ID-7-P-2317 Three Dimensional EDS Elemental Mapping of Nonmetallic Inclusions of a Japanese Sword with an FIB-SEM

 Japanese swords are made of raw steel produced by smelting iron sand. The raw steel made by the Tatara method contains less P and S, and higher concentrations of nonmetallic inclusions (NI) than modern steel. By analyzing NIs the source of the iron sand and the heat process used during processing Japanese swords have been investigated. The purpose of this study is to reveal three dimensional (3D) distribution of NIs to know solidification process. The 3D distribution was observed by an FIB-SEM. A sample was repeatedly sliced to expose a new surface for analyzing with an EDS automatically. The 3D distribution was reconstructed from the acquired EDS data to analyze solidification processes of the NIs.

 The sample was a Japanese sword with the signature of Bizen Osafune Katsumitsu (property of M. Kitada). It was made in Japan in the 16th century. An EPMA was used to draw elemental maps of NI one cross-section (CS) of the sword. An FIB-SEM, JIB-4601F (JEOL), and an EDS (Oxford Instruments) were used to determine the 3D distribution of NIs. The SEM condition was as follows; the accelerating voltage (Acc. V.): 10 kV, probe current (P.C.): 14 nA, whereas the FIB processing condition; Acc. V.: 30 kV, P.C.: 10 nA, and ion dose: 250 nC/µm².

 An overall image of the CS after etching is shown in Figure 1. A BEI and elemental maps of O, Al, Si, and Ti in this CS are shown in Figure 2. The distribution of these elements is almost the same as show in Fig.2, which indicates that the NI was oxide. The NI in a red circle of Fig.1 was analyzed with the FIB-SEM. The 3D BEI reconstructed by the MIP method is shown in Figure 3(a). A superimposed 3D elemental map of Al, Si, and Ti is shown in Figure 3(b). The distribution of the NIs was clearly observed three dimensionally. As to the distribution of the elements in the NIs, Al rich areas were wrapped in Si rich areas, whereas Ti rich areas were relatively isolated.

 No cracks were observed in NI. This observation suggests that NI was melted during high temperature forging and then solidified during cooling. When the inclusion was cooled, Ti oxide precipitated first from molten oxide, successively the remaining molten oxide solidified as alminosilicate glass. In the area close to metallic iron with high heat-conductivity, Si rich glass solidified first followed by solidification of Al rich glass. This solidification sequence resulted in the observed Al rich areas wrapped in the Si rich areas, where Ti oxide particles were randomly distributed. The elemental distribution in the NIs suggests that the source of the iron sand was the one rich in ilumenite (FeTiO₃).

Reference: M. Kitada, Fine structures of a Japanese Sword Fabricated in the Late Muromachi Era (16th Century), Uchida-Roukakuho Tokyo (2008)27-36.