On the basis of segregation study in crystal growth from a thin melt layer in presence of the submerged heater the possibility to obtain the uniform material along the height of the ingot is shown. Numerically in modeling of solidification of 200 mm in a diameter Sb doped Ge the accurate solution with account for convection was found in the central part of the domain to coincide with the one dimension problem for the melt layer beginning from 40 mm. Condition to neglect with convection in mass transfer are to be more rigorous: the melt layer should be less than 20 mm. In this case, one may use Tiller’s equation obtained to calculate the axial impurity distribution in approach of the diffusion-control segregation. The analysis of attempts to describe experimental data of crystal growth by use of the simplified equations has shown their validity in case of account for actual crystal growth rate or change in the melt layer thickness during the run, as in expression find by Marchenko et al. The above said makes it possible for to describe analytically the axial distribution of impurity in the ingot, particularly, for B and P in silicon and to recommend the amount of its concentration over the height. The uniform material in the very end of the solidification process of the rest portion of the ingot can be obtained by the variation of the growth rate due to change in the cooling rate with time.

The work is purely metrological in nature. The developed methods for measuring the effective magnetic anisotropy field HAeff and the line width of ferromagnetic resonance ∆H of magnetically-axis hexagonal ferrites in the working frequency range of the microwave range of electromagnetic waves are presented. Methods allow to determine HAeff in the ranges of 10—23 kOe and 28—40 kOe and ΔН in the range of 0.5÷5 kOe. The first technique (measurement technique in free space in the three-millimeter wavelength range) is implemented in the frequency range 78.33—118.1 GHz. The second technique (the technique using the microstrip transmission line) is implemented in the frequency range from 25 to 67 GHz.
Testing of the methods on polycrystalline samples of hexagonal barium and strontium ferrites (both nominal composition and complex ones) with a high degree of magnetic texture, comparison of measurement results with results obtained using standard measurement techniques on spherical samples showed their high accuracy and reliability.

In this work we have used contact and contactless techniques to measure the electrical resistivity of single crystal silicon wafers with porous layers of variable thickness synthesized on the surface. The porous layers have been synthesized on the surfaces of single crystal wafers with well pronounced microroughness pattern, either textured or grinded. We have used the classic four-probe method with a linear probe arrangement as the contact measurement technique, and the resonance microwave method based on microwave absorption by free carriers as the contactless measurement technique. Electrical resistivity distribution over the specimen surface has been mapped based on the measurement results. We have demonstrated a general agreement between the electrical resistivity distribution patterns as measured using the contact and contactless measurement techniques. To analyze the electrical resistivity scatter over the specimen surface area we have simulated the field distribution in the electrolyte during porous layer formation in a non-planar anode cell. The regularities of the electrical resistivity spatial distribution in different types of specimens are accounted for by specific porosity formation mechanisms which in turn are controlled by the initial microroughness pattern and the field distribution pattern in the electrolyte for each specific case.

The phosphorus concentration profiles in germanium in In0,01Ga0,99As/In0,56Ga0,44P/Ge heterostructures with gallium co-diffusion, that were obtained during first cascade of a multicascade solar cell formation were analyzed. The diffusion of phosphorus took place from the layer In0,56Ga0,44P together with the diffusion of gallium in a strongly gallium-doped germanium substrate, which determined the features of the diffusion process. First of all, co-diffusion of gallium and phosphorus leads to formation of two p—n junctions. Fick’s laws cannot be used for diffusion description. Distribution of the P diffusivity (DP) in the depth of the sample was determined by two methods —Boltzmann—Matano (version of Sauer—Freise) and the coordinate-dependent diffusion method. It is shown that when we have used the coordinate-dependent diffusion method, the DP values are more consistent with the known literature data due to taking into account the drift component of diffusion. The tendency of DP to increase at the heterostructure boundary and to decrease at approaching to the main p—n junction is observed for both calculation methods. DP increase in the near-surface region of the p—n-junction, whose field is directed to the interface of the heterostructure, and the decrease in the region of the main p-n junction, whose field is directed in the opposite direction, as well as the observed growth of DP with the electron concentration, leads to the conclusion that diffusion in this case takes place as the part of negatively charged VGeP complexes, as in the case of P diffusion alone.