Direct compositional analysis of AlGaAs/GaAs heterostructures by the reciprocal space segmentation of high-resolution micrographs

A detailed description of a combined reciprocal and real space technique for the mapping of layer compositions across interior interfaces from high-resolution electron micrographs is presented. The analysis is based on the reciprocal space extraction of chemically sensitive image information encoded in lattice images of AlGaAs/GaAs heterostructures taken under optimized imaging conditions. Analysis procedures include centering a set of apertures around chemically sensitive reflections in the Fourier transform of lattice images and performing an inverse transformation, thus extracting composition related information from experimental micrographs. It is demonstrated that this approach is characterized by the same spatial resolution as real space techniques but by improved capabilities with respect to analysing images characterized by a minor signal-to-noise ratio. For illustration purposes the stability of AlAs/GaAs multiple quantum wells grown under low-temperature conditions against thermal treatment as expressed by interfacial roughness parameters is investigated.