Nitride-based, nanostructured, light-emitting devices

点击翻译按钮获取中文摘要

Nearly all of the projects depend to a significant extent on the availability of a FIB, either directly or indirectly via expected experimental results. If such a possibility for the nanofabrication of structures and the preparation of samples for analytical investigations is given, the expectation of novel experimental results will be justified, which also influence theoretical projects and modelling. Quantum dots are a central topic of the research group. The required structures for single QD studies can be prepared without undue delay and in sufficient numbers only if the FIB is available locally in Bremen. The same holds for TEM sample preparation, which is a basic requirement for a successful realization of the intended devices. The FIB will be fully exploited with regard to the large number of projects directly depending on this equipment. A focussed ion beam system is needed not only to write patterns, but also to enable the preparation of accurately specified areas for TEM investigations. In advanced systems taking into consideration here, the Ga-ion beam is not only used for patterning, but the surface can be imaged by the same beam with high resolution in a nearly non-destructive manner before, during and after processing. This is well possible for nitrides being a rather resistive material which contrasts, e.g., to the case of II-VI semiconductors where ion damage would occur unavoidably. Detailed arguments 1. Mesa etching is a precondition for the investigation of single quantum dots. The understanding of the energy transfer into the dots is a basic requirement for modelling the microscopic processes which determine quantum-dot lasing. The µ-PL spectroscopy efforts in project I-2 directly depend on the availability of mesa structures. 2. In order to achieve a three-dimensional confinement of the optical wave inside a semiconductor microcavity, micropillars with large height-to-diameter aspect ratios have to be prepared. The required precision for this task is only provided by a FIB. These taylored structures are needed for the optimization of device performance and fundamental studies of light-matter interaction. 3. The FIB technique greatly expedites the preparation of specimens for TEM, thus reducing the feed-back time for the optimization of the processing parameters. This will be of great benefit especially for the laser projects I-1 and I-2, as well as for the structural characterizations in II-1 and 11-2. 4. An essential requirement for the TEM investigations is the ability to prepare thin areas of a specimen from accurately specified sample positions, e. g. interfaces of quantum wells or quantum dots. This is difficult to achieve by conventional ion milling. In a "state of art" FIB, the secondary ions can be used to image the specimen and continuously monitor the preparation process without additional damage at least for nitridebased compounds. 5. Through injection of gases close to the point of impingement of the ion beam, it is possible either to enhance the etch rate or deposit layers onto the specimen surface, depending on the gas, the beam intensity and the scanning parameters. 6. The ability to vary the beam current over several orders of magnitude is a special feature of the FIB apparatus, allowing for a final thinning down to less than 100 nm over areas as large as 20 µm or even more. Such thin foils with nearly parallel surfaces and hence a large electron-transparent area allow for the determination of the density of lattice defects with a much greater statistical significance than is attainable with conventional ion milled specimens. The relevance for all device projects is evident. 7. In contrast to all other TEM preparation methods, FIB processing makes it possible to produce thin films with a known thickness. The examination of series of specimens with different thicknesses will be of great value for the determination of strain-relaxation effects. In addition, the specimen thickness must be known for image simulation.