ABSTRACT
A b stra c t........................................................................................... 173 1. In tro d u c tio n ................................................................................... 173 2. Epitaxial Growth and Structure of AlGaN layers, Heterostructures
and Quantum W e l l s ........................................................................ 175 3. Optical properties of A l G a N ........................................................ 182 4. Photoluminescence (PL) Bands in A lG aN ..................................... 184 5. Recombination Mechanisms in A lG a N ......................................... 193 6. Electrical Properties of A l G a N .................................................... 204 7. Band Offsets, Schottky Diodes and MIS Structures in AlGaN . 221 8. Nonuniformities of Electrical and Recombination Properties
in AlGaN F ilm s ............................................................................... 224 9. C o n c lu s io n s ................................................................................... 225
Acknowledgem ents........................................................................ 226 References ........................................................................................ 226
1. INTRODUCTION
One of the most exciting developments in semiconductor physics and technology in recent years is the breathtaking race between the GaN, the SiC and the II-VI chalcogenides materials systems for the dominance in the field of shortwavelength optoelectronics and high power/high frequency/high temperature electronics. If ever it turns out that the GaN system is winning
in this race it will largely happen because of the existence of a variety of direct wide bandgap ternary and quaternary InGaAIN solid solutions that allow one to create heterojunctions, quantum wells and superlattices , combine electronics and optoelectronics applications and to move much further into the blue and UV regions of the spectrum as compared to the SiC and II-VTs (see e.g. [1]). AlGaN ternaries are important members of this family of solid solutions and they will be the subject of the present chapter. To the best of my knowledge Lyutova and Bamitskaya [2] were the first to demonstrate formation of solid solutions in the GaN-AIN system. Since no suitable lattice matched bulk substrates were (and to some extent are) in existence, heteroepitaxial growth techniques had to be developed. In the late seventies Hagen et al. [3] and Baranov et al. [4] reported successful growth of AlGaN films on sapphire using hydride vapor phase epitaxy (HVPE). Yoshida et al. [5] used reactive ion molecular beam epitaxy (RI-MBE) to deposit AlGaN films on sapphire in the entire Al composition range. Low pressure metalorganic chemical vapor deposition (LPMOCVD) growth of AlGaN on sapphire was first reported by Khan et al. [6]. It seems Koide et al. [7] were the first to grow AlGaN on sapphire in the whole range of compositions using atmospheric pressure MOCVD (APMOCVD). This group was also the first to demonstrate remarkable benefits of the low temperature (LT) AIN buffer to the crystal perfection of the AlGaN layers deposited on sapphire by APMOCVD [8,9]. Later several groups demonstrated growth of good quality AlGaN films in a wide range of compositions on sapphire by MOCVD using LT AIN, GaN or AlGaN buffers (see e.g. [10-18]). Epitaxial growth of AlGaN on SiC in the whole range of compositions was described e.g. in [19]. Electron cyclotrone assisted molecular beam epitaxy (ECR-MBE) has been used to prepare AlGaN films in the entire range of compositions either on sapphire or on SiC substrates (see e.g. [20,21]). Good quality AlGaN/GaN heterostructures, quantum wells (QW’s) and superlattices (SL’s) were grown by a number of techniques on sapphire and SiC (see e.g. [16,17,22-28]). Controlled donor doping by Si or Ge [12,13,16,19,21,29] and acceptor doping by Mg [16,21,30-32] have been demonstrated. Enhancement of electron mobility due to the two dimensional (2D) gas formation at the AlGaN/GaN interfaces was observed for structures grown on sapphire and SiC (see e.g. [22,23]). The difference in energy of formation between the wurtzite structure (the dominant polytype) and the cubic zinc blend structure is not high in AlGaN and growth of cubic AlGaN films on cubic substrates has been demonstrated using nonequilibrium growth techniques like MOCVD or MBE (see e.g. [33,34]). Device work was going hand in hand with the work on growth and characterization of AlGaN films. Prototype devices of almost every type, such as laser diodes (LDs), light emitting diodes (LEDs), modulation doped field effect transistors (MODFET’s), solar-blind photodetectors have been
successfully built to confirm the enormous device potential of the AlGaN materials system (the reader will find a much more detailed discussion in Dr. Nakamura’s and Prof. Shur’s chapters in this volume). In what follows below I tried to address some AlGaN materials related issues relevant to device operation. In particular, we’ll look at growth and structure of AlGaN films, mechanisms of optical absorption and photoluminescence, at recombination mechanisms, at doping by n-type and p-type impurities, at mechanisms of conductivity and compensation mechanisms in AlGaN films. We’ll also discuss the problems related to 2D electrons, properties of Schottky barriers, AlGaN/GaN interfaces and interfaces with dielectrics and finish with discussion of electrical nonuniformities observed in AlGaN. My aim was to concentrate on the results obtained for AlxGai_xN ternaries with composition different from GaN. However, it must be said that, even for GaN to say nothing of AlGaN, the state of understanding of quite a number of problems is very far from being satisfactory. Therefore, I had, by necessity, to use the existing data on GaN to interpret the AlGaN situation, but at the same time I tried not to slip into writing a GaN review which would be counterproductive considering that some excellent reviews on the topic have already been published (see e.g. [1,35]). Even though the state of our understanding of the properties of AlGaN is nowhere close to what it should be, the number of papers on AlGaN is fairly impressive and I sincerely apologize to the colleagues whose work I might have overlooked. I also have to apologize to the reader for perhaps excessive use of our own results that are still in press. My only excuse in doing so is that in my, necessarily subjective, view these results provide some insight into the issues being discussed.
