Bronze smelting activities during the Zhou dynasty: Analysis of slags from the Xincun and Baidian foundry sites in Central China

doi:10.21203/rs.3.rs-2857759/v1

Download PDF

Research Article

Bronze smelting activities during the Zhou dynasty: Analysis of slags from the Xincun and Baidian foundry sites in Central China

https://doi.org/10.21203/rs.3.rs-2857759/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

The highly developed metallurgy in Central China during the Zhou dynasty represents an important stage in the development of mining and metallurgical civilisation. The Xincun and Baidian foundry sites are important sites found in Central China, with the main remains of the Wei and Jin states, respectively. Elemental analyses and microstructural examinations of slags from both the sites were conducted in this paper, aiming to reconstruct the technological landscape of bronze alloys and smelting processes at both the sites, and to further explore the differences or/and similarities between the bronze processing activities at the two sites in different historical contexts. The results showed that the co-smelting of tin ore with metallic copper was still used in the alloying process at both the Xincun and Baidian foundry sites. Also, the smelting temperatures at both the sites were relatively high with the addition of calcareous fluxes to the process. In comparison, the Baidian foundry site was found to have a more diverse type of alloying technologies and more advanced metallurgical process than the Xincun foundry site, indicating a more advanced metallurgy in the Jin state than in the Wei state. These findings provide valuable insights into metallurgy in these areas and have important implications for future research in this field.

Central China

Slag

Bronze production

Alloying technology

Smelting condition

The large number of splendid bronzes excavated in Central China during the Shang and Zhou period have shaped the unique cultural character and artistic style of the Chinese Bronze Age. It has attracted widespread international academic attention and discussion (Nickel 2006; Bagley 2009; Pollard et al. 2018). The advanced smelting and casting technology used in the bronze production played a pivotal role in the ancient Chinese metallurgy. Studies of the bronze alloy technology of the Central Plains, the political heartland of China during the pre-Qin period, are important in exploring associations between metal resources, technological choices, and social change. Many scholars have conducted research on metal remains to explore the smelting processes and bronze alloying technologies.

The emergence of the Bronze Age in Central China can be traced back to the Erlitou period, c. 1750 − 1520 BCE (Su et al. 1995). Most of slags found from foundry sites of this time were molten copper slags, and bronzes were made by adding tin ore to copper metal (Li and Xu 2007; Liang et al. 2005; Liang 2019). A technological shift towards the use of tin ingots in alloying took place during the Shang dynasty (Liang et al. 2005 & 2009; Cui et al. 2022). However, it is possible that lead-tin symbiotic ore was still used as an alloy ingredient in the early Shang dynasty (Huang et al. 2011). The use of tin ingots to make bronzes continued during the Zhou period, and the presence of secondary cassiterite and stannous calcium (iron) silicate in the slag was taken as evidence for the use of tin ingots (Zhang et al. 2021). The research on bronze casting in Central China has been concentrated on the early stage of the Chinese Bronze Age up to the Shang dynasty, whilst relatively fewer studies on the Zhou dynasty resulted in limit information about the technological advancements during this period. In this study, metallurgical remains from the Xincun and Baidian foundry sites were analysed using a variety of analytical methods, to help better understand the development of metallurgical technology in Central China during the Zhou dynasty.

1.1. Archaeological background

The Xincun foundry site is located in the middle east of Xincun village in Hebi city, Henan province. It was the capital of the Wei state during the Western Zhou period (1046 − 771 BCE) (Fig. 1) (Gao et al. 2020). Based on stratigraphic association and pottery typology, the Wei occupation of this site can be divided into two phases: the 1st being in the late Western Zhou period, and the 2nd being in the Eastern Zhou period (770 − 256 BCE), dating to transition between Spring and Autumn period and Warring States period. The metallurgical remains used in this study were from ash pits of the 2nd phase.

The Baidian foundry site is an important part of the Houma foundry, the largest Bronze Age foundry found to date in China. It is located in the northwest of Baidian village in Houma city, Shanxi province, also the capital of the Jin state during the Eastern Zhou period (770 − 256 BCE) (Fig. 1). The foundry at Baidian was discovered during a highway construction in 2003 and excavation began in the same year (Shanxi Provincial Institute of Archaeology 2012). The excavations revealed a large number of ash pits, tombs and ditches on the site, which was dated to the transition between Spring and Autumn period and Warring States period based on pottery found at the site.

Despite the fact that both the Xincun and Baidian foundry sites are of the same date, they were of different importance. Due to the invasion of the Di (狄人), the Wei state moved its capital eastwards and its control over the Xincun region was greatly diminished (Chen 2019). The Baidian foundry, in contrast, was at the political heart of the area during the Eastern Zhou period and was continued to be used uninterruptly with a large scale of bronze production. The elemental analyses and microstructural examinations were conducted in this study to explore the bronze alloying technologies and smelting processes at both the sites, and further to identify the differences or/and similarities in bronze processing activities at the two sites in different historical contexts.

2.1. Materials

A total of 12 samples were selected for this study, including two slags and one crucible fragment from the Xincun foundry site, and four slags and five crucible/furnace wall fragments from the Baidian foundry site (Fig. 2). The slags were mostly irregular in shape and different in size, with vitrification observed on all the surfaces. The smaller crucible samples consisted of vitrified inner layer and outer layer of orange grass-clay mixture (samples H210:1; BH2:08; BH17:020; BH20:043). The samples of furnace wall fragments were relatively large, and sample H15:07 had slightly vitrified.

2.2. Analytical methods

In this study, a Leica DVM6M digital microscope, a Leica DMi8 metallurgical microscope, and electron scanning microscopy coupled with energy dispersive X-ray spectrometry (SEM-EDX) were used for surface and microscopic observation, metallographic analysis, and chemical composition analysis of these samples.

Metallographic samples were first prepared by cutting the slags to include all the different layers as far as possible, such as the vitrified and unvitrified parts. The samples were mounted in epoxy resin, grounded with sandpaper of different grit sizes (600–1200), and then polished to a finish of 0.25µm. A Thermo Scientific Quattros SEM with a Bruker Xflash 6160 EDX was used for elemental analysis of various phases and metal prills in the samples. The analyses were run at an accelerating voltage of 20 kV at a low vacuum (50 Pa) with a working distance of 8.5 mm, and a scanning time of 90 seconds. All SEM images in this paper are backscattered electron images.

3.1. Xincun samples

Microstructural and compositional analysis of slags H91:4 and H98:3 shows that the two contain metal prills of different sizes, quantities and compositions. The metal inclusions in slag H91:4 are copper prills with approximately 99.4% Cu (Fig. 3a), and trace amounts of tin (Sn) and lead (Pb) as impurities. Almost all of the metal prills in slag H98:3 are Cu-Sn-Pb alloys, with the tin content varying over a wide range (10.7–40.1%) and is to some extent reflected in the metallographic structure of the metal prills (Figs. 3b&c). Some of the metal prills in H98:3 have higher lead contents, up to 28.3% in one of the prills. And sulphide inclusions with 20% S are present in the α-phase of the bronze matrix, suggesting that they are derived from copper (Fig. 3d).

Table 1

The SEM-EDX results of analyses on the labelled areas (Fig. 3) of metal inclusions in slags from the Xincun foundry site (wt%) (data: C Wang )
Sample	Area of analysis	Identification	Cu	Sn	Pb	Fe	As	S
Slag H91:4	1	copper prill	99.4	0.1	0.5	-	-	-
Slag H98:3	2	alloy prill	72.9	21.5	4.4	-	1.2	-
	3	alloy prill	84.2	10.7	5.1	-	-	-
	4	sulphide inclusion	72.2	4.6	-	3.2	-	20.0

The different areas of the matrix of the slag and crucible samples were analysed by SEM-EDX and the results are shown in Table 2. The microstructure of crucible H210:1 showed the presence of a vitrified layer, a partly vitrified region, and a non-vitrified matrix. Iron (Fe)-rich particles were found in both the vitrified layer and the partly vitrified region. The vitrified layer is characterised by high silicon (Si), aluminum (Al), moderate iron, and low calcium (Ca) content, which is generally consistent with the composition of the matrix, but with a slightly higher iron content. The elevated iron in vitrified layer is probably resulted from the direct contact between the crucible wall and molten metal, metallurgical raw material or flux. H210:1 has not developed a slag lining, and no metallic prills were observed within the layer. But the bright area in the SEM image (Fig. 4a) (contain 65.4% Pb, 15.2% Fe, 10.6% As, 8.8% Sn) within the crucible may correspond to a melting or alloying processing.

Table 2

The SEM-EDX results of matrix of the slag and crucible samples from the Xincun foundry site (wt%) (data: C Wang )
Sample	Areas of analysis	SiO₂	Al₂O₃	FeO	CaO	K₂O	MgO	Na₂O	SnO₂	PbO	TiO₂
Slag H91:4	Matrix	65.4	17.9	5.7	1.7	4.0	1.9	1.8	0.4	-	1.2
Slag H98:3	Central area	26.0	7.3	3.2	26.8	0.7	2.6	1.1	30.8	1.4	0.1
Slag H98:3	Outer area	38.9	7.9	3.1	36.2	1.5	6.7	0.2	4.0	0.8	0.7
Crucible H210:1	Vitrified layer	69.5	14.9	7.0	1.6	3.3	1.4	1.2	0.4	0.2	0.5
	White irregular particle	12.7	3.9	81.4	0.4	0.4	1.2	-	-	-	-
	partly vitrified	69.7	14.0	6.2	1.9	3.3	2.0	1.7	0.4	0.2	0.5-
	non-vitrified matrix	69.6	15.9	5.1	1.6	3.5	2.0	1.8	0.2	0.1	0.2

Composition analyses of the slag matrix revealed that slags H91:4 and H98:3 are highly heterogeneous. The vitrified matrix of slag H91:4 was found to mainly contain silicon (Si), aluminum (Al), and iron (Fe), similar to the composition of the vitrified layer of the crucible sample. Silicate with high copper (Cu) content were found in the glassy matrix (Fig. 4b), indicating the heterogeneous redox conditions inside the smelting container. The presence of cuprous oxide (Cu₂O) in the copper prill suggests that the copper oxide (CuO) may not have been fully reduced. Of course, cuprous oxide may have also resulted from the oxidation of copper. Bright granular clusters and black needle-like clusters were identified within the silicate (Fig. 4b), which is probably unreacted ore remains, including malayaite and lead-bearing aluminosilicates (Rovira et al. 2009).

Slag H98:3 showed a higher calcium (Ca) content (26.8–36.2%) than the crucible sample, which could be attributed to the gangue of the ore or fluxes used during the smelting (Li et al. 1995). The matrix of this sample is a Si-Ca-Sn-Al-Mg glass, which varies in chemical compositions at different areas and contains multiple phases. Well-developed cassiterite crystals, biconical tin oxide crystals, cellular calcium tin silicate and malayaite (CaSn(SiO₄)O) were found present in the matrix and encircle metal prills. Additionally, calcium-rich phases were distributed in areas with few metal prills. Malayaite is typically formed through the reaction between cassiterite and calcium-rich pyroxenes or wollostonites (CaSiO₃), and could be taken as evidence for the use of tin ores (Rovira 2007). Many of the biconical tin oxide crystals had hollow cores enclosing bronze prills, which is a typical feature of secondary cassiterite formed during the oxidation of bronze (Renzi and Rovira 2016; Rademakers and Farci 2018) (Figs. 4c- f).

Table 3

The SEM-EDX results of analyses on the labelled areas (Fig. 4) of different phases in slags from the Xincun foundry site (wt%) (data: C Wang )
Sample	Areas of analysis	Identification	SiO₂	Al₂O₃	FeO	CaO	K₂O	MgO	Na₂O	CuO	SnO₂	PbO	TiO₂
Slag H91:4	1–3	1.malayaite	29.9	2.5	1.4	20.6	-	0.3	1.0	2.7	40.5	1.1	-
		2.leading-bearing aluminosilicates	54.9	22.6	3.2	2.2	-	0.6	10.4	2.3	0.5	3.2	-
		3.silicate arrounding metal	53.0	16.1	4.2	4.8	0.4	0.6	7.6	5.3	2.2	4.2	1.6
Slag H98:3	4–8	4.cassiterite	13.7	2.6	1.3	5.3	-	1.5	1.5 Cl:0.2	1.9	71.5	0.8	-
		5.tin oxide	14.2	3.2	1.0	7.6	-	1.8	1.9	-	69.5	0.8	-
		6.malayaite	19.4	1.6	0.6	20.0	-	1.0	1.1	-	55.6	0.6	-
		7.calcium tin silicate	21.9	10.8	5.8	27.6	-	2.5	1.0	-	30.5	-	-
		8.calcium-rich phase	38.6	6.8	2.4	40.5	0.9	9.9	0.9	-	-	-	-

3.2. Baidian samples

SEM-EDX analysis of the composition of metal prills in these four slag samples from the Baidian foundry site revealed the presence of three types of metal prills: copper matte, tin bronze, and Cu-Sn-Pb ternary alloys.

1) copper matte was only found in slag BH2:07. The metal prill was partially oxidised, and barium (Ba) was detected in some of the area (Fig. 5a).

2) the metal prills in slag BH20:044 were all tin bronzes with dendritic structure of α+ (α + δ). One of the prills had corroded due to chloride ions (Cl), resulting in copper loss and the enrichment of tin. Iron (Fe) and antimony (Sb) were found present in some of the prills, in which antimony could have originated from the ore and iron possibly from either the ore, the flux or the soil (Fig. 5b).

3) most of the metal prills in slags BH2:09 and BH20:045 were found to be Cu-Sn-Pb ternary alloys. Bronze prills in slag BH20:045 were found to have tin contents between 1.0 and 98.2%. Most of the prills were small and contained high tin (Sn). Some of the bronze prills contained small amounts of arsenic (1.2–4.9%) and trace amounts of antimony (0.8–1.1%). One metal prill had an uneven composition with a clear boundary between the lead-rich phase and the alloy phase. The lead-rich phase was found to contain up to 87.4% lead (Pb) and 4.4% silver (Ag) (Fig. 5c). Bronze prills in slag BH2:09 were found to contain 2.1–9.8% tin with the presence of arsenic (As, 1.1–2.8%) and antimony (0.3–0.9%). Trace amounts of iron (0.1–0.5%) were also found in some of the prills. The arsenic containing metal prills were present in the central area of the sample, mostly in association with lead prills (Fig. 5d).

Table 4

The SEM-EDX results of analyses on the labelled areas (Fig. 5) of metal inclusions in slags from the Baidian foundry site (wt%) (data: C Wang )
Sample	Areas of analysis	Identification	Cu	Sn	Pb	Fe	As	S	Other
Slag BH2:07	1–3	1.copper matte	68.5	-	-	4.0	-	27.5	-
		2.oxidised area	59.0	-	-	2.6	-	28.0	O:10.4
		3.barium-containing area	63.9	-	-	24.2	-	6.1	Ba:5.8
Slag BH20:044	4–6	4.tin bronze	26.3	60.0	-	11.2	-	0.5	Sb:2.0; Cl:0.2
		5.corroded area	59.9	-	-	1.3	-	-	Cl:12.9; O:25.9
		6.copper-rich area	95.5	3.1	-	-	-	-	Cl:1.4
Slag BH20:045	7–10	7.alloy phase	29.7	37.8	29.1	0.2	3.3	-	-
		8.silver-containing area	3.0	3.2	87.4	-	2.0	-	Ag:4.4
		9.lead-rich phase	1.6	2.7	94.3	-	1.4	-	-
		10.sulphide inclusion	72.1	1.3	7.1	-	-	19.5	-
Slag BH2:09	11–12	11.bronze prill	86.3	8.1	4.7	-	-	-	Sb:0.9
Slag BH2:09	11–12	12.arsenic bearing prill	8.4	3.9	75.3	0.5	2.4	-	Sb:0.9; Cl:8.7

The bulk composition of the slags and crucible/furnace wall samples were analysed with SEM-EDX and the results are shown in Table 5. The crucible and furnace walls mainly contain silicon (Si), aluminum (Al), iron (Fe) and calcium (Ca). The irregular white particles present in the SEM image of the crucible/furnace walls (Fig. 6a) were found to mainly contain iron and titanium (Ti), pointing to probably ilmenite. Notably, the vitrified layers of crucibles BH2:08 and BH20:043, as well as furnace wall H17:017 show significantly higher calcium content than other areas within each of the samples (Table 5). In smelting process, calcium oxide (CaO) decomposed from the gangue in the ore, the flux, or calcium feldspar particles in the furnace wall under high-temperature calcination environment can easily react with impurities, forming silicate minerals that adhere to the inner wall of the furnace (Yu 2010). Therefore, the elevated calcium in the vitrified layer could have resulted from the ore smelting or the addition of calcareous fluxes. Moreover, small amounts of tin (﹤1%) were detected in crucibles BH17:020 and BH20:043, indicating their possible association with tin metallurgy. Furthermore, the bright area in crucible BH20:043 (Fig. 6b) was found to contain 68.8% Ba, 21.6% Fe, and 9.6% Sn, pointing to the smelting of tin ores.

Table 5

The SEM-EDX results of the slag, crucible, and furnace wall samples from the Baidian foundry site (wt%) (data: C Wang )
Sample	Area of analysis	SiO₂	Al₂O₃	FeO	CaO	K₂O	MgO	Na₂O	SnO₂	PbO	TiO₂
Slag BH2:07	Matrix	67.1	13.5	4.4	8.4	2.9	1.8	1.4	-	-	0.5
Slag BH2:09	Matrix	56.8	13.8	4.1	13.3	4.0	2.7	2.5	0.4	1.9	0.6
Slag BH20:044	Matrix	30.3	7.3	15.5	41.7	1.7	2.4	0.6	-	0.1	0.5
Slag BH20:045	Matrix	63.5	13.1	4.5	11.3	2.9	2.3	1.4	0.1	0.1	0.7
Crucible BH2:08	Vitrified layer	61.2	17.5	6.0	6.2	3.9	2.5	1.7	0.1	-	0.8
	matrix	63.0	17.9	6.7	4.2	3.6	2.2	1.4	0.1	-	0.8
	White irregular particle	16.9	7.9	43.1	1.2	1.2	3.4	0.5	-	-	25.8
Crucible BH17:020	Vitrified layer	60.4	20.3	8.3	1.2	4.3	2.7	0.9	0.7	0.3	0.9
Crucible BH17:020	matrix	58.8	20.6	9.1	2.0	4.8	2.7	0.9	0.3	-	0.9
Crucible BH20:043	Vitrified layer	61.4	14.6	4.8	10.8	3.2	2.3	2.0	-	-	0.9
	matrix	67.4	13.2	4.9	6.2	3.8	1.8	1.8	0.2	-	0.8
	White irregular particle	25.6	6.3	23.8	4.3	1.2	2.7	0.9	-	-	35.2
Furnace wall H15:07	Vitrified layer	65.3	14.0	5.7	7.1	2.8	2.1	1.8	-	0.4	0.7
	matrix	65.2	14.5	5.2	7.6	3.0	1.9	1.6	-	-	1.0
	White irregular particle	13.8	3.6	79.9	0.9	0.4	0.7	0.4	-	-	0.2
Furnace wall H17:017	Inner wall	59.8	14.3	5.3	13.0	3.2	2.2	1.7	-	-	0.7
	matrix	62.1	14.6	5.6	9.8	2.9	2.2	1.7	-	-	1.1
	White irregular particle	9.7	2.6	36.9	2.3	0.3	0.6	0.3	-	-	47.3

The slag samples from the Baidian foundry site show heterogeneous microstructure with different phases, and SEM-EDS analysis of the slag matrix shows that they contain mainly silicon (Si) and aluminum (Al), with a small amount of iron (Fe) and calcium (Ca). Acicular, rhombohedral or skeletal shape crystals of tin oxide (containing approximately 70.9% Sn) and copper&lead-rich silicates are embedded in the matrix of slag BH20:045 (Fig. 7a). It has generally been believed that acicular and skeletal tin oxide crystals tend to form through the oxidation of tin and bronze (Rademakers and Farci 2018). Grey phases containing a significant amount of tin were discovered in slag BH2:09. Microanalysis of this phase gave a chemical composition similar to that of stokesite (CaSn(Si₃O₉)▪2H₂O) (Xue et al. 2017). Furthermore, a large number of prismatic lead silicate crystals are also distributed in this area (Fig. 7b).

The matrix of slag BH20:044 within this batch of samples shows a significant level of calcium and iron, bearing several iron-rich phases such as iron silicates, hematite (contain 30.0% Fe, 70.0% O, corresponding to Fe₂O₃), and iron-rich tin-calcium silicates (Figs. 7c&d). The iron silicate mainly contains iron (64.6%), oxygen (31.4%) and silicon (4.0%), thus consisting mainly of 3FeO▪SiO₂. Barite (BaSO₄), pyroxene, calcium-rich phase and magnesium-rich phase were also observed in the matrix. (Fig. 7e). The higher iron contents detected in the high tin bronze prills than in the copper-rich phases (Fig. 7f) suggest that most of the iron may have been brought into the alloy via the tin materials used. And the barite inclusion is usually considered a direct remnant from ore (Mueller et al. 2004; Yu et al. 2021), hence indicating possible smelting of tin ores.

Table 6

The SEM-EDX results of analyses on the labelled areas of different phases (Fig. 7) in the slags from the Xincun foundry site (wt%) (data: C Wang)
Sample	Areas of analysis		SiO₂	Al₂O₃	FeO	CaO	K₂O	MgO	Na₂O	CuO	SnO₂	PbO	Other
Slag BH20:045	1	1.copper&lead-rich silicates	24.7	2.4	0.8	7.0	1.3	0.6	0.2	39.1	4.1	13.8	MnO:6.0
Slag BH2:09	2–3	2.probably stokesite	32.7	3.9	1.0	18.7	-	0.7	1.0	-	37.6	4.4	-
Slag BH2:09	2–3	3.lead silicate	11.2	2.4	1.4	6.2	-	0.6	-	-	4.0	74.2	-
Slag BH20:044	4–13	4.iron silicate	8.4	2.6	75.3	10.7	-	2.1	0.9 Cl:0.2	-	-	-	-
		5.tin-calcium silicate	7.1	1.8	4.0	29.7	-	0.6	0.9	-	55.9	-	-
		6.iron-rich tin-calcium silicates	26.3	6.9	16.5	26.0	-	1.3	0.3	-	22.6	-	-
		7. copper-tin containing silicates	46.7	8.4	11.7	13.2	1.3	4.9	0.4	7.3	6.2	-	-
		8.pyroxene	34.3	4.4	12.4	43.4	0.4	4.5	0.7	-	-	-	-
		9.barite	8.7	-	2.3	4.9	0.7	2.0	-	-	-	-	SO₃:39.2; BaO:35.5; TiO₂:6.3
		10.calcium-rich phase	7.4	1.1	4.6	76.5	0.2	2.9	0.4	-	-	-	SO₃:4.5; BaO:1.8; TiO₂:0.6
		11.magnesium-rich phase	56.0	1.5	4.4	11.9	-	23.8	0.6	1.8	-	-	-
		12.hematite	6.5	1.4	82.7	4.3	-	4.7	0.4	-	-	-	-
		13.copper-rich phase	22.8	3.6	2.4	12.1		1.5	2.2	47.2	1.9		SO₃:0.7 MnO:5.4

4.1. Alloying technology as reflected in the slags from the Xincun and Baidian foundry sites

Some differences were observed between the copper slags from the Xincun and Baidian foundry sites. Although both slags consisted of pure copper, the copper is present as copper prills in the Xincun slag, whilst it is present as copper matte prills in the Baidian slag, indicating different raw materials or smelting processes used in the production of copper between those two sites. The nature of copper slag needs to be discussed, since copper smelting, refining, and melting can all generate copper-rich slag free from alloying elements. The Xincun slag H91:4 had a heterogeneous matrix with relatively low iron but high silicon contents. The presence of unreacted copper-bearing minerals in the slag matrix (Fig. 4b) suggests that this slag probably resulted from smelting of natural ores. The absence of sulfur (S) in the slag indicates the likelihood of oxidised ores or ores with little sulphied being smelted at the site. The presence of copper matte in Baidian slag BH2:07 suggests possible use of sulphide ore-copper matte-copper smelting process at the site. This process is described in Eibner‘s model, i.e. the matte resulted from the roasting was then smelted to metallic copper (Eibner 1982). However, the copper matte found here was too small both in size and amount to reflect the true grade of the copper matte produced, therefore, it is difficult to conclude that a sulphide ore-copper matte-copper smelting process was used at the Baidian foundry site. A direct reduction of sulphide-bearing oxide ores or mixed ores was also possible (Li et al. 2020).

The copper smelting process represented by the copper smelting slag of Xincun and Baidian foundry sites provided a foundation for bronze production at the sites. Previous researchers have suggested several ways of bronze production: co-smelting a mixture of copper-tin ores; alloying of copper with mineral cassiterite (SnO₂ present in nature); melting metallic copper and tin in right proportions; remelting of waste bronze, with the possible addition of copper/tin metal or ore (Sloley and Lucas 1948; Rovira et al. 2009; Rademakers and Farci 2018).

It is noteworthy that a large number of tin-bearing phases are retained in the matrix of Xinxun slag H98:3. Rademakers et al. (2018) reconstructed bronze production processes by identifying tin oxide and cassiterite in slag through experimental archaeology. While the remains of unreacted mineral cassiterite can be easily identified, indicating the use of tin ore. The presence of partially re-crystallised mineral cassiterite which are typically composed of euhedral cluster crystals, could also point to the use of tin ore. Therefore, it is likely that the clustered crystals in this sample are the mineral cassiterite. Additionally, malayaite in slag is often resulted from the reaction between cassiterite and calcium-rich pyroxenes and can be taken as an indication of the use of tin ores (Rovira et al. 2009). Furthermore, malayaite occurs in tin-rich contact metamorphic skarn deposits and is also a hydrothermal metamorphic product of cassiterite or other tin-bearing minerals. No phases associated with copper smelting were found in slag H98:3, so it has probably resulted from a co-melting process using tin minerals and metallic copper.

The slags from the Baidian foundry site reflect the diversity of alloying technologies compared to the Xincun foundry site. The enrichment of copper&lead-rich silicates in the matrix of slag BH20:045 may be indicative of the use of copper-lead symbiotic ores (as opposed to metal) in the alloying process, which would be more reactive to form silicates (Julia 2020). For the alloy prills, antimony (Sb) was not detected in the lead-rich phase metal prill, but a significant amount was found in the high-tin prills. Thus, antimony is more likely originated from the tin materials, e.g. tin ores or tin ingots. However, the presence of a large lead particle suggests the possible use of lead ingots (Fig. 5c). Lead ingots were reported to have been found in the Houma foundry (Shanxi Provincial Institute of Archaeology 2012).

In slag BH2:09, the metal prills were found to be of ternary alloys, but none containing extraordinarily high-tin. The extraordinarily high-tin prills indicate that metallic tin or cassiterite was freshly added to the melt, which appeared not the case for this slag (Rademakers and Rehren 2016; Liu et al. 2020). Remelting of waste bronzes for alloying could have produced such a slag. Additionally, the irregular shape of the metal prills in this sample may be related to melting (Cui et al. 2021). The presence of stokesite and lead silicate in the matrix indicates possible addition of tin (lead) minerals to recycled bronzes in the melting process.

As mentioned previously, the presence of barite in the matrix of Baidian slag BH20:044 indicates the use of tin ore. Higher iron content detected in the high tin bronze prills than in the copper-bearing phase implies that most of the iron may have been introduced from the tin materials into the alloy. Iron minerals tends to react with cassiterite under high temperature at reducing conditions during roasting to form tin&iron-rich compounds, which may attribute to the presence of iron-rich calcium-tin silicate in slag BH20:044 (Annemie 1996; Su 2014). As tin is preferentially oxidised over copper, there should not be a situation in which copper is oxidised while tin is still present as a metal (Rovira 2007; Dungworth 2000). This suggests that the copper-rich phase present in the slag had formed before tin was added to the charge (Li et al. 2007). During melting, metallic copper can turn into cuprite due to poorly controlled redox conditions (Liu et al. 2020). The copper-rich phase was likely the outcome of such an operation. Slag BH20:044 may represent a process in which tin ores were added to the molten copper to make bronzes. The bright area in the SEM image of crucible BH20:043 (Fig. 6b) appeared to have a similar composition to a barite particle present in slag BH20:044 (Fig. 7e), suggests that the crucible could have been used for an activity from which slag BH20:044 was produced.

4.2. Possible use of flux in the slags of the Xincun and Baidian foundry sites

The significant level of calcium detected in Xincun slag H98:3 and Baidian slag BH20:044 is worth noting, and in the latter the high iron content was also noted. Both furnaces and crucibles at these sites have low calcium content. It is therefore possible that the calcium comes from an external addition, i.e. a gangue in the ore or calcareous fluxes. However, the contribution of the calcium content of the gangue to the slag composition cannot be conclusive due to the lack of original ore in this study. Calcareous fluxes were reported to have been used in the iron smelting industry of the pre-Qin period, and similar levels of calcium (to that reported here) were found in iron-smelting slags (Qin et al. 2016; Du et al. 2011). In contrast, such a high calcium content (Tables 2&5) has rarely been found in copper slags, and no calcareous fluxes such as limestone (CaCO₃) or dolomite (CaMg(CO₃)₂) have been found from bronze smelting and casting sites.

The different chemical and mineralogical compositions observed in the two very different areas (Figs. 4e&f) of Xincun slag H98:3 indicate its heterogenous feature. The central area of the slag has a high tin content and promoted the formation of some tin-bearing phases such as tin oxide, malayaite and stannous calcium silicate. While the outer area is more enriched with calcium and magnesium than the central area (Fig. 8). A slag found at a Bronze Age mining site at Mušiston (Tajikistan) was reported similar to this situation. The slag from Mušiston (Tajikistan) contained higher tin and zinc contents in the central area, and much richer in iron in the outer area than in the centre. Berger suggested that the predominance of tin and zinc (compounds) in the centre together with the numerous metallic inclusions limits the zone of ore reaction to this area, thus the sample may have experienced a low smelting temperature or insufficient smelting time (Berger et al. 2022; Hauptmann 2007). Based on these points, the central area of slag H98:3 is probably closer in chemical composition to the original tin ore than its outer area, and the original tin ore type could have been a calcium-rich vein type. The higher calcium content in the outer area may not be entirely derived from the ore (Berger et al. 2022). The calcium content of the crucible was only 1.6%, which is much lower than that in the slag, indicating an introduction of considerable amount of external calcium during the smelting process. The poor homogeneity of the slag may be due to the uneven distribution of tin in the ore, resulting varying degrees of reaction between the ore and the calcareous fluxes in different areas (Berger et al. 2022).

The Baidian slags show higher calcium content than the Xincun slags, with most slags containing approximately 10% calcium. In particular, slag BH20:044 contains an extraordinarily high level of both calcium and iron in the matrix (41.7% CaO, 15.5% FeO). The calcium oxide in the slag is predominantly present in the glassy phase and there is a large amount of calcium-rich pyroxene dispersed in the matrix. Presently it is difficult to ascertain whether this attribute deliberately added flux or to the gangue in the ore, but there is a greater likelihood that it is the former. The iron and silicon in slag BH20:044 facilitated the formation of iron silicate phases, hematite and iron-rich tin-calcium silicate phases. Such iron-rich phase in the slag is believed to have formed during smelting or refining iron-containing copper and/or tin ores. Hauptmann (2003) argued that early Bronze Age smelters used high-grade copper ores containing silicon and iron to facilitate the smelting reaction, referring this phenomenon to self-fluxing, suggesting that people were already aware of the role of silicon and iron in the slag-making process. In contrast, a smelting slag of the Zhou dynasty found at the Yangxin Dalupu site in Hubei province contained more than 70% FeO, less than 12% CaO and 25% SiO_2, respectively. This suggests that smelters at that period have intentionally added iron ore to lower the slag viscosity, and facilitate its discharge from the smelting furnace (Zhu 2021). Despite of a higher iron content (15.5%, Table 5) found in slag BH20:044 than in other samples, the phase identification and iron content alone are insufficient to point to a deliberate use of iron ore fluxes. The iron could have also been impurities present in the tin ore.

4.3. Smelting conditions in the Xincun and Baidian foundry sites

Viscosity is a critical property of slag material that determines the separation of silicates from metal melt. Exploring the range of viscosities of slags offers insight into the smelting process for different types of slag. The addition of fluxes is a usual way to decrease viscosity and increase fluidity. However, the presence of a large amount of metal prills in the centre of the Xincun slag H98:3 indicates that the viscosity of the slag was not low enough to facilitate complete separation of the metal from slag.

The model proposed by Bachmann (1989) was used in this study to calculate the viscosity index (K) of slag, i.e. the ratio of the viscosity-decreasing to viscosity-increasing oxides based on the following equation:

K= (CaO + MgO + FeO + TiO ₂ + Na₂O + K₂O + PbO) /(SiO₂ + Al₂O₃ + SnO)

A higher viscosity index (K) generally corresponds to lower slag viscosity and higher fluidity at a certain temperature. The calculation of the viscosity index based on the analytical results (Tables 2&5) shows high values for Xincun slag H98:3 and Baidian slag BH20:044. Table 7 shows that the central and outer areas of Xincun slag H98:3 have viscosity index values of 0.6 and 1, respectively, which explains the abundance of alloy prills in the centre of the slag. In contrast, the viscosity index of Baidian slag BH20:044 is as high as 1.7, indicating a significantly more efficient separation of metal from silicate slag compared to Xincun slag H98:3. Additionally, the viscosity index values for the other samples from both Xincun and Baidian foundry sites are less than 0.5, points to high viscosity of those slag samples.

Table 7

The viscosity index (K) results of the slags from the Xincun and Baidian foundry sites (data: C Wang )
Sample	Xincun foundry site			Baidian foundry site
	H91:4	H98:3		BH2:07	BH2:09	BH20:044	BH20:045
		Centre area	Outer area
Viscosity index (K)	0.2	0.6	1.0	0.2	0.4	1.7	0.3

The viscosity of the slag is also affected by smelting temperature. However, the complex chemical composition of the charge and disequilibrium conditions during the smelting process results in varied solidification temperatures for different slags or at different zones (Kresten 1986). We attempted to use the ternary state diagram (Fig. 9) to estimate the smelting temperature at these two sites. Data used in the figure are that of the high-silica region (slag matrix), except for Xincun H98:3 and Baidian slag BH20:044. However, since the liquidus temperature in the high-silica region is higher than what is typical for ancient smelting, incomplete melting may occur (SÁEZ et al. 2003). The high silica content in the slag contributes to its high viscosity, which hinders homogenisation of the mixture, thus explaining the inhomogeneity observed in the slag.

As shown in Fig. 9, Xincun slag H98:3 and Baidian slag BH20:044 showed relatively high smelting temperatures, but the presence of Na₂O, K₂O, Cu₂O and PbO in the slag may alter the estimates (Hauptmann 2007; Ströbele et al. 2010). Therefore, the smelting temperature depicted in Fig. 9 may fluctuate slightly. A large amount of malayaite present in slag H98:3 is often produced through the reaction between cassiterite and wollastonite (CaSiO₃) or calcium-rich pyroxene, which are frequently formed during the smelting process (Rovira et al. 2009). As the transformation of wollastonite to pseudowollastonite (Ca₃Si₃O₉) begins at roasting temperatures above 1100°C, wollastonite almost completely disappears when the temperature exceeds 1200°C, and is transformed to mainly pseudowollastonite and silica (Sheng et al. 2015). The calcium-rich phase may be a product of the transformation process, and the absence of wollastonite in this slag suggests a smelting temperatures of exceeding 1200°C.

Similarly, a large amount of calcium-rich pyroxene and high calcium content were observed in the matrix (41.7% CaO) of Baidian slag BH20:044, with a smelting temperature of up to 1500°C, which is even higher than that of Xincun slag H98:3. This suggests that the slag from both the Xincun and Baidian foundry sites were probably formed at extremely high temperatures. However, the formation of slag melts was reported to possibly up to about 1100–1300℃ in early metallurgy (Hauptmann 2007). Experimental tests have shown that higher temperatures inside the furnace indicate forced ventilation in the smelting system, and air input enhanced the combustion reactions, and led to temperature increases in the furnace (Rovira 1999). Therefore, it is doubtful that the Zhou dynasty bronze smelting process could have reached such high temperatures, and more research is needed to confirm or refute this hypothesis.

The atmosphere of the furnace is another important factor affecting the smelting process. A strong reducing atmosphere is necessary for the reduction of minerals in the ores to metals, without remnant unreacted ore components (except for quartz) (Berger et al. 2022). The analysis of the Xincun slag H91:4 suggested a weak reducing atmosphere during the smelting of the copper ores, resulting in incomplete ore reduction and leading to a heterogeneous microstructure with a mixture of malayaite and lead-bearing aluminate particles present (Fig. 4b) (Elin et al. 2010). The presence of cuprous oxide (Cu₂O) in the copper prill indicates an incomplete reduction of cupric oxide (CuO). Certainly, it is also possible that oxidation occurred during the final stage of smelting or cooling of the melt, causing some of the copper to oxidise to Cu₂O.

In the Xincun slag H98:3, the well-developed chain-like clusters from the recrystallisation of cassiterite, appeared to retain within the original mineral grains (Fig. 4c), indicating an inadequate reducing atmosphere during the smelting process (Rademakers et al. 2018). The presence of CaO and SiO₂ drives the formation of the stannate and stannous silicate phases under weak reducing atmospheres, therefore, the reduction of stannate and stannous silicates requires a stronger reducing atmosphere (Su 2014). Thus, the stannous-calcium-silicate phase in slag H98:3 may indicate a weak reduction. Additionally, oxidising atmosphere is evidenced by the presence of tin oxides around the metal prills (Rademakers et al. 2018), which could have entrapped in the slag and oxidised during cooling.

Iron oxidation or silicate phases are often used to infer smelting atmospheres due to the prevalence of iron-bearing phases in the slag and their variation in oxidation state (Stolarczyk 2013). Iron-bearing phases such as iron silicate (3FeO ▪SiO₂) and pyroxene, both containing Fe²⁺, were observed in the Baidian slag BH20:044 (Fig. 7c), indicating a strong reducing atmosphere during the smelting (Katarzyna et al. 2021). However, the presence of hematite, which contains Fe³⁺, in slag BH20:044 (Fig. 7e), indicating a weak reducing atmosphere. This, on the one hand, may be due to the heterogeneous feature of smelting using an open crucible that may have created weak reducing atmospheres in some areas (Rovira 2004). On the other hand, oxidation may have occurred during the cooling process, with some of iron silicates being oxidised, and precipitated to hematite (Hauptmann 2007; Berger et al. 2022).

In this paper, slags and technical ceramics from the Xincun and Baidian foundry sites of different historical context were studied to explore the smelting processes and bronze alloying technologies at both the sites. Comparison of the differences or/and similarities in bronze processing activities between the two sites was made. The results show that:

The Baidian foundry site appears to have more diverse alloying technologies than the Xincun foundry site. In the Xincun foundry site, it appears that only tin ores were used for alloying, whilst in the Baidian foundry site, the bronzes seem to have been produced using a variety of materials including tin metal, tin ore, and waste bronzes. But this observation may be biased by a small number of the samples from Xincun, and may not be representative of the site.

The smelting technologies at the two sites are similar in some aspects. It appears that calareous fluxes were used at both the sites in the smelting process, and the low viscosity index of some slag samples from both the Xincun and Baidian foundry sites resulted in metal prills present in the slag. This reflects a close relationship and influence in the use of metallurgical technology among different states of Central China during the Eastern Zhou period.

The raw copper used at the Xincun foundry site was probably oxide ores or ores with very little sulphide. The copper matte in slag from the Baidian foundry site was, however, too small both in size and amount to be confidently identified the type of ore and the smelting process.

The smelting temperatures at both the sites appear quite high, as shown in the SiO₂-FeO-CaO ternary phase diagram, which could have reached up to 1500°C at Baidian. However, whether such high temperatures were achieved during the Zhou dynasty smelting process needs to be further studied. Additionally, the reduction atmosphere at the Baidian foundry site appears stronger and more stable than at the Xincun foundry site, indicating a more advanced metallurgical process at Baidian. This reflects the strong effect of the political landscape on bronze production.

In summary, this study provides new data on archaeometallurgy of the Zhou dynasty in Central China, and shows the similarities and/or differences of smelting technologies and alloying technology between the Wei and Jin states during the Eastern Zhou period. This contributes to a better construction of the evolution sequence of smelting and casting technologies of bronze production in the pre-Qin period of Central China. Isotopic analysis is required in the future studies to trace the raw materials used in the bronze production from the Xincun and Baidian foundry sites, and to explore the changes in the source of raw materials used with time.

Acknowledgements

This research was supported by the National Natural Science Foundation of China (No. 52072220). The authors would like to thank Professor Rongyu Su of the Institute for the history of natural sciences, Chinese academy of sciences for providing the samples for this study and his support to this project.

Author contribution

Conceptualization: Cong Wang; methodology: Cong Wang; formal analysis and investigation: Cong Wang; writing-original draft preparation: Cong Wang; writing-review and editing: Cong Wang, Quanyu Wang, Zhenlong Gao; supervision: Cong Wang, Quanyu Wang.

Funding

This study was supported by the National Natural Science Foundation of China (No. 52072220)

Competing interests

The authors have no competing interests to declare that are relevant to the content of this article.

Annemie A (1996) Elemental composition and microstructure of Early Bronze Age and Medieval Tin Slags. Mikrochim Acta 124: 89-98. https://doi.org/10.1007/BF01244961
Bachmann H-G, Lutz C, Thiemann U (1989) Schlackenviskositäten. In: Hauptmann A, Pernicka E, Wagner GA (Hrsg ) (eds) Archäo-metallurgie in der Alten Welt – Old World archaeometallurgy. Der Anschnitt, Beiheft 7. Verlag Glückauf GmbH, Bochum, pp 137–140.
Bagley RW (2009) Anyang Mold-Making and The Decorated Model. Artibus Asiae69(01): 39-90.
Berger D, Brügmann G, Friedrich R et al (2022) Shiny bronze in glassy matter: an inconspicuous piece of slag from the Bronze Age mining site of Mušiston (Tajikistan) and its significance for the development of tin metallurgy in Central Asia. Archaeological and Anthropological Sciences(14): 150. https://doi.org/10.1007/s12520-022-01606-2
Chen K (2019) The archeological research on the Wei in the Western Zhou dynasty: Based on the sites and tombs of Wei region. PHD Thesis. Zhengzhou University, Zhengzhou.
Cui CP, Tang YY, Tian W et al (2021) A study on copper slag unearthed from Eastern Zhou Dynasty remains at the Xiwubi site in Jiangxian County, Shanxi Province. Journal of National Museum of China(08): 79-87.
Cui CP, Dai XM, Tian W et al (2022) Copper smelting technology in the southern Jin region during the Xia and early Shang Dynasties: the example of the Xiwubi site in Jiangxian County, Shanxi. Archaeology(07): 98-108.
Du N, Li JX, Zhang GM et al (2011) Investigation and study of iron smelting sites in the northeastern part of the old city of Qi, Linzi, Shandong. Journal of Jiangxi University of Science and Technology, Vol. 32(6): 12-15.
Dungworth D (2000) A note on the analysis of crucibles and moulds. Hist Metall 34(2): 83–86.
Eibner C (1982) Kupferbergbau in Österreichs Alpen. Prähistorische Archäologie in Südosteuropa (1): 399–408.
Elin F, Rui JCS, João CSM et al (2010) Smelting and recycling evidences from the Late Bronze Age habitat site of Baiões (Viseu, Portugal). Journal of Archaeological Science(37): 1623-1634. https://doi.org/10.1016/j.jas.2010.01.023
Gao ZL, Niu HB, Han CH (2020) 2014 Excavation of Western Zhou cemetery at Xidapo, Xincun site, Hebi, Henan pronvince. Huaxia Archaeology(03): 3-22.
Hauptmann A (2003) Rationales of Liquefaction and Metal Separation in Eariliest Copper Smelting; Basics for Reconstructing Chaalcolithic and Early Bronze Age Smelting Processes, Archaeometallurgy in Europe: Proceedings of the Internationg Conference, pp 459-468.
Hauptmann A (2007) The archaeometallurgy of copper: evidence from Faynan, Jordan. Springer, Berlin.
Huang J, Wei GF, Song GD et al (2011) Analyses on Remains of Smelting and Casting Copper at Xiaoshuangqiao Site. Nonferrous Metals63(01): 147-152.
Julia ML, Lgnacio MR, Pere CM et al (2020) Traditions and innovations: versatility of copper and tin bronze making recipes in Iron Age Emporion (L＇Escala, Spain). Archaeological and Anthropological Sciences(12): 124. https://doi.org/10.1007/s12520-020-01081-7
Katarzyna D, Mateusz S, Kamil N (2021) Reconstruction of copper smelting technology based of 18-20th-Century slag remains from the old copper Basin, Poland. Minerals(11): 926. https://doi.org/10.3390/min11090926
Kresten P (1986) Melting points and viscosities of ancient slags: a discussion. Hist. Metall(20): 43–45.
Li Q, Wei GF, He ZJ et al (2020) Reasearch on smelting process of Tongniujing mining and smelting site in Huaining county, Anhui province. Nonferrous Metals(07): 105-110.
Li XC (1993) A further study on the sources of ores for the central plains bronze vessels of the Shang-Zhou period. Studies in the History of Natural Science Vol. 12(03): 264-267.
Li YX, Hong YR (1995) Slagery of copper smelting in antiquity. Sciences of Conservation and Archaeology(01): 28-34.
Li YX, Xu H (2007) Preliminary study of the smelting and casting relics excavated from the Erlitou site. (Scienci for Archaeology, Vol. 2). Science Press, Beijing, pp:59-82.
Liang HG, Li YX, Sun SY et al (2005) Study of the arsenic-bearing slag excavated from the Yuanqu City site. Nonferrous Metals(04): 127-130.
Liang HG, Li YX, Sun SY et al (2009) Preliminary study of the metal particles and mineral composition within the furnace wall and slag excavated from the Yuanqu City site. Sciences of Conservation and Archaeology21(04): 18-30.
Liang HG (2019) Reflections on some special phenomena in the metallurgical relics excavated from the Erlitou site. Cultural Relics in Southern China(05): 60-67.
Liu SR, He XL, Chen JL et al (2020) Micro-slag and “invisible” copper processing activities at a Middle-Shang period (14th-13th century BC) bronze casting workshop. Journal of Archaeological Science(122): 105-222. https://doi.org/10.1016/j.jas.2020.105222
Mueller R, Rehren T, Rovira S (2004) Almizaraque and the early copper metallurgy of southeast Spain: new data. Madrider Mitteilungen(45): 33–56. https://discovery.ucl.ac.uk/id/eprint/93613/
Nickel L (2006) Imperfect Symmetry: Re-Thinking Bronze Casting Technology in Ancient China. Artibus Asiae66(01): 5-39.
Pollard AM, Rawson J, Liu R (2018) Some Recently Rediscovered Analyses of Chinese Bronze from Oxford. Archaeometry60(01): 118-127. https://doi.org/10.1111/arcm.12346
Qin Z, Chen JL, Zhang H (2016) Research on the steel production system of the Warring States Qinhan Iron Smelting Site Group in Wugang and Xiping areas of Henan. Cultural Relics of Central China(01): 109-117.
Rademakers FW, Rehren T (2016) Seeing the forest for the trees: assessing technological variability in ancient metallurgical crucible assemblages. Journal of Archaeological Science(7): 588–596. https://doi.org/10.1016/j.jasrep.2015.08.013
Rademakers FW, Farci C (2018) Reconstructing bronze production technology from ancient crucible slag: experimental perspectives on tin oxide identification. Journal of Archaeological Science: Reports(18): 343-355. https://doi.org/10.1016/j.jasrep.2018.01.020
Renzi M, Rovira S (2016) Metallurgical vessels from the Phoenician site of La Fonteta (Alicante, Spain): a typological and analytical study. In: Körlin G, Prange M, Stöllner T, Yalçın Ü (eds) From on the occasion of 40 years research in archaeometallurgy and archaeometry. Der Anschnitt, Beiheft 29. Verlag Maria Leidorf, Rahden/Westf., pp 143–166.
Rovira S (1999) Una propuesta metodológica para el estudio de la metalurgia prehistórica: el caso de Gorny en la región de Kargaly (Orenburg, Rusia). Trabajos de Prehistoria(56): 85-113. https://doi.org/10.3989/tp.1999.v56.i2.277
Rovira S (2004) Tecnología metalúrgica y cambio cultural en la Prehistoria de la Península Ibérica. Norba Revista de Historia (17): 9–40.
Rovira S (2007) La producción de bronces en la Prehistoria. In: Molerai Marimon J, Farjasi Silva J, Rourai Grabulosa P, Pradelli Cara T (eds) Avances en Arqueometría 2005. Actas VI Congreso Ibérico de Arqueometría. Universitat de Girona, Girona, pp 21–35.
Rovira S, Ruiz IM, Renzi M (2009) Experimental co-smelting to copper-tin alloys. Metals and societies: 407-413.
SÁEZ R, Nocete F, Nieto MJ (2003) The extractive metallurgy of copper from CABEZO JURÉ, Huelva, Spain: chemical and mineralogical study of slags date to the third millenium B.C. The Canadian Mineralogist(41): 627-638. https://doi.org/10.2113/gscanmin.41.3.627
Shanxi Provincial Institute of Archaeology (2012) Houma baidian zhutong yizhi (Bronze foundry at Baidian, Houma). Science Press, Beijing, pp:1-5.
Sheng HY, Lv L, Liang B et al (2015) Effect of calcination temperature on mineralization of CO₂ with calcium silicate. Multipurpose Utilization of Mineral Resources(05): 76-80.
Sloley RW, Lucas A (1948) Ancient Egyptian Materials and Industries. The Journal of Egyptian Archaeology(34): 125. https://doi.org/10.2307/3855372
Stolarczyk T (2013) Wyniki Bada ´n Reliktów Dawnego Górnictwa Miedzi na Terenie Pogórza Kaczawskiego w Roku 2012; Zagozdzon, P.P., Madziarz, M., Eds.; Dzieje Górnictwa—Element Europejskiego Dziedzictwa Kultury: Wrocław, Polska, Vol.5, 339–351. (In Polish).
Ströbele F, Wenzel T, Kronz A et al (2010) Mineralogical and geochemical characterization of high-medieval lead-silver smelting slags from Wiesloch near Heidelberg (Germany). Archaeological Anthropological Science(2): 191–215. https://doi.org/10.1007/s12520-010-0039-7
Su RY, Hua JM, Li KM et al (1995) The metal technology of early ancient China. Shandong Science and Technology Press, Jinan, pp:38.
Su ZJ (2014) Research on the behavior of tin removal from tin-bearing iron ores by reduction roasting. MA Thesis, Central South University, Changsha.
Xue Y, Li GW, Yang GM (2017) Mineralogy and crystallography of stokesite from inner Monglia China. The Canadian Mineralogist(55): 63-74.
Yu LW (2010) Civil Engineering. Soutueast University Press, Nanjing, pp:42.
Yu TY, Jin ZY, Tan JZ et al (2021) Scientific and technological analysis reveals the metallurgical mining economic model in the historical period of Southwest Henan and the enlightenment to the archaeology of metal resources in the Bronze Age. Cultural Relics in Southern China(03): 153-161.
Zhang J, Gao XP, Ding SC et al (2021) The preliminary research on copper casting technology and metal resources changes of Guanzhuang Site, Xingyang, Henan Pronvince. Cultural Relics in Southern China(03): 162-173.
Zhu YF (2021) Study on ore blending technology of ancient copper mines in the middle and the lower reaches of the Yangtze River. MA Thesis, University of Science and Technology of China, Anhui.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Bronze smelting activities during the Zhou dynasty: Analysis of slags from the Xincun and Baidian foundry sites in Central China

Status:

Version 1

Abstract

Figures

1. Introduction

1.1. Archaeological background

2. Materials and analytical methods

2.1. Materials

2.2. Analytical methods

3. Results

3.1. Xincun samples

3.2. Baidian samples

4. Discussion

4.1. Alloying technology as reflected in the slags from the Xincun and Baidian foundry sites

4.2. Possible use of flux in the slags of the Xincun and Baidian foundry sites

4.3. Smelting conditions in the Xincun and Baidian foundry sites

5. Conclusions

Declarations

Acknowledgements

Author contribution

Funding

Competing interests

References

Additional Declarations

Status:

Version 1