It was recalled that recent transmission electron microscopic observations had shown that defects, which exhibited a contrast that was similar to that expected for antiphase domain boundaries, were observed within the massive -phase in quenched samples of alloys that were based on TiAl. It was also shown that a thin region of 90 domain was associated with such boundaries. In the present work, a mechanism was proposed for the formation of these complex boundaries. Image calculations which were based upon this model were shown to be in agreement with experimental observations. It was concluded that these domains nucleated from a metastable face-centered cubic phase, and did not form directly from the high-temperature equilibrium hexagonal -phase. During annealing, these boundaries became crystallographic before individual domains shrank, so that the antiphase domain boundaries were eventually removed. Similar observations were made of the structures of antiphase domain boundaries in other intermetallics, and the structures were shown to vary from pure antiphase domain boundaries in CuAu, to complex antiphase domain boundaries with various widths in the other alloys. It was suggested that the structures of antiphase domain boundaries in the L10 structure were a function of the c/a ratio and of the ordering energy.

X.D.Zhang, Y.G.Li, M.J.Kaufman, M.H.Loretto: Acta Materialia, 1996, 44[9], 3735-47