Search

Two shallow hole traps dominate the deep level transient spectroscopy (DLTS) data for strain-relaxed Si0.7Ge0.3 layers grown on Si(100) by ultrahigh vacuum chemical vapor deposition at temperatures ⩽560 °C. The trap energy levels are at Ev+0.06 and Ev+0.14 eV and trap concentrations are ⩽5×1014 cm−3 in relaxed layers having threading dislocation densities of 2–4×107 cm−2. A logarithmic dependence of the filling rate indicates that these traps are associated with extended defects and this is confirmed by their absence in a sample having no dislocations. The annealing temperature of the DLTS peaks is consistent with the interpretation of these traps as states of defect complexes at dislocations, rather than intrinsic dislocation states or isolated defect complexes. The trap concentrations are proportional to the oxygen concentration in the film, suggesting that oxygen may be a constituent of the defect complex.

Sensitive curvature measurements were performed on 40 A period W/C multilayer structures on Si substrates using a two beam laser reflection technique. A compressive stress of approximately 1530 MPa was measured in these sputtered multilayer films. Thermal annealing to 500 °C resulted in very little strain relaxation in the multilayers but x‐ray diffraction data show a slight increase of the multilayer period. Significant strain relaxation, though, was observed when a 400 A W buffer layer was included. Thermal annealing of these samples to 400–500 °C resulted in large strain relaxation due to the growth of α‐W crystals in the buffer layer. Large oxide formation on air annealed samples did not significantly change the strain state of the films.

We demonstrate that depositing Ta diffusion barriers under ultra‐high vacuum conditions without in situ oxygen dosing allows for variations both in microstructure and in the concentration of chemical impurities that severely degrade barrier performance. The effects of deposition pressure, in situ oxygen dosing at interfaces, hydrogen and oxygen contamination, and microstructure on diffusion barrier performance to Cu diffusion for electron‐beam deposited Ta are presented. 20 nm of Ta diffusion barrier followed by a 150 nm Cu conductor were deposited under ultra‐high vacuum (UHV, deposition pressure of 1×10−9 to 5 ×10−8 Torr) and high vacuum (HV, deposition pressure of 1×10−7 to 5×10−6 Torr) conditions onto 〈100〉 Si. In situ resistance furnace measurements, Auger compositional depth profiling, secondary ion mass spectrometry, and forward recoil detection along with scanning and transmission electron microscopy were used to determine the electrical, chemical, and structural changes that occurred in thin‐film Ta diffusion barriers upon annealing. Undosed HV deposited Ta barriers failed from 560 to 630 °C, while undosed UHV barriers failed from 310 to 630 °C. For UHV Ta barriers, in situ oxygen dosing during deposition at the Cu/Ta interface increased the failure temperatures by 30–250 °C and decreased the range of failure temperatures to 570–630 °C. Undosed UHV Ta barriers have no systematic relationship between failure temperature and deposition pressure, although correlations between breakdown temperature, oxygen and hydrogen concentrations, and microstructural variations were measured.

We have studied the formation of As precipitates in doped GaAs structures that were grown by molecular beam epitaxy at low substrate temperatures and subsequently annealed. We find that the As precipitates form preferentially on the n side of such fabricated GaAs pn junctions. As the coarsening process proceeds, there is a gradual increase in the amount of As in precipitates in the n‐GaAs region and a decrease in the p‐GaAs region; the depletion region between the pn junction becomes free of As precipitates. These observations can be understood qualitatively based on the charge states of the As interstitial and using thermodynamic arguments in which the crystal attempts to minimize the chemical potential during the anneal. The presence of the excess As results in a stable Be profile even to anneals of 950 °C. Finally, a temperature cycling technique to grow arbitrarily thick GaAs epilayers containing As precipitates was demonstrated.

The epitaxial growth of GaAs and (Al,Ga)As doped with boron (from diborane) and silicon was investigated to examine the effect of boron on (DX) centers in silicon‐doped material. The addition of diborane to the growth of GaAs and (Al,Ga)As results in the superlinear incorporation of boron into the solid with a concurrent reduction in the growth rate. Boron incorporation also decreases as the growth temperature is increased. Additionally, the AlAs mole fraction increases with increasing diborane during (Al,Ga)As growth. The DX center was not eliminated in (Al,Ga)As by the addition of boron. The thermal stability of these materials was also investigated.

Secondary ion mass spectroscopy has been used to quantitatively determine the carbon concentration in nominally undoped GaAs grown by metalorganic vapor phase epitaxy from TMG (13C 99%) and AsH3. Both an increase in the V/III ratio and the addition of supplemental gas phase radicals reduced the carbon incorporated from the TMG. Higher V/III ratios are proposed to increase the surface concentration of AsHx species. Supplemental gas phase t‐butyl radicals, produced from the decomposition of azo‐t‐butane, are proposed to attack AsH3, also resulting in an increase in the surface concentration of AsHx species. Higher surface concentrations of AsHx are then proposed to reduce carbon incorporation by enhancing the desorption of carbon‐containing species.

The debated question of the superluminal speed of information remains open to different interpretations in spite of recent, apparently conclusive, contributions. A different point of view is here considered, which is based on experimental results of microwave propagation experiments, in far-field conditions. They support the hypothesis of the existence of fast waves of the Zenneck type. (c) 2006 American Institute of Physics.

The outdiffusion of Be implanted into GaAs has been found to be identical after capless or capped (Si3N4 or SiO2 ) rapid thermal annealing (RTA) at 900–1000 °C and to depend on the Be dose and its proximity to the surface. The outdiffusion is more pronounced when the Be implant is shallow ( 1×1015 cm−2 ). It is demonstrated that the Be outdiffusion is driven by the presence of a highly damaged surface layer. Auger results suggest the formation of a BeOx compound at the surface of a high‐dose (1×1016 cm−2 ) Be‐implanted sample that underwent capless RTA at 1000 °C/1 s. It appears that BeOx formation occurs when the outdiffused Be interacts with the native Ga/As oxides during annealing. All the Be remaining in the GaAs after a >900 °C/2 s RTA is electrically active.

We report a detailed deep level transient spectroscopic study in p‐type Mg‐ and Zn‐doped GaAs epitaxial layers grown by metal‐organic vapor phase epitaxy. Dependence of deep level structures on doping concentrations and growth temperatures has been investigated. Over a wide range of growth conditions, four hole traps and an electron trap ranging in activation energy from 0.18–0.79 eV were measured in GaAs:Mg while only a single hole trap has been observed in GaAs:Zn.The presence of a certain trap and its concentration in GaAs:Mg depends mainly on the doping concentration in the layers. The total trap concentration in the GaAs:Mg decreases rapidly with doping concentration for p>4×1017 cm−3. The physical and chemical origins of several of these traps have been identified. The Mg‐doped GaAs always exhibited a greater concentration of midgap trap levels than the Zn‐doped material, regardless of dopant concentration or growth temperature. The overall defect structure and dopant incorporation characteristics i...

We have explored the potential of proton‐induced x‐ray spectroscopy as a quantitative analytical technique to determine the composition of a homogeneous complex target. Direct production of x ray by protons in the low MeV range was considered using the binary encounter approximation and published data on the atomic number dependance of the stopping power. Also included is a correction scheme for secondary excitation processes if radiation is generated in the target capable of fluorescence excitation. The formalism of the correction scheme is closely related to the one commonly employed in electron microprobe spectroscopy. Our data indicate that the accuracy achievable in IIXS is of the order of ±5%, if one considers the two important corrections: (1) the difference in the stopping power properties of the composite target compared to the pure element target; (2) secondary excitation processes if required.

A Sm2.48Lu0.26(Pb,Pt)Fe5O12 garnet film composition grown earlier1 by LPE on NdGaG substrates oriented (100) and (111) is now diluted (in steps) to 56 mol% of a hypothetical La1.35Lu1.7Fe5O12 garnet. Most of the results reported are for films grown at 830 °C, but the effects of changing growth temperature for the compositions with the largest La–Lu substitutions are also discussed. La–Lu substitutions in fluxed melts significantly decrease film growth rates, and La must be present in melts in amounts in excess of those desired in the final crystal film. Both Pb from the fluxed melt and Pt from the crucible via the melt are incorporated into films in minor quantities. The ratio of rare earths and Pb to Fe and Pt in films is greater than the stoichiometric 0.6. The films support bubbles in the 0.5μm diameter range and have characteristic length l values of 0.037 to 0.062μm. Saturation magnetization values are 1700 to 2000 g (×10−4T). Domain wall damping coefficients are reduced in proportion to the amount o...