A Yellow Tianqiuping Bottle Vase with Anhua Decoration and a Chenghua Mark: XRF Analysis and Microscopy

A yellow porcelain vase of the tianqiuping (天球瓶) form, featuring concealed anhua incised decoration (dragon, phoenix, and clouds) and a six-character Chenghua mark (object CHENGHUA-YEL-001), was examined using portable X-ray fluorescence (pXRF) and optical microscopy. The yellow coating is a high-lead, low-temperature glaze (Pb: 48.97 wt.% according to the instrument readings), with no detectable antimony, zirconium, vanadium, titanium, cobalt, or copper. The base, which bears the cobalt mark, exhibits a geochemical profile tentatively consistent with Jingdezhen materials (Rb/Sr ≈ 5) and contains no identified modern technological markers. The readings for Cd, Sn, Zn, Se, As, S, and other elements are interpreted as known artefacts associated with a high-lead matrix. Microscopy revealed a multilayer coating structure, including a stepped glaze edge, angular delamination of the upper layer along the crazing network, and bubbles in the underlying layer visible through the yellow coating. Ochre-brown darkening of the yellow glaze within the incised lines and pronounced iridescence were also observed. These features are consistent with published descriptions of early Ming imperial yellow glazes. Several questions remain unresolved: the low iron content measured at the surface (Fe: 820 ppm), the chromium reading (approximately 1,000 ppm), and the typological uniqueness of the vase form. The analytical and microscopic findings are consistent with the proposed Chenghua-period attribution but do not prove it. A programme of non-destructive verification is proposed. Keywords: Ming dynasty; Chenghua period; imperial yellow; anhua; tianqiuping; lead glaze; multilayer structure; iridescence; portable X-ray fluorescence; spectral artefacts. 1. Introduction Yellow monochrome porcelain of the Chenghua period (1465–1487) is documented by museum and archaeological evidence, although only a very small number of complete objects have survived. Museum corpus. No more than approximately twenty yellow monochrome dishes bearing Chenghua marks and attributed to the period are known; most are held in museums (Bonhams, 2020; Krahl, Sotheby's, 2017). Archaeological evidence. According to R. Krahl, Liu Xinyuan, director of the excavations at the imperial kilns in Jingdezhen, showed J. Thompson a box full of yellow sherds from a Chenghua-period archaeological layer and reportedly stated that there were many such boxes. Yellow coatings used in complex decorative programmes of the Chenghua period are illustrated in A Legacy of Chenghua (1993). The closest technological parallel. A pair of dishes decorated with incised anhua dragons, dated to the late Tianshun–early Chenghua period, features an egg-yolk-yellow glaze applied over a transparent crazed glaze (Bonhams, 2020). Archaeometric evidence. The colourant in Ming imperial yellow glaze is Fe₂O₃ dissolved in a lead-glass matrix, at concentrations of 1.81–4.87 wt.% (Hao et al., 2024). A transparent high-temperature layer, 150–450 μm thick and containing numerous bubbles, lies between the yellow glaze and the ceramic body (Hou et al., 2022). A slip layer has been described in Chenghua-, Hongzhi-, and Zhengde-period samples on the basis of TEM and EPMA analyses. Early Republican-period imitations also employed a slip layer, but their yellow colour was produced using antimony and tin (Zheng, Wu, Wu, 2008). The aim of this study is to determine whether the chemical and microstructural characteristics of the vase are consistent with these criteria and to identify what further evidence is required to substantiate the proposed attribution. 2. Object The object is a porcelain vase of the tianqiuping form, with a globular body and a tall cylindrical neck, covered with a yellow lead glaze. The anhua decoration, incised beneath the glaze, consists of a dragon, a phoenix, clouds, and a ruyi band around the neck. The base has an unglazed foot band and a white high-temperature glaze bearing a six-character Chenghua mark in underglaze cobalt blue. The archival reference is CHENGHUA-YEL-001; within the measurement series, the object is designated “Vase 4”. Surface condition: dense crazing across the yellow surface, pronounced iridescence, and a concentration of the yellow colour to an ochre-brown tone within the incised lines. The object was previously described by the author on the basis of gemmological microscopy (Dronova, Zenodo 20785217). The present study adds elemental data for the first time. 3. Materials and Methods Measurements were carried out in situ on 3 November 2026 using a portable XRF analyser (serial number 940487). Optical observations were made using macro- and microphotographs of the surface and glaze edge. Parameter Value Analytical mode Geochem (3-Beam) Extra; 30 seconds per point Point 1 (Daily ID 6, 13:59:53) Yellow glaze on the broad body; Sample ID 4, “vase yellow side” Point 2 (Daily ID 7, 14:00:51) Unglazed band at the base, with part of the white high-temperature glaze and the edge of the cobalt mark included; Sample ID 4, “vase yellow bottom” Measurement No. 5 in the same report (“vase yellow”) belongs, according to its Sample ID, to a different object (Sample ID 3) and was excluded from the present analysis. The reported values are elemental concentrations calculated by the instrument's built-in algorithm; uncertainties are given as ±3σ. The Geochem mode is based on a light-element matrix model and therefore provides semi-quantitative data when applied to high-lead glaze compositions. Spectra and readings from other glazes measured during the same session (12 objects, 19 measurement points) were used to assess possible spectral artefacts. 4. XRF Results The yellow coating is a high-lead glass containing no detectable antimony or industrial pigment elements. The measurement at the base produced a composition tentatively consistent with Jingdezhen materials and showed no identified modern markers. Table 1. Instrument readings (ppm unless otherwise indicated; ±3σ) and assessment of reliability Element Yellow glaze (ID 6) Base, mixed measurement point (ID 7) Assessment Pb 48.97% ± 0.15 450 ± 21 Genuine signal Si 16.53% 23.39% Genuine; absolute values are inaccurate Al 4,500 ± 1,200 3.68% Genuine K 4,440 ± 360 1.740% Genuine; underestimated in the yellow glaze (Section 6.4) Ca ND (<460) 2.106% — Fe 820 ± 170 4,600 ± 170 Genuine; low value in yellow glaze remains unresolved (Section 8) Mn 230 ± 180 749 ± 98 Genuine Co ND (<19) 63 ± 53 At the detection limit; measurement includes the edge of the mark Rb / Sr ND / ND 397 / 81 Genuine signals Cr 1,020 ± 220 ND (<110) Unresolved Cu ND (<37) 25 ± 22 No copper detected Sb ND (<10) ND (<93) Not detected Zr ND (<3) 65 ± 7 Not detected in the yellow glaze Ti ND (<510) 770 ± 560 Not detected / at the detection limit V ND (<62) ND (<210) Not detected As 3.249% 142 ± 20 Predominantly an artefact in the yellow glaze S 28.58% 9,080 Artefact in the yellow glaze Cd 1,420 ± 290 30 ± 27 Artefact Sn 850 ± 320 55 ± 35 Artefact Zn / Se 2,960 / 290 65 / ND Probably artefacts Nb / Mo / Hf / W 1,264 / 287 / 2,820 / 4,510 21 / 9 / 40 / 44 Artefacts At measurement point 2, the close-up photograph shows that the measurement field includes both the rough ceramic body and the glossy white glaze (LE 67%; poor contact). Consequently, the composition of the ceramic body itself cannot be determined from this measurement. The Pb reading of 450 ppm and As reading of 142 ppm at this point are most likely traces originating from the yellow glaze. 5. Spectral Artefacts Associated with the High-Lead Matrix An uncritical reading of Table 1 — for example, interpreting Cd at 1,420 ppm, Se at 290 ppm, and Zn at 2,960 ppm as genuine concentrations — could lead to the false conclusion that the object contains modern cadmium yellow. Examination of the spectrum indicates that these readings are artefacts. 5.1. Pulse Pile-Up Peaks: Cd and Sn At count rates of approximately 20,000 counts per second on the lead lines, the spectrum contains a triplet of sum peaks at approximately 21.1, 23.2, and 25.3 keV, with the middle peak being the most intense. These correspond to Pb Lα + Lα, Pb Lα + Lβ, and Pb Lβ + Lβ, respectively. The second peak coincides with Cd Kα (23.2 keV), while the third coincides with Sn Kα (25.2 keV). Such peaks disappear when the X-ray tube current is reduced (Lee, 2016; WAAC, 2009). In three different glazes measured during the same session and containing approximately 50% Pb, the apparent “cadmium” concentration is nearly identical (1,420–1,450 ppm). In contrast, cadmium was not detected in a glaze containing 7.4% Pb (below 19 ppm). 5.2. Sulphur and Arsenic Sulphur (S: 28.58%). The apparent sulphur signal results from overlap between the Pb M lines (2.35 keV) and S Kα (2.31 keV). A comparable case was reported for high-lead Corning C reference glass, where portable XRF incorrectly indicated approximately 30% SO₃ while underestimating light elements. Arsenic (As: 3.25%). This reading results from the overlap of Pb Lα (10.55 keV) and As Kα (10.54 keV) (Rowe et al., 2015). The As/Pb ratio is 0.066, matching the ratios in two other glazes containing approximately 50% Pb (0.065 and 0.066), which is characteristic of a lead-related artefact. The small peak at approximately 11.35 keV is Pb Lη, not As Kβ. 5.3. Other Lead-Related Signals The readings for Zn, Se, Mo, Nb, Hf, and W are consistently reproduced in all glazes containing approximately 50% Pb and are explained by escape and sum peaks. A hafnium reading of 2,820 ppm in the presence of zirconium below 3 ppm is physically implausible for a genuine material composition. A high-lead matrix distorts the pXRF determination of trace elements and requires dedicated calibration. This limitation is also documented in comparisons of pXRF and LA-ICP-MS analyses of lead-rich glass tesserae. 6.