A relationship between growth conditions of silicon wires and their morphology, composition, and electroconductivity is presented. In these crystals recently a lattice parameter change and a visible light emission have been found being almost the same as those observed in porous silicon. The crystals were grown by a method of gas-phase reaction in a sealed tube. Using electron microscopy, X-ray microprobe analysis, ion mass spectrometry, Auger electron spectrometry, and temperature dependencies of electroconductivity, it is shown that the crystals grown in different parts of the tube have various shapes, compositions, and electroconductivities. The study of the grown crystals allows to draw conclusions on the mechanisms of crystal growth and doping, and to find a method for controlling these processes. Varying the growth conditions, one is able to change (i) shape of crystals (cylindrical or prismatic), (ii) size of crystals in the ranges from 1 pm to a few cm (axial) and from 0.1 to 100 pm (transverse), (iii) specific electroconductivity from a few to 200 9 -l cm-' with various degrees of compensation.

In this paper the results of an X-ray study of free-standing filament-like crystals of GexSi1-x solid solution are presented. The crystals were grown by a method of vapour transport reactions. The parameter x varied from 0.004 to 0.271. Independent measurements of both the lattice parameter and the composition have shown that the lattice parameter of filament crystals with diameters less than 1 mu m is less than that of bulk crystals of the same composition. For example, in the crystals of 0.5 mu m diameter with x=0.271 the measured lattice parameter differed from its bulk value by 0.0047 AA. The effect is shown to be not connected with high concentrations of impurities of defects. It is explained by the disturbance of atomic bonds in the uppermost few layers of the crystal and accompanying reconstruction of these layers, which gives rise to a mechanical stress in the crystal bulk. From these data the strain decay length was calculated and was found to vary from 10 to 18 AA. The effect observed is unexpectedly large. We connect this value of the observed change with the cylindrical shape of the crystals. As we did not observe deviations from Vegard's law in the bulk crystals with diameters greater than 1 mu m, we have discussed the possible origin of such deviations observed by other investigators.

No abstract

No abstract

No abstract