Library

Notes: We have investigated the gas-source molecular beam epitaxy growth of GaAs on a (001) Ge surface, off-cut 6° towards the [110]. Initiation of GaAs on Ge with the typical procedure of using a self-terminating As layer invariably produces poor GaAs surface morphology and generates a plethora of antiphase boundaries, despite the large miscut of the (001) surface. However, initiation of GaAs growth with approximately 1 ML (monolayer) of Ga results in single-domain films with excellent surface morphology. We conclude that the Ga monolayer prevents the Ge surface from forming a high step density surface which is established when Ge is exposed to As2.

Notes: The capture kinetics and trapping properties of a dislocation related electron trap detected in strain-relaxed, compositionally graded Ge0.3Si0.7/Si grown by rapid thermal chemical-vapor deposition are investigated by deep-level transient spectroscopy (DLTS). The volume DLTS trap concentration scales linearly with the areal threading dislocation density, as determined by electron-beam-induced current measurements on samples with different compositional grading rates, indicating that the detected trap is most likely associated with dislocation core states in these graded structures. The dislocation related trap exhibits both the logarithmic dependence of DLTS peak height on fill pulse time tp, and broadened DLTS peaks which typically characterize carrier trapping at dislocations. These effects are quantified and analyzed to gain insight into the trapping properties of dislocations in GeSi/Si heterostructures and to investigate the effects of dislocation related carrier trapping on DLTS measurements. It is demonstrated that the peak broadening, as characterized by the dimensionless broadening parameter FWHM/Tp, where FWHM and Tp are the full width at half-maximum of the DLTS peak and the DLTS peak temperature, respectively, monotonically decreases with decreasing fill pulse duration, and approaches point-defectlike behavior for tp〈100 μs.The observed broadening is asymmetric about Tp, and occurs predominantly on the low-temperature side of the DLTS peak. This asymmetric broadening is shown to shift the "apparent'' trap activation energy, as determined by Arrhenius analysis, from EC−0.6 eV to EC−0.9 eV (relative to the bulk conduction-band edge) as tp decreases from 5 ms to 50 μs. These observations are explained by the presence of a dislocation related distribution of energy levels within the GeSi band gap and the consequent fill-pulse-dependent local band bending. The lowest-energy states within this distribution are preferentially filled with electrons for short fill pulse times. The Arrhenius-determined "apparent'' activation energy is hence interpreted as being a measure of the average energy of the filled defect states, weighted by the density of states distribution in this energy band and by the related fill-pulse-dependent local band bending. It is further demonstrated that the minority-carrier capture cross section may be enhanced by the presence of an attractive coulombic barrier for minority carriers at the dislocations, and we use the logarithmic capture equations to derive a value of 4×10−12 cm2 for this "effective'' capture cross section. © 1995 American Institute of Physics.

Notes: Photoluminescence (PL) spectroscopy, cathodoluminescence (CL) spectroscopy and imaging, and preferential defect etching and optical microscopy have been used to characterize compositionally graded Si1−xGex alloy layers grown by molecular beam epitaxy. Si1−xGex capping layers grown on the compositionally graded layers have low threading dislocation densities, and both PL and low-beam energy CL spectra show bound exciton luminescence features identical with those observed in bulk Si1−xGex alloys and relatively weak dislocation related D-band features. Increasing the beam energy increases the relative strength of the D bands in the CL spectra, indicating that they are associated with the misfit dislocations in the compositionally graded layer. This has been confirmed by combined chemical etching and PL spectroscopy measurements. The misfit dislocations can be observed by monochromatic CL imaging at a high-beam energy using a narrow band pass filter centerd on the D4 band.

Notes: A procedure for the fabrication of two-dimensional carrier (electron and hole) gases in modulation doped GeSi/Si heterostructures is presented. The best 4.2 K mobilities measured for the two-dimensional electron and hole gases are 180 000 cm2/V s and 18 000 cm2/V s, respectively. Recently, two-dimensional hole gases with mobilities as high as 55 000 cm2/V s have been obtained. The carrier gases are fabricated on top of relaxed, compositionally graded GexSi1−x buffer layers with low threading dislocation densities (≈106 cm−2). Experimental evidence indicates that the function of the graded buffer is to promote dislocation propagation while suppressing nucleation. A comparative analysis is carried out for two dimensional electron gases in GeSi/Si/GeSi and in AlGaAs/GaAs structures. Although molecular beam epitaxy is used to grow the samples, the principle discussed here is independent of growth technique.

Notes: We study the electrical activity of threading dislocation defects in relaxed GeSi films with a novel, high-resolution optical technique. A near-field scanning optical microscope is used to measure spatially resolved photoresponse while simultaneously imaging the surface topography. We have convincingly established that shallow topographic depressions in these films are electrically active threading dislocations. The apparent sizes of the dislocations in the photovoltage images are in agreement with estimates based on the junction geometry and the near-field optical excitation spot size. We can clearly observe photoresponse changes at ≤100 nm lateral scale, a tenfold improvement from far-field optical techniques. This higher resolution is due to reduction of the excitation volume and of the carrier lifetime near defects. © 1996 American Institute of Physics.

Notes: The Si substrates were patterned and etched to create arrays of squares separated by 10 μm trenches. Some edges of these squares were ion implanted with a large dose of Ge to create damaged regions that act as misfit dislocation nucleation centers during molecular beam epitaxy of a 700 nm Si0.90Ge0.10 lattice mismatched layer. Low temperature photoluminescence spectra of the material were obtained with an illuminated area selected so that only one type of edge implant was studied at a time. Sub-band gap luminescence peaks associated with dislocations, D lines, were found only in the squares with implants, consistent with electron microscopy results that showed that the control squares had few misfit dislocations. A large D1 luminescence peak was observed in samples that consisted of intersecting arrays of misfit dislocations, while only a small peak was present in specimens implanted such that dislocations lay in only one in-plane direction. These results are interpreted to mean that the D1 peak is intrinsic to the dislocations themselves and not to impurity gettering. These observations further infer that D1 luminescence is related to dislocation kinks. © 1998 American Institute of Physics.

Notes: Strain-relaxed Ge0.3Si0.7/Si, grown by rapid thermal chemical-vapor deposition, has been investigated with deep-level transient spectroscopy (DLTS) and bias-dependent electron-beam-induced current (EBIC). A single electron trap and several hole traps have been detected in these samples. The apparent electron capture cross section is found to be ∼2×10−13 cm2, which is several orders of magnitude larger than the apparent hole capture cross sections (∼10−17 cm2), and is responsible for the detection of the minority-carrier electron trap even under reverse-bias majority-carrier capture conditions. All observed traps which were investigated as a function of filling pulse time exhibit logarithmic capture kinetics, as expected for extended defects, and the bias-dependent DLTS peak height and EBIC relative defect contrast are consistent with the spatially varying dislocation density. Moreover, the trap concentration, as determined by DLTS, is correlated to the dislocation density, as determined by EBIC measurements. Based on a comparison of Arrhenius plots, the observed logarithmic capture kinetics, the correlation of trap density to dislocation density, and the observed bias dependence, the electron trap appears to be related to dislocation core states, while two of the hole traps appear to be related to either dislocation core states or Cottrell atmospheres. © 1995 American Institute of Physics.

Notes: The luminescence properties of 3 μm thick, strongly emitting, and highly porous silicon films were studied using a combination of photoluminescence, transmission electron microscopy, and Fourier transform infrared spectroscopy. Transmission electron micrographs indicate that these samples have structures of predominantly 6–7 nm size clusters (instead of the postulated columns). In the as-prepared films, there is a significant concentration of Si—H bonds which is gradually replaced by Si—O bonds during prolonged aging in air. Upon optical excitation these films exhibit strong visible emission peaking at ≈690 nm. The excitation edge is shown to be emission wavelength dependent, revealing the inhomogeneous nature of both the initially photoexcited and luminescing species. The photoluminescence decay profiles observed are highly nonexponential and decrease with increasing emission energy. The 1/e times observed typically range from 1 to 50 μs. The correlation of the spectral and structural information suggests that the source of the large blue shift of the visible emission compared to the bulk Si band gap energy is likely to be due to quantum confinement in the nanometer size Si clusters. The electron-hole recombination process, on the other hand, remains unclear.

Notes: A detailed study of the electrical and defect properties of ion-implanted erbium in silicon shows that erbium doping introduces donor states. The concentration of erbium related donors as a function of implant dose saturates at 4×1016 cm−3 at a peak implanted Er-ion concentration of (4–7)×1017 cm−3. The defect levels related to erbium in silicon are characterized by deep level transient spectroscopy and identified as E(0.09), E(0.06), E(0.14), E(0.18), E(0.27), E(0.31), E(0.32), and E(0.48). The dependence of the photoluminescence on annealing temperature for float zone and for Czochralski-grown silicon show that oxygen and lattice defects can enhance the luminescence at 1.54 μm from the erbium. Temperature-dependent capacitance-voltage profiling shows donor emission steps when the Fermi level crosses EC − ET = 0.06 eV and EC − ET = 0.16 eV.