Growth of Crystals

It can be inferred from currently available experimental data that specific features of the formation of point defects and nonuniform impurity distribution in thin layers of binary semiconductors result from specific features of the incorporation of host and impurity atoms at kink sites. In particular, statistical correlations between elementary events occurring at step kinks may play an important role in the growth of semiconductor materials by MBE (molecular-beam epitaxy) and CVD (chemical vapor deposition).

The kinetics of crystal growth and their dependence on external factors provide information critical for understanding the growth mechanism and for controlled production of crystals with a given composition and the required properties. Crystal growth is investigated using both kinetic and morphologic principles. The former relates the growth rate to the experimental conditions (temperature, source material composition, vapor composition and pressure, etc.). The latter examines the influence of these factors and the growth rate on the habit and morphology of the crystal. The present work studies relationships among the growth conditions, growth rate, and crystal morphology for sublimation in a closed system.

Lateral epitaxy of semiconducting materials has been developing for the last 15-20 years as a method of fabricating promising structures for micro-and optoelectronics [1, 2]. In particular, it has been used to fabricate permeable-base UHF transistors [4] and low-loss optical waveguides [4] and to reduce the cost of epitaxial layers for solar energy applications [5].

The epitaxial system Si—porous Si—Si substrate possesses unusual properties for microelectronic structures and is promising for use in instruments, especially after converting the three-layer system into a Si-on-insulator structure via oxidation of the porous sublayer [1-8]. Anodic etching regimes enabling the surface layer of p+-type Si to be converted into a porous material with a significant pore volume are known. The physical properties of porous Si, mainly mechanical and optical, differ significantly from those of single-crystalline Si. In particular, the Young’s modulus of Si with 54% porosity is 1.7.10¹⁰N/m²[9], which is an order of magnitude less than that of a single crystal. The Poisson coefficient is 0.09, which is also much less than for compact crystalline Si, which has a Poisson coefficient of 0.26. Thus, porous Si is much weaker than the starting compact substrate material. Its strength decreases with increasing porosity. However, the crystal structure of Si within the porous layer is nearly identical to that of Si without pores. It has been noted [9] that the x-ray rocking curves of porous and nonporous Si are very similar. Immediately after anodic etching the lattice constant of porous Si perpendicular to the interface is hundredths of a percent greater than that of the substrate and depends on the porosity and the average pore size, increasing as they increase.

Heterostructures A^IIB^VI /A^IIIB^Vare valuable materials for semiconductor electronics that are employed especially in the fabrication of various photoactive devices.

Long-range shear (or compression—expansion) stresses (LSSs) can arise during epitaxial growth of a layer at the stage where stresses due to the lattice misfit parameters are relieved [1]. These stresses are related to the fact that the Burgers vectors of similar misfit dislocations (MDs) have screw components [2, 3]. The LSSs can influence the microstructure of the epitaxial layer and should be considered when choosing the growth conditions. The generation of such stresses has been examined [2-4] for strictly singular interfaces and in more detail for the (001) interface. However, substrates misoriented relative to singular orientations are widely used in research and industry to grow semiconductor heterostructures.

Mismatched heteroepitaxy is a special type of epitaxy for semiconducting solid solutions that has recently become very popular. Investigations of this type of epitaxy have progressed in two directions: 1) the preparation of practically strain-free junctions in order to accommodate layers and substrates with different lattice parameters and 2) preparation of strained (pseudomorphic) layers [1]. Strained layers with elastic strain “frozen” into them possess novel properties that are not found in strain-free layers. Their availability increases the variety of materials at our disposition [2]. If the structure of the layers is perfect enough, new practical properties may be utilized in conjunction with, for example, modification of the heterojunction band structure or the creation in InGaAs(111) layers of an intrinsic piezoelectric field [3].

Eutectic thin films of various microstructure are widely employed in metallurgy, machinery, and microelectronics [1]. Liquid-phase growth methods are used for preparing them as commonly as vapor-phase methods. An example is the preparation of protective coatings by rapid cooling of a melt layer [1, 2].

Amorphous phases crystallize in a peculiar manner because both kinetic parameters (nucleation rate and growth rate of the crystalline phase) that control the process increase with increasing temperature. Thus, the amorphous phase converts to the crystalline one faster as the release rate of latent heat of crystallization increases (a system with a positive feedback). As a result, explosive crystallization can occur in which the conversion rate jumps by several orders of magnitude and then remains more or less constant [1].

During the growth of crystals from solution, the crystallizing substance reaches the growth surface by diffusing through the solution bulk. As a rule, the solution moves relative to the crystal. The diffusion conditions in the boundary layer have a strong influence on the perfection of the growing crystal since they can cause morphological instability to develop. This often leads to liquid inclusions.

Zonal inhomogeneity or growth striation of crystals growing by the layer-by-layer mechanism from solution consist of macroscopic crystal layers that are parallel to a growing face and have slightly differing lattice constants. Significant strains develop at the boundaries between these layers (zone boundaries). These strains can be large enough to produce dislocations. Striation is a defect that degrades the optical quality of crystals.

The crystallographic orientation of shaped crystals grown by the Stepanov method is determined by the seed orientation. The original orientation persists if the crystal grows ideally enough. However, crystals often contain structural defects such as dislocations and their pileups, slip bands, and intergrain boundaries. Each of these defects is connected with a certain distortion of the crystal lattice. As a result, the orientation of the grown crystal differs from the initial one and varies throughout the volume. The orientation changes little during growth if the crystal contains few defects. The crystal orientation can change quite significantly if the defect density is high. Until now, studies of the change of crystal orientation during pulling have been limited to an examination of twinning effects.

Recently LiF-RF₃ (R = rare-earth element, REE) systems have been widely studied and discussed. The interest in these systems arises from the application of LiRF₄ single crystals and lithium yttrium fluoride (LYF) in quantum electronics [1, 2]. This family contains members that exhibit excellent processing characteristics in addition to a high capacity for isomorphous replacement by trivalent REE ions. These characteristics include, among others, relatively low melting points, congruent melting for several phases, and the ability to use inert media other than actively fluorinating ones. An investigation of LYF generation spectra showed that they can compete favorably with the most popular materials used in quantum electronics in those instances where high-power radiation is not required. These facts suggested that LYF could become one of the first multicomponent fluorides to be used as commercial laser crystals. Crystal production preceded investigations of phase equilibria in LiF -RF₃ systems (see below). Since the early 1970s LiRF4 crystals were included in searches for new REE compounds needed to solve problems in quantum electronics.

Oxide crystals for solid-state lasers are grown, as a rule, by crystallization from the melt. Their structure, in particular, such properties as the cation distribution over crystallographically nonequivalent atomic positions and the shape (distortion) of the coordination polyhedra, plays a decisive role in determining the laser properties of the crystals and the conditions for their successful growth.

Axial low-frequency small-amplitude vibrations of a growing crystal can have a significant effect on the fluid dynamics and heat transfer in the melt and can enhance the structural perfection of crystals on a macroscopic and microscopic scale [1, 2]. In general, few methods are available for controlling the fluid dynamics during crystal growth. These are primarily well known methods of influencing internal growth parameters such as temperature gradients, variation of the crystal and/or crucible rotation rate, and maintenance of a certain melt level [3, 4]. Methods of external control include various types of magnetohydrodynamic (MHD) influences on the melt [5]. However, this is applicable only to a limited range of materials that possess metallic conductivity in the molten state.

The nonlinear optical material LiNbO₃is transparent over the wide spectral range 0.35-4 µm. Due to recent advances in producing periodically poled LiNbO3 crystals, applications based on quasi-phasematched (QPM) conversion effects such as second-harmonic generation [1-3] and optical parametric oscillation [4] have become possible.