No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Interpretation of the Electric Field Dependent Thermal Emission Data of Deep Traps
Abstract
The paper considers the consequencies of the simultaneous thermal
emission and recapture of carriers in the edge layer of a reverse biased
junction on the interpretation of the thermal emission probability
measurements. It is concluded that the contribution of the capture
process if properly taken into account proves the validity of the
Poole-Frenkel model of the electric field dependence.
We investigated the impact of mesa etching profiles on the emission spectra of In(Ga)As quantum dot ridge waveguide lasers grown by metal-organic chemical vapour deposition. The mesa etching was terminated: (i) well before the waveguide, (ii) directly at the waveguide, (iii) after the waveguide forming tilted sidewalls and (iv) after the waveguide forming vertical sidewalls. We found a dramatic impact of the etching profiles on the spectral intensity modulation of the longitudinal modes. The spectral hole burning effect due to the Fabry–Pérot cavity resonances causes strong modulation of the lasing spectrum, if the etching profile is terminated at the waveguide, or when the mesa sidewalls are tilted. In addition, deep-etched-through mesas with vertical sidewalls demonstrate extra spectral features induced by the high Q-factor modes originating due to the total internal reflection at the vertical sidewalls. In contrast, no intensity modulation is found in the shallow mesa devices, due to the weak effective refractive index step. The present results indicate extended opportunities for the emission spectrum control characteristic of quantum dot lasers.
DOI:https://doi.org/10.1103/PhysRev.54.647
It is shown theoretically that correlation methods of deep level transient spectroscopy give results sensitive to thermal capture effects. In thermal emission studies, the apparent trap depth depends on the free-carrier concentration.
The most common electron trap in nitrogen doped GaP (0.45 eV below the conduction band) was studied in detail. It was established that the trap concentration quadratically depends on the N concentration and nearly linearly on the net donor concentration. Recombination enhanced anneal of this trap is accompanied by an increase of the nearest‐neighbor nitrogen pair luminescence. A tentative model is suggested to identify this trap as a nitrogen split interstitial pair on phosphorous site.
Electric field effects on the thermal emission of traps in a diode have been studied. Calculations were performed and compared with experimental data on deep centers in GaAs. The results are consistent with a thermal equivalent of the optical Franz‐Keldysh effect.
A small capture pulse followed by an emission pulse was used to study the carrier emission of defect centers in the presence of electric field. The capture pulse is used to fill a localized group of centers with electrons in n‐type semiconductors and a reverse pulse causes these centers to emit electrons at an electric field determined by the reverse‐bias pulse. The enhancement of the emission due to the presence of a high electric field can be determined readily from the change of capacitance transients detected using a conventional deep level transient spectroscopy setup. In this technique, a low dc bias is used during the measurement of the capacitance transients and thus results in a high sensitivity of detection. Calculation and experimental results are given for the E c -0.35 eV center in GaAs.
For large enough electric fields, the emission of carriers from a deep level in a semiconductor is dominated by a tunneling mechanism. This mechanism is very sensitive to the parameters S and h̸ω describing the interaction of the deep level with vibronic modes of the lattice. A best‐fit method is used to yield agreement between the theoretical and experimental dependence of the emission rate on electric field and temperature. This is done for levels E3 and EL2 (’’O’’ level) in GaAs and for the ZnO center in GaP. The best‐fit values obtained for S and h̸ω are found to be consistent with other estimates deduced independently from photocapacitance, luminescence, and electron capture cross section measurements.