Light-emitting diodes (LEDs) are devices designed to efficiently convert electrical energy into electromagnetic radiation, most of which is visible to the human eye. Some of the disciplines involved in the understanding and utilization of LEDs are reviewed, with emphasis on the III-V semiconducting compounds and GaP LEDs in particular. Salient features of photometry, the physics of electrical injection and luminescence, and the design of LEDs are discussed in detail, followed by a survey of prominent applications for the various LEDs.

Light-emitting diodes

All defects which are dominated by short-range forces belong to the family of ‘deep’ impurities and exhibit distinctly different properties from the familiar shallow donors and acceptors, where the decisive term is the Coulomb potential. Whereas formation of the shallow states relates to a small part of the Brillouin zone and can be described within the effective mass theory, the opposite is true of the deep states. However, it has recently been shown that the formation of deep localized states in semiconductors can be described with speed and accuracy, and in a self-consistent manner, by exploiting the localized character of the main part of the defect potential. It is possible to project the interaction between the localized potential and the rest of the crystal upon a limited number of localized functions, spanning the range of the potential, with an uncertainty which is small compared to the magnitude of the forbidden gap. This is achieved without truncating proper characterization of the ele...

Deep levels in semiconductors

This review covers results of recent O.D.M.R. measurements of recombination processes in layered semiconductors, II–VI compounds, at dep traps and in amorphous compounds, and includes consideration of experimental aspects.

Optically detected magnetic resonance (O.D.M.R.) investigations of recombination processes in semiconductors

Donor-acceptor pair recombination in GaN

Non-radiative transitions affect many aspects of semiconductor performance. Normally they reduce device efficiency by suppressing luminescence, creating defects, reducing carrier lifetimes, or enhancing diffusion during operation. The present review surveys both the theoretical and practical understanding of non-radiative transitions. It includes general theoretical results and the associated ideas, with the emphasis on phonon-induced and defect Auger processes. Most of the purely formal aspects are omitted, but the points of principle where uncertainties remain are discussed. The review also covers the relation between basic theoretical studies and practical applied work on device degradation. This includes a description of the atomic processes involved in the more important mechanism of device deterioration and the theoretical understanding of the mechanism of these underlying processes. Finally, there is a survey of models proposed for 'killer' centres.

/pdf/non-radiative-transitions-in-semiconductors-2c1mre1fqm.pdf

Non-radiative transitions in semiconductors

We have observed a large number of sharp lines in photoluminescence spectra of ZnSe at 4.2\ifmmode^\circ\else\textdegree\fi{}K, leading to a broad band at lower energy. The peak energy of this band increases from \ensuremath{\sim}2.70 to \ensuremath{\sim}2.72 eV with large increase in excitation intensity. The energies and intensities of the sharp lines are consistent with recombinations between distant donor-acceptor pairs. This is the first report of such a spectrum from a II-VI compound or any direct gap semiconductor. Two high-energy pair spectra (HEPS) occur with similar patterns of sharp lines mutually displaced by 2.0\ifmmode\pm\else\textpm\fi{}0.2 meV. A lower-energy pair spectrum (LEPS), with peak near 2.69 eV, also often occurs in ZnSe. Sharp lines have not been resolved in the LEPS. A high-energy series of broad bands, whose properties suggest free-to-bound recombinations, is readily seen in association with the LEPS, but not with the HEPS. The thermal-quenching rate is greater for the LEPS than for the HEPS. This difference in behavior can be understood if the donor and acceptor in the HEPS have comparable ionization energies, while these energies differ considerably in the LEPS. The identities of the donors and acceptors are unknown.

Pair Spectra and "Edge Emission" in Zinc Selenide

Two sets of sharp emission lines, associated with the photoluminescence spectrum of the isoelectronic nitrogen trap in GaP, are unambiguously identified as the recombination of a second exciton bound to this center with an energy of 10 meV. This is the first observation of an excitonic molecule bound at a defect. The two electrons and two holes within this complex combine to form two antisymmetric states of angular momentum ${J}_{t}=0$ and ${J}_{t}=2$. The ${J}_{t}=2$ state is split 0.16 meV by the cubic crystal field, and the ${J}_{t}=0$ state lies 0.17 meV above the center of gravity of this doublet. Transitions from these states to the $A$ and $B$ states of a single exciton bound to nitrogen are seen. The complex Zeeman splittings predicted by this model agree in detail with experiment. The excitonic molecule is stable at low temperature; at 1.5\ifmmode^\circ\else\textdegree\fi{}K, the excitonic molecule emission lines increase as the square of the single exciton intensity with increase in pumping power. However, nonradiative Auger recombination reduces the over-all nitrogen emission by a factor of 3 below the intensity at 4.2\ifmmode^\circ\else\textdegree\fi{}K. The binding energy of the second exciton at the nitrogen trap is nearly equal to that of the single exciton. This remarkable fact may be possible only for an isoelectronic trap.

J. L. Merz

Papers

Pair Spectra and "Edge Emission" in Zinc Selenide

Excitonic Molecule Bound to the Isoelectronic Nitrogen Trap in GaP

Luminescence and optical absorption of implanted nitrogen in GaP

A new quasiparticle bound to the isoelectronic trap nitrogen in GaP: A rebuttal