Self-absorption of photons generated by prior luminescence processes affects both the steady-state and transient characteristics of LEDs. In this study we confine ourselves to one of the most important steadystate parameters -external quantum efficiency. It is shown that the photon self-absorption increases the injected carrier concentration, and consequently the emitted photon flux. This results in an increase in LED external quantum efficiency. A detailed analysis of this phenomenon is presented, and a simplified evaluation of its influence on the external quantum efficiency is given. Our calculation also takes into account multiple internal reflection of light generated inside the LED. The obtained results are applied to a diffused GaAs (Zn, Te) surface-emitting LED. It is estimated that the feedback carrier excitation (re-excitation) due to selfabsorption increases the LED external quantum efficiency by about 60% and multiple internal reflections by about 12%.
IntroductionIn an LED with a direct bandgap structure, a large number of photons generated by luminescence processes are subsequently reabsorbed within the LED' analysed the contribution of the re-excitation to the external quantum efficiency of GaAs-Al^Gaj _ x As heterostructures [4]. The influence of re-excitation on diffusion lengths of minority carriers and the minority-carrier concentration distribution in GaAs epitaxial layers excited by intensive light beam has been studied by Epifanov et al. [5]. The influence on emission spectrum shape has also been studied [6].In our paper a theoretical model of re-excitation is studied. The main purpose has been to describe the influence of the re-excitation on the LED external quantum efficiency. The presented theory also includes multiple reflections of light generated in the diode, as this effect cannot be neglected in many LED structures.
Theory of re-excitationIn any analysis of re-excitation we need to determine the photon generation rate g(x, e), where x is the position and e is the photon energy. This function represents the number of photon generated per unit volume per unit time, and may be evaluated by a self-consistent method. In the first approximation we write the generation rate as g (1) (x, e), and do not consider re-excitation. This function £(1) is then used to determine the photon absorption rate v(x, e) giving the number of photons absorbed per unit volume per unit time. This value of photon absorption rate is then used to find the second approximation to the generation rate g^2 ) (x,e) from a modified continuity equation, and taking into account the first approximation of the photon absorption rate. The procedure may be repeated to find a new photon absorption rate and a third approximation to the generation rate. It may be shown that the third approximation does not contribute tothe generation rate significantly. For practical purposes then, it is sufficient to confine this procedure to the first two steps and to consider the function g (2) as the resultant function, the super...