Microstructural characterization of official and imitative nummi of Vth century A.D

Metallography is an important tool that provides useful data on the fabrication technology, thermo-mechanical history of the object and on the nature of alloy employed. This research is part of a project aimed at reinforcing numismatic classification and description with chemical and microstrustural investigations. In this paper the attention has been focused on three bronze coins: Two Roman nummi struck under Arcadius/Honorius/Theodosius II in Rome (RIC, X, nn. 1271-1283, sample C9) and under Majorian in Ravenna (RIC, X, n. 2621,sample C26); the results of analyses on an italic imitation issued during the second half of V century AD are also presented, in order to evaluate possible connections between official and unofficial coins. The composition has been determined by XRF (Kevex 770) equipped whit a secondary target of Gd. The spectrometer operated at the following conditions: 55 kV, 1.00 mA. The microstructure of coins were investigated on metallographic cross-sections by light microscopy (Leica DM 100) and by SEM (Leica Cambridge Stereoscan 440) and analysed using the X-ray micro-analysis EDS (EDAX Philips) coupled to SEM. The EDS compositional profile is also obtained on the same coins to determine the element distribution and the concentration profile from the surface to the bulk. For EDS compositional profile the SEM operated at 25 kV. The semi-quantitative determination of element concentration was carried out standarless with the ZAF correction. In Tab.1 the bulk (b) and the surface (s) compositions (Wt%), determined whit EDS and XRF, are reported. The disagreement between XRF and SEM results are due to the different surface and bulk composition [2]. The high amount of Pb on the surface layer can be attributed to its preferential migration on the surface. The thickness and morphology of corrosion depend on chemical and physical properties of environment where coin was buried ; then the values obtained by SEM on the bulk (reported in tab 1) are reliable and show that specimens consist of a bronze alloy Cu-Sn with high rates of Pb. The content of Sn, shown in fig. 3, higher of 5%, identifies a western production [1, 10, 11], in contrast with what happens in the eastern mints. The micrographs show large grains flattened, evidence of the original cast microstructure with dendritic segregation; some slip lines were detected, confirming the plastic deformation (Fig 4, Fig 6, Fig 8). In Fig 5, 7, 9 BSE images show the directional preference of Pb. These characteristics can be attributed to identical production technologies for all the samples: The hypothesis is that the flans were obtained for solidification in the mold and then hammered to the desired thickness. Next, the plastic deformation could not be too strong because the Cu-Sn alloy whit an high amount of Pb is too brittle for further deformation. After being reduced in thickness, the flan was heated and coined. The process of hammering is evidenced by the shape of the grains that are not defined and polygonal but reflect a previous dendritic phase, which shows a fusion process. The presence of microsegregation zones shows the low working on metal surface after the melting process. XRF technique, instead, has not delivered the expected results in this work; in fact, the obtained data do not reflect the true composition of the sample but only the surface's layer composition of material