Spontaneous emission spectrum in n-type GaAs of different doping concentrations excited by a 40 kV electron beam at 300 K
TL;DR: In this paper, the spontaneous emission spectrum in n-type GaAs excited by a 40 kV electron beam at 300 K was theoretically determined for different values of the doping concentration.
Abstract: The spontaneous emission spectrum in n-type GaAs excited by a 40 kV electron beam at 300 K is theoretically determined for different values of the doping concentration. The procedure of determination is justified through reasonable agreement of the doping dependence of the peak emission energy and the linewidth of spontaneous emission as obtained from the computed cathodoluminescence spectra with that observed experimentally. Moreover, it is shown that the doping dependence of the low energy tail slope of the computed cathodoluminescence spectrum and the appearance of humps on the high energy side of this spectrum beyond a certain level of doping are also in agreement with the experimental observation. An attempt is also made to explain the origin of such humps.
References
More filters
TL;DR: In this paper, the Laplace transform of the distribution of excitation with depth was used to determine the diffusion length and the surface recombination velocity of excess carriers in semiconductors with high efficiency for radiative recombination.
Abstract: In semiconductors having high efficiency for radiative recombination, the voltage dependence of cathodoluminescence may be used to determine the diffusion length and to estimate the surface recombination velocity of excess carriers. Theoretical calculations are based on a knowledge of the Laplace transform of the distribution of excitation with depth as determined from the target absorption correction in electron probe microanalysis or, alternatively, on a Gaussian approximation to the distribution of excitation with depth. Experimental results with accelerating voltages of 5–50 kV indicate values of diffusion length in n‐type GaAs ranging from 3.0 μ at low‐carrier concentration (5.1×1016 cm−3) to 0.65 μ at high‐carrier concentration (3×1018 cm−3). The estimated accuracy for diffusion lengths between 0.5 and 4 μ is ±30%. Information is also obtained on the surface recombination velocity and the thickness of a ``dead layer'' at the surface.
225 citations
TL;DR: In this paper, the salient features of radiative recombination near the band edge in GaAs are determined by electron beam excitation of donor and acceptor doped crystals ranging in impurity content from 10 16 to 10 20 /cm 3.
Abstract: The salient features of radiative recombination near the band edge in GaAs are determined by electron beam excitation of donor and acceptor doped crystals ranging in impurity content from 10 16 to 10 20 /cm 3 . The results indicate a strong influence of impurities and allow a consistent interpretation of recombination mechanism over the whole impurity range. At optimum doping (2 × 10 18 to 10 19 /cm 3 ) p-type GaAs at 77°K should exhibit laser action under high intensity electron bombardment.
158 citations
TL;DR: In this paper, the authors derived the hole diffusion length for Te-doped GaAs with electron concentrations ranging from 2×1016 to 6.5×1018 cm−3.
Abstract: Hole diffusion lengths L were determined for several Te‐doped GaAs crystals with electron concentrations ranging from 2×1016 to 6.5×1018 cm−3. The values of L were obtained by fitting experimental photoluminescence spectra to a theoretical expression which contains the measured absorption coefficients. This expression was derived from radiative recombination statistics, taking into account the reabsorption of emitted photons and the diffusion of minority carriers. The idea used by van Roosbroeck and Shockley of introducing the measured absorption coefficient by means of the principle of detailed balance is shown here to be valid for both nondegenerate and degenerate n‐type GaAs. It was found that the hole diffusion lengths are nearly independent of electron concentration n for n 2×1018 cm−3 values for diffusion length decrease rapidly with increasing n. This decrease is attributed to the formation of additional defects associated with donor complexes or precipitates, or both. The diffusion lengths determined in this work are in good agreement with those found by Wittry and Kyser from the electron‐beam excitation method but are larger than those obtained by Aukerman et al. from short‐circuit current measurements on n‐type surface barrier diodes subjected to high‐energy electron bombardment.
150 citations
TL;DR: Theoretical curves of the voltage dependence of cathodoluminescence have been obtained in this paper for the case where the intensity varies superlinearly with the net carrier generation rate.
Abstract: Theoretical curves of the voltage dependence of cathodoluminescence have been obtained for the case where the intensity of cathodoluminescence varies superlinearly with the net carrier generation rate. Calculations are based on numerical integration of the distribution of excess carriers assuming a Guassian approximation to the distribution of excitation with depth. The theoretical curves make possible the measurement of diffusion lengths using the method of voltage dependence of cathodoluminescence in specimens where the intensity of cathodoluminescence is not linearly proportional to the specimen current (e.g., p‐type GaAs). Experimental results with accelerating voltages of 5–50 kV using a defocused electron beam and selected area technique indicate values of electron diffusion length in p‐type GaAs ranging from 3.2 μ at low carrier concentration (6.9×1016 cm−3) to 0.6 μ at high carrier concentration (3.76×1019 cm−3).
84 citations
65 citations