- Email: [email protected]
Study on the property of the production for Fengdongyan kiln in Early Ming dynasty by INAA and EDXRF
Nuclear Instruments and Methods in Physics Research B 381 (2016) 52–57
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research B journal homepage: www.elsevier.com/locate/nimb
Study on the property of the production for Fengdongyan kiln in Early Ming dynasty by INAA and EDXRF L. Li a, Y. Huang a,b, H.Y. Sun a,b, L.T. Yan a, S.L. Feng a, Q. Xu a, X.Q. Feng a,⇑ a b
Institute of High Energy Physics, Chinese Academy of Sciences, 19 Yu Quan Lu, Beijing 100049, China University of Chinese Academy of Sciences, Chinese Academy of Sciences, Beijing 100049, China
a r t i c l e
i n f o
Article history: Received 22 March 2016 Received in revised form 20 May 2016 Accepted 26 May 2016 Available online 2 June 2016 Keywords: Official kiln Imperial porcelain Civilian porcelain Fengdongyan kiln INAA EDXRF
a b s t r a c t A lot of official wares carved ‘‘Guan” or the dragon patterns were excavated on the strata of Ming dynasty of the Fengdongyan kiln site at Dayao County. The imperial porcelain was fired in Hongwu and Yongle eras. However, the emergence of this imperial porcelain has triggered academic debate about the property of Fengdongyan kiln in the Early Ming dynasty. Based on the differences of the official kiln management, some scholars have determined that the property of the production for this kiln was the civilian kiln. According to the historical textural records and typology, others preliminary confirmed that Fengdongyan kiln was the official kiln. In this paper, the elemental compositions of body and glaze in imperial and civilian porcelain are study by INAA and EDXRF for determining the property of the production for this kiln in Early Ming dynasty. After the processing of experimental data by geochemical analysis and principal component analysis, the result show that the raw materials for making body and glaze in imperial porcelain are similar with those of the civilian porcelain and the degrees of elutriation for body can be slightly different in HW-M period of Ming dynasty. The analytical results support the view that the Fengdongyan kiln is civilian not official. Ó 2016 Elsevier B.V. All rights reserved.
1. Introduction Fengdongyan kiln, an important ancient kiln of Longquan, located in the valley 1.5 km from the north of Dayao County, in Longquan city (see Fig. 1). From September 2006 to January 2007, the Fengdongyan kiln was excavated by the Zhejiang Province Institute of Cultural Relics and Archaeology, the School of Archaeology and Museology in Peking University and the Longquan Museum. An area of over 1600 square metres was explored and a mass of exquisite Longquan celadon shards in Yuan and Ming dynasty were unearthed. In particular, abundant civilian porcelain and a lot of official wares carved ‘‘Guan” or the dragon patterns in the Hongwu (1368–1398 AD) and Yongle (1403–1424 AD) eras of the Ming dynasty were excavated on the strata of Ming dynasty [1]. In archaeology, due to the lack of specimens, the question about the existence of an official kiln in Longquan has only been modestly investigated. However, the emergence of the imperial porcelain caused academic debate about the property of Fengdongyan kiln in the Early Ming dynasty. At present, there are two main ⇑ Corresponding author. E-mail address: [email protected] (X.Q. Feng). http://dx.doi.org/10.1016/j.nimb.2016.05.027 0168-583X/Ó 2016 Elsevier B.V. All rights reserved.
views about Fengdongyan kiln. Some believe that it is the civilian kiln, while others think that it is the imperial kiln. ‘‘Longquan Kiln archaeological excavations expert meetings and press conference” was held in Longquan on January 18–19, 2007 [2]. Combined with the definition of official kiln, most experts believed that Fengdongyan kiln should belong to the civilian kiln, and the property of the production for this kiln was following government orders and producing the imperial porcelain. But according to ancient literature of the imperial porcelain [3], a large number of imperial porcelain was fired in Longquan kiln from Hongwu to Tianshun periods of the Ming dynasty. In addition, Longquan kiln and Jingdezhen imperial kiln have always been mentioned together. So some scholars believe that Fengdongyan kiln is similar to the official kiln of Jingdezhen. Therefore, the identification of the Fengdongyan kiln is a problem that cannot be answered by only typological considerations. It is well known that the contents of major, minor and trace elements of the porcelain body and glaze depend on its raw material and manufacturing technology, which can be used to indicate the age of the porcelain and its provenance [4–6]. Different contents mean different recipes of body and glaze. In the ancient, due to strict hierarchy, the porcelain fired at the official kiln was used for the emperor. So the recipes and the dispersion degree of body or glaze
L. Li et al. / Nuclear Instruments and Methods in Physics Research B 381 (2016) 52–57
53
Fig. 1. The overall view of Fengdongyan kiln.
in the official kiln were different from those of civilian kiln. The civilian and official productions of Ru Celadon have been researched focusing on their different techniques and glaze formulae [7]. The results mean that Ru official celadon and Ru folk porcelain were made from almost the same body materials but each has its own different glaze formula. So according to studying on the variation of raw materials and techniques between the civilian and imperial porcelain, the property of Fengdongyan kiln is discussed in our work. In this paper, the chemical composition of the civilian and imperial Longquan celadon porcelain of Fengdongyan kiln were explored by use of Instrumental Neutron Activation Analysis (INAA) and Energy Dispersive X-ray Fluorescence (EDXRF). INAA is a promising method of elemental analysis. It has been broadly used in the archaeological field since E.V. Sayre’s work [8]. The
EDXRF technique is a highly sensitive, non-destructive, multielement analysis method with good repeatability [9]. The two techniques have been widely used to characterise the ingredients of ancient ceramics, glass, biomedicine and so on. From a statistical analysis of the elemental data, the raw material and the development of porcelain firing technology are discussed. The information is used to display the inheritance relationship of the raw materials used for the body and glaze in the civilian and imperial porcelain. 2. Samples and experimental 2.1. Samples One hundred and sixty-four shards excavated at Fengdongyan kiln were selected as specimens. According to the typology, they
54
L. Li et al. / Nuclear Instruments and Methods in Physics Research B 381 (2016) 52–57
2.2. INAA experiment
Table 1 Detail information on the celadon shards in Fengdongyan kiln site. Name in the paper
Date
Amount
Type
HW-M
25 46
E-M
Hongwu era of Ming dynasty (1368–1398 AD) Yongle era of Ming dynasty (1403–1424 AD) Early Ming dynasty
M-M
Middle Ming dynasty
50
Imperial porcelain Imperial porcelain Civilian porcelain Civilian porcelain
YL-M
43
were identified as the products of imperial and civilian porcelain, and they were typologically classified into four cultural periods (see Table 1, Figs. 2 and 3). The first group is referred to as HW-M in this paper. It was fired during the Hongwu period (1368–1398 AD) of Ming dynasty. The second group is referred to as YL-M in this paper and was fired during the Yongle (1403–1424 AD) of Ming dynasty. These two groups belong to imperial porcelain concentrated in the southwest of Fengdongyan kiln in Ming dynasty. The third and fourth groups are referred to as E-M and M-M in this paper, and they were fired during the Early Ming dynasty and during the Middle Ming dynasty, respectively. The third and fourth groups were the civilian porcelain excavated at the south of Fengdongyan kiln.
For one group, a sample with 30 mm 10 mm was cut from each shard. After abrading away the glaze, the pure body was washed in an ultrasonic cleaner with tap water and then three times in deionised water, and then dried at 105 °C for 8 h. After those steps, the pure body was ground into a powder of 74 lm size in an agate mortar. About 200 mg of each powder sample were placed into the miniature neutron source reactor of the Chinese Institute of Atomic Energy simultaneously with quality control standards. The quality control standard is the Chinese national certified reference material of soil (GBW07402). The radioactivity of the sample was measured by an HPGe detector connected to a multi-channel analyser. Under these conditions, the contents of 26 elements (V, Mn, Dy, Ga, Ho, Sb, La, Sm, Yb, W, U, Sc, Cr, Co, Zn, Rb, Cs, Ba, Ce, Nd, Eu, Tb, Lu, Hf, Ta and Th) were determined. The validation of the analytical method applied in this work was carried out by analysis of GBW07402. The results showed that the experimental values were well in agreement with the certified ones (see Table 2). 2.3. EDXRF experiment For another group, samples with 30 mm 10 mm were cut from these sherds. After the cross-section was polished, it was washed and dried in the same procedure as in INAA. The EDXRF experiments were performed on an EDAX Eagle III spectrometer
Fig. 2. The appearances of imperial porcelain shards excavated from Fengdongyan kiln.
Fig. 3. The appearances of civilian porcelain shards excavated from Fengdongyan kiln.
55
L. Li et al. / Nuclear Instruments and Methods in Physics Research B 381 (2016) 52–57 Table 2 Contents of 32 elements in the certified reference material GBW07402 by INAA. Element
Experiment value
Reference value
Element
Experiment value
Reference value
V/ppm Mn/ ppm Dy/ppm Ga/ppm Ho/ppm Sb/ppm La/ppm Sm/ ppm Yb/ppm W/ppm U/ppm Sc/ppm Cr/ppm
59.4 501
62.0 510
Co/ppm Zn/ppm
8.07 41
8.70 42
4.21 10.8 0.90 1.48 163 17.4
4.40 12.0 0.93 1.30 164 18.0
Rb/ppm Cs/ppm Ba/ppm Ce/ppm Nd/ppm Eu/ppm
84.9 4.52 900 399 197 2.8
88.0 4.90 930 402 210 3.0
2.06 1.29 1.51 10.3 51.1
2.00 1.08 1.40 10.7 47.0
Tb/ppm Lu/ppm Hf/ppm Ta/ppm Th/ppm
0.87 0.31 6.22 0.62 16.5
0.97 0.32 5.8 0.78 16.6
of Institute of High Energy Physics, CAS, Beijing of China. The spectrometer with a Mo tube and a 125 lm Be window has an incident beam angle of 65° and an emergence angle of 60°. The detector is a liquid-nitrogen-cooled Si (Li) crystal with a 160.3 eV at Mn Ka. There is a vacuum in the chamber and the diameter of X-ray beam spot was selected to be 1 mm to fit the body thickness dimensions. The voltage and current of X-ray tube are operated at 40 kV and 250 lA, respectively. The software employed for spectrum deconvolution and analysis is the program VISION32. The elemental abundances of Na2O, MgO, Al2O3, SiO2, K2O, CaO, TiO2, MnO, Fe2O3, Cu, Zn, Rb and Sr were quantified by fundamental parameter (FP) method. In order to get a better measure of the precision and accuracy of the data in this work, ancient ceramic samples of known composition were selected to compare the expected and observed elemental concentrations measured on the ceramic reference samples used in the analysis. The results show that the experimental values are well in agreement with the certified ones (see Table 5). The average values of each elemental composition in porcelain body and glaze are displayed in Tables 6 and 7, respectively. The data of Na2O and MgO are provided as suggestive only because of the poor fluorescence yields and low counting statistics obtained for characteristic X-ray radiation.
Table 4 REE parameter of ancient porcelain body.
HW-M YL-M E-M M-M
La/Sm
Tb/Yb
La/Yb
dEu
3.99 3.87 3.18 3.86
1.15 1.21 1.18 1.18
7.86 7.99 6.21 7.53
0.31 0.29 0.27 0.29
(9), the YL-M group (9), the E-M group (9) and the M-M group (9). The average values of each element in the bodies of the imperial and civilian porcelain in four cultural periods are shown in Table 3. It can be seen that the average contents of other trace elements were close to each other in HW-M, YL-M, E-M and M-M groups except for V and Ba. The scatter plots of V and Ba contents are exhibited in Fig. 5. All samples are located in two plot regions. Samples of HW-M group are concentrated in the upper right region while those of the YLM and E-M groups are distributed in the lower left region, and a few data plots of the M-M group overlap with those of the HWM groups. It means that the raw material of the imperial porcelain body in Hongwu period is similar with that of the civilian porcelain body in Middle of Ming dynasty, while the same raw material of the body are used in imperial porcelain in Yongle period and civilian porcelain. Rare earth elements (REEs) are located in IIIB column of the periodic table of and their atomic numbers range from 57 to 71. Their crystal chemical and geochemical properties are similar because their atomic and ionic radiuses are close. Due to relative stability, REE can be used as a tracer source in the archaeological research [10,11]. The patterns and some parameters of REEs in the porcelain bodies of HW-M, YL-M, E-M, and M-M groups are shown in Fig. 4 and Table 4. In Fig. 4, the REEs patterns of imperial porcelain (HW-M and YL-M groups) match those of the civilian porcelain (E-M and M-M groups) very well. In Table 4, the variation range of (La/Sm), (Tb/Yb), (La/Yb), and dEu of the four groups are 3.18–3.99, 1.15–1.21, 6.21–7.99 and 0.27–0.31, respectively. They are similar to each other. So it can be inferred that the kaolin for imperial porcelain body are the same as that of the civilian porcelain body.
3.2. The multivariate statistical analysis of EDXRF data for imperial and civilian porcelain
3. Results and discussion 3.1. The geochemical interpretation of INAA data for imperial and civilian porcelain Due to the expensive cost and long measuring time, only 36 shards were selected as specimens for INAA: The HW-M group
The average values of each element in the bodies of imperial and civilian porcelain are shown in Table 6. It can be seen that the K2O content of the porcelain body in HW-M group is similar to those of E-M group and lower than those of M-M group, while the values of K2O in YL-M group are lower than those of other three
Table 3 The average values of 32 elements in body by INAA. Elements
HW-M
YL-M
E-M
M-M
Elements
HW-M
YL-M
E-M
M-M
V/ppm Mn/ppm Dy/ppm Ga/ppm Ho/ppm Sb/ppm La/ppm Sm/ppm Yb/ppm W/ppm U/ppm Sc/ppm Cr/ppm
13 ± 2 490 ± 35 8±1 24 ± 2 1.3 ± 0.1 0.18 ± 0.03 58 ± 4 9.1 ± 0.3 4.4 ± 0.2 3.9 ± 0.7 4.4 ± 0.2 6.1 ± 0.3 3.3 ± 1.0
6±1 478 ± 60 10 ± 1 22 ± 2 1.5 ± 0.1 0.21 ± 0.04 66 ± 8 10.6 ± 0.5 4.9 ± 0.2 2.5 ± 0.4 4.3 ± 0.1 5.9 ± 0.5 3.5 ± 0.5
5±2 471 ± 110 10 ± 1 23 ± 2 1.6 ± 0.1 0.26 ± 0.04 53 ± 17 10.2 ± 1.3 5.1 ± 0.4 2.2 ± 0.4 4.3 ± 0.3 5.5 ± 0.8 3.0 ± 0.3
7±3 463 ± 49 8±1 23 ± 2 1.4 ± 0.1 0.25 ± 0.07 56 ± 10 9.0 ± 0.7 4.5 ± 0.4 2.7 ± 0.3 4.6 ± 0.3 5.6 ± 0.7 3.7 ± 1.6
Co/ppm Zn/ppm Rb/ppm Cs/ppm Ba/ppm Ce/ppm Nd/ppm Eu/ppm Tb/ppm Lu/ppm Hf/ppm Ta/ppm Th/ppm
1.7 ± 0.2 84 ± 10 319 ± 17 6.6 ± 0.3 514 ± 76 103 ± 5 54 ± 3 1.0 ± 0.1 1.3 ± 0.1 0.6 ± 0.1 6.2 ± 0.3 2.3 ± 0.1 38 ± 1
1.4 ± 0.2 93 ± 25 278 ± 8 5.6 ± 0.4 327 ± 34 103 ± 11 62 ± 5 1.0 ± 0.1 1.6 ± 0.1 0.7 ± 0.1 6.3 ± 0.2 2.3 ± 0.2 39 ± 1
1.2 ± 0.3 83 ± 21 296 ± 34 5.4 ± 0.4 276 ± 58 89 ± 21 56 ± 9 0.9 ± 0.2 1.6 ± 0.2 0.7 ± 0.1 6.5 ± 0.2 2.5 ± 0.4 42 ± 3
1.1 ± 0.4 74 ± 19 307 ± 17 5.9 ± 1.3 435 ± 92 102 ± 14 52 ± 5 0.9 ± 0.1 1.4 ± 0.1 0.6 ± 0.1 6.2 ± 0.3 2.3 ± 0.3 38 ± 2
56
L. Li et al. / Nuclear Instruments and Methods in Physics Research B 381 (2016) 52–57
Table 5 The quantitative results for ancient ceramic sample by EDXRF.
Experimental value The certified value
Na2O %
MgO %
Al2O3 %
SiO2 %
K2O %
CaO %
TiO2 %
MnO %
Fe2O3 %
Cu ppm
Zn ppm
Rb ppm
Sr ppm
0.65 0.44
0.62 0.70
24.34 23.90
67.54 67.50
2.27 2.30
0.54 0.62
0.88 0.95
0.022 0.026
2.77 2.70
46 27
54 59
119 113
109 103
Table 6 The average values of 13 elemental compositions in body by EDXRF. Elements
HW-M
YL-M
E-M
M-M
Elements
HW-M
YL-M
E-M
M-M
Na2O/% MgO/% Al2O3/% SiO2/% K2O/% CaO/% TiO2/%
0.59 ± 0.17 0.43 ± 0.06 20.4 ± 0.5 70.7 ± 0.7 5.32 ± 0.34 0.056 ± 0.009 0.23 ± 0.04
0.77 ± 0.18 0.38 ± 0.07 20.5 ± 0.5 70.9 ± 0.8 5.23 ± 0.42 0.077 ± 0.016 0.15 ± 0.03
0.71 ± 0.18 0.39 ± 0.05 20.3 ± 0.8 70.9 ± 1.0 5.35 ± 0.46 0.082 ± 0.014 0.14 ± 0.03
0.72 ± 0.20 0.38 ± 0.06 19.6 ± 0.7 70.9 ± 0.8 6.13 ± 0.30 0.093 ± 0.017 0.16 ± 0.04
MnO/% Fe2O3/% Cu/ppm Zn/ppm Rb/ppm Sr/ppm
0.05 ± 0.01 2.11 ± 0.12 24 ± 2 44 ± 4 326 ± 23 40 ± 5
0.05 ± 0.01 1.79 ± 0.11 24 ± 2 45 ± 8 304 ± 21 30 ± 5
0.05 ± 0.01 1.87 ± 0.14 24 ± 3 46 ± 8 319 ± 24 31 ± 5
0.06 ± 0.01 1.86 ± 0.13 26 ± 3 41 ± 6 340 ± 18 36 ± 7
Table 7 The average values of 13 elemental compositions in glaze EDXRF. Elements
HW-M
YL-M
E-M
M-M
Elements
HW-M
YL-M
E-M
M-M
Na2O/% MgO/% Al2O3/% SiO2/% K2O/% CaO/% TiO2/%
0.64 ± 0.18 0.68 ± 0.10 12.8 ± 0.5 70.0 ± 1.7 5.84 ± 0.48 7.06 ± 1.73 0.21 ± 0.03
0.68 ± 0.16 0.67 ± 0.11 13.1 ± 0.7 70.7 ± 1.6 5.81 ± 0.32 6.24 ± 1.22 0.18 ± 0.03
0.64 ± 0.16 0.59 ± 0.15 12.8 ± 0.8 70.4 ± 2.1 6.11 ± 0.63 6.53 ± 1.95 0.19 ± 0.03
0.66 ± 0.21 0.69 ± 0.18 12.7 ± 0.5 70.1 ± 1.6 6.28 ± 0.32 6.81 ± 1.25 0.20 ± 0.03
MnO/% Fe2O3/% Cu/ppm Zn/ppm Rb/ppm Sr/ppm
0.33 ± 0.08 2.17 ± 0.13 31 ± 4 60 ± 16 269 ± 16 467 ± 97
0.34 ± 0.05 2.12 ± 0.23 31 ± 4 54 ± 17 268 ± 13 424 ± 56
0.33 ± 0.08 2.14 ± 0.30 30 ± 4 48 ± 20 280 ± 26 477 ± 131
0.38 ± 0.10 1.91 ± 0.35 31 ± 4 52 ± 25 283 ± 23 456 ± 73
cultural periods. The average value of CaO in the imperial porcelain is slightly lower than those of civilian porcelain. The average content of TiO2 in the HW-M group is (0.23 ± 0.04)% higher than that (0.14–0.16)% in other three groups. The abundance change of Fe2O3 in the imperial porcelain body is similar to that of TiO2. The composition of Fe2O3 in the porcelain clay in the Longquan district is less than that of the imperial porcelain body, so some other clay material, probably a Zijin soil was added in the bodies of the imperial and civilian porcelain. It can not only improve the body strength, but also increase the colour of black or grey in porcelain body. A detailed discussion of this has been published elsewhere [12]. The average concentrations of SiO2, Al2O3, MnO and the trace elements Cu, Zn, Rb, Sr are close to each other in the imperial and civilian porcelain body, as shown in Table 6.
Fig. 4. The chondrite-normalised REE pattern.
In this work, principal components analysis (PCA) was used to study the raw material used for the body in the imperial and civilian porcelain. The main objective of PCA is to reduce the dimensionality of the observations [13]. A factorial analysis diagram from the composition of bodies made in the imperial and civilian porcelain is shown in Fig. 6. The data for K2O, CaO, TiO2, Fe2O3 and Rb are used and the eigenvalue sum of Factors 1 and 2 accounts for 73.92% of the total variance. Factor score 1 (F1) mainly includes the changes of TiO2 and Fe2O3 and Factor score 2 (F2) mainly represents the content of K2O and Rb. Some samples in the HW-M group are located in the lower area and the others overlap with the three groups in Fig. 6. The reason for this difference could be different elutriation degrees of the same raw materials or the different raw materials used. Combining with the major elements (Table 6) and the trace elements (Table 3), the most of element contents are similar in the four groups. So we conclude that different elutriation degree
Fig. 5. Scatterplot of V and Ba in bodies of imperial and civilian porcelain in Ming dynasty.
L. Li et al. / Nuclear Instruments and Methods in Physics Research B 381 (2016) 52–57
57
tration of TiO2, Fe2O3, V and Ba in HW-M group. It means that the raw materials for making body in imperial porcelain are similar with those of the civilian porcelain and the degrees of elutriation can be slightly different in HW-M group. The average contents of the glaze in major and trace elements are similar, which displays that the raw materials of the glaze are close to each other between imperial and civilian porcelain. Based on the result of composition in body and glaze, Fengdongyan kiln in Early Ming dynasty should be the civilian kiln. 4. Conclusion
Fig. 6. PCA for the contents of elemental composition in body of imperial and civilian with EDXRF.
In this paper, INAA and EDXRF have been applied to analyse the elemental composition of imperial and civilian porcelain at the Fengdongyan kiln. Comparing the elemental composition of the imperial porcelain with those of civilian porcelain, the contents of TiO2, Fe2O3, V and Ba in Hongwu period are slightly different. After eliminating the influence of the change in manufacturing technology between different periods by the analysis of REEs, we could conclude that the mineral sources of producing kaolin for imperial porcelain body are the same to that of civilian porcelain body. According to the PCA analytical results, it means that the raw materials for making body and glaze in imperial porcelain are similar with those of the civilian porcelain and the degrees of elutriation for the body can be slightly different in HW-M period of Ming dynasty. Those analytical results support the interpretation put forth by some archaeologists, i.e. that the Fengdongyan kiln is civilian not official. Acknowledgements The authors greatly appreciate the Zhejiang Province Institute of Cultural Relics and Archaeology, which provided ancient Chinese celadon excavated from the Fengdongyan site of the Longquan kiln. This work was financially supported by the National Natural Science Foundation of China (11575207 and 10875137).
Fig. 7. PCA for the contents of elemental composition in glaze of imperial and civilian with EDXRF.
of the same raw materials is the main reason between the HW-M group and the others. The raw material of making ceramic was purified by the elutriation, and the impurity was removed, especially iron impurities. So the chemical composition of body and glaze can be changed as different degrees of elutriation. From the above, the raw materials of the HW-M group have the similarity with the civilian porcelain. In Fig. 6, samples of the YL-M group overlap with those of E-M and M-M, which reveal that the raw materials for making the porcelain body in YL-M group are the same as the civilian porcelain. The elemental compositions in the porcelain glaze of the four groups are shown in Table 7. There were no evident differences of average concentration of major and trace elements between each other. The similar conclusion can also be get from Fig. 7. The contents of Al2O3, SiO2, K2O, CaO, TiO2, MnO, Fe2O3, Rb and Sr are used and the eigenvalue sum of Factors 1 and 2 accounts for 58.93% of the total variance. Fig. 7 displays that the sample plots in the four groups are scattered in a large region and the distribution of data points in the HW-M and YL-M group are relatively concentrated. It means that the raw materials of glaze in the four groups are close to each other and the similar processing for making glaze is used between imperial porcelain and civilian porcelain. In view of analysis results of INAA and EDXRF, there are no obvious differences of contents of the most major and trace elements between imperial and civilian porcelain body except the concen-
References [1] Zhejiang Institute of Cultural Relics et al., Porcelain Excavated from Fengdongyan Kiln Site at Dayao Country of Longquan City, Cultural of Relics Publishing House, Beijing, 2009, p. 1. [2] Y.M. Shen, D.S. Qin, W.B. Shi, The academic forum summary of archaeological excavation at Fengdonggyan site of Longquan kiln, Cultural Relics 5 (2007) 93– 96. [3] H. Huo, J.H. Chen, M.H. Hao, et al., On Longquan celadon fragments form the accumulation pit of Hexia site in Chuzhou District, Huai ’an City of Jiangsu, Southeast Culture 2 (2010) 38–45. [4] D. Zhu et al., PIXE study on the provenance of Chinese ancient porcelain, Nucl. Instrum. Methods B 249 (2006) 633–637. [5] B. Ma, L. Liu, S.L. Feng, et al., Analysis of the elemental composition of Tang Sancai from the four major kilns in China using EDXRF, Nucl. Instrum. Methods B 319 (2014) 95–99. [6] G.X. Xie, S.L. Feng, X.Q. Feng, et al., The dating of ancient Chinese celadon by INAA and pattern recognition methods, Archaeometry 51 (4) (2009) 682–699. [7] Y.Z. Ding, H. Li, X.M. Sun, et al., Ru celadon ware production in Qingliangsi kiln for civilian and official uses focusing on their different techniques and glaze formulae, Palace Mus. J. 3 (2013) 62–73. [8] E.V. Sayre, R.W. Dodson, Neutron activation study of Mediterranean potsherds, Am. J. Archaeol. 61 (1975) 35–41. [9] D.N. Papadopoulou et al., Comparison of a portable micro-X-ray fluorescence spectrometry with inductively coupled plasma atomic emission spectrometry for the ancient ceramics analysis, Spectrochim. Acta B 59 (2004) 1877–1884. [10] X.Q. Jia, J.Q. Dong, S. Han, et al., INAA analysis for Jun porcelain and Modern Chinese Jun porcelain, J. Isot. 17 (3) (2004) 129. [11] I.L. Yong, Provenance derived from the geochemistry of late Paleozoic early Mesozoic mudrocks of the Pyeongan Supergroup, Korea, Sediment. Geol. 149 (4) (2002) 219–235. [12] L. Li, L.T. Yan, S.L. Feng, et al., Elemental characterization by EDXRF of imperial Longquan celadon porcelain excavated from Fengdongyan kiln, Dayao County, Archaeometry 57 (6) (2015) 966–976. [13] W. Härdle, L. Simar, Applied Multivariate Statistical Analysis, Springer, Berlin, 2003, p. 323.
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research B journal homepage: www.elsevier.com/locate/nimb
Study on the property of the production for Fengdongyan kiln in Early Ming dynasty by INAA and EDXRF L. Li a, Y. Huang a,b, H.Y. Sun a,b, L.T. Yan a, S.L. Feng a, Q. Xu a, X.Q. Feng a,⇑ a b
Institute of High Energy Physics, Chinese Academy of Sciences, 19 Yu Quan Lu, Beijing 100049, China University of Chinese Academy of Sciences, Chinese Academy of Sciences, Beijing 100049, China
a r t i c l e
i n f o
Article history: Received 22 March 2016 Received in revised form 20 May 2016 Accepted 26 May 2016 Available online 2 June 2016 Keywords: Official kiln Imperial porcelain Civilian porcelain Fengdongyan kiln INAA EDXRF
a b s t r a c t A lot of official wares carved ‘‘Guan” or the dragon patterns were excavated on the strata of Ming dynasty of the Fengdongyan kiln site at Dayao County. The imperial porcelain was fired in Hongwu and Yongle eras. However, the emergence of this imperial porcelain has triggered academic debate about the property of Fengdongyan kiln in the Early Ming dynasty. Based on the differences of the official kiln management, some scholars have determined that the property of the production for this kiln was the civilian kiln. According to the historical textural records and typology, others preliminary confirmed that Fengdongyan kiln was the official kiln. In this paper, the elemental compositions of body and glaze in imperial and civilian porcelain are study by INAA and EDXRF for determining the property of the production for this kiln in Early Ming dynasty. After the processing of experimental data by geochemical analysis and principal component analysis, the result show that the raw materials for making body and glaze in imperial porcelain are similar with those of the civilian porcelain and the degrees of elutriation for body can be slightly different in HW-M period of Ming dynasty. The analytical results support the view that the Fengdongyan kiln is civilian not official. Ó 2016 Elsevier B.V. All rights reserved.
1. Introduction Fengdongyan kiln, an important ancient kiln of Longquan, located in the valley 1.5 km from the north of Dayao County, in Longquan city (see Fig. 1). From September 2006 to January 2007, the Fengdongyan kiln was excavated by the Zhejiang Province Institute of Cultural Relics and Archaeology, the School of Archaeology and Museology in Peking University and the Longquan Museum. An area of over 1600 square metres was explored and a mass of exquisite Longquan celadon shards in Yuan and Ming dynasty were unearthed. In particular, abundant civilian porcelain and a lot of official wares carved ‘‘Guan” or the dragon patterns in the Hongwu (1368–1398 AD) and Yongle (1403–1424 AD) eras of the Ming dynasty were excavated on the strata of Ming dynasty [1]. In archaeology, due to the lack of specimens, the question about the existence of an official kiln in Longquan has only been modestly investigated. However, the emergence of the imperial porcelain caused academic debate about the property of Fengdongyan kiln in the Early Ming dynasty. At present, there are two main ⇑ Corresponding author. E-mail address: [email protected] (X.Q. Feng). http://dx.doi.org/10.1016/j.nimb.2016.05.027 0168-583X/Ó 2016 Elsevier B.V. All rights reserved.
views about Fengdongyan kiln. Some believe that it is the civilian kiln, while others think that it is the imperial kiln. ‘‘Longquan Kiln archaeological excavations expert meetings and press conference” was held in Longquan on January 18–19, 2007 [2]. Combined with the definition of official kiln, most experts believed that Fengdongyan kiln should belong to the civilian kiln, and the property of the production for this kiln was following government orders and producing the imperial porcelain. But according to ancient literature of the imperial porcelain [3], a large number of imperial porcelain was fired in Longquan kiln from Hongwu to Tianshun periods of the Ming dynasty. In addition, Longquan kiln and Jingdezhen imperial kiln have always been mentioned together. So some scholars believe that Fengdongyan kiln is similar to the official kiln of Jingdezhen. Therefore, the identification of the Fengdongyan kiln is a problem that cannot be answered by only typological considerations. It is well known that the contents of major, minor and trace elements of the porcelain body and glaze depend on its raw material and manufacturing technology, which can be used to indicate the age of the porcelain and its provenance [4–6]. Different contents mean different recipes of body and glaze. In the ancient, due to strict hierarchy, the porcelain fired at the official kiln was used for the emperor. So the recipes and the dispersion degree of body or glaze
L. Li et al. / Nuclear Instruments and Methods in Physics Research B 381 (2016) 52–57
53
Fig. 1. The overall view of Fengdongyan kiln.
in the official kiln were different from those of civilian kiln. The civilian and official productions of Ru Celadon have been researched focusing on their different techniques and glaze formulae [7]. The results mean that Ru official celadon and Ru folk porcelain were made from almost the same body materials but each has its own different glaze formula. So according to studying on the variation of raw materials and techniques between the civilian and imperial porcelain, the property of Fengdongyan kiln is discussed in our work. In this paper, the chemical composition of the civilian and imperial Longquan celadon porcelain of Fengdongyan kiln were explored by use of Instrumental Neutron Activation Analysis (INAA) and Energy Dispersive X-ray Fluorescence (EDXRF). INAA is a promising method of elemental analysis. It has been broadly used in the archaeological field since E.V. Sayre’s work [8]. The
EDXRF technique is a highly sensitive, non-destructive, multielement analysis method with good repeatability [9]. The two techniques have been widely used to characterise the ingredients of ancient ceramics, glass, biomedicine and so on. From a statistical analysis of the elemental data, the raw material and the development of porcelain firing technology are discussed. The information is used to display the inheritance relationship of the raw materials used for the body and glaze in the civilian and imperial porcelain. 2. Samples and experimental 2.1. Samples One hundred and sixty-four shards excavated at Fengdongyan kiln were selected as specimens. According to the typology, they
54
L. Li et al. / Nuclear Instruments and Methods in Physics Research B 381 (2016) 52–57
2.2. INAA experiment
Table 1 Detail information on the celadon shards in Fengdongyan kiln site. Name in the paper
Date
Amount
Type
HW-M
25 46
E-M
Hongwu era of Ming dynasty (1368–1398 AD) Yongle era of Ming dynasty (1403–1424 AD) Early Ming dynasty
M-M
Middle Ming dynasty
50
Imperial porcelain Imperial porcelain Civilian porcelain Civilian porcelain
YL-M
43
were identified as the products of imperial and civilian porcelain, and they were typologically classified into four cultural periods (see Table 1, Figs. 2 and 3). The first group is referred to as HW-M in this paper. It was fired during the Hongwu period (1368–1398 AD) of Ming dynasty. The second group is referred to as YL-M in this paper and was fired during the Yongle (1403–1424 AD) of Ming dynasty. These two groups belong to imperial porcelain concentrated in the southwest of Fengdongyan kiln in Ming dynasty. The third and fourth groups are referred to as E-M and M-M in this paper, and they were fired during the Early Ming dynasty and during the Middle Ming dynasty, respectively. The third and fourth groups were the civilian porcelain excavated at the south of Fengdongyan kiln.
For one group, a sample with 30 mm 10 mm was cut from each shard. After abrading away the glaze, the pure body was washed in an ultrasonic cleaner with tap water and then three times in deionised water, and then dried at 105 °C for 8 h. After those steps, the pure body was ground into a powder of 74 lm size in an agate mortar. About 200 mg of each powder sample were placed into the miniature neutron source reactor of the Chinese Institute of Atomic Energy simultaneously with quality control standards. The quality control standard is the Chinese national certified reference material of soil (GBW07402). The radioactivity of the sample was measured by an HPGe detector connected to a multi-channel analyser. Under these conditions, the contents of 26 elements (V, Mn, Dy, Ga, Ho, Sb, La, Sm, Yb, W, U, Sc, Cr, Co, Zn, Rb, Cs, Ba, Ce, Nd, Eu, Tb, Lu, Hf, Ta and Th) were determined. The validation of the analytical method applied in this work was carried out by analysis of GBW07402. The results showed that the experimental values were well in agreement with the certified ones (see Table 2). 2.3. EDXRF experiment For another group, samples with 30 mm 10 mm were cut from these sherds. After the cross-section was polished, it was washed and dried in the same procedure as in INAA. The EDXRF experiments were performed on an EDAX Eagle III spectrometer
Fig. 2. The appearances of imperial porcelain shards excavated from Fengdongyan kiln.
Fig. 3. The appearances of civilian porcelain shards excavated from Fengdongyan kiln.
55
L. Li et al. / Nuclear Instruments and Methods in Physics Research B 381 (2016) 52–57 Table 2 Contents of 32 elements in the certified reference material GBW07402 by INAA. Element
Experiment value
Reference value
Element
Experiment value
Reference value
V/ppm Mn/ ppm Dy/ppm Ga/ppm Ho/ppm Sb/ppm La/ppm Sm/ ppm Yb/ppm W/ppm U/ppm Sc/ppm Cr/ppm
59.4 501
62.0 510
Co/ppm Zn/ppm
8.07 41
8.70 42
4.21 10.8 0.90 1.48 163 17.4
4.40 12.0 0.93 1.30 164 18.0
Rb/ppm Cs/ppm Ba/ppm Ce/ppm Nd/ppm Eu/ppm
84.9 4.52 900 399 197 2.8
88.0 4.90 930 402 210 3.0
2.06 1.29 1.51 10.3 51.1
2.00 1.08 1.40 10.7 47.0
Tb/ppm Lu/ppm Hf/ppm Ta/ppm Th/ppm
0.87 0.31 6.22 0.62 16.5
0.97 0.32 5.8 0.78 16.6
of Institute of High Energy Physics, CAS, Beijing of China. The spectrometer with a Mo tube and a 125 lm Be window has an incident beam angle of 65° and an emergence angle of 60°. The detector is a liquid-nitrogen-cooled Si (Li) crystal with a 160.3 eV at Mn Ka. There is a vacuum in the chamber and the diameter of X-ray beam spot was selected to be 1 mm to fit the body thickness dimensions. The voltage and current of X-ray tube are operated at 40 kV and 250 lA, respectively. The software employed for spectrum deconvolution and analysis is the program VISION32. The elemental abundances of Na2O, MgO, Al2O3, SiO2, K2O, CaO, TiO2, MnO, Fe2O3, Cu, Zn, Rb and Sr were quantified by fundamental parameter (FP) method. In order to get a better measure of the precision and accuracy of the data in this work, ancient ceramic samples of known composition were selected to compare the expected and observed elemental concentrations measured on the ceramic reference samples used in the analysis. The results show that the experimental values are well in agreement with the certified ones (see Table 5). The average values of each elemental composition in porcelain body and glaze are displayed in Tables 6 and 7, respectively. The data of Na2O and MgO are provided as suggestive only because of the poor fluorescence yields and low counting statistics obtained for characteristic X-ray radiation.
Table 4 REE parameter of ancient porcelain body.
HW-M YL-M E-M M-M
La/Sm
Tb/Yb
La/Yb
dEu
3.99 3.87 3.18 3.86
1.15 1.21 1.18 1.18
7.86 7.99 6.21 7.53
0.31 0.29 0.27 0.29
(9), the YL-M group (9), the E-M group (9) and the M-M group (9). The average values of each element in the bodies of the imperial and civilian porcelain in four cultural periods are shown in Table 3. It can be seen that the average contents of other trace elements were close to each other in HW-M, YL-M, E-M and M-M groups except for V and Ba. The scatter plots of V and Ba contents are exhibited in Fig. 5. All samples are located in two plot regions. Samples of HW-M group are concentrated in the upper right region while those of the YLM and E-M groups are distributed in the lower left region, and a few data plots of the M-M group overlap with those of the HWM groups. It means that the raw material of the imperial porcelain body in Hongwu period is similar with that of the civilian porcelain body in Middle of Ming dynasty, while the same raw material of the body are used in imperial porcelain in Yongle period and civilian porcelain. Rare earth elements (REEs) are located in IIIB column of the periodic table of and their atomic numbers range from 57 to 71. Their crystal chemical and geochemical properties are similar because their atomic and ionic radiuses are close. Due to relative stability, REE can be used as a tracer source in the archaeological research [10,11]. The patterns and some parameters of REEs in the porcelain bodies of HW-M, YL-M, E-M, and M-M groups are shown in Fig. 4 and Table 4. In Fig. 4, the REEs patterns of imperial porcelain (HW-M and YL-M groups) match those of the civilian porcelain (E-M and M-M groups) very well. In Table 4, the variation range of (La/Sm), (Tb/Yb), (La/Yb), and dEu of the four groups are 3.18–3.99, 1.15–1.21, 6.21–7.99 and 0.27–0.31, respectively. They are similar to each other. So it can be inferred that the kaolin for imperial porcelain body are the same as that of the civilian porcelain body.
3.2. The multivariate statistical analysis of EDXRF data for imperial and civilian porcelain
3. Results and discussion 3.1. The geochemical interpretation of INAA data for imperial and civilian porcelain Due to the expensive cost and long measuring time, only 36 shards were selected as specimens for INAA: The HW-M group
The average values of each element in the bodies of imperial and civilian porcelain are shown in Table 6. It can be seen that the K2O content of the porcelain body in HW-M group is similar to those of E-M group and lower than those of M-M group, while the values of K2O in YL-M group are lower than those of other three
Table 3 The average values of 32 elements in body by INAA. Elements
HW-M
YL-M
E-M
M-M
Elements
HW-M
YL-M
E-M
M-M
V/ppm Mn/ppm Dy/ppm Ga/ppm Ho/ppm Sb/ppm La/ppm Sm/ppm Yb/ppm W/ppm U/ppm Sc/ppm Cr/ppm
13 ± 2 490 ± 35 8±1 24 ± 2 1.3 ± 0.1 0.18 ± 0.03 58 ± 4 9.1 ± 0.3 4.4 ± 0.2 3.9 ± 0.7 4.4 ± 0.2 6.1 ± 0.3 3.3 ± 1.0
6±1 478 ± 60 10 ± 1 22 ± 2 1.5 ± 0.1 0.21 ± 0.04 66 ± 8 10.6 ± 0.5 4.9 ± 0.2 2.5 ± 0.4 4.3 ± 0.1 5.9 ± 0.5 3.5 ± 0.5
5±2 471 ± 110 10 ± 1 23 ± 2 1.6 ± 0.1 0.26 ± 0.04 53 ± 17 10.2 ± 1.3 5.1 ± 0.4 2.2 ± 0.4 4.3 ± 0.3 5.5 ± 0.8 3.0 ± 0.3
7±3 463 ± 49 8±1 23 ± 2 1.4 ± 0.1 0.25 ± 0.07 56 ± 10 9.0 ± 0.7 4.5 ± 0.4 2.7 ± 0.3 4.6 ± 0.3 5.6 ± 0.7 3.7 ± 1.6
Co/ppm Zn/ppm Rb/ppm Cs/ppm Ba/ppm Ce/ppm Nd/ppm Eu/ppm Tb/ppm Lu/ppm Hf/ppm Ta/ppm Th/ppm
1.7 ± 0.2 84 ± 10 319 ± 17 6.6 ± 0.3 514 ± 76 103 ± 5 54 ± 3 1.0 ± 0.1 1.3 ± 0.1 0.6 ± 0.1 6.2 ± 0.3 2.3 ± 0.1 38 ± 1
1.4 ± 0.2 93 ± 25 278 ± 8 5.6 ± 0.4 327 ± 34 103 ± 11 62 ± 5 1.0 ± 0.1 1.6 ± 0.1 0.7 ± 0.1 6.3 ± 0.2 2.3 ± 0.2 39 ± 1
1.2 ± 0.3 83 ± 21 296 ± 34 5.4 ± 0.4 276 ± 58 89 ± 21 56 ± 9 0.9 ± 0.2 1.6 ± 0.2 0.7 ± 0.1 6.5 ± 0.2 2.5 ± 0.4 42 ± 3
1.1 ± 0.4 74 ± 19 307 ± 17 5.9 ± 1.3 435 ± 92 102 ± 14 52 ± 5 0.9 ± 0.1 1.4 ± 0.1 0.6 ± 0.1 6.2 ± 0.3 2.3 ± 0.3 38 ± 2
56
L. Li et al. / Nuclear Instruments and Methods in Physics Research B 381 (2016) 52–57
Table 5 The quantitative results for ancient ceramic sample by EDXRF.
Experimental value The certified value
Na2O %
MgO %
Al2O3 %
SiO2 %
K2O %
CaO %
TiO2 %
MnO %
Fe2O3 %
Cu ppm
Zn ppm
Rb ppm
Sr ppm
0.65 0.44
0.62 0.70
24.34 23.90
67.54 67.50
2.27 2.30
0.54 0.62
0.88 0.95
0.022 0.026
2.77 2.70
46 27
54 59
119 113
109 103
Table 6 The average values of 13 elemental compositions in body by EDXRF. Elements
HW-M
YL-M
E-M
M-M
Elements
HW-M
YL-M
E-M
M-M
Na2O/% MgO/% Al2O3/% SiO2/% K2O/% CaO/% TiO2/%
0.59 ± 0.17 0.43 ± 0.06 20.4 ± 0.5 70.7 ± 0.7 5.32 ± 0.34 0.056 ± 0.009 0.23 ± 0.04
0.77 ± 0.18 0.38 ± 0.07 20.5 ± 0.5 70.9 ± 0.8 5.23 ± 0.42 0.077 ± 0.016 0.15 ± 0.03
0.71 ± 0.18 0.39 ± 0.05 20.3 ± 0.8 70.9 ± 1.0 5.35 ± 0.46 0.082 ± 0.014 0.14 ± 0.03
0.72 ± 0.20 0.38 ± 0.06 19.6 ± 0.7 70.9 ± 0.8 6.13 ± 0.30 0.093 ± 0.017 0.16 ± 0.04
MnO/% Fe2O3/% Cu/ppm Zn/ppm Rb/ppm Sr/ppm
0.05 ± 0.01 2.11 ± 0.12 24 ± 2 44 ± 4 326 ± 23 40 ± 5
0.05 ± 0.01 1.79 ± 0.11 24 ± 2 45 ± 8 304 ± 21 30 ± 5
0.05 ± 0.01 1.87 ± 0.14 24 ± 3 46 ± 8 319 ± 24 31 ± 5
0.06 ± 0.01 1.86 ± 0.13 26 ± 3 41 ± 6 340 ± 18 36 ± 7
Table 7 The average values of 13 elemental compositions in glaze EDXRF. Elements
HW-M
YL-M
E-M
M-M
Elements
HW-M
YL-M
E-M
M-M
Na2O/% MgO/% Al2O3/% SiO2/% K2O/% CaO/% TiO2/%
0.64 ± 0.18 0.68 ± 0.10 12.8 ± 0.5 70.0 ± 1.7 5.84 ± 0.48 7.06 ± 1.73 0.21 ± 0.03
0.68 ± 0.16 0.67 ± 0.11 13.1 ± 0.7 70.7 ± 1.6 5.81 ± 0.32 6.24 ± 1.22 0.18 ± 0.03
0.64 ± 0.16 0.59 ± 0.15 12.8 ± 0.8 70.4 ± 2.1 6.11 ± 0.63 6.53 ± 1.95 0.19 ± 0.03
0.66 ± 0.21 0.69 ± 0.18 12.7 ± 0.5 70.1 ± 1.6 6.28 ± 0.32 6.81 ± 1.25 0.20 ± 0.03
MnO/% Fe2O3/% Cu/ppm Zn/ppm Rb/ppm Sr/ppm
0.33 ± 0.08 2.17 ± 0.13 31 ± 4 60 ± 16 269 ± 16 467 ± 97
0.34 ± 0.05 2.12 ± 0.23 31 ± 4 54 ± 17 268 ± 13 424 ± 56
0.33 ± 0.08 2.14 ± 0.30 30 ± 4 48 ± 20 280 ± 26 477 ± 131
0.38 ± 0.10 1.91 ± 0.35 31 ± 4 52 ± 25 283 ± 23 456 ± 73
cultural periods. The average value of CaO in the imperial porcelain is slightly lower than those of civilian porcelain. The average content of TiO2 in the HW-M group is (0.23 ± 0.04)% higher than that (0.14–0.16)% in other three groups. The abundance change of Fe2O3 in the imperial porcelain body is similar to that of TiO2. The composition of Fe2O3 in the porcelain clay in the Longquan district is less than that of the imperial porcelain body, so some other clay material, probably a Zijin soil was added in the bodies of the imperial and civilian porcelain. It can not only improve the body strength, but also increase the colour of black or grey in porcelain body. A detailed discussion of this has been published elsewhere [12]. The average concentrations of SiO2, Al2O3, MnO and the trace elements Cu, Zn, Rb, Sr are close to each other in the imperial and civilian porcelain body, as shown in Table 6.
Fig. 4. The chondrite-normalised REE pattern.
In this work, principal components analysis (PCA) was used to study the raw material used for the body in the imperial and civilian porcelain. The main objective of PCA is to reduce the dimensionality of the observations [13]. A factorial analysis diagram from the composition of bodies made in the imperial and civilian porcelain is shown in Fig. 6. The data for K2O, CaO, TiO2, Fe2O3 and Rb are used and the eigenvalue sum of Factors 1 and 2 accounts for 73.92% of the total variance. Factor score 1 (F1) mainly includes the changes of TiO2 and Fe2O3 and Factor score 2 (F2) mainly represents the content of K2O and Rb. Some samples in the HW-M group are located in the lower area and the others overlap with the three groups in Fig. 6. The reason for this difference could be different elutriation degrees of the same raw materials or the different raw materials used. Combining with the major elements (Table 6) and the trace elements (Table 3), the most of element contents are similar in the four groups. So we conclude that different elutriation degree
Fig. 5. Scatterplot of V and Ba in bodies of imperial and civilian porcelain in Ming dynasty.
L. Li et al. / Nuclear Instruments and Methods in Physics Research B 381 (2016) 52–57
57
tration of TiO2, Fe2O3, V and Ba in HW-M group. It means that the raw materials for making body in imperial porcelain are similar with those of the civilian porcelain and the degrees of elutriation can be slightly different in HW-M group. The average contents of the glaze in major and trace elements are similar, which displays that the raw materials of the glaze are close to each other between imperial and civilian porcelain. Based on the result of composition in body and glaze, Fengdongyan kiln in Early Ming dynasty should be the civilian kiln. 4. Conclusion
Fig. 6. PCA for the contents of elemental composition in body of imperial and civilian with EDXRF.
In this paper, INAA and EDXRF have been applied to analyse the elemental composition of imperial and civilian porcelain at the Fengdongyan kiln. Comparing the elemental composition of the imperial porcelain with those of civilian porcelain, the contents of TiO2, Fe2O3, V and Ba in Hongwu period are slightly different. After eliminating the influence of the change in manufacturing technology between different periods by the analysis of REEs, we could conclude that the mineral sources of producing kaolin for imperial porcelain body are the same to that of civilian porcelain body. According to the PCA analytical results, it means that the raw materials for making body and glaze in imperial porcelain are similar with those of the civilian porcelain and the degrees of elutriation for the body can be slightly different in HW-M period of Ming dynasty. Those analytical results support the interpretation put forth by some archaeologists, i.e. that the Fengdongyan kiln is civilian not official. Acknowledgements The authors greatly appreciate the Zhejiang Province Institute of Cultural Relics and Archaeology, which provided ancient Chinese celadon excavated from the Fengdongyan site of the Longquan kiln. This work was financially supported by the National Natural Science Foundation of China (11575207 and 10875137).
Fig. 7. PCA for the contents of elemental composition in glaze of imperial and civilian with EDXRF.
of the same raw materials is the main reason between the HW-M group and the others. The raw material of making ceramic was purified by the elutriation, and the impurity was removed, especially iron impurities. So the chemical composition of body and glaze can be changed as different degrees of elutriation. From the above, the raw materials of the HW-M group have the similarity with the civilian porcelain. In Fig. 6, samples of the YL-M group overlap with those of E-M and M-M, which reveal that the raw materials for making the porcelain body in YL-M group are the same as the civilian porcelain. The elemental compositions in the porcelain glaze of the four groups are shown in Table 7. There were no evident differences of average concentration of major and trace elements between each other. The similar conclusion can also be get from Fig. 7. The contents of Al2O3, SiO2, K2O, CaO, TiO2, MnO, Fe2O3, Rb and Sr are used and the eigenvalue sum of Factors 1 and 2 accounts for 58.93% of the total variance. Fig. 7 displays that the sample plots in the four groups are scattered in a large region and the distribution of data points in the HW-M and YL-M group are relatively concentrated. It means that the raw materials of glaze in the four groups are close to each other and the similar processing for making glaze is used between imperial porcelain and civilian porcelain. In view of analysis results of INAA and EDXRF, there are no obvious differences of contents of the most major and trace elements between imperial and civilian porcelain body except the concen-
References [1] Zhejiang Institute of Cultural Relics et al., Porcelain Excavated from Fengdongyan Kiln Site at Dayao Country of Longquan City, Cultural of Relics Publishing House, Beijing, 2009, p. 1. [2] Y.M. Shen, D.S. Qin, W.B. Shi, The academic forum summary of archaeological excavation at Fengdonggyan site of Longquan kiln, Cultural Relics 5 (2007) 93– 96. [3] H. Huo, J.H. Chen, M.H. Hao, et al., On Longquan celadon fragments form the accumulation pit of Hexia site in Chuzhou District, Huai ’an City of Jiangsu, Southeast Culture 2 (2010) 38–45. [4] D. Zhu et al., PIXE study on the provenance of Chinese ancient porcelain, Nucl. Instrum. Methods B 249 (2006) 633–637. [5] B. Ma, L. Liu, S.L. Feng, et al., Analysis of the elemental composition of Tang Sancai from the four major kilns in China using EDXRF, Nucl. Instrum. Methods B 319 (2014) 95–99. [6] G.X. Xie, S.L. Feng, X.Q. Feng, et al., The dating of ancient Chinese celadon by INAA and pattern recognition methods, Archaeometry 51 (4) (2009) 682–699. [7] Y.Z. Ding, H. Li, X.M. Sun, et al., Ru celadon ware production in Qingliangsi kiln for civilian and official uses focusing on their different techniques and glaze formulae, Palace Mus. J. 3 (2013) 62–73. [8] E.V. Sayre, R.W. Dodson, Neutron activation study of Mediterranean potsherds, Am. J. Archaeol. 61 (1975) 35–41. [9] D.N. Papadopoulou et al., Comparison of a portable micro-X-ray fluorescence spectrometry with inductively coupled plasma atomic emission spectrometry for the ancient ceramics analysis, Spectrochim. Acta B 59 (2004) 1877–1884. [10] X.Q. Jia, J.Q. Dong, S. Han, et al., INAA analysis for Jun porcelain and Modern Chinese Jun porcelain, J. Isot. 17 (3) (2004) 129. [11] I.L. Yong, Provenance derived from the geochemistry of late Paleozoic early Mesozoic mudrocks of the Pyeongan Supergroup, Korea, Sediment. Geol. 149 (4) (2002) 219–235. [12] L. Li, L.T. Yan, S.L. Feng, et al., Elemental characterization by EDXRF of imperial Longquan celadon porcelain excavated from Fengdongyan kiln, Dayao County, Archaeometry 57 (6) (2015) 966–976. [13] W. Härdle, L. Simar, Applied Multivariate Statistical Analysis, Springer, Berlin, 2003, p. 323.