On the diffusion of lattice matched InGaAs/InP microstructures

Abstract

Photoluminescence and high-resolution x-ray diffraction ͑HRXRD͒ studies of the diffusion in lattice matched InGaAs/InP quantum wells show that at high temperatures intermixing can be modeled by Fick's law with an identical diffusion rate for both the group III and group V sublattices. This results in materials that remain lattice matched for all compositions created by the diffusion. At lower temperatures, the photoluminescence shows that the diffusion process changes and HRXRD shows that strained layers are produced within the structure. This may be due to the presence of the miscibility gap within the InGaAsP phase diagram.

Photoluminescence and x-ray diffraction studies of the diffusion behavior of lattice matched InGaAs/InP heterostructures

“…Given that the HRXRD data presented in this work show that the two sublattices diffuse at different rates at all temperatures below 900°C the diffusion lengths quoted in Ref. 4 cannot be correct.…”

“…In Ref. 4 it was stated that for these samples annealed at 800, 850, and Experimental 004 x-ray rocking curves for a sample from wafer B annealed at 800°C for various times. From qualitative modeling of HRXRD the broad feature to the right of the substrate peak for samples annealed for 1000 and 2400 s is indicative of tensile strain in the well.…”

“…3 Photoluminescence and high resolution x-ray diffraction ͑HRXRD͒ characterization of the intermixing of a 10 nm lattice matched InGaAs/InP quantum well at various temperatures 4 clearly showed that the miscibility gap had no noticeable effect on the diffusion at a temperature of 900°C. But at 600°C, the HRXRD rocking curves showed that strain had developed in the annealed specimen.…”

Characterization of interdiffusion around miscibility gap of lattice matched InGaAs∕InP quantum wells by high resolution x-ray diffraction

“…Previous studies, that also used high resolution x-ray diffraction ͑HRXRD͒ as a tool of characterization of diffusion, showed that a model such as Fick's laws with a constant coefficient of diffusion could not be taken for granted for lattice matched InGaAs/ InP heterostructures. 5,6 And indeed, another study on the a priori straightforward InGaAs/ GaAs system clearly showed that indium is not a simple marker of diffusion, and that diffusion cannot be modeled with Fick's laws and a constant coeffi-cient of diffusion. 7 Hence, the latter model is highly unlikely to be suitable at all to the InGaAs/ InP system, which is further complicated by interdiffusion on two sublattices ͑of group III and V atoms͒ and the presence of a miscibility gap in the InGaAsP quaternary diagram.…”

Concentration dependent interdiffusion in InGaAs∕GaAs as evidenced by high resolution x-ray diffraction and photoluminescence spectroscopy

“…Despite this, there have been reports in the literature that have suggested that the diffusion might be concentration or strain dependant. 1,3,4 It is also true that departure from the above linear variation can be difficult to detect because the nonlinearity is not always dramatic 5 and also because of a lack of data or scatter in the data points.…”