Beyond the Average Sherd: Low Manganese, Raw-Material Selection, and the Technological Ideal of Sky-Blue Ru Ware

A Preliminary XRF-Based Study and Testable Hypothesis on High-Quality Northern Song Celadon Abstract The celebrated sky-blue (tianqing, 天青) appearance of Northern Song Ru ware represents one of the most sophisticated achievements in historical Chinese ceramic technology. Yet much of the available archaeometric knowledge of Ru ceramics derives from excavated fragments, including material representing different production stages, firing outcomes, and glaze colours. This paper examines the possibility that exceptionally successful sky-blue vessels represent a chemically and technologically selected subset of Ru production rather than simply the average composition of excavated material. Three non-destructive X-ray fluorescence (XRF) measurements from a privately held, sky-blue glazed vessel are presented as a preliminary case study. Two measurements on the base report manganese concentrations of 69 and 89 ppm, with iron concentrations of 0.866 and 0.856 wt%, respectively. Potassium concentrations are 5.816 and 4.170 wt%, while titanium is reported at 0.135 and 0.096 wt%. A third measurement on the edge shows a substantially different elemental profile. These observations are discussed in relation to published research on official Ru glazes, particularly the work of Ding et al. (2020), which identifies low manganese and titanium contents, together with distinctive potassium- and calcium-bearing glaze compositions, as technologically significant characteristics of official celadon production. The central hypothesis is that careful raw-material selection, controlled impurity levels, glaze formulation, and reducing firing conditions may have contributed jointly to the production of the most successful sky-blue Ru ceramics. Importantly, low manganese alone cannot establish either the colour-forming mechanism or the historical attribution of an individual object. The present study proposes a comparative research programme capable of distinguishing this technological-selection hypothesis from alternative explanations. Keywords: Ru ware; Ruyao; Northern Song dynasty; tianqing; sky-blue celadon; manganese; iron redox; XRF; Qingliangsi; ceramic provenance; glaze technology; archaeometry. 1. Introduction: The Problem of the Archaeological Average Ru ware occupies a distinctive position in the history of Chinese ceramics because of its refined shapes, carefully controlled manufacture, and characteristic blue-green to sky-blue glaze. Scientific studies of Ru ceramics have established important compositional characteristics of the official Qingliangsi production tradition. However, a methodological problem remains: excavated archaeological assemblages are not necessarily representative of the most successful finished vessels. A kiln-site assemblage may contain production waste, broken objects, underfired or overfired pieces, and fragments of otherwise successful vessels. Consequently, the average composition of excavated material should not automatically be interpreted as the optimum technological composition. This paper introduces the concept of a technological selection hypothesis: that the most successful sky-blue Ru vessels may occupy a restricted region within the broader compositional and firing-technology distribution. The hypothesis is not that every successful vessel must contain less manganese than every archaeological sherd. Rather, it proposes a relationship between impurity control, iron oxidation state, glaze microstructure, and visual outcome that can be tested quantitatively. 2. Scientific Background: Manganese, Titanium, and Iron 2.1 Iron and the origin of celadon colour Iron is a principal chromophore in traditional celadon glazes. Its optical behaviour depends on concentration, oxidation state, coordination environment, and the structure of the surrounding glass. Reducing kiln conditions favour the formation of ferrous iron (Fe²⁺), which contributes to the characteristic blue-green appearance of many celadons. Ferric iron (Fe³⁺) contributes different absorption characteristics, frequently associated with more yellowish or brownish coloration. Nevertheless, the resulting colour cannot be predicted from the total iron concentration alone. Glaze thickness, phase separation, scattering, crystallisation, and firing history must also be considered. 2.2 The possible role of manganese Manganese is a multivalent transition element whose behaviour in silicate glasses depends on oxygen availability, temperature, composition, and cooling history. Under suitable conditions, manganese and iron may participate in coupled oxidation-reduction reactions, including: Mn³⁺ + Fe²⁺ ⇌ Mn²⁺ + Fe³⁺ This reaction provides a possible mechanism through which manganese could influence the distribution of iron oxidation states. However, its equilibrium direction and practical importance cannot be assumed without knowing the glaze composition and firing conditions. Manganese may also contribute directly to optical absorption. Therefore, a lower manganese concentration could be advantageous for obtaining a particular pale blue appearance, but it is not sufficient by itself to guarantee that colour. 2.3 Evidence from Palace Museum research Ding Yinzhong and colleagues (2020) investigated the technological relationship between official Ru, Southern Song Guan, and related celadon traditions. Their study reported an average glaze composition for 35 official Ru samples from Qingliangsi that included: K₂O: 4.21 wt% CaO: 12.03 wt% TiO₂: 0.12 wt% MnO: 0.10 wt% Fe₂O₃: 1.91 wt% These values represent glaze oxide compositions rather than elemental concentrations in the ceramic body. The authors associated the relatively low titanium and manganese concentrations of official-type glazes with the technological conditions favourable to the formation of the desired sky-blue appearance. They also emphasised the importance of potassium- and calcium-bearing raw materials in producing the characteristic thick, jade-like glaze. This published research provides a basis for investigating whether impurity selection was an intentional or empirically developed component of the official Ru technological tradition. 3. Preliminary XRF Case Study 3.1 Object and measurement context The present case study concerns a privately held vessel described by its owner as having a sky-blue Ru-type glaze. The reported collection history places the object in a private collection in the United States from approximately 1930. The object is said to have been stored with other ceramics and to have rested for an extended period on an old rubber support. This provenance is presently owner-reported and has not been independently verified through documentary archival research. Three XRF measurements were provided: two from the base and one from the edge. The available reports indicate a measurement duration of 45 seconds for each acquisition. The instrument model, calibration procedure, analytical atmosphere, beam geometry, reference standards, and precise relationship of each measurement spot to the glaze-body interface have not yet been documented. Accordingly, the results should be regarded as preliminary surface-sensitive measurements. 3.2 Elemental results Element Base 1 Base 2 Edge Aluminium (Al) 5.05% 4.44% 13.27% Silicon (Si) 39.28% 34.36% 31.63% Potassium (K) 5.816% 4.170% 2.411% Calcium (Ca) 7.599% 6.503% 8750 ppm Titanium (Ti) 1350 ppm 960 ppm 5740 ppm Manganese (Mn) 69 ppm 89 ppm ND (<570 ppm) Iron (Fe) 8660 ppm 8560 ppm 1.356% Zinc (Zn) 125 ppm 2091 ppm 2558 ppm Sulfur (S) 3120 ppm 3830 ppm 5890 ppm Table 1. Selected elemental concentrations transcribed from three supplied XRF reports. Values are reported as elemental concentrations, not oxide equivalents. ND indicates not detected above the stated analytical threshold. 3.3 Reproducibility of low manganese measurements The two base measurements report manganese concentrations of 69 and 89 ppm. Their arithmetic mean is 79 ppm, or 0.0079 wt% elemental manganese. The agreement is noteworthy, although the two measurements should not be treated as independent determinations of the entire vessel. Both spots may sample similar surface layers or overlapping material volumes. Furthermore, the reports do not establish whether the measured manganese originates predominantly from the glaze, body, or a mixture of both. The edge measurement does not establish similarly low manganese: its reported detection threshold is substantially higher, at 570 ppm. Consequently, the evidence supports low reported manganese in two base measurements, not a demonstrated uniform manganese concentration throughout the vessel. 3.4 Differences between base and edge The edge shows higher aluminium, titanium, and iron concentrations, together with lower potassium and calcium concentrations. These differences are consistent with the possibility that the edge measurement contains a greater contribution from the ceramic body. Alternatively, they may reflect differences in surface composition, geometry, glaze thickness, or analytical response. Without polished cross-sections or validated layer-specific measurements, the contributions cannot be separated reliably. This distinction is essential when comparing the present measurements with published analyses of archaeological Ru glazes. 4. A Technological Selection Hypothesis for Exceptional Sky-Blue Ru Ware The proposed hypothesis consists of four related processes. First: selection of suitable raw materials. Historical potters may have selected clays, feldspathic materials, and calcium-bearing ingredients according to their observed firing behaviour, plasticity, and final colour. Second: control of colouring impurities. Low concentrations of manganese and titanium in glaze raw materials may have reduced unwanted colour contributions and increased the reproducibility of the desired blue appearance. Th