1. Introduction

The phenomenon of self-assembly of nanoclusters of impurity atoms interacting with lattice defects in the crystal lattice of semiconductor is likely to be one of the promising and novel methods for the development of nanoscale structures.

The possibility of building clusters of impurity atoms in semiconductors is currently being actively investigated by many researchers [1-5]. The interest in the assembly of nanoclusters is largely motivated by the desire to develop nanosize structures and control their parameters, which in turn will have given the possibility to develop novel elements for micro- and nanoelectronics. One can evidence that shaping of clusters of various nature significantly depends on solubility, diffusion parameters of impurity atoms and thermodynamic conditions as well. However, virtually no major research has been conducted over lately on the possibility to regulate the parameters of such clusters.

Clusters of impurity atoms in silicon were largely obtained by the technique of homo-epitaxial growth or molecular-beam epitaxial growth. The authors of the present research work propose the technology of obtaining of self-building impurity clusters by diffusion. The above method in contrast to the existing method of molecular-beam epitaxy has certain advantages and does not require complex and expensive equipment. The method allows:

● to produce nanoscale structures throughout the entire bulk of the crystal;

● to set easily the structure, composition, distribution and ordering of nanoscale clusters;

● to obtain magnetic nanoclusters with adjustable magnetic moment, i.e. a novel magnetic semiconductor material;

● to control the charging state of nanoclusters in the broad range (N+(-)n, где n>3), thus producing multiply charged centers in the semiconductor, which might serve as the basis for promising novel material for nanophotonics.

The possibility of building clusters of impurity atoms of Ni in silicon and controlling their parameters is currently investigated in the present research article. The choice of the impurity atom is preconditioned by the fact that firstly, its solubility and the diffusion coefficient in silicon is comparatively higher than that of other elements of the Fe group, and secondly, Ni will most probably turn out to be one of the most important metals for the semiconductor industry in the near future.

2. Main Body

2.1. Theoretical Analysis

Over the past few years, there has been a widespread interest among experts in the field of nanotechnology and nanoelectronics all over the world in the technology of self-organizing impurity clusters with manageable structures and magnetic properties. In this respect one can note some interesting results related to implanting Co and Ge ions on Si, ion implantation in other semiconductor materials [6-8]. As we have heard, the technology of self-organization of clusters of impurity atoms by using the diffusion technologies currently is not sufficiently studied. Diffusion technology for producing nanoscale structures is not only a more affordable and cheap technique, allowing large-scale production, but also gives the ability to synthesize nanoscale structures of various type (2D, 1D, 0D), as well as specify the distribution and density over the bulk of the crystal.

Therefore, the main objective of this research is to demonstrate that under certain doping conditions one can witness shaping of clusters which also gives us the opportunity to manage parameters of those clusters.

2.2. Experimental Part

Our group had developed a special technique for doping, the so-called "low-temperature doping" of semiconductors. This method of doping is based upon the diffusion process which is carried out in stages by gradually increasing temperature ranging from room temperature to the diffusion temperature [9].

The native sample and dopant respectively (pure metallic Nickel of certain weight (to be determined by ampoule volume)) are put into evacuated quartz ampoules (pressure ∼10^-6 mm.of mercury column), which are inserted into the diffusion furnace at Т=300K. Shortly thereafter starting from T = 300 K the furnace temperature at the location of the ampoule is gradually increased at a rate of 5°C / min and preheated to the temperature on intermediate processing and this temperature is retained to a certain period, then the furnace temperature is increased quickly at the rate t = (15÷20) C/min. and so up to the diffusion temperature (1523C) and the samples were retained under this temperature during 40 minutes. After having completed all the above stages the ampoules are removed from the furnace and cooled off. Such doping conditions ensure embedding maximum concentration of Ni in the Si lattice and uniform doping through the entire bulk of the crystal.

For the experiment we used single crystalline silicon of both n- and p-type with phosphorus concentration N_P=10¹³÷10¹⁷cm^-3 and the boron concentration N_B=10¹⁴÷10¹⁷cm^-3. The native samples had a minimal dislocations density which was N_D<10² cm².

Also, we used single crystalline silicon with various concentration of oxygen atoms N_O2=10¹⁷ cm^-3 and N_D