Search

The mopholological, structural and chemical properties of InxGa1–xN nanopillars directly grown on Si (111) substrates, by molecular beam epitaxy, were investigated employing transmission electron microscopy related techniques. Single crystalline, single phase nanopillars were observed exhibiting a low density of crystal defects, which contribute to good crystal quality. Initial nanostructures merge through subgrain boundaries to form final nanopillars. Energy dispersive X-ray analysis revealed a very low InN mole fraction near the interface with the substrate, owing to high desorption rates from the elevated growth temperature, and gradually higher In incorporation rates near the tips of the nanopillars. This compositional fluctuation is maintained due to poor segregation of indium adatoms along the c-axis of the nanopillars towards the Si interface. A second species of long and narrow nanopillars was found In-free. (© 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)

Freestanding c-plane GaN layers were grown by hydride vapor phase epitaxy on γ-LiAlO2(100) substrates. The layers appear transparent and colorless. However, they contain a large number of dodecagonal pinholes (V-pits), which are surrounded by brownish GaN pyramids. It is shown that the V-pit formation occurs at the top of N-polar inversion domains, which appear in the Ga-polar GaN matrix. The GaN layers contain threading dislocations, which are partly attracted to the V-pit surface in order to minimize the dislocation energy. The appearance of brown GaN pyramids around the V-shaped pinholes seems to be a direct consequence of impurity incorporation, which might stem from the LiAlO2 substrate. (© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)

ZnTe nanowires were grown via a vapour-liquid-solid pro- cess. Nano-sized droplets of a gold-based eutectic act as catalysts. The comprehensive transmission electron microscopy studies reveal that the nanowires are single crystals with numerous stacking faults and twins. The dimension of the wires is several micrometers in length and a few tens of nanometers in diameter. At the sidewall of the nanowires additional nanocrystals of ZnO embedded in an amorphous layer are identified. The formation process of the nanowires can be understood as a two-step process. The first step is the one-dimensional growth along the wire axis by consuming all the material deposited near the droplet. In a second step, facets are formed due to lateral growth of the nanowire. (© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)

III-nitride semiconductor nanostructures are subject of intense studies with respect to their optoelectronic, structural and chemical properties. Important parameters for the wavelength of the emitted light are the chemical composition and the dimensionality of the nanostructures. Transmission electron microscopy is used to determine these characteristics at a nanometer scale. In this work, the information provided by diffraction contrast images of (In,Ga)N/GaN quantum wells (QWs) is studied. Experimental dark-field images alternatively using the 0001 and the 0002 reflection show a different contrast regime. In order to understand the contrast, one has to calculate the intensity Ig of the individual diffracted beam g. The intensity of the 0002 beam is a function of the sum of the atomic scattering amplitudes of the group III and the group V element. Consequently, the 0002 reflection is strain sensitive. According to the kinematical theory the 0001 reflection is forbidden. However, it is excited in the experiment. Therefore, dynamical effects have to be taken into account. The corresponding intensity is calculated by the Howie-Whelan equations. It turns out that the intensity of the 0001 beam strongly depends on the In content of (In,Ga)N. A good agreement of intensity profiles of 0001 dark-field images compared to the theory is found. (© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)

The study reports on the structure of thin ternary NiGaxSi2–x silicide layers grown on Si(001) by Ni/Ga codeposition at room temperature followed by subsequent annealing. The influence of both the Ga content and the annealing temperature on the phase formation and orientation of the layers was investigated for the first time. Addition of Ga leads to a decrease of the disilicide formation temperature down to less than 500 °C. Island formation is observed for the Ga content of 0.17 after annealing above 900 °C. The NiSi2-xGax islands grow with a NiSi2–xGax(110) ‖ Si(001) texture. In contrast, closed epitaxial NiSi2–xGax layers of A-type are obtained for the Ga content of 0.28 after annealing at 900 °C. (© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)

The growth properties of InN quantum dots on GaN (0001) surfaces by molecular beam epitaxy are being investigated. The dependence of the dimensions and density of the dots on the nucleation temperature and their evolution during growth at a constant substrate temperature are described. It is shown that both dimensions and density can be accurately controlled through nucleation temperature and deposition time. In the range from 400 °C to 450 °C, the formation of InN quantum dot structures of small dimensions and high density is feasible. InN dots with less than 3nm height, less than 22 nm diameter and with density higher than 1.7×1011 cm–2 have been obtained. Finally, vertical alignment of dots was observed for a multi-period InN quantum dot structure in GaN matrix, grown at 430 °C. (© 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)

In this paper we have investigated the effect of CF4 plasma treatment on AlGaN/GaN HEMTs and compared it with Ar plasma treated devices. We have investigated the electrical characteristics of the devices using transistor characteristics as wells as CV measurements in order to verify the impacts of the different treatments on the electron concentration in the 2DEG. Furthermore, a TEM-EDX analysis was performed to find out if any fluorine incorporation in the AlGaN barrier during the CF4 plasma treatment occurs. No significant impact of the fluorine on the threshold voltage could be determined, We conclude that fluorine incorporation is not responsible for the threshold voltage shift. (© 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)