Search for content and authors |
Optical properties of InN films and the influence of surface contaminations |
Christoph Cobet 4,5, Patrick Vogt 4, Munise Rakel 4,5, Ruediger Goldhahn 2,6, Massimo Drago 4, Colm Stephens 1, Antje Vollmer 3, Wolfgang Richter 4, Norbert Esser 5 |
1. Trinity College, Deptartment of Physics, College Green, Dublin Dublin 2, Ireland |
Abstract |
Up to the late 90’s, InN properties were measured mostly on sputtered thin films of low quality. The fundamental bandgap of these samples was assumed to be around 2eV. But new theoretical investigations, as well as experimental progress in the epitaxial growth of more perfect layers by MBE and MOVPE, have shown that the fundamental band gap is more likely located around 0.7eV. Therewith, the bandgaps of group III–Nitrides could cover a very wide spectral range from the far ultraviolet (AlN) to the near infrared (InN). But the reasons for the contradicting results concerning the fundamental band gap and the VIS-VUV dielectric functions are still under discussion. Here we report on measurements of the dielectric function of hexagonal InN in a broad spectral range from 0.5-12eV by means of ellipsometry. An a-plane InN(11-20) layer grown by molecular beam epitaxy in the Cornell University, was utilized to determine the ordinary and the extraordinary part of the dielectric tensor. We find a huge anisotropy between both components. The specific absorption structures differ concerning the energy position or disappear in the extraordinary component. The ellipsometric measurements also indicate a strong influence of surface contaminations, which finally could effect the position of the measured band gaps as well. In order to analyze the surface degradation in more detail, we performed SXPS before and after a thermal annealing of hexagonal MOVPE-grown InN samples in UHV. The InN(0001) surface after transfer into UHV gives rise to clear C1s and O1s core-level contributions, the latter originating from hydroxides due to the water contamination of the surface in air. Thermal annealing at 573K is sufficient to remove the carbon and hydroxide components. Stable oxide contributions could not be removed even at the highest annealing temperature (733K) possible for InN. These results may explain partly, at the end, the wrong conclusions drawn about the fundamental band gap. |
Legal notice |
|
Related papers |
Presentation: poster at E-MRS Fall Meeting 2005, Symposium A, by Christoph CobetSee On-line Journal of E-MRS Fall Meeting 2005 Submitted: 2005-05-27 13:57 Revised: 2009-06-07 00:44 |