Rare-earth oxide single crystals for a magnetic refrigeration are reviewed The magnetic refrigeration below 20 K using the Carnot cycle is progressed surprisingly for the purpose to produce a liquid helium (4.2 K) or superfluid helium (1.8 K) replaced by a gas refrigeration. The magnetic refrigeration is mainly used for the paramagnetic properties of antiferromagnetic materials such as rare-earth garnet or perovskite single crystals. The point of research and development are to improve growth conditions and to control physical properties of the single crystals as well as to obtain a large size bulk single crystals.
Figures etched by KOH aqueous solutions were observed on a Si crystal surface, and the relations between the etch pits and surface defects were discussed. Surface roughness depended on the crystal orientation and concentration of the etching solution. Etching 4°off the (100) surface by 17% solution at 92℃ showed a folded pattern, consisting of rows small overlapping L shape pits. While by etching exactly on the (100) surface, this pattern was not seen. A lapped wafer with mechanical damage had an accelerated etch rate. This alkali etching is very sensitive to surface defects such as cracks and scratches, as delineated by square pits on the (100) surface. Surface contamination was appeared as small hillocks. A KOH etched (111) surface observed by an atomic force microscope revealed triangular pits and atomic steps with two atomic layers deep in spite of being exposed to air for several hours.
In our study of molten Si, the energy dispersive X-ray diffraction technique revealed the reduced intensities structure factor and temperature dependent change of the nearest interatomic distance and the coordination number as compared with the previous findings. A small angle scattering technique showed that there is no evidence of cluster formation in Si melt near melting temperature against an expectation. The extended X-ray absorption fine structure spectroscopy revealed that impurity level gallium atoms are tightly bound by three Si atoms in Si melt.A measurement of Si melt density after melting gave a clear tendency of relaxation with an equilibration period of 3 hours. The influence of the relaxation time to grown crystals was studied and found evident in formation of point defects. The density measurement after relaxation revealed that the anomalous change takes place in the temperature range up to 15 degress higher than the melting point. In this region, thermal expansion was calculated about an order of magnitude lager than that in the higher temperature region. The property anomaly in the same temperature range was also found for surface tension and viscosity. Addition of 0.1% boron did not show any effect while that of 0.1% gallium or antimony wiped out the anomaly. We found that in highly antimony doped Si melt a volatile species of Sb_2O forms in proportion to the square of Sb concentration and evaporates. Also determined are the evaporation rates of SiO and Sb_2O. Electrical resisticity, thermal diffusivity and spectral emissivity were also studied and it was deduced that four electrons from each Si atom were delocalized liberated into the bulk of the melt. Conventional computer simulation techniques were closely checked and the result was compared with that by the X-ray fluoroscopic observation of Si melt in a crucible. They agreed with each other qualitatively. The calculation using the newly determined evaporation rate revealed that the experimentally found radial oxygen distribution in the crystal agreed well with the calculation result. The simulation algorism was improved by analyzing a three dimensional mesh system and introducing a computation technique called k-ε model. The result of the calculation based on our new algorism was found remarkably consistent with the experimentally determined temperatures at some localities of Si melt in the crucible.
The practical application of GaAs on Si technology into combining optoelectronic and high-speed devices is our hopefully rewarding long-term objective. To achieve this, the defect density in GaAs on Si needs to be reduced to values obtainable in bulk or homoepitaxial GaAs (≦10^4/cm^2). We discuss in the present paper recent progress of GaAs on Si technology from the viewpoint of how to be able to suppress the threading dislocation density in GaAs layers on the basis of the results which were obtained throughout the course of investigation in OTL. In particular, the effects of new materials for buffer layers and insertion layers, doped-impurities, growth area confinement, substrate orientation and high-temperature annealing on the reduction of threading dislocation generation and propagation, are described. These results are discussed by considering two groups of misfit dislocations at the interface regions between GaAs and Si.
Issues of the heteroepitaxy of group III nitrides, especially heteroepitaxy of nitrides on highly lattice-mismatched substrates using a buffer layer and heteroepitaxy of nitrides on GaN is discussed. In spite of the large lattice-mismatch between nitrides and sapphire, high quality nitrides have been obtained. Growth process of the group III nitride on sapphire by using a buffer layer was understood by in-situ observation of the reflection high-energy electron diffraction from the epitaxially grown GaN on sapphire by molecular beam epitaxy. It is also found that nitride alloys grown on thick GaN was coherently grown even though the thickness exceeds the critical layer thickness.
The properties of substrate lattice-matching to InGaAlN, which has progressed in light emitting devices in the wavelength shorter than that of green light, and high-power transport devices operated at high temperature, are described. The lattice constant, crystal structure, cleavability and its direction of substrates is explained in comparison with InGaAlN. Of the sapphire substrates widely used at present, the (011^^-0) plane is shown to be the most suitable substrate commercially available. The GaN growth on a (0001) 6H-SiC substrate with polarity is introduced and the substrate polarity is described to severely affect the crystalline quality of an epitaxially grown film. This paper also reviews (101) NdGaO_3 and (111) MgAl_2O_4 as substrates nearly lattice-matched to GaN. The In_<0.22>Ga_<0.78>N on a house-made (0001) ZnO substrate is reported as the only attempt of lattice-matching growth in InGaAlN.