Introduction

The Han Dynasty (206 BC–220 AD) was a period of transformative change in China, profoundly impacting the funeral system among other aspects. The prevalence of Mingqi—crafted items specifically created for burial with the deceased—signified this cultural shift. Pottery towers, a common form of Mingqi, have been unearthed in numerous locations across China. However, the discovery of twenty polychrome pottery towers at the Daqu burial site in 2013 marked a unique occurrence, representing the first instance of finding such a large collection of polychrome towers outside the traditional center of the Han Dynasty, namely Shaanxi and Henan provinces [1, 2].

Situated 3 km southeast of Anding Town in the Daxing District of Beijing, as shown in Fig. 1, the Daqu burial site has unearthed a total of 32 tomb. Archaeological clues, including tomb architecture (brick tomb, Fig. 1) and burial products, dates 31 of these tombs to the Later Eastern Han period [2]. This is supported by thermoluminescence dating of pottery shards, suggesting a firing date of approximately B.P. 1875 ± 185 years. The excavated pottery towers, ranging from 33 to 72 cm in height, could be assembled into structures comprising two or three individual towers (Fig. 2). The pottery surface features painted beams, purlins, bracket systems, and animal heads, accompanied by abstract polychrome wall paintings utilizing black, white, and red pigments. These artifacts offer valuable insights into ancient construction methods and decorative arts, reflecting the Han Dynasty's beliefs in immortality.

Fig. 1
figure 1

Geological map of Beijing municipality, position of Daqu Burial site, and photographs of tombs

Full size image
Fig. 2
figure 2

Examples of pottery towers and their archaeological numbers

Full size image

Previous studies have explored typology and stylistic aspects of Han Dynasty pottery towers. The crafting of polychrome pottery towers with substantial volume and delicate aesthetics requires a high level of professional expertise. Consequently, the proprietors of these pottery towers have been identified as members of the middle and petty bourgeoisie [3,4,5,6]. In central China during the Han Dynasty, the building of exquisite pottery was a marker distinguishing the wealthy from ordinary people: “In the capital of the Han Dynasty, Chang'an city, a pottery cooker was priced at two hundred coins. Acquiring a complete set of pottery Mingqi required a minimum of one thousand coins. This amount roughly equates to around 10 dan (石) [a unit of weight in ancient China, with a possible range of 13–30 kg] of rice, a cost beyond the means of ordinary people.” [7].

Scientific analysis can provide a wealth of archaeological information to enhance the understanding to ancient society. Methods such as optical microscopy (OM), X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FT-IR), and X-ray fluorescence (XRF) can unveil crucial information about raw materials, processing methods, firing atmosphere, firing temperature, and burial context. A comprehensive analysis can open avenues for discussing the origin, techniques, allocation, circulation, usage, mending, and abandonment of pottery, enabling a deeper exploration of technology levels, customs, organization, and inter-regional exchanges in ancient society [8,9,10,11,12]. However, scientific analysis of Han Dynasty pottery is currently limited [13, 14].

In this study, we conducted a detailed mineralogical and geochemical analysis of pottery samples from the Daqu burial site. The results provide insight into the materials and techniques employed in crafting pottery towers in the Beijing region during the late Eastern Han Dynasty, which will contribute to our understanding of ancient society.

Materials and methods

Sample selection

A variety of ceramics were excavated at the Daqu burial site, including bowls, cups, towers, bricks, and fragments. These potteries exhibit distinct textures. Notably, pottery towers exhibited a fine texture and relatively high strength, while ellipse-type dishes and certain pots displayed a rougher texture, with some parts showing surface disintegration. Five unrestorable shards, whose original positions could not be determined, were selected for analysis. They were from 4 different tombs. Based on their shape and the position at which they were unearthed, Samples 1 to 3 were traced to polychrome pottery towers, the origin of Sample 4 remained uncertain, and Sample 5 was identified as part of a coarse ellipse-shaped dish. Photographs and sample information are detailed in Fig. 3 and Table 1. These samples had previously undergone analysis using differential scanning calorimetry-thermal gravity analysis (DSC-TG), Fourier transform infrared spectroscopy, and dilatometry (DIL) [15].

Fig. 3
figure 3

Photographs of pottery samples chosen for analysis from the Daqu tomb and polished cross-section samples

Full size image
Table 1 Sample description
Full size table

Analytical procedures

Following ultrasonic washing and drying of the painted pottery fragments, the samples underwent slicing (approximately 30 µm thick) and double polishing for microscopic observation. Photographs of polished pottery samples are shown in Fig. 3. Additionally, 1.0 g of each sample was pulverized for subsequent XRD and XRF analyses.

L*a*b* values of the pottery cross sections were measured using a 3nh NH310 portable colorimeter (Guangdong Sanenshi Intelligent Technology Co., LTD, China).

A German Axioskop 40 was utilized for polarized and reflected light observations and photography of the pottery slices. Petrographic characteristics, mineral composition, and structural configuration of each sample were documented, and the contents of mineral components were estimated. A Keyence VHX-6000 microscope (Japan) facilitated the observation of the section structure of the pottery samples at magnifications ranging from 50× to 2000×.

Qualitative and semi-quantitative analysis of the composition of the pottery samples was conducted using a D8 Advance X-ray diffractometer (Germany) equipped with a monochromized Cu radiation source. The 2θ scanning angle ranged from 3° to 70°.

For infrared analysis, the potassium bromide pellet method was employed, and Fourier transform infrared spectroscopy (FT-IR) was performed using a Nicolet Is 5 spectrometer (Thermo Fisher, United States). Spectra were recorded at a resolution of 2 cm−1, with 64 scans added and averaged before Fourier transform. Spectra covering 400 cm−1 to 4000 cm−1 range were previously published. [15] This study focused on absorption peaks occurring at a wave number lower than 1600 cm−1.

An Idax EDAX Orbis X-ray Fluorescence Spectrometer (Ametek, the United States) was employed for qualitative and quantitative analysis of the chemical composition of the pottery samples, utilizing an Rh target X-ray tube operated at 40 kV and 500 μA. MBH 32XN7A was used as standard reference sample. Major oxides (SiO2, Al2O3, K2O, CaO, MgO, Fe2O3, TiO2, MnO, and P2O5) were measured.

Results

Optical microscopy

Upon examination, all pottery samples demonstrated petrographic characteristics consistent with those of typical sedimentary loess found in northern China, predominated by quartz with less frequent feldspar and minimal inclusions (Fig. 4, Additional file 1: Fig. S1; Table 2). Samples 1–4 displayed no optical activity, while Sample 5 exhibited low optical activity.

Fig. 4
figure 4

Photomicrograph of a thin section from pottery samples. Photographs in the first and second columns of each sample are captured under the same field of view

Full size image
Table 2 Main compounds and optical activity of samples
Full size table

The mineral compositions of Samples 1–3 were identical, while Sample 4 closely resembled them but lacked the presence of biotite. Calcite in the four samples was likely introduced as inclusions or pigments after firing, as it appeared on or near the pottery surface. Sample 5 shared the main mineral components with Samples 1–3, with additional minor minerals such as chlorite.

All samples showed oblong pores, potentially generated after burning straw or other plants added as temper in the clay paste (Additional file 1: Fig. S2). Clay and silt content were lower in Samples 1–4 and higher in Sample 5, suggesting raw materials may have been derived from alluvial soils or subjected to repeated panning [16].

The color of clay paste, indicative of firing atmosphere, ranged from yellowish brown to brownish black, suggesting a valence of Fe at + 3 and thus oxidation during pottery firing [17]. The proportion of hematite to magnetite in the pottery provided additional evidence of oxidation conditions during the firing process, with the hematite content surpassing that of magnetite [16].

Samples exhibited a sandwich structure with a darker core and lighter surface, indicating incomplete oxidation inside the body [18]. Organic materials were not fully burned out, leaving carbon residue in the middle of the pottery. Sample 4, lighter in color, suggested nearly complete combustion of organic matter.

X-ray diffraction

XRD analysis identified the primary mineral phases as quartz, plagioclase, microcline, and muscovite across all samples, as detailed in Table 3 and Fig. 5. Sample 5 exclusively exhibited the presence of kaolinite, montmorillonite, and calcite. Quantitative results were derived from peak areas. However, it is important to note that these results may be influenced by factors such as microstructure and mass absorption, thus requiring careful interpretation [19]. Nonetheless, they provide a valuable reference point.

Table 3 Mineralogical composition of the shards according to XRD analysis
Full size table
Fig. 5
figure 5

XRD patterns of pottery samples

Full size image

Infrared spectroscopy

FT-IR spectra, previously published in Chinese [15], were further analyzed using second derivative (Fig. 6). Peak assignments were made based on relevant literature [20] and the RRUFF database, and mineral abundance was estimated based on the ratio of peak intensity to the intensity of the quartz absorption peak (1080 cm−1). Results are presented in Table 4. All shards were similar and could be divided into two groups. As shown in Table 4, quartz, feldspar, and muscovite were present in all infrared spectra. Sample 5, however, featured additional minerals such as chlorite, calcite, and kaolinite.

Fig. 6
figure 6

FT-IR patterns of pottery samples [15] and their second derivative

Full size image
Table 4 Mineralogical composition of the shards according to the infrared analysis
Full size table

X-ray fluorescence

Elemental analysis exhibited consistent distribution of major elements, classifying the clay as general fusible clay [5] based on its composition—SiO2 (65.80–67.09 wt%), Al2O3 (14.44–16.12 wt%), and total flux content (17.21–20.50 wt%)—with low elemental Ca content (less than 6%) suggesting the use of non-calcareous clay [21] (Fig. 7; Additional file 1: Table S1).

Fig. 7
figure 7

XRF patterns of pottery samples

Full size image

The main differences in the distribution of elemental content were observed in the K content of Sample 3 and the elemental Mg content of Sample 5. The high content of P and K in Sample 3 could be attributed to plant inclusions at the sampling location, as biomass materials are rich in these elements. The overall elemental compositions in all five samples supported the speculation of consistent soil sources.

Discussion

Mineralogical results

The combination of three analytical methods—optical microscopy in polarized light, XRD, and FT-IR—yielded comprehensive results, as these methods complemented each other. XRD effectively identified crystalline minerals, while FT-IR was effective for analyzing both crystalline and non-crystalline materials [20].

All samples exhibited consistency in their main components, including quartz, plagioclase, microcline, muscovite, hematite, and magnetite. Further categorization based on minor components revealed that Sample 4 shared similarities with Samples 1, 2, and 3, with the absence of biotite being the distinguishing factor. In contrast, Group 2 comprised only Sample 5, characterized by the presence of chlorite, montmorillonite, calcite, and kaolinite. This grouping aligns with the classification based on pottery texture.

Firing techniques

Insights into firing techniques were gleaned from the formation and disappearance of minerals at specific temperature ranges. Phase transitions, indicative of firing temperature, were evident in the mineral composition. For example, feldspars typically melt around 1100 °C [22, 23], while orthoclase decomposes at 1050 °C [20]. The decomposition temperature of calcite ranges from 650 to 750 °C, with complete decomposition occurring at 800–850 °C [24,25,26]. Chlorite exists below 700 °C [27], and the decomposition of kaolinite and the formation of metakaolinite occur in the temperature range of 500–650 °C [28,29,30].

The observed variation in minor components among samples was attributed to differences in firing temperature. Samples 1–4 exhibited firing temperatures between 800 °C and 1100 °C, while Sample 5's firing temperature was below 650 °C, a conclusion corroborated by polarizing microscope analysis. Sample 5 displayed low optical activity, while samples 1–4 exhibited no optical activity in the matrix, and quartz exhibited less angular structure, indicating firing temperatures exceeding 900 °C [31]. The differences in minor components, including the absence of chlorite, kaolinite, montmorillonite, and calcite, were fully explained by variations in firing temperature.

Origin of polychrome pottery towers

Despite minor differences in minor mineral components and elemental contents, likely due to variations in firing temperature and uneven plant inclusions, respectively, all samples exhibited similarities in bulk composition, element content, plant inclusions, and firing atmosphere. The negligible differences in raw materials among the five samples suggest a consistent origin for pottery in the Daqu burial site.

The coarse potteries (Sample 5), characterized by low value, vulnerability, weight, and ease of production, were improbable products of inter-regional trade [32]. Consequently, the fine pottery towers (Samples 1–4) were likely crafted locally. This inference aligns with the notion proposed by Tite et al. [33] that large and heavy pottery types frequently found at a site are typically produced nearby. Notably, three pottery building samples (Sample 1–3) from different tombs and periods displayed nearly identical raw materials and techniques. Given the similarities in architectural structures and consistency in decorative details [15], it is reasonable to suggest that the polychrome towers were crafted following the same tradition [34], perhaps originated from the same workshop. Their high consistence in quality demonstrates the advanced techniques of ancient workers.

These findings contribute to the understanding of the late Han Dynasty. Variations in the quality and volume of potteries in the Daqu burial site may be linked to their different roles in the funeral system. Local producers likely designed and offered various Mingqi potteries for distinct purposes, emphasizing the importance of comparing the usage and quality of different potteries in other districts to explore customs during the late Eastern Han Dynasty and the interconnectedness between the central and northeastern regions of ancient China.

Conclusions

This study investigated materials and techniques employed for pottery in Han Dynasty at the Daqu burial site, marking the first use of scientific technologies on ancient Beijing potteries. Mineralogical and geochemical analyses revealed similarities in main components, elemental composition, plant inclusions, and firing atmosphere across all samples. Minor differences, particularly in minor minerals, were attributed to variations in firing temperature.

Archaeological information of pottery materials and firing techniques were obtained through OM, XRD, FT-IR, and XRF analyses. The results indicate that all pottery in the Daqu burial site utilized similar raw materials and techniques, and delicate polychrome pottery towers from different tombs likely originated from the same local workshop. Further studies on the economic and social conditions in the Beijing region during the Han Dynasty can build upon these results.