Using 2D integral breadth to study plastic relaxation in a quasi-lattice-matched HgCdTe/CdZnTe heterostructure

Abstract

Micro-Laue diffraction has been used to record cross-section profiles on a quasi-lattice-matched HgCdTe/CdZnTe heterostructure as a function of the stress induced by a flexion machine. The heterostructure may be decomposed into four different regions according to depth. Sufficiently far from the interface, the CdZnTe substrate is undisturbed by the HgCdTe layer, while the region situated 10 µm beneath the interface presents an in-plane lattice parameter adjustment to the +0.02% mismatched layer. The layer has a 2 µm critical thickness and, beyond, misfit dislocations induce a large peak broadening whose main direction changes with depth. The same occurs over the whole heterostructure once flexion-induced plastification has started. Consequently, the usual full width at half-maximum or integral breadth is no longer relevant, and only a newly defined and rotationally invariant 2D integral breadth correctly measures the plastification-induced peak broadening. Taking into account only the critical thickness region, a 15.1 ± 0.7 MPa tensile HgCdTe elastic limit was measured, slightly overestimated because of the initial compressive layer stress. It was observed that the plastic onset of the substrate perfectly matches the elastic limit of the layer, despite the fact that the substrate elastic limit is expected to be four times higher: a striking demonstration of the propagation of threading dislocations. The `plastification easiness' is found to be 2.4 times smaller deep inside the substrate than in the layer critical thickness region, while in the substrate lattice adjustment region, the plastification easiness goes from the substrate to the layer value with a 22–25 MPa transition interval. This novel method using the 2D integral breadth allows for easy critical thickness measurement as well as precise plastic onset determination and plastification easiness assessment. It is a quite general method, since it may be applied to the vast class of epitaxial layers for which the critical thickness is larger than the micro-Laue beam size (currently 250 nm).