Carbon clustering in Si1−xCx formed by ion implantation

Silicon-carbon alloys were formed by multiple energy implantation of C⁺ ions in silicon and in Silicon on Sapphire (SOS). The ion fluence ranged between 5 × 10¹⁶ − 3 × 10¹⁷ ions/cm² and the energy between 10–30 keV in order to obtain constant carbon concentration into a depth of 100 nm. The carbon atomic fraction (x) was in the range 0.22–0.59 as tested by Rutherford backscattering spectrometry (RBS). Thermal annealing of the implanted films induced a transition from amorphous to a polycrystalline structure at temperatures above 850°C as detected by Infrared spectrometry (IR) in the wavenumber range 600–900 cm⁻¹. The optical energy gap and the intensity of the infrared signal after annealing at 1000°C depended on the film composition: they both increased linearly with carbon concentration reaching a maximum at the stoichiometric composition (x = 0.5). At higher carbon concentration the IR intensity saturated and the optical energy gap decreased from the maximum value of 2.2 to 1.8 eV. The behaviour at the high carbon content has been related to the formation of graphitic clusters as detected by Raman spectroscopy.     

Effects of secondary electrons emitted from surroundings on defect formation in silica glass under γ-ray irradiation

We have investigated the effects of secondary electrons and photons emitted from surrounding materials on defect formation in silica glass under γ-ray irradiation. SiO₂ (silica) glass plates and those sandwiched in a pair of various material disks (carbon, stainless steel or lead) were irradiated by γ-ray, and the optical absorption spectra (UV–vis spectra) of the silica glass plates before and after the irradiation were examined. UV–vis spectra of the glass plates after the irradiation showed three absorption bands peaked around 2 eV, 4 eV and 5.8 eV being assigned to color centers relate metal impurities (Al and Ge) and oxygen-deficient centers like E′ center, respectively. All three bands were found to grow with γ-ray irradiation dose and saturated at higher doses, and absorbance of the bands at the saturation for the sandwiched glass plates was higher than that for the bare glass plate. Moreover, the saturated absorbance was higher for the glass plate sandwiched with heavier materials. Employing Monte Carlo N-Particle (MCNP) code for the simulation of the photon–electron transport process, enhanced energy deposition and numbers of secondary electrons and photons emitted from sandwiching material disks to a silica glass plate were calculated. The higher deposition energy correlates well to the higher saturated absorbance, indicating that the secondary electrons and photons emitted from the disks clearly enhanced the defect formation in the sandwiched silica glass plates. This suggests the existence of the dose effect above a critical does, i.e. the irradiation with higher dose will result in higher saturated absorbance.     

Thermal stability of radiation-induced optical centers in non-stoichiometric spinel crystals

The successive anneals of the neutron-irradiated non-stoichiometric spinel crystals MgO · 2.2Al₂O₃ leads to incremental change of optical spectra that demonstrates two main bands whose intensity and spectral position depends on annealing temperature. While temperature increases from 450 to 750 K, one of the bands shifts from photon energy of 4.2–5.1 eV. Another one shifts in the opposite direction from 6.4–5.6 eV. This effect can be attributed to coagulation of radiation induced defects near cation vacancies and change of the energy levels and transitions in F⁺- and F-centers in neutron-irradiated crystals. The final position of these two bands 5.1 and 5.6 eV corresponds to transitions in F⁺- and F-centers in non-stoichiometric spinel, respectively. The investigation of optical centers induced at subsequent UV-illumination of neutron-irradiated crystals annealed to temperature of 750 K, and comparison with as-grown crystals, shows the existence of residual concentration mainly of the antisite defects and partially of the anionic vacancies in the neutron irradiated and annealed crystals. The existence of two temperature stages where optical centers at antisite defects are effectively destroyed may indicate the presence of spatially correlated and isolated antisite defects in irradiated spinel.     

Defect creation caused by the decay of cation excitons and hot electron–hole recombination in wide-gap dielectrics

Insufficient radiation resistance of construction materials is the Achilles heel for thermonuclear energetics. In wide-gap dielectrics, Frenkel defects are created not only because of the knock-out (impact) mechanism but also because of the decay of the electronic excitations formed during the irradiation (i.e. due to nonimpact mechanisms). The processes of the defect creation at the irradiation of highly pure LiF single crystals at 6–8 K by 1–30-keV electrons, X-rays, or synchrotron radiation of 12–70 eV have been investigated. The annealing processes of these defects in a temperature range up to 200 K have been studied as well. In LiF, creation has been revealed for the following: (1) F–H pairs caused by the decay of anion excitons or by the recombination of electrons and holes, (2) F′–H–V_K and F–I–V_K defect groups at the decay of cation excitons (62 eV), or (3) 20-keV electron irradiation. The mechanism of defect creation at the recombination of hot holes and hot electrons has been discussed for -SiO₂ crystals with an energy gap between the subbands of a valence band. One of the possible ways to suppress this mechanism (“luminescent defence”) is doping a material by luminescent impurities able to capture a part of the energy of hot carriers before their relaxation and recombination (e.g. in MgO:Cr).     

Growth of amorphous carbon films by carbon atom bombardment in the energy range 10–500 eV

Molecular dynamics (MD) simulations were carried out of the energetic impact of C atoms on amorphous carbon (a-C) substrates in the energy range 10–500 eV. The results show that a densification of the deposited material occurs up to 150 eV energy but with a low sp³ bond content. At 500 eV impact the substrate material shows a significant decrease in density and a porous a-C structure forms.     

Radiation-induced luminescence in magnesium aluminate spinel crystals and ceramics

Radioluminescence (RL) and thermoluminescence (TL) in spinel crystals and ceramics were investigated to elucidate the radiation-induced electronic processes in single crystals grown by Verneuil and Czochralski methods as well as transparent and translucent ceramics. Both RL and TL spectra demonstrate a UV-band related to electron–hole recombination luminescence at intrinsic defects; green and red luminescence are identified with emission of Mn²⁺- and Cr³⁺-ions, respectively. The kinetics of growth of different RL luminescence bands depending on dose at the prolonged X-irradiation shows the competitive character of charge and energy transfer between defects and impurity ions. The dependence of RL intensity on the temperature of the sample was measured in the range of 300–750 K and compared with TL for different emission bands. The variety of maxima in the temperature dependence of RL and in the glow curves of TL measured for different luminescence bands in spinels of different origins and crystalline forms is used to show that charge carrier traps and luminescence centers are not isolated defects but are complexes of defects and impurities. The formation, structure and properties of these complexes depend on the processing conditions.     

Influence of annealing temperatures and time on the photoluminescence properties of Si nanocrystals embedded in SiO2

We report on the effects of annealing conditions on the photoluminescence from Si nanocrystal composites fabricated by implantation of Si ions into a SiO₂ matrix, followed by thermal treatment in a nitrogen atmosphere. The evolution of the photoluminescence under different annealing temperatures (900–1100 °C) and annealing time (0.5 up to 5 h) were systematically studied for the implanted samples. After annealing the spectra presented two photoluminescence bands: one centered at 610 nm and another around 800 nm. Combined with transmission electron microscopy, we conclude that the photoluminescence behavior of the two bands suggests different origins for their emissions. The 610 nm band has its origin related to matrix defects, while the 800 nm band can be explained by a model involving recombination via quantum confinement effects of excitons in the Si nanocrystals and the interfacial states recombination process confined in the interfacial region between nanocrystals and SiO₂ matrix.     

Low-temperature irradiation effects in lithium orthosilicates

The emission spectra of lithium orthosilicates (Li₄SiO₄₎ ceramics have been measured in the range of 1.8–5.8 eV under irradiation by 6–30 eV photons or 1–30 keV electrons at 6–300 K. The tunnel recombination phosphorescence, as well as luminescence, stimulated by 1.5–2.5 eV photons has been detected in the sample preliminarily irradiated at 6 or 80 K. The main peaks of thermally stimulated luminescence (TSL) in the irradiated ceramics have been observed at 72, 118 and 265 K. The creation spectra of the 118 K TSL peak, as well as the excitation spectrum of photostimulated luminescence (PSL) span the region of the intrinsic absorption of a lithium orthosilicate (9–30 eV). The intensity of PSL and the TSL peaks in Li₄SiO₄ ceramics prepared in hydrogen/argon atmosphere is several times lower than that in the mainly investigated Li₄SiO₄ ceramics prepared in the atmosphere of dry argon. The optical characteristics of Li₄SiO₄ are compared with the ones known for Li₂O and SiO₂. Low-temperature luminescent methods are promising for the investigation of electron–hole processes and radiation defects serving as the traps for tritium released in D–T fusion reactor blanket systems.     

Sputtering of hydrocarbons by ion-induced breaking of chemical bonds

While the physical sputtering of atoms caused by keV and MeV ions has been studied extensively both by molecular dynamics (MD) simulation and experiments, the mechanisms leading to atom and molecule erosion at energies 1–100 eV are not very well understood. We now describe how low-energy hydrogen ions can cause erosion of carbon atoms and hydrocarbon molecules by entering the region of a carbon–carbon chemical bond and hence breaking it, a process we call ‘swift chemical sputtering'. In the particular case of hydrogen bombardment of carbon-based materials, we further show that this can lead to erosion yields far exceeding those expected for a physical sputtering process.     

RBS analysis of GaAs and InP after electron beam annealing

The decomposition of GaAs and InP surfaces during scanned electron-beam rapid thermal annealing (RTA) has been investigated. The molecules evaporated from the uncapped surfaces during annealing were collected on Si substrates and analysed using 2.0 and 3.5 MeV ⁴He Rutherford backscattering (RBS). The evaporation behaviour was determined in the temperature ranges 600–830° C (GaAs) and 450–630° C (InP). Deep-level transient spectroscopy, current-voltage and capacitance-voltage measurements were used to characterize the electrical properties of the annealed samples. Comparison of RBS and electrical measurements yields an optimum annealing temperature for the chosen annealing technique.     

Mean excitation energy for the stopping power of light elements

We have evaluated the mean excitation energy or I value for Coulomb excitations by swift charged particles passing through carbon, aluminum and silicon. A self-consistent Kramers–Kronig analysis was used to treat X-ray optical spectra now available from synchrotron light sources allowing us to carry out Bethe’s original program of evaluating I from the observed dielectric response. We find that the K and L shell are the dominant contributors to I in these light elements and that the contribution of valence electrons is relatively small, primarily because of their low binding energy. The optical data indicate that Si and Al have nearly equal I values, in contrast to Bloch’s Thomas–Fermi result, I ∝ Z. The optically based I values for C and Al are in excellent agreement with experiment. However, the dielectric-response I value for Si is 164 ± 2 eV, at variance with the commonly quoted value of 173 ± 3 eV derived from stopping-power measurements.     

Silicon detector response to heavy ions at energies of 1–2 MeV/amu

The response of silicon detectors has been measured for He, O, S, Cl, Br, Ag and Pb ions in the energy range 1–2 MeV/amu. Following deliberate, long exposures of the detector, a transient effect was observed for 140 MeV Br ions, in which the pulse height decreased with increasing ion dose and then partially recovered within an hour of the final exposure. Using brief, consecutive exposures, the effective energy for creating a detectable electron–hole pair was determined using the pulse height difference method. The energy deposited by ions in the ‘dead-layer' at the detector surface and energy loss via non-ionizing events was taken into account. For ions with atomic numbers 2Z17 and energies above the Bragg peak, the effective energy was found to decrease linearly with increasing electronic stopping power at first, and then to level off at 3.52 eV/electron–hole pair. For intermediate mass ions (17<Z40), at energies close to the Bragg peak, increases slightly (2%) with increasing stopping power. For the heaviest ions studied (Z40), whose energies are below the Bragg peak of the stopping power curve, increases strongly (10–20%), even though the electronic stopping power is approximately constant.     

Injection Dependence of Transient Annealing in Neutron-Irradiated Silicon Devices

Studies have been performed to explore accurately the injection level and temperature dependence of transient annealing in neutron-irradiated P- and N-type silicon. In P-type material, the annealing factor in the 0 to 0.1 second time interval is very sensitive to the minority carrier injection level. For example, by varying the injection level from 10-5 to 10-1 the annealing factor at 0.001 second can be reduced from 10 to approximately 2. In contrast to the P-type results, the injection dependence observed in N-type silicon is very small and, furthermore, is in the opposite sense; i.e., an increase in the injection level causes an increase in the annealing factor. However, this study shows that this seemingly different behavior can be correlated on the basis of the hole-to-electron ratios of the different material types and resistivities. Annealing measurements performed in the temperature range from 180°to 300°K reaffirm the 0.3 eV activation energy previously found in P-type silicon and establish a value of 0.17 eV for N-type silicon.     

Low energy nitrogen ion doping into GaAs using combined ion-beam and molecular-beam epitaxy method

Low energy nitrogen (N) ions were irradiated during the epitaxial growth of GaAs using combined ion beam and molecular beam epitaxy (CIBMBE) method as a function of N⁺ ion acceleration energy (E_a) and N⁺ ion beam current density (I_N). E_a was varied from 70 to 170 eV I_N from 900 pA/cm² to 75 nA/cm². GaAs growth rate was fixed to 1 μm/h. In 2 K photoluminescence (PL) spectra of the samples with I_N = 3 nA/cm² and E_a = 70–100 eV, two sharp emissions at 1.508 eV (X₁) and 1.495 eV (X₂), which have been attributed to the emissions of excitons bound to isolated N atoms, and another one at 1.443 eV (X₅) were observed. These results show that nitrogen (N) atom in GaAs becomes optically active as an isoelectronic impurity at least in as-grown condition. For N⁺ ion-irradiated samples with rather high I_N, e.g., with I_N = 75 nA/cm² and E_a = 100 eV, a broad emission together with multiple sharp ones were observed after furnace annealing at 750°C which were ascribed to emissions of excitons bound to nitrogen-nitrogen (N---N) pairs.     

Luminescent nanoclusters array fabricated by pulsed laser beam irradiation

Ordered luminescent nanoclusters array in the form of grating structures are fabricated on silicon (1 0 0) surface by Q-switched Nd:YAG laser beam irradiation of second harmonic wavelength (532 nm) in vacuum. Blue-green photoluminescence (PL) spectrum from the ordered nanoclusters array exhibits two asymmetrical peaks at 2.58 eV and 2.88 eV in the blue-green region corresponding to the bimodal distribution of nano size clusters. The size of the nanoclusters is estimated from the three dimensional quantum-confined model incorporating Gaussian size distribution. When subjected to rapid thermal annealing at 710 °C for 10 min in N₂ atmosphere there is an enhancement of the PL intensity without any change in the peak emission energy and broadening suggesting that the origin of PL is related to quantum confinement effect in Si nanocrystallite. The surface morphology of the irradiated surface varies considerable with the number of laser shots, laser fluence and ambient conditions.     

Reactive ion etching of PECVD silicon nitride in SF6 plasma

The reactive ion etching of PECVD silicon nitride thin films has been investigated using SF₆ plasma. Effects of variations of process parameters such as pressure (50–350 mTorr), RF power (50–250 W), gas flow rate (3–130 sccm) and additions of O₂ and He (0–50%) in SF₆, on the PECVD silicon nitride etch rate and selectivity to the AZ 1350J photoresist were examined. An etch rate of 1 μm/min has been obtained under the condition of 150 mTorr, 100 W and 60 sccm. Experimental results also indicated a maximum etch rate at approximately 30% O₂ while addition of He showed only dilution effect. A nitride/photoresist selectivity ranging from 1 to 3:1 has been obtained.     

A molecular dynamics study of the clustering of implanted potassium in multiwalled carbon nanotubes

We have studied the low energy irradiation of carbon nanotubes (CNT) with K ions using classical molecular dynamics simulations with analytical potentials. The studied CNTs had diameters of about 0.5–1.2 nm and single or multiple walls. The average penetration depth and probabilities to introduce an impurity atom into CNT were studied with simulations on irradiating the CNT with single K ion. The number of potassium clusters, their average sizes and the damage produced into the CNT due to the irradiation were studied using multiple K ion irradiations. We found that the K ions are mobile in CNTs right after the implantation event and that they cluster together. For CNTs with 1–3 coaxial tubes, the highest ratio of K atoms in clusters per total number of K ions was obtained by using an irradiation energy of about 100 eV. Also the least damage per K ion was found to be produced into the CNT with this energy when those energies high enough for the ion to penetrate the outermost wall of the CNT were considered.     

The formation of compound layers in silicon by ion beam synthesis

The technique of ion beam synthesis (IBS) using high doses of energetic ions has been successfully implemented to produce a variety of compounds, the physical properties of which are dependent on the implanted species and range from insulators, e.g. SiO₂, through semiconductors, e.g. SiC, to conductors, e.g. CoSi₂. In this paper we study the evolution of these compounds and compare and contrast their methods of formation. To demonstrate the versatility of the technique we look at three examples of IBS layers: (1) To date most of the interest in IBS has concentrated on the production of buried oxide layers for silicon-on-insulator (SOI) device applications. Recently it has been shown that by using a series of sequential implants and high-temperature anneals the defect density in the silicon overlayer can be dramatically reduced. To study how this process occurs, we followed the redistribution of the implanted species during implantation and annealing using both ¹⁶O⁺ and ¹⁸O⁺. (2) Buried CoSi₂ layers can be fabricated in (100) single-crystal silicon by implanting high doses of energetic cobalt ions at elevated temperatures. For the higher doses (≥ 4 × 10¹⁷ O⁺/cm² at 350 keV), a continuous coherent layer of CoSi₂ grows epitaxially during implantation. For lower doses, precipitates of both A- and B-type CoSi₂ are observed. After annealing at 1000° C for 30 min, single-crystal aligned layers are produced for the higher doses, while for lower doses discrete octahedral A-type precipitates are formed. (3) The microstructures of synthesized SiC layers are more complex than analogous synthesized oxide or silicide layers. Unlike buried oxide layers, the carbon concentration at the peak of the implanted distribution does not saturate at a value equivalent to that in the stoichiometric compound, but continues to rise, reflecting the lower diffusivity of the C in the synthesized compound layer. To achieve chemical segregation of the implanted carbon, very-high-temperature (≥ 1300°C), long-time (typically 20 h) anneals are required. At the interface with the silicon substrate the synthesized layer grows with a degree of epitaxy. This is also found to occur during implantation if the temperature is ≥ 650° C.     

On the use of silicon as thermal neutron filter

A simple formula is given which allows to calculate the contribution of the total neutron cross-section including the Bragg scattering from different (hkl) planes to the neutron transmission through a solid crystalline silicon. The formula takes into account the silicon form of poly or mono crystals and its parameters. A computer program DSIC was developed to provide the required calculations. The calculated values of the total neutron cross-section of perfect silicon crystal at room and liquid nitrogen temperatures were compared with the experimental ones. The obtained agreement shows that the simple formula fits the experimental data with sufficient accuracy. A good agreement was also obtained between the calculated and measured values of polycrystalline silicon in the energy range from 5 eV to 500 μeV. The feasibility study on using a poly-crystalline silicon as a cold neutron filter and mono-crystalline as a thermal neutron one is given. The optimum crystal thickness, mosaic spread, temperature and cutting plane for efficiently transmitting the thermal reactor neutrons, while rejecting both fast neutrons and gamma rays accompanying the thermal ones for the mono crystalline silicon are also given.     

Effect of oxygen on ion-beam induced synthesis of SiC in silicon

The properties of Si-structures with a buried silicon carbide (SiC) layer created by high-dose carbon implantation into Cz–Si or Fz–Si wafers followed by high-temperature annealing were studied by Raman and infrared spectroscopy. The effect of additional oxygen implantation on the peculiarities of SiC layer formation was also studied. It was shown that under the same implantation and post-implantation annealing conditions the buried SiC layer is more effectively formed in Cz–Si or in Si (Cz-or Fz-) subjected to additional oxygen implantation. So we can conclude that oxygen in silicon promotes the SiC layer formation due to SiO_x precipitate creation and accommodation of the crystal volume in the region where SiC phase is formed. Carbon segregation and amorphous carbon film formation on SiC grain boundaries were revealed.