Si Adatom Diffusion Research Articles

Peculiarities of axial and radial Ge–Si heterojunction formation in nanowires: Monte Carlo simulation

The process of axial and radial Si–Ge heterostructure formation during nanowire growth by vapor–liquid–solid (VLS) mechanism was studied using Monte Carlo (MC) simulation. It was demonstrated that radial growth can be stimulated by adding chemical species that decrease the activation energy of precursor dissociation or the solubility of semiconductor material in catalyst drop. Reducing the Si adatom diffusion length also leads to Si shell formation around the Ge core. The influence of growth conditions on the composition and abruptness of axial Ge–Si heterostructures was analyzed. The composition of the GexSi1–xaxial heterojunction (HJ) was found to be dependent on the flux ratio, the duration of Si and Ge deposition, and the catalyst drop diameter. Maximal Ge concentration in the HJ is dependent on Ge deposition time owing to gradual changing of catalyst drop composition after switching Ge and Si fluxes. The dependence of junction abruptness on the nanowire diameter was revealed: in the adsorption-induced growth mode, the abruptness decreased with diameter, and in the diffusion-induced mode it increased. This implies that abrupt Ge–Si HJ in nanowires with small diameter can be obtained only in the chemical vapor deposition (CVD) process with negligible diffusion component of growth.

One-Dimensional Defect-Mediated Diffusion of Si Adatoms on theSi(111)−(5×2)−AuSurface

Using scanning tunneling microscopy, we determine that the one-dimensional diffusion of Si adatoms along the Si(111)-(5 x 2)-Au surface reconstruction occurs by a defect-mediated mechanism. Distinctive diffusion statistics, especially correlations between sequential adatom displacements, imply that the displacements are triggered by an interaction with a defect that is localized to the adatom. The defect is intrinsic and thermally activated. The measured diffusion statistics are modeled accurately by a Monte Carlo simulation. The measured adatom diffusion activation barrier is 1.24 +/- 0.08 eV.

Molecular dynamics simulation of microcrystalline Si deposition processes by silane plasmas

Molecular dynamics (MD) simulations have been performed to examine enhanced surface diffusion of Si adatoms during silane(SiH 4)-based plasma enhanced chemical vapor deposition (PECVD) processes. Such high surface diffusion, if it actually takes place, has been known to account for the growth of microcrystalline silicon( μc-Si) in the PECVD process. Focused in the present study is the motion of a silicon (Si) adatom on a hydrogenated Si surface and the surface diffusion coefficient of Si adatoms on the fully hydrogenated (111) Si surface at 600 K was evaluated from MD simulations. The obtained diffusion coefficient is much larger than those of typical clean Si surfaces known in the literature. The interatomic potential functions for Si–H systems used for the simulations, which we have developed for this study based on ab initio calculations of the interatomic energies, are also presented.

The Calculation of Adatom Diffusion Process on Si and Ge Surfaces from First-Principles

We study Si adatom diffusion process near Si/Ge stepped surfaces from first-principles. Dependence of surface growth morphology on growing conditions such as temperature is not clearly understood. We calculate Ehrlich-Schwoebel(ES) barrier of stepped surfaces from first-principles and analyze adatom diffusion process to understand growth mechanism of the surfaces. The configuration of the surfaces with ES barrier would play a key role in the diffusion process because it controls hopping rates in the presence of step edges. Our results are likely to help correctly access the adatom diffusion process on the surfaces.

Adsorption and diffusion of Si adatom near single-layer steps on Si surface

By use of the empirical tight-binding (ETB) method, the adsorption and diffusion behaviours of single silicon adatom on the reconstructed Si(100) surface with single-layer steps are simulated. The adsorption energies around the SA step, nonrebonded SB step, rebonded SB step and rough SB step with a kink structure are specially mapped out in this paper, from which the favourable binding sites and several possible diffusion paths are achieved. Because of the rebonded and kink structures, the SB step is more suitable for the attachment of Si adatom than the SA step or defective surface.

Low-temperature silicon epitaxy on hydrogen-terminated Si(001) surfaces

Si deposition on H terminated $\mathrm{Si}(001)\text{\ensuremath{-}}2\ifmmode\times\else\texttimes\fi{}1$ surfaces at temperatures $300--530\phantom{\rule{0.3em}{0ex}}\mathrm{K}$ is studied by scanning tunneling microscopy. Hydrogen apparently hinders Si adatom diffusion and enhances surface roughening. The post-growth annealing effect is analyzed. Hydrogen is shown to remain on the growth front up to at least $10\phantom{\rule{0.3em}{0ex}}\mathrm{ML}$. Si deposition onto the $\mathrm{H}∕\mathrm{Si}(001)\text{\ensuremath{-}}3\ifmmode\times\else\texttimes\fi{}1$ surface at $530\phantom{\rule{0.3em}{0ex}}\mathrm{K}$ suggests that dihydride units further suppress Si adatom diffusion and increase surface roughness.

A combined three-dimensional kinetic Monte Carlo and quantum chemistry study of the CVD of Si on Si(1 0 0) surfaces

The gas phase deposition of Si on a Si(1 0 0) surface in the presence of hydrogen was studied with three-dimensional kinetic Monte Carlo (KMC) simulations. The KMC model here proposed explicitly accounts for the 2×1 Si surface reconstruction and includes the reactions of adsorption of Si and H 2, their diffusion on the surface and the desorption of molecular hydrogen. Several parameters, such as the energy of interaction of adsorbed Si and H atoms, as well as the kinetic constants for the desorption of H 2 were calculated with density functional theory using B3LYP functionals. In particular, the activation energies for the diffusion of H along and across the dimer rows were calculate to be 2.3 and 1.8 eV, respectively. The KMC model was used to simulate the molecular beam epitaxy of Si. The activation energy and pre-exponential factor for the diffusion of Si adatoms along the dimer rows giving the best agreement with experimental data were 0.825 eV and 10 13 s −1. The TPD of hydrogen from Si(1 0 0)2×1 surfaces was simulated with success using a literature kinetic constant but could not be reproduced using the single site kinetic constant calculated through quantum chemistry. Finally, it was found that adsorbed H can significantly influence the morphology evolution of the film during the growth process.

Facet evolution in selective epitaxial growth of Si by cold-wall ultrahigh vacuum chemical vapor deposition

Si epitaxial layers were selectively grown on local-oxidation-of-silicon-patterned Si (100) substrates by cold-wall ultrahigh vacuum chemical vapor deposition. The Si windows were aligned along the [110] direction on Si (100) surface. As growth temperature increased from 550 to 650 °C, the development of (111) facets was dramatically suppressed, and the Si growth on sidewall facet planes was decreased. It is believed that surface diffusion of Si adatoms plays an important role in the morphological evolution of selective epitaxial growth (SEG). We propose a model to explain our experimental observations, and to clarify the effect of growth temperature on the facet morphology in terms of the surface mass transport and mass accumulation processes on facet surfaces. (211) facet formation between (311) and (111) facets in Si SEG is reported, and the stability of the (211) plane is also discussed. Finally, we investigated the changes in facet morphology with Si layer thickness, which supports our model for the facet evolution observed in Si SEG.

Molecular dynamics simulation of ion bombardment on hydrogen terminated Si(001)2×1 surface

Molecular dynamics simulations were performed to investigate H2+ and SiH3+ ion bombardment of hydrogen terminated Si(001)2×1 surfaces. Normal incidence ion bombardment effects on dangling bond generation, adatom diffusion, and nucleation were studied as a function of incident energy between 10 and 40 eV. The dangling bond generation rate due to H2+ impacts at 20 and 40 eV was about twice that of SiH3+. However these effects appeared to be insignificant compared to probable neutral radical effects under typical plasma-enhanced chemical vapor deposition conditions. The enhanced diffusion of Si adatoms due to ion bombardment was observed to be minor in comparison with thermal diffusion and the disruption of ledge sites due to SiH3+ ion bombardment is not significant, with ion incident energies up to 40 eV. Ion bombardment in the incident energy range between 10 and 20 eV can contribute the modification of surface kinetics without bulk damage.

Si adatom diffusion on Si (100) surface in selective epitaxial growth of Si

The growth temperature dependence of Si adatom diffusion on Si (100) surface was systematically investigated in a cold wall ultrahigh vacuum chemical vapor deposition system. Si epitaxial layers were selectively grown on local oxidation of silicon patterned Si substrates. By cross-sectional transmission electron microscopy analysis, the increase in Si growth rate on the Si (100) surface near the edge of (311) facet was observed at various growth temperatures. This can be understood as a consequence of the mass transport from the sidewall (311) facet to the top (100) surface. Based on a simple diffusion model, the surface diffusion lengths of Si adatoms along the [110] direction were estimated to be about 70, 140, and 200 nm at 550, 600, and 650 °C, respectively. The calculated activation energy of 1.4 eV for diffusion of Si adatoms from the temperature dependence of the diffusion length was relatively higher than those in previous reports obtained under ultrahigh vacuum environment. This difference was discussed in terms of the discrepancy in the surface states by hydrogen adsorption on the Si surface.

Migration of Si adatom on strained Si(1 1 1) surfaces

Quantum chemical calculations were performed to indicate the lowest energy path and the activation energy for Si adatom diffusion on a Si(1 1 1) surface and to evaluate the dependency of the activation energy on strains imposed on the Si surface. The activation energy barrier for the adatom diffusion was estimated to be 0.8 eV for a non-strained surface. The energy barrier increases on Si surfaces under compressive strain and, conversely, decreases under tensile strain. However, the energy variations are only within 0.1 eV under strains ranging between ±1%. The influence of strain on the 7×7 reconstruction is discussed in light of these theoretical results.

Role of hydrogen for adsorption and diffusion of a Si adatom on monohydride and dihydride Si(0 0 1) surfaces

We investigate adsorption and diffusion of a Si adatom on monohydride and dihydride Si(0 0 1) surfaces using first principles calculations. We find that adsorption and diffusion of a Si adatom on hydrogenated Si(0 0 1) surfaces are drastically altered from those on a clean surface in several ways. Both positions and number of adsorption sites vary with different coverages of hydrogen, leading to completely different diffusion behavior from the clean surface. The Si adatom is stabilized by capturing surface hydrogen without potential barrier during adsorption. The H atom hops back and forth exothermally to the adatom during diffusion and thus, the diffusion species is no longer Si adatom itself. It reveals different diffusion features that typical potential energy surface with a single adatom fails to describe. Differences in the adsorption and diffusion of Si adatom between monohydride and dihydride phases are discussed. The role of hydrogen as a surfactant on hydrogenated surface will be further discussed based on our calculations.

AFM and RHEED study of Ge/Si(001) quantum dot modification by Si capping

Atomic force microscopy (AFM) images of 6 Ge equivalent monolayers (EML) deposited on Si(001) at 500°C by solid source molecular beam epitaxy reveal the formation of ∼2 nm high islands with a height/base dimension ratio close to 1/10. Clear differences are observed, by both reflection high-energy electron diffraction and AFM, according to the Si capping conditions of such hut clusters. At room temperature the impinging Si adatoms homogeneously cover the Ge clusters due to very low Si and Ge surface diffusions nearly preserving the corrugation of uncovered Ge clusters. In contrast, at 500°C and a low Si deposition rate (∼2 EML min −1), a surface smoothing occurs after deposition of a thickness as low as 5 Si EML. A Si adatom diffusion from the apex towards the bottom of the islands concomitant with a Ge lateral segregation lead to a rapid collapse of the buried islands. This scenario is supported by the observation of a (2×9) superstructure, characteristic of the formation of SiGe alloy.

Surface electronic structure modifications due to buried quantum dots

Coherently strained Ge islands have been grown on Si(100) substrates by means of molecular beam epitaxy and subsequently covered by a 10-nm-thick Si cap layer. In situ scanning tunneling microscopy revealed surface protrusions (up to about 0.4 nm) whose height depends on the size of the buried Ge dots. From the surface deformation, the in-plane strain within the capping layer was calculated. Evidence for directed diffusion of Si adatoms away from the highly strained regions was found. On the protrusions, a lowering of the surface band gap was measured using locally resolved scanning tunneling spectroscopy. This can be explained by changes of the electronic structure of the silicon surface induced by the inhomogeneous strain around the buried dots.

Energetics and kinetics for Si–Ge intermixing on Ge-adsorbed hydrogenated Si(100) surfaces

We present first-principles total-energy calculations for Ge adsorption on the H-terminated Si(100) 2×1 surface. The configuration with Si atoms at the topmost surface has the lowest energy among the various configurations for all Ge coverages considered, implying that Si segregation takes place in Ge growth on H/Si(100). Since Si segregation is a kinetic process, although driven by energetics, we consider an atomistic model for the place exchange between surface Ge and subsurface Si atoms based on the previous studies on the Si-adatom diffusion on H/Si(100). We argue that kinetics as well as energetics is important in understanding the Si–Ge intermixing near the interface.

Theory of adsorption and diffusion of Si adatoms on H/Si(1 0 0) stepped surface

First-principles calculations on the formation energy of single height steps on the hydrogen terminated Si(1 0 0) surface and the adsorption and diffusion of Si adatoms on the stepped surfaces are reported. It is found that the formation energy of the S A step is lower than that of the S B step, leading to a relatively straight S A step and a relatively rough S B step. The S A step acts as a poor sink for Si adatoms whereas the S B step acts as a strong sink. In the initial stage of the homoepitaxial growth of the hydrogen terminated Si(1 0 0) surface, one-dimensional islands may be formed similarly to on the bare Si(1 0 0) surface.

First-principles calculation on diffusion of Si adatoms on H/Si(001)-(2×1) surface

We report the result of first-principles calculations on the Si adsorption on the monohydride terminated Si(001)-(2×1) surface. It is found that the Si adatom spontaneously segregates one H atom from a surface Si dimer during adsorption, and further captures the remaining H atom of the same Si dimer during surface migration, leading to the most stable adsorption geometry. The Si adatom diffusion process on the hydrogenated Si(001) surface is complicated by the transfer of H atoms between the adatom and the Si dimer atoms. The migration of the Si adatom is extremely reduced compared with that on the bare Si(001) surface.

Surface preparation of Si substrates for epitaxial growth

We have improved and evaluated hydrophilic and hydrophobic chemical treatments to prepare Si(001) and Si(111) wafers for epitaxial growth substrates. In the hydrophilic case, we found that a simple chemical treatment in a mixture of H 2SO 4 and H 2O 2 solutions has the same effect as the traditional RCA-based methods. However, the boron contamination of 5×10 18 atoms cm −3 on the resultant surface after removal of its oxide layer is difficult to prove in nature and blocks surface diffusion of Si adatom at the initial stage of crystal growth. In the hydrophobic case, improvement of preparation facilities in order to free a sample from oxygen allows a surface to be obtained that is clean enough for crystal growth without the boron problem. In both cases, there is no difference in the carbon contamination concentration of 2–4×10 18 atoms cm −3. These results demonstrate that the hydrophobic surface has advantages. We further studied the relationship of surface contamination and surface structures of hydrophobic Si(001) and Si(111). With increasing annealing temperature in UHV, the Si(001) surface exhibited c(4×4) in the narrow range of 580–750°C, and eventually, a very weak SiC transmission pattern appeared above 750°C. The simultaneous disappearance of c(4×4) and appearance of a SiC pattern indicates the association of the c(4×4) reconstruction with carbon. The carbon concentration of 2.5–4×10 18 atoms cm −3 is insufficient for carbon atoms to form a component of the c(4×4) structure, and therefore the carbon must have another effect. A consequence of this conclusion is that carbon is not responsible for C-type defects, and we speculate that boron may be the origin. The Si(111) surface showed ( 3 × 3 ) at between 400°C and 600°C, and the reconstruction is independent of the carbon contamination.

Adsorption and Diffusion of Si Adatom on Hydrogenated Si(100) Surfaces

We present first-principles total-energy calculations which provide a detailed picture of adsorption and diffusion of a Si adatom on hydrogenated Si(100) surfaces. We find that the adatom spontaneously substitutes for the H atom upon adsorption. We also find that the pathways and barriers of the adatom diffusion are sensitive to H coverage. Calculated results are consistent with H-induced variation in morphology of overlayers observed in epitaxial growth.

Molecular dynamics study of adatom diffusion on Si(100) surface — importance of the exchange mechanism

The Si adatom diffusion on a Si(100) surface is studied at temperatures between 1000 and 1300 K through long-time-scale empirical molecular dynamics simulations. Exchange is found to occur with a non-negligible frequency, roughly 1 30 to 1 80 as often as hopping occurs. This has only a minor effect on the diffusion constants parallel to the dimer row, but has a major effect on those perpendicular to the dimer row. In the simulations, a variety of complex exchange events involving up to three surface atoms were observed. The dynamical effects are also discussed by comparing with the static calculations.