Decomposition of quasicrystalline phase formed in a rapidly solidified Al-Fe-V-Si alloy has been studied by TEM. The as-cast microstructure varies through the thickness of melt-spun ribbon; microeutectic precipitation of the bcc silicide near the wheel side, formation of globular quasicrystalline icosahedral phase with the microeutectic silicide phase at the middle of the ribbon, and the decomposition of quasicrystalline phase near the air side of the ribbon. During heating, as observed by annealing studies and by in-situ TEM studies, quasicrystalline phase decomposes into various phases such as aluminum, silicide and other unidentified phases. It has been shown that the preferential sites for the transformation are either at the center of quasicrystalline particles or at the quasicrystal/matrix interface, depending on the location of quasicrystalline particles.

Transmission electron microscopy (TEM) has been employed to examine the stability of the (presumed) icosahedral T2 (A16 Cu Li3) phase. The T2 phase was found to be unstable either when irradiated by the electron beam or during in-situ heating. In addition, certain specimen preparation techniques (e.g., ion-beam thinning) also led to the decomposition of the T2 phase. When the T2 particles were formed during conventional aging of aluminum-rich Al-Li-Cu based alloys, the transformation products were invariably microcrystalline. Individual microcrystals have been identified as the aluminum rich ∝-solid solution which, in certain instances, contained the δʹ (Al3Li) phase. TB (A17.5 Cu4Li) and T1 (Al2CuLi) particles were also found. When the T2 phase was prepared by casting alloys of the proposed stoichiometry of T2, then the transformation products were more complex, although certain reaction products have been identified as the a solid solution, TB and T1.

We have determined the structures and thermal stability of AI(6-x) CuxZr2 (0≤ × ≤ 1) powders prepared by mechanically alloying elemental powder mixtures. In the as mechanically alloyed condition the structure is L12 for all ×. After heating to 750°C the alloy is single phase D023 for 0 < × < 0.2, is a two-phase mixture of D023 and L12 for 0.2 < × < 0.6, and remains in the Ll2 phase for 0.6 < × < 1. Al5CuZr2 is stable in the Ll2 structure up to 1300°C. The lattice parameter of the annealed L12 phase is Independent of ×.

Neutron diffraction experiments on rare-earth transition metal magnetic alloys Er2Fe14B and Er2Fe17 have been carried out at temperatures above and below the ordering temperature (Tc). An anomalously large magnetic moment is observed at the crystallographic j2 site in Er2Fe14 B which is the intersection point of the major ligand lines in the crystal structure. The interatomic Fe-Fe distances are in the range of strong ferromagnetic bonds (≥2.66 Å). The analogous f site in Er2Fe17 does not develop as large a magnetic moment. In addition, the same sites show strong preference for Fe atoms in the respective substituted compounds. Due to poor phase stability of Er2 (CoxFe1-x)14 B compounds, iron substitution has been studied in detail in Er2(CoxFe1-x)17 alloys for site specific order and lattice distortion effects. However, a nonlinear change in the c lattice parameter observed in the neutron diffraction results cannot be explained on the basis of site preference alone.

The neutron refinement results indicate iron rich compositions in Er2(CoxFe1-x)17 materials, which is related to random substitution of Fe dumbbell pairs on the rare earth sites in the lattice. However, extensive electron microscopy (selected area electron diffraction and high resolution imaging) of Er2Fe17 and Er2 (Co .40Fe.60)17 failed to reveal any microscopic inhomogeneity. At this stage, the concept of direct negative exchange interaction between dumbbell Fe atom pairs at very short distances of −2.38 A is invoked and its effect is correlated with phase stability. We further suggest that the addition of boron in Er2Fe14B suppresses the substitution of dumbbell Fe pairs as a result of which Tc is raised significantly.