Search

Band engineered Si/O superlattices are promising channel materials for ultimately scaled complementary metal oxide semiconductor devices. Theoretical calculations have indicated that insertion of O monolayers into Si lattice creates anisotropy in the Si band structure with an enhanced carrier mobility in the channel direction. However, the experimental demonstration of such superlattices is not straightforward, as it requires processes for the deposition of a monolayer of O on Si, and to continue the Si epitaxy thereon. In this work, we investigate processes for the fabrication of Si/O superlattices using the chemical vapor deposition technique. Ozone is used for the O deposition with control of the deposited O content at the monolayer level. The SiH 4 based Si deposition is performed at low temperatures (500 °C–550 °C) in order to obtain a confined O monolayer in the Si superlattice. Si deposited on O layer of 1.2 O monolayers is completely amorphous. This indicates that this O layer is extremely uniform as it prevents any epitaxial ordering with the substrate. Reducing the O content to 0.9 O monolayers in combination with a slower Si deposition favors epitaxial ordering of Si on O/Si. Finally a defect free Si epitaxy is demonstrated on the O monolayer. Such ordering of epitaxial Si on the O layer provides a promising outlook for the growth of superlattices.

Ge (1− x ) Sn ( x ) is receiving a growing interest in the scientific community, as it has important applications in opto-electronic devices, ( as stressor) Source/Drain materials for Ge and SiGe MOSFETS. It is predicted that at 10% Sn concentration or even lower, unstrained Ge (1− x ) Sn ( x ) will exhibit a direct band gap. Moreover, in strained Ge (1− x ) Sn ( x ) the expected concentration of Sn for this cross-over is even lower. As the theoretical Sn incorporation in Ge (1− x ) Sn ( x ) is less than 1%, and Ge (1− x ) Sn ( x ) is prone to relaxation, routes towards the growth of metastable strained films has been extensively explored. Although Ge (1− x ) Sn ( x ) films (with x up to 10%) have been grown using various methods like molecular beam epitaxy, CVD growth etc. there remain issues with tendency of these layers to relax. Detailed studies on the relaxation mechanisms and effects on the Sn-atoms require suitable characterization techniques. Various techniques have been used to study the surface of the film, crystallography or concentration of Sn in the film but none of them provides information at the atomic scale as they average over many layers and atoms. Atom probe tomography (APT) analysis, on the other hand, is one such method that can provide atomic scale resolutions (∼0.3 nm) due to its ability to perform atom by atom analysis. In this paper we explore the use of APT for characterizing Ge (1− x ) Sn ( x ) layers. We comment on the difference of field evaporation values of Ge and Sn in Ge (1− x ) Sn ( x ) layer by taking a closer look at the co-evaporation of the two elements and comment on the accuracy of depth reconstruction of APT for Ge (1− x ) Sn ( x ) layer. Comparing the Sn-distributions and their local surroundings we saw a tendency for the Sn to locally enrich forming Sn clusters. Higher order clusters were observed for the relaxed sample.

One hour 500 °C air annealing induced movement of implanted Au in Si have been studied for 32 keV Au implantation in Si, in the fluence range of 1 × 10 15 – 1 × 10 17 ions cm - 2 . Samples were characterized using Rutherford backscattering spectrometry and cross-sectional transmission electron microscopy. The results indicate that, depending on the initial state of Au in the matrix, there is a clear difference in the diffusion behaviour of Au in Si. When Au is precipitated as gold-silicide nanoclusters inside the Si matrix, annealing is found to cause diffusion of Au into the bulk Si. Compared to this, for a random atomic distribution of Au in an amorphous Si matrix, annealing is found to result in out-diffusion of Au towards the surface.

Thin Mn films were deposited on GaAs(100) surfaces at room temperature by a thermal evaporation system followed by annealing at 500 °C for times ranging from 2 to 8 h in nitrogen atmosphere. X-ray diffraction shows that for samples annealed at 500 °C, the interfacial reaction results in the formation of different phases such as Mn2As, MnAs and Mn3Ga. Rutherford Backscattering Spectrometry indicates a diffused layer of Mn–Ga–As system along with some minor oxygen content. Magnetization study done at 10 K shows M–H curve comprising a ferromagnetic and antiferromagnetic part and M–T measurement done from 10 to 300 K shows a transition around 45 K which may be related to the presence of GaMnAs alloy along with the presence of the above binary phases.

Many studies have been done on low energy (1–200 keV) and high dose (1016–1017) implantation of Mn in GaAs. This study is an attempt to incorporate Mn ions in GaAs through implantation of 1 MeV Mn+1 ions in semi-insulating GaAs substrates at doses of 3×1015/cm2 and subsequent annealing. This was done to find out if any alloy of Mn–Ga–As, or binary compounds of Mn–Ga or Mn–As form due to annealing of Mn+1 ions implanted in GaAs substrates. High Resolution XRD (HRXRD) performed before annealing shows a possibility of Ga–Mn–As alloy formation, and after annealing at 800°C, except for GaAs main peaks no other phase peaks were detected. Scanning electron microscopy (SEM) shows nanostructures of various dimensions which are thought to be formed due to the defects generated due to implantation. Fourier Transform Infrared (FTIR) study shows the shift in bandgap due to Mn doping. Raman spectroscopy shows the red shift in LO and TO peak positions of GaAs after annealing, which indicates the presence of disorder and damage due to implantation. Resistivity measurement shows a thermally activated semiconductor character of charge conduction with an activation energy of 51 meV and this activation may have occurred through the transitions from impurity band to valence band. Large positive (∼25%) magnetoresistance and a mixture of ferromagnetic and paramagnetic behavior obtained in the magnetization measurement indicate the presence of ferromagnetic MnAs nanoclusters embedded in paramagnetic GaAs:Mn matrix.

We have carried out scanning tunneling microscopy experiments under ultrahigh vacuum condition to study the roughness of pristine as well as ion-bombarded Si(1 0 0) surfaces and of ultrathin Ge films deposited on them. One half of a Si(1 0 0) sample (with native oxide layer) was irradiated at room temperature using 45 keV Si − ions at a fluence of 4 × 10 15 ions/cm 2 while the other half was masked. STM measurements were then carried out on the unirradiated as well as the irradiated half of the sample. Root-mean-square (rms) roughness of both the halves of the sample has been measured as a function of STM scan size. Below a length scale of ∼30 nm we observe surface smoothing and surface roughening is observed for length scales above this value. However, the surface is self-affine up to length scales of ∼200 nm and the observed roughness exponent of 0.46 ± 0.04 is comparable to earlier cases of ion sputtering studies where only roughening [J. Krim, I. Heyvart, D.V. Haesendonck, Y. Bruynseraede, Phys. Rev. Lett. 70 (1993) 57] or only smoothing [D.K. Goswami, B.N. Dev, Phys. Rev. B 68 (2003) 033401] was observed. Preliminary results involving morphology for Ge deposition on clean ion-irradiated and pristine Si(1 0 0) surfaces are presented.

High fluence low energy negative ion implantation has been used to synthesize embedded metal nanoclusters of Au , Ag and Sb in silica glass. The Au - and Ag -implanted samples showed peaks, corresponding to surface plasmon resonance (SPR) in the optical absorption (OA) spectra, confirming the formation of metallic nanoparticles in the matrix. No SPR peak was observed in case of Sb -implanted samples which is attributed to the absence of pure metallic precipitates which could be detected in the OA spectrum. Low frequency Raman scattering (LFRS) measurements also confirm this. Cross-sectional transmission electron microscopy has been used to infer about the size distribution of the nanoparticles. Sequential implantations of Au and Ag or Au and Sb have been found to result in SPR peaks at locations in between those for nanoparticles of the constituent atoms, indicating the formation of alloy nanoparticles in the system. In case of the Au + Ag system, Rutherford backscattering spectrometry has been used to infer about the composition of the nanoparticles in terms of the concentrations of the metallic constituents. A direct, one-to-one correspondence between the SPR peak position and composition has been observed.

Embedded Au nanoclusters (NCs) in silica glass were formed by ion beam synthesis using 32 keV Au− implantation to a fluence of 4 × 1016 ions cm−2. Subsequently the NCs were irradiated with swift heavy ions in the form of 100 MeV Au8+ at ∼100 K in the fluence range of 5 × 1012–1 × 1014 ions cm−2. Samples were analyzed using Rutherford backscattering spectrometry, transmission electron microscopy and optical absorption spectroscopy. At lower irradiation fluence the high energy heavy ion irradiation has been found to result in a loss in Au due to an outward movement of the NCs together with a growth in size. At the highest irradiation fluence, almost 80% Au was lost, with only few large NCs seen which had moved to the surface. These were found to be of a deformed non-spherical shape. The amount of Au lost has been found to linearly increase with irradiation fluence indicating the movement of Au to be not dominated by diffusion. Results indicate the importance of nuclear energy loss in sputtering at very high energies.

Si(1 0 0) substrates were implanted with C n - (n = 1, 3, 5) ions at 5 keV/atom for fluence ranging between 1 × 1014 and 1.6 × 1015 atoms cm−2. The implanted samples were analysed using channeling Rutherford backscattering spectrometry with growth in the surface peak intensity representing growth in damage. At lower fluence cluster ion implantation has been found to result in nonlinear damage production, the nonlinearity increasing with cluster size. Increase in fluence has been found to result in lower nonlinearity. The present values of nonlinearity are much higher for those obtained with C clusters in Si at MeV energies which is primarily because of beam induced annealing effects, coming from electronic energy loss. The cross section for damage production has also been estimated which for all the three values of n yield a consistent result.

ZnS nanoparticles have been observed on catcher foils due to 2 MeV Au ion irradiation of ZnS films thermally evaporated on Si(1 0 0) substrates. The structure and size distribution of nanoclusters collected were studied using transmission electron microscopy for irradiation fluences in the range of 1 × 10 11 –1 × 10 15 ions cm −2 . The nanoclusters were found to have a hexagonal wurtzite structure. Optical absorption measurements on similarly deposited ZnS on silica glass indicate the film to be also composed of hexagonal wurtzite ZnS. Based on this and available data we argue that the observed nanoparticles on the catcher foils are the results of shock waves induced emission of material chunks with the same atomic coordination as in the target.

Embedded Au nanoparticles have been produced in ordinary silica glass by high dose 32 keV Au − ion implantation. Samples implanted with doses higher than 1 × 10 16 cm − 2 are found to show an absorption band peaked around 535 nm, corresponding to the surface plasmon resonance (SPR) of the Au nanoclusters in silica glass. An increase in the SPR peak intensity with increase in dose has been observed upto a dose ∼4 × 10 16 cm − 2 , after which the absorption intensity shows a saturation. This saturation in the SPR intensity has been explained as coming due to a break up of larger Au nanoparticles formed near the surface by displacement spikes induced by subsequently incident Au ions against their regrowth from the movement of Au atoms towards the surface. Further high dose O implantation at 32 keV has been found to result in a drop in absorption. On the contrary, a 2 MeV O implantation has been found to result in a growth in nanoparticle size resulting in an enhanced absorption.

Changes in composition in passivated mercury cadmium telluride (Hg1–xCdxTe or MCT) due to annealing have been investigated using Rutherford backscattering spectroscopy (RBS). Different passivated structures such as anodic oxide/MCT, CdTe/MCT and ZnS/MCT annealed at 80 °C in 10−6 Torr (parameters used during device packaging) have been studied and compared with the as-deposited samples. It is found that in as-deposited samples of anodic oxide/MCT and ZnS/MCT, the interface between the passivating layer and the MCT substrate is very sharp and composition of the capping layer remains constant throughout. Composition of the anodic oxide layer is estimated to be Hg:Te:Cd:O::0.4:1:0.25:3.35 while that of ZnS is Zn:S::0.9:1.1. In an annealed anodic oxide/MCT sample, the composition of the upper layer is different from that in the as-deposited sample and the interface is comparatively wider, while in the annealed ZnS/MCT sample the thickness of the interfacial region is found to be more than 50 nm and the composition of the remaining ZnS layer remains unaltered. The composition of the CdTe layer in both cases (as-deposited and annealed) is Cd:Te::1:1 and the interface thickness of the as-deposited sample is found to be ~35 nm; which on annealing, increases to ~50 nm.

Implantation of 32 keV Au− ions, at the rate of ∼2 × 1013 ions cm−2 s−1, to a dose of 8 × 1016 ions cm−2, into a thermally grown oxide layer on Si(1 0 0), has been found to produce Au nanoclusters with a bimodal size distribution. Rutherford backscattering spectrometry and transmission electron microscopy have been used to characterize the samples. The as-grown nanocluster distribution has ∼6 nm size clusters in the near surface region with a tail of smaller clusters extending into the oxide matrix. A further O implantation (dose in the range of 5 × 1015 to 5 × 1016 atoms cm−2) has been found to result in a breakup of larger Au nanoclusters. This can be understood in terms of the formation of displacement spikes coming from the second implantation. Annealing, both the as-implanted and subsequent high dose O implanted samples, at 800 °C for 1 h has been found to result in a growth in the nanocluster size the later showing a more uniform distribution.

Crystalline superlattices consisting of alternating periods of Si layers and O-atomic layers are potential new channel materials for scaled CMOS devices. In this letter, we investigate Chemical Vapor Deposition (CVD) for the controlled deposition of O-atoms with O-3 as precursor on Si(100) substrates and Si epitaxy on the O-layer. The O-3 reaction at 50 degrees C on the H-terminated Si results in the formation of Si-OH and/or Si-O-Si-H surface species with monolayer O-content. Defect-free epitaxial growth of Si on an O-layer containing 6.4E+14 O-atoms/cm(2) is achieved from SiH4 at 500 degrees C. (C) 2013 The Electrochemical Society. All rights reserved.

The performance, strengths and limitations of RBS and PIXE for the characterization of trace amounts of Cl in TiN thin films are critically compared. The chlorine atomic concentration in ALD grown TiN thin films on Si is determined for samples grown at temperatures ranging from 350°C to 550°C. We show that routine Rutherford backscattering spectrometry measurements (1.5 MeV He+) and PIXE measurements (1.5 MeV H+) on 20 nm thick TiN films allow one to determine the Cl content down to 0.3 at% with an absolute statistical accuracy reaching 0.03 at%. Possible improvements to push the sensitivity limit for both approaches are proposed.

This paper has been withdrawn keeping in view of publication elsewhere with some appropriate modifications., This paper has been withdrawn

The damage growth and surface modifications in Si(100), induced by 25 keV Si$_5^-$ cluster ions, as a function of fluence, $\phi$, has been studied using atomic force microscopy (AFM) and channeling Rutherford backscattering spectrometry (CRBS). CRBS results indicate a nonlinear growth in damage from which it has been possible to get a threshold fluence, $\phi_0$, for amorphization as $2.5\times 10^{13}$ ions-cm$^{-2}$. For $\phi$ below $\phi_0$, a growth in damage as well as surface roughness has been observed. At a $\phi$ of $1\times 10^{14}$ ions-cm$^{-2}$, damage saturation coupled with a much reduced surface roughness has been found. In this case a power spectrum analysis of AFM data showed a significant drop, in spectral density, as compared to the same obtained for a fluence, $\phi < \phi_0$. This drop, together with damage saturation, can be correlated with a transition to a stress relaxed amorphous phase. Irradiation with similar mass Cs$^-$ ions, at the same energy and fluence, has been found to result in a reduced accumulation of defects in the near surface region leading to reduced surface features., Comment: 7 pages, 4 figures

Two exponents, $\delta$, for size distribution of $n$-atom clusters, $Y(n)\sim n^{-\delta}$, have been found in Au clusters sputtered from embedded Au nanoparticles under swift heavy ion irradiation. For small clusters, below 12.5 nm in size, $\delta$ has been found to be 3/2, which can be rationalized as occurring from a steady state aggregation process with size independent aggregation. For larger clusters, a $\delta$ value of 7/2 is suggested, which might come from a dynamical transition to another steady state where aggregation and evaporation rates are size dependent. In the present case, the observed decay exponents do not support any possibility of a thermodynamic liquid-gas type phase transition taking place, resulting in cluster formation., Comment: 4 pages, 3 figures

Ultraviolet (UV) photoluminescence (PL) data taken on a double Au implanted Si matrix are reported. This has been studied over a wide temperature range of 28–220 K. At low temperature, the spectrum shows four peaks corresponding to a zero-phonon line (ZPL) and three low-energy phonon-assisted lines. At 28 K the ZPL has an energy of 3.362 eV with a lifetime of ∼240 ps suggesting transitions across a direct gap. The temperature variation of the ZPL intensity indicates an activation energy ∼10 meV comparable to the binding energy of excitons in crystalline Si (c-Si). For T > 100 K, it shows a redshift which can be understood as coming from exciton–phonon interaction. At lower temperatures the corresponding line broadening shows evidence of interaction with low-energy phonons of average energy ∼7 meV as suggested earlier. All the above results indicate the emission to be coming from Si nanoclusters (NCs) embedded in the matrix.

We have studied the influence of substrate surface roughness on the growth of Ge on self-affine fractal Si surfaces. Self-affine fractal surfaces can be prepared by irradiating surfaces of solid materials with an energetic ion beam, which modifies the surface morphology predominantly via an erosion process. By tuning the ion beam energy, surface parameters such as surface roughness (σ0) and correlation length (ξ), can be controlled. Values of σ0 and σ0/ξ affect the effective free energy of the surface. When materials are grown on surfaces with different values of surface roughness parameters, the morphology of the deposited layer is expected to be influenced. In comparison with an earlier study where 2 MeV Si ion irradiation was used, we have produced surfaces with larger values of σ0 and σ0/ξ by irradiating the surfaces with 45 keV Si ions. We have made a comparison for Ge growth on Si surfaces irradiated with 2 MeV and 45 keV Si ions and have observed remarkably different growth behaviour. Upon Ge deposition, in the former case, the surface roughness decreases at smaller length scales and remains practically unchanged at larger length scales. This indicates that Ge has grown preferentially within the valleys of the rough surface. In the latter case, where σ0 is large and ξ is small on the substrate surface, the roughness increases at all length scales by almost a factor of 2, upon Ge deposition, indicating the lack of preferential growth within the surface valleys. Following Ge deposition, in the former case the roughness exponent increases from 0.82 to 0.99, while in the latter case it remains practically unchanged (~0.83). Roughness parameters of the substrate surface are found to have a significant influence on the morphology of the overlayers.

Au implantation at 32 keV into Si(100), in a fluence range of 1 x 10(15)-1 x 10(17) cm(-2), has been used to produce a gold-rich damaged Si layer of thickness around 30 nm. Local recrystallization of this layer, induced by 1.5 MeV Au irradiation, to a fluence of 1 x 10(15) cm(-2), has been studied using Raman scattering, photoluminescence (PL) and x-ray photoemission spectroscopy (XPS). For a sample with a low energy Au fluence of 5 x 10(15) cm(-2), the MeV Au irradiation has been found to result in Si nanocrystal (NC) formation. The size of the NCs, as estimated from the PL data, has been found to be about 4 nm, which agrees well with the result of a thermal spike model calculation. Annealing of the sample at 500 degrees C resulted in an enhanced PL signal, without any significant shift in peak position, indicating an increase in the local concentration of the NCs. In the case of samples with an initial Au fluence above 1 x 10(16) cm(-2), the MeV Au irradiation has been found to result in better overall recrystallization of the amorphous layer, with silicide formation as observed by XPS. However, there was no PL signal, indicating the absence of Si NCs in the system. The results suggest that the initial amorphizing Au fluence plays a crucial role in Si NC formation induced by MeV ion irradiation.

Nanocrystalline Sn particles have been formed in silica glass through 50 keV Sn− implantation followed by annealing in N2 at 650 °C for 30 min. Samples prepared this way have been annealed in air for 1 h, separately at four different temperatures, 400, 600, 800, and 1000 °C, each at a given temperature. Annealing at temperatures higher than 400 °C has been found to result in oxidation of the Sn nanoparticles (NPs) and formation of the SnO2 phase as confirmed from optical absorption (OA), transmission electron microscopy, and Raman scattering measurements. For the sample annealed at 600 °C, Raman scattering data showed three bands at about 525, 629, and 771 cm−1, the last two corresponding to the A1g and B2g classical Raman modes of rutile SnO2. Increase in annealing temperature resulted in an increase in the intensities of the A1g and B2g modes showing better crystallinity. Also, the A1g peak shifted toward a higher wave number with a steady decrease in the intensity at 525 cm−1. This is in line with the ...

Thermal annealing induced redistribution behavior of Au (3 MeV, 6.0×1015 ions cm−2), implanted into SiO2∕Si(100) substrates, has been investigated using Rutherford backscattering spectrometry and cross-sectional transmission electron microscopy (XTEM). Sequential annealing in the temperature range of 550−750 °C has been found to result in rejection of Au atoms from Si toward the SiO2∕Si interface. Above 750 °C a significant fraction of the implanted Au atoms has been found to back diffuse into deeper regions, well below the projected range Rp of Au. Direct annealing of a sample at 850 °C also shows a similar anomalous diffusion of Au into deeper regions, well below Rp. However, direct annealing at 1050 °C has been found to result in an enhanced accumulation of Au at a deeper layer. XTEM micrographs for the same sample show the presence of Au-rich nanoparticles and dislocations, decorated with Au-rich nanoparticles in this region. Trails of Au-Si liquid nanodroplets along with dislocations extending into t...

ZnS films deposited on Si have been irradiated with Au ions at 35 keV, 2, and 100 MeV. Sputtered particles, collected on catcher foils during irradiation, were analyzed using transmission electron microscopy. For the case of 35 keV Au irradiation, no nanoparticle (NP) could be observed on the catcher foil. However, NPs 2–7 nm in size, have been observed on the catcher foils for MeV irradiations at room temperature. For particle sizes ≥3 nm, the distributions could be fitted to power law decays with decay exponents varying between 2 and 3.5. At 2 MeV, after correction for cluster breakup effects, the decay exponent has been found to be close to 2, indicating shock waves induced ejection to be the dominant mechanism. The corrected decay exponent for the 100 MeV Au irradiation case has been found to be about 2.6. Coulomb explosion followed by thermal spike induced vaporization of ZnS seems to be the dominant mechanism regarding material removal at such high energy. In such a case the evaporated material can ...