Abstract
Composition analysis at the nm-scale, marking the onset of clustering in bulk metallic glasses, can aid the understanding and further optimization of additive manufacturing processes. By atom probe tomography, it is challenging to differentiate nm-scale segregations from random fluctuations. This ambiguity is due to the limited spatial resolution and detection efficiency. Cu and Zr were selected as model systems since the spatial distributions of the isotopes therein constitute ideal solid solutions, as the mixing enthalpy is, by definition, zero. Close agreement is observed between the simulated and measured spatial distributions of the isotopes. Having established the signature of a random distribution of atoms, the elemental distribution in amorphous Zr(59.3)Cu(28.8)Al(10.4)Nb(1.5) samples fabricated by laser powder bed fusion is analyzed. By comparison with the length scales of spatial isotope distributions, the probed volume of the bulk metallic glass shows a random distribution of all constitutional elements, and no evidence for clustering is observed. However, heat-treated metallic glass samples clearly exhibit elemental segregation which increases in size with annealing time. Segregations in Zr(59.3)Cu(28.8)Al(10.4)Nb(1.5) > 1 nm can be observed and separated from random fluctuations, while accurate determination of segregations < 1 nm in size are limited by spatial resolution and detection efficiency.