Luminescence efficiency measurements of silicon nanoclusters

A. J. Kenyon, P. F. Trwoga, C. W. Pitt, G. Rehm
1998 Applied Physics Letters  
We present the results of what we believe to be the first study of the power efficiency of room temperature photoluminescence from thin films of silica containing silicon nanoclusters. Films were prepared by plasma enhanced chemical vapor deposition from silane and nitrous oxide precursors. Luminescence was excited using the 476 nm line of an argon-ion laser. We have measured power efficiencies for samples that exhibit luminescence solely due to radiative recombination of quantum confined
more » ... ntum confined excitons. Efficiencies around 0.04% are reported. Following Canham's report of visible luminescence from porous silicon, 1 there has been an explosion of interest in light emission from novel forms of nanoscale silicon. There have been a number of reports in the literature of visible and near-IR emission from silicon nanoclusters. [2] [3] [4] [5] [6] Typically, such clusters lie in the sub-100-Å diameter regime and exhibit a broad red luminescence band similar to that reported from porous silicon. Reports have been published of nanoclusters deposited onto silicon substrates and embedded within dielectric matrices such as silica. There has been much debate over the nature of the luminescence mechanism in this material and contradictory reports of its optical properties, but there is a growing consensus that, in common with porous silicon, quantum confinement of excitons within the nanoclusters plays a significant role. 7-10 Previous work by this group has addressed the nature of the luminescence mechanism and has indicated the presence of two distinct processes: radiative recombination of confined excitons within the silicon nanoclusters and defect luminescence from the surrounding matrix. 7,8 A single unique mechanism does not provide a good enough explanation of the luminescence properties of nanoscale silicon. However, whatever the mechanism, this remains a technologically important material as it makes a significant contribution to the search for a silicon-based light emitting material. For this reason, it is imperative to obtain measurements of luminescence efficiency from nanoclustered silicon. To date, despite the number of reports of luminescence from silicon nanoclusters, there have been no reported studies of luminescence efficiency from this class of material. In this letter we present the results of a study undertaken to measure photoluminescence power efficiencies of thin films of silica containing nanoclustered silicon. The films were produced by plasma enhanced chemical vapor deposition ͑PECVD͒: details of their growth can be found in Ref. 7. Films were 1-2 m thick and were grown on silicon substrates. Photoluminescence was excited using the 476 nm line of an argon-ion laser, and for the purposes of recording spectra luminescence was dispersed through a single-grating Bentham M300 monochromator and detected using a photo-multiplier tube. Spectra were corrected for the spectral response of the optical system and the whole apparatus was computer controlled. For measurements of power efficiency, the total incident laser power was measured using an optical power meter, the specular laser reflection was measured directly, and diffuse laser scatter estimated by measuring the laser light intensity away from the specular reflection, knowing the area of the power meter head and film-to-detector distance, and hence calculating the total scattered power. Account was taken of the reflection of laser scatter at the thin film/silicon interface by measuring the reflectivity of a clean silicon wafer at the laser wavelength. Photoluminescence from the silicon nanoclusters was collected using a BK7 glass lens which focused the light onto the power meter head through a filter which served to remove the laser scatter. The transmission factors of the lens and the filter were measured at 820 nm using a gallium arsenide semiconductor laser. This wavelength was chosen as being representative of the peak of the photoluminescence band from silicon nanoclusters, and could therefore be used to estimate losses in the lens and filter. Once again, the reflectivity of the silicon substrate was taken into account in order to calculate the total luminescence yield, including that emitted into the substrate. Figure 1 details the geometry of the measurements and the luminescence collection arrangement. Figure 2 illustrates a photoluminescence spectrum typical of a silica film containing silicon nanoclusters. Indicated are the two bands which we ascribe to quantum confinement ͑low energy band͒ and defect luminescence ͑high energy band͒. Further details of these assignments can be found in Ref. 7. For this study, it was proposed to measure the luminescence efficiency of the quantum confinement band in isolation, as this is the more useful of the two luminescence mechanisms and the only one that can be assigned to the presence of silicon nanoclusters. For this reason, the samples selected were those that exhibited very little or no defect luminescence. We have shown in previous work that it is possible to control the relative intensities of the two luminescence bands through careful choice of deposition and postprocessing parameters. 7 Consequently, the samples studied a͒ Electronic
doi:10.1063/1.121921 fatcat:2xx5it6kqvfhhel4dv3xeziqkm