N. L. Shwartz

3D-model of epitaxial growth on porous {111} and {100} Si surfaces

A 3D Monte Carlo model of epitaxial and annealing processes on (111) and (100) surfaces of diamond-like crystals was developed. Simulations of epitaxial growth on porous Si(111) and Si(100) surfaces using this model were carried out. Distinct relief is formed over sealing pores due to diffusion differences on these surfaces. A smooth solid layer is formed on the (111) surface whereas pyramidal pits faceted by (111) planes are created over pores on the (100) surface. The deposited dose necessary for pore sealing is one order of magnitude greater on the Si(100) than on the Si(111) surface.

A Monte Carlo simulation of the processes of nanostructure growth: The time-scale event-scheduling algorithm

The SilSim3D program package is developed to conduct Monte Carlo simulations of the kinetics of growth, evaporation, and annealing of thin layers on solid substrates on the basis of an original algorithm of scheduling events on a real time scale. The model allows the simulation of various nanoelectronic technological processes in multicomponent physical and chemical systems of more than 107 particles in time intervals comparable with the actual experimental times. The initial stages of atomic layer deposition and the growth of silicon nanowhiskers are simulated as an example.

SiOx layer formation during plasma sputtering of Si and SiO2 targets

Deposition of SiOx layers of variable composition onto silicon wafers was performed by co-sputtering of spaced Si and SiO2 targets in argon plasma. Coordinate dependences of the thickness and refractive index of separately deposited Si and SiO2 layers and the SiOx layer grown during co-sputtering of targets were determined using optical techniques. It was shown that the SiOx layer composition is not equal to a simple sum of thicknesses of separately deposited Si and SiO2 layers. The coordinate dependences of the Si and SiO2 layer thicknesses were calculated. To fit the calculated and experimental data, it is necessary to assume that no less than 10% of silicon is converted to dioxide during co-sputtering. A comparison of the coordinate dependences of the IR absorbance in SiO2 and SiOx layers with experimental ellipsometric data confirmed the presence of excess oxygen in the SiOx layer. Taking into account such partial oxidation of sputtered silicon, composition isolines in the substrate plane were calculated. After annealing of the SiOx layer at 1200°C, photoluminescence was observed in a wafer area predicted by calculations, which was caused by the formation of quantum-size Si nanocrystallites. The photoluminescence intensity was maximum at x = 1.78 ± 0.3, which is close to the composition optimum for ion-beam synthesis of nanocrystals.

Lattice Monte Carlo model of SiOx layers

A kinetic Monte Carlo model of amorphous SiOx layers has been developed on the basis of a partially filled diamond-like lattice with allowance for different valences of silicon and oxygen. In the starting model layers of stoichiometric composition SiO2, about 50% of lattice sites are randomly occupied by silicon and oxygen atoms, which tend to form chains of vertex-sharing tetrahedra SiO4 upon long-term model annealing. The amount of perfectly coordinated atoms reaches ∼65% after short-term annealing and slowly increases with time, indicating the formation of a metastable state. In the most ordered final state, more than 90% of atoms are perfectly coordinated. The presence of excess silicon in SiOx layers increases the fraction of perfectly coordinated oxygen atoms up to 95%. As the SiO2 matrix tends to the final ordered state, excess silicon atoms assemble to form nanoclusters.

Interlayer atomic diffusion as the reason for self-assembled quantum dot formation

Abstract Self-assembled 3D-islands formation during epitaxial growth was investigated using a kinetic Monte Carlo model. Schwoebel barriers for the explanation of 3D growth kinetics were suggested. The diffusion hop probability ratio for atom hops up to hops down χ , leading to 2D, 3D and Stranski–Krastanov growth modes, was obtained. The necessity of atomic flow from the island edge to the upper layers for 3D-island formation was demonstrated. Island size equalization during the growth process was observed for certain χ values. The new layer nucleus on the top of the island becomes the trap for atoms detached from the edges of the lower layer. Atomic flow to the upper layer increases and that is the reason for the equalization of separate 3D-island lateral sizes.

Simulated growth of GaAs nanowires: Catalytic and self-catalyzed growth

The kinetic lattice Monte Carlo model of GaAs nanowire growth by the vapor-liquid-crystal mechanism is suggested. The catalytic and self-catalyzed growth of nanowires on the GaAs (111)B surface is simulated. The dependence of the morphology of the growing nanowires on the growth parameters is demonstrated. Upon self-catalyzed growth with gallium drops serving as the growth catalyst, the growth rate of the nanowires linearly depends on the arsenic flow in a wide range of arsenic flow rates. The decreasing dependence of the self-catalyzed growth rate of the nanowires on the initial gallium drop diameter is less steep, and the optimal growth temperature is higher than that for catalytic growth. It is shown that self-catalyzed growth is more sensitive to the ratio between the gallium and arsenic flow rates than catalytic growth.

Mechanisms of nanowhisker formation: Monte Carlo simulation

Nanowhisker formation on substrates activated by catalyst drops is studied by Monte Carlo simulation. Dependences of the whisker growth rate on diameter are investigated for various growth modes. The influence of deposition conditions on whisker morphology is examined. It is shown that straight thin whiskers of uniform thickness can be obtained only using a catalyst having a large contact angle with the whisker material. In such a physicochemical system, variation of growth conditions can result in nanotube formation. An atomic mechanism for the formation of a hollow whisker is proposed. Ranges of model growth conditions suitable for the growth of nanowhiskers and nanotubes are determined.

Causes of the stability of three-bilayer islands and steps on a Si (111) surface

AbstractThe initial stages of growth of Ge and Si layers on a singular Si (111) surface result in an unusual morphology of the growth surface if the layers are deposited at a low rate; i.e., triangular islands with a height of as much as three atomic layers are formed. A simulation based on the Monte Carlo method has been used to show that an additional barrier with a height of 0.5–0.6 eV, serving to incorporate atoms into dimerized bonds at the edges of the triangular islands, brings about enhanced growth of the islands in relation to their height and a change in the triangles’ orientation. According to the suggested hypothesis, the increase in the islands’ height and the limitation of their height to three bilayers are due to the effect of the edge dimers, whose orientation changes when the height of a step perpendicular to the

Monte Carlo simulation of the formation of AIIIBV nanostructures with the use of droplet epitaxy

\langle \bar 1\bar 12\rangle