Left: the band alignment of GaAs, BeTe, ZnSe, and CdS, assuming a valence-band offset between GaAs and ZnSe of 1 and between ZnSe and CdS. A huge band offset as large as has been achieved by combining BeTe and CdS. Right: energy band diagram of the conduction band for the SL. The arrow shows the ISB-T process from the ground state (e1) to the second state (e2) in the SLs.
The in situ RHEED oscillations measured along the  azimuth during the initial stages of the growth of SLs on the GaAs substrate with ML BeTe buffer. Long pronounced RHEED oscillations can be detected by inserting a 2 ML ZnSe interlayer (top) as well as a 0.5 ML ZnSe intermediate layer (middle). However, the RHEED oscillations disappear and the specular spot intensity decreases permanently without the ZnSe interlayer (the bottom).
(002) HRXRD patterns of superlattices for S1–S6. The dotted lines are drawn as a guide to the eyes to show the envelope. The lattice mismatch with the GaAs substrate for each SL sample is given and indicated next to its corresponding scan curve. The arrows mark the modification in the envelope.
Top: absorption spectra, calculated from transmission spectra and measured for sample S3 using -polarized and -polarized lights. A strong ISB-T of occurs in S3 with a 1 ML ZnSe interlayer; however, the ISB-T disappears from the SLs (sample S4) with a 0.5 ML ZnSe interlayer. The insert at the upper right shows the multipass waveguide geometry for the intersubband absorption measurements. Bottom: absorption spectra for S1–S6 obtained by taking -log of the transmission ratio of -polarized to -polarized light to eliminate the influence of the background. The ISB absorptions shift the towards the shorter wavelength with a decrease in the thickness of the ZnSe interlayer in S1–S3. The full width at half maximum values of the absorption spectra are given. The ISB absorptions disappear in samples S4, S5, and S6 with the 0.5 ML ZnSe interlayer.
Cross-sectional high-resolution transmission electron microscopy image of sample S3. A sharp interface is formed between the barrier and the well layer in SLs with the thickness of ZnSe interlayer ML. The fluctuation in layer thickness is about 1 ML.
HRTEM images of S4 and S5 reveal that the 0.5 ML ZnSe interlayer results in different properties in the SL structure. A rough interface with a greater thickness ( ML) is formed.
RHEED patterns taken along the  azimuth at the end of the SL growth: (a) 7 ML CdS/2 ML BeTe heterostructure without the ZnSe interlayer; (b) A typical RHEED pattern ( azimuth) with ZnSe interlayer. The streaky pattern, as well as the Kikuchi bands, confirm a 2D growth with a smooth surface. (c) RHEED intensity oscillations recorded from the onset of growth along the  azimuth. The growth mode degrades immediately as soon as the deposition of BeTe starts on the CdS surface. The streaky pattern changes to a spotty one.
Schematics of the interface properties with an integer ZnSe IL (a) and 0.5 ML ZnSe IL (b). A broad transition region is formed in SLs with 0.5 ML ZnSe, while a sharp interface is formed in those with ZnSe IL ML. (b) Also presents the growth model for CdS (or BeTe) growth on BeTe (or CdS) surfaces covered with 0.5 ML ZnSe. The two-dimensional ZnSe islands become the nucleation zone for the subsequent growth of BeTe (or CdS) on CdS (or on BeTe) to maintain the constant 2D growth mode.
The structure parameters for the designed superlattices. We also show the results of ISB-T. In S1–S3, ISB-Ts are observed (O), while is S4–S7, ISB-Ts disappeared (X).
Structural parameters of superlattices obtained from HRTEM images. For comparison, the structural parameters calculated from XRD dynamical simulations are also shown.
Elastic constants, lattice mismatch, and strain for BeTe, ZnSe, and CdS. The mismatch for BeTe with respect to ZnSe and CdS is defined as and . The strain for BeTe, grown on ZeSe or CdS, is defined as and . Same rules are also applied to the mismatch and strain for ZnSe (CdS) growing on BeTe (or ZnSe) and CdS (BeTe).
Article metrics loading...
Full text loading...