The metallography and corrosion of an ancient chinese bimetallic bronze sword

The metallography and corrosion of an ancient chinese bimetallic bronze sword

G Model CULHER-3490; No. of Pages 7 ARTICLE IN PRESS Journal of Cultural Heritage xxx (2018) xxx–xxx Available online at ScienceDirect www.scienced...

3MB Sizes 1 Downloads 16 Views

G Model CULHER-3490; No. of Pages 7

ARTICLE IN PRESS Journal of Cultural Heritage xxx (2018) xxx–xxx

Available online at

ScienceDirect www.sciencedirect.com

Case study

The metallography and corrosion of an ancient chinese bimetallic bronze sword Huang Wei a,∗ , Winfried Kockelmann b , Evelyne Godfrey b,c , David A Scott d a

China Numismatic Museum, Beijing, China Rutherford Appleton Laboratory, ISIS Facility, Chilton, UK Uffington Heritage Watch, Faringdon, Oxfordshire, UK d UCLA/Getty Conservation Program, Cotsen Institute of Archaeology, University of California, Los Angeles, USA b c

a r t i c l e

i n f o

Article history: Received 26 April 2018 Accepted 10 October 2018 Available online xxx Keywords: bimetal sword Chinese bronze Corrosion

a b s t r a c t A bimetal bronze sword, unearthed from Hunan Province China, dating to the Warring States period (476–221BC) was examined analytically by optical metallography, neutron diffraction and X-ray diffraction. The results indicate that the tin content of the blade is 16%, not higher as previously reported. The bimetallic bronze sword possesses a typical cast structure with annealing on the edge of spine. Distinct corrosion compounds pyromorphite was identified, with pseudomorphic malachite having replaced the original alpha phase. Redeposited copper and unusual phosphorus-containing corrosion products were analysed to investigate the mechanism of formation. © 2018 Elsevier Masson SAS. All rights reserved.

1. Introduction The earliest bimetallic artifacts found in Hebei province and Beijing are bronze Yues with meteoritic iron blades dating to the Middle Shang Period (14th century BC). Various bimetallic objects such as a bronze dagger with an iron blade, an iron Ben with a bronze hilt, an iron sword with a bronze hilt and an iron knife with a bronze hilt dated from Western Zhou Dynasty (11–8th century BC) to early Spring and Autumn period (770–476 BC) were found in Henan and Shaanxi province [1], from the central plain and surrounding area. Iron as a precious material was often used together with bronze tools and weapons before the advent of its mass production in the Late Bronze Age of ancient China. Recently two typical iron-bronze bimetallic artifacts, iron-blade bronze chipping knife and iron-blade bronze dagger axe (Ge in Chinese), were found in Liangdaicun village, Han cheng city, Shanxi province, which are examined technically by metallography [2], the two bimetallic artifacts are dated to early Spring and Autumn period (770–476 BC). There are two kinds of bimetallic artifact in Chinese archaeological finds, iron-bronze combined mechanically, and bronze with different tin contents on corresponding parts of a sword. During the Warring States Period, (476–221BC), iron-bronze bimetallic artifacts were replaced by bimetallic bronze swords. These swords were cast together in two different alloys for the back and the blade.

∗ Corresponding author.

This kind of bimetallic bronze sword was only found in South China, within the area of the Chu culture and its southward neighbours. Some bimetallic bronze swords excavated at Jiangxi, Zhejiang, Hunan and Guangdong were subjected to preliminarily alloy compositional analyses [3,4,5] showing higher tin contents for the blade as compared with the back. Four bimetallic bronze sword fragments collected by the Shanghai Museum were examined using X-ray fluorescence; the tin content in the blade was found to be 22–24 wt%, while the back had 8–15 wt% tin [6,7]. Simulation experiments designed by the author were used to reconstruct the casting processes: the back of the blade is cast first followed by annealing the tenon slightly, and then the piece-mould of the blade with finished back inset is prepared for pouring. The back and blade would be integrated into a completed sword after the liquid metal solidification and shrinkage (see Fig. 1, the middle part is finished firstly and then inserting to the piece-mould with the blade hollow). Two bronze bimetallic swords dating to early Warring States Period (476–221BC) unearthed from Xintai city, Shandong province, were analyzed by X-ray, CT, XRF and metallography, which show the low and high tin bronze alloy are used respectively on spine and blade for every sword [8]. However, little attention has been paid to the corrosion phenomena of bimetallic swords, which provides a good opportunity to investigate the mechanism of deterioration of different bronze alloys with variable tin content under the same environmental conditions. More case studies are needed to increase the technological information on these swords from outside the central plain area in ancient China.

https://doi.org/10.1016/j.culher.2018.10.004 1296-2074/© 2018 Elsevier Masson SAS. All rights reserved.

Please cite this article in press as: W. Huang, et al., The metallography and corrosion of an ancient chinese bimetallic bronze sword, Journal of Cultural Heritage (2018), https://doi.org/10.1016/j.culher.2018.10.004

G Model CULHER-3490; No. of Pages 7

ARTICLE IN PRESS

2

W. Huang et al. / Journal of Cultural Heritage xxx (2018) xxx–xxx

Fig. 1. The casting method for bimetal sword (after Haiping, 2002).

Fig. 2. Bimetal bronze sword excavated in Hunan Province.

Fig. 3. The colorful surface of the blade (yellow corrosion at the seam, left above).

Fig. 5. Redeposited copper in the crevice between the back and the blade (the corroded alpha + delta eutectoid and lead particles can be seen). Magnification 20X.

content than the other sections. Fig. 5 It is worth noting that some unusual yellowish green corrosion products are found in the seam between the back and blade. On the other side of the sword, layered corrosion products are clearly visible, the cuprite is possibly surrounded or covered by blue and green corrosion from loose to compact green layers outwards. Soil minerals and corrosion products formed on the back section illustrate a light green color on the outmost layer. Fig. 4. The cross section for sampling.

Figs. 2–4 shows a remnant of the bimetallic bronze sword studied here dating to Warring States period, excavated in the Hunan Province by Chinese archaeologists. Distinct colors of the corrosion phases are observed on all parts of the sword: green, yellowish green and blue, with some soil inclusions in different areas. Higher magnification illustrates that the glazed green corrosion shiny layer appeared only on the surface of the blade, which contains more tin

2. Experimental details Sampling on the sword was carried out, and samples were embedded in resin, followed by grinding, polishing and etching (by using FeCl3 ·6H2 O in C2 H5 OH with HCl, Na2 S2 O3 and Na2 S2 O5 miscible liquids for colorful metallography) to investigate the microstructure. The colorful metallography can show the distinct phases clearly, such as in Figs. 6 and 7.The back and blade samples show a typical casting structure with ␣ phase, (␣ + ␦) eutectoid and some dispersed lead particles. The re-crystallized grains on the

Please cite this article in press as: W. Huang, et al., The metallography and corrosion of an ancient chinese bimetallic bronze sword, Journal of Cultural Heritage (2018), https://doi.org/10.1016/j.culher.2018.10.004

G Model CULHER-3490; No. of Pages 7

ARTICLE IN PRESS W. Huang et al. / Journal of Cultural Heritage xxx (2018) xxx–xxx

3

Fig. 6. The corroded alpha phase in the middle area of the image. Magnification 20X.

edge of the tenon section indicate slight re-heating treatment after casting. A blade piece was sampled to be examined by time-of-flight neutron diffraction, which provides a large volume and high penetration measurement of the bulk phase compositions non-

destructively [9,10,11]. The phases in the alloy substrate and corrosion products on the surface can be tested simultaneously, and tin contents of the alloy are calculated in terms of chemical formula [12]. Neutron diffraction analysis was performed on the General Materials Diffractometer (GEM) of the ISIS neutron

Please cite this article in press as: W. Huang, et al., The metallography and corrosion of an ancient chinese bimetallic bronze sword, Journal of Cultural Heritage (2018), https://doi.org/10.1016/j.culher.2018.10.004

G Model CULHER-3490; No. of Pages 7

ARTICLE IN PRESS

4

W. Huang et al. / Journal of Cultural Heritage xxx (2018) xxx–xxx

Fig. 7. Layered corrosion products on the blade surface (Magnification 20X Polarized Light).

source at the Rutherford Appleton Laboratory, UK. Neutrons are produced by stopping 800 MeV protons in a tungsten target. ISIS is a pulsed source operating at 40 Hz, thus delivering 40 polychromatic pulses of neutrons per seconds to the sample. The pulsed operation allows determining the wavelength of each diffracted and detected neutron in the polychromatic pulse via a time-of-flight (TOF) measurement. A neutron detector at a scattering angle 2 measures the neutron TOF, i.e. the time that passes from the generation of the neutron until its capture in the detector. GEM has about 7000 individual scintillation detector elements viewing the sample from different angles.

The total 16% tin content in the blade sample is different from the composition reported by Haiping [6,7], which exhibits an average 23% tin content in the blade by XRF and a high result of 17–19% tin content of the blade was also reported by Shifan [5]. Arguably, the 23% tin content, obtained by X-ray fluorescence energy dispersive spectrometry, was due to the limited penetration depth of the Xray method, i.e. the composition is from the tin-rich layer on the corroded surface of the blade that originates from copper depletion to the environment. The phases and elements distributed in the sword substrate are very unevenly distributed based on the neutron diffraction data and metallographic examination. XRF or SEM-EDS have the same short penetration depths of a few␮m in a small area on the corroded surface of the bronze alloy. The compositional data from the surface methods are less representative in revealing the alloy constituents reliably. More practical case studies are needed to explore the compositions of bimetal alloy quantitatively, for example by using non-destructive neutron diffraction. Furthermore, a 23% tin content of blade is not good for the mechanical properties of sword as a weapon [13]. Although high tin contents can improve the hardness of bronze alloys, 23% or more tin generates brittleness for weapons while decreasing the tensile strength. Ancient Chinese mirrors often contain tin contents of more than 20%, which enhances light reflection by the shiny silver-coloured surface. However such high tin bronze mirrors are fragile. So the tin content of 16–18% for the blade of the Hunan sword should be reasonable. The lead content of the blade is 7%. Lead can only improve the fluidity of the melt, and has no other purpose with regards to the mechanical properties of the sword, since lead is generally not added to bronze if it is hammered or worked to shape. For ancient Chinese bronze, lead is usually added for the facility of alloy flowing during casting.

3. Metallography and bulk composition 4. Corrosion phenomena The neutron diffraction data given in Table 1 demonstrate that two types of alpha phases with 14% and 10% tin contents, respectively, are present in the sampled bronze substrate, indicating coring during cooling to room temperature during solidification. The coring means a process of the non-average elemental distribution in alloy. The delta phase of the (␣ + ␦) eutectoid is a common phase in cast bronze. It is noted that a small amount of epsilon phase, often found in tinned surfaces of ancient copperbased alloys, is present in the blade sample of the bimetal sword, which is seldom seen in bronze alloy analysis. Usually, attention is given to copper, tin and lead contents. Hence the appearance of the epsilon phase may be relevant with regards to pure tin being added into the copper during the fabrication of the bronze alloy. Possibly, the ␦ phase transforms to (␣ + ␧) at room temperature with abnormal cooling conditions (Michelle, 2007), but this is seldom if ever observed.

Cuprite was observed as main corrosion product. Malachite was not identified by neutron diffraction but it was found that the amount of malachite is smaller than 0.2%. As shown in Fig. 3 and Fig. 7, three kinds of patina were found on the surface of the sword: glazed and compact green, yellowish with soil minerals and powdery blue corrosion products. These products were analyzed by X-ray diffraction, the results being shown in Table 2, which indicates that malachite, cuprite, cassiterite and quartz are the major constituents in the corrosion. However, the yellowish green corrosion products sampled between the seam of the back and the blade found iron, lead and phosphorus-containing compounds in addition to malachite, cuprite, quartz, and romarchite. Diverse selective corrosion and redeposited copper were observed in the microstructure of the sword sample. The

Table 1 Phase composition and tin contents from neutron diffraction data of the blade. Identified phases: weight fraction (%) Alpha-Cu/Sn (1) Alpha-Cu/Sn (2) Delta phase Epsilon phase Lead Cuprite Malachite

Calculated tin content: weight fraction (%) 57 13 20 0.5 7 2.5 < 0.2

Alpha-Cu/Sn(1) Alpha-Cu/Sn(2) Delta Epsilon

14.1 % Sn calculated 10.2 % Sn calculated 33 % Sn on phase diagram 38% Sn on phase diagram

Total tin Total lead

16 wt% Sn 7 wt% Pb

(TOF-ND: GEM diffractometer; data analysis: GSAS program for Rietveld analysis).

Please cite this article in press as: W. Huang, et al., The metallography and corrosion of an ancient chinese bimetallic bronze sword, Journal of Cultural Heritage (2018), https://doi.org/10.1016/j.culher.2018.10.004

G Model

ARTICLE IN PRESS

CULHER-3490; No. of Pages 7

W. Huang et al. / Journal of Cultural Heritage xxx (2018) xxx–xxx

5

Table 2 Corrosion products of the bimetal sword examined by X-ray diffraction. Sampling zone

Corrosion products identified by XRD

Spine surface Blade surface Seam between spine and blade(greenyellow corrosion)

Malachite, cassiterite, quartz Malachite, cuprite, cassiterite, quartz Pyromorphite, libethenite, malachite, quartz, goethite, phosphogartrellite, cuprite, romarchite, lanarkite, lepidocrocite

(Rigaku R-axis Spider machine at UCLA).

Fig. 9. Pseudomorphic structure on the surface of sword blade surface. Magnification 20X.

Fig. 8. XRD pattern for green and yellow corrosion from different parts.

phosphorus-contained corrosion products are rarely reported from Chinese contexts. 5. Selective corrosion: type I structures Ancient Chinese bronze corrosion features were suggested by Chase [14] to show type I corrosion, namely the removal of the tin-rich phase to form tin oxide or mixed tin oxide and cuprite corrosion products on the surface [15]. In type II, the alpha phase of the alloy is dissolved and the delta remains in an uncorroded state. As Robbiola [16] pointed out, in an oxygenated corrosive medium, the main phenomenon involved in bronze corrosion is selective dissolution of copper from the copper solid solution (␣phase) in type I, while a coarse structure with a three-layer morphology is assumed to be controlled by mass transportation of negative species from the soil (mainly oxygen and chloride ions) characterizes type II. In the bimetallic sword we have a typical type I corrosion structure, cuprite mixed with cassiterite and copper carbonates such as malachite, shown in Fig. 2, Fig. 8, Fig. 7. The pseudomorphic dendritic structure shown in Fig. 6, Fig. 9 indicates that the preservation of structure within the corrosion may originate from an epitaxial relation between substrate and product, with a corresponding retention of shape [17]. The malachite crust in Fig. 7 shows a well preserved pseudomorphic dendrites pattern. Few examples of such structures have been published. One exception is a Han

dynasty (206BC–220BC) buckle reported by Scott [17], page 84, plate 11). The dendritic malachite on the higher tin zone in the metal corrosion could indicate the stable environment. In South China, excavated bronzes with thick greenish malachite covered is common, where the underground water is plentiful that has advantages for the formation of copper carbonates. In this example, the dense malachite corrosion not only protects the original cast structure itself in a good burial condition, but also hinders the massive oxidation of cuprite in the inner layer with the low tin content. In the meanwhile, it is noted here that the corroded (␣ + ␦) eutectoid appeared in the alloy substrate, but in the outer layer, corroded ␣solid solution is the main corrosion observed on the surface exposed to the environment overlaying a totally mineralized corrosion layer. A Chinese bronze vessel fragment dating to the first millennium BC studied by Gettens [18], exhibited a similar corrosion structure to the bimetallic sword: in the intermediate penetration zone the eutectic has been corroded leaving dendrites of the alpha phase unattacked, while on the outer edge of the intermediate penetration zone the alpha dendrites are attacked, and from here outward only mineral corrosion products are found. This corrosion phenomenon suggests that a passivating layer formed in type I corrosion which controlled the oxygen and soil ions migration into the inner alloy substrate. Between the seam of the blade and back, corroded (␣ + ␦) eutectoid has created much redeposited copper, which will be discussed below. As distinguished from type II corrosion structure and Getten’s research, no copper chloride and lead carbonate compounds were found in the X-ray and neutron diffraction analysis, yet the unusual corrosion product pyromorphite was found here. 6. Redeposited copper: mechanism and optimum forming condition Redeposited copper is found in many archaeological copperbased artifacts from all over the world, studied and reviewed by many scholars. Scott [9] reviewed the mechanisms for the formation of redeposited copper, from decomposition of the delta phase, the dissolution of cuprite inclusions within the bronze or corrosion from the dissolution of cuprite crusts formed during burial, or voids occupied by inclusions and lead globules which become replaced by copper. Jin bronze (West Zhou Dynasty, 1027 BC–771BC) fragments excavated from southwest Shanxi were studied by Wang [19], which showed that redeposited copper is very common both in the bronze core and in external corrosion layers, displayed as fine particles within the cuprite, in filled cracks and surface bands. These

Please cite this article in press as: W. Huang, et al., The metallography and corrosion of an ancient chinese bimetallic bronze sword, Journal of Cultural Heritage (2018), https://doi.org/10.1016/j.culher.2018.10.004

G Model CULHER-3490; No. of Pages 7

ARTICLE IN PRESS

6

W. Huang et al. / Journal of Cultural Heritage xxx (2018) xxx–xxx

phenomena are explained by the author as follows: preferential removal of tin seems to have been the first stage followed by the formation of cuprite, metallic copper was deposited from solution when cupric or/and cuprous ion solutions were saturated under reducing conditions. Bosi [20] provides a review and some case studies on Chinese and Italian artifacts, which determined that in most cases unalloyed copper was located in the core of the objects far from the outer corrosion products layer. Three types were propose • irregularly shaped forms, that pseudomorphically replaces other phases, which is closely related to long-term corrosion processes; • globular, that are surrounded by a layer of copper sulphide formed during roasting of the copper ore; or; • large, irregularly shaped masses, that display a twinned microstructure, which is due to incomplete mixing of copper and tin during the casting of the alloy. From the metallography of the bimetallic sword, large redeposited copper illustrates type (a) structure described by Bosi [20]. The metallic copper attributed to type (a) is considered a result of destannification by Wang and Merkel [21]. In the sword, copper has also formed in pre-existing cracks or voids. A great deal of redeposited copper assembled in the seam between blade and back, this narrow gap was formed by the incomplete combination during casting. In fact, destannification is just a step in the formation of metallic copper. Copper or secondary copper reduced from cuprite would replace the original alpha phase and eutectoid structure in situ if there was not enough space to escape. The movable redeposited copper will selectively dwell into the cracks or voids which provides a relatively large volume to accommodate, as exhibited in the case of the bimetallic sword, which indicates that the void could be one of the key conditions for the formation of redeposited copper. The redeposited copper and its various morphologies found in archaeological metal objects contribute to the authentication of works of art. Ying [22] suggested that redeposited copper is formed in a long natural corrosion process which cannot be simulated in a laboratory. But in recent years, Wang [23] claimed that the deposition of regenerated copper on simulated archaeological bronzes was simulated under experimental conditions. Copper (1) chloride could be reduced to pure copper if reduction times were sufficiently prolonged. However, the metallographic images of regenerated copper shown by the author did not exhibit the morphology that has been found in archaeological objects, such as the replacement of the eutectoid, the alpha phase, or voids by free copper. This simulation experiment has yet to be verified by further case studies. 7. Unusual corrosion products: phosphorus-containing compounds Cuprite, quartz, tin oxide as either cassiterite or romarchite, basic copper carbonates such as malachite are usual corrosion products found in archaeological bronze objects. The yellowish-green corrosion (Fig. 3 left above, Fig. 4 left) found in the junction of the back and blade was examined by X-ray diffraction; the dominant species are pyromorphite with some libethenite. Libethenite Cu2 (PO4 )(OH), is the most common copper phosphate found as a corrosion product of bronzes excavated from cremation sites, associated with decomposing bone and some other source of phosphorus [9]. Pyromorphite Pb5 (PO4 )3 Cl, a yellowish-green corrosion product was found in a Han Dynasty (c. 175BC) bronze An [24], in lead beads of the Early Bronze Age in Scotland [25], and a Roman coin [26]. Pyromorphite is uncommon in archaeological materials. The

mineral pyromorphite has been found in the oxidation zone of a polymetallic vein in Argentina [27], the appropriate formula could be Pb5 (PO4 , AsO4 )3 (Cl,OH)·xH2 O suggested by recent research [28]. Pyromorphite corrosion may be related to phosphatecontaining soil ascribed to degrading bones. Pyromorphite could be formed in the soil as a weathering product of lead-bearing minerals or inclusions. Phosphate amendments to the soil could induce further formation of pyromorphite [29], which is effective at immobilizing lead and reducing the associated bioaccessibility [21]. It is speculated that the bimetallic sword was buried in a tomb where the bone deteriorated to produce phosphates in a complex burial context. The phosphate and lead in the soil with chloride ions, for example, react to generate pyromorphite under environmental conditions distinct from that of lead carbonate produced in some Chinese bronzes. The lead dominantly transformed into pyromorphite and lanarkite Pb2 O(SO4 ) in the bimetallic sword case, whereas no cerussite was found. 8. Conclusions We studied the bulk composition, microstructure and corrosion phases of a bimetal bronze sword from Hunan Province, dating to the Warring States period (476–221 BC). The 16% tin content of the blade is not as high as that examined in a previous study which found 23% tin for the blade of the sword. It is necessary to examine a larger volume of core metal using preferably non-invasive analytical techniques such as neutron diffraction. The corrosion phenomena illustrated in the examination of the bimetallic sword suggest an unusual burial environment over a long period, with decomposed bone in the tomb generating insoluble pyromorphite. In this context, selective corrosion – both alpha phase and eutectoid had been corroded with the pseudomorphic original structure having clearly being replaced by malachite. Voids or cracks provided the optimum condition for the migration and redeposition of metallic copper in this sword. No bronze disease was found, so chemical treatment is not necessary in this case. In conclusion, the sword is stable, without active corrosion in terms of the analysis, so keeping the bimetallic sword in a dry and clean atmosphere would be appropriate for preserving the beautiful glazed green corrosion layers. Acknowledgments Thanks are due the conservation laboratory at the Cotsen Institute of Archaeology at University of California, Los Angeles (UCLA), The Getty Conservation Institute and Research Institute for providing the laboratory facilities and documental resources. The author would like to thank the China Scholarship Council and Peking University for financial support for his research at UCLA from 2007–2008. The samples were provided by Professor Zhou Weirong from the China Numismatic Museum, which is gratefully acknowledged. References [1] H. Rubin, K. Jun, The History of Chinese Science and Technology: Mining and Metallurgy (in Chinese), Science Press, Beijing, 2007, pp. 357–358. [2] C. Jianli, The manufacturing technology of bronze iron metallic artifacts unearthed from tomb 27 Liangdancun village site, Sci. China. E. Technol. Sci. 9 (2009) 1574–1581. [3] C. Peifen, Compositions and related casting technology of ancient bronze weapon and mirror, J. Shanghai Museum 1 (1981) 143–150 (Shanghai People Press). [4] Tangkun H, 1983. Several Compound Material Technologies of Metal Artifacts with Sharp Cutting Edge in Ancient China. Proceedings of the First Symposium on the History of Technology Ancient China. [5] P. Shifan, H. Jueming, W. Yuzhu, Bimetal bronze swords excavated from Jiangxi and their examination, Cult. Relics Central China 3 (1994) 101–103.

Please cite this article in press as: W. Huang, et al., The metallography and corrosion of an ancient chinese bimetallic bronze sword, Journal of Cultural Heritage (2018), https://doi.org/10.1016/j.culher.2018.10.004

G Model CULHER-3490; No. of Pages 7

ARTICLE IN PRESS W. Huang et al. / Journal of Cultural Heritage xxx (2018) xxx–xxx

[6] L. Haiping, T. Derui, Study on bimetallic bronze swords in ancient China, Sci. Conserv. Archaeol. 14 (2002) 319–334. [7] Haiping L, Derui T 2003. A study of the making of bronze weapons in the Wu and Yue states during the Eastern Zhou period in China. In Scientific Research in the Field of Asian Art: Proceedings of the First Forbes Symposium at the Freer Gallery of Art, Edited by Paul Jett with Janet G. Douglas, Blythe McCarthy, and John Winter, Archetype Publications in association with the Freer Gallery of Art, Smithsonian Institution, 47-52. [8] D. Zhongming, The technical study of bronze bimetallic sword dating to East Zhou period unearthed from Xintai City, Shangdong province, Sci. Conserv. Archaeol. 1 (2012) 75–86. [9] Kockelmann W., Godfrey E., 2019 (forthcoming). Neutron Diffraction (ND). In S.L. López Varela (Ed.), The SAS Encyclopaedia of Archaeological Sciences (Wiley-Blackwell). [10] H. Wei, W. Kockelmann, E. Godfrey, DA. Scott, W. Xiaohong, Non-destructive phase analysis of song dynasty iron coins by TOF neutron diffraction, Acta. Sci. Nat. Univ. Pekinensis 46 (2) (2010) 245–250. [11] M. Modlinger, E. Godfrey, W. Kockelmann, Neutron diffraction analyses of Bronze Age swords from the Alpine region: Benchmarking neutron diffraction against laboratory methods, J. Archaeol. Sci. Rep. 20 (2018) 423–433. [12] W. Kockelmann, S. Siano, L. Bartoli, et al., Applications of TOF neutron diffraction in archaeomety, Appl. Phys. 83 (2) (2006) 175–182. [13] D. Hanson, W.T. Pell-Walpole, The Chill-Cast Tin Bronzes, Edward Arnold, London, 1951, pp. 243. [14] W.T. Chase, Chinese Bronzes: Casting, Finishing, Patination, and Corrosion, in: DA Scott, J Podany, BB Considine (Eds.), Ancient and Historic Metals – Conservation and Scientific Research, Proceedings of a Symposium Organized by the J. Paul Getty Museum and the Getty Conservation Institute, Getty Conservation Institute, Malibu, 1994, pp. 85–117. [15] M. Taube, A.H. King, W.T. Chase, Transformation of ancient Chinese and model two-phase bronze surfaces to smooth adherent patinas, Phase Trans. 81 (2) (2007) 217–232. [16] L. Robbiola, J.M. Blengino, C. Fiaud, Morphology and mechanisms of formation of natural patinas on archaeological Cu-Sn alloys, Corrosion Sci. 40 (12) (1998) 2083–2111.

7

[17] D.A. Scott, Copper and Bronze in Art: Corrosion, Colorants, Conservation, The Getty Conservation Institute, Los Angeles, 2002. [18] R.J. Gettens, The corrosion products of an ancient Chinese bronze, J Chem. Educ. 28 (1951) 67–71. [19] Q. Wang, J.F. Merkel, Studies on the redeposition of copper in Jin Bronzes from Tianma-Qucun, Shanxi, China, Stud. Conserv. 46 (4) (2001) 242–250. [20] C. Bosi, G.L. Garagani, V. Imbeni, C. Martini, R. Mazzeo, G. Poli, Unalloyed copper inclusions in ancient bronze artifacts, J. Mater. Sci. 37 (2002) 4285–4298. [21] A.P. Schwab, K. Lewis, M.K. Banks, Lead stabilization by phosphate amendments in soil impacted by paint residue, J. Environ. Sci. Health Part A 41 (2006) 359–368. [22] J. Ying, S. Rongyu, H. Jueming, L. Ren, H. Yunlan, X. Menglong, Studies on mechanism of the redeposition of copper in corroded ancient bronzes, Sci. Conserv. Archaeol. 11 (2) (1999) 31–40. [23] Wang J, Xu C, Guocheng L, 2006. Formation processes of CuCl and regenerated Cu crystals on bronze surfaces in neutral and acidic media. Appl Surface Sci 252, 6294–6303. Kockelmann, W., Godfrey, E. (in press 2018). Neutron Diffraction. SAS Encyclopaedia of Archaeological Sciences. [24] X. Zhang, An unusual corrosion product, pyromorphite, from a bronze and a technical note, Stud. Conserv. 47 (1) (2002) 76–79. [25] M. Davis, F. Hunter, A. Livingstone, The corrosion, conservation and analysis of a lead and cannel coal necklace from the early bronze age, Stud. Conserv. 40 (4) (1995) 257–264. [26] G.M. Ingo, T. Decaro, C. Ricciccu, S. Khosroff, Uncommon corrosion phenomena of archaeological bronze alloys, Appl. Phys. Mater. Sci. Process. 83 (2006) 581–588. [27] I.L. Botto, V.L. Barone, J.L. Castiglioni, Characterization of a natural substituted pyromorphite, J. Mater. Sci. 32 (1997) 6549–6553. [28] R.L. Frost, S.J. Palmer, A Raman spectroscopic study of the phosphate mineral pyromorphite Pb5 (PO4 )3 Cl, Polyhedron 26 (2007) 4533–4541. [29] J. Cotter-Howells, Lead Phosphate Formation in Soils, Environ Pollut 93 (s) (1996) 9–16.

Please cite this article in press as: W. Huang, et al., The metallography and corrosion of an ancient chinese bimetallic bronze sword, Journal of Cultural Heritage (2018), https://doi.org/10.1016/j.culher.2018.10.004