|
Structural investigation of InGaN/GaN heterostructures and quantum wells for long wavelength emission
P. Ruterana
CIMAP, 6 boulevard du Marechal Juin, Université de Caen, Caen, France
InGaN/GaN multiple quantum wells (MQW) heterostructures are at the basis of highly efficient light-emitting diodes (LED) and laser diodes in the near UV to the green range [1]. However, bridging the green gap still remains an important challenge [2]. Different mechanisms have been reported in the literature in order to explain the decrease of emission efficiency when the indium composition is increased in order to reach longer wavelengths. One of them is the quantum confined Stark effect (QCSE) which introduces a large spatial separation between electrons and holes [3]. Others are the Auger effect which impacts the internal quantum efficiency (IQE) mainly under high excitation regime [4], and the non-radiative recombination centers which could be predominant for highest indium contents due to possible generation of high densities of defects [5]. In this alloy, reports have also shown that there could be strain relaxation by generation of defects such as stacking faults, dislocations [6] or trenches which may constitute non-radiative recombination centers [7]. More recently, it was reported that the relaxation of the local strain through I1 basal stacking faults (BSF) was at the origin of a type threading dislocations (TDs) from QW layers to the surface [8].
In this work, we investigate the local structure of indium rich heterostructures produced by metalorganic chemical vapor deposition (MOCVD) and molecular beam epitaxy (MBE) using transmission electron microscopy (TEM) in order to explain the strain relaxation mechanisms. Our observations point out to important differences between the two growth techniques, surfaces play a critical role in MOVPE whereas dislocation glide is found to be acting in MBE. For the indium rich QWs, we find new hexagonal defects that are at the origin of a type threading dislocations going from the quantum well to the surface. Interestingly, they become predominant when the nominal indium composition is beyond 20%. For these native defects, the measured displacement vector is along <10-10> directions and it is very small. Therefore, they do not correspond to the conventional stacking faults of the wurtzite structure.
This work is supported by the Laboratory of excellence GaNex and Région Normandie.
[1] M. R. Krames, O. B. Shchekin, R. Mueller-Mach, G. O. Mueller, L. Zhou, G. Harbers, and M. G. Craford, J. Disp. Technol. 3, 160 (2007).
[2] T. Mukai, M. Yamada, and S. Nakamura, Jpn. J. Appl. Phys. 38, 3976 (1999).
[3] W. Liu, D. G. Zhao, D. S. Jiang, P. Chen, Z. S. Liu, J. J. Zhu, X. Li, F. Liang, J. P. Liu, S. M. Zhang, H. Yang, Y. T. Zhang, and G. T. Du, Superlattices Microstruct. 88, 50 (2015).
[4] C. Weisbuch, M. Piccardo, L. Martinelli, J. Iveland, J. Peretti, and J. S. Speck, Physica Status Solidi (A) 212, 899 (2015).
[5] T. Li, A. M. Fischer, Q. Y. Wei, F. A. Ponce, T. Detchprohm, and C. Wetzel, Appl. Phys. Lett. 96, 031906 (2010).
[6] M. Zhu, S. You, T. Detchprohm, T. Paskova, E. A. Preble, D. Hanser, and C. Wetzel, Phys. Rev. B 81, (2010).
[7] F. Wu, Y.-D. Lin, A. Chakraborty, H. Ohta, S. P. DenBaars, S. Nakamura, and J. S. Speck, Appl. Phys. Lett. 96, 231912 (2010).
[8] J. Smalc-Koziorowska, C. Bazioti, M. Albrecht, and G. P. Dimitrakopulos, Appl. Phys. Lett. 108, 051901 (2016).
|
