Compositional dependence of local atomic structures in amorphous Fe100-xBx (x=14,17,20) alloys studied by electron diffraction and high-resolution electron microscopy

Local atomic structures of rapidly quenched amorphous Fe100-xBx(x = 14,17,20) alloys have been investigated comprehensively by means of high-resolution electron microscopy (HREM), nanobeam electron diffraction (NBED), and electron diffraction atomic pair distribution function (PDF) analysis. In HREM images, crystalline cluster regions with a bcc-Fe structure extending as small as 1 nm were observed locally as lattice images, while NBED with a probe size as small as 1 nm revealed an existence of local clusters with structures of bcc-Fe and also of Fe-boride in all the as-formed alloys. Atomic PDF analyses were performed for these alloys by precise measurements of halo-electron diffraction intensities using imaging-plate and energy-filtering techniques. From the interference functions, atomic structure models were constructed for the Fe-B amorphous structures with the help of reverse Monte Carlo calculation. From Voronoi polyhedral analysis applied to these structure models, it was confirmed that atomic polyhedral arrangements with bcc and icosahedral clusters of Fe, and trigonal prisms of Fe and B, are formed in these amorphous structures, and the fraction of bcc-Fe clusters increases with the Fe content, while the fraction of trigonal prisms increases with the B content. The direct observation of local cluster structures of bcc-Fe and Fe-boride by HREM and NBED is an indication of "nanoscale phase separation" driven in the course of amorphous formation of these alloys, and the constructed structures based on the experimental PDFs with different B contents are inconsistent with a local structure scheme expected from the "nanoscale phase separation" model. The present study demonstrates that the structure model of nanoscale phase separation stands for the amorphous alloy structures where the phase separation fatally occurs in the crystallization stage.