Introduction

Lacquer culture originated in China1, and later spread to East Asia and the rest of the world2,3. Although extensive literature has documented the techniques and processes involved in lacquerware production4,5,6, there are no records of coiled wood core lacquerware. In recent years, coiled wood core lacquerware has gradually attracted widespread attention, with research mainly focused on lacquerware unearthed from southern China7,8,9,10. The coiled wood core lacquerware unearthed in Xuzhou has, however, become the only example discovered so far in northern China.

From December 2020 to April 2021, an archaeological excavation was conducted in coordination with the Xuzhou Wenmiao Block construction project, covering an area of 900 square meters. The excavation unearthed over 1600 ceramic and metal artifacts, including a lacquered wooden object found in the Ming dynasty stratum of the northern partition beam in area TN1E111, which confirms its burial during or after that period but does not allow for a precise determination of its manufacturing date. When it was unearthed, the entire lacquerware was inverted on the ground and severely damaged, with only the base, parts of the walls, and dried lacquer film remaining (Fig. 1a). The base is circular, with a diameter of 132.8 mm and a thickness of ~1.7 mm, made from long, narrow wooden strips. It shows clear marks of ringed feet, marks of the a fabric layer’s previous existence, and slight deformation (Fig. 1b). The surrounding area is decayed, with three cracks visible. Only eight wooden strips remain on the walls, none intact. Both the inner and outer surfaces of the lacquerware are coated with lacquer, with the inner side being red and the outer side black. There is an unclear seal inscription (Fig. 2).

Fig. 1: Current condition of the lacquerware sample upon excavation.
figure 1

a Side view. b Top view.

Full size image
Fig. 2
figure 2

Red seal inscription on the lacquerware base.

Full size image

This study uses modern analytical techniques to investigate the materials and manufacturing processes of this lacquerware, aiming to provide a scientific basis for the restoration and conservation of ancient lacquerware, while also offering important physical data for the study of China’s lacquerware manufacturing history.

Methods

Samples

This study selects samples from the core base, strip-shaped core used to form the wall of the lacquerware, lacquer ash, and lacquer film for analysis. For convenience in documentation, the sample codes consist of an English part and a numerical part (Table 1).

Table 1 Sample names and codes
Full size table

Digital X-ray imaging analysis

The base of the lacquerware was supported by an acrylic plate, and digital radiographic images were acquired using a digital X-ray imaging system (Selenia™, Hologic Inc., Danbury, Connecticut, USA). The imaging was performed under the following conditions: tube voltage of 39 kVp, tube current-time product of 11.6 mAs, compression thickness of 13.6 cm, with the X-ray tube positioned at 0°, and no compression paddle used. A silver (Ag) filter was applied during imaging.

Optical microscopy (OM) analysis

For microscopic analysis, the sample C02 was first dehydrated in a graded series of polyethylene glycol aqueous solutions (PEG 1500, analytical grade), ranging from 10% to 90% (w/w), with each concentration applied for 24 h. After dehydration, the specimen was embedded in molten PEG 1500. Thin sections were prepared using a fully automated rotary microtome (RM2255, Leica Microsystems, Wetzlar, Germany) along the cross, radial, and tangential planes. Section thickness was 15 μm for cross sections and 10 μm for radial and tangential sections. The prepared slides were examined using a biological microscope (BX51, Olympus Corp., Tokyo, Japan). Microscopic images were captured for documentation. Wood anatomical features were compared with references including Timbers of China, Wood Identification Atlas, and the InsideWood database for species identification.

Extended depth-of-field microscopy analysis

The cleaned core and lacquer film samples were observed using an extended depth-of-field digital microscope (VHX-6000, KEYENCE, Osaka, Japan) at magnifications of 50×, 100×, 200×, and 300×, with illumination provided by a ring-shaped LED light source. For cross-sectional observation of the stratigraphy of the lacquer film, the samples were embedded in a fast-curing epoxy resin system (Buehler, Lake Bluff, Illinois, USA) prior to imaging.

Scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDS) analysis

The microstructure of the lacquer foundation layer was examined using a field-emission scanning electron microscope (Apreo S, Thermo Fisher Scientific, USA). The sample was sputter-coated with gold and observed at an accelerating voltage of 8 kV, using secondary electron detection, a working distance of 10 mm, and a horizontal field width of 104 μm. Elemental composition analysis was performed using an energy-dispersive X-ray spectrometer (X-Max 20, Oxford Instruments, Oxford, UK), equipped with a SuperATW window, a 20 mm² active area, and an energy resolution better than 127 eV, enabling detection of elements from Be (Z = 4) to U (Z = 92). Comparative analysis was conducted using modern brick powder and tile powder samples. Particle size measurements of the filler materials were conducted using Nano Measurer software (Version 1.2), and the statistical distribution of particle sizes was plotted using OriginPro 2023 (Version 10.0, OriginLab Corporation, Northampton, Massachusetts, USA).

X-ray diffraction (XRD) analysis

X-ray diffraction (XRD) analysis was performed to determine the mineralogical composition of the lacquer foundation layer using a Smart-LAB diffractometer (Rigaku, Tokyo, Japan). The specific testing conditions were as follows: tube current of 150 mA, tube voltage of 40 kV, scanning range from 5° to 90° (2θ), scanning speed of 10°/min, and step size of 0.01°. The instrument operated at a maximum power of 9 kW, utilizing a copper (Cu) target and a standard Z sample stage for measurements. Data processing was carried out using OriginPro 2023 (Version 10.0, OriginLab Corporation, Northampton, Massachusetts, USA) to analyze and interpret the spectral results.

Fourier transform infrared spectroscopy (FT-IR) analysis

Fourier transform infrared (FT-IR) spectroscopy was conducted using an FT-IR spectrometer (LUMOS, Bruker, Bremen, Germany). The potassium bromide pellet method (KBr, AR, Sinopharm Chemical, China) was employed. Both the extracted fabric and lacquer film samples were analyzed. The scanning range was 4000–600 cm⁻¹, with a resolution of 4 cm⁻¹ and 24 scans per sample. Modern reference samples of cotton, hemp, wool, and silk were also tested for comparison. All spectra were processed using the smoothing function in OriginPro 2023 (Version 10.0, OriginLab Corporation, Northampton, Massachusetts, USA) with a span of 60 points.

Laser micro-confocal Raman spectroscopy (RS) analysis

Raman analysis was conducted at room temperature under darkroom conditions using a confocal Raman spectrometer (inVia, Renishaw, UK) with a 785 nm excitation laser. The test sites were examined under magnifications of 50× using the microscope. The laser power was set to 0.05% of the maximum output (300 mW), and the actual laser power reaching the sample surface was ~45 μW. The tested samples included lacquer film from the lacquerware (RF01 and BF01), as well as black lacquer film added carbon (archaeological sample). Data processing was carried out using OriginPro 2023 (Version 10.0, OriginLab Corporation, Northampton, Massachusetts, USA) to analyze and interpret the spectral results.

Pyrolysis-gas chromatography-mass spectrometry (Py-GC/MS) analysis

Pyrolysis-gas chromatography-mass spectrometry (Py-GC/MS) analysis was performed using a multifunctional pyrolyzer (EGA/PY-3030D, Frontier Lab, Japan) coupled with a GC/MS system (7890B-7000C, Agilent Technologies, USA). Separation of analytes was achieved on an HP-5 capillary column (5% diphenyl/95% dimethylpolysiloxane, 30 m × 0.25 mm × 0.25 μm). Less than 1 mg of sample was mixed with 5 μL of a 10% tetramethylammonium hydroxide (TMAH) methanol solution (analytical grade, Aladdin Reagent Co., Shanghai, China) in a stainless-steel sample boat, and introduced into the pyrolysis chamber. Online methylation derivatization was performed simultaneously with pyrolysis at 500 °C for 12 s. The interface temperature between the pyrolyzer and the GC inlet was set to 300 °C. The GC injector temperature was maintained at 300°C with a split ratio of 50:1. Helium was used as the carrier gas at a constant flow rate of 1.0 mL/min.

The GC oven temperature was initially held at 50 °C for 2 min, then ramped at 4 °C/min to 300 °C, and held for 5 min. The mass spectrometer operated in electron ionization (EI) mode at 70 eV, with an ion source temperature of 230°C and a quadrupole temperature of 150 °C. Full-scan acquisition was performed over a mass range of 10–600 m/z. Pyrolysis products were identified by interpreting mass spectra and comparing them with the NIST library, and further supported by the ESCAPE (Expert System for Characterization using AMDISPlus Excel) method introduced by the Getty Conservation Institute (GCI) in the Recent Advances in Characterizing Asian Lacquer (RAdICAL) workshop.

Results

Wooden core

Based on digital X-ray radiography, it is revealed that the base of the lacquerware is constructed from wooden slats of varying widths, and with a small point in the center of the base plate that does not pass through the base12. This point may have been left by a drawing compass used to cut out the circular wooden panel13. The outermost layer of the base is wrapped with fine wooden strips to secure the entire structure (Fig. 3).

Fig. 3
figure 3

Digital X-ray radiograph of the lacquerware base. The image shows a small point at the center of the base plate, highlighted by a red square.

Full size image

Each strip-shaped core used to form the wall of the lacquerware was individually observed under the extended depth-of-field microscope (Fig. 4). The edges of the strip-shaped core samples are straight, with no signs of bonding, suggesting that each strip underwent processing such as cutting and shaving. The cross-section of the core samples is rectangular, with varying sizes between the different pieces, ranging from 0.724 cm × 2.170 cm to 1.432 cm × 2.474 cm.

Fig. 4
figure 4

Cross-sectional view of sample C01 under extended depth-of-field microscopy.

Full size image

Each strip-shaped core exhibits varying shades of color on the same lateral surface (Fig. 5), possibly caused by extended overlap with adjacent cores underground. The overlapping areas likely retained more of the wood’s original color, while the regions in direct contact with the burial soil darkened due to prolonged environmental exposure. This observation supports the inference that the wooden strips were originally overlapped during the construction of the lacquerware core.

Fig. 5
figure 5

Side view of sample C01 under extended depth-of-field microscopy. scale bar: as indicated; ruler tick: marks denote distance (mm); color contrast: marks the boundary between lighter overlapping zones.

Full size image

Microscopic observation of wood anatomical structures (Fig. 6), in combination with comparisons to standard wood identification charts and relevant literature13, it can be determined that the core is Cunninghamia lanceolata (cedar) from the Cupressaceae family14.

Fig. 6: Microstructural three-cut views of core sample.
figure 6

a Cross-sectional view. b Radial view. c Tangential view.

Full size image

Foundation Layer

Based on SEM observations, the lacquer ash particles were found to exhibit irregular morphologies, yet discernible projections allowed for further particle size analysis (Fig. 7). Therefore, the Feret diameter was adopted to estimate particle sizes from the projected outlines15. As shown in Fig. 8, the particle sizes vary widely, with a maximum of 67.97 μm, a minimum of 2.28 μm, and an average of 14.04 μm, indicating a highly uneven distribution. This suggests that raw lacquer might have been used as a binding agent during preparation, leading to cohesion among the particles after curing16.

Fig. 7: SEM image of lacquer ash (the red square indicates the EDS detection area).
figure 7

a 200×. b 500×.

Full size image
Fig. 8: Particle size distribution.
figure 8

a Particle size distribution of modern tile powder. b Particle size distribution of modern lacquer ash. c Particle size distribution of archaeological lacquer ash samples).

Full size image

According to Xiushilu (髹饰录), the Wanqi (垸漆)technique involves the application of lacquer ash, prioritizing horn and porcelain powders, followed by bone ash and clam powder, with brick and tile powders as alternate choices17. The results of EDS (Table 2) showed that the foundation layer mainly contains O, Si, Ca, Al, K and Mg, which are common elements from clay minerals and calcite18. The analysis results from XRD (Fig. 9) confirm that the main components of the lacquer ash from this coiled wood core lacquerware are SiO2 and CaCO3, indicating that brick powder or tile powder were predominantly used, with no evidence of animal bone ash19,20 or other additives21. These findings are also consistent with the location where the lacquerware was unearthed. The coiled wood core lacquerware was discovered at the underground city site in the Wenmiao Street area of Xuzhou, which was a civilian artifact, not part of a high-status tomb’s burial goods.

Fig. 9
figure 9

XRD pattern of A01.The diffractogram shows characteristic reflections assigned to silica (SiO₂) and calcium carbonate (CaCO₃).orange: SiO₂; blue:CaCO₃; x-axis: 2θ (°); y-axis: intensity (a.u.).

Full size image
Table 2 EDS results of lacquer ash (Wt%)
Full size table

On the two opposite sides of a single strip-shaped core used to form the wall of the lacquerware (i.e., the sides extending from the shorter edges of the cross-section), lacquer ash was adhered, with noticeable mark visible on the lacquer ash. The width of the impressions was measured to be between 0.293 mm and 0.322 mm (Fig. 10), which directly corresponds to the fabric extracted from the base of the lacquerware. The extracted spinning yarn, composed of a single thread made of multiple fibers twisted in a “Z-twist”22 fashion and measuring approximately 0.35 mm in diameter, was observed under an extended depth-of-field microscope. (Fig. 11). This indicates that the marks on the lacquer ash were caused by fiber layer, thus confirming that the “Biaobu (裱布)”23.

Fig. 10
figure 10

Image of the side of a single strip-shaped core with marks caused by the fabric layer under extended depth-of-field microscopy. scale bar: as indicated; ruler tick: marks denote distance (μm).

Full size image
Fig. 11: Image of ancient spining yarn extracted from the lacquerware base under extended depth-of-field microscopy.
figure 11

a 50×. b 200×.

Full size image

The organization structure of the fiber is plain weave, In traditional lacquerware manufacturing, fabric was often applied over the wooden core to enhance structural stability, with silk, cotton, and ramie being the most common materials24. However, due to the limited quantity and poor preservation of the fabric sample, precise fiber identification could not be achieved(Fig. 12). The diameters of the yarn are around 0.200–0.300 mm, which is consistent with ramie25 The density of the fabric was measured, and it was found to be ~5-6 threads per centimeter (Fig. 13). Considering the low thread count and the direction of the yarn, it is inferred that the fabric used to enhance structural stability was likely made of coarse bast fiber fabric. Bast fibers (e.g., hemp, ramie,)is native to most parts of China26, and was officially recorded in lacquerware production guidelines during the Qianlong period as a suitable material for cloth layers27.

Fig. 12
figure 12

FTIR spectra comparison: extracted fabric sample versus modern reference standards for cotton, hemp, wool, silk.

Full size image
Fig. 13
figure 13

Image of the side of a base core with marks caused by the fabric layer under extended depth-of-field microscopy.white line segment: 1 cm (scale); red dots: stitches (needle holes).

Full size image

Lacquer film

In this lacquerware, red lacquer was applied to the interior and black lacquer to the exterior. Based on extended depth-of-field and SEM imaging of the red and black lacquer film samples (Fig. 14), the red lacquer film comprises two distinct layers: a dark brown primer layer (70.82 μm) and a red pigmented lacquer layer (55.96 μm). A smooth and straight interface between the two indicates that polishing was performed.

Fig. 14: Cross-sectional microscopic images of lacquer film samples.
figure 14

a OM image of RF01 sample, b SEM image of RF01 sample, c OM image of BF01 sample, d SEM image of BF01 sample).

Full size image

The FT-IR results (Fig. 15) show that both the red and black lacquer film samples exhibit a series of absorption peaks corresponding to raw lacquer at wavenumbers of 3400 cm⁻¹, 2920 cm⁻¹, 2620 cm⁻¹, and 1453 cm⁻¹. Among them, the absorption peak around 3400 cm⁻¹ corresponds to the hydroxyl group on the phenolic ring of the lacquer phenol; the range 2800–2900 cm⁻¹ shows the asymmetric and symmetric stretching vibration peaks of methyl and methylene groups; 1500–1620 cm⁻¹ corresponds to the aromatic ring skeleton vibration; and 1453 cm⁻¹ corresponds to the bending vibration peak of the methylene group. Absorption peaks around 778–779 cm⁻¹ are caused by in-plane bending vibrations of 1,2,3-trisubstituted groups in lacquer phenol, indicating that the lacquer phenol mainly consists of a 1,2,3-trisubstituted structure. The main infrared absorption peaks of the lacquer film samples match well with the infrared characteristic absorption peaks of lacquer phenol (one of the main components of raw lacquer). Therefore, it can be concluded that the film-forming material for all the lacquer films is raw lacquer28.

Fig. 15
figure 15

Infrared spectrum of lacquer film sample. black curve: black lacquer film; red curve: red lacquer film; x-axis: wavenumber (cm⁻¹); y-axis : intensity (a.u.).

Full size image

The Raman analysis results show that the red lacquer film sample has clear and concise characteristic peaks (Fig. 16a), with a peak at 254 cm⁻¹ corresponding to the cinnabar characteristic peak29, indicating that the colorant used in the red lacquer layer is cinnabar. The Raman spectrum of the black lacquer film sample (Fig. 16b) does not exhibit the distinct characteristic absorption peaks of carbon black commonly seen in ancient black lacquer films30, suggesting that no chromogenic substances were likely added. The black coloration is presumed to originate primarily from the raw lacquer itself.

Fig. 16: Raman spectrum of lacquer film sample.
figure 16

a RF01 sample. b BF01 sample and black lacquer samples from other excavated site.

Full size image

Lacquer identification was achieved by Py-GC/MS. The discrimination between the three types is based in the detection of marker compounds for each lacquer31: 1.2-dimethoxy-3-pentadecenylbenzene for Toxicodendron vernicifluum; 1,2-dimethoxy-3-heptadecenylbenzene and1,2-dimethoxy-3-heptadecenylbenzene for Toxicodendron succedaneum; 1,2-dimethoxy-3-(10-phenyldecyl)benzene and 1,2-dimethoxy-3-(12-phenyldodecyl)benzene for Melanorrhoea usitate32. The compound class distribution of the lacquer film from the coiled wood core sample is visualized using the ESCAPE protocol, as shown in Fig. 17. The Py-GC/MS analysis identified a range of characteristic degradation products of raw lacquer, including substituted phenolic derivatives and their oxidation products, saturated and unsaturated hydrocarbons, alkylbenzenes, simple phenols, and aliphatic carboxylic acids. These pyrolysates are consistent with the thermal decomposition profile of urushiol-type lacquers. The analysis revealed the simultaneous presence of both 3-pentadecylcatechol and 3-heptadecylcatechol, with the former occurring at a significantly higher concentration than the latter. Notably, the lacquer used in this lacquerware contains a mixture of urushiol and laccol, suggesting that urushi and laccol raw lacquer may have been mixed. While T. succedaneum has traditionally been associated with lacquers from Vietnam and Taiwan, this species has also been recorded in southern China33, particularly in Guangxi Province where Donglan lacquer trees have been identified as Rhus succedanea34. This species has also been identified in several Chinese export lacquerwares35, supporting its historical use in southern China. Therefore, it is likely that the lacquer used in this lacquerware originated from within China.

Fig. 17
figure 17

Pyrolysis products of the coiled wood core lacquer film.

Full size image

A significant amount of glycerol, methyl monocarboxylate and methyl dicarboxylate was detected in the lacquer film (Fig. 18), suggesting that vegetable oil may have been incorporated into the lacquer solution36. Previous studies have shown that different drying oils exhibit specific P/S (palmitic/stearic) and A/P (azelaic/palmitic) ratios37, which can be used to identify the type of vegetable oil present. The ratio of azelaic acid to palmitic acid (A/P) was 0.29, confirming that drying oils were added to the coiled wood core lacquer film. The ratio of palmitic acid to stearic acid (P/S) was 1.06, which is consistent with tung oil (1–1.2), suggesting that tung oil was the drying oil added to the lacquer film38.

Fig. 18
figure 18

Distribution of fatty acids in the coiled wood core lacquer film.

Full size image

Alkyl phenyl alkenoates (APAs), a well-established marker of heated tung oil, were notably absent in our sample. This absence can likely be attributed to the processing method employed in the preparation of the tung oil. Existing literature indicates that APAs typically form when drying oils are subjected to elevated temperatures during processing39. Consequently, the absence of APAs in our sample suggests that the oil was not exposed to high-temperature treatment, and raw tung oil may have been utilized instead.

Discussion

Based on the results obtained from a series of scientific analyses, the regular arrangement of the wooden strip samples suggests a unique core-forming craft, providing insight into the manufacturing process of the lacquerware in Xuzhou. The lacquerware unearthed from Xuzhou is a coiled wood core lacquerware, which can be divided from the outer to the inner layers as follows: black lacquer film layer, foundation layer (composed of lacquer ash and fabric), wooden core, foundation layer (lacquer ash and fabric), and red lacquer film layer.

Figure 19 is a schematic cross-sectional diagram of the restored wooden strip of the lacquerware. The base of the lacquerware(Fig. 19a) is constructed from wooden slats of varying widths, with the outermost layer wrapped in fine wooden strips to reinforce and secure the structure. Each strip-shaped core used to form the wall of the lacquerware that comes into contact with the inner red lacquer film is labeled as surface ①. Following the ordering method in Fig. 19b, the surfaces are named in a clockwise direction as surfaces ②, ③, and ④. According to the extended depth-of-field microscope observations, the coiled wood core manufacturing process can be explained as follows: Surface ① and its opposite (surface ③) show marks caused by the fabric layer; surface ② and its opposite (surface ④) exhibit a distinct color boundary with varying shades. The formation of this phenomenon corresponds closely to the coiled wood core manufacturing process: surfaces ① and ③ constitute the inner and outer surfaces of the lacquerware, adjacent to the foundation layer; surfaces ② and ④, which are stacked with adjacent Cunninghamia lanceolata (cedar) wooden strips, exhibit color variation likely resulting from extended overlap during burial, where areas in direct contact with soil appear darker, while overlapping regions retain the original wood tone, forming a distinct boundary between light and dark areas on the same surface40.

Fig. 19: Schematic cross-sectional diagram of the restored core of the lacquerware.
figure 19

a The base of the lacquerware. b The wall of the lacquerware.

Full size image

The coiled wood core, exemplifying the traditional Quandie (圈叠) core-forming technique in lacquerware production in China, is formed by joining narrow wooden slats for the base, with a small point in the center of the base plate, and successively stacking strips of tough and resilient wood for the walls41. The coiled wood core technique differs from the rolled wooden core13. In coiled wood core, multiple thin wooden strips are layered and stacked, which disperses the internal stress of the wood, helps to overcome the anisotropy of the wood, and makes it less prone to deformation and cracking. The material range is more extensive, as ordinary wood can be used for the strip stacking, avoiding the limitations of using high-quality materials like camphor wood or nanmu wood that were previously required for making the wooden strips. This alleviates the problem of material shortages. Additionally, the appearance of the coiled wood core craft has provided more diverse shapes in lacquerware development. Unlike the bending wood technique, which can only produce cylindrical shapes like boxes, goblets, or cups42,43, coiled wood core is more suited for creating small, curved forms such as bowls, basins, and plates, as well as detailed molding of flowered edges in lacquerware.

The foundation layer of the lacquerware is composed of lacquer ash and fabric. The lacquer ash primarily consists of brick powder or tile powder, while the fabric—used to enhance structural stability—is inferred to be a coarse bast fiber fabric.

The entire lacquerware adopts the Suxiu (素髹) technique, which is characterized by the application of monochrome lacquer finishes without additional decorative patterns44. Red lacquer was applied to the interior surface, while black lacquer was applied to the exterior.

The raw lacquer used in this lacquerware contains a mixture of urushiol and laccol; tung oil was also detected in the film. The red lacquer film comprises two distinct layers: a dark brown primer layer and a red pigmented surface layer. The smooth and well-defined interface between them suggests that polishing was performed between layers. The primer layer, traditionally referred to as Caoqi (糙漆)45, serves an essential function in lacquerware production by leveling the underlying lacquer ash layer and providing a stable foundation for the subsequent pigmented lacquer and any decorative applications46. In contrast, the black lacquer on the exterior was applied as a single-layer coating, without a visible primer.

Overall, this study provides a solid scientific foundation for understanding the historical evolution of lacquerware manufacturing process. In particular, it offers compelling evidence supporting the coiled wood core craft as a distinctive and sophisticated method of core-forming. The findings contribute significantly to the reconstruction of ancient technological systems and the broader appreciation of traditional Chinese craftsmanship.