0541.4235.83
Al2O36.4510.176.429.0116.5916.9117.5018.1613.2013.56
CaO34.3721.5730.1127.5016.2321.244.462.3113.8324.48
MgO1.232.141.481.392.302.481.711.541.641.90
K2O1.372.281.441.934.072.544.383.652.703.05
Na2O--1.15-------
Fe2O32.914.832.733.967.577.526.597.795.765.56
TiO20.380.700.330.560.860.910.880.930.710.70
LOI29.1322.9532.6422.969.343.904.398.2820.4914.70
Total99.8799.6399.8199.8399.4899.7999.8699.7199.7799.79
Rare elements (elements/ppm)V-172.64-----28.26--
Cr---195.56----155.04181.51
Mn374.87775.73299.96490.44758.29636.34585.701086.81788.19766.31
Ni----87.6398.93- 98.75--
Cu87.19120.03-81.24130.37145.10134.43113.57130.21-
Zn-136.3892.28-199.45186.43135.77191.69139.94181.69
Rb-115.5576.3894.40198.13181.90177.48165.22101.06131.04
Sr623.24350.52649.34601.29551.96682.60206.16163.65356.34591.46
Zr31.48196.79-56.4680.54112.41217.30412.84205.43102.30
Table 3. PXRD results—mineral/phase composition of clay (raw and heated at 1100 °C) and ceramic samples.
Table 3. PXRD results—mineral/phase composition of clay (raw and heated at 1100 °C) and ceramic samples.
OriginMineral/PhaseGeneral Formula *Reference [25]
Raw mineralsquartzSiO2#06-175
kaoliniteAl2Si2O5(OH)4#29-1488
illite(Si4−xAlx)4(Al2−yMgy)4O10(OH)2(x + y)K+#02-0056
muscovite(Si3Al)4(Al2)6O10(OH)2K+#34-0175
chlorite(R2+,R3+)3(Si4−xAlx)O10(OH)2 (R2+ = Mg2+,Fe2+,Mn2+; R3+ = Al3+,Fe3+,Cr3+)#83-1381
Raw and recrystallized mineralcalciteCaCO3#06-6528
Raw mineral and new-formed high-temperature phase (?)anorthiteCa(Al2Si2O8)#41-1486
New-formed high-temperature phasesgehleniteCa2Al2SiO7#04-016-0209
akermaniteCa2Mg[Si2O7]#35-0592
* The general chemical formulas of the phyllosilicates are presented according to Bergaya et al., 2006 [27].
Table 4. Thermal analysis results—CaO (measured and calculated) and calcite.
Table 4. Thermal analysis results—CaO (measured and calculated) and calcite.
SampleTinfl Calcite/
T°C 		MLCO2/
%		CaOcalcite/wt%
Calculated		CaOtotal/wt%
Measured		CaOsilicates/amorphous wt%
Calculated		Calcite/%
Calculated
R1815.221.4727.3234.377.0548.63
R2793.510.4213.2621.578.3123.60
R3812.519.4827.7930.112.3249.47
R4816.514.5318.4927.509.0132.91
C1712.72.403.0516.2313.185.43
C2728.91.531.9421.2419.033.45
C3701.10.651.454.463.011.45
C4---2.312.31-
C5787.83.965.0413.838.798.97
C6766.57.068.9824.4815.5015.98
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Kostova, B.; Todorova, S.; Mihaylova, K.; Lalov, C.; Berberova, R. A Provenance Study of Ceramic Artifacts from the Area of Makariopolsko Village, NE Bulgaria. Geosciences 2025, 15, 193. https://doi.org/10.3390/geosciences15060193

AMA Style

Kostova B, Todorova S, Mihaylova K, Lalov C, Berberova R. A Provenance Study of Ceramic Artifacts from the Area of Makariopolsko Village, NE Bulgaria. Geosciences. 2025; 15(6):193. https://doi.org/10.3390/geosciences15060193

Chicago/Turabian Style

Kostova, Bilyana, Svetlana Todorova, Katerina Mihaylova, Chavdar Lalov, and Ralitza Berberova. 2025. "A Provenance Study of Ceramic Artifacts from the Area of Makariopolsko Village, NE Bulgaria" Geosciences 15, no. 6: 193. https://doi.org/10.3390/geosciences15060193

APA Style

Kostova, B., Todorova, S., Mihaylova, K., Lalov, C., & Berberova, R. (2025). A Provenance Study of Ceramic Artifacts from the Area of Makariopolsko Village, NE Bulgaria. Geosciences, 15(6), 193. https://doi.org/10.3390/geosciences15060193

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop