Archaeometric Research of Boian Pottery Decoration from the Settlement of Hârșova-Tell

Abstract

This paper presents the results of archaeometric research on the white and red materials used to decorate some ceramic vessels belonging to the Boian culture, the Vidra and Spanţov phases from the 5th millennium BC, discovered at Hârşova-Tell, on the right bank of the Danube, Romania. Digital microscopy, wavelength-dispersive X-ray fluorescence spectroscopy, X-ray diffraction, and Fourier transform infrared spectroscopy were used to identify the morphology, the crystalline phases, chemical formula, and percentage content of each crystalline phase in the white decoration materials and what type of pigment was used to prepare the red paint. The results obtained reveal that the white decorating material in the pottery is mostly made of calcite. It was added after the pottery was fired. Sometimes, powders from burnt bones were also used. The red pigment is hematite. The work contributes to the completion of a puzzle related to the artistic vision of the members of the Boian communities in decorating ceramics with culture-specific motifs, preserving the pattern, and adapting to the local natural resources.

1. Introduction

The use of physical investigation methods in archaeological research helps uncover the technical choices made by diverse human societies throughout history. These methods provide insights into the materials and techniques used by various communities across different time periods. In the archeological ceramics field, the natural sources of clays and decorative materials, the recipes used for the preparation of the ceramic paste, and the temperature of firing ceramics were the materials and techniques considered for achieving various chromatic effects and even to answer the question of what the vessels were used for.

The study focuses on examining ceramic fragments with white and/or red decoration from the Hârșova-Tell (Figure 1), a settlement on the right bank of the Lower Danube in the western-central region of Dobrogea (Figure 2). This site is one of the most impressive archaeological deposits in Europe with a stratigraphy exceeding 12 m. It exhibits continuous habitation from the Boian culture, the Vidra phase, to the period of the Cernavodă I culture, covering the entire Eneolithic Weak; traces of habitation have been identified after this period.

The Boian culture emerged in the present-day territory of Romania, encompassing Muntenia, southeastern Transylvania, and Dobrogea. It was defined and subdivided into four phases based on ceramic stylistics: Bolintineanu, Giulesti, Vidra, and Spanțov [1]. The last two phases of the Boian, the Vidra and Spanţov cultures, from around 4900–4500 BC [2], show the stability of the communities that lived in compact and permanent settlements; the inhabitants were sedentary and involved in farming and animal husbandry and had a strong and vivid connection with their ancestors. The dwellings, surface or semi-buried, had a rectangular shape. Tell-type settlements have a layered structure because of the destruction of dwellings, sometimes caused by deliberate fire, and the building of new ones over the remains of the old ones [3,4]. The appearance of the dwellings and material life is identical to that of the Boian communities on the left bank of the Danube and was based on the raw materials in the surroundings.

In the last two phases of the Boian culture, the distinctive decorative pottery features include the use of white paste inlays for decorative effects on a black polished background. This artistic technique is also found in related cultures in southeastern Europe, such as the Maritsa I–IV culture, in Karanovo (in the south of the Balkan Mountains), in Gradeshnitsa (in the northwest of Bulgaria), and in Polyanica (in northeastern Bulgaria) [2], and could be dated back to the late 6th millennium BC [3,4].

Changes in ceramic styles seem to correspond to the adoption of new techniques based on chipped tools, the demographic growth that will result in the expansion of inhabited areas, advances in pyrotechnics, and even the discovery of the first metallurgical process [2].

Much less common is the situation of applying red pigment over white material or white material over red pigment, reported at Gălățui-Movila Berzei/Romania [5], Nanov-Vistireasa 3/Romania [6], or Radovanu/Romania [7]. During the final phase of the Boian culture, the practice of excision and filling with white paste is linked to graphite painting [7]. In some cases, it is also connected to the application of red pigment on the inner lip of vessels [6]. Additionally, the practice of painting directly on the vessel with white paste reappears [8]. This technique became a common practice in Gumelnița culture [9,10], which succeeded the Boian culture in the same area.

2. Materials and Methods

2.1. Selection and Indexing of Samples

In the Hârșova-Tell site, the living levels belonging to the Boian culture are more than 8 m deep. This fact influenced research strategies to focus on excavating small areas in the southern base of the tell. The goal was to recover materials from specific archaeological layers (Figure 3).

The ceramic fragments in this research, which belong to the layer attributed to the Boian culture (Figure 3), the Vidra and Spanțov phases (around 4900–4600 BC), from the archaeological ensembles of Hârșova-Tell, were discovered during the excavations from the 2016–2017 campaigns [11,12]. Three housing structures were studied in these campaigns. These were from the category of non-burned structures and were located on larger surfaces or only corners of the surfaces. The houses investigated were numbered chronologically, from top (the latest) to bottom (the earliest): 111 = 112, 113, and 114. This succession of dwellings was dug in an archaeological deposit with a thickness of about 50–65 cm, which has been understood as an outdoor place where community members engaged in different domestic activities, such as cutting wood or horns, storing bundles of reeds, and discarding scraps and garbage, a fact proven by the number of bones, the large number of limestone pieces, sometimes with burn marks, or the presence of coprolites (Figure 3) [11,12].

Out of the thousands of ceramic fragments in the lot, only 11 have been found to encompass the entire stratigraphic layout of the site and had the most well-preserved white and red decorative elements. These ceramic fragments are indexed with B2.P.10.1.-5 and B2.P.11.1.-6 (

Table 1. Representative macro and microscopic images of the studied samples. Each ceramic fragment’s macro images are marked with a 1 cm scale line.

Ceramic Fragment Index and Description	Microscopic Image Localization on the Sample Surface	Microscopic Images (x40)
B2.P.10.1 is a fragment (4.5 × 3.5 cm) from the body of a vessel, of unknown shape, made from coarse paste (with ceramoclastes), reductive fired to a gray color. The applied exterior decor is of the excised type, applied in wide fields, filled with white paste.
B2.P.10.2 is a fragment (2 × 1.7 cm) from the lip of a small bowl—a miniature vessel with thin walls (0.4 cm), made from fine clay, incompletely fired in an oxidizing environment. The walls have been mechanically polished both on the interior and exterior. On the outside, the lip is covered with red pigment and delineated from the body by an incised line filled with white paste, applied before the red painting.
B2.P.10.3 (inside) is a fragment (3.7 × 1.7 × 0.4 cm) of a cup lip with threshold, made of fine ceramic, mechanically polished on the outside and inside, fired to grey. The narrowed inner edge of the vessel was painted white, over which red paint was applied.
B2.P.10.4 is a fragment (4 × 3.5 cm) from the body of a vessel of unknown shape, made of coarse paste (with ceramoclastes), reductive fired. The exterior surface is decorated with grooves and wider spaces, currently rough, but most likely originally filled with white paste. The interior surface is smooth and shows traces of red pigment.
B2.P.10.5 is a fragment (6.3 × 6.5 × 0.7 cm) of a lip from a bowl with a slightly flared edge, made from semi-fine paste (with ceramoclastes), reductive fired to gray. The exterior is decorated with grooves that delineate narrow excised bands, which no longer retain the white paste. On the rougher flared edge on the interior, red pigment has been applied, which has migrated to the exterior.
B2.P.11.1 is a fragment (11 × 5.2 cm) from the body of an unknown-shaped vessel, made of coarse paste (with ceramoclastes), reductive fired. The exterior surface is decorated with closely spaced horizontal grooves, which exhibit incisions. These have been mechanically polished and covered with varnish. The incrustations between the grooves have been filled with white paste, which is extremely degraded. The interior has been mechanically smoothed. It shows traces of deposits from the soil.
B2.P.11.2 is a fragment (5.1 × 3.5 cm) from the body of a vessel, unknown shape, semi-fine paste (with small ceramoclaste), fired to gray. On the outside, it has a decoration consisting of horizontal successive grooves, with the spaces between them filled with white paste.
B2.P.11.3 is a fragment (4.5 × 4 cm) from a miniature vessel, with a short lip, made of semi-fine clay (with small ceramoclaste), fired in a reducing environment. The exterior surface is decorated with closely grooves, with the spaces between them excised and filled with white clay. There is a deposit in some places on the lip, also present on the polished inner surface.		a
		b
B2.P.11.4 is a fragment (5 × 4.1 cm) from the body of an unknown-shaped vessel, made of semi-fine, porous paste (containing fine ceramoclasts and fine plant matter), fired in an oxidizing environment, smoked, and decorated on the exterior with deep incisions filled with gray paste very degradated.
B2.P.11.5 is a fragment (4.5 × 2.5 cm) from the body of an unknown-shaped vessel, made of porous semi-fine paste (containing ceramoclasts and fine plant material), reduced fired to gray, decorated on the exterior with deep successive incisions filled with white paste.
B2.P.11.6 is a fragment (4.2 × 2.9 cm) from the body of an unknown-shaped vessel, made of semi-fine, porous clay (containing ceramoclasts and fine plant material), incompletely oxidized during firing, decorated with grooves that delineate fields filled with rich and good preserved white paste.

). Thus, fragment B2.P.10.2 comes from a destroyed dwelling, 111 = 112 (dwelling framed in the Boian culture, Spanțov phase). The other ceramic fragments in this study come from the area surrounding dwelling 114 including the outdoor occupational area.

Out of the 11 ceramic fragments selected, only 3 had enough white material for analysis using X-ray diffraction (XRD) and wavelength-dispersive X-ray fluorescence (WDXRF), which are B2.P.10.1, B2.P.11.2, and B2.P.11.6. According to the styles of ceramics, the fragment B2.P.10.1 is from the Spanțov phase of the Boian culture, while the other two, B2.P.11.2 and B2.P.11.6, are from the Boian-Vidra levels. Also, all 11 samples could be analyzed using digital microscopy and FTIR spectrometry, a method that did not require relatively large sampling quantities. The samples underwent archaeometric analysis to gather precise details on the crystalline phases, chemical formula, and percentage content of each crystalline phase in the white decoration materials and what type of pigment was used to prepare the red paint used to decorate the vessels that the ceramic fragments originated from.

The hypotheses regarding ceramic decoration techniques were as follows: until now, it has been thought that the technology of Boian ceramics relied on firing vessels at low temperatures in a simple furnace. The red pigment paint was applied post-firing, whereas the white paste was used both before and after firing. The challenge in determining when the white paste was applied arises from the additional firings the vessels underwent, particularly those crafted from reddish ceramic material. Studies on ceramic pieces found in Vlădiceasca (Boian–Vidra phase) show that white paste made from calcite or hydroxyapatite was used before firing [13], while in two other pieces discovered in the Sultana Ghețărie site, the white paste was applied post-firing [4].

In this context, the subsequent archaeological questions can be addressed:

• What types of materials were utilized in the decoration of the Boian ceramics studied?
• When were the decorative materials applied: before or after the ceramics were fired?
• Are there similarities and differences to what is seen in other contemporary sites?

2.2. Methods of Analysis

2.2.1. Digital Microscopy

The initial step involved examining the decorated surfaces under a Levenhuk DTX RC4 digital microscope at 40x magnification. A cold corona-type light source was utilized for enhanced brightness in the areas of interest. Through digital microscopy, only the observation of the morphology of the decorated surfaces at a magnification of approximately x40 was achieved. The particle sizes were not visually determined or analyzed using any graphical software.

2.2.2. Sampling for X-Ray Analysis

For X-ray analyses, WDXRF (wavelength-dispersive X-ray fluorescence spectroscopy) and XRD (X-ray diffraction), around 3 g of white powdered material, was sampled only from the areas decorated with white material of the ceramic fragments B2.P.10.1, B2.P.11.2, and B2.P.11.6. A diamond-tipped hand cutter was used for sampling. Then, the sampled powders were mortared up to the limit size given by their hardness. All sample preparation operations took precautions to avoid sample contamination or phase changes. For WDXRF analysis, the powder was mixed with binder BOREOX (FLUXANA GmbH and Co.KG, Bedburg-Hau, Germany). The homogeneous mixture was pressed with the TestChem LPR-250 laboratory press (Sp. z o.o., Radlin, Poland) with a force of 250 kN, making a 4 g sample pellet with a diameter of 32 mm and a height of 3 mm. For XRD analysis, the remaining powder parts were deposited on glass slides (20 × 20 × 0.5 mm).

Wavelength-Dispersive X-Ray Fluorescence Spectroscopy (WDXRF)

The elemental composition and percentage of basic oxides in white decoration materials were determined using the X-ray fluorescence spectrometer WDXRF Supermini200 Rigaku (Tokyo, Japan). This is a benchtop sequential spectrometer equipped with a wavelength dispersive technique and operated by a computer system. It comes equipped with accessories for analyzing solid and liquid samples, including an automatic sample changer with 12 positions. The system is equipped with two detectors (PC and SC) and three analyzing crystals (with automated exchange): LiF (200) for heavy elements (²²Ti—⁹²U) and PET and RX 25 for light elements (⁸O—¹²Mg and ¹³Al—²¹Sc).

The X-ray tube has a Pd anode with a power of 200 W (voltage 50 kV, intensity 4 mA). The detection limit is 1 ppm–10 ppb and the precision is <0.1–0.5%. The optimal experimental setup had the following parameters: EZ Scan analysis, in a vacuum, oxide powder type sample, and pellet covered with a Myler polypropylene film with a thickness of 6 μm. The total time to scan a sample was 30 min.

The experimental data were processed and interpreted using a preset algorithm in the spectrometer software SQX Analysis by Scatter Fundamental Parameters. SQX is a semi-quantitative analysis program that calculates theoretical X-ray intensities using the fundamental parameter method and an internal database, eliminating the need for standard reference materials. The calculation is based on sequential scan measurements from ⁹F to ⁹²U. SQX analysis is designed for elemental analysis in unknown samples.

X-Ray Diffraction (XRD)

X-ray diffraction (XRD) analysis was conducted using the Rigaku Ultima IV X-ray diffractometer (Rigaku, Tokyo, Japan) with Cu--Kα radiation (λ = 0.15406 nm) in order to identify the crystalline phases, chemical formula, and percentage content of each crystalline phase in the white decoration materials.

The optimal experimental setup has the following parameters: Bragg–Brentano beam geometry; anode voltage 40 KV; anodic current intensity 30 mA; continuous scanning at a speed of 1° × min⁻¹, at a step of 0.01°; and angular range (2Ꝋ) = 10–100°.

The interpretation was conducted using the software PDXL 2.2 based on the patterns (ICDD cards—contained in the PDF-4+ database, release 2021) from category S (star patterns—extremely high-quality data), respectively, to I (indexed patterns—indexed high-quality data).

2.2.3. Fourier Transform Infrared Spectroscopy (FTIR)

The FTIR analysis was performed with the Vertex 80 spectrometer (spectral range: 350–8000 cm⁻¹; spectral resolution: 0.2 cm⁻¹, accuracy: 0.1% T), which is equipped with an automatic beam-splitter changer, controlled by software, which allows the extension of the spectral range without affecting the vacuum of the spectrometer.

This vibrational technique offers several advantages: a small sample amount, easy and rapid sample preparation, and short analysis time. The decoration materials on the samples were only found in small quantities on the surface. Sampling for FTIR analysis involved extracting a small part of the material from the most representative area of each ceramic fragment using a spatula with a sharp tip.

3. Results

3.1. Microscopic Images: The Morphology of the Areas Decorated with White and Red Materials

The white decorative material from seven ceramic fragments was analyzed: deposited in a wide decorative field (B2.P.10.1, B2.P.11.6), in narrow, incised (B2.P.10.2) or excised spaces (B2.P.11.1, B2.P.11.2, B2.P.11.4; B2.P.11.5), on the outer surface of some vessels from the black fired category. In the case of samples B2.P.10.1 and B2.P.11.6, it is observed that the white material has a fairly high degree of homogeneity and the lack of inclusions and large-grained sand, present in the other samples, can be noted.

The analysis of red-pigmented decoration materials was conducted on four samples: B2.P.10.2, B2.P.10.3, B2.P.10.4, and B2.P.10.5, with the mention that in two cases, it was used for decorating the inner lip of vessels from the fine (B2.P.10.3) or semi-fine (B2.P.10.5) ceramic categories. In the case of ceramic fragment B2.P.10.4, we believe that the red pigment comes from the inner surface of the vessel, possibly used for storing or preparing the red color. In the case of B2.P.10.2, the red pigment was applied to decorate the outer thickened lip of a bowl.

The ceramic vessels samples B.2.P.10.2 and B2.P.10.3 feature a noticeable combination of red and white in their decoration. In sample B.2.P.10.2, it can be seen that the incision was filled with white fine homogeneous material followed by decoration with red pigment. In the case of sample B2.P.10.3, there is a red color overlay on the white material with sand.

3.2. WDXRF Results

The elemental determinations by WDXRF reveal the massive presence of silicon accompanied by some amounts of elements like aluminum, calcium, sodium, iron, magnesium, and others (

Table 2. Elemental composition (WDXRF) of the white material analyzed, expressed in (mass± SD)%, normalized to 100%.

No.	Element	Mass Percent [%]
		B2.P.10.1.		B2.P.11.2.		B2.P.11.6.
		Value	SD	Value	SD	Value	SD
1	Na	-	-	-	-	0.7705	0.5885
2	Mg	0.4969	0.1120	1.4900	0.1388	2.6781	0.1750
3	Al	13.4478	0.0561	10.0332	0.0666	6.1886	0.1171
4	Si	31.8551	0.0739	25.7762	0.0768	16.8406	0.0812
5	P	0.1515	0.0130	-	-	2.1044	0.0215
6	S	0.0220	0.0082	0.1421	0.0103	0.3551	0.0138
7	Cl	-	-	0.1231	0.0109	0.1093	0.0137
8	K	1.6344	0.0296	2.1269	0.0348	2.5532	0.0443
9	Ca	0.9152	0.0196	10.3836	0.0301	23.9749	0.0441
10	Ti	0.1315	0.0243	0.3633	0.0339	0.2225	0.0353
11	Mn	-	-	0.0434	0.0162	0.1366	0.0264
12	Fe	1.2227	0.0120	2.1860	0.0160	2.6677	0.0279
13	Ni	-	-	0.1128	0.0108	-	-

), which indicates the presence of oxides rich in these elements (

Table 3. Oxides composition (WDXRF) of the white material analyzed, expressed in (mass± SD)%, normalized to 100%.

No.	Oxide	Mass Percent [%]
		B2.P.10.1.		B2.P.11.2.		B2.P.11.6.
		Value	SD	Value	SD	Value	SD
1	Na2O	-	-	-	-	1.0386	0.7934
2	MgO	0.8240	0.1857	2.4705	0.2302	4.4405	0.2902
3	Al2O3	25.4092	0.1061	18.9574	0.1258	11.6932	0.2212
4	SiO2	68.1481	0.1582	55.1435	0.1643	36.0274	0.1737
5	P2O5	0.3472	0.0299	-	-	4.8219	0.0493
6	SO3	0.0549	0.0206	0.3548	0.0257	0.8866	0.0344
7	Cl	-	-	0.1231	0.0109	0.1093	0.0137
8	K2O	1.9687	0.0356	2.5620	0.0419	3.0755	0.0533
9	CaO	1.2805	0.0274	16.4577	0.0421	33.5454	0.0617
10	TiO2	0.2194	0.0406	0.6061	0.0566	0.3711	0.0589
11	MnO	-	-	0.0560	0.0209	0.1764	0.0341
12	Fe2O3	1.7481	0.0172	3.1254	0.0229	3.8140	0.0398
13	NiO	-	-	0.1435	0.0138	-	-

).

3.3. XRD Results

The elemental concentration obtained from each sample analyzed with WDXRF was used to interpret the XRD data. The software enabled the refinement of the candidate phases. The quantitative analysis was performed using the Rietveld method of Whole Powder Pattern Fitting (WPPF) based on crystal structure parameters.

The XRD analyses (

Table 4. The results of the XRD analysis (WPPF): the crystalline phases, the chemical formula, and the percentage content of each crystalline phase, obtained with the help of ICDD cards (PDF-4+, release 2021), category S (star patterns—extremely high-quality data), respectively I (indexed patterns—indexed high-quality data).

No.	Phase Name /DB Card Number	Formula	Mass Percent [%]
			B2.P.10.1	B2.P.11.2	B2.P.11.6
1	Quartz	SiO2	62.0	30.7
	01-070-7344
	04-007-0522				13.6
2	Calcite, syn	Ca(CO3)	-		57.0
	01-083-4602
	01-078-4614			45.5
3	Ilmenite HP, syn, iron titanate	FeTiO3	4.7		-
	01-075-1212
	01-075-1204			8.2
4	Aluminum Titanium Oxide	Al2O5Ti	-	3.7	-
	01-076-8801
5	Nickel Titanate	NiTiO3	-	11.9	-
	04-009-3622
6	Amesite-2H2	Mg2Al2SiO5(OH)4	8.2	-	-
	01-076-0534
7	Calcium Phosphate	Ca3(PO4)2	10.0	-	-
	04-018-9895
8	Rossiantonite	Al3(PO4)(SO4)2(OH)2(H2O)	13.4	-	-
	04-020-1265
9	Berlinite, syn	AlPO4	-	-	6.9
	04-009-5761
10	Sanidine	K(AlSi3O8)	-	-	3.1
	01-077-0982
11	Potassium Calcium Magnesium Phosphate	KCa9Mg(PO4)7	-	-	14.0
	04-017-5717
12	Sodium Oxide	Na2O	-	-	5.4
	04-007-0576
13	Potassium Calcium Phosphate	KCaP3O9	0.86	-	-
	04-015-3680

, Figure 4) show significant quartz percentages in all samples. Important values of calcite were identified in samples B.2.P.11.2 and B2.P.11.6. Calcium and potassium phosphate content were found only in sample B2.P.10.1. Certainly, the chemical compositions of decorative materials reflect those of mineral-rich rocks and clays from the Hârșova area (Dobrogea), which were used as raw materials during that period and are not identical from one sample to another. These contain, in addition to quartz and calcite (white materials), respectively, hematite (red pigment) and many other minerals such asrossiantonite, berlinite, ilmenite, amesite, sanidine, aluminum titanium oxide, nickel titanate, and sodium oxide.

3.4. FTIR Spectra

All samples were analyzed using the FTIR spectroscopy method. The infrared bands found in FTIR spectra are situated within the 450–1500 cm⁻¹ range and these values are listed on each FTIR spectrum chart and in

Table 5. Characteristic infrared bands for the compounds found in the samples analyzed by FTIR.

Mineral/Crystaline Phase	IR Bands (Wavenumber) [cm−1]	Samples
Quartz [14,15,16,17,18]	475, 525, 690, 779, 799, 805, 1035	all
Calcite [4,19,20]	713, 875, 1460	B2.P.11.2 *, B2.P.11.3-5, B2.P.11.6 *
Calcium Phosphate [21]	1037	B2.P.10.1 *
Amesite [22]	930	B2.P.11.1, B2.P.11.4
Muscovite [23]	754	B2.P.11.4
Hematite [24,25]	470, 537	B2.P.10.2-5

* confirmed by XRD analyses.

. Generally, the values are in accordance with the experimental results presented in the references.

The molecular vibrations determined by infrared radiation are affected by the bond strength between the atoms and the atoms’ mass. The absorption bands shift to lower wavenumbers as the atomic mass increases.

The most representative oxide compound found in all samples is quartz, identified by the vibrations of the Si–O–Si group. The significant spectral band at 1035 cm⁻¹ is due to asymmetric stretching vibrations of Si–O–Si. The bands at around 800 cm⁻¹ are due to symmetric stretching vibrations of Si–O–Si, while the band at 475 cm⁻¹ is for Si–O–Si out-of-plane deformations.

Calcite is present in the FTIR spectra of samples taken from the white decoration material (

Table 6. FTIR spectra of the decorative white material and FTIR spectrum of the black slip from sample B2.P.11.1.

FTIR Spectrum of the White Material and Black Slip
B2.P.10.1	B2.P.11.2
B2.P.11.1	B2.P.11.1 (black slip)
B2.P.11.3	B2.P.11.4
B2.P.11.5	B2.P.11.6

) due to the vibration of the C–O bond from the (CO₃)²⁻ group.

Similar to the XRD result, calcium phosphate was only found in the FTIR spectrum of sample B2.P.10.1. This was identified by the bending vibration of P–O bonds from the (PO₄)³⁻ group.

Amesite, identified by the vibration of the Si–H bond, and muscovite, identified by the vibration of Al,Si–OH bonds, have also been found in the white decorative materials.

In the FTIR spectrum of the material taken from the glossy black slip on sample B2.P.11.1, only the presence of quartz is identified.

The FTIR spectra of samples taken from small red-painted areas of the ceramic fragments (

Table 7. FTIR spectra of the red pigmented material.

FTIR Spectrum of the Red Material
B2.P.10.2.	B2.P.10.3.
B2.P.10.4	B2.P.10.5

) showed the presence of hematite (α-Fe₂O₃) due to vibration of the FeII–O bond.

4. Discussions

All of the samples analyzed contain quartz (SiO₂). In microscopic images (Table 1), it can be seen that quartz is predominantly present as sand with varying grain sizes, likely added for technological reasons. Also, quartz is an important compound of the rocks used to obtain the white paste. The lower quartz content of samples B2.P.11.2 (30.7%) and B2.P.11.6 (13.6%) compared to B2.P.10.1 (62%) suggest that this may serve as either a white material extender or an abrasive element during the mortaring of the white decorative material, particularly when calcined bones are used (sample B2.P.10.1).

Moreover, B2.P.10.1 is the only sample that contains calcium phosphate (Ca₃(PO₄)₂) together with calcium and potassium phosphate (KCaP₃O₉). The transformation of hydroxyapatite into calcium phosphate occurs around the temperature of 600 °C [26]. Since the white material in this sample does not have calcite (CaCO₃) or calcium oxide (CaO), calcium phosphate cannot be derived from calcareous rocks containing corals or shells. We can conclude that a sample of white decorative material B2. P.10.1 contains calcined bone powder.

Numerous studies have shown that ground bones were used in decorative white materials in prehistoric times, with varying frequency based on cultural and chronological factors. For example, in ceramic samples from Vădastra culture, calcium phosphate was reported [13]. It was also found in a ceramic sample from Gălățui–Movila Berzei (Boian–Giulești phase, contemporary with Vădastra II) [27], but more recent studies did not find any calcium phosphate [5]. Hydroxyapatite was identified in percentages of up to 8%, along with calcite and quartz, in samples from Sultana, Vidra, and Vlădiceasca sites during Boian Vidra phase [4]. This practice is considered rare and is also observed during the Spanțov phase. It has been demonstrated that ceramic samples from Radovanu were decorated with a paste made from white calcite and hydroxyapatite following the firing process [7]. The bone calcination technique was as follows: mainly sourced from sheep or cattle, the bones were burned in an open fire at temperatures over 700 °C until they became ash. After that, the rough ash blended with quartz sand was ground into a powder, to which other white materials were added depending on the desired decorative quantity and shades [4]. Some archaeological discoveries related to the Boian culture have revealed pits with intentional burn marks alongside ceramic fragments, animal bones [3], and sometimes tools [6]. The Chalcolithic period is when calcined bones began to be used in the preparation of white decorative material, suggesting a shift in the thinking of prehistoric communities in the Lower Danube region [15]. However, we cannot precisely determine the significance of bone calcination, i.e., whether it was for passage rituals, fertility, healing, or seasonal practices. It can be admitted that, whatever the purpose of the pits where the bones were incinerated, they could have given community identity and a sense of belonging to a group.

Except for sample B2.P.1.10, the white decorative material on all other ceramic fragments is calcite (CaCO₃), confirmed by XRD and/or FTIR analyses. This is found in large quantities in samples B2.P.11.2 and B2.P.11.6. It is highly probable that the calcite originates from the area near the site, as the right bank of the Danube is made up of calcareous formations. The limestone boulders were prob ground and blended with organic material. The resulting white paste was then applied to spaces with varying widths of the ceramic vessel. Due to the presence of calcite, and not calcium oxide, we can say that the white paste was applied after burning, as the calcite would have transformed into calcium oxide around 800 °C [28]. Analyses of ceramic samples from 13 sites in Bulgaria dating from the Neo-Eneolithic period revealed that calcite or calcium oxide was commonly used to produce white decoration paste. These were mixed with quartz or sediments rich in silicates or alumino-silicates, probably added for better adhesion on the walls of ceramic vessels. Other materials such as gypsum, dolomite, and talc have rarely been used and prove a good knowledge and use of local resources [29].

Ilmenite (FeTiO₃) is present in the white material from sample B2.P.10.1 at a percentage of 4.7%. It is a dark-colored titanium and iron oxide that can undergo thermal or chemical transformation into white rutile (TiO₂) [30]. This can be found in the metamorphic and volcanic rocks of the Dobrogea relief.

Amesite (Mg₂Al₂SiO₅(OH)₄) is contained in the white material of the following samples: B2.P.10.1 (XRD analyse), B2.P.11.1, and B2.P.11.4 (FTIR spectra). It is commonly found in metamorphic rocks, often occurring alongside calcite and diopside. Amesite, together with talc and chrysotile, was found in the white decorative material on ceramic fragments analyzed by XRD from the Djulyunitsa site in Bulgaria, dating 6100–5900 BC [31]. The role of amesite is to decrease porosity, which enhances waterproofing due to the presence of magnesium in the composition of this compound.

Rossiantonite (Al₃(PO₄)(SO₄)₂(OH)₂(H₂O)) was found in sample B2.P.10.1 at a concentration of 13.4%. This is a colorless hydrated aluminum hydroxy-phospho-sulfate that is found in fluvial sands, often in association with quartz. Through heat treatment, it undergoes transformations in the following stages: it loses the hydration water molecules up to 149 °C and in the range 975–1100 °C it eliminates SO₂, resulting in berlinite (AlPO₄) [32], as we find it in sample B2.P.11.6. In this case, the presence of rossiantonite in sample B2.P.10.1, which contains water molecules and hydroxyl groups, indicate that the white paste was derived from unfired raw material and applied after firing the vessels, somewhat clarifying the problem of the presence of calcium phosphate.

Resuming the above idea about berlinite (AlPO₄) found in sample B2.P.11.6 at a percentage of 6.9%, this aluminum phosphate is directly related to rossiantonite, which transforms into berlinite between 975 and 1100 °C after some thermal treatments, releasing H₂O and SO₂ [32]. Berlinite is transparent, with possible variations in color such as pale gray or pink colors.

Potassium, calcium, and magnesium phosphate (KCa₉Mg(PO₄)₇) is present in a percentage of 14% in white material from sample B2.P.11.6. This is a white compound like any alkaline and alkaline-earth metal phosphate. So far, no studies have been published that indicate it in the composition of ceramics or white materials used for decoration.

Sodium oxide (Na₂O) is found in 4.8% of the white material from sample B2.P.11.6. It is a white powder typically found in combination with other minerals. At a temperature above 400 °C, it turns into sodium peroxide (Na₂O₂). This compound plays a significant role in the process of vitrification [33].

In samples containing red-pigmented material, FTIR analysis confirms it as hematite. Digital microscopy reveals that in two cases, the pigment is found on the inner lip of a fine ceramic vessel (fragment B2.P.10.3, characteristic decorative detail of Boian ceramics [4,11]) and a semi-fine ceramic vessel (B2.P.10.5). Additionally, the ceramic fragment B2.P.10.4 shows the pigment on the inner surface, likely used for storing or preparing the red color. In the case of B2.P.10.2, the red pigment was used to decorate the outer thickened lip of a bowl.

In the case of the material with a reddish granular appearance on sample B2.P.11.3 (image a in Table 1), analyzed to determine if it is a red pigment or a patina, FTIR analysis did not provide conclusive results.

Sample B2.P.11.1 displays areas of black film, likely serving a decorative purpose. FTIR analysis confirmed that it is mainly composed of quartz. We consider that a more detailed examination of the composition and structure of this slip through XRD and WDXRF is needed. This will be coordinated with future studies on the ceramic mass (clay) of the samples.

5. Conclusions

The archeometric analysis of these ceramic samples from Hârșova-Tell provides answers to the issue of the diversity of ceramic decoration techniques used around 4900–4600 BC. These techniques are becoming better understood through recent investigations [4,5,6,8,13,15,27,29,34,35,36]. These partially answer the archaeological questions raised.

The correlation of the analyses’ results on our samples from Hârșova-Tell with those already published, carried out on samples from different sites in the Neo-Eneolithic period (Boian–Vidra and Spanţov phases) located in the space between the Southern Carpathians and the Danube and in the south of the Danube, demonstrates that the same community utilized various natural raw materials and techniques to obtain the white material for decoration.

The extensive use of calcite combined with various oxidic compounds like quartz, silicates, or alumino-silicates, possibly added to improve the adhesion of decorative materials on ceramic vessel walls, is significant. From time to time, the technique of calcination of bones was also used to obtain white paste rich in calcium phosphate.

This study indicates the significant presence of sand quartz in various percentages and granulations, as expected. It plays the role of filler in the white material but also as an abrasive element during mortaring.

The other identified compounds, such as rossiantonite, berlinite, ilmenite, and amesite, came from magma and metamorphic rocks of Dobrogea. These range in color from white to light gray.

The red-painted areas from our samples were only analyzed using FTIR spectroscopy, and the spectra indicated that the pigment utilized was hematite according to specific cultural practices in that period.

The XRD and FTIR analysis results reveal that the crystalline structures of the compounds indicate that both white and red paint decorations were added to the ceramic vessels after firing.