Search

A combination of high resolution Laplace deep level transient spectroscopy (LDLTS) and direct capture cross-section measurements has been used to investigate whether deep electronic states related to interstitial-type defects introduced by ion implantation originated from point or extended defects, prior to any annealing. n-type silicon was implanted with doses of 1×10[sup 9] cm[sup -2] of silicon, germanium, or erbium, and comparison was made with proton- and electron-irradiated material. When measured by LDLTS at 225 K, the region of the implant thought to contain mostly vacancy-type defects exhibited a complex spectrum with five closely spaced defect-related energy levels, with energies close to E[sub C]-400 meV. The region nearer the tail of the implant, which should be dominated by interstitial-type defects, exhibited a simpler LDLTS spectrum with three closely spaced levels being recorded, again with energies centered on E[sub C]-400 meV. Annealing at 180 °C did not completely remove any of the defect peaks, suggesting that the energy levels were not due to the simple vacancy-phosphorus center. Direct electron capture cross-section measurements revealed that the defects in the tail of the implanted volume, prior to any annealing, were not simple point defects, as they exhibited nonexponential capture properties. This is attributed to the presence of extended defects in this region. By contrast, defects with the same activation energy in proton- and electron-irradiated silicon exhibited point-defect-like exponential capture. © 2003 American Institute of Physics. [ABSTRACT FROM AUTHOR]

We have developed a high resolution technique for examining minority carrier emission from defect states in semiconductors called Laplace minority carrier transient spectroscopy (LMCTS). The experiment uses the same analytical approach to the capacitance transient as Laplace deep level transient spectroscopy (LDLTS), but minority carriers are injected into the depletion region by application of a suitable light pulse. The combination of LDLTS and LMCTS means that detailed emission properties of closely spaced majority and minority carrier traps across the whole band gap can be now characterized. The technique has been used to study minority carrier traps in gas source molecular beam epitaxy-grown Si/Si[SUB0.86]Ge[SUB0.14] strained quantum wells. Initially the technique was evaluated by comparing LMCTS of a hole trap associated with the gold-hydrogen complex in n-type silicon with LDLTS of the same trap in p-type silicon. Both techniques confirm that this level consists of two states, as previously suggested in the literature. LMCTS was then applied to an n-type multiquantum well Si/SiGe layer. We have been able to measure directly the emission rate of holes out of SiGe quantum wells using LMCTS. The emission rate exhibited only slight temperature dependence, in strong contrast to that of holes which are thermally emitted from isolated point defects. We show that in the particular case of LMCTS, a temperature invariant emission rate out of quantum wells is to be expected, and this is consistent with theoretical predictions. [ABSTRACT FROM AUTHOR]

We have carried out high resolution Laplace deep level transient Spectroscopy (DLTS) and conventional DLTS on silicon implanted with very low doses of either silicon, germanium, erbium, or ytterbium, and compared the results to those from electron-irradiated silicon. DLTS spectra of all the samples initially look very similar, and a peak at 95 K appears in all spectra which may be due to the vacancy-oxygen (VO) defect. We have carried out detailed measurements of the capture cross section and activation energy of this defect using Laplace DLTS. We show that, when the mass of the implanted ion exceeds that of silicon, the defect has a much smaller electron capture cross section than that expected for the VO defect, and a smaller activation energy. Hydrogen has been introduced, either by wet chemical processing or plasma, to all samples to observe the hydrogenVO interactions resulting in VOH. By using high resolution DLTS we are able to establish that, after hydrogenation, the VOH defect exists with an identical emission rate in the silicon-implanted silicon and the electron-irradiated silicon, but not in the silicon implanted with heavier ions. We conclude that the peak at 95 K in the DLTS spectra in the case of the heavier ions is due to a different defect, confirming earlier reports in the literature. This defect is negatively charged, unlike VO, which is acceptor-like. We are also able to observe VOH in samples where VO is not present, after these samples have been annealed. We attribute this to release of V and H atoms from other defects during annealing. [ABSTRACT FROM AUTHOR]

The decay of the photoluminescence at 1.54 μm from erbium-implanted silicon has been recorded over nearly three decades of intensity. Two components of the decay are observed at 7.5 K, one with a decay time constant of ∼40-160 μs, and the other with a characteristic time of between 800 and 1200 μs. It is found that the proportions of fast and slow components can vary depending upon the amount of implantation-induced damage, and this variation correlates with a broadening on the high energy side of the erbium related emission. The temperature dependence of the fastest decay is not consistent with it being due to an Auger process involving free carriers, and it is suggested that extended defects in the layers are responsible for this part of the decay curve. The broadening of the erbium line is attributed to the overlap of the dislocation-related line D1 with the erbium emission. Selective chemical etching and scanning electron microscopy show that there are extended defects present in samples with a short fast decay component. © 2001 American Institute of Physics. [ABSTRACT FROM AUTHOR]