Thomas Research Group

Chemical Etching of Porous Silicon

Our work, over the past four years, has been in support of the quantum confinement model. We have principally worked with the chemical etching system (HNO₃ and HF) rather than the more common electrochemical system and have drawn attention to the chaotic behaviour of the kinetics and the hillock structure (see SFM image to the left; the image size is 20 x 20 µm) which is unique to the chemical system. Of central importance is a series of experiments involving the photoinduced chemical etching of porous silicon. We have shown how the chemical reaction can be controlled by light and that the frequency of the incident light affects the final emission characteristics. By etching while irradiating the sample with light of a shorter wavelength, the resulting material photoluminesces also at a shorter wavelength. These ideas support the quantum model since the photogeneration of holes inside a quantum dot is dependent upon the local bandgap. As the particle becomes smaller as a result of the etching process, the bandgap widens until absorption no longer occurs and the reaction stops. Shorter wavelength light (higher energy) will continue to be absorbed until the particle is even smaller (larger bandgap). Such a smaller particle will also tend to photoluminesce at shorter wavelength, since the bandgap is wider.

 Another aspect of this work is found in the epi-fluorescence microscope images of the regions irradiated by the laser to induce the photoetching process (the image on the right is about 200 µm across at the bottom). The vast majority of poSi material is produced electrochemically and so the chaotic kinetics of the chemical system are not commonly observed. However, we have found that the chemical system gives rise to spatial variations in the etching conditions so that there are spatially separated regions which emit different colors of lights. The overall spectrum is similar in width to those observed traditionally, but it is clear that this is actually made up from a distribution of emitters. While in electrochemically etched material these emitters are uniformly distributed across the material, in the chemical etching system they can be locally less dispersed, so that regions which emit yellow are separate from those which emit orange or red. Presumably, this suggests that the quantum emitters in the yellow region are monodisperse in size and are smaller than those in the orange or red emitting regions. This pre-sorting of particle size by the chemical reaction provides us with a unique opportunity to try to answer the important poSi question regarding the correlation of emitter size with emission wavelength.

We are working to convert our scanning probe microscopes (SPM: we have one from Burleigh Instruments and another from Topometrix) so that they can function as scanning near-field optical microscopes (SNOM) but operate specifically in a fluorescence collection mode. In this manner, the light can be collected from a 50 nm region and directed into a spectrometer to acquire the spectrum. We expect that the spectrum will be much narrower than the complete spectrum from the entire sample. Another possibility modifies this new probe even further by taking the probe tip and coating with a transparent conductor, such as In₂O₃. Now this probe tip can function as an scanning tunneling microscope (STM) tip. While an STM image can be acquired in normal fashion (others have shown that one can still obtain atomic resolution in this manner), a short voltage spike on the tip can inject a burst of high energy electrons into a particular object. Since this can induce electroluminescence, the probe tip can pick up the emitted fluorescence and acquire the spectrum. We hope that this may prove to be the critical experiment to correlate size with emission wavelength.