Hydrogen implantation in InGaNAs studied by spectroscopic ellipsometry
Introduction
InxGa1−xNyAs1−y alloys have attracted wide interest in the past few years due to their interesting optical and electronic properties. A rapid decrease of the band-gap energy has been measured with increasing nitrogen incorporation [1], [2]. Lattice matched and highly strained InxGa1−xNyAs1−y layers are of high interest for GaAs based solar cell and 1.3–1.55 μm laser diode applications, respectively [3], [4]. By probing the influence of hydrogen on the optical properties of InxGa1−xNyAs1−y, new insights into the bandstructure and nitrogen-related localized states in this alloy can be derived. Hydrogen implantation in this material system has been studied using mainly photoluminescence (PL) as characterization technique [5], [6]. The main result has been the blueshift of the PL peak upon hydrogen implantation, which almost compensates the nitrogen-related redshift of the lowest direct transition energy [5], [6].
In this study we use spectroscopic ellipsometry (SE) to investigate the effect of hydrogen implantation on the optical properties of InxGa1−xNyAs1−y. SE is a well-known technique for determination of thin film dielectric function spectra, properties of electronic interband transitions, optical phonon modes, and free carriers. We have recently used SE for thorough investigation of the optical and infrared-optical properties of GaNyAs1−y [7], [8], [9], GaNyP1−y [10], [11], BxGa1−xAs [9] and InxGa1−xNyAs1−y [12].
Section snippets
Experimental details and model analysis
Hydrogen implantation was performed on an InxGa1−xNyAs1−y/GaAs single quantum-well (SQW, dwell∼12 nm, dcap∼25 nm) sample grown by metalorganic vapour-phase epitaxy at a temperature of 560 °C on a (001) GaAs substrate. Trimethylindium, trimethylgallium, (1,1)-dimethylhydrazine, and tertiarybutylarsine have been used as precursors. Nitrogen and indium compositions can be estimated to 1.7% and 5.3%, respectively, using the lattice constant (determined by high resolution X-ray diffraction) and the
Results and discussion
Fig. 1 shows real and imaginary parts of the pseudodielectric functions of the as-grown and the hydrogen-implanted sample. In general, there is a good agreement between the experimental data (dotted lines) and the best-fit modelled spectra (solid lines). The structure at ∼1.43 eV, which is clearly visible in all spectra, is due to the fundamental band-gap transition E0 of the GaAs layers/substrate material. The vertical arrows mark the lowest transition energies of the as-grown and the
Conclusion
The lowest transition energy of InGaNAs quantum-wells is drastically blue-shifted upon hydrogen implantation, which was demonstrated by SE and PL. In a subsequent annealing step at 300 °C the original transition energy of the as-grown layer can be almost recovered. If the annealing temperature is raised to 650 °C the band-gap energy remains ∼50 meV above the value of the as-grown layer, which corresponds to the material-typical annealing-related blueshift. Optical constants n(E) and k(E) of an
Acknowledgements
This work has been supported by Deutsche Forschungsgemeinschaft under contracts G0 629/4 and 5. Additionally, we wish to thank Mrs G. Benndorf, Dr K. Otte and Dr F. Frost for fruitful discussions.
References (20)
- et al.
J. Cryst. Growth
(1998) Solid State Commun.
(1996)- et al.
Jpn. J. Appl.Phys.
(1992) - et al.
Jpn. J. Appl. Phys.
(1996) - et al.
IEEE Photonics Technol. Lett.
(1998) - et al.
Phys. Rev. B
(2001) - et al.
Phys. Rev. B
(2003) - et al.
Appl. Phys. Lett.
(2000) - et al.
J. Appl. Phys.
(2001) - et al.
Phys. Rev. B
(2003)