Optical properties of InN—the bandgap question
The recent controversy on the bandgap of InN is addressed, with reference to optical data on single crystalline thin film samples grown on sapphire. The optical absorption spectra deduced from transmission data or spectroscopic ellipsometry are consistent with a lowest bandgap around 0.7 eV in the l...
Gespeichert in:
Veröffentlicht in: | Superlattices and microstructures 2005-07, Vol.38 (1), p.38-56 |
---|---|
Hauptverfasser: | , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | The recent controversy on the bandgap of InN is addressed, with reference to optical data on single crystalline thin film samples grown on sapphire. The optical absorption spectra deduced from transmission data or spectroscopic ellipsometry are consistent with a lowest bandgap around 0.7 eV in the low doping limit. Further, these data from a number of different independent authors and samples give values for the absorption coefficient within a factor 2 well above the absorption edge, supporting an intrinsic direct bandgap process. The presence of Mie resonances due to In inclusions in the InN matrix affects the shape of the absorption above the edge, but is less relevant for the discussion of the bandgap for pure InN. The alternative model of a deep level to conduction band transition requires the presence of a deep donor at a concentration close to 10
20 cm
−3; in addition this concentration has to be the same within a factor 2 in all samples studied so far. This appears implausible, and no such deep donor could so far be identified from SIMS data in the highest quality samples studied. The line shape of the photoluminescence spectra can be quite well reproduced in a model for the optical transitions from the conduction band states to localized states above the valence band, including the Coulomb effects of the impurity potentials. A value of 0.69 eV for the bandgap of pure InN is deduced at 2 K. For samples that appear to be only weakly degenerate n-type two narrow peaks are observed in the photoluminescence at low temperature, assigned to conduction band—acceptor transitions. These peaks can hardly be explained in the deep level model. Recent cathodoluminescence data on highly n-doped InN films showing that the emission appears to be concentrated around In inclusions can also be explained as near bandgap recombination, considering the plausible enhancement due to interface plasmons. Finally, recent photoluminescence data on quantum structures based on InN and InGaN with a high In content appear to be consistent with moderate upshifts of the emission from a 0.7 eV value due to electron confinement. |
---|---|
ISSN: | 0749-6036 1096-3677 1096-3677 |
DOI: | 10.1016/j.spmi.2005.04.006 |