Hetero-epitaxy of compound semiconductors on largely lattice mismatched substrates is discussed. To reduce dislocation density, microchannel epitaxy that is a growth technique consisting of selective area epitaxy and epitaxial lateral over-growth has been proposed. It is shown that the key issue of this technique is how to increase the lateral growth rate compared with the vertical growth rate, which means how one can enhance the growth anisotropy. In the case of liquid phase epitaxy, one can get a large anisotropy by using a facet on the top and atomically rough surfaces on the sides. With very small interface supersaturation, wide and flat microchannel epitaxial layers have been successfully obtained for GaAs and InP on Si substrates. Wide dislocation free areas have been obtained outside of a dislocated region just above the microchannel. Microchannel epitaxy by molecular beam epitaxy is also described.
This article describes the growth feature of alkali halide single crystalline films in order to discuss the heteroepitaxy of ionic crystals. With alkali halides heteroepitaxy is realized even for the lattice misfit as large as several tens %. Growth mode is classified into three groups according to the lattice misfit. Lattice strain was discussed for both continuous and discontinuous interfaces. A perfect pseudomorphism is not observed even for the continuous interface and a few % strain exists even for the discontinuous interface.
Recent developments of the research on metallic multi-layers, relating to epitaxy and lattice distortion, are briefly surveyed. Main topics are structural analysis using X-ray diffraction, superstructures with monolayers, ultrathin films with unstable crystal phase fabricated utilizing epitaxy, and progresses in supermodulus effect study.
Lattice-mismatch strain was induced in (001)-oriented thin films of La-214 superconducting oxides. For the compressed films of La2-xSrxCuO4 and La2-xBaxCuO4 on LaSrAlO4 substrates, the superconducting transition temperature (Tc) reached 44 K and 47 K, respectively, which are higher than the values for bulk samples. Moreover, the Tc-x phase diagrams for the films did not show a local minimum at x∼0.125, the so-called “1/8 anomaly”. The compressive strain expands the c-axis via the Poisson effect and seems to suppress formation of the low-temperature tetragonal phase. We speculate that the both effects result an increase in the bond length between Cu and the apical oxygen, which is responsible for the enhancement of Tc. We also found that the lower residual resistivity gives a higher Tc. Based on the experimental results, we suggest that reduction of spin fluctuation is effective to increase Tc in the high-temperature superconductors.
Heteroepitaxy of 3C-SiC on Si(001) in gas source molecular beam epitaxy has been carried out by a combination of the formation of a SiGeC buffer layer and subsequent 3C-SiC growth. The SiGeC buffer layer was formed by the use of dimethylgermane ((CH3)2GeH2) to convert the surface region chemically into 3C-SiC. It was found that single-crystalline 3C-SiC with a smooth surface could be obtained reproducibly at as low as 650oC, and Ge atoms could be introduced successfully in the carbonized layer. Single crystalline 3C-SiC showing a single-domain (3×2) surface superstructure could be grown at as low as 910oC on the SiGeC buffer layer. By the precise control of the disilane (Si2H6) flux, a smooth surface without pits and Si-islands was realized successfully. The activation energy of growth rate became lower from 27.4 kcal/mol in the three-dimensional growth to 12.7 kcal/mol in the step-flow growth with the increase of growth temperature. Growth mechanisms in the 3C-SiC/Si heteroepitaxial system and the effects of Ge incorporation in the buffer layer are discussed.
Heterovalent ZnSe/GaAs(001) interface structure was controlled by preparing various initial surface structures of GaAs(001) and by pre-growth deposition of Zn or Se. It was found that density of stacking faults and other kind of defects in the ZnSe film strongly depended on the interface structure. In order to reduce the defect density, it was important to avoid chalcogenization of GaAs surface and segregation of excess As. A ZnSe film with lowest density of defects was obtained by Zn treatment on the GaAs(001)-(2×4) surface.
When the Si/SiO2 interface is formed by thermal oxidation of Si, quite a large strain is induced by the volume expansion from Si to SiO2 (approximately 2.25 times). This strain should have a great effect on the formation of the Si/SiO2 interface. Here, we review our recent theoretical studies of the strain effect on the atomical flatness and electrical properties of the interface by using the first-principles calculation method. We also show that the Si emission at the interface is an important mechanism for releasing the strain.
A review of ultra trace element analysis with total-reflection X-ray photoelectron spectroscopy (TRXPS) and its applications to the semiconductor surface are descrided. TRXPS is a method for improving the detection sensitivity by increasing the peak to background ratio for photoelectron spectrum. The detection limit of TRXPS was found to be 9×1010 atoms/cm2 for contamination on Si wafers. The improvement is 40 to 100 times of the normal-type XPS.
The development of a method to mask bitter taste has widely been required in pharmaceutical science and food science. We found that a lipoprotein (PA-LG) composed of phosphatidic acid (PA) and β-lactoglobulin (LG) inhibits, selectively and reversibly, the frog taste nerve responses and human taste sensation to various bitter substances without affecting sweetness, saltiness and sourness. It was suggested that PA is a key material for the inhibition of bitterness because PA-containing lipoproteins show inhibitory action irrespective of species of proteins. As for the mechanism of bitter suppression, PA-LG would appear to mask the receptor site for bitterness on taste receptor membrane. Inhibitory action for bitterness by PA itself without protein was also examined in human. The result showed that PA selectively inhibits bitterness. The masking ability of PA for bitterness was satisfactory in practical use, although it was less than that of PA-LG. PA has been already in use as a masking additive for bitter taste of drugs and foods.