Articles you may be interested inElectrical spin injection into InGaAs/GaAs quantum wells: A comparison between MgO tunnel barriers grown by sputtering and molecular beam epitaxy methods Room temperature 1.54 μ m light emission of erbium doped Si Schottky diodes prepared by molecular beam epitaxy Appl.

No abstract

We estimated the energy-transfer rate between an erbium (Er) 4f shell and a Si host by using two independent measurements. The first method involved measuring the temperature dependence of the decay time of Er 4f-shell luminescence and obtaining the energy-transfer rate by assuming that the energy transfer is assisted by nonradiative multiphonon processes and that thermal quenching is due to an energy back-transfer mechanism. The estimated value was 2×108 s−1. The second method involved measuring the time response for luminescence intensity after pulsed host photoexcitation. Although some tens of μs luminescence delay after host excitation has been reported, we found that the slow response time of the detection system may cause spurious delay. We measured the luminescence decay curve using a system with a fast response time and analyzed the data, taking into account the system response time. The energy transfer rate was estimated to be at least 107 s−1. This estimation is consistent with the result obtained by the first method, mentioned above, confirming a rather large energy-transfer rate between the Er 4f shell and Si host.

An experimental study has been made of the detection of structural defects in n‐type normalInP crystals by electrochemical etching under illumination. Under illumination, presence of structural defects such as mechanical damage or grown‐in defects on the crystal surface led to a decrease of dissolution rate of the surface. This enabled structural defects to be revealed as etch hillocks. Surface damage introduced during handling processes of the crystal was detected with high sensitivity. By observing etch hillocks, detailed information was obtained regarding fine structure and distribution of dislocations or dislocation clusters in nominally undoped crystals. It was found that novel defects, which were not involved in nominally undoped crystals false(n=1–5×1016 cm−3false) , were involved in highly sulfur‐doped crystals false(n=4×1018 cm−3false) with low dislocation densities false(EPD⪝103 cm−2false) .

Various oxide films, including SiO2, ZnO, BaTiO3, SrTiO3, and BaFe12O19, are deposited by 13.56 MHz radio frequency mode (rf-mode) sputtering with 2.45 GHz electron cyclotron resonance (ECR) plasma stimulation. The rf plasma is effectively cooled by the ECR plasma as a result of the shielding effect on the rf electric field on the target surface. This causes the self-bias voltage of the target to decrease and the rf-mode sputtering is effectively performed as in a direct current mode (dc-mode) system. There are two stages in the deposition of the oxide films: an oxygen supply-determining stage controlled by the ECR plasma density, and a sputtered-particle supply-determining stage controlled by the rf sputtering power. In the former stage, the film deposition rate is dramatically enhanced by the ECR plasma irradiation onto the substrate surface (ten times higher than without ECR plasma). In the latter stage, the film deposition rate does not change with oxygen plasma density but with the sputtering power. In this stage, the film deposition uniformity depends only on the uniformity of the sputtered particles. Fully oxidized films can be easily deposited by this technique. Changes of film composition are mainly due to anisotropy of the angle at which sputtered atoms of each component are ejected from the target surface, and also to the sticking efficiency of each component on the film surface with ECR plasma irradiation.