Excitonic fine structure in GaAs/AlGaAs (111) quantum dots grown by droplet epitaxy

Entanglement plays a crucial role in many protocols for quantum cryptography and in various approaches for quantum computation. Semiconductor quantum dots (QDs) have been proposed as a source of polarization-entangled photons which can be integrated in an electrically driven solid state device [1]. The generation process relies on the biexciton-exciton radiative cascade. However, e-h exchange interaction often induces a fine structure splitting (FSS) between the bright exciton states and destroys quantum correlation. In the case of QDs grown on commonly used (100) substrates, this degeneracy lifting is caused by the C2v symmetry stemming from asymmetric interfaces, strain anisotropy, piezoelectric fields and shape elongation [2,3]. A viable alternative relies on growing QDs on the higher symmetry (111) substrate. Theoretical investigations have shown that QDs with C3v symmetry should exhibit zero FSS [4,5]. The most studied III-V materials do not grow in the Stranski-Krastanov mode on a (111) surface, however different techniques such as droplet epitaxy [6] or the use of patterned substrates [7] have demonstrated to be able to overcome this limitation. In our work we focus on GaAs/AlGaAs (111) QDs grown by droplet epitaxy. Polarization resolved single dot photoluminescence measurements on hexagonal QDs are presented. Charged and bi-excitonic complexes are consistently identified by means of power and polarization dependence analyses. A broad FSS energy distribution is observed, with an average value smaller than the one reported for QDs grown on (100) substrates using the same technique and emitting at similar wavelengths. The phase distribution of the polarization axis evidences the absence of systematic anisotropies. These results are in agreement with previous studies on similar samples [8]. Recent advances in fabrication have proven the ability to obtain atomically flat substrates and to gradually tune the shape from hexagonal to triangular by changing the growth parameters. This lays the groundwork for a systematic investigation of the impact of geometry on the excitonic fine structure, with the goal of finding the best conditions for vanishing FSS. [1] O. Benson, C. Santori, M. Pelton, Y. Yamamoto, in: Physical Review Letters 84, 2513 (2000). [2] G. Bester, S. Nair, A. Zunger, in: Physical Review B 67, 161306 (2003). [3] R. Seguin, A. Schliwa, S. Rodt, K. Poetschke, U. W. Pohl, D. Bimberg, in: Physical Review Letters 95, 257402 (2005). [4] R. Singh, G. Bester, in: Physical Review Letters 103, 063601 (2009). [5] A. Schliwa, M. Winkelnkemper, A. Lochmann, E. Stock, D. Bimberg, in: Physical Review B 80, 161307 (2009). [6] E. Stock, T. Warming, I. Ostapenko, S. Rodt, A. Schliwa, J. A. Toefflinger, A. Lochmann, A. Toropov, S. Moshchenko, D. Dmitriev, V. Haisler, D. Bimberg, in: Applied Physics Letters 96, 093112 (2010). [7] Y. Sugiyama, Y. Sakuma, S. Muto, N. Yokoyama, in: Applied Physics Letters 67, 256 (1995). [8] T. Mano, M. Abbarchi, T. Kuroda, B. McSkimming,, A. Ohtake, K. Mitsuishi, K. Sakoda, in: Applied Physics Express 3, 065203 (2010).