We introduce a calibration method to quantify the impact of external mechanical stress on the emission wavelength of distinct quantum dots (QDs). Specifically, these emitters are integrated in a cross-section of a semiconductor core wire and experience a longitudinal strain that is induced by an amorphous capping shell. Detailed numerical simulations show that, thanks to the shell mechanical isotropy, the strain in the core is uniform, which enables a direct comparison of the QD responses. Moreover, the core strain is determined in situ by an optical measurement, yielding reliable values for the QD emission tuning slope. This calibration technique is applied to self-assembled InAs QDs submitted to incremental elongation along their growth axis. In contrast to recent studies conducted on similar QDs submitted to a uniaxial stress perpendicular to the growth direction, optical spectroscopy reveals up to ten times larger tuning slopes, with a moderate dispersion. These results highlight the importance of the stress direction to optimize the QD optical shift, with general implications, both in static and dynamic regimes. As such, they are in particular relevant for the development of wavelength-tunable single-photon sources or hybrid QD opto-mechanical systems.

We use strain to statically tune the semiconductor band gap of individual InAs quantum dots (QDs) embedded in a GaAs photonic wire featuring very efficient single photon collection efficiency. Thanks to the geometry of the structure, we are able to shift the QD excitonic transition by more than 20 meV by using nano-manipulators to apply the stress. Moreover, owing to the strong transverse strain gradient generated in the structure, we can relatively tune two QDs located in the wire waveguide and bring them in resonance, opening the way to the observation of collective effects such as superradiance.Epitaxial semiconductor quantum dots (QDs) embedded in nanophotonic structures are very efficient single photon sources (see [1][2][3] and [4] for a review). However their use in quantum information protocols involving more than two sources has been hindered by the dispersion in energy of different QDs. This dispersion is due to their intrinsically random self-assembly fabrication process [5], so that two QDs are never alike. QD energy tuning can be achieved using temperature [6], electric field [7-9], or material strain [10][11][12][13][14]. Temperature tuning is limited to fine tuning. Electrical control is also suitable for fine tuning and can reach shifts up to 25 meV [9]. Strain tuning can be used for fine tuning [10,11,13,14] (see [15] for a review) and, as temperature and electrical tuning, can enable two-photon interferences with two different QDs [6,8,11]. Interestingly, it offers the additional possibility to generate ultra-large shifts up to 500 meV as demonstrated by Wu et al [12] using a diamond anvil cell. Such cells are however limited to bulk systems and are unsuitable to QDs embedded in photonic environments. Fine strain tuning is usually realized by bonding the bulk QD structure on piezoelectric actuators [10-12, 14, 15], imposing limitations on the structure geometry. Remarkably, Kremer et al [13] managed to achieve up to 1.2 meV strain tuning for a QD embedded in a nanowire antenna using this bonding technique.In this paper, we demonstrate large static strain tuning (up to 25 meV) of QDs embedded in a photonic waveguide allowing efficient light extraction [1]. Four years ago, these photonic structures were used by some of us as mechanical oscillator to demonstrate strain-mediated optomechanical coupling [16]. In the present work, the strain is produced statically using a nanomanipulator enabling the realization of bright and broadly tunable quantum light sources. In addition, since the generated strain field features a very large gradient across the wire diameter [17], our method allows us to bring in resonance two QDs contained in a waveguide and opens interesting perspective for the observation of collective spontaneous emission effects such as superradiance [20].Our system is based on the GaAs waveguide shown in Fig.1a and described in detail in [1]. The waveguide from sample S1 that we have used for the first experiment (see Fig.2) is 12 µm high. The top diameter has been measured to 1.69 µm, and...