Growth and Characterisation of II-VI Quantum Dot Resonant Tunneling Diodes

The work presents the growth by molecular beam epitaxy (MBE) and characterization of CdSe quantum dot (QD) resonant tunneling diodes (RTDs) based on the group II-VI (Zn,Cd,Be,Mn)Se material system. The key requirement for observing interesting physics like single electron transport, giant spin-splitting, Fermi edge singularity (FES), and a quantum phase transition (QPT) is to find the right CdSe coverage and tunnel barrier parameters. The very narrow range of these parameters requires high accuracy and a detailed and systematic understanding of the growth. At first, the pseudomorphic growth of ZnSe on GaAs substrate is achieved using growth parameters taken from earlier work in our group. Nomarski microscopy is used in the study of defects and the optimization of GaAs buffer and ZnSe layers. A high Se-to-Zn flux ratio of 5 for the initial 1-minute growth of the ZnSe-based layers on top of the GaAs buffer significantly reduces the density of defects such as stacking faults and dislocations. The highly strained Zn\(_{0.71}\)Be\(_{0.29}\)Se tunnel barrier layers are calibrated using the X-ray interferometry (XRI) technique and their direct reflections. The barrier layer calibration samples grown with identical growth conditions on different days showed nearly identical parameters in the dynamic simulation of X-ray diffraction (XRD) data, proving the stability of growth conditions. The CdSe growth rate is calibrated just using the XRI technique due to its high lattice mismatch (7 %) with the GaAs substrate and the very low layer thickness. The reciprocal space map (RSM) of the asymmetric (115) reflex of the sample showed pseudomorphic growth of the layer structure. The morphological study of a sample with 2 monolayers (ML) thick CdSe layer using scanning electron microscopy and atomic force microscopy showed the formation of CdSe QDs with a density of around 2 \times 10\(^{10}\) / cm\(^2\). The photoluminescence and scanning transmission electron microscope (STEM) images of similar samples also showed the presence of QDs. The CdSe layer thickness obtained by XRI calibration between ZnSe cladding layers is defined as a nominal thickness or coverage. A large reduction of about 35 % is observed in CdSe thickness when the nominal thickness of CdSe is stacked between Zn\(_{0.71}\)Be\(_{0.29}\)Se layers. The observed deviation of the CdSe coverage from the nominal value increases with the Be content in the tunnel barrier layers. In a series of samples with increasing Be content in the barrier layers while keeping the nominal CdSe thickness constant, we observed a decreasing CdSe thickness together with a reduction of the Be content in the barriers. Another series of samples with fixed barrier parameters (5 nm Zn\(_{0.71}\)Be\(_{0.29}\)Se) and varied CdSe coverage also shows a correlation in which the CdSe thickness observed is 65 % of the nominal values. Geometric phase analysis (GPA) of STEM images from representative samples from these series revealed a deviation in agreement with the XRD studies. The vertical lattice constant of the CdSe-rich layer of these samples is significantly affected by the Be content in the barrier. The thickness of the CdSe-rich layer in the GPA analysis showed a similar value in both samples (CdSe in ZnSe and in Zn\(_{0.71}\)Be\(_{0.29}\)Se) contradicting the possibility of CdSe desorption and indicating another mechanism for explaining the above correlations. Homogenous intermixing or diffusion between CdSe and barrier layers at a length scale comparable to the barrier thickness may reduce the phase shift as observed in the XRD studies and is consistent with the STEM GPA analysis. A possible local strain relaxation in the QDs is not observed in the RSM data. It may also cause such deviation in observed CdSe layer thickness, but it cannot explain the observed reduction of strain and Be content in the barrier layers. The real mechanism cannot be concluded using the XRD and the STEM techniques used, since different compositions of (Zn,Cd,Be)Se can create the deviation. (Zn,Mn)Se quantum well (QW) RTD samples are reproduced to compare them with older samples to check the reproducibility of the parameters. Then we replaced the QW by the CdSe QD layer and analyzed the data of the 2\(\Theta\) - \(\omega\) scan of the (004) reflections. The Bragg peak of the (Zn,Be)Se barrier layers is modified by the CdSe layer. The angular position of the minimum in intensity and the ratio of the peak intensities within the Bragg reflection of the (Zn,Be)Se barrier layers help to determine the CdSe coverage with 0.1 monolayer (ML) accuracy. We have grown RTD samples, where we vary the CdSe coverage systematically. XRD data from a series of RTD samples where the CdSe coverage is varied in steps of 0.2 ML with constant barrier parameters showed the reproducibility and accuracy of the CdSe thickness and barrier parameters. The current-voltage (I-V) characteristics of RTD samples with nominal CdSe coverage between 1.3 ML and 2.3 ML showed the following results: A low CdSe coverage RTD sample (1.3 ML in ZnSe) showed few resonance peaks in I-V at high bias voltage. A high CdSe coverage RTD sample (2.3 ML in ZnSe) showed numerous resonance peaks and cannot be used for I-V studies of the resonances of a single QD. A CdSe coverage between 1.7 ML and 2 ML in ZnSe layers is found to be optimum for getting few resonances at low bias voltage. We further studied the influence of the position of QD layer with constant CdSe coverage between the barrier layers with symmetric 5 nm / 5 nm, as well as asymmetric 4.7 nm / 5.3 nm or 5.3 nm / 4.7 nm thicknesses. The XRD data of these samples showed marginal differences in the position of the intensity minimums of the (Zn,Be)Se layer reflection because the thickness differs only by 0.3 nm. The introduction of isoelectronic Mn atoms into the barrier layers facilitated the formation of a bosonic bath of magnons in the QD RTD structure. The I-V characteristics show a giant Zeeman splitting and finally the first realization of a QPT in a magnetic CdSe QD RTD with 3.5 % Mn in the tunnel barriers.