Erbium in Silicon

Because of its basic electronic properties, silicon does not permit to construct the classical light emitting devices known from III-V- or II-VI- compounds. On the other hand, Si is the dominant material in microelectronics and many future applications depend on the integrability of optoelectronics in large scale integrated circuits, compatible with Si technology - a research area of strong world-wide efforts. One possible approach to circumvent the intrinsic limitations of Si may consist in the incorporation of "optical dopants" into Si, and here in particular of rare earth elements which are utilised already in commercially available solid state lasers and amplifiers. Erbium is used in silica fibres for optical amplifiers in high performance fibre- optical data communication systems since Er has an optical transition within its 4f shell at a wavelength of 1.54 micro-m, close to the minimum damping of the fibre. This Er wave length is practically independent of the host material because if the efficient screening of these states by the outer valence electrons. In this project, we intend to investigate the physical principles and the possibilities to invoke Er in Si for the realisation of electroluminescent devices working at room temperature with reasonable efficiency and output. Excitation of the Er by electron-hole pairs, the conventional approach to produce light in a forward biased diode, has been proven to be very efficient at low temperatures but it exhibits strong thermal quenching at temperatures above 150 K. The quenching is attributed to the back-transfer via some additional co-activator impurity state. We will investigate this process in detail and we will try other co-dopants in order to find ways to suppress the backtransfer process. At present, the most promising approach seems to rely on the excitation of Er by hot carriers in reverse biased diodes. In order to prevent thermal quenching of the luminescence, the devices presented in the literature were produced with very large concentrations of oxygen in addition to that of Er. We infer that thermal quenching is suppressed in such devices by the formation of Si-O-Er precipitates with nanometer dimensions. Since these precipitates are the key for room temperature operation we intend to characterise them in detail in order to understand their excitation mechanism, their spectra, and their loss processes since these are the factors governing the light output of a device. In view of the observed inhomogeneous broadening in such samples, this task requires the possibility for the direct excitation of individual precipitates, either by spatially resolved spectroscopy or by selective excitation of Er states. Both approaches will be pursued in this project with the aim to investigate quantitatively the influence of the precipitate structure on its luminescence properties. From a quantitative understanding we hope to be able to develop criteria for the optimisation of a device. Finally, we will undertake to fabricate prototype devices of light emitting diodes and test structures of lasers for room temperature operation. Conservative estimation of the threshold parameters shows that lasing may be achieved for very carefully prepared resonators.