1 Introduction

Considering the availability of advanced characterisation techniques like neutron diffraction, synchrotron X-ray diffraction and fluorescence, advanced field emission Scanning Electron Microscopy with Energy Dispersive X-ray Analysis (EDX) facility, investigating ancient coins using these advanced technologies would be useful in establishing historical interactions between east and west on trade and consequent cultural exchanges as well as the technological status of the era and so on. Studies of similar kind have been reported earlier [1–3]. Following the above objective, an ancient coin reportedly unearthed around the river bed of Tamirabharani located at the district of Tirunelveli in the state of Tamil Nadu of India has been investigated, for chemical and microstructural analysis. Present study of the coin aims at gathering information such as the chemical composition and microstructure that will help in deciphering the manufacturing methods that were in vogue during the era the coin was in use as well as corroborate any possible trade exchanges between South India and Rome.

2 Experimental

Figure 1 represents the photographs of the coin taken on both the sides. It can be observed from these photographs that the coin exhibits irregular surface. Further, detailed observations made using video microscope indicate presence of projections and depressions on the surface (Fig. 2). The coin might have been excavated and the irregular surfaces observed would have been due to corrosion of the coin that occurred below the soil or due to improper handling during the time of excavation. Alternatively, irregular surface finish of the coin might have taken place during the manufacturing stage of the coin itself.

Fig. 1
figure 1

Photograph of a tin coin (two sides)

Full size image
Fig. 2
figure 2

Photograph of the tin coin observed under video microscope showing irregular surface (projections and depressions)

Full size image

In order to establish the chemical composition of the coin, X-ray fluorescence (XRF) measurements were carried out using standard X-ray tube based XRF system [Philips make] and further using synchrotron radiation facility available at Raja Ramanna Centre for Advanced Technology (RRCAT), Indore. The incident monochromatic energy (calibrated with standard Si sample) used for this measurement was 19 keV. The XRF measurements were carried out in reflection mode using Vortex—EX (SII make) detector. The distance between the sample and the detector was maintained at 50 mm, and the angle between the incident X-ray beam and the sample was fixed at 45°, and the sample to detector angle was fixed at 45°. Further, X-ray diffraction (XRD) data was gathered using the synchrotron facility whose details were discussed elsewhere [4]. By careful metallographic preparation of the thin edge (~2 mm) of the coin, microstructural analysis was carried out using optical microscope and Scanning Electron Microscope (SEM).

3 Results and Discussion

3.1 XRF Technique for Chemical Analysis

The preliminary analysis of the coin believed to be belonging to the ‘Sangam Period’ was analysed using XRF technique which provides semi quantitative information about the concentration of the elements. The XRF spectrum was analyzed using the standard code of PyMCA [5]. The elements identified include Sn, Fe, Ni, Pb, Zn and Cu. This suggests that the coin belongs to Sangam period as literature reports that usage of tin bronze was in vogue in south India during the 3BC–2AD [2]. One side of the coin is relatively smooth and flat as compared to the other rough side. Further detailed analysis was carried out using synchrotron source (Fig. 3) which confirmed the elements present in the coin, as shown in Table 1. As the elements can be present in different phases which may be either solid solutions or intermetallics, it is prudent to examine the phases present. Among the different elements observed, Fe and Mn were on the higher side whereas Ni, Zn and Pb were in traces. Therefore the formation of intermetallics of Sn with Fe or Mn is likely [6]. In order to establish the various phases present in the coin, X-ray diffraction analysis was carried out using synchrotron facility.

Fig. 3
figure 3

XRF spectrum of the coin gathered from synchrotron radiation. The major peak belongs to Sn. Other distinct peaks are identified as Mn, Fe, Cu, Ni, Zn and Pb. (Color figure online)

Full size image
Table 1 Elements present in the coin
Full size table

3.2 XRD Analysis

XRD measurements were carried out using the synchrotron facility Indus-2 beam line 12 at the wavelength of 0.94825 Å. Figure 4a shows the X-ray diffraction pattern of the coin. It is found that, there are two types of peaks, one type is very sharp with low FWHM and the others are relatively broadened peaks. The peaks were identified using the ICDD JCPDF [7] and the sharp, high intensity peaks were well matched with pure Sn (ICDD #04-0673) (Tetragonal crystal system with space group of I41/amd (141)). In order to get the accurate lattice parameters, the XRD data was analysed by Rietveld method [8] with JANA2000 [9] program. The background was modeled with Legendre polynomials with 15 terms and pseudo voigt profile peak function was used for the refinement. The typical Rietveld plot of Sn coin was shown in Fig. 4b. The plot consists of observed, calculated and difference in X-ray powder diffraction profile of tin coin. Crosses indicate the observed data and the solid line shows the calculated profile. All possible Bragg reflections are indicated with vertical markers beneath. The refined lattice parameters of the Sn coin were found to be a = 5.82599(12) and c = 3.17824(7) Å. The lattice parameter of pure Sn was 5.8313 and 3.1815 for ‘a’ and ‘c’ respectively reported in the literature [10]. In comparison, the data observed for tin in the present study exhibited a reduction of the lattice parameters which may be attributed to trace amounts of impurity elements like Fe, Mn, Ni, Pb, Zn forming solid solution with tin, considering that the XRD results do not suggest the presence of other intermetallic compounds [6].

Fig. 4
figure 4

(a) XRD pattern gathered from Synchrotron facility indicating presence of Sn as major element and Cu6Sn5. (b) Typical Rietveld plot of synchrotron X-ray diffraction data shows the peaks match well with characteristic Sn lines

Full size image

Further, it is observed that very small intensity peaks (less than 1 % compared with highest major peak intensity) were analysed for possible line compounds such as A3B type. It is observed that the broadened, very weak intensity peaks at the 2θ value of, 18.237, 19.372 and 23.296 match with the line of Cu6Sn5 (Monoclinic, C2/c). Therefore, all these three peaks are assigned to Cu6Sn5 compound. Further, based on the intensity ratio of Cu6Sn5 line with reference to the 100 % line of Sn, the estimated amount of Cu6Sn5 is <1 % in quantity.

3.3 Microstructural Analysis by Conventional Metallography

A wooden fixture (Fig. 5) was employed for polishing the edge of the tin coin for the microstructural analysis. The coin was held between two wooden blocks using a mini vice. In-situ polishing was adopted to polish the thin edge of the coin. This ensured that the faces of the coins are not damaged, since the coin is not moulded in any resin. In-order to reveal the microstructure, well-polished surface of the coin was etched using an etchant with the following composition: 2 % HCl plus 5 % HNO3 in methanol. Figures 6 & 7 show the microstructures of the coin in two different magnifications. The microstructure consists of two phases which are bright and dark in contrast. Further, non uniformity in grain size was observed. This suggests that hot working methods were possibly employed for the manufacture of the coin, as non-uniform metal flow (or strain) during hot working operations is probably the major source of evolution of grains of different sizes [11].

Fig. 5
figure 5

Photograph depicting a wooden fixture to hold thin edge of the coin for metallographic polishing

Full size image
Fig. 6
figure 6

Optical micrograph of tin coin exhibits two-phase microstructure

Full size image
Fig. 7
figure 7

Higher magnification of the same region reveals the details of the different microstructural entities

Full size image

3.4 Scanning Electron Microscope (SEM)

Scanning electron microscopy was carried out on the well-polished and etched tin coin, using SEM (Model No. XL30ESEM, M/s FEI, The Netherlands) operating at a voltage of 30 kV. The photo micrograph depicts the typical microstructure of the coin showing a two-phase structure (Fig. 8). On observing at higher magnification (as shown as inset in fig. 8), some of the grains show bright and dark regions. Energy Dispersive Analysis of X-rays (EDX) (Fig. 9) indicates Sn alone and no peak corresponding to Cu was found. Since no difference in composition of the individual microstructural entity could be recorded, the labyrinth structure (as seen in Fig. 8 inset) observed in the microstructure is attributed to be due to not-fully broken cast structure. Further, as the Cu6Sn5 phase is estimated to be less than 1 %, the Cu content is so low that it was not detected in the EDX spectrum.

Fig. 8
figure 8

SEM micrograph of the tin coin showing bright and dark regions in the microstructure

Full size image
Fig. 9
figure 9

Comparison of spectra from bright and dark regions referred in Fig. 8.shown as black and red colors respectively in overlap mode. (Color figure online)

Full size image

Further, in order to understand the two-phase microstructural observation and also to explain the non-revelation of the chemical difference among the two distinct microstructural entities, the phase diagram of Sn-Cu has been referred which is shown in Fig. 10. A line corresponding to this composition (Sn with less than 1 %Cu) would fall in the two phase region: (i) Sn and (ii) eutectic made of intermetallic Cu6Sn5 plus tin [12]. Based on the lever rule, the amount of intermetallic which separates out from the melt as eutectic must be <1 % and Sn would be more than 99 % in volume fraction. Since the volume fraction of the compound Cu6Sn5 is <1 % and the particles were expected to be finely dispersed in the matrix, their unique identification by imaging and spot analysis through EDX would be difficult. Therefore, the presence of Cu is not distinctly seen in the EDX spectra of different morphological microstructural entities.

Fig. 10
figure 10

Phase diagram shows the presence of eutectic of Sn and Cu6Sn5 at the extreme tin rich end (99.1 %); possible coin composition is marked by dotted line on the diagram

Full size image

Presence of Cu6Sn5 in small quantity strengthens the soft tin matrix [13]. Tin being a low melting metal, additions of Cu may lead to strengthening of the matrix as fine distributions of intermetallic Cu6Sn5 form. Presence of intermetallics is considered to provide through transient liquid phase bonding during hot working [14]. It is speculated that during hot working, the non uniformity in deformation is generated due to the presence of occluded small quantity of Cu6Sn5, leading to non-uniform grain size distribution (Figs. 6 and 7).

4 Conclusion

  1. (i)

    The chemical composition of the coin obtained using XRF technique revealed the presence of tin as the major element and the presence of other elements like iron, nickel, copper less than 1 %.

  2. (ii)

    XRD results confirmed the presence of Sn in the coin as the major phase and a small amount of Cu6Sn5. This is due to the composition chosen for the coin being close to the Sn-Cu6Sn5 eutectic point (99 % Sn).

  3. (iii)

    Metallography showed the presence of grains with non-uniformity in size distribution. Based on the microstructural studies, it is concluded that the coin has been manufactured by casting followed by hot working method.

  4. (iv)

    The contrast suggesting presence of two phases is attributed to possible texture effect that arises from labyrinth structure due to casting, since the regions in distinct contrast did not exhibit any difference in the chemical composition as revealed by the EDX spectra.