Correlation of interface morphology and composition in GaInP/GaAs with growth conditions

Abstract

Ternary (GaIn)P materials ststems grown on GaAs have attracted a lot of attention for laser applications, especially due to the low recombination velocities at the interface [1]. The physical properties of the interface are greatly influenced by the interface morphology, which can be controlled by either the growth temperature or growth interruptions or the introduction of interlayers, consisting e.g. of GaP. The interface morphologies of the semiconductor quantum wells can be characterized by the quantitative evaluation of high resolution high angle annular dark field (HAADF) images in scanning transmission electron microscopy (STEM). In the present work, quantitative evaluation of HAADF imaging in STEM is used to correlate the interface morphology and composition in (GaIn)P grown on GaAs with the growth conditions. The (GaIn)P/GaAs QWs were grown with metal organic vapor phase epitaxy (MOVPE) on GaAs (001) substrate at temperatures of 525°C and 625 °C, respectively, with different growth interruption times with or without GaP interlayer. In order to be able to compare different samples, a carefully applied method to gain reliable results from high resolution STEM micrographs was used. Also, to derive the chemical composition maps, the chemical sensitive background intensity is subtracted after image normalization as shown in Figure 1. From the composition maps, the interface features are revealed and then correlated with the growth conditions. The growth interruptions can significantly affect the composition fluctuation and the interface morphology. At higher temperature of 625 °C, with the two GaP monolayers between (GaIn)P and GaAs substrate, shorter growth interruption time leads to intermixing at the interface while a longer growth interruption results in a sharp interface. Also, without the GaP buffer layer, platelet islands can be observed at the interface. At lower temperature of 525 °C the GaP buffer layer has less influence. Hence, the quantitative evaluation of HAADF STEM images can reveal the interface morphologies, which also have important influence on the optoelectronic properties.