Abstract

The mechanisms operating in green-emitting InGaN/GaN quantum well (QW) structures to give bright luminescence have been the subject of much debate in recent years. For example, there have been a vast number of investigations performed to determine whether indium clustering in InGaN QWs is a valid mechanism for exciton confinement. Despite these efforts, conclusive evidence for or against indium nano-clustering have yet to be proven. More recently, substantially higher optical efficiency has been observed for commercial green LED structures which exhibit long-range thickness fluctuations, compared with LEDs that have uniform InGaN layers. It has been suggested that these thickness fluctuations may be a key to quantum confinement. However, for this assumption to be valid, it must be assumed that the disk-like regions of InGaN exist, and that the number density of these regions must be higher than the dislocation density.

Similar thickness variations have also been observed in Cambridge- grown InGaN/GaN quantum wells. Using AFM and several TEM techniques, the actual structure responsible for the observed well-width fluctuations has been determined, while the variation in the indium composition and the positions of the threading dislocations relative to the QW structure has also been examined. This presentation will explore these results and outline a possible new mechanism which may explain the bright luminescence observed for green-emitting InGaN QWs.