Epitaxial growth of nitride quantum dots
Andre Strittmatter
Introduction
The concept of quantum dots (QD) as initially proposed by Arakawa and Sakaki in 1982 was motivated by the quest for semiconductor materials providing higher radiative efficiencies, lower power consumption, and better temperature stability for light emitters such as laser diodes [1]. In the most simple picture quantum dots are three-dimensional boxes out of a semiconductor material confining charge carriers in discretized electronic quantum states. Such confinement is achieved if the dimensions of the boxes are less than the de-Broglie wavelength of the charge carriers. This quantum-mechanically defined quantity, however, is of the order of a few nm up to 20-30 nm for III/V semiconductors. Not only the small sizes are challenging for a materials scientist to realize, but also controlling densities of such tiny objects over several orders of magnitude.
Attempts to engineer quantum dots by means of patterning and etching largely failed, as defect generation deteriorates their optoelectronic properties. A groundbreaking alternative, the self-organized growth of quantum dots, was established around the mid-1990s in GaAs semiconductors and related alloys [2]. Most widely adopted today for the growth of quantum dots is the Stranski- Krastanow mode where coherently strained layers grow initially two-dimensional before nanometer-sized islands are formed upon exceeding a critical layer thickness [3]. Coherently strained islands of 10-30 nm base length, 1-5 nm height, and with densities from 106-10n cm~2 are controlled by parameters such as temperature, layer thickness, and growth interruption. Another technique, droplet epitaxy, matured recently for site-controlled growth of single QDs and quantum dot molecules [4].
The successful realization of quantum dot-based laser diodes with lower laser threshold current density and improved temperature stability validated their utility for optoelectronic applications [1, 5]. Quantum dots grown by the Stranski-Krastanow method are also demonstrated for InP and related alloys, and in the SiGe/Si material system. Interfacial strain due to lattice mismatch between different compounds is also present in the nitride material system. The limits of 2.5% strain at AlN/GaN interfaces and 11% strain at InN/GaN interfaces makes the SK growth mode likely to be available for the nitrides too. It is also fairly easy to motivate research on nitride quantum dots, as the prospect of increased internal quantum efficiencies will eventually lead to better performances of optoelectronic consumer devices such as lamps, projectors, traffic lights, and the like. Another prospective area is single nitride-based quantum dots for novel applications such as quantum computing and quantum cryptography which require generation of single and entangled photons on demand. For operation of single-photon emitters at room temperature a separation of the quantum- mechanical states of charge carriers well above the kinetic energy of 25 meV at room temperature is a prerequisite. The nitride semiconductor family allows for combination of wide-gap matrix materials and narrow-gap quantum dots which would provide the necessary energy spacing within the quantum dots. Nitride-based quantum dots are therefore considered as potential candidates for room-temperature single-photon emitters.
This chapter is devoted to concepts for realizing nitride quantum dots by epitaxial methods. It starts with GaN quantum dots grown on AlN, as these quantum dots can be grown using the Stranski-Krastanow regime and important parameters controlling their properties are identified. It covers both the molecular beam epitaxy (MBE) as well as the metalorganic vapor phase epitaxy (MOVPE) method, and will also summarize results for non-planar and semipolar QDs. Next, an attempt is made to present successful growth techniques for InGaN-based QDs. This comprises Stranski-Krastanow growth mode, droplet epitaxy, and surface pre-treatments using various chemicals. Of course, the intensively discussed subject of spontaneous QD formation within InGaN layers will be touched. InN quantum dots are treated separately, as it is probably the most challenging part of nitride growth. The chapter ends with current strategies for site-controlled growth of InGaN QDs, which is mainly done using growth of hexagonal pyramids and subsequent InGaN overgrowth of the apexes.
