Microstructures produced during the epitaxial growth of InGaN alloys

Abstract

Effects due to phase separation in InGaN have been identified as having major effects on the performance of devices, in particular light-emitting diodes (LEDs) and injection lasers. However, the complexity of the various materials phenomena that can occur has led to a great deal of recent confusion. Much of this confusion can be eliminated by considering the experimentally measured materials properties in the context of the set of physical phenomena occurring during epitaxial growth, including coupling that exists between the various effects. Spinodal decomposition is expected to produce phase separation due to the miscibility gap in InGaN alloys. However, the actual occurrence of this phenomenon has been disputed due to the complexity of real systems. For example, the region of solid immiscibility for InGaN is strongly dependent on elastic strain. In addition, the strain, itself, affects properties such as the bandgap energy. Complicating the analysis of these phenomena is that the solid composition can be affected by elastic strain due to the well-known thermodynamic phenomenon of “compositional pulling”. An additional factor must be considered if the experimentally observed phenomena are to be understood. Thin, lattice mismatched epitaxial layers are coherent with the substrate (or underlying layer). Thus, the actual growth process for the formation of lattice mismatched layers, namely the Stranski–Krastanov (S–K) formation of islands, must be included in any realistic growth model. By considering all the phenomena together, including the coupling between them, it becomes clear that several separate mechanisms exist for phase separation. The focus of this paper is the analysis of the thin (2–3 nm), coherent InGaN layers used in the quantum well structures used for virtually all LEDs and lasers produced by the S–K mechanism. By considering these coupled phenomena together it is possible to arrive at a coherent interpretation of the various materials properties measured using techniques such as high resolution transmission electron microscopy, X-ray diffraction, and optical techniques as well as the device characteristics.

Introduction

InGaN and AlInGaN alloys have assumed increasing importance during the last decade. This is because the bandgap of AlInGaN can be tuned over the entire deep UV–visible–near IR range from 0.7 (bandgap of InN) to 6.2 eV (bandgap of AlN). These alloys are essential for the fabrication of blue, green, and white light-emitting diodes (LEDs) [1], [2]. However, the performance of yellow and, especially, red LEDs is poor, probably because of materials quality issues. AlInGaN alloys are also important for solar-blind detectors [3] and high power FETs [4]. Because of the wide range of bandgap tunability, they are also currently being investigated for multi-junction solar cells [5]. In principle, all of the bandgaps required for the most efficient combination of solar cells could be obtained using a single alloy system.

On the other hand, these alloys have also proven to be most difficult to understand and control. The basic properties of the alloys are often masked by the large defect density induced by the lack of a native substrate. This typically means that they are grown on sapphire (or SiC) substrates having a dissimilar crystal structure, lattice spacing, and thermal expansion coefficient [1], [2]. This results in epitaxial AlInGaN layers having dislocation densities of 10⁸–10¹⁰ cm⁻², in addition to stacking faults, twins, and other defects near the sapphire-epilayer interface [6]. GaN and InGaN alloys for device applications have been grown by molecular beam epitaxy (MBE) and hydride vapor phase epitaxy (HVPE) techniques, but are nearly always grown by organometallic vapor phase epitaxy (OMVPE, also called MOCVD) for commercial LEDs and lasers. Typically, a thin GaN or AlN buffer layer is grown first at a low temperature followed by high temperature (>1000 °C) growth of a GaN or AlGaN layer, followed by the growth of InGaN at temperatures ranging from 700 to 900 °C [7]. Typical LED structures consist of multiple quantum wells with well thicknesses of 2–3 nm and In contents ranging from 15% (blue) to 20% (green) InN, and even higher for yellow and red LEDs. The performance of these LEDs is stunningly good, considering the density of threading dislocations [8]. In other III/V semiconductor materials systems a dislocation density of 10⁸ cm⁻² would be sufficient to kill the radiative recombination efficiency [9], [10]. For LED performance, a number of materials issues turn out to be of critical importance. Issues relating to OMVPE growth, defect generation, and p-type doping, all major materials issues, will not be dealt with here. Instead, this paper will concentrate on issues related to the basic AlInGaN, or more frequently InGaN alloys, including the control of solid composition, factors leading to non-uniform alloy composition, and deviations from an ideal, random distribution of the cation species. OMVPE will be the main growth technique considered, since it is virtually the sole technique used for the commercial production of LED and injection laser devices. However, critical data obtained for materials grown by MBE will also be considered, where appropriate.

The objective of this paper is to examine the growth of these complex materials as a whole in terms of the basic material phenomena occurring during growth in an effort to better understand the phenomena controlling the microstructure, in particular phase separation. A better understanding of these basic phenomena is vital in our search for tools to control the microstructure and, hence, the materials properties and to, literally, allow the design of these alloys in complex structures for specific device applications. The complex issues involved in understanding the microstructure of InGaN epilayers are typically not considered together. Frequently, experimental observations are considered only in terms of a specific, individual phenomenon, which has led to some confusion. The approach taken here will be to review the basic materials phenomena individually and then to consider the coupling between the various phenomena, which makes the analysis of results more complex. This will be followed by a discussion of the salient materials and device properties in terms of the complex growth phenomena. This allows a consistent interpretation of much of the existing data.