Phase separation and crystallization in amorphous Pd-Si-Sb

It is found that replacing some of the Si in a Pd-Si-Sb glass with Sb can make it more resistant to crystallization than the pure binary material. This is shown by the calorimetric results, in which the crystallization temperature increases and the glass temperature decreases with increasing Sb substitution, and also by the finding that certain alloys, such as Pd_80.5 Si_16.5Sb₃ may be quenched into amorphous rods (by the gradient-quench method) up to 0.5 mm thick.

Calorimetry and small-angle X-ray scattering data show that certain alloys in this system phase separate in the amorphous state. The two phases thus produced independently crystallize. Isothermal measurements indicate that the rate of crystallization of one of the phases is interface-limited, with most nuclei initially present. The other crystallization process does not fit an Avrami law, possibly due to a transformation of the crystalline phase.

A method called “gradient-quenching” (GQ) was used to prepare samples from the melt in such a way that the quench-rate was non-uniform, so all stages of crystallization in slowly-quenched alloys are displayed. Nucleation was frequently found to take place on voids and inclusions.

The use of the calorimetric and GQ methods in conjunction to gather information about crystallization processes in metallic glasses is discussed.